Oxford Latin Syntax 1
Oxford Latin Syntax 1
Oxford Latin Syntax 1
T H E OX F O R D
LATIN SYNTAX
..................................................................
VOLUME I
HARM PINKSTER
1
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
# Harm Pinkster 2015
The moral rights of the author have been asserted
First Edition published in 2015
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2014957587
ISBN 978–0–19–928361–3
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
PREFACE
When towards the end of the twentieth century I started working on this Syntax
I wanted to have it published in 2014, one hundred years after Kühner-Stegmann’s
Satzlehre was published in two volumes as part of the Ausführliche Grammatik der
lateinischen Sprache. It was—and is—my ambition to present an up-to-date successor
to that monumental work. I mean up-to-date in several ways: Since 1912/14 new
editions of most Latin texts have been published, which are based on a better knowledge
of the manuscripts. In addition, texts unknown at that time have since been published
and studied intensively. One also needs to take account of the numerous linguistic
studies published since then, especially the large number devoted to authors and texts
other than the ‘classical’ ones. And, finally, new methods and models have been
developed in linguistics in general which make it possible to look at well-known facts
from a different angle and to present them in a different way. Several excellent general
grammars have been published in the recent past, and I have used them with great
profit: Touratier (1994), Lavency (19972), the Grammaire fondamentale series started
by Serbat, and the Sintaxis del latín clásico coordinated by Baños (2009). Special
mention must be made of the four volumes published by Baldi and Cuzzolin (2009–
11). Apart from Kühner-Stegmann, my main sources of inspiration in organizing this
Syntax were Quirk et al.’s A Comprehensive Grammar of the English Language, Givón’s
Syntax, and above all Simon Dik’s The Theory of Functional Grammar.
The most conspicuous difference between this Syntax and most Latin grammars
(Kühner-Stegmann is in fact an exception) is that it is not organized along the
traditional (indeed ancient) lines of word classes and morphological categories.
There is not, for instance, a chapter on the syntax of the adverb, nor one on all the
uses of the genitive. However, the index functions very much like Kühner-Stegmann’s
and entries for the various uses of the adverbs and the genitive can be found there.
The most conspicuous similarities are the wealth of examples and the attention
devoted to textual critical problems.
The role within the discipline of a comprehensive syntax has changed considerably
since 1912/14. At that time the users of such a syntax, both Latinists and scholars
working in other fields, had a detailed knowledge of the language. Exceptions to the
rules were more interesting than the rules themselves. Due to changes in the average
user’s level of knowledge, I have tried to give more attention to the rules and have
added translations to the examples in the main text.
The lexical resources are now much better: many articles of the Thesaurus Linguae
Latinae offer detailed grammatical information; the Oxford Latin Dictionary is also
extremely useful for the grammatical information its entries contain; there are specialized
dictionaries for most authors; and finally, there are searchable electronic corpora. So, in
giving examples, I have tried to make a distinction between things that can easily be found
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
xxxii Preface
in one of these resources (for instance, the case of the object of the verb studeo ‘to devote
oneself [to]’) and things that cannot (for instance, historic presents).
I have profited from the generous help of various institutions and persons. My own
Universiteit van Amsterdam granted periods of sabbatical leave in the early stage of
the work, which were essential for getting it started. Wolfson College, All Souls
College, and St John’s College at Oxford offered housing and honoured me with
visiting fellowships. The late Anna Morpurgo Davies played an essential role in
introducing me to academic life in Oxford and, more specifically, to the Oxford
University Press. Oxford has been my second University since then. The University of
Chicago, especially its Department of Classics, has accepted me as a visitor and
allowed me to use its facilities; it has also administered a generous grant from the
Salus Mundi Foundation (Prof. A. Richard Diebold), which made it possible to
employ two graduates from that University over a long period of time. So the UoC
became my third University. I thank these institutions, but above all the people
affiliated with them, many of whom were or have become dear friends, for their
support. I am grateful to the late Prof. Diebold for his confidence in my work.
Several persons read the entire text and gave many suggestions for improving both its
content and its format: Jim Adams, Guus Bal, Helma Dik, Hannah Rosén, Esperanza
Torrego, and above all of Olga Spevak. Guus Bal, Roland Hoffmann, and Olga Spevak
also provided many corrections of the proofs. Other friends and colleagues read
chapters and/or gave advice on individual matters. I mention here Pierluigi Cuzzolin,
Andy Dyck, Gerd Haverling, Daan den Hengst, Nigel Holmes, Caroline Kroon, David
Langslow, Adriana Manfredini, Anna Morpurgo Davies, John Penney, Rodie Risselada,
Josine Schrick, Peter White, and Jaap Wisse. Chapter 4 was the subject of a meeting
with Spanish colleagues in Madrid. Akke Pinkster cleaned the electronic version of
technical imperfections and produced the Index locorum. I thank all of these people for
their support and their help.
Thanks to the grant mentioned above, a number of graduate students from the
University of Chicago have assisted me by checking examples, finding translations, and
commenting on the argumentation and wording. Three of them must be singled out for
the duration and extent of their help: Aaron Seider in the initial stage, and Jeremy
Brightbill and Branden Kosch until the completion of this volume. Jeremy also did a
magnificent job in producing the Index of grammatical terms and Latin words;
Branden (in cooperation with Helma Dik) did a thorough revision of the entire text.
Branden also read the entire proofs; Jeremy read the proofs of the Bibliography. It was a
privilege to work with them, and I am very grateful to them and the other assistants.
I thank the staff of OUP for their competent and cheerful engagement, especially
John Davey (now retired), Julia Steer, Vicky Sunter, Kate Gilks, and Emma Turner, as
well as Jess Smith (copy-edidting) and Lesley Rhodes (proofreading).
Finally, I thank Willy van Wetter, my daughters Fenne and Akke and their families,
and my friends for their encouragement, patience, and support.
This volume is dedicated to the memory of Machtelt Bolkestein and Simon Dik, with
whom I would have loved to discuss numerous issues which I found and still find difficult.
Amsterdam, Oxford, Chicago
May 2014/2015
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
ABBREVIATIONS
Eng. English
Fr. French
It. Italian
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
CHAPTER 1
Introduction
1.1 Latin
Latin is an Indo-European language which was originally spoken by the Latini ‘the
Latins’, the inhabitants of the region called Latium, present-day Lazio, in Italy.
Among the towns of Latium, the town Roma ‘Rome’, founded in 753 BC according
to tradition, emerged as the strongest. It became the capital of the area and, over the
course of centuries, of the entire Roman Empire. With the political and geographical
expansion of Rome, its language, Latin, expanded as well and became the language of
the Empire, certainly of its Western part, and ousted the pre-existing languages of the
peoples who were incorporated into the empire. When from c. AD 400 onwards the
Western part of the empire collapsed, the new Germanic rulers adopted Latin. It also
continued to be spoken and written in the former African provinces until these
became part of the Arabic-speaking world, in the course of the seventh century. In
the European provinces, with the exception of Britain and certain other areas, Latin
continued to be the primary language and gradually evolved into the modern
Romance languages. The earliest documents in Latin that we have date from c.650 BC;
the transition from Latin to the individual Romance languages occurred at various
speeds and very gradually: until well into the eighth century AD, the language used by
the inhabitants of the former provinces still closely resembled Latin, and most
speakers probably considered themselves Latin-speaking.1
1
See Banniard (1997) and the chapter on ‘Periodization’ in Wright (2002: 36–48).
Introduction
manuscripts of Late Antiquity and the Middle Ages: for example, the papyrus
containing a fragment of Cic. Ver. 2.3–4, dated to 20 BC (CPL 20). The two main
groups both have their advantages and limitations for syntactic analysis. The first
group, transmitted via manuscripts, constitutes the richest source, not only because of
its size, but above all because of its diversity: it ranges from high literature to
handbooks on farming and aqueducts, from highly to modestly educated writers.
The second group, directly transmitted, contains inscriptions, graffiti, and ostraca
(potsherds), which can be roughly divided into two groups: those that were intended
for public use and are mostly formulaic (legal texts and funerary inscriptions, for
instance) on the one hand, and individual documents (graffiti, for instance, and letters
on papyri or thin wooden leaves), written by or on behalf of people with various levels
of linguistic training, on the other. The advantage of these documents is that they
have not been manipulated or altered over the course of transmission. Such is not the
case with manuscripts, which were copied from earlier manuscripts, some over and
over again. This was done with varying degrees of accuracy, by scribes and scholars
whose knowledge of Latin also varied, and who were often intent on making the text
(more) intelligible.
Another important source of information are the descriptions and characterizations
of Latin by Romans and Greeks, notably by grammarians and writers on rhetoric.
In conclusion, although the sources may not be ideal, the amount of data and their
accessibility result in a quality of Latin grammars not inferior to those of modern
languages. The lack of native speakers is sufficiently compensated by the amount of
the data.2
2
For a characterization of the various sources, see Baldi (1999: ch. 5).
3
The term is taken from Adams (2013: 856–62), where the problem is discussed in detail.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
have originated in Latin, their common ancestor, which are, however, invisible in the
sources that we have or which can only be seen in retrospect, taking into account the
outcome of the evolution of the Romance languages. As for the initial question, some
of our sources do give us access to spoken Latin: Plautus’ comedies, for example, are a
(stylized) form of speaking and so are the conversational parts of Cicero’s dialogues.
Petronius’ Cena Trimalchionis is another example of stylized conversation. We also
possess recorded texts (or texts based on protocols), such as the Passio sanctorum
Scillitanorum (the trial took place in AD 180)4 and the Acta of the conference at
Carthage in 411.5 There are, furthermore, features of orality in Augustine’s Ser-
mones.6 Thus, as far as syntax is concerned, there is no need to assume that we are
missing much of importance.
4
See the edition by Ruggiero (1991).
5
See the edition by Lancel (1972–91) and Lancel’s introduction I.309–16, 342–6.
6
Marti (2005) deals with this both in a general sense and in reference to the Sermones of Augustine in
particular.
7
For the ‘relation of the delivered and published speeches’, see Dyck (2013: 25–6), with references.
8
On the linguistic properties of these, see Halla-aho (2009, 2011), with references.
9
For text type, see § 2.13. For characterizations of a number of text types, see the contributions in Part
IV of Clackson (ed.) (2011).
10
For the Christian authors, see Fredouille (1996).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Introduction
11
See the references in note 18 and Adams (2013: 12–22).
12
For female speech, see Adams (1984, 2005a). For a broader perspective on ‘feminine discourse’, see
Dutsch (2008). For the language of old men in Terence, see Karakasis (2005: chs 2 and 3). For general
discussion and references, see Clackson (2011). For a survey of the literature, see Fögen (2004).
13 14
See Adams (2013: 16), with references. The quotation is taken from Adams (2013: 848).
15
See Adams’ conclusion (2007: 727–8).
16
The term is used by Rosén (1999) in the subtitle of that work.
17
For the problems involved in using this term, see Adams (2007: 13–17).
18
For the ideas of Roman writers and scholars about their own language, see Fögen (2000) and Müller
(2001). For the role of Greek in Roman society, see Kaimio (1979).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
inspiration they looked to the great Greek poets and dramatists, enriched their style
through imitation of Greek forms of expression, and experimented with Latin syntax
in other ways. For us this only becomes visible on a larger scale through the works of
Lucretius and Catullus, younger contemporaries of Cicero, and it reaches its culmin-
ation in Virgil, Horace, Ovid, and other Augustan poets. These developments in
poetry in turn formed a source of inspiration for historians from Sallust and Livy
onwards and culminated in the almost experimental prose of Tacitus’ major works.
For Christian authors the language of the Bible was a point of reference, not only in
their translations, but also in their own work.
19
This periodization follows Haverling’s periodization and terminology (1988: 20–3), also used in
Cuzzolin and Haverling (2009). For discussion, see Adamik (2015).
20
Most recently, Landfester (1997: 44ff.). The distinction is not made in Clackson (ed.) (2011) either.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Introduction
Likewise, it is not difficult to make a further distinction within the Silver Latin period,
notably between the earlier period and the later part of it with its archaizing
authors, such as Fronto, Gellius, and Apuleius. In addition, within the Late Latin
period (in fact already before AD 200, with the early publications of Tertullian), there
is a marked difference between ecclesiastical and pagan authors. In general, when
further specification is needed beyond what is provided by the chronological division
adopted here, other or additional labels are used. If more precise indications are
needed to characterize the text(s) for which specific features are discussed, this will be
done in an ad hoc manner.
Finally, there are texts dating from before Early Latin as defined above. These texts
(for which Cuzzolin and Haverling use the term ‘archaic’) will generally be ignored
because of their often fragmentary state and/or because of the uncertainties of
their date.21
21
The most extensive discussion is Hartmann (2005).
22
For interventions by scribes, see Haverling (2006b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
23
See Nocentini (2005).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Introduction
and morphological categories can be used: for example, which expressions can be
used as the attribute of a noun (answer: [among other things] adjectives and noun
phrases in the ablative). Those users who prefer the ‘traditional’ organization are
invited to use the Index, in which traditional terminology is used alongside less
familiar terms.
This Syntax will appear in two volumes. The first volume treats ‘the simple clause’;
the second will deal with complex sentences and with discourse phenomena. The
organization of the material will be the same in both volumes. A preliminary table of
contents of Volume II is attached to the global table of contents of this volume.
A detailed table of contents, which will be similar to the one in the present volume,
will be presented in Volume II.
Since this is a reference work, which is not expected to be read from cover to cover,
description of the Latin material has been standardized as much as possible. A typical
section will start with a few introductory lines followed by a description of a number
of examples. The examples that are explicitly discussed are provided with a transla-
tion. Sometimes the description of a phenomenon needs some additional comment or
explanation; this is done in indented sections in small type, referred to in the text as
‘Note’. Occasionally an ‘Appendix’, in small type, is added at the end of a section to
discuss issues that are more or less narrowly related to the main text. In the body of
the text SMALL CAPITALS are used to mark the place where terms and concepts are
defined. References to these definitions are contained in the Index of grammatical
terms and Latin words. The footnotes are almost exclusively reserved for references to
other sources of information.
Translations
1.13 Translations
I have used existing English translations wherever possible. The main source for
translations is the Loeb editions. However, I have modified existing translations when
in my view they do not correspond precisely enough to the Latin structure and/or
when they do not reflect my linguistic analysis of the Latin text.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Introduction
CHAPTER 2
This chapter introduces the terms used most frequently in this Syntax to describe the
structure of Latin sentences and texts. Details will follow in later chapters.
This clause centres on the verb form occentabunt. The verb occento ‘to serenade at or
to’ requires two PARTICIPANTs who are directly implicated in the action of the verb: a
person who performs the serenade (in this example: illi ‘those people’) and an entity
the serenade is directed to (in this example: ostium ‘door’). Accordingly, the verb
occento is a TWO-PLACE VERB; that is, it requires two words or phrases to complete its
meaning. For elements of a clause that are semantically obligatory in this way, the
term ARGUMENT will be used. The arguments illi and ostium in combination with the
verb, occentabunt, form the core or the NUCLEUS of the clause (further details and
complexities follow in § 2.7 and in Chapter 4).
In this particular example, the future tense of occentabunt signals that the nucleus
refers to a STATE OF AFFAIRS that will come into being at a time later than the
communicative situation in which the speaker utters the words. Furthermore, the
speaker presents his utterance as a fact by his choice of mood: he uses the indicative
mood of the verb.
The clause in (a) also contains a number of words that are not required by
occentabunt. The word noctu localizes the state of affairs in time. For this type of
non-nuclear elements that specify in some way the state of affairs described by the
core of the clause, the term SATELLITE will be used (a more traditional label is
‘adverbial’). Three further examples of satellites are given in (b) and (c). In (b),
modo is a satellite that refers to the time and pugnis a satellite that refers to the
instrument used in the action. In (c), per iocum indicates the manner.
(b) Nonne hic homo modo me pugnis contudit?
(‘Didn’t this man here beat me up with his fists just now?’ Pl. Am. 407)
(c) Equidem haec vobis dixi per iocum.
(‘I for one said these things to you in jest.’ Pl. Poen. 541)
CLAUSE
NUCLEUS SATELLITE(S)
VERB ARGUMENT(S)
The contribution to the meaning of the entire clause by the words at ‘but’ and enim
‘y’know’ in (a) is different. Ex. (a) is part of an EXCHANGE between the slave Toxilus
and the pimp Dordalus. In the preceding context, the slave tries to convince the pimp
of the advantages of buying a certain girl: he will become popular as a result. In
response the pimp says he is not up for having drinking parties in his house. This is
what prompts the utterance quoted as (a) above. At connects the slave’s words to what
precedes (it is a CONNECTOR) and marks them as an objection (in more common
terminology at is called an ‘adversative coordinating conjunction’). Enim expresses an
appeal by the slave to shared knowledge between the two of them: ‘you know what’s
going to happen if you don’t let them in, don’t you?’ For this type of element the term
INTERACTIONAL PARTICLE will be used (in more traditional terminology it is called a
‘causal coordinating conjunction’).
Ex. (a) centres on a finite verb form. However, there are also clauses that do not
contain a finite verb form or any verb form at all. Both are exemplified by (d). The
clause illum me vivo corrumpi is one of the arguments of the two-place verb sinam
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(the other argument [‘I’] is incorporated in the verb form sinam). The verb form of
the clause is an infinitive; its subject illum is in the accusative. This type of clause,
often used to complete the sense of verbs of speaking or thinking, is known as an
ACCUSATIVE AND INFINITIVE clause. Me vivo, by contrast, is a clause with no verb form at
all. It consists of an adjective vivo ‘alive’ and one argument me ‘me’, both in the
ablative case. This construction is known as an ABLATIVE ABSOLUTE clause. It functions
as a satellite with respect to illum corrumpi, since it expresses ancillary information
that is not directly implicated in the action of the verb.
(d) . . . sine. / # Non sino neque equidem [[illum {me vivo} corrumpi]] sinam.
(‘Let it be. # No, I won’t, and I won’t let him be corrupted while I’m alive.’
Pl. Bac. 418–19)
The clauses discussed above are part of another clause. For that reason they are called
SUBORDINATE or EMBEDDED. Neque equidem . . . sinam is called the MAIN clause in relation
to its subordinate clause illum . . . corrumpi; illum . . . corrumpi is the main clause in
relation to its subordinate clause me vivo. Together, the main and subordinate clauses
form what in this Syntax is called a COMPLEX CLAUSE.
Whereas the accusative and infinitive and the ablative absolute clauses in (d) have
no finite verb form, most types of subordinate clauses do have a finite verb. Two
major types of subordinate clauses can be distinguished, illustrated by (e) and (f ).
(e) Postquam peperit, pueros lavere iussit nos.
(‘After she gave birth, she told us to wash the boys.’ Pl. Am. 1102)
(f ) Miser est qui amat.
(‘Wretched is the man who is in love.’ Pl. Per. 179)
In (e) we have a finite subordinate clause that indicates when the order to wash the
boys was given: it contains ancillary information with respect to the main clause
(it is a satellite). The precise temporal relation is indicated by the SUBORDINATOR
postquam, which serves as a link between the main clause and a subordinate clause.
Postquam belongs to the subordinate clause without fulfilling a function of its own
in that clause. In more traditional terminology it is called a temporal ‘subordinating
conjunction’.
In (f ) the subordinate clause qui amat indicates the entity that is wretched. Miser
est requires by its meaning an entity that is miser: qui amat is therefore an argument.
Just like postquam, qui is the link between the two clauses. However, qui is also an
argument in its subordinate clause: amat requires an entity that is in love. Qui is a
RELATIVE PRONOUN; qui amat a RELATIVE CLAUSE.
The difference between a subordinate and a main clause becomes manifest when
the complex clause itself is transformed into an accusative and infinitive clause: the
finite subordinate clause remains a finite clause, while the main clause becomes an
accusative and infinitive clause. This is shown in (g). The accusative and infinitive
clause se redhibere corresponds to a finite clause redhibebo.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(g) Quid si igitur reddatur illi unde empta est? # Minime gentium. / # Dixit se
redhibere, si non placeat.
(‘Then how about returning her to the man I bought her from? # Not at all. # He said
he’d take her back if I didn’t like her.’ Pl. Mer. 418–19)
Exx. (d)–(g) show various ways of combining clauses to form complex clauses, where
one clause is part of another clause, either as a satellite or as an argument. Another
form of clause combining is illustrated by (h). Ex. (h) has two finite clauses which are
linked by the COORDINATOR et ‘and’ and so form what is called a COMPOUND CLAUSE in
this Syntax (a more traditional term for et is ‘coordinating conjunction’). In contra-
distinction to the complex clauses dealt with above, neither of the clauses contains the
other or is contained in the other. In (d) above we also had two finite clauses: non sino
and neque . . . sinam. Here the two clauses were linked by neque, which combines the
functions of a negation word and a coordinator. Another example of two coordinated
clauses is (i). When a compound clause is transformed into an accusative and
infinitive clause, all coordinated clauses become accusative and infinitives.
(h) {Odi} et {amo.}
(‘I hate and I love.’ Catul. 85.1)
(i) Quia {ego hanc amo} et {haec med amat} . . .
(‘Because I love her and she loves me.’ Pl. As. 631)
1
The terms ‘multiple’, ‘compound’, and ‘complex’ are taken from Quirk et al. (1985: 987).
2
See Longacre (2007), who defines sentences as combinations of clauses. For the definition of
‘sentence’, see also Bodelot (2007).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Ex. (a) is a grammatically complete clause with all the arguments required by emit
present. In actual use (either oral or written) it is marked as a complete unit of commu-
nication. In (a), this follows from the use of the period. Although in spoken Latin there
may have been something comparable to the falling tone at the end known from many
languages, including English, we are not able to recover much information about Latin
intonation.3 Ex. (b) is a question (or exclamation) in reaction to the surprising preceding
statement. It is an incomplete clause but a complete unit of communication (in modern
texts the question mark is used to indicate this; in spoken Latin the intonation may have
been different from (a), for example in having a rising tone, as in English). Ex. (c) is also an
incomplete clause, but like (a), it is a statement (marked by a period). Such complete units
of communication are called sentences in this Syntax.
A clause (simple or multiple) may coincide with a sentence. This is the case in (a)
above and in the examples (d)–(i) cited at the end of the preceding section. Sentences
may also be shorter than a complete clause, as is shown by (b) and (c) above. This is
further illustrated by (d)–(g). In (d), the sentence Vobis . . . volt consists of a relative
clause that is uttered as a reply to the preceding question and which is the subject of the
verb that has to be supplied from that question (est ‘is’). In (e), the one-word sentence
consisting of the noun malum ‘evil’ is a reply to the preceding question, from which
something like dabo ‘I will give’ must be supplied to govern the object malum.
Similarly, the adjective scitula in (f ) goes with the noun facie in the preceding question.
In these three instances, the preceding question contains an element that offers a slot to
be filled by the answer. In (g), by contrast, while the speaker probably expects an
answer to his suggestion (the question mark in the OCT may be exaggerated), the
precise form of the reaction is unpredictable. This is even more true for nugae in (h), a
one-word sentence serving as a comment on the preceding words of the interlocutor.
(d) Quis est? / # Vobis qui multa bona esse volt.
(‘Who is it? # Someone who wants you to have lots of good things.’ Pl. Truc. 116–17)
(e) Si huius miseret, / ecquid das qui bene sit? # Malum.
(‘If you feel pity for her, will you give her anything from which all will go well for her?
# A thrashing.’ Pl. Cur. 518–19)
(f) Qua sunt facie? # Scitula.
(‘Of what appearance are they? # Pretty.’ Pl. Rud. 565)
(g) Fortasse tu huc vocatus es ad prandium, / ill’ qui vocavit, nullus venit? #
Admodum.
(‘Perhaps you were invited here for lunch and the one who invited you didn’t come? #
Exactly.’ Pl. Rud. 142–3)
3
Ancient grammatical and rhetorical treatises manifest a keen interest in intonation, and they also
recognize various types of sentences, but they do not seem to pay attention to the relationship between
these areas. Augustine (Doctr. chr. 3.3.6) gives instructions to pronounce information-requesting and
rhetorical questions differently in order to avoid ambiguity (as pointed out by Branden Kosch, p.c.). For
ideas about intonation in Antiquity, see Luque (2006).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Latin has four different types of sentences (SENTENCE TYPES), three of which are
illustrated above. When using a DECLARATIVE sentence the speaker asserts a state
of affairs, as is illustrated by the answers in (d)–(g) and the second sentence of (h).
Ex. (d)–(g) also illustrate INTERROGATIVE sentences (or ‘questions’). Ex. (h) illustrates
the IMPERATIVE sentence type. Finally, Latin has EXCLAMATORY sentences.
The number of utterances that can be used as sentences is infinite. However, there
are certain words that cannot be used as sentences. Examples are the preposition ad
‘to’, the coordinator et ‘and’, the connector nam ‘for’, and the interactional particle
enim ‘for (as you know)’.
Application of the content of the preceding paragraph to the Latin material is not
unproblematic, not least because we only have written material, and the spaces and
punctuation marks that were in use served a different purpose from punctuation in
modern languages. Capitals or something similar were not used to mark the begin-
ning of sentences in a systematic way. (For details on the evidence see Chapter 24.4)
The punctuation in our printed texts is the result of modern interpretation and varies
from one national tradition to another. For the purposes of this Syntax, capitals and
periods are used to demarcate words and sequences of words as sentences. Although
the texts used for illustration are essentially those from the Teubner series and the
Corpus Christianorum series, the punctuation has been adapted whenever necessary.
Sentences start with a capital and end with a period or question mark. In practice this
means that many semicolons have been replaced by periods.
Appendix: In actual pronunciation, sentences, certainly those of a greater length,
were segmented into smaller units, which may, but need not, coincide with gram-
matical units (words, phrases, or clauses). These segments may also be sequences of
grammatical units between which no immediate syntactic relation exists. A very
prosaic reason for this segmentation is the need to take a breath, but it may also serve
other purposes, such as giving emphasis and the production of certain rhythmical
structures. We may assume that such segments were also marked by intonation. For
such segments the term COLON will be used. Details are discussed in Chapter 24.
The term PHRASE will be used for combinations of two or more words that behave as a
unit with respect to other elements of the structure to which they belong. In this
Syntax, four types of phrases are distinguished: noun, adjective, and adverb phrases
on the one hand, and prepositional phrases on the other. The term ‘verb phrase’ will
not be used in this Syntax. A NOUN PHRASE consists of a noun (the HEAD) and one or
4
A good introduction to pause and punctuation in Latin texts is ch. I of Parkes (1992).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
more elements modifying it, belonging to various classes. In principle, the noun
phrase as a whole has the same structural potential as the head by itself.
The elements modifying a noun are called ATTRIBUTEs, but sometimes also in a more
general way: MODIFIERs. At this point, it suffices to give an example of a modifying
possessive adjective, as in (a), and another with a noun in the genitive, as in (b). In (a),
the possessive adjective nostrum functions as the attribute of hortum, which is the
head of the noun phrase hortum nostrum. In (b), the genitive noun horti modifies the
noun ostium, which is its head. In both cases, omission of the modifying element does
not make the remainder of the structure ungrammatical.
(a) Abii illac per angiportum ad hortum nostrum clanculum.
(‘I secretly went that way through the alley to our garden.’ Pl. Mos. 1045)
(b) Ostium quod in angiporto est horti, patefeci fores . . .
(‘I opened the wings of the garden door that is in the alley.’ Pl. Mos. 1046)
Simple illustrations of an ADJECTIVE PHRASE and an ADVERB PHRASE are (c) and (d). We
see the degree adverb valde ‘very’ modifying the adjective bonis ‘good’ and the adverb
bene ‘well’, respectively. Here, too, omission of the modifying element does not lead to
an ungrammatical expression.
(c) Explicat orationem sane longam et verbis valde bonis.
(‘He unrolls a speech that is very long and characterized by very fine words.’ Cic.
Agr. 2.13)
(d) Rem te valde bene gessisse rumor erat.
(‘There is a report that you have had a highly successful campaign.’ Cic. Fam. 1.8.7)
With PREPOSITIONAL PHRASEs the situation is different. In (a) above, ad hortum nostrum
is a prepositional phrase. Ad seems to be the ‘dominant’ element in the phrase, since
the accusative case form of hortum nostrum depends on it. However, neither the
preposition ad nor the noun phrase hortum nostrum can be omitted without making
the remaining structure ungrammatical. (The relation between the preposition and
the other constituent in a prepositional phrase is discussed further in § 12.23 and
in § 12.25.)
A familiar definition of WORD is: the smallest linguistic segment in an utterance that
can (in principle) be moved independently. In this respect, a word differs from a
morpheme (a word-internal segment) and from a phrase (a segment containing two
or more words). More importantly, words may stand at the beginning of a sentence,
which distinguishes them from CLITICs. Most clitics have a meaning of their own,
but they are bound to a word that functions as their ‘host’ with which they form a
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
phonological unit. Although clitics have a certain freedom as to which words they are
attached to, they can never stand in a sentence-initial position (for details see § 3.28).
Modern editions use a space to separate individual words. The Romans often
marked word boundaries in some way or other, but there was both individual and
diachronic variation (see Chapter 24). Sometimes the order and relative position of
words in a phrase is such that one may hesitate whether a given segment is a word or a
phrase. In the course of time, a phrase may develop into a word. This can be
illustrated with the segment magno opere / magnopere in its meaning ‘greatly’ or
‘particularly’. It is rare in Early Latin, but there are a few instances in Plautus. In the
OCT it is printed as two words in (a) (manuscripts vary). There are also instances
with the order reversed, as in (b). Obviously therefore, in Plautus’ time it was still
a phrase. After Plautus, the reverse order is not attested any more (except with a
different meaning). Cicero has two instances of separation—(c) and (d)—but in the
instances where the words occur together, editions of his texts vary in printing magno
opere or magnopere. This is also the case in editions of later authors, most of which do
not feature instances of separation.5
(a) Edictum est magno opere mihi, ne quoiquam hoc homini crederem . . .
(‘And I was ordered explicitly not to entrust this to anyone . . . ’ Pl. Per. 241)
(b) Vos omnis opere magno esse oratos volo . . .
(‘I want to appeal earnestly to you all . . . ’ Pl. Cas. 21)
(c) Cum puerorum igitur formas et corpora magno hic opere miraretur . . .
(‘When, therefore, he was greatly admiring the figures of the boys, and their bodies, . . . ’
Cic. Inv. 2.2)
(d) . . . magnoque opere abs te peto cures ut is intellegat . . .
(‘ . . . I earnestly request you to make him realize . . . ’ Cic. Fam. 13.34.1)
2.6 Constituents
Clause (a) below, repeated from the preceding section, may in a first analysis
(ignoring -que) be divided into four parts, indicated by brackets.
(a) . . . [[magnoque opere]] [[abs te]] [[peto]] [[cures ut is intellegat]] . . .
(‘ . . . I earnestly request you to make him realize . . . ’ Cic. Fam. 13.34.1)
These four parts belong to different categories: a noun phrase, a prepositional phrase,
a verb, and a subordinate clause, respectively. They also fulfil different functions in
the clause. Magno opere, for example, is a manner expression and the subordinate
clause functions as the object of peto. As a neutral term for referring to ‘parts’ of a
5
For details, see TLL s.v. opus 854.10ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
more complex structure without specifying to which category they belong, or which
function they fulfil, the term CONSTITUENT has become common in contemporary
linguistics; it will be used throughout this Syntax.6
In (a), some of the constituents have an internal structure of their own. The noun
phrase magno opere, for example, consists of two constituents and the cures clause
also consists of two constituents: cures and the clause ut is intellegat, which in turn can
be analysed in more detail. A syntactic argument for regarding magno opere as one
constituent in its clause (and not two) is that it could be replaced by one word (for
example, valde ‘strongly’); similarly, the clause cures ut is intellegat could be replaced
by the single word auxilium ‘help’. These replacements would still result in a correct
grammatical structure: valde abs te auxilium peto and would leave the syntactic
structure of the sentence unchanged.
The number of arguments a particular verb requires depends upon its meaning. The
term VALENCY is used to describe this. The verb occento is two-place or BIVALENT; as
shown above, it requires an argument that performs the action (an agent), and a
second argument on which the action is performed (the patient). The verb ambulo ‘to
walk’ is a one-place or MONOVALENT verb: it requires only an agent to complete its
meaning. The verb do in its basic meaning ‘to give’ is three-place or TRIVALENT (one
entity gives another entity to a third entity). And, finally, pluit ‘it rains’ does not
require any element at all to complete its meaning; hence it is ZEROVALENT.
If a verb has more than one meaning the number and the type of arguments may
vary accordingly. The verb dico, for example, has several distinct meanings, each with
6
The term is not found frequently in scholarship in Classics; for an exception, see Habinek (1985:
passim).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
its own constellation of arguments or, in the terminology of this Syntax, its own VERB
FRAME. This is illustrated by (a)–(d). In (a) dico means ‘to speak’, ‘to give a speech’; in
this meaning it requires one argument (the agent) (note the parallelism with vivendi).
In (b) it means ‘to say’ (two arguments), in (c) ‘to tell’ (three arguments), in (d) ‘to
appoint’ (also three arguments, but of a different type and with a different pattern of
case forms).
(a) . . . idem erant vivendi praeceptores atque dicendi . . .
(‘ . . . the same men were teachers of ethics and of speaking . . . ’ Cic. de Orat. 3.57)
(b) Egone istuc dixi? # Ita.
(‘I said that? # Yes.’ Pl. Bac. 806)
(c) Egone istuc dixi tibi? / # Mihi quidem hercle.
(‘I said that to you? # Yes, to me, by Hercules.’ Pl. Mer. 761–2)
(d) Postero die dictator . . . magistrum equitum dicit L. Tarquinium . . .
(‘On the following day the dictator . . . named as his master of the horse Lucius
Tarquinius . . . ’ Liv. 3.27.1)
The frame or frames a verb is associated with do(es) not only depend on the meaning
of the verb. Even with one and the same meaning, a verb may have multiple frames.
The lexical properties of the argument(s) play a role as well. Thus, with a human first
argument the verb incipio ‘begin’ is usually bivalent (‘a person begins something’) but
with another type of first argument, inanimate autumnus ‘autumn’ for example,
incipio is monovalent, as in (e).
(e) Nam desinit aestas, / incipit autumnus media sub Virgine utrimque.
(‘For at the middle of the Virgin summer on one side ceases and autumn on the other
begins.’ Man. 2.176–7)
The verb facio in its meaning ‘to make, to produce’ will normally be considered
bivalent (‘someone produces something’), but with an emotion noun as its patient,
such as dolorem ‘grief ’, the resulting causative expression requires a recipient in the
dative case form, as in (f).
(f ) Sed augeo commemorando dolorem et facio etiam tibi.
(‘But I make the pain worse by dwelling on it, and give you pain also.’ Cic. Att. 11.8.2)
Conversely, the verb do ‘to give’ is a typical trivalent predicate, requiring a recipient in
the dative, but not so if the second argument is something like motus ‘movements’
(further details are discussed in Chapter 4).7
Valency is essentially a quantitative notion: it indicates how many arguments a
verb normally requires. However, the arguments also have a semantic relationship
7
For factors influencing the presence or absence of a recipient constituent with a number of verbs, see
Baños (1998: 28–39).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
with respect to the verb (functioning, for example, as the agent) and a syntactic
function (functioning, for example, as the subject). Semantic and syntactic functions
are dealt with in § 2.12.
Appendix: The notion of valency is relatively recent, but can be traced back to
ancient notions of transitivity and intransitivity. It was introduced into general
linguistics in a systematic way by Tesnière (1959) and into Latin linguistics by
Dressler (1970).8 Although the notion is present in some form or other in most
contemporary linguistic models, there remain a number of difficulties in its practical
application (see §§ 4.1–4.6). It has been suggested that assigning one or more precise
valencies to every verb is impossible in view of the number of borderline cases and
that it is better to regard valency as a scalar concept: some entities are more
obligatory or optional than others.9 Similarly, some scholars consider the strict
division between arguments and satellites impracticable and/or unnecessary.
Objective tests to determine the valency of predicates have been developed for
several languages, including Latin.10
The special role assigned to the verb (or a comparable entity) is justified because the
structure of the nuclear predication is determined by its meaning. This does not
exclude the possibility of verbless utterances that in all other respects look like ‘normal’
clauses. Nor does this say anything about the prosodic status of verbs in general or
of particular verbs. The fact that notably the monosyllabic forms es ‘you are’ and est
‘he/she/it is’ were elided and are also often found in the shadow of salient words,
resembling the behaviour of clitics, is also irrelevant for the concept of valency.11
The notion of valency is also applicable to non-verbal categories, such as adjectives and
nouns. The adjectives similis ‘alike’ and dissimilis ‘unlike’ in (g) are both bivalent; the
same holds for the noun dux ‘guide’, ‘leader’ in (h), whose valency is like that of duco.
(g) Sic dicitur similis homo homini, equus equo, et dissimilis homo equo.
(‘Thus a human being is said to be like a human being, and a horse to be like a horse,
and a human being to be unlike a horse.’ Var. L. 10.4)
(h) Cuius legationis Divico princeps fuit, qui bello Cassiano dux Helvetiorum
fuerat.
(‘The leader of the deputation was Divico, who had been commander of the Helvetii
in the campaign against Cassius.’ Caes. Gal. 1.13.2)
Nouns, adjectives, and verbs that share a particular semantic feature often have the
same valency and allow the same types of elements to be combined with them. This
phenomenon of TRANSCATEGORIAL PARALLELISM is illustrated by (i)–(k). The verb cupio
‘to desire’ governs an infinitival clause, as in (i). The same type of embedding is
8
References in LSS § 1.2. For Latin valency from a typological perspective, see Lehmann (2002).
9
For the application of the concept of scalarity to valency, see Himmelmann (1986).
10
See Happ (1976: 346–428); LSS § 2.1; Baños et al. (eds) (2003), especially the contribution by de la
Villa (2003); Torrego et al. (eds) (2007), with references.
11
On the prosodic status of verbs, see Fortson (2008: ch. 9).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
found with the noun cupido ‘desire’, as in (j), and with the adjective cupidus ‘desirous’,
as in (k).
(i) Cum hoc . . . cupivit . . . bellum componere . . .
(‘With him . . . he wished to arrange a truce . . . ’ Nep. Han. 6.2)
(j) . . . si tanta cupido / bis Stygios innare lacus . . .
(‘ . . . if there is such a great desire / to swim the Stygian lake twice . . . ’ Verg.
A. 6.133–4)
(k) . . . cupidus falsis attingere gaudia palmis . . .
(‘ . . . yearning to touch with unreal hands his heart’s delight.’ Prop. 1.19.9)
The verb and its arguments refer to some situation in the observable or imaginary
world in which we live. The structure of such situations may be quite diverse. Human
beings may be actively involved, exercising their will over what is taking place, or
something may take place or be the case without some form of active involvement.
Further, the situation may or may not involve change (for example, John grew old or
John was old), etc. From Aristotle onwards, terms have been proposed to describe and
characterize the most important types of situations. As a general term for the various
types of situations this Syntax will use the term STATE OF AFFAIRS (SoA). The following
distinctions will be made:12 (i) SoAs that are controlled or not controlled by some
entity denoted by the first argument (usually a human being, but also animals or
forces of nature may be depicted as controllers, exercising their will over some
situation); (ii) SoAs that do or do not imply change (or, in other words, SoAs that
are ‘dynamic’ or ‘non-dynamic’). The combination of these two features CONTROL and
CHANGE results in four types of states of affairs, as indicated in Figure 2.2.
English examples of these four types are: John went home (action), John stayed
home (position), John grew up in London (process), and John knows Greek (state).
Dynamic Non-dynamic
(+ change) (– change)
Controlled ACTION POSITION
(+ control) (go) (stay)
Not
PROCESS STATE
controlled
(grow) (know)
(– control)
12
Essentially this follows LSS § 2.4 with adaptations from Haverling (2000: 22–31), where further
references to the literature can be found. Some of the English examples are taken from her book. See also
Haverling (2010a: 284–340). The term ‘state of affairs’ is taken from Dik (1997: I.105–26).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
States of affairs that are controlled can occur in the imperative and may be used in
subordinate clauses depending on verbs of ordering and wishing, like impero and
volo. They can furthermore be combined with manner and instrument expressions
and also with beneficiaries. This is briefly illustrated by (a)–(c). In (a), abi is an
imperative form. It is combined with the manner satellite cito. In (b), dormitum ut
abeas depends on volo, to be understood from the preceding question. Ex. (c)
illustrates a beneficiary satellite (maioribus natu). Details and problems follow in
§ 6.30 (imperative) and in § 10.42 (process adjuncts).
(a) Abi cito.
(‘Off with you, quickly!’ Pl. Cist. 781)
(b) Numquid vis? # Dormitum ut abeas.
(‘Do you want anything? # Only that you go off and sleep.’ Pl. Ps. 665)
(c) Ut maioribus natu assurgatur . . .
(‘ . . . as, for example, rising out of respect to elders . . . ’ Cic. Inv. 1.48)
The absence or presence of the feature ‘change’ is especially relevant for the combin-
ability of an SoA with various expressions of time and for the interpretation of certain
tense forms. In fact, the binary distinction +/– dynamic has to be refined further.
Dynamic states of affairs either may be brought or come to a natural end (in that case
they are called TERMINATIVE or ‘telic’) or may not (NON-TERMINATIVE or ‘non-telic’).
English examples are John is walking along the beach (non-terminative) and John is
walking to school (terminative). A further distinction can be made into MOMENTANEOUS
and NON-MOMENTANEOUS states of affairs. An example of a momentaneous SoA is John
finds a book on the beach. The distinction +/– momentaneous operates especially within
the type of terminative states of affairs, but a non-terminative SoA like John coughed
is momentaneous as well. Non-momentaneous SoAs are called ACCOMPLISHMENTS,
momentaneous ones ACHIEVEMENTS. These distinctions are represented in Figure 2.3.
State of Affairs
– dynamic + dynamic
John is lying on the beach.d
– terminative + terminative
John is walking along the beach.e
– momentaneous + momentaneous
John is walking to school.f John finds a book on the beach.g
2.10 Satellites
Clauses often contain more than just the verb and its obligatory arguments. Ex. (a),
repeated here from § 2.2, contains apart from its nucleus a connector (at), an
interactional particle (enim), and a satellite (noctu).
(a) At enim illi noctu occentabunt ostium . . .
(‘Well, they’ll serenade your door at night . . . ’ Pl. Per. 569)
13
See Vester (1983: 22–7); Haverling (2010a: 305).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Satellites
Several types of satellites can be distinguished. In the first place, there are satellites
that add details to, or specify in some way, the SoA described by the verb and its
arguments: in (a) noctu situates the action illi occentabunt ostium in time. In the same
way, one could add that they will do this magna voce ‘in a loud voice’, in platea ‘on the
street’, or suis sodalibus ‘for the benefit of their comrades’. For this type of specifying
satellites, this Syntax uses the term ADJUNCT. Some of these adjuncts occur only, or
predominantly, with states of affairs of a specific type. Instrument adjuncts, for
example, occur typically with controlled states of affairs. Time adjuncts like noctu,
by contrast, have no restrictions.
Another type of satellite is illustrated by (b) and (c). Certe in (b) expresses the
speaker’s subjective evaluation of the content of the clause. The clause ut . . . veniamus
in (c) indicates the status of the following information in the ongoing conversation
and how that information has to be understood. The two expressions do not give
further details about the SoA; they relate rather to the communicative situation of the
speaker and hearer. For this type of satellite this Syntax uses the term DISJUNCT.
Further discussion will follow in §§ 10.97–10.107.
(b) Certe hic insanu’st homo.
(‘This man is certainly mad.’ Pl. Men. 282)
(c) Ut vero iam ad illa summa veniamus, quae vis alia potuit . . . homines unum
in locum congregare . . .
(‘To come, however, at length to the highest achievements of eloquence, what other
power could have been strong enough . . . to gather . . . humanity into one place . . . ’
Cic. de Orat. 1.33)
A composite clause based on exx. (a)–(c) can be represented in a graph as in Figure 2.4,
showing the hierarchical position of the various components. The verb is the most central
CLAUSE
NUCLEUS SATELLITE(S)
Figure 2.4 The hierarchical structure of the clause (more elaborate version)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(ii) CAUSE for entities that bring about an SoA without wilfully controlling it, as
in (d). Satellites may also function as cause, as in (e).
(d) Eius amor cupidam me huc prolicit per tenebras.
(‘Love for him is driving my eager self out here through the darkness.’ Pl. Cur. 97)
(e) Is amore misere hanc deperit mulierculam . . .
(‘He’s dying wretchedly with love for this woman . . . ’ Pl. Cist. 131)
(iii) PATIENT for entities undergoing an SoA, as in (a)–(d). The term ‘patient’
covers both so-called AFFECTED objects (as in John restored the house) and
so-called EFFECTED objects (as in John built the house). The term EXPERIENCER is
used for entities that undergo a physiological or psychological SoA, as hanc
mulierculam in (e).
(iv) RECIPIENT and ADDRESSEE for entities that function as third arguments in three-
place states of affairs referring to handing over or communicating something,
respectively. Examples are (f) and (g), respectively.
(f ) Diabolus Glauci filius Clearetae / lenae dedit dono argenti viginti minas . . .
(‘Diabolus, the son of Glaucus, has given twenty silver minas as a gift to the
madam Cleareta . . .’ Pl. As. 751–2)
(g) Em tibi! Hic mihi dixit tibi quae dixi.
(‘Here you go! He told me what I told you.’ Pl. Mil. 365)
(v) ASSOCIATIVE for entities that actively participate in a two-place SoA together
with a principal agent, as with depugno (cum) ‘to fight (with)’ in (h). This
function is also fulfilled by satellites, in combination with verbs that do not
require them by their meaning, as in (i).
(h) Ter depugnavit Caesar cum civibus, in Thessalia, Africa, Hispania.
(‘Three times Caesar fought it out with his countrymen, in Thessaly, Africa,
and Spain.’ Cic. Phil. 2.75)
(i) . . . ut cum exercitu / hinc profectus sum ad Teloboas hostis eosque ut
vicimus.
(‘ . . . since I went away with the army from here to our enemy, the Teloboians,
and defeated them.’ Pl. Am. 733–4)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(vi) DIRECTION, PLACE, and SOURCE for entities that function as arguments in SoAs
implying motion or rest, like the source argument in (j). These functions are
also fulfilled by satellites.
(j) Puer aberravit inter homines a patre.
(‘The boy strayed from his father among the crowd.’ Pl. Men. 31)
The terms mentioned in (i) and (iii)–(v) are especially useful for the description
of states of affairs in which human beings are involved, as the terms themselves
suggest. Many linguists use different labels, and also sometimes a more exten-
sive set of categories.14 In this Syntax, this particular set of labels will help
explain certain facts about Latin; it is not suggested that they cover all semantic
relations between the verb and its arguments. More discussion will follow in
Chapter 4.
The most common semantic functions of satellites (apart from agent, cause, and
associative—see above) are:
(vii) BENEFICIARY (the entity to whose advantage or disadvantage an action is
undertaken).
(viii) INSTRUMENT (the tool, usually inanimate, used in performing an action).
(ix) MANNER (indicates the way in which an SoA is carried out or takes place).
(x) REASON (indicates why an SoA is carried out or takes place).
(xi) PURPOSE (indicates with what goal in mind an action is performed).
(xii) Several spatial specifications.
(xiii) Several temporal specifications.
A fuller list can be found in the table of contents of Chapter 10 and Chapter 16. The
number of functions (and their nomenclature) varies in the literature.
Arguments and satellites can furthermore be described in terms of their SYNTACTIC
FUNCTIONS. The only syntactic function on which Latinists seem to agree is that of
the SUBJECT, commonly defined as the constituent in a finite clause with which the
verb form shows agreement.15 This definition is then often extended to non-finite
constructions, such as the accusative and infinitive clause and the ablative absolute
clause. Examples are (k)–(m). In the active clause (k) laudat agrees with its subject
pater: it is third person singular. In the passive clause (l) the verb agrees with filius.
In the (bracketed) accusative and infinitive clause in (m) there is another type of
agreement (number, case, and gender) between the subject filium and the infinitive
laudatum esse.
(k) Pater filium laudat.
(‘The father praises his son.’)
14
See for example Dik (1997: I.117–24) and Givón (2001: 117–61). Part of the terminology is taken
from them, and some definitions are adaptations from these sources.
15
For discussion of the status of the subject in Latin, see Lavency (1994).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Most Latinists will call filium in (k) the (DIRECT) OBJECT of the clause. Some will restrict
the use of the term object to instances precisely like (k), that is, to the second
argument in the accusative case that can also be used as the subject in a passive
counterpart, as in (l). As a consequence, in this view, the second arguments in (n) and
(o) do not count as objects: pecunia mea in (n) is in the ablative and passivization is
excluded; magnos articulorum dolores in (o) is in the accusative, but passivization
is nevertheless excluded. Another instance of an accusative which is not passivizable is
me in (p) with the deponent verb sequor. Other Latinists will take only the accusative
as the defining criterion.
In this Syntax, the term ‘object’ will be used for all four types of second
argument, (k) and (n)–(p). If necessary, for instances like (n) the term ‘ablative
object’ is used.
The term INDIRECT OBJECT is commonly used for the third argument of verbs of transfer
or communication, that is, for an argument that functions semantically as a recipient
or an addressee (see above). Examples are (q) and (r). Some scholars also use it for the
second argument of verbs that govern a dative, such as faveo, as in (s). In this Syntax,
however, that constituent is called a ‘dative object’.
(q) Prandium uxor mihi perbonum dedit . . .
(‘My wife gave me a very good lunch.’ Pl. Mos. 692)
(r) Em tibi, hic mihi dixit tibi quae dixi.
(‘Here you go! He told me what I told you.’ Pl. Mil. 365)
(s) Hominum nobilium non fere quisquam nostrae industriae favet.
(‘There is hardly one member of the old families who looks kindly on our activity.’
Cic. Ver. 5.182)
Adopting the traditional term ‘indirect object’ raises the question of what to call other
third arguments—for example, the ablative constituent with verbs of supplying, as
in (t). (NB: this verb is also used with an accusative + dative [object + indirect object]
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
pattern.) There is no generally accepted term available. In this Syntax, the description
of such structures will be given in an ad hoc way.16
(t) Ubi igitur est crimen? Quod eum Pompeius civitate donavit.
(‘Wherein then does the accusation consist? In this: that Pompeius has honoured him
with citizenship.’ Cic. Balb. 7)
As for the syntactic functions of satellites, two types of satellites are distinguished in
§ 2.10, those functioning as ADJUNCT and those functioning as DISJUNCT.
The last syntactic function of clause constituents to be discussed in this section is
the SECONDARY PREDICATE. Two examples are (w) and (x), the first with a present
participle, the second with an adjective. The terminology for these constituents varies
from language to language and from author to author.18 They are optional constitu-
ents, which makes them resemble satellites. While they often refer to properties of
nouns and other nominal constituents, they are not part of the noun phrase. In (w),
for instance, flentes is related to principes but it is not an attribute—for which see
§ 2.4. Secondary predicates also occur with implied constituents (especially subjects),
such as ‘we’ in (x). In (y), flens is related to the relative pronoun quae, the subject of its
clause. Details are discussed in Chapter 21.
(w) Postridie in castra ex urbe ad nos veniunt flentes principes.
(‘The next day their leaders came from the city to our camp, crying.’ Pl. Am. 256)
(x) . . . scandentem moenia Romanae coloniae [et] Hannibalem laeti spectamus.
(‘ . . . but now that Hannibal is scaling the walls of a Roman colony, we look on with
indifference.’ Liv. 22.14.7)
16
In LSS the term ‘complement’ was used, but this has its own disadvantages. In this Syntax,
complement will be used differently, as described immediately below.
17
The terms are taken from Quirk et al. (1985). 18
They are called ‘praedicativa’ in LSS: ch. 8.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(y) Is amore misere hanc deperit mulierculam / quae hinc modo flens abiit.
(‘He’s dying wretchedly with love for this woman who went off crying a moment ago.’
Pl. Cist. 131–2)
The third type of function used in this Syntax are PRAGMATIC FUNCTIONS. Three
factors play an important role in the structure of a sentence and in the choice of
specific expressions. The first is the role of a particular constituent in its surround-
ing context—for example, whether it is already present in the preceding context
or will be continued in the following context. The second factor is the speaker’s
or writer’s estimate of what the addressee already knows. The third factor con-
cerns the importance attached to a particular constituent by the speaker and/or
addressee for their communicative interaction. Constituents of a clause therefore
differ in their form and degree of TOPICALITY and FOCALITY: normally a clause
contains at least one entity that the speaker considers to some extent known to,
or at least accessible for, the addressee (the TOPIC of the clause), and which is
therefore a good starting point for providing new or unexpected information (the
FOCUS of the clause).
In this Syntax, various types of TOPIC and FOCUS are distinguished. These prag-
matic functions are especially relevant for word order, the choice between active
and passive voice, and the selection of anaphoric expressions. Exx. (z)–(ab) illus-
trate the use of these terms for describing word order phenomena. In (z), Terentia,
Cicero’s wife, is of course well-known to Atticus, the addressee of the letter, so it is
unproblematic for Cicero to take her as the starting point for the new information
given in the remainder of the sentence. The initial position in the sentence is very
common for topical constituents. In the immediately following sentence—here
(aa)—there is no need to mention Terentia, duly introduced, again. That Cicero’s
daughter Tullia joins in for the greetings is an additional (see et ‘also’) and salient
piece of information and placed at the end of the sentence. Ex. (ab) illustrates a
specific focalizing structure to emphasize the role of Clodium (by contrast with
alios) (a so-called cleft construction).
(ab) Neque tu eras tam excors tamque demens ut nescires Clodium esse qui
contra leges facere, alios qui leges scribere solerent.
(‘And yet you were not so senseless and so infatuated as not to know that it was
Clodius’ part to act in defiance of the laws, and the business of others to formulate
them.’ Cic. Dom. 48)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Details on pragmatic functions (and some of the problems involved) are given in
Chapters 5, 23, and 24.19
In other studies in which systematic attention is paid to the pragmatic structure of
clauses and sentences other terms are used for ‘topic’ and ‘focus’, for example ‘theme’
and ‘rheme’. The term ‘focalization’ also plays a role in narratological studies, in a
sense that is not entirely unrelated to its use in this Syntax, but with which it should
not be confused.
19
For a description of Latin word order along pragmatic lines, see Spevak (2010a) and Hoffmann
(2010).
20
See, for example, Dokkum (1900).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(iv) COMMUNICATIVE PURPOSE, which results in various TEXT TYPES, such as narrative
texts, argumentative texts, didactic texts, poetic and prose texts of various
types (this coincides partly with ‘literary genre’). The effect of text type on
the linguistic features of texts is vastly underestimated in Latin scholarship;
one example: quoniam ‘since’ occurs some hundred and forty times in
Lucretius and ‘only’ less than fifty times in Livy’s much longer work.
Lucretius’ is an argumentative text, Livy’s a narrative; almost all his instances
occur in the orations, the most argumentative parts of his work;
(v) STYLE, defined as the set of individual choices made by the speaker/writer(s)
within limits imposed by items (i)–(iv), such as periodic vs non-periodic
sentences, archaic vs contemporary expressions, analogy vs anomaly, adopt-
ing foreign elements (for example, Greek) vs purism.21
The coherence of discourse derives in the first place from the fact that the ‘world’ it
describes has its own internal coherence, which is manifested in the semantic
relations that exist between the individual sentences and paragraphs. These must
share some overarching DISCOURSE TOPIC. But there are also specific linguistic devices
that enhance discourse coherence. In (a) at the beginning of this chapter, the
connector at was mentioned. Connectors play an important role in marking inter-
sentential relations and relations between other discourse units, such as paragraphs.
A few further devices were mentioned in the paragraph on pragmatic functions
in § 2.12. Anaphoric expressions, active/passive variation, the use of the tenses, etc.
contribute to discourse coherence. Further discussion will be found especially in
Chapter 24.
21
For a discussion of such ‘discourse decisions’ see Dik (1997: II.415–22).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
CHAPTER 3
This chapter serves as an introduction to the various word classes (also called ‘lexical
categories’ or ‘parts-of-speech’) of Latin and to the morphosyntactic and morpho-
semantic categories that are relevant to them. It covers only those aspects that are
relevant to the chapters on syntax that follow; references indicate the subsequent
sections where each word class is discussed in greater detail. As an introduction to
word classes, its contents will for the most part be familiar to readers of Latin,
although some of the terminology may be new.1
In the first description of the Latin language that has come down to us (Varro’s
De lingua Latina 6.36), Varro observes that from a ‘root’ word like lego ‘I read’,
which is not itself derived from another ‘root’, four types of ‘declined words’ can be
formed:
(i) those that signal tempora ‘times’ but not casus ‘cases’—leges ‘you will read’
and lege ‘read’ (both are finite verb forms in modern terminology);
(ii) those that signal cases but not times—lectio ‘a reading’ and lector ‘a reader’
(both nouns);
(iii) those that signal both times and cases—legens ‘reading’ and lecturus ‘being
about to read’ (both participles);
(iv) those that signal neither case nor time—lecte and lectissime ‘choicely’ and
‘most choicely’ (both adverbs).
Case and tense (in modern terminology) have been essential criteria for the classifi-
cation of Latin words ever since.2 In this chapter a threefold distinction is made
between classes of words that have (mainly) ‘nominal’ characteristics, classes of words
that have (mainly) ‘verbal’ characteristics, and classes of words that lack these
characteristics.
1
For a discussion of word classes from a typological perspective, see Bisang (2011).
2
Varro’s linguistic ideas stand in a long tradition of Greek philosophers and grammarians. See
Swiggers and Wouters (2001).
Latin in the Classical period had six cases, here indicated vertically, and two
numbers (numerus singularis and numerus pluralis), here indicated horizontally.
The word dominus is the nominative singular form; the word dominōs, the accusative
plural. In the paradigm of dominus it is easy to identify the part that all the forms
have in common, the so-called ‘root’ or ‘stem’ domin-. One could say that this part
carries the lexical meaning ‘master’. (Determining the stem of rex is slightly more
complicated—see below.) However, in the remaining portion of the forms (the
so-called INFLECTIONAL ENDINGS) it is not possible to identify, for example, a nominative
element and a singular element, or an accusative and a plural element. In Latin
the information concerning case and number is ‘fused’ into one form, as in other
so-called ‘inflecting languages’. In Table 3.1 there are twelve boxes (6 2) for each
noun, but the actual number of distinct forms is eight, and the boxes that coincide are
not the same between the two paradigms. In the paradigm of dominus the datives and
ablatives (singular and plural) are identical, so one might ask what is the reason for
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
distinguishing them. The distinction is based on the paradigm of rex, where the dative
and ablative singular are different (rēgī and rēge respectively); further confirmation is
found in other nominal classes. Nevertheless, the dative and ablative case forms are
never distinct in the plural; the reason for distinguishing these two cases is that the
singular forms are (sometimes) formally distinct.
In order to indicate that the eight words in the paradigms in Table 3.1 are all
different forms of something that means dominus or rex linguists have introduced the
term LEXEME: the words in question are forms of the lexemes dominus and rex (note
that the nominative singular form is used as the label for a noun lexeme).
As Table 3.1 shows, dominus and rex differ in the form of their inflectional endings.
The genitive singular of the lexeme dominus, for example, has the ending -ī; that of
the lexeme rex, -is. On the basis of their inflectional endings nouns are traditionally
divided into five so-called DECLENSIONS (with a number of subclasses). Details follow
in § 3.6.
As for the functions of the inflectional categories number and case, NUMBER is a
grammatical category by which the language user indicates whether the noun he is
using concerns one entity or more than one. Number, therefore, is a semantic
category. Latin has singular and plural forms of nouns. Dominus in (a) concerns
one master, domini in (b) multiple masters. This singularity and plurality is also
reflected in the verb forms misit and absunt, respectively, a phenomenon called
‘agreement’ or ‘concord’ (see Chapter 13).3
(a) Dominus me boves mercatum Eretriam misit.
(‘My master has sent me to Eretria to buy oxen.’ Pl. Per. 322)
(b) . . . abeunt lavatum, perstrepunt, ita ut fit domini ubi absunt.
(‘ . . . they all rushed out to take their baths, chattering away, as happens when the
masters are absent.’ Ter. Eu. 600)
Interestingly, singular and plural forms of a lexeme in the same case are, with very few
exceptions (see note below), always formally distinct. The reason must be that
‘number’ is an important semantic category: it makes a difference whether we are
talking about one dominus or several domini. Not surprisingly, the distinction
‘singular’ : ‘plural’ is well preserved in the Romance languages, whereas the cases
(largely) disappeared.
Not every noun has singular and plural forms. Abstract nouns, such as iustitia
‘justice’, mass nouns, such as aurum ‘gold’, and collective nouns, such as proles
‘offspring’, are never or rarely found in the plural. The Latin term for such nouns is
singularia tantum ‘only singular’. By contrast, some nouns only occur in the plural,
even if only one entity is involved. The Latin term for these nouns is pluralia tantum;
examples are insidiae ‘ambush’, tenebrae ‘darkness’, and reliquiae ‘remains’.4
3
For the generic use of singular and plural nouns, see § 11.118.
4
For a list of ‘defective’ nouns, see G.-L.: 35; K.-H.: 500–19.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The singular/plural distinction is not available for the nominative cases of diēs ‘day’
or ‘days’ and rēs ‘thing’ or ‘things’, although it may be indicated by other modifiers
within the noun phrase.
Two other structures in which case indicates the role of the nouns involved are
illustrated by (d) and (e). In (d) domini is part of the noun phrase nomen domini;
in (e) it is part of the adjective phrase domini similis.
(d) Sed nomen domini quaero quid siet.
(‘But I’m trying to remember what the owner’s name is.’ Pl. Mos. 661)
(e) Domini similis es.
(‘You are similar to your master.’ Ter. Eu. 496)
In (d) domini is the genitive singular form, which makes it clear that it is domini that
modifies nomen ‘name’ and not the other way around. In (e) genitive singular domini
depends on similis ‘similar’.
Although case provides a certain amount of information that contributes to a
correct understanding of the structure in which it occurs, it is of a much more abstract
character than the information number provides. Case is a MORPHOSYNTACTIC category,
as opposed to number, which is a MORPHOSEMANTIC category. (For detailed discussion
of the function of case, see Chapter 12.)
neuter. While the lexeme bonus has separate forms for three genders, there are also
adjectives with one set of forms for the feminine and masculine and another for the
neuter (see § 3.7). Note that the singular nominative masculine form is used as a label
to identify adjectival lexemes.
It is not difficult to make a distinction for the forms of the lexeme bonus between
the stem (bon-) and the inflectional endings. However, just as with nouns, it is not
possible to isolate within the endings separate elements for number, case, and gender.
Like nouns, adjectives belong to various declensions. Table 3.2 shows one of them,
the so-called first and second declension, which contains masculine and neuter forms
of the -o stem and feminine forms of the -a stem. There is another rather complex
declension, containing -i stem and consonant stem adjectives. See § 3.7 for details.
Turning now to the function of these inflectional categories, although they bear the
same names as the equivalent categories with nouns, their function is not entirely
the same. The expression of these categories in the inflection of adjectives serves the
purpose of showing to which noun or comparable constituent the adjective belongs.
This is illustrated by (a)–(c). Ex. (a) is a textually uncertain example, accepted in this
form by most editors.5 Despite vetus being separated from dominus by villae, the
ending (singular, nominative, masculine/feminine) shows that it modifies dominus,
itself a singular and nominative form (and intrinsically masculine—see below). In (b)
the fact that quartus is a singular, nominative, and masculine form helps to disam-
biguate diēs as singular and nominative (the verb ending is also singular). In (c) the
fact that the form of the possessive adjective meus is singular, nominative, and
masculine makes it clear that the canis ‘dog’ involved is a male one.
(a) . . . cum tu vetus (eius cj. Baiter) villae dominus sis cuius paulo ante fuerat
Caesar.
(‘ . . . when you were the old established master of the villa in which the master had
once been Caesar.’ [Cic.] Sal. 20)
5
Reynolds (OCT) prints eius while reporting that the mss. read vetus. The pre-OCT editors report that
almost all mss. have veteris, which would then go with villae, but this is currently rejected by all editors.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(b) Itaque expositis tridui disputationibus quartus dies hoc libro concluditur.
(‘Accordingly, now that the discussions of three days have been set out in full, the
fourth day is comprised in this book.’ Cic. Tusc. 4.7)
(c) Et praeter eos agnos meus est istic clam mordax canis.
(‘And besides these lambs my dog is secretly in there, a real biter.’ Pl. Bac. 1146)
Exx. (a)–(c) concern adjectives belonging to a noun phrase. Adjectives that function
as a constituent of a clause (for example, as subject complement) behave in a similar
way. This is illustrated by (d) and (e). In (d) the subject complement bonus is singular,
nominative, and masculine, as is the subject hic homo. In (e) the speaker is talking to a
woman, present on the stage. As a consequence, the subject complement bona is
feminine in agreement with the implicit feminine subject of the clause.
(d) Bonus est hic homo, mea voluptas.
(‘This man is a good one, my darling.’ Pl. Poen. 1214)
(e) Bona si esse vis, bene erit tibi.
(‘If you’re willing to be good, you’ll have a good time.’ Pl. Mer. 510)
Except in the case of the so-called substantival use of adjectives and other comparable
word classes (see § 3.5), their inflectional endings depend on the nouns or other
nominal constituents to which they are related. The technical term for this phenom-
enon is AGREEMENT or CONCORD. Details are discussed in Chapter 13.
Whereas adjectives and other words are inflected for gender, the inflectional
endings of nouns do not contain an element of gender. There are a number of
masculine and feminine suffixes such as -tor and -trix that can be used to derive
nouns from verbal stems (e.g. laudator ‘a male person who praises’, laudatrix ‘a
female person who praises’), but the declension of these lexemes does not contain
specific masculine or feminine forms. The gender of nouns is an intrinsic property
and to a large extent arbitrary.6
Grammatical gender of nouns is partly related to sex (male animate entities are
masculine; female animate entities are feminine), partly to semantic properties, but
mostly to formal properties. As for semantic properties, the names of months and
winds are masculine and so are most names of rivers and mountains that could be
considered familiar; names of countries, islands, cities,7 and plants and trees are
usually feminine. A few illustrations are given in the Supplement.8
Supplement (modifiers added to indicate gender):
Ianuarius (sc. mensis ‘month’) ‘January’; Auster imbricus ‘the rainy south wind’;
flavus Tiberis ‘the yellow Tiber’; Vesuvius noster ‘our Vesuvius’;
6
For the relationship between gender and sex in Latin, see Löfstedt (1963a).
7
For the names of cities, see Biville (1998: 833).
8
For more instances, fluctuation, and change, see G.-L.: 10–11; K.-H.: 258–72; Sz.: 10–12; Adams
(2013: 383–452).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
. . . dilectus tota Italia habiti ‘levies held throughout Italy’; Latonia Delos ‘Latona’s
Delos’; disciplinis erudita Corinthus ‘Corinth with her polished culture’; alba popu-
lus ‘white poplar’.
A number of nouns have the same forms for male and female animate beings.
The technical term is ‘common gender’. An example is canis ‘dog’ quoted in (c).
There the modifier meus by its masculine form supplies additional information.
Another solution is the addition of mas ‘male’ or femina ‘female’, as in bos mas ‘bull’.
Examples of nouns denoting human beings are civis ‘citizen’ and comes ‘companion’.
As for formal properties, a few examples may suffice. (i) Nouns with a nominative
singular form in -a, such as mensa ‘table’ (that is, nouns belonging to the so-called
first declensional class—see § 3.6), are feminine, unless one of the semantic properties
mentioned above applies. So Anchisa (better known as Anchises), nauta ‘sailor’,
and Hadria ‘the Adriatic’ are masculine. (ii) Nouns with a nominative singular
form in -us, such as hortus ‘garden’ and fructus ‘fruit’, are masculine, unless one of
the semantic properties mentioned above applies. So Corinthus and populus ‘poplar’
in the Supplement above are feminine (there are a few more exceptions, which are not
relevant to this exposition). (iii) Nouns with a nominative singular form in -um,
such as bellum ‘war’, are neuter, unless one of the semantic properties mentioned
above applies. So, for example, amoena Stephanium ‘pretty Stephanium’ (a woman’s
name in Greek) with a feminine adjective. (iv) Nouns with a nominative singular
form in -or, such as labor ‘toil’, are masculine, unless one of the semantic properties
mentioned above applies. So arbor ‘tree’ is feminine.
The neuter gender has a number of special functions of a different character. First
there is the use of singular neuter modifiers with various linguistic entities other than
nouns, illustrated by (f )–(i). In (f ) the letter ‘E’ is taken as a neuter form, modified by
plenissimum. In (g) the determiner istuc modifies the quotation taceo, which is treated
as a neuter word. In (h) novum modifies the indeclinable adverb mane (see also
§§ 11.22–11.23). In (i) the possessive adjective tuum modifies the infinitive amare,
which is also treated as a neuter word (for a detailed discussion, see § 11.8).
(f ) . . . ut Iota litteram tollas et E plenissimum dicas . . .
(‘ . . . in dropping the letter I and substituting a very full E . . . ’ Cic. de Orat. 3.46)
(g) Taceo. # Si tacuisses, iam istuc “taceo” non gnatum foret.
(‘I’m quiet. # If you'd been quiet, your “I’m quiet” wouldn’t have been born.’
Pl. Poen. 262)
(h) . . . dum mane novum, dum gramina canent . . .
(‘ . . . while the day is young, while the grass is hoar . . . ’ Verg. G. 3.325)
(i) Ita tuom conferto amare semper . . .
(‘Always handle your love in such a way . . . ’ Pl. Cur. 28)
clause as subject or object. Examples are (j) (an infinitive) and (k) (an accusative and
infinitive clause). The infinitive and the clause are treated as a neuter constituent.
For details concerning this common use, see § 15.126 and § 15.101, respectively.
(j) Facilitatis, liberalitatis . . . signa proferre perutile est.
(‘It is very helpful to display the tokens of good-nature, kindness, . . . ’ Cic. de Orat. 2.182)
(k) . . . te meminisse id gratum est mihi.
(‘ . . . I’m thankful that you remember it.’ Pl. Capt. 414)
3.6 Nouns
The category of nouns is usually further subdivided into COMMON NOUNS (nomina
appellativa) and PROPER NAMES (nomina propria). Proper names typically refer to a
particular individual entity, such as a person (Marcus), a river (Tiber), or a hill
(Aventinus). By contrast, common nouns when not determined in some way do not
refer to individual entities but denote entities as part of a class (homo ‘man’, amnis ‘river’,
mons ‘mountain’). Common nouns (henceforth typically referred to simply as ‘nouns’)
are discussed first. Proper names are then dealt with at the end of this section.
As was said in § 3.3, nouns are assigned to different declensions on the basis of their
inflectional endings. The five classes are presented below in their traditional order.
The lexemes that serve as examples are given in their singular nominative and
genitive forms.9
(i) First declension, also called -a stem declension, containing lexemes like puella
(nominative singular), puellae (genitive singular) ‘girl’ and mensa, mensae ‘table’.
(ii) Second declension, also called -o stem declension (for historical reasons that
are not relevant here), containing lexemes of various types: (a) dominus,
domini ‘master’ and hortus, horti ‘garden’; (b) puer, pueri ‘boy’ and ager, agri
‘field’; (c) bellum, belli ‘war’.
(iii) Third declension, usually divided up into consonant stems of various types
and vowel stems: (a) consonant stems: (i) consul, consulis ‘consul’, nomen,
nominis ‘name’, and labor, laboris ‘toil’; (ii) cinis, cineris ‘ashes’, flos, floris
‘flower’, and corpus, corporis ‘body’; (iii) princeps, principis ‘chief ’, rex, regis
‘king’, and aetas, aetatis ‘age’; (b) vowel stems: (i) collis, collis ‘hill’, vulpes,
vulpis ‘fox’, and mare, maris ‘sea’; (ii) grus, gruis ‘crane’.
(iv) Fourth declension, also called -u stem declension: fructus, fructūs ‘fruit’ and
cornū, cornūs ‘horn’.
(v) Fifth declension, also called -e stem declension: dies, diei ‘day’, res, rei ‘thing’.
The first and second declension share a number of endings: for example, the plural
genitive in -rum and the plural dative and ablative in -is. The two declensions can
therefore be regarded as belonging to one ‘macro’ declension, as is seen with
adjectives like bonus as well (see Table 3.2 in § 3.4).10
9 10
For details, see G.-L.: 13–34; K.-H.: 296–471. See Dressler (2002).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Common nouns denote entities of four types. The first type concerns entities that are
located in space and time and are observable, such as vir ‘man’, equus ‘horse’, lapis
‘stone’, domus ‘house’, fluvius ‘river’, sonus ‘sound’, aqua ‘water’, aurum ‘gold’,
pecunia ‘money’, exercitus ‘army’, and gens ‘clan’ (‘first-order’ entities). Nouns of
this type are often called ‘concrete’. They can be subdivided into a number of
semantic classes: animate nouns (vir and equus) vs non-animate nouns (lapis,
domus); count (vir, lapis, and sonus) vs mass (aqua, aurum, and pecunia); individual
(vir, lapis) vs collective (exercitus, gens). This type of nouns is often regarded as the
most prototypical. The remaining three types of nouns are ‘abstract’. The second type
concerns entities that are located in time and denote an event or a situation that is
observable. Examples are dolor ‘grief ’, adventus ‘arrival’, admiratio ‘astonishment’,
nuntius ‘message’, and pietas ‘respect’ (‘second-order’ entities). The third type concerns
entities that are located neither in space nor in time and denote mental constructs, such
as opinio ‘opinion’, memoria ‘memory’, iudicium ‘judgement’ (‘third-order’ entities).
Many of the nouns belonging to these two classes are derivationally related to verbs that
indicate an event or a situation. The fourth type concerns entities that denote a property
of some entity, such as color ‘colour’, modus ‘size’, and magnitudo ‘size’ (‘zero-order’
entities). Property nouns are often derivationally related to adjectives.11
Some nouns that denote concrete countable entities, like domus ‘house’ and malum
‘apple’, may occur with all sorts of modifiers (colour, ownership, provenance, quality,
etc.), but they are perfectly understandable without any specification. Other nouns
like the kinship word pater ‘father’ are inherently RELATIONAL, in the sense that one is
always the father of someone. So in (a), when the slave Tranio says to his young
master Philolaches: ‘pater adest’, pater is understood as ‘your father’, as becomes clear
further on by the addition of tuos. Noun phrases related to these nouns can also be
expressed in the dative, like mihi in (b) below (for details see § 10.96).
(a) Philolaches. # Quid est? # <Et> ego et tu—# Quid ‘et ego et tu’? # Periimus.
/# Quid ita? # Pater adest. # Quid ego ex te audio? # Absumpti sumus. / Pater
inquam tuos venit.
(‘Philolaches. # What’s the matter? # You and I—# What, “you and I”? # We’ve died.
# How so? # Your father’s back. # What must I hear from you? # We’re done for. I’m
telling you, your father has come.’ Pl. Mos. 364–6)
(b) Ampsigura mater mihi fuit, Iahon pater.
(‘My mother was Ampsigura and my father Iahon.’ Pl. Poen. 1065)
Comparable are nouns that denote a part (including mental capacities) of the body.
These entities are INALIENABLE and the nouns that are used for them are in many
languages favoured in or excluded from certain constructions. So one does not
normally expect the possessive dative construction with caput ‘head’, as in (c), nor
11
The distinction of four types of entities goes back to Lyons’ distinction of first- and second-order
entities (1968: 346–9). For its application to Latin, see Spevak (2014a: ch. 1).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
a possessive adjective, as in (d). On the other hand, one does find such nouns in the
sympathetic dative construction, exemplified by (e) (for details, see § 10.96).
(c) ?Mihi caput est.
(‘I have a head.’—Mihi caput infirmum est is permissible, see Cels. 1.4.1)
(d) ?Meum caput dolet.
(‘My head hurts.’—Caput dolet is permissible, see Pl. Am. 1059)
(e) . . . etiam <mi> misero nunc malae dolent.
(‘ . . . my jaws are still hurting, dear me.’ Pl. Am. 408)
Just like the verbs from which they are derived, abstract verbal nouns may require one
or more entities to be expressed explicitly or to be understood in the context. Thus
profectio ‘departure’, navigatio ‘sailing’, and desperatio ‘despair’ usually denote some-
one’s departure, the sailing of someone (by boat), and the despair of someone
concerning somebody or something, respectively (although a discussion in abstracto
is possible). A distinction can be made between monovalent, bivalent, and trivalent
nouns, with which one, two, or three ‘adnominal arguments’ are required, respect-
ively. Actor verbal nouns like actor ‘performer’ and dux ‘guide’ likewise normally
require an entity that might be the second argument of the verb to which these nouns
are related. Further examples of nouns with a valency are facultas ‘ability’ and
necessitas ‘need’. Details concerning the valency of nouns are given in § 11.70.
The nouns discussed above often develop more ‘concrete’ meanings. One may think
of words like oratio, which rarely denotes the act of speaking12 and more often
denotes the product of speaking (an oration) or other comparable objects. In these
cases these nouns have no valency.13
Proper names have a number of formal properties of their own (archaisms, many loan
words). However, on the whole they fit in with the morphology of common nouns.14
Although they typically refer to definite individuals, they can be used in the same way
as common nouns. For example, they can be modified by the same range of categor-
ies, as in (f ) by an indefinite determiner (for this term, see § 3.8). They can also be
used in the plural, as in (g) (see also § 11.4).15
(f ) Et tamen non alienum est dignitate tua . . . habere aliquem in consiliis
capiendis Nestorem . . .
(‘And yet it detracts nothing from your prestige to have a Nestor to consult . . . ’ Cic.
Fam. 9.14.2)
(g) Eius testamentum deporto trium Ciceronum signis obsignatum . . .
(‘I am bringing his will home, witnessed by three Ciceros . . . ’ Cic. Att. 7.2.3)
12 13
See TLL s.v. 877.68ff. See Rosén (1983) and Spevak (2014a: 33–5).
14
For the formal properties of proper names, see Biville (1998).
15
See Orlandini (1995: 148–68) and Biville (1998: 832–3).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
3.7 Adjectives
Latin adjectives belong to two declensional classes whose inflectional endings corres-
pond to the first, second, and third declensions of nouns. The first class, with endings
from the first and second declensions, is shown in Table 3.2. The other class contains
endings from the third declension. As with the nouns of the third declension, the
stems of the adjectives in this class are quite diverse. Usually a distinction is made
between -i stem adjectives and consonant stem adjectives. The -i stem adjectives (for
example facilis ‘easy’ and acer ‘sharp’), have ten different forms (out of a possible
thirty-six), the consonant stems (for example felix ‘lucky’) only nine.16 Singular and
plural forms of an adjectival lexeme in the same case are always distinct.
Another common division of adjectives is according to the number of different
singular nominative forms for the inflectional category gender: three endings (fem-
inine, masculine, and neuter), e.g. bonus; two endings (non-neuter and neuter), e.g.
facilis; and one ending, e.g. felix.
A noun can be modified by more than one descriptive adjective. If these adjectives
belong to the same class they are usually coordinated. If they belong to different
classes they cannot be coordinated but are juxtaposed. An example of juxtaposition is
(b). For further examples, see § 11.39 and § 11.75. For coordination of adjectives, see
Chapter 19.
(b) Contra haec Pompeius naves magnas onerarias quas in portu Brundisino
deprehenderat adornabat.
(‘To meet this Pompeius fitted out some large merchant-ships which he had seized in
the port of Brundisium.’ Caes. Civ. 1.26.1)
16
For details, see G.-L.: 37–43; K.-H.: 533–50.
17
This classification is based on Spevak (2014a: 58), who uses Hetzron (1978: 178) and Risselada
(1984). For other classifications see Dixon (1977), Givón (2001: 81–4), and Wetzer (1996).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Adjectives at the beginning of the scale usually stand closer to their head noun
(onerarias in (b) is an exception) and more often follow that noun. They are often
non-gradable (see below).
Another important semantic parameter is that of ‘temporal stability’: the distinc-
tion between adjectives that indicate a permanent and those that indicate a non-
permanent or transient property.18 Adjectives at the beginning of the scale usually
indicate a permanent, those at the end a non-permanent property. Adjectives that
indicate a non-permanent property can be used as a secondary predicate. An example
is maestus ‘sad’ in (c). Adjectives that indicate an evaluation cannot be used as
secondary predicates. For details, see Chapter 21.
(c) Mnesilochus eccum maestus progreditur foras.
(‘There, Mnesilochus is coming out with a sad look on his face.’ Pl. Bac. 611)
Whereas some adjectives indicate only a property of the constituent they are related
to (so, for example, maestus in (c)) other adjectives require a second entity. This is, for
instance, the case with cupidus ‘eager (for)’, ‘desirous (of)’, which requires another
entity in the genitive, as in (d). Cupidus is a bivalent adjective. Valency of adjectives is
discussed in § 4.99.
(d) Leno ad se accipiet auri cupidus ilico.
(‘Eager for the money, the pimp will receive him at once.’ Pl. Poen. 179)
Adjectives of amount (multus ‘many’, paucus ‘few’) share a number of properties with
(some) descriptive adjectives: they are gradable and can be used as secondary
predicates. In addition, they can be coordinated with some descriptive adjectives,
although they are usually juxtaposed. For details, see § 11.33.
From many adjectives comparative and superlative forms can be derived that
indicate that the relevant property applies to the entity in a higher or in the highest
degree of intensity. Examples of the POSITIVE, COMPARATIVE, and SUPERLATIVE DEGREE OF
COMPARISON of a few adjectives are given in Table 3.3.
Adjectives in the comparative degree are inflected as adjectives of the third declension;
those in the superlative degree as adjectives of the first and second declension.
18
For this parameter, see Givón (2001: I.50–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
19
For a fuller list, see K.-H.: 565–6; TLL s.v. magis 61.24ff.; 72.58ff. For the notion of gradability, see
Bertocchi (2002); Hoffmann (1987: ch. 5).
20
See K.-H.: 566–9; TLL s.v. magis 61.5ff.; 72.44ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The other pronouns have in common the fact that they can be used both inde-
pendently (substantively) and as determiners (attributively). In this Syntax the terms
that are used are (independent) PRONOUN and DETERMINER.21 These two uses are
illustrated for the demonstrative pronoun/determiner ille ‘that’ by (a) and (b) respect-
ively. With a few exceptions the forms are the same for the independent and the
determinative use. Exceptions are, among others, the singular nominative masculine
forms quis? ‘who?’ and qui (vir)? ‘which (man)?’ and the singular nominative/accusa-
tive neuter forms quid? ‘what?’ and quod (facinus)? ‘what (deed)?’.
(a) Illa illum censet virum / suom esse . . .
(‘She believes he’s her husband . . . ’ Pl. Am. 134–5)
(b) Sed mos numquam <ille> illi fuit patri meo . . .
(‘But that father of mine never had that habit . . . ’ Pl. Am. 46)
The three lexemes in (ii) are classed together for etymological reasons: idem can
plausibly be analysed as < *is + dem; for ipse this is less straightforward but generally
accepted.23 However, in their actual use they are quite diverse. In this Syntax the three
will be designated as anaphoric (is), identifier (idem), and intensifier (ipse). Details are
discussed in § 11.26, § 11.31, and § 11.32 respectively.
The singular genitive and dative forms that are typical of the pronouns/determiners
mentioned above are also found with a number of words which for the remainder of
their forms behave like adjectives of the first and second declension (bonus).24 These
words are the numeral unus ‘one’ (it also means ‘only’, ‘alone’ and can be used like
solus ‘alone’); the indefinite determiner ullus (see § 11.27); the identifiers alius ‘other’
and alter ‘the other of two’ (see § 11.31); the quantifiers nullus ‘not any’ (see § 8.33)
and neuter ‘neither’ (see § 11.37); the totality expression totus ‘whole’ (see § 11.38);
21
On determiners, see Touratier (2010: 131–2).
22 23
See, for example, G.-L.: 57–62. See de Vaan (2008: s.v.).
24
An exception is the singular nominative/accusative neuter form aliud.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
3.9 Numerals
The class of numeral adjectives traditionally contains three subclasses, of which two
contain real quantifiers (the cardinal and distributive numerals); the third one (the
ordinal numerals) contains adjectives that indicate position (in time or place), such as
primus ‘first’. Examples of cardinal numerals are unus ‘one’, viginti ‘twenty’, and
ducenti ‘two hundred’; of distributive numerals, singuli ‘one each’, viceni ‘twenty
each’, and duceni ‘two hundred each’. For their syntactic properties, see § 11.33.
Of the cardinal numerals only unus ‘one’, duo ‘two’, tres ‘three’, and the hundreds
beginning with ducenti ‘two hundred’ are inflected for number, case, and gender.
The neuter noun milia ‘thousands’ is a plural form and only inflected for case.
The distributive numerals are inflected as adjectives of the first and second declension
(bonus).
3.10 Verbs
Verbal lexemes have a wide variety of forms with very diverse formal, semantic, and
syntactic properties. A first distinction is between so-called FINITE and NON-FINITE verb
forms. Finite verb forms are inflected for a number of semantic categories that are
typical of verbs. Non-finite verb forms, while sharing the meaning of the stem and
some categories with the finite verb forms, have also some of the nominal features
discussed in §§ 3.1–3.9. As a first illustration two forms of the lexeme amo ‘to love’
will suffice.
(a) . . . amabat patriam . . .
(‘ . . . he loved his country . . . ’ Cic. Att. 4.6.1)
(b) . . . boni cives amantes patriae . . .
(‘ . . . good citizens who love our country . . . ’ Cic. Att. 9.19.3)
The word amabat in (a) is inflected for four different semantic categories: (i) for
PERSON/NUMBER: third person singular; (ii) for TENSE: past; (iii) for MOOD: indicative; (iv)
for VOICE: active. These categories are discussed in § 3.11. Moreover, (v), amabat is a
form of the stem that indicates SIMULTANEITY, the so-called infectum stem. Amantes in
(b) corresponds to amabat in being (iv) active and (v) simultaneous. In addition, it is
a participle, an adjective-like word that is inflected for number, case, and gender, in
accordance with the third declension of adjectives. In (a) amabat governs patriam,
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Verbs
which is in the usual accusative case; in (b) patriae is in the genitive case, which
is typical for certain adjectives that require a second argument (see auri cupidus in
(d) in § 3.7).
Verbs can be zerovalent (pluit ‘it is raining’), monovalent, requiring one argument
(pater ambulat ‘the father is walking’), bivalent, requiring two arguments (pater amat
filium ‘the father loves his son’), or trivalent, requiring three arguments (pater filio
librum dat ‘the father gives his son a book’). For the concept of valency, see § 2.8. For a
classification of verbs according to the number of arguments they require and the
form of these arguments, see Chapter 4.
The combination of a verb and its argument(s) refers to a certain situation. It is
useful to distinguish a number of types of situations or, in the terminology of this
Syntax, types of ‘states of affairs’. Verbs can be classified for the type of states of affairs
in which they can occur. In § 2.9 four different types of states of affairs are distin-
guished: actions, positions, processes, and states.
Latin verb forms of the active voice are built on two different stems: the INFECTUM and
PERFECTUM stem. The situation is more complicated for the passive voice: whereas for
the infectum stem so-called SYNTHETIC forms are used, comparable to the active forms
of the infectum stem, there are no special perfectum stem forms for the passive.
Instead so-called PERIPHRASTIC forms consisting of the passive perfect participle and
forms of the verb sum ‘to be’ are used. This is shown in Table 3.4.
Besides different stems in the active, there are also a number of special endings used
with the perfect stem, notably the person/number endings of the indicative of the
perfect tense. All this shows that the distinction between the two stems is central to
the structure of the verb.
There is much discussion about the semantic correlates of the distinction between
the infectum and perfectum stem. Whereas some scholars consider the distinction an
‘aspectual’ one (‘non-complete’ vs ‘complete’), others describe the distinction in terms
of ‘simultaneity’ and ‘anteriority’. The latter is the position taken in this Syntax. For
discussion, see § 7.4. For the remainder of this section the distinction infectum/
perfectum will be ignored.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
On the basis of these stems, Latin verb forms are marked for four inflectional
categories: person/number, voice, tense, and mood, which are discussed in that order
below. Just as with nominal forms, it is to some extent possible to make a distinction
between the stems and the inflectional endings: in the words delebat and delebatur in
Table 3.4 it seems reasonable to separate between dele- and the other parts; in deletus
dele- may be isolated; in deleverat the stem is delev- (for complications, see below).
However, it is not possible to isolate formal correlates for each of the four inflectional
categories. Person/number is the most recognizable element, as can be seen from the
present tense forms of the verb deleo ‘to destroy’ in Table 3.5.
Table 3.5 shows that there are different sets of endings for the active and the passive
or, to put it differently, that the information concerning person/number and voice are
combined in one ending. The first person singular shows variation between the indica-
tive and the subjunctive. The -m and -r endings are used elsewhere in the paradigm for
the first person singular; the -o ending is more restricted. The other endings are also used
for the imperfect and future tense forms, as well as for the active perfect, pluperfect, and
future perfect forms (with the exception of the perfect indicative).
It is much more difficult—if not impossible—to isolate in the forms information
concerning tense and mood. In the imperfect form delebat in Table 3.4 we can
separate dele-ba-t, and take ba as the bearer of the information ‘imperfect tense’ +
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Verbs
‘indicative mood’, but in the present form delet in Table 3.5 there is nothing between the
stem dele- and the person/number/voice ending -t. The total meaning of the form can
only be established by taking into account the other forms in the paradigm of deleo.25
Latin verbs can be assigned to four different inflectional classes (one of them
consists of two subclasses) on the basis of their stem and the inflectional endings
used in the infectum forms (for the perfectum they all use the same endings). The
special term that is used for a verbal inflectional class is CONJUGATION. Apart from
these four conjugations there are a number of irregular verbs, notably sum ‘to be’,
possum ‘to be able’, eo ‘to go’, volo ‘to want’. Whereas all conjugations (and to a large
extent also the irregular verbs) have the same person/number/voice endings, they
differ in other respects. These differences are partly determined by properties of the
stem, partly by use of different endings. Two examples will suffice. (i) In Table 3.5 the
forms of the present subjunctive have an -a- between the stem and the person/
number/voice ending. This is the case in two other conjugations, but not in the so-
called first conjugation, which contains verbs with a stem ending in -a, such as the
verb amo ‘to love’. Instead of *ama-am, *ama-as, etc. its present subjunctive forms
are amem, ames, etc. (ii) For the future indicative some conjugations have endings in
-bo, -bis, etc., others endings in -am, -es, etc. This is illustrated by Table 3.6. The -ī stem
verbs (fourth conjugation) adopted the endings of the third conjugation (the conso-
nant and -ı stems), and did the same in the case of the imperfect forms.
Table 3.6 The formation of future active and imperfect active forms
-a stem -e stem Consonant -ı̌ stem -ī stem
stem
1 sg. fut. act. ama-b-o dele-b-o em-a-m capi-a-m audi-a-m
‘I will love’ ‘I will destroy’ ‘I will buy’ ‘I will take’ ‘I will hear’
(Early L. audi-b-o)26
2 sg. fut. act. ama-bi-s dele-bi-s em-e-s capi-e-s audi-e-s
(Early L. audi-bi-s)
1 sg. impf. act. ama-ba-m dele-ba-m em-eba-m capi-eba-m audi-eba-m
(Early L. audi-ba-m)27
25
The best discussion of this aspect of inflectional morphology is Matthews (1972).
26
For a survey of Early Latin (and later) future forms of the fourth conjugation, see K.-H.: 727–8.
27
For a survey of Early Latin (and later) imperfect forms of the fourth conjugation, see K.-H.: 724–5.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
second person forms, in the singular and the plural (see Table 3.5). For the so-called
future imperative forms there are second and third person singular and plural forms
(there is no passive second plural). For the other finite verb forms the situation is as in
Table 3.5: first person singular: the speaker; second person singular: the addressee;
third person singular: an animate or inanimate entity which is not the speaker or the
addressee; first person plural: a plurality including the speaker; second person plural:
more than one addressee or a plurality including the addressee(s); third person plural:
a plurality of animate or inanimate entities, not including the speaker or the
addressee. In the case of the first and second person the identity of the subject is
known. The use of the first and second person personal pronouns as explicit subjects
serves different purposes from mere identification (see § 9.2). In the case of third
person singular and plural forms, the identity of the subject must be expressed
explicitly or be inferred from the context or situation. In addition, the third person
singular is used with zerovalent verbs in which no subject is involved, such as pluit ‘it
is raining’ in (a) (see § 4.90). It is also used in subjectless passive expressions (the so-
called impersonal passive—see § 5.21), as in (b).
As stated above, the information concerning VOICE (genus verbi) is combined with that
of person/number, so all finite verb forms are marked for voice. Latin has two voices,
ACTIVE and PASSIVE. Not all verbs have forms of both voices, and some have one voice
for the infectum forms, the other for the perfectum forms. For those verbs that have
both voices for the entire paradigm, the choice of a specific voice is related to which
argument of the verb is used as subject of the clause. This is illustrated by (c) and (d).
In (c)—active—the action of ‘praising the son’ is presented from the perspective of the
father, the subject of the sentence; in (d)—passive—the same action is presented from
the perspective of the son, the subject of the sentence (for discussion, see § 5.5).
Latin has a considerable number of verbs that only have finite verb forms in the
passive voice. As far as their meaning is concerned, they do not differ from the active
form of verbs that have forms of both voices. Such verbs that are ‘passive in form, but
active in meaning’, as it is often formulated, are called DEPONENTS (verba deponentia).
Deponents are found in all four conjugations. Some of them are one-place verbs;
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Verbs
some have a second argument, mostly in the accusative, but also in other cases and in
prepositional phrases. Examples are given in Table 3.7. Some of these verbs have
alternative active forms and some occasionally have a passive meaning (especially in
the perfect participle); individual verbs also show fluctuation over time (see § 5.33).
Deponents have present and future participles of the active voice, as well as an active
future infinitive (for example: cunctans ‘delaying’, cunctaturus ‘going to delay’, and
cunctaturus esse ‘to be going to delay’).
With a small number of verbs, forms of one voice are used for the infectum stem,
forms of the other for the perfectum stem. This is shown in Table 3.8. These verbs are
called SEMIDEPONENTS.28
With many verbs that have only active finite forms there is one exception: the third
person singular used in impersonal passive clauses (see (b) above and gaudetur in Table 3.8).
Latin has six TENSES, derived from three infectum and three perfectum stem forms:
present ((tempus) praesens), imperfect (praeteritum imperfectum), future (futurum
(simplex)), perfect (praeteritum perfectum), pluperfect (praeteritum plusquamperfec-
tum), future perfect (futurum exactum). They are presented in Table 3.9. The English
translations of the present, imperfect, and perfect tenses do not precisely cover the
values of Latin forms. Depending on the context, a non-progressive translation of the
present and imperfect (‘destroys’ and ‘destroyed’) is more appropriate; similarly, for
28
A list of these verbs can be found in K.-St.: I.97–9.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
the perfect the translation ‘destroyed’ is often more appropriate. For discussion of the
values of the tenses, see §§ 7.11–7.33.
Latin has three MOODS: the indicative (modus indicativus), the subjunctive
(modus subiunctivus or coniunctivus), and the imperative (modus imperativus). The
three moods are shown for the present in Table 3.5. The imperative mood is the
most restricted. Apart from the present imperatives there are only the so-called
future imperatives (see § 7.65). Latin verbs have no future and future perfect sub-
junctive forms. If in a certain context a subjunctive is required, a combination of the
future participle and a subjunctive form of the verb sum ‘to be’ serves as a substitute.
There are imperfect and pluperfect subjunctive forms.
The indicative mood serves as a signal that the speaker or writer commits himself
to the reality or factuality of an event (see § 7.12), as in (e).
(e) In illisce habitat aedibus / Amphitruo . . .
(‘In that house there lives Amphitruo.’ Pl. Am. 97–8)
It is much more difficult to describe the value of the subjunctive. When used in
independent sentences the interpretation depends on the type of sentence: in imperative
sentences the subjunctive signals that it is or was desirable or necessary that a certain state
of affairs come into effect, be prevented, or stop (the so-called deontic use of the
subjunctive), as in (f ). In declarative and interrogative sentences, the subjunctive signals
that a certain state of affairs is possible, likely, or counterfactual (the so-called epistemic
use of the subjunctive, also called the ‘potential’ use), as in (g). For discussion, see § 7.38.
(f ) Taceas, me spectes.
(‘Be quiet and watch me.’ Pl. As. 680)
(g) Videas corde amare inter se.
(‘You can see that they love each other from the heart.’ Pl. Capt. 420)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Verbs
By contrast, in subordinate clauses the subjunctive often does not signal one of the
values just mentioned, but serves as a sign of subordination. An example is the use of
the subjunctive in indirect questions, as in (h). In conclusion, it is difficult to establish
one clear autonomous value for the subjunctive. For extensive discussion, see
§§ 7.128–7.164.
(h) Opservabo quam rem agat.
(‘I’ll observe what he’s up to.’ Pl. Am. 270)
Non-finite verb forms can be subdivided into substantival and adjectival non-finite
verb forms on the one hand and supines on the other. Substantival (in varying degree)
are the present, future, and perfect infinitives and the gerund; adjectival are the
present, future, and perfect participles and the gerundive.
3.15 Infinitives
Latin has a total of six infinitive forms: the present (simultaneous), future (posterior),
and perfect (anterior) each have an active and a passive infinitive. These are presented
in Table 3.10. Deponents have only three infinitives.
The simple infinitives are not inflected, nor is the periphrastic deletum iri. The two
remaining infinitives consist of a participle and the infinitive of the verb sum ‘to be’.
The participial part is inflected for number, case, and gender (as an adjective of the
first and second declension, e.g. bonus).
The simple forms are treated as neuter singular, as is illustrated by (a), where the
neuter singular forms totum and hoc agree with the infinitive philosophari, which
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
functions as subject of the third singular verb form displicet. Infinitives in such
functions keep their verbal characteristics, for example their valency: in (b) petere
governs its second argument honorem. For further discussion of modification of the
(present) infinitive, see § 11.8. For further discussion of infinitives functioning as
subject, see § 9.7; as object, see § 9.15.
(a) Nam quibusdam . . . totum hoc displicet philosophari.
(‘Certain people disapprove of this, the study of philosophy, altogether.’ Cic. Fin. 1.1)
(b) Petere honorem pro flagitio more fit.
(‘To seek public office as a reward for criminal behaviour is the custom.’ Pl. Trin. 1035)
3.16 Gerund
The gerund is the name for a paradigm of four inflected forms, formed from the
infectum stem of (most) verbs with an infix -nd- and the endings of the second
declension of nouns. This is shown in Table 3.11.
In principle, all types of verbs have a gerund, whatever their valency, their morphology,
and the type of state of affairs in which they participate. Thus one finds zerovalent
pluendo (corresponding to pluit ‘it rains’), monovalent ambulando (of the verb ambulo
‘to walk’, semantically an action) and algendo (of algeo ‘to feel cold’, a state), bivalent
delendo (of deleo ‘to destroy’, an action, governing a second argument in the accusative
case) and utendo (of utor ‘to use’, an action, a deponent verb governing a second
argument in the ablative case). However, some verbs lack a gerund, for instance sum
‘to be’ and its compounds, possum ‘can’, and volo ‘to want’ and related verbs.
Although there is no external proof as in the case of the present infinitive, which
can function as head of a noun phrase and be modified by a determiner or another
modifier (see example (a) in § 3.15), gerunds are usually taken as singular neuter
forms. Gerunds can be found in a variety of contexts, in argument, satellite, attribute
and other positions, just like ‘normal’ nouns. They can be governed by prepositions as
well. However, unlike the (present) infinitive, they cannot function as subject or
object (for discussion of this, see § 5.42). It appears, then, that the gerund and the
present infinitive active are in complementary distribution.29 Indeed, the forms of the
29
For contexts in which the gerund and the infinitive are in competition, see § 15.137.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Verbs
gerund have from Antiquity onwards been described as genitive / dative / accusative /
ablative forms of the infinitive. The main difference between the gerund and the
present infinitive active is that only the latter can be modified (see above).
We have seen that the gerund and the infinitive have a number of nominal properties.
Nevertheless, they must not be equated with verbal nouns proper. This may be
illustrated by comparing certain properties of a gerund and a verbal noun derived
from the same verb. There are several instances in which a gerund and a verbal noun
are used in the same context. Examples are (a)30 and (b)–(c).
(a) Quid ergo, si duas causas latitandi habuit vel plures, inter quas etiam
fraudandi creditores? . . . si plures causae sint latitationis, inter quas est et
fraudationis causa . . .
(‘What, then, are we to say, if the debtor has two or more grounds for hiding, among
them being fraud on his creditors? . . . if he has several reasons for hiding, one being
fraud on his creditors . . . ’ Ulp. dig. 42.4.7.6)
(b) . . . qui . . . cupiditate inimicos ulciscendi arderent . . .
(‘ . . . who were fired with a desire for vengeance on their enemies . . . ’ Liv. 29.6.7)
(c) Hos . . . ultionis cupiditas ad virtutem accendit.
(‘The former are fired to brave action by eager desire for vengeance.’ Tac. Hist. 2.77.3)
When one compares the behaviour of ulciscendi and ultionis in more detail, the
verbal properties of ulciscendi become more apparent (it is found with a second
argument, as in (b)), while its nominal properties appear to be more restricted. The
head nouns with which it occurs are nouns like cupiditas (bivalent nouns), whereas
ultionis is also found in a combination like solacium ultionis (‘the comfort derived
from revenge’ Apul. Met. 2.28.5). In addition, ultio can be modified by determiners,
adjectives, and relative clauses, all of which are excluded with the gerund.
30
Aalto (1949: 120–4, 133) gives many examples of gerunds and nouns in coordinate and parallel
structures as a proof of the substantival properties of the gerund.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
3.19 Participles
Most Latin verbs have three participles: present active, future active, perfect passive.
For verbs with both an active and a passive voice the system of participles is therefore
defective, as is shown in Table 3.12. There are three participles for deponents as well,
all with an active meaning, as with proficiscor ‘to depart’ in Table 3.12.
The future and perfect participles are inflected as adjectives of the first and second
declension (bonus), the present participle as adjectives of the third declension. The
future and perfect participles are used as elements of the future and perfect
infinitive, in combination with the infinitive esse. The perfect participle is also
used with forms of the verb sum ‘to be’ to form periphrastic finite forms of the
perfectum stem (see Table 3.4). The future participle can be used with forms of sum,
among other things as a substitute for lacking future subjunctive forms. The present
participle does not become common in periphrastic expressions until very Late
Latin (see § 7.78).
When used on their own, participles may fulfil various functions at the clause and
noun phrase levels: (i) as subject complement, as in (a); (ii) as an equivalent of a finite
verb form in a subordinate participial (ablative absolute) clause, as in (b); (iii) as
secondary predicate, as in (c); and (iv) as modifier at the noun phrase level, as in (d).
In addition, (v), a participle can be used substantively, as in (e). Further examples can
be found in later chapters.
(a) Omnino est amans sui virtus.
(‘I grant that Virtue loves herself.’ Cic. Amic. 98)
(b) Sed mirus invaserat furor . . . ut pugnare cuperent, me clamante nihil esse
bello civili miserius.
(‘But a strange madness was abroad . . . . They were possessed with the lust of battle,
while I cried aloud that nothing is worse than civil war.’ Cic. Fam. 16.12.2)
(c) . . . prius quam istam adii atque amans ego animum meum isti dedi . . .
(‘ . . . before I came to her and, being in love, gave her my heart . . . ’ Pl. As. 141)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Verbs
(d) Canum vero tam fida custodia tamque amans dominorum adulatio . . . quid
significat aliud nisi se ad hominum commoditates esse generatos.
(‘Then think of the dog, with its trusty watchfulness, its fawning affection for its
master . . . —what do these qualities imply except that they were created to serve the
conveniences of man?’ Cic. N.D. 2.158)
(e) Nam ubi amans complexu’st amantem . . .
(‘Yes, when a male lover embraces his female lover . . . ’ Pl. Ps. 1259)
For each of these functions, parallels with an adjective instead of a participle could be
given, for some more easily than for others; the most typically adjectival one is (d).
Participles vary in the degree to which they can fulfil each of the five functions
illustrated above. Of the four regular participles, the active future participle is the
least ‘adjectival’ one: it is uncommon, at least in Classical Latin, for it to be used in the
functions illustrated by (b), (c), and (d). The active present and passive perfect
participles can—depending on their meaning—fulfil all five functions. The perfect
participles of deponent verbs are uncommon in the functions illustrated by (a), (d),
and (e).
Depending on the meaning of the verb, some participles can in a specific context
denote a permanent property, thus behaving like an adjective, while in other contexts
they maintain their verbal properties as denoting an action or a process that is
anchored to the time of the main verb. The present participle amans in (a) and (d),
for instance, denotes a permanent property, whereas the present participles in (b) and
(c) denote ongoing actions. In their adjectival behaviour such participles share other
features with ‘normal’ adjectives. This is illustrated below for the present participle.
For the perfect passive participle similar examples could be given.31
If a present participle denotes a permanent property and has a gradable meaning, it
can be used in any of the three degrees of comparison, as is shown in (f ) and (g) for
the comparative and superlative, respectively. Note in (f ) the presence of equino
armento, the entity with which the cow is compared (an ablative of comparison).
(f) Omnis tamen externi frigoris tolerantior equino armento vacca est . . .
(‘Cows, however, endure every outdoor cold better than horses . . . ’ Col. 6.22.2)
(g) . . . L. Caesar, vir fortissimus et amantissimus rei publicae . . .
(‘ . . . Lucius Caesar, a man most courageous and devoted to the Republic . . . ’ Cic.
Catil. 4.13)
These participles may be negated by the prefix in- ‘un-’, as in (h) and (i), and in that
way become regular adjectives.32 (See also § 8.51.)
31
For further discussion, see § 4.102.
32
For criteria to distinguish pure adjectives from pure verbal forms, such as participles and compo-
nents of periphrastic forms like laudatus est, see Eklund (1970: 13–14) and Dietrich (1973). For a critique
of these criteria, see Hoffmann (1997a: 10–15).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Adverbs may be derived from them as from regular adjectives—for instance, amanter
‘lovingly’ and toleranter ‘with fortitude’.
Such present participles can govern a constituent in the genitive, the typical case for
internominal relations, even though the corresponding verb governs a different case
(see the introduction to § 3.10). Examples are marked in italics in (a), and (f)–(i).
Another example is (j). For further examples, see § 4.102.
(j) Piaculum est miserere nos hominum rei male gerentum.
(‘It would be a sin for us to take pity on those who mismanage their affairs.’
Pl. Truc. 223)
In principle, all two- and three-place verbs, including deponents, have gerundive
forms.33 They are formed from the infectum stem with an infix -nd- and an ending
identical to the ending of adjectives of the first and second declension (bonus):
delendus ‘to be destroyed’, patiendus ‘to be tolerated’. The gerundive shares with
the participles agreement with the noun (phrase) to which it is related. It was indeed
described as a future passive participle by Latin grammarians but was not used as such
before the third century AD. Although it adapted to some new uses, the form was used
less and less in the course of time and, unlike the gerund, did not continue into the
Romance languages. Some verbs have no gerundive: for instance, sum ‘to be’ and its
compounds, possum ‘to be able’, and volo ‘to want’ and related verbs.34
There are several types of context in which gerundives can be used, though not all
gerundives occur in all of them; for this chapter it will suffice to mention three
contexts. Gerundives can, firstly, be used in combination with the auxiliary esse (‘to
be’) to form a passive deontic expression, as in (a) and (b)—the latter a deponent verb.
Further details concerning this use can be found in § 5.37–5.41 and § 7.84.
(a) Idem nunc vobis Naso legendus erit.
(‘You should read the same Naso now.’ Ov. Rem. 72)
33
For a number of notable exceptions, see § 5.39, Appendix.
34
For ancient discussions of the gerundive, see Aalto’s historical survey (1949: ch. I). For the
diachronic development, see § 5.42. For the relationship between the various uses of the gerundive and
the semantic value of the -ndo- morpheme, see Joffre (1995, 2002), who regards the various uses as
contextually determined variants of a basic value ‘incomplete’.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Verbs
The second use of gerundives is in combination with nouns or noun phrases to form
gerundival clauses which function to some extent as abstract verbal nouns. An
example is (c), where the gerundival clause pecoris pascendi—a concrete noun
combined with a gerundive that is in agreement with it—is coordinated with the
noun phrase agri culturae. Both agri culturae and pecoris pascendi are attributes of
disciplinam. Gerundival clauses are discussed in § 15.139, § 16.105, and Chapter 17.
(c) . . . habere utramque debet disciplinam, et agri culturae et pecoris
pascendi . . .
(‘ . . . he ought to have a knowledge of both pursuits, agriculture and cattle-raising . . . ’
Var. R. 2. pr.5)
In both these uses the gerundives do exhibit the adjectival property of agreement, but
otherwise they behave as verbal forms. They can, for instance, be used with various
types of arguments (depending on the valency of the verbs) and with satellites
(manner adverbs, for example).
The third type of context in which (some) gerundives are used consists of the
standard adjectival uses as attribute and subject or object complement. Sometimes it
can be difficult to tell whether a given form should be analysed as gerundive or
adjective. Examples of the gerundive used as an attribute and as a subject complement
are (d) and (e), respectively.35
(d) Et Hasdrubalem propediem adfore cum manu haudquaquam contemnenda.
(‘And Hasdrubal, they said, would soon arrive with a force by no means to be
despised.’ Liv. 30.7.10)
(e) . . . quamquam est in dicendo minime contemnendus, prudentia tamen
rerum magnarum magis quam dicendi arte nititur.
(‘ . . . though no mean speaker, he yet relies rather on his knowledge of higher politics
than on the art of oratory.’ Cic. de Orat. 1.214)
In such cases, the gerundive resembles adjectives in -bilis, for example contem(p)tibilis
‘contemptible’, derived from the same verbal stem as contemnendus. Comparative
forms of gerundives are very rare, even with those that have almost become adjec-
tives36 and adverbs ending in -e should not be derived from them according to some
Latin grammarians.37 The prefix in- is used to form negative forms, such as
35
Gerundives of deponent verbs are very rare in this function.
36
This was already noted by Augustine (V.495K). Aalto (1949: 104) gives a few superlatives: infan-
dissimus (Varius ap. Quint. Inst. 3.8.45); admirandissimus (Salv. Ep. 8.2); laudandissimus is found in
Oribas. Syn. 6.13.3. See also Neue-W.: II.241.
37
Pompeius (V.36K); Cledonius (V.65). According to them, one should not say horrende resonat, but
horrendum. As a translation of Greek adverbs horrende is found in Vulg. Sap. 6.6, 17.3. The TLL has
another late example of horrende (ACT. Petr. 15, p. 62.1). Mirande is found in Lucr. 4.419 and 4.462.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
intolerandus ‘unbearable’ from the gerundive tolerandus; such words lack the verbal
properties mentioned above. The normal negation word with gerundives is non.
There is much discussion about the relationship between the gerund and the
gerundive and about which of the two was the oldest form (see also § 5.42). They
are both found from the earliest preserved texts onwards, in varying frequency.38
In speculations about their prehistorical development some scholars take the ger-
undive as the original form,39 others the gerund. On the whole, the latter position
seems stronger.40
3.21 Supines
Latin has two verb forms, the so-called first and second supines, that from a historical
point of view must be regarded as nominals, but have no nominal properties of the
type mentioned: they cannnot be modified by determiners or adjectives. They resem-
ble accusative and ablative forms of abstract verbal nouns of the fourth declension,
ending in -um and -u respectively. Their stem is the same as that of the perfect
participle: deletum ‘with the purpose of destroying’ and deletu ‘in respect to destroy-
ing’. It is sometimes not easy to decide whether a form in -u is a second supine or the
ablative of a verbal noun that functions as a respect expression (see § 10.91).41
The syntactic and semantic properties of the two supines are quite diverse (for
details, see § 16.111). For the purpose of this chapter two typical examples will suffice.
Ex. (a) contains a verb of movement with a first supine irrisum ‘to ridicule’, which
together with its second argument dominum functions as a purpose clause; the first
supine has clear verbal properties. The addition of satellites is also possible. The
second supine is illustrated by (b). Here factu facilem is a modifier of the head noun
rem (note the coordination with other modifiers). Facilis is an evaluative adjective; the
verb facio has a very general meaning; semantically, rem is also the second argument
of factu. The addition of further arguments and satellites is very rare.
(a) . . . nunc venis etiam ultro irrisum dominum.
(‘ . . . now you’re even coming of your own accord to ridicule your master.’ Pl. Am.
587)
(b) Bonam atque iustam rem oppido imperas et factu facilem.
(‘That’s an absolutely fine and reasonable suggestion—and easy enough to carry out.’
Ter. Hau. 704)
Table 3.10 contains a future passive infinitive delet-um iri ‘to be going to be des-
troyed’, which is the passive counterpart of the first supine deletum with the infinitive
ire ‘to go’. The history of the passive infinitive is uncertain.42
38 39
See Maltby (2005). Notably Risch (1984).
40 41
See Kircher-Durand (2008). See Kroon (1989a) and Lambertz (1996).
42
See Pinkster (1985b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Another word class that is relatively easy to identify are prepositions: they always are
part of a prepositional phrase, usually with a noun or noun phrase in a particular case.
However, the boundary between adverbs and other particles is not very clear. In
particular, connectors and interactional particles share certain features with adverbs.
In § 3.27 a number of remaining particles are discussed that are not easily assigned to
a specific category.
3.23 Adverbs
ADVERBS constitute a very heterogeneous word class, both in respect to their forms and
in respect to their meanings and syntactic functions. As far as meaning and syntax are
concerned, the ancient term ‘adverb’ relates to its function as modifier of a verb,
illustrated by (a), which is comparable to the way an adjective can function in relation
to a noun, as in (b).
(a) Nunc, spectatores, valete et nobis clare plaudite.
(‘Now, spectators, farewell and give us your loud applause.’ Pl. Men. 1162)
(b) Verum si voletis plausum fabulae huic clarum dare . . .
(‘But if you want to give this play your loud applause . . . ’ Pl. Rud. 1421)
For a manner adverb like clare the ancient terminology seems adequate, but for
many other adverbs it is not. Time adverbs like numquam, for instance, situate the
whole state of affairs, and not only the verb, in time, as in (c). Fortasse ‘perhaps’ in
43
For the various uses of the term, see Pinkster (1972: 135–6).
44
For this criterion, see Pinkster (1972: 136–7); Schrickx (2011: 9–24, 151).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(d) denotes the speaker’s judgement on the content of the sentence; in the termin-
ology of this Syntax it is a disjunct. For further details, see the various sections of
Chapter 10.
(c) Sed tamen me / numquam hodie induces . . .
(‘Still, you’ll never get me today to . . . ’ Pl. As. 493–4)
(d) Fortasse haec tu nunc mihi non credis quae loquor.
(‘Perhaps you don’t believe what I’m saying now.’ Pl. Ps. 888)
Whereas the functions illustrated above may be regarded as the core functions of
adverbs, adverbs may fulfil other functions as well. Some can act as a cohesive device
between sentences, such as tamen in (e) (for details, see Chapter 24); others can be
used (alongside other uses) as subject or object complement in their clause, like sic ‘so’
in (f) (for details, see § 9.36).
(e) Cupis me esse nequam: tamen ero frugi bonae.
(‘You wish me to be useless; still, I’ll be useful.’ Pl. Ps. 468)
(f) Ne dixis istuc. # Ne sic fueris.
(‘Don’t say that. # Don’t be like that.’ Pl. As. 839)
Adverbs that denote degree can be used as modifiers of gradable adjectives and
adverbs. One example will suffice. In (g), valde ‘extremely’ modifies the manner
adverb bene ‘well’.
(g) Rem te valde bene gessisse rumor erat.
(‘There is a report that you have had a highly successful campaign.’ Cic. Fam. 1.8.7)
From the point of view of morphology, adverbs can be divided into productive and
non-productive groups. Typical examples of productive adverbs are those derived
with the suffix -e from descriptive adjectives of the first and second declension, such
as clar-e in (a), and those derived with the suffix -(i)ter mainly from adjectives of
the third declension, such as felic-iter ‘luckily’ from felix. In a similar way adverbs
can be formed from participles, for instance aman-ter ‘lovingly’ from amans.
Productive adverbs typically denote manner. They are mainly formed from adjectives
towards the end of scale (a) in § 3.7. If they have a gradable meaning, comparative
and superlative forms are possible, such as clarius ‘more clearly’ and clarissime ‘most
clearly’. Adverbs derived from two- and three-place adjectives may require an argument.
An example is dignius ‘more worthily’ in (h), where it governs the ablative cruce.
Another productive category of words that behave like adverbs are those derived from
adjectives of amount (see § 10.63). The actual forms are the same as the singular
accusative neuter forms and indeed are often called ‘adverbial accusatives’ or ‘internal
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
objects’.45 However, an example like (i) shows that the latter term is incorrect: the
verb already has its second argument position filled by nullum adulescentem. These
words can also be used as degree expressions with adjectives and adverbs, as in (j). For
the use of similar forms of descriptive adjectives, see § 10.45.
(i) Nullum adulescentem plus amo.
(‘There’s no young man I love more.’ Pl. Mer. 540)
(j) . . . non multum fuit molesta servitus . . .
(‘ . . . slavery wasn’t very troublesome for me . . . ’ Pl. Capt. 272)
Apart from these productive formations, there are numerous unproductive, yet
analysable adverbs, such as crebro ‘frequently’, originally an ablative form from the
adjective creber; acerva-tim ‘in heaps’ from the verb acervo; dextra ‘on the right-hand
side’, short for dextra manu, a noun phrase in the singular ablative ‘right hand’. Next
there are many pronominal adverbs such as ubi ‘where’ and tum ‘then’, for which the
etymology is more or less clear.46 Finally, there are many adverbs of which the
etymology is less certain, such as clam ‘secretly’, palam ‘openly’, and saepe ‘often’.
Some words function both as adverb and as preposition. An example is clam
‘secretly’, ‘without the knowledge of ’, which is used in both ways from Early Latin
onwards. The history of these homonymous pairs is diverse. Palam is used as an
adverb from Early Latin onwards; as a preposition ‘openly in the presence of ’ it is first
attested in Horace and remains rare. Supra ‘on top (of)’ is attested from Early Latin
onwards in both uses. Extra ‘outside’ is attested as a preposition from Early Latin
onwards; as an adverb from Cicero onwards. Intra ‘inside’ as a preposition is common
from Early Latin onwards; as an adverb it is rare (it is first attested in de Bello
Hispaniensi). For a survey of adverbs/prepositions, see § 12.24. On the relationship
between adverbs and prepositions, see also § 12.25.
3.24 Prepositions
In this Syntax the term PREPOSITION is used in its traditional way to cover prepositions
proper that precede the noun or noun phrase, such as a/ab ‘from’,47 the preposition
cum that is usually preposed, but follows in certain well-defined combinations (cum
patre ‘with the father’, but mecum ‘with me’), and the few prepositions that follow
their noun or noun phrase, such as tenus ‘as far as’, ‘up to’. An alternative term
covering all these groups would be ADPOSITION. For details concerning the placement
of prepositions, see Chapter 23.
The Latin grammarians state that prepositions had no accent of their own but
functioned as proclitics and formed a prosodic unit with the word they precede.
45 46
So, for instance, K.-St.: I.280–2. See K.-H.: 1003–27.
47
For very few exceptions, see TLL s.v. 3.51ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Further evidence comes from punctuation, which seems to show that it was not
necessary to separate a preposition from the following word. There are a few instances
of punctuation after a group consisting of a preposition and one following word,
as in ex. (a).48
(a) . . . DE HAC · RE· . . .
(‘about this matter’ Tab. Vindol. II.211 (c. AD 92–7))
For the formation of prepositions, see § 12.24. For the relationship between preposi-
tions and adverbs, see § 12.25.
Latin has three types of clause and sentence linking devices: connectors, coordinators,
and subordinators. The traditional name for them is ‘conjunctions’, usually divided
into ‘coordinating’ and ‘subordinating’ conjunctions.
(i) ‘Connectors’ are words that indicate the semantic relation between successive
sentences; examples are at ‘but’, igitur ‘therefore’, ‘so’, nam ‘for’, and quippe
‘indeed’. For details concerning sentence connection, see Chapter 24.
(ii) ‘Coordinators’ are words or expressions that link clauses, phrases, or words;
examples of ‘conjunctive’ coordinators are atque (ac) ‘and’, et ‘and’, the clitic
-que ‘and’; of ‘adversative’, sed ‘but’; of ‘disjunctive’, aut ‘or’; notably et and
sed can also be used to connect sentences. For details concerning coordin-
ation, see Chapter 19.
(iii) ‘Subordinators’ are words that serve to make one clause a part of another;
examples of ‘temporal’ subordinators are cum ‘when’, postquam ‘after’; of
‘causal’, quia ‘because’; of ‘conditional’, si ‘if ’. For details concerning sub-
ordinators, see §§ 14.21–14.23.
The morphology of these words is obscure, with the exception of the subordinators,
some of which are analysable (postquam for example) or are etymologically clear
(cum < quom, the stem of relative pronouns, for example).
This term covers a group of particles traditionally included among the ‘(coordinating)
conjunctions’, as devices that link sentences in the way connectors do. Examples are
enim ‘y’know’, ergo ‘then’, and nempe ‘of course’. The difference between these words
48
For the lack of accent, see Allen (1973: 24–5) and Probert (2002); for punctuation, see Wingo (1972:
16) and Adams (1996).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
and connectors and adverbs is that they invite the addressee to subscribe to the point
of view of the speaker.49
There is a residual class of particles which are difficult to subclassify. It contains words
like adeo ‘to such an extent’ and etiam ‘even’, solum ‘only’, which are used in
combination with other words or phrases and have no independent position in
their clause. They are called ‘subjuncts’ in Quirk et al. (1985).50
3.28 Clitics
The term ‘clitic’ is used in various ways. In this Syntax it is used in a narrow sense for
forms that cannot be used independently and have to be attached to a host word,
with which they form one prosodic unit (sometimes called ‘bound clitics’).51 Latin has
a limited number of clitics (or, as some would prefer, ‘enclitics’). Apart from -met and
-pte, which are attached to certain personal and possessive pronouns,52 and -nam,
which is attached to question words,53 Latin has -ne, -que, and -ve, which can in
principle be attached to any word. These three make a distinct semantic contribution
to the content of the clause. The fact that the combination of the host word and the
clitic behaves as one prosodic unit appears from the position of a host–clitic com-
bination in verse. In (a) and (b), the accent (and the ictus) is on the penultimate
syllable of the unit, instead of on grávi, and cálathis.54 For the hosts to which these
clitics are attached, see Chapter 23.
(a) haec est / vita solutorum misera ambitione gravíque.
(‘Such is the life of men set free from the burden of unhappy ambition.’ Hor. S.
1.6.128–9)
(b) . . . bellatrix, non illa colo calathísve Minervae / femineas adsueta manus . . .
(‘ . . . a warrior maid, never having trained her woman’s hands to Minerva’s distaff or
basket of wool . . . ’ Verg. A. 7.805–6)
Exx. (c)–(e) illustrate the relative mobility of -que in a prepositional phrase which
-que coordinates with the preceding context. It may be attached to the preposition at
49
These particles are called ‘conversation management particles’ by Kroon (2011).
50
Another term is ‘focus particles’ (Kroon 2011).
51
For a discussion of clitics in a broader sense, see Janson (1979: 90ff.) and Spevak (2006). For the shift
of the accent to the penultimate syllable of the new prosodic unit (even if the vowel of that syllable is
short), see Probert (2002). (NB: one cannot be entirely sure about this for Cicero’s time—ibidem.)
52
The material can be found in Neue-W.: II.361ff., 373.
53 54
For clitic -nam see § 6.21. See Leumann (1977: 238–40).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
the beginning, to the noun at the end if it is not modified in some way, or to the
modifier of the noun in the middle. However, it cannot precede the phrase.
(c) . . . Cumas contulisse se dicitur inque ea urbe senio et aegritudine esse
confectus.
(‘ . . . he withdrew, it is said, to Cumae, and in that city was brought to the grave by old
age and distress of mind.’ Cic. Tusc. 3.27)
(d) . . . et apud populum . . . , cum pro Milone diceret, clamore convicioque
iact<at>us est in senatuque a Catone aspere et acerbe inim<icor>um
magno silentio est accusatus . . .
(‘ . . . in the public meeting he was shaken by noise and abuse as he spoke in defence of
Milo, and in the senate he was accused harshly and vehemently by Cato, amidst the
great silence of his enemies . . . ’ Cic. Fam. 1.5b.1)
(e) . . . fruitur praesentibus ab iisque vitiis quae paulo ante collegi abest
plurimum . . .
(‘ . . . he finds his true enjoyment in the present. Also he is entirely free from the vices
that I instanced a few moments ago . . . ’ Cic. Fin. 1.62)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
CHAPTER 4
Verb frames
This chapter deals with verb frames, described in § 2.8 as the constellation or
constellations of arguments related to a verb in accordance with its meaning or
meanings. The chapter begins with some general observations on the way the valency
of a verb can be established and the difficulties involved.1 After this the various frames
will be presented. The first type of frame that will be discussed is that of one-place
(monovalent) verbs, that is, verbs that require only one argument. Next come two-
place (bivalent) and three-place (trivalent) verbs. Finally, attention will be paid to
verbs that do not require an argument to complete their meaning (such as pluit ‘it is
raining’) and are for that reason called zerovalent. Separate sections are devoted to the
copula sum ‘to be’ and other copular verbs as well as to auxiliary (modal and phasal)
verbs. The concept of valency is also relevant to adjectives and nouns. Adjectives are
discussed in §§ 4.99–4.104. Nouns are discussed in the chapter on the noun phrase
(see § 11.70).
Verbs not only differ in the number of arguments they require, but also in the type
of arguments that may fill the position or positions in the frame. In this chapter,
nominal arguments will be the focus of attention. The use of the noun as the second
argument of the verb dico ‘to tell’ in ex. (a) provides an illustration. A survey of the
types of arguments that can be used in the various positions in a verb frame is given in
Chapter 9. Clausal arguments, such as the accusative and infinitive clause functioning
as second argument of dico in (b), are discussed in Chapters 15 and 17.
(a) O Saturio, opportune advenisti mihi. / # Mendacium edepol dicis, atque hau
te decet.
(‘O Saturio, you’ve come to me in the nick of time. # You’re telling a lie and it isn’t
right of you.’ Pl. Per. 101–2)
(b) Nempe uxor ruri est tua, quam dudum dixeras / te odisse [aeque] atque
anguis.
(‘Surely your wife is in the country, who you said a while ago you hate as much as you
hate snakes.’ Pl. Mer. 760–1)
1
For a typological treatment of Latin valency, see Lehmann (2002).
Verb frames
Within the sections on the various (one-place, two-place, three-place) verb frames a
further distinction is made according to the formal marking of the arguments, by
cases and/or by prepositions. Since the formal marking of the arguments is sometimes
related to the meaning of the verb, further subclassifications according to the meaning
of the verbs are made when this may help to understand why a particular marking is
used. This means that much information on ‘the use of case X or preposition Y with
verb Z’ that one finds in traditional grammars is found in this chapter and not in a
chapter on the uses of the individual cases and prepositions.
Obviously, only a small selection of verbs can be discussed and none of them
completely. For individual verbs the professional Latinist should always consult the
lemmata of the Oxford Latin Dictionary or the Thesaurus Linguae Latinae.
The first practical problem involved in defining the valency of a verb and establishing the
verb frame(s) in which it is used stems from the fact that we are dealing with a corpus in
which clauses normally do not occur in isolation. In Latin, more than in English,
arguments can be left unexpressed when they have been properly introduced into the
discourse. This problem is especially relevant for second and third arguments and
receives some attention in Chapter 9 on arguments. For establishing the valency of a
verb such contextually given but unexpressed constituents should be taken into account.
Once this problem is solved, the next task is to make a distinction between arguments
and satellites (see § 2.2). Arguments are those constituents which are required by the
meaning of the verb, in the sense that the clause would become ungrammatical (or the
verb would turn out to have a different meaning) if that particular constituent were
eliminated and could not be inferred from the context or situation. Usually it is not very
difficult to distinguish arguments from adjuncts, that is, satellites that specify in some
way the state of affairs expressed by the nucleus (the verb and its argument(s)).
However, not all adjuncts can be combined randomly with any nucleus. Instrument
and manner adjuncts, for example, are almost entirely restricted in their use to
controllable states of affairs (see § 10.42), and there are also restrictions on the use of
adjuncts of duration and of time within which (see § 10.31 and § 10.36). These types of
adjuncts are ‘sensitive’ to the meaning of the nucleus and, in that respect, resemble
arguments.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Prepositional phrases with de ‘about’ are another example. They are often used as
adjuncts in combination with states of affairs that refer to the exchange of informa-
tion, as in (a). In this example dixit and dixerit govern a second argument in the
accusative (quod and aliquid, respectively). De ambabus and de me are adjuncts.
However, when there is no such argument, as with rogare in (b), the de phrase might
be regarded as an argument instead.
(a) Nimiae voluptati est quod in extis nostris portentum est, soror, / quod[que]
haruspex de ambabus dixit. # Velim de me aliquid dixerit.
(‘What was shown in the entrails of our sacrifice, my sister, gives me enormous joy,
and also what the soothsayer said about both of us. # I wish he’d said something
about me.’ Pl. Poen. 1205–6)
(b) Iam de istoc rogare omitte. Non vides nolle eloqui, / ne suarum se mise-
riarum in memoriam inducas?
(‘Stop asking about this now. Can’t you see she doesn’t want to tell, so that you won’t
remind her of her misfortunes?’ Pl. Per. 642—NB: for the text, see Woytek ad loc.)
2
For the verbs which are used with ab in this way, see TLL s.v. ab 18.71ff. For the description of this
use of peto, see TLL s.v. 1961.71ff.
3
This is my paraphrase of the description of the TLL s.v. peto 1961.71ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
Whereas peto in the preceding section can still be said to have its own meaning when
combined with the nouns mentioned, this is not the case, or is less clear-cut, with a
number of two- and three-place verbs of a rather general meaning that are used with
abstract (often verbal) nouns as their second arguments in more or less fixed (some-
times called ‘analytic’ or ‘periphrastic’) combinations. In these combinations, further
arguments do not depend on the verb but rather on the noun. These verbs are called
4
SUPPORT VERBS in this Syntax. Often the meaning of the complex expression comes close
to that of a simple verb that is semantically and formally related to the abstract noun.
This is illustrated in (a)–(d) for combinations of verbs with an object constituent.5
(a) Nunc mihi certum est alio pacto Pseudolo insidias dare / quam in aliis
comoediis fit, ubi cum stimulis aut flagris / insidiantur.
(‘Now I’m resolved to ambush Pseudolus in a different way from how it happens in
other comedies, where they lie in ambush with cattle-prods or whips.’ Pl. Ps. 1239–41)
(b) . . . sic onere adsueto vacuus dat in aere saltus . . .
(‘So, empty without its accustomed burden, it gives leaps into the air . . . ’ Ov. Met. 2.165)6
(c) . . . ad tibicinis modos saltantes, haud indecoros motus more Tusco dabant.
(‘ . . . while dancing to the strains of the flautist they performed not ungraceful
evolutions in the Tuscan fashion.’ Liv. 7.2.4)
(d) Tardius messim primam eius facere oportebit . . .
(‘It will be best to make the first cutting rather late . . . ’ Col. 2.10.28)
In (a), the verb insidior could have been used instead, as it is two lines below. Do may
have been used to make the involvement of the agent more explicit, or for the sake of
4
The term ‘support verbs’ is an adaptation from French ‘verbes de support’. In German the term
‘Funktionsverb’ has become common. Quirk et al. (1985: 750–2) use the term ‘eventive object’. For
terminology and a survey of noun/verb combinations, see Rosén (1981: 130–59). For support verbs, see
also Flobert (1996) and Hoffmann (1996, forthc.). For tests for distinguishing support verb constructions,
see Langer (2004), Ros (2005: 417–18), and Baños (2012). For the use of another type of verbs that ‘are
notoriously poor in semantic content’ in Virgil, see Görler (1999).
5
For more examples, see TLL s.v. do 1686.33ff.
6
For Ovid’s practice, see Bömer ad Ov. Met. 2.165.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
variation. In (b), saltat would have signified continuous jumping instead of incidental
jumps. In (c), movebant alone might indicate involuntary movement; this would
exclude the addition of a manner adjunct (indecore), whereas the support verb
construction has the advantage of allowing the adjective indecoros. Likewise, in (d)
the combination messim facere instead of metere allows the addition of primam.
In (b) and (c), the two-place expressions can be regarded as variations on a one-
place frame, a form of ‘externalizing’ shown below in § 4.10 for cognate objects. In
fact, these instances are often regarded as cognate objects. The most frequent support
verbs are affero ‘to bring about’, ago ‘to make’, ‘to do’, capio ‘to take’, do ‘to give’, facio
‘to make’, fero ‘to carry’, gero ‘to bear’, and habeo ‘to have’, all of which govern an
accusative object (and are thus passivizable).7 However, other verbs are used in this
way as well, such as utor ‘to use’ + ablative and various verb + prepositional phrase
combinations: sum + in + ablative and versor + in + ablative ‘to be involved in’; and
venio + in + accusative ‘to come to (mind)’, ‘to be included in (a category)’, etc.
An interesting instance of variation in the translation of a Greek support verb can be
seen in Vulg. I Macc. 10.15, where the Greek text f Œı R å is translated
labores quos laboraverunt. The Vetus Latina has either habuerunt or passi sunt.
Supplement (in alphabetical order by verb):
Aperite hasce ambas fores / prius quam pultando assulatim foribus exitium affero.
(Pl. Capt. 831–2); . . . mentio a te facta pacis suspicionem multis attulit immutatae
voluntatis. (Cic. Phil. 12.18); Ille contra haec omnia. / Ruri agere vitam, semper parce
ac duriter / se habere. (Ter. Ad. 44–6); . . . civium curam ago? (Liv. 6.15.11); . . . aut ne
laborem capias quom illo uti voles? (Pl. Mer. 146); Eorum una pars, quam Gallos
obtinere dictum est, initium capit a flumine Rhodano . . . (Caes. Gal. 1.1.5); Cur
autem (sc. ista auguratio) de passerculis coniecturam facit . . . ? (Cic. Div. 2.65);
Antonius solus contumelia naturae vilitatem auro fecit. (Plin. Nat. 33.50); Si . . . res
publica aliqua faciat venditionem . . . (Ulp. dig. 21.1.1.4); Simul Alcumenae, quam vir
insontem probri / Amphitruo accusat, veni ut auxilium feram. (Pl. Am. 869–70);
Quid mi opu’st decurso aetatis spatio cum <m>eis / gerere bellum . . . (Pl. St.
81–2); . . . ego referam sermones eos quos de agri cultura habuimus nuper . . . (Var.
R. 1.1.7); Proximis diebus habetur extra urbem senatus. (Caes. Civ. 1.6.1); . . . quae
difficilem curationem habeant . . . (Cels. 5.26.1.C); Interiores simplicius et antiquius
permutatione mercium utuntur. (Tac. Germ. 5.3);
Quo quid absurdius quam aut res sordidas atque deformis deorum honore adficere
aut homines iam morte deletos reponere in deos, quorum omnis cultus esset futurus
in luctu. (Cic. N.D. 1.38); Quod cum isti renuntiaretur de basi ac litteris, existimavit
homines in oblivionem totius negotii esse venturos, si etiam basim tamquam
indicem sui sceleris sustulisset. (Cic. Ver. 4.79); (sc. philosophos) quod in veri
investigatione versentur . . . propterea iustos esse. (Cic. Off. 1.28);
7
For ago and facio, see López Moreda (1987). For do, see Martín Rodríguez (1996a). For affero, see
Hoffmann (1996: 204–10, 2005). For accipio, see Brunet (2008). For capio, see García González (2014).
For the use of support verbs with sermonem and verba, see Roesch (2001). For bellum gero, see Baños
(2013).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
One and the same verb can be found with a variety of abstract nouns, so for example
do ‘to give’ with copiam ‘grant power/ability’, negotium ‘charge’ or ‘commission’,
pugnam ‘fight a battle’, and spatium ‘give way’ or ‘make room’. In these combinations
do does not have its literal meaning of ‘transfer’, and if a dative argument is present it
cannot easily be regarded as the ‘recipient’, Moreover, the nouns themselves lose their
independence to some extent in that they relatively often precede the verb and are
rarely modified by a relative clause or by adjectives other than those of quantity.
However, in contrast with actual idioms the resulting constructions remain fully
grammatical.
Conversely, one and the same abstract noun may be used with a number of the
verbs mentioned, so for example spem ‘hope’ with affero, capio, do, facio, habeo, and
pono alongside spero ‘to hope’. In comparison with the verb spero, spem capio can be
regarded as an ingressive and spem do as a causative correlate of the simple verb spero.
Apart from semantic factors that may favour the use of a complex expression there
may also be a syntactic factor involved, as illustrated in (e), where an ablative absolute
clause would not have been possible with the simplex verb meto.
(e) Messi facta spicilegium venire oportet . . .
(‘When the harvest is over the gleaning should be let . . . ’ Var. R. 1.53.1)
This type of expression is frequent in didactic texts, but not restricted to these. It has
been maintained that it is typical of colloquial Latin, but it is quite common in
Cicero’s philosophical works as well.8
A verb and its object, when used together often, may develop into an idiom and
create the possibility, and often the need, of adding another participant. Here are a
few examples of two types of such an expansion of complex expressions. Firstly, the
whole constellation may look like a three-place frame with a third argument that
formally seems to be a recipient in the dative, even though semantically it functions as
an experiencer rather than a recipient, as in (f). Along the same lines such complex
expressions often occur with constituents which formally look like satellites, but seem
to be obligatory in the combination. An example is (g). (For examples of object
incorporation see § 4.79.)
(f) STOMACHUM·MIHI·NULLU<M> UNCQUAM·FECIT·NISI·QUOD M<ORTUA EST>
(‘She was never an annoyance to me, except in dying.’ CIL X.8192 (Puteoli))
(g) Primo haec pudice vitam, parce ac duriter / agebat . . .
(‘At first she lived a virtuous life, sparing and thrifty . . . ’ Ter. An. 74–5)
8
See Hoffmann (1996: 203). Schøsler (2003: 409–11) has a brief comparison of Cicero’s de Inventione
and an Early French and a Modern French translation. She shows that the Latin original and the Modern
French translation have far fewer support verb constructions than the Early French one, and that support
verb constructions are very unstable over time. For the use of verbal nouns in combination with a support
verb instead of simple verbs, especially in didactic texts, see Langslow (2000: 408–16), with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
. . . maximas tibi omnes gratias agimus, C. Caesar, maiores etiam habemus. (Cic.
Marc. 33); Ego me iniuriam fecisse filiae fateor tuae . . . (Pl. Aul. 794); . . . cui maxi-
mam fidem suarum rerum habeat (Cic. Ver. 2.131);
The same phenomenon is also very common with the copula in combination with
dative nouns functioning as subject complement (the so-called predicative dative—
see § 9.34). This combination in turn governs a noun or noun phrase in the dative, as
nobis in (h). The combination auxilio fuit behaves like the verb auxilior ‘to give help’,
which also governs a dative. Ex. (i) is comparable, with auxilio functioning as a
secondary predicate. The same analysis is possible in (j), but with mitto a dative may
also be used for the recipient when there is no predicative dative. Also, in cases like
(k), one may consider the combination of sum and the noun curatio a support verb
construction (see § 11.71).
(h) Mi pater, tua pietas plane nobis auxilio fuit.
(‘My father, your piety has clearly helped us.’ Pl. Poen. 1277)
(i) . . . si Romanus auxilio suis venisset . . .
(‘ . . . if the Roman had come to the help of his countrymen . . . ’ Sal. Jug. 81.3)
(j) Qua re nuntiata Caesar omnem ex castris equitatum suis auxilio misit.
(‘Upon report of this Caesar sent the whole of the cavalry from the camp to assist his
men.’ Caes. Gal. 4.37.3)
(k) Quid tibi hanc curatio est rem, verbero, aut muttitio?
(‘Why are you interfering in this matter, you whipping stock, or why are you
muttering?’ Pl. Am. 519)
Quite a few verbs appear to be used in more than one argument configuration, not only
when a different meaning of the verb corresponds to each configuration (see § 2.8 on
dico), but also when the verb’s meaning appears to be the same. On the one hand, one-
place verbs can under certain conditions be expanded with a second argument, as is
illustrated by the verb somnio ‘to dream’. Conversely, two-place verbs can be used
without their second argument; this is illustrated by the verb bibo ‘to drink’.9
Some verbs that at first sight seem to be one-place verbs are sometimes combined
with a second constituent, which behaves like a second argument in many respects,
while the meaning of the verb seems to be the same in both situations. This is, for
example, the case with the verb somnio ‘to dream’, which can also be used in
9
For a very complete and careful study of two-place verbs in Cicero, see Lebreton (1901a: ch. IV).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
combination with the accusative noun somnium in the meaning ‘to dream a dream’. In
this construction somnium resembles a second argument functioning as the object, like
vinum in bibo vinum ‘to drink wine’. The question therefore arises whether we should
assign two valencies and two verb frames to somnio or only one. In this Syntax one
frame will be assigned if a verb is used in multiple combinations, but with the same
meaning.10 Determining whether the meaning is the same to some extent depends on
intuition, but corroboration can be found by studying the constituents with which the
verb is combined and by comparing verbs with their (near-)synonyms and antonyms.
Furthermore, statistical data (which combinations are the most frequent in our cor-
pus?) play an important role. In order to explain the other configurations in which a
verb is used, an attempt is made to describe under which conditions they are used. In
this particular case, somnio is considered a one-place verb; the explanation for the use of
somnio somnium is given in § 4.10 in the paragraph on the so-called cognate object.
Other types of EXPANSION of one-place verbs are discussed there as well.
The reverse situation (REDUCTION of the number of arguments) can be illustrated by
the verb bibo, already mentioned in the preceding paragraph. This verb is typically
used with an animate subject and a liquid as its object (vinum, for example), so it is
considered a two-place verb. In the OLD this is the first type of use recognized; it is
labelled ‘transitive’. Examples are (a) and (b).
(a) Nauteam / bibere malim . . .
(‘I’d rather drink puke . . . ’ Pl. As. 894–5)
(b) Deinde aquam tepidam, ut supra scriptum est, bibere.
(‘ . . . after that drink tepid water as described above.’ Cels. 1.3.22)
However, not infrequently bibo is used without a second argument, as in (c). The
relationship between the first argument and the verb is the same in both frames and
the meaning of bibo is also the same.
(c) (sc. oves) . . . eo sapore cupidinem bibendi pascendique concipiunt.
(‘ . . . the taste of it makes them conceive a desire to drink and eat.’ Col. 7.3.21)
Grammars and dictionaries often call this the ABSOLUTE use of the verb: it is the act of
drinking as such, without further specification of the entity drunk. The OLD calls this
use ‘intransitive’. Preferred contexts for the absolute use of two-place verbs are
definitions, proverbs, and enumerations, and infinitival and gerundial forms are
relatively frequent.11 Two more examples are (d) and (e).
10
For a different approach, in which more variants of valency are accepted for one and the same verb
although there is no difference in meaning involved, see Happ (1976: 445–9).
11
For contexts in which absolute use of two-place verbs is common, see Dressler (1970: 32), Happ
(1976: 239–61, 438–41) (who uses the term ‘Ellipse’), and Sznajder (1998). For a detailed discussion of the
absolute use of two-place amo in Ovid and the conditions for its correct interpretation, see Hoffmann
(1997a: 57–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) C. deinde Piso statarius et sermonis plenus orator, minime ille quidem tardus
in excogitando, verum tamen voltu et simulatione multo etiam acutior quam
erat videbatur.
(‘Then there was Gaius Piso, an orator of the stationary or quiet type; his manner of
speaking was wholly conversational; by no means slow in invention, yet by coun-
tenance and expression he gave the appearance of greater acumen than he possessed.’
Cic. Brut. 239)
(e) Nam et in grammatice et in arte medendi aut gubernandi videmus bona
humillimis quibusque contingere.
(‘For both in the scholar’s art and in the art of healing or of navigating, we notice that
goods fall to the lot of the very lowest sort of men.’ Sen. Ep. 87.15)
Supplement:
Senatorum enim urna copiose absolvit, equitum adaequavit, tribuni aerarii con-
demnarunt. (Cic. Q. fr. 2.5.4); Nemo apud eos (sc. Hunnos) arat . . . (Amm. 31.2.10);
FACITIS·VOBIS·SUAVITER·EGO CANTO. (CIL IV.3442a (Pompeii)); . . . illum identidem
monere ut caveret . . . (Cic. S. Rosc. 110); St! / Tace atque abi. Nec paro neque
hodie coquetur . . . (Pl. Cas. 149–50); Ad quem fruendum non modo non retardat,
verum etiam invitat atque adlectat senectus. (Cic. Sen. 57);
Bibo is also found with other entities as its arguments.12 Instead of an animate first
argument we find, for example, metreta ‘jar’ and hortus ‘garden’. In combination with
a liquid as its object (e.g. oleum ‘oil’) bibo means ‘absorb’. In this meaning it can be
used absolutely as well. Especially in poetry, the range of second arguments can
extend from liquids to other concrete entities, so that bibo comes to mean something
like ‘swallow’. Finally, poets can use an abstract entity for the second argument, as in
(f) and (g). Note that in (f) the interpretation is facilitated by the use of aure. This use
is manifestly two-place. There is no need to set up separate frames for these cases of
bibo. The specific interpretations are entirely brought about by the combination of
these ‘abnormal’ entities with bibo.
(f) . . . sed magis / pugnas et exactos tyrannos / densum umeris bibit aure
volgus.
(‘ . . . but the crowd, packed shoulder to shoulder, drink in more eagerly with their
ears the tales of battles and banished tyrants.’ Hor. Carm. 2.13.30–2)
(g) Nec non et vario noctem sermone trahebat / infelix Dido longumque bibebat
amorem / . . .
(‘No less did unhappy Dido prolong the night with varied talk and drink deep
draughts of love . . . ’ Verg. A. 1.748–9)
Other verbs that denote ‘consuming liquid’ are sorbeo ‘to suck up’ and haurio ‘to
swallow’. With these two verbs, which denote a special manner of drinking, the
12
See OLD s.v. bibo §§ 7–10.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
absolute use is uncommon (see note below). Bibo, by contrast, is the neutral form of
drinking and perhaps for this reason it is used relatively often without a further
specification of the entity drunk. The existence of other verbs with roughly the same
meaning which are clearly two-place verbs may be taken as support for considering
bibo a two-place verb as well.
An instance of an absolute use of sorbeo is Pl. Mos. 791: Simul flare sorbereque hau
factu facile est (‘Whistling and drinking at the same time is a difficult thing to do.’).
TLL mentions two instances of the absolute use of haurio (‘to drink’, ‘to swallow’) in
Curtius (7.5.15) and Pliny the Elder (Nat. 13.139). In both cases, the context supplies
the necessary information.
The OLD distinguishes a second type of intransitive use of bibo, as illustrated in (h)
and (i).
(h) DUM·VIXI / BIBI·LIBENTER·BIBITE·VOS / QUI·VIVITIS
(‘While I lived, I drank freely. Drink, you who live.’ CIL III.6825.2–4 (Pisidia, AD
100))
(i) Iam bis bibisse oportuit.
(‘You ought to have had two drinks already.’ Pl. Bac. 759)
In these examples the object of bibo is understood by convention: ‘to drink intoxi-
cating liquor’ (OLD) (CONVENTIONAL REDUCTION of arguments). In English, drink in
this sense is used in a number of special expressions: ‘John drinks too much’, ‘John is a
drinker’. The Latin noun bibitor ‘drinker’, ‘toper’ is used in the same specialized way.
In this case it is doubtful whether the meaning of bibo is the same as in (a)–(c), so one
might regard this use as a separate one-place frame.
This section has shown that determining the valency of a verb is not always an easy
task. In one of the earliest and most complete discussions of valency in Latin, a
distinction is introduced between ‘obligatory’ and ‘facultative’ arguments, to deal
with, for example, the de expressions in § 4.3.13 Other scholars think that there is a
gradient between highly obligatory and weak arguments.14 Finally, there are some
scholars who are highly critical of using the concept of valency at all.15 In this chapter
due attention will be paid to borderline cases.
13
See Happ (1976: 226–38).
14
For this approach, see Hopper and Thompson (1982). For an application to Latin, see Johnson
(1991) and Spevak (2010a: 115–25).
15
See, for example, Serbat’s discussion (1981: 150–4) of Happ (1976) and Jacobs (1994).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
One-place verbs
80
70
60
50
40
30
20
10
0
Plautus Cato Cicero Virgil
sum 1-pl 2-pl 3-pl
16
See also Happ’s summary of his Ciceronian sample (1976: 473–6).
17
Some scholars use the term ‘unipersonal’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
exclusively in the third person singular form and without a personal subject. The two
types of one-place verbs (‘personal’ and ‘impersonal’) are discussed separately.
Personal one-place verbs have only one argument, which fulfils the function of subject
in the clause and with which the verb agrees in person/number (and gender, where
applicable). As will be seen, many verbs that usually require only one argument can be
extended in some way with a second argument or stand in a systematic semantic and
formal relationship to other uses of the same lexeme. One-place verbs can be active or
deponent.18
Prototypical one-place verbs—that is, ones that almost always occur in one-place
constructions—are mainly of the following types:
(i) Processes in which human beings (for some verbs also non-human entities) are
involuntarily involved: the first argument fulfils the semantic function of
patient. To some of these verbs an (animate) agent or (inanimate) force may
be added.
(a) Cave ne cadas . . .
(‘Watch out so you don’t fall . . . ’ Pl. Mos. 324)
(b) Et quod in stipulatione est ‘sive quid ibi ruet’, non videri sibi ruere quod
aut vento aut omnino aliqua vi extrinsecus admota caderet, sed quod
ipsum per se concideret.
(‘In his view, the phrase in the stipulation “if something falls on it” applies not to
material that comes down because of wind or any externally applied force, but to
what collapses for intrinsic reasons.’ Alf. dig. 39.2.43.pr)
Supplement (in alphabetical order by verb):
. . . sus usque adeo pinguitudine crescere solet . . . (Cato Orig. 39); . . . ut homines,
cum viderent, aut ipsam videre se Cererem aut effigiem Cereris non humana
manu factam, sed de caelo lapsam arbitrarentur . . . (Cic. Ver. 5.187); . . . ut fame
senatores quinque morerentur. (Cic. Att. 6.1.6); M. Marcello circa mortem, cum
periit ab Hannibale, defuit in extis. (Plin. Nat. 11.189); AGER . . . QUEI ROMAE PUBLICE
2
VENIEI<T> VENIERITVE. (CIL. I .585.75 (Lex Agr., Rome, 111 BC));
(ii) States in which animate beings or inanimate entities are involved: the first
argument fulfils the semantic function of patient. Verbs in -sco, like calesco ‘to
become hot’, which denote a process in which an animate or inanimate entity is
involved as a patient, are derivationally related to some of them.
18
As is shown in § 5.5, passivized two-place verbs are often to be interpreted as agentless and so, in a
sense, are also one-place. However, because we are dealing with a productive phenomenon, they are
excluded from this section.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
One-place verbs
(iii) Psychological and physiological states which animate beings (rarely other
entities, e.g. with sudo) involuntarily undergo as experiencers.
(e) Tu amas, ego esurio et sitio.
(‘You’re in love; I am hungry and thirsty.’ Pl. Cas. 725)
(f) Pro monstro extemplo est quando qui sudat tremit.
(‘It is immediately an omen when someone who is sweating is shivering.’ Pl. As. 289)
Supplement (in alphabetical order by verb):
Familiae male ne sit, ne algeat, ne esuriat. (Cato Agr. 5.2); . . . leges . . . Lycurgi
laboribus erudiunt iuventutem venando currendo, esuriendo sitiendo, algendo,
aestuando. (Cic. Tusc. 2.34); . . . qui valetudinis vitio furerent . . . (Cic. Div.
1.81); . . . sanu’ sim anne insaniam! (Ter. Eu. 556); Si quando incidit pecus in
spem nascentis, hoc deprehenditur signo: ove, cum comederit, dormiente pro-
tinus, capra sternuente crebrius. (Plin. Nat. 19.40); . . . stomachor vero cum
aliorum non me digna in me conferuntur. (Cic. Planc. 35);
(v) Movements (moving in a specific way) in which animate beings are involved as
agents, inanimate entities as patients.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
(vi) Activities in which two or more animate beings are voluntarily involved as
agents and associatives (see § 4.38 and § 4.69).
(k) Tum isti Graeci palliati, capite operto qui ambulant / qui . . . / constant,
conferunt sermones inter sese . . .
(‘Then those Greeks in their cloaks, who wander around with their heads
covered, who . . . stand together, palaver among each other . . . ’ Pl. Cur. 288–90)
(l) Ii denique qui tum armati dies noctesque concursabant . . .
(‘Finally, these very men who at that time were running about armed day and
night . . . ’ Cic. S. Rosc. 81)
Supplement:
Heri aliquot adulescentuli coiimus in Piraeo / in hunc diem ut de symbolis
essemus. (Ter. Eu. 539–40);
(vii) Verbs of existence (also in use as copular verbs—for more details see § 4.91).
One-place verbs
meaning ‘to cry’, implies the production of sound and the expression of an emotion,
and this it has in common with communication and emotion verbs. We therefore find
it with an accusative and infinitive clause which is typical of communication and
cognition verbs (see § 15.97), as in (a). Similarly, we find furo ‘to rage’, hence ‘to
behave like a madman’, used as a cognition verb, again with an accusative and
infinitive, as in (b). Sternuo ‘to sneeze’ involves the production of a sound and so
Catullus uses it with approbationem ‘approval’ as its object, as in (c).
(a) Nam me discedere flevit . . .
(‘For she wept at my leaving . . . ’ Verg. Ecl. 3.78)
(b) Furebat a Racilio se contumaciter urbaneque vexatum.
(‘He was infuriated by Racilius’ insulting and witty attack.’ Cic. Q. fr. 2.1.3)
(c) Hoc ut dixit, Amor sinistra ut ante / dextra sternuit approbationem.
(‘As he said this, Love on the left, as before on the right, sneezed goodwill.’ Catul. 45.8–9)
At least some of the instances of the so-called adverbial accusative (see § 10.45) can be
explained along these lines. An example of furo is (d). Infanda may be taken as: ‘in his
rage he produced unspeakable utterances’ and not adverbially as ‘he manifested an
unspeakable rage’ or ‘he raged in an unspeakable manner’.
(d) At fessi tandem cives infanda furentem / armati circumsistunt . . .
(‘But at last, outworn, his citizens in arms surround the man hurling, in his rage,
unspeakable invective . . . ’ Verg. A. 8.489–90)
Supplement:
Stabat acerba fremens . . . (Verg. A. 12.398); . . . insanire putas sollemnia me . . . (Hor.
Ep. 1.1.101); . . . vix credenda furentem . . . (Sil. 15.452);
Furthermore, verbs may be expanded with an argument because they are in some way
associated with another straightforward (group of) two-place verb(s). For example,
since tremo ‘to tremble’ may occur as a physiological symptom of emotion, poets first
(and then Livy and others) use it with an object noun phrase in the accusative case,
like timeo ‘to fear’. Examples are (e) and (f).
(e) Quid est enim cur ego hominem aut feram, quid est cur sagittam aut
lanceam tremam?
(‘Why should I tremble at a man or wild beast, why should I tremble at an arrow or
spear?’ Sen. Nat. 6.32.2)
(f) . . . in eadem civitate in qua magistri equitum virgas ac secures dictatoris
tremere atque horrere soliti sint.
(‘ . . . in that very state in which masters of the horse had been used to tremble and
shudder at the rods and axes of the dictator.’ Liv. 22.27.3)
The examples above illustrate the expansion of the number of arguments on the basis
of specific semantic features of the verb. Another productive way of expanding the
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
number of arguments similar to those above but with the verb used in its ‘normal’
meaning is the use of the so-called COGNATE object19 in the accusative case. As the term
‘cognate’ indicates, the meaning of the noun phrase is closely related to the meaning
of the verb. It ‘externalises’ the meaning of the verb. Most of the verbs involved belong
to controlled and non-terminative states of affairs. Examples are (g) and (h).
(g) Prius quam istam pugnam pugnabo, ego etiam prius / dabo aliam pugnam
claram et commemorabilem.
(‘Before fighting this fight, I’ll also put up some other impressive and memorable
fight.’ Pl. Ps. 524–5)
(h) Hunc, oro, sine me furere ante furorem.
(‘Let me first, I beg, give vent to this madness.’ Verg. A. 12.680)
The term ‘cognate object’ is here used in a limited sense for what others sometimes call
the figura etymologica within the category of the ‘internal object’ (German: ‘Akkusativ
des Inhalts’). The ‘internal’ object is often defined in the same way as the term ‘effected’
object is used in this Syntax (see § 4.20), but in reality it covers more.20
This use of the cognate object is frequent in Plautus and in poetically inspired texts
but is certainly not restricted to these. Just as with many two-place verbs that govern
an accusative object, the cognate constituent with these one-place verbs can be the
subject in the passive. However, this is not a very frequent phenomenon. A good
example of a passive construction is (i).
(i) Haec est pugnata pugna usque a mani ad vesperum . . .
(‘This fight was fought there without interruption from morning till evening . . . ’ Pl.
Am. 253)
Although formally, and judging from its behaviour under passivization, the accusative
constituent looks like a normal object, it has distinctive semantic properties. The noun
phrase does not refer to some real, existing or future, entity, but specifies in some way or
other either the way an event takes place or the result produced by the action.
• Often there is an attributive adjective or another type of attribute.21
• The noun (phrase) has a more specific meaning than the meaning of the verb
itself (e.g. Graecanicam Pyrricam saltare ‘to dance a Greek Pyrrhic (dance)’, cf.
Apul. Met. 10.29.4). A good example of this is the noun servitutem in the
expression servitutem servire. A specific way of servire is meant, as is shown
by . . . aliud est servum esse, aliud servire . . . ‘ . . . to be a slave and to be in
servitude are different . . . ’ (Quint. Inst. 5.10.60).
19
On the cognate object, see Rosén (1981: 101–27, 1996b) and de la Villa (2007, 2014). For the type of
verbs, see Baños (2014a: 772–6).
20
For a survey of ‘figura etymologica’ combinations, see Müller (1908: 12–34).
21
For cognate accusatives without an attribute, see Landgraf (1881: 14–18).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
One-place verbs
• The cognate object contains information about the result of the action (e.g.
miserabile certamen currere ‘to run a lamentable match’ cf. Stat. Theb. 3.116).
Sometimes the use of the cognate object is triggered by coherence between the situation
described in the clause and the preceding context. In (i) above, haec pugna refers to the
previous context, summarizing it as a ‘battle’. It is therefore topical information and a
suitable candidate to be subject in its clause (hence pugno is in the passive). Another
constellation in which constituents occur in the passive is the ablative absolute, as in (j).
Depugnato by itself is impossible and needs ‘support’.
(j) Sed metuo ne sero veniam depugnato proelio.
(‘But I’m afraid that I’m coming too late, after the battle’s been fought.’ Pl. Men. 989)
Supplement:
Active: . . . dum vitam / vivas. (Pl. Per. 494–5); Pulmoneum edepol nimis velim
vomitum vomas. (Pl. Rud. 511); . . . consimilem luserat / iam olim ille ludum . . . (Ter.
Eu. 586–7); . . . cur non eosdem cursus . . . cucurrerunt? (Cic. Agr. 2.44); . . . prima leo,
postrema draco, media ipsa, Chimaera / ore foras acrem flaret de corpore flammam?
(Lucr. 5.905–6); . . . ut suum gaudium gauderemus. (Cael. Fam. 8.2.1); . . . multa basia
basiare . . . (Catul. 7.9); Lynceus ipse meus seros insanit amores. (Prop. 2.34.25);
Vides autem quam malam et noxiosam servitutem serviturus sit . . . (Sen. Dial.
7.4.4); . . . qui . . . pro servis servitutem servierunt . . . (Leg. pub. (Font. iur. p. 47) 5);
Passive: . . . illa militia militatur . . . (Pl. Per. 232); Proelium factum depugnatumque
pro castris . . . (Cato Hist. 82—NB: depugnatum is coordinated with factum); . . . pro
qua mihi sunt magna bella pugnata . . . (Catul. 37.13); . . . mea sunt populo saltata
poemata saepe . . . (Ov. Tr. 2.519);22
Whereas in the examples above the object constituents are not only semantically but
also morphologically related to the governing verbs, there are also instances where
there is only a semantic relation between the object and the governing verb. Examples
are (k) and (l).
(k) Multa ego possum docta dicta et quamvis facunde loqui . . .
(‘I could say many wise words, and that as eloquently as you like.’ Pl. Trin. 380)
(l) Pergam quo coepi hoc iter.
(‘I’ll be on my way along here.’ Ter. Hec. 194—tr. Brown)
Supplement: Soleo hercle ego garrire nugas. (Pl. Aul. 830); Sed metuo ne sero veniam
depugnato proelio. (Pl. Men. 989); . . . semper longam incomitata videtur / ire viam
. . . (Verg. A. 4.467–8); . . . populus frequens / laetum theatris ter crepuit sonum.
(Hor. Carm. 2.17.25–6);
22
The fullest survey of examples is Müller (1908). For its use in comedy, and especially its frequency in
the ‘Langvers’ in Plautus, see Haffter (1934: 10–43).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
(m) Sed memento, si quid saevibunt senes, / suppetias mihi cum sorore ferre.
(‘But remember to come to my assistance with your sister if the old men are a bit
grumpy.’ Pl. Epid. 658–9)
(n) Ain tu, laeta’st? # Non tam ipso quidem / dono quam abs te datum esse: id
vero serio triumphat.
(‘What do you say? Is she delighted? # Not so much with the gift itself but with the
fact it comes from you. That really and truly is a triumph for her.’ Ter. Eu. 392–3)
Occasionally, one-place verbs are used with a second argument in parallel contexts,
next to regular two-place verbs. A striking example is (n).
(o) Dies noctesque me ames, me desideres, me somnies, me expectes, de me
cogites, me speres, me te oblectes, mecum tota sis.
(‘Day and night love me, miss me, dream of me, look for me, think of me, hope for
me, delight in me, be entirely with me.’ Ter. Eu. 193–5)
Appendix: This means of ‘externalizing’ the meaning of the verb is also available with
two-place verbs, when they are used as one-place verbs, that is, when their normal
object is neither expressed nor can be understood from context (see § 4.21).
Another way of externalizing the meaning of the verb is the so-called cognate
manner adjunct in the ablative (see § 10.44).
With a number of one-place verbs that refer to processes involving liquid material,
an adjunct in the ablative may be added that specifies the type of liquid (commonly
labelled ‘instrumental’). This is the case with e.g. fluo ‘to flow’ and mano ‘to flow’ (if
combined with an ablative noun usually translated as a state ‘to be wet’), stillo ‘to drip’,
and sudo ‘to sweat’ (also zero-place pluit ‘it is raining’—see § 4.90). It occurs especially
in poetry, as in the Ennian example (p), but see also (q) from Cicero. (Mainly) poets
also venture a two-place use of these verbs with the liquid as an accusative object
(usually labelled ‘cognate’), as in (r) and (s), even in the passive, as in (t).
One-place verbs
The two-place frame is especially common when these verbs are used metaphorically.
This is illustrated in (d)–(g) for esurio and sitio.
(d) Pater liberos esurit . . .
(‘The father wants to eat his children.’ [Quint.] Decl. 12.27)
(e) Nil ibi quod nobis esuriatur erit.
(‘Nothing will be there that excites my hunger.’ Ov. Pont. 1.10.10)
(f) Fastidit vinum quia iam sitit iste cruorem.
(‘That man does not care for wine, since now he is thirsty for blood.’ Vers. pop. in
Suet. Tib. 59.1)
(g) Beati qui esuriunt et sitiunt iustitiam . . .
(‘Happy are those who hunger and thirst after righteousness.’ Vulg. Mat. 5.6)
23
See Baños (2007: 27–31).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
ablativus causae). Prepositional expressions (notably with de) are found with some of
them as well. The most representative verbs are given in Table 4.1. One may hesitate
whether the ablative constituent is an argument or a cause adjunct (see § 10.79)
combined with a one-place frame. The latter position is taken in this Syntax. A few
illustrations are given in (h)–(k).
Table 4.1 One-place emotion verbs that combine with an ablative constituent
Emotion angor doleo maereo delector gaudeo laetor
‘to be ‘to grieve ‘to be ‘to be ‘to be ‘to delight
distressed’ (at)’ sad’ delighted’ glad’ (in)’
24
For in + ablative expressions, see Baños (2007: 31–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
One-place verbs
(l) Nam epistula Leonidae quam ad me misisti quid habet, quaeso, in quo
magno opere laetemur?
(‘As for that letter of Leonides’ you sent me, pray, what does it contain that
we could be very delighted at?’ Cic. Att. 14.16.3)
(m) Propter virtutem enim iure laudamur et in virtute recte gloriamur.
(‘For we are praised on account of virtue justly and we rejoice in our virtue
rightly.’ Cic. N.D. 3.87—Note the parallelism with propter virtutem)
The verb laboro ‘to suffer’ is usually mentioned in the same context, and indeed
combinations with noun phrases in the ablative are common, as in (n) and (o). These
expressions are rather cause adjuncts, as with the verb aegroto ‘to be ill’. The part of
the body where the pain is felt may be specified with a prepositional phrase (in Cicero
with the preposition ex), as in (p).
(n) Vestrum nemo est quin intellegat populum Romanum . . . hoc tempore
domestica crudelitate laborare.
(‘There is no one among you who does not know that the Roman people . . .
are suffering today from cruelty towards their own citizens.’ Cic. S. Rosc. 154)
(o) Nec vero quisquam stultus non horum morborum aliquo laborat.
(‘Yet there is no foolish man but is afflicted by some one of these diseases.’
Cic. Fin. 1.59)
(p) Nam cum ex renibus laboraret, ipso in eiulatu clamitabat falsa esse illa quae
antea de dolore ipse sensisset.
(‘For upon an attack of kidney trouble, even amid his shrieks, he kept on
crying out that the opinions he had himself previously held about pain were
false.’ Cic. Tusc. 2.60)
(ii) Some verbs denoting the production of a smell, a taste, or a sound (oleo ‘to smell’,
sapio ‘to taste’, sono ‘to sound’) have both a one-place frame, with the person or object
producing the smell, etc. in the nominative, as in (q), and a two-place frame (‘to smell
of ’, etc.) with a second argument in the accusative case referring to the kind of smell,
taste, or sound produced, as in (r). There are also metaphorical instances. This two-
place frame is often dealt with in the same way as the cognate object, but one
important difference is that the constituent occurring as the second argument in
the two-place frame may also occur as the only argument in the one-place frame, as
the comparison of (q) and (r) shows. The source of the smelling may also be expressed
by a satellite in the ablative (rare), as in (s). Another difference between these two
types of frame is that, whereas verbs whose second argument is a cognate object can
be passivized, those discussed in this section cannot.
Verb frames
(iii) With verbs of motion several expression types in the accusative are used. One is a
satellite indicating the distance covered (see § 10.23), another is the cognate object
construction as in (t), a third one is a so-called ‘locative’ object, exemplified by (u).
(t) . . . passim / quisque suas avidi ad lacrimas miserabile currunt / certamen . . .
(‘ . . . everywhere, each running eager to find their own sorrow, a pitiful contest.’ Stat.
Theb. 3.115–17)
(u) . . . cernebant imperatorem longam viam sub gravium armorum onere
currentem . . .
(‘ . . . they saw the emperor running a long way under the burden of heavy arms . . . ’
Paneg. 3.6.4)
The difference between the two is that the ‘locative object’ is interchangeable with a
prepositional expression or compound (per viam currentem or percurrentem ‘along
the road’) whereas this alternative does not exist for the cognate object. Like cognate
objects, they rarely occur in passive constructions.
Supplement:
. . . itque reditque viam totiens. (Verg. A. 6.122); . . . sciat indociles currere lympha
vias. (Prop. 1.2.12); Nempe abruptis turbata procellis / nocte natat caeca serus
freta . . . (Verg. G. 3. 259–60);
25
For parallels, see Stockert ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
One-place verbs
Passive: . . . quia Oceanus navigari non potest. (Sen. Suas. 1.1); Qui sustinet hamos /
novit quae multo pisce natentur aquae. (Ov. Ars 1.48–9); . . . campus curritur . . .
mare navigatur . . . (Quint. Inst. 1.4.28); . . . via curritur . . . (August. Serm. 159.1.1);
NB: autocausative passive (for the term, see § 5.19): . . . nunc Satyrum, nunc agres-
tem Cyclopa movetur. (Hor. Ep. 2.2.125);
(iv) With some verbal lexemes a one- and two-place frame go side by side in a
particular semantic relationship. Moveo ‘to move’, for example, can be used of
animate and inanimate entities in the meaning ‘to make movements’. As such, it is
one-place and may be compared to verbs like tremo ‘to tremble’. An example is (v).
The two-place frame, in its meaning ‘to create movement in someone or something’,
behaves like quatio ‘to shake’ and is exemplified by the passives in (w) and (x).
(v) Terra dies duodequadraginta movit.
(‘The earth trembled for thirty-eight days.’ Liv. 35.40.7)
(w) . . . sedatoque eam tumultu moveri vetuisse puerum . . .
(‘ . . . she stilled the uproar and commanded that the boy should not be disturbed . . . ’
Liv. 1.39.2)
(x) . . . terram . . . quamvis levissimo dente moveri satis est.
(‘ . . . the soil is stirred sufficiently with the lightest plough-point.’ Col. 2.2.25)
In the case of two-place moveo the first argument is an agent, voluntarily involved in
an action, the second a patient undergoing that action. With one-place moveo,
however, the argument is rather a patient involved in some (involuntary) process.
This use of moveo can be described as the ‘decausative’ counterpart of two-place
moveo. More examples are given in § 5.30 (intransitivization). However, the second
arguments of the two-place frame cannot always occur as the single argument of the
one-place frame (and vice versa).
Some of the verbs involved, for example moveo, also occur in the autocausative
passive construction (see § 5.19) and in the autocausative reflexive construction, as
well as in the same meaning without the reflexive pronoun (see § 5.31).
Verb frames
As is mentioned in § 4.8, the reason for calling the verbs to be discussed in this section
‘impersonal’ is that they are usually restricted to third person singular verb forms
without a personal subject. They often have a subordinate clause (an accusative and
infinitive clause or a finite clause of some form) as their subject.
The one-place modal verb oportet ‘it is proper’ is used with either an infinitive or with a
clause as the subject. It is usually an accusative and infinitive, rarely a finite subjunctive
clause (for a Late Latin ut clause see § 15.69). Nominal subjects are rare, but see (e) and
(f). Unlike licet, oportet has no two-place frame with a dative experiencer.27
(e) . . . quae adsolent quaeque oportent [v.l. oportet] signa esse ad salutem omnia
huic esse video.
(‘ . . . all the usual symptoms which ought to exist toward recovery, I perceive in her.’
Ter. An. 481)
(f) Latitudines autem earum ita oportere fieri videntur . . .
(‘The width of the colonnades should be arranged as follows . . . ’ Vitr. 5.9.2)
26 27
See TLL s.v. liceo 1362.53ff. For nominal subjects, see TLL s.v. oportet 745.38ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
One-place verbs
28
The most complete collection of examples and a defence of ‘impersonal’ est can be found in
Svennung (1922: 78–81).
29
For a convincing alternative reading and additional examples, see Gratwick (2002). See also Adams
(2005b: 94).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
ut: Si est ut velit redducere uxorem, licet. (Ter. Hec. 501; cf.: 558, 637; Manere
adfinitatem hanc inter nos volo, / si ullo modo est ut possit. [723–4], 796; Ph.
925); . . . non erat ut fieri posset mirarier . . . (Lucr. 5.979); Namque si lex perite fuerit
scripta, erit ut sine captione uterque ab utroque liberetur. (Vitr. 1.1.10); Sin autem
(aedificia) inpedientur ab angustiis aut aliis necessitatibus, tunc erit ut ingenio et
acumine de symmetriis detractiones aut adiectiones fiant, uti non dissimiles veris
symmetriis perficiantur venustates. (Vitr. 6.3.11); See also § 15.28;
quod: Quod ultra mihi molesti sitis, non est. (Sen. Con. 10.pr.1);
The third person singular form of sum is also used in expressions of the type recte est
‘it is OK’ and sero est ‘it is late’, where no subject is present and the adverbs
characterize ‘the situation’. Examples are (b)–(d).30 A dative beneficiary can be
added, as in (b). (For est with an expression of distance, see § 4.84.)
(b) Ubi tu lepide voles esse tibi, ‘mea rosa,’ mihi dicito . . .
(‘When you want to have a lovely time, say to me, “my rose,” . . . ’ Pl. Bac. 83)
(c) Apud Herum recte erat.
(‘All was well at Herus.’ Cic. Q.fr. 3.1.1)
(d) Sero est: si sexta tibi placet, venibo.
(‘It is late. If the sixth hour suits you, I will come.’ Pompon. com. 65)
Supplement:
Belle erit . . . (Petr. 46.2 (Echion speaking)); . . . in libertate est ad patrem in patria
<domo>. Bene est . . . (Pl. Capt. 699–700); Patria est ubicumque est bene. (Pac. trag.
inc. 92); Fuit enim periucunde. (Cic. Att. 13.52.1); De Attica pergratum mihi fecisti
quod curasti ut ante scirem recte esse quam non belle fuisse. (Cic. Att. 14.16.4); Sed
tamen suaviter fuit . . . (Petr. 65.11 (Habinnas speaking));
With a dative beneficiary (in alphabetical order by adverb): Sed hoc mihi aegre
est . . . (Pl. Capt. 701); . . . mi bene est et tibi male est. (Pl. Mos. 52); . . . ne qui deterius
huic sit quam quoi pessume est. (Pl. Capt. 738); I hac mecum intro, ubi tibi sit lepide
victibus, vino atque unguentis. (Pl. Bac. 1181); Animo male est. (Pl. Cur. 312);
Numquam tam male est Siculis . . . (Cic. Ver. 4.95); . . . cum meliuscule tibi esset . . .
(Cic. Fam. 16.5.1);
Hoc de genere nihil te nunc quidem moneo (sero est enim . . . ) (Cic. Q.fr. 1.2.9);
Appendix: In Plautus and Terence impersonal est is used in combination with
a prepositional phrase with absque ‘without’ in some form of conditional clause,
as in (e).
(e) Nam apsque foret te, sat scio in alto / distraxissent . . . satellites tui me
miserum foede . . .
(‘Had it not been for you, I know well enough that your attendants would
have . . . dragged me apart in the sea in a frightening way . . . ’ Pl. Trin. 832–3)
30
On these expressions, see Ripoll (2007: 146–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Two-place verbs
In Late Latin, the third person singular active forms of habeo ‘to have’ and facio ‘to do’
are used as ‘impersonal’ one-place verbs with an accusative argument. Examples are
(b)–(e). In (b) librum elephantinum resembles the ‘sujet réel’ in French il y a un livre.
Ex. (c) shows an extent of space argument (see § 4.84); (d) an extent of time
argument.31
(b) . . . habet in bibliotheca Ulpia in armario sexto librum elephantinum, in
quo . . .
(‘ . . . in the sixth chest in Ulpian’s library, there is a book consisting of ivory tablets, in
which . . . ’ Hist. Aug. Tac. 8.1)
(c) Habebat autem de eo loco ad montem Dei forsitan quattuor milia . . .
(‘From that place to the mountain of God was perhaps four miles.’ Pereg. 1.2)
(d) Pater eius . . . ex quo hinc profectus est habet annos XIIII . . .
(‘It has been fourteen years since her father left here . . . ’ Hist. Apoll. RA 31)
(e) Numquam fecit tale frigus. Numquam fecit tales aestus.
(‘Never did such cold occur. Never did such heat occur.’ August. Serm. 25.3.3)
31
For more instances of personal and impersonal forms of habeo, see TLL s.v. habeo 2461.56ff.; for
facio, TLL s.v. facio 99.45ff. For a discussion of the prehistory of Spanish hay, see Luque (1978). Other
developments, and Greek influence on these, are discussed by García Hernández (1992). For habeo in the
Peregrinatio, see Stengaard (2008).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verb frames
in any conventional setting. In addition to what is said about bibo in § 4.5, this may be
illustrated by an example of the verb moveo ‘to move’ in its meaning ‘to move
somebody or something from its original position to another position’.32 In (a),
movi governs castra and is two-place, with a satellite indicating from where the
camp was moved. In (b), castra is not expressed. In military contexts like this,
when the commander is the subject, as Pompeius is, castra ‘camp’ is easily understood
as the entity to be moved. This use of moveo, which we might call an instance of
CONVENTIONAL REDUCTION of arguments, has to be distinguished from one-place moveo
in its meaning ‘to make movements’, as in (c) (see § 4.11), and from its one-place use
meaning ‘to depart’ (see § 5.31).
(a) Ita confirmato illo ex eo loco castra movi.
(‘Having encouraged him so far, I struck camp from there.’ Cic. Fam. 15.2.8)
(b) Haec autem scribebam prid. Non., XIIII die postquam ille Canusio moverat.
(‘ . . . I am writing this on the 6th, thirteen days after he marched from Canusium.’
Cic. Att. 9.1.1)
(c) Terra dies duodequadraginta movit.
(‘The earth trembled for thirty-eight days.’ Liv. 35.40.7)
32
Joffre (1995: 44), following Feltenius (1977), deals with such cases as instances of intransitivization
(see § 5.30).
33
See Happ (1976: 234–6) for Latin and Larjavaara (2000) for French.
34
See TLL s.v. moveo 1546.9ff. and Oakley ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The prototypical two-place frame refers to an action, in which a (typically human) agent
and a (human or non-human) patient are involved (see § 2.9). This type is very well
represented in our corpus, much of which deals with the actions of human beings.35
However, there are many other semantic frames of arguments, more easily found outside
the literary works that form the basis of older grammars. Table 4.2 and the following
quotations serve as illustrations of the many frames that are possible (for definitions of
semantic functions, see § 2.12). It is impossible to give a representative classification of all
the frames found. Attention will be given to a few more remarkable cases.
35
They are ‘high’ on the ‘transitivity scale’ of Hopper and Thompson (1982). For the semantic frames
of two-place verbs, see Dik (1997: I.120–4) and Givón (2001: I.126–36).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
However, the accusative objects of two-place verbs need not be either affected or
effected. Especially in the case of inanimate subjects and objects and of abstract
entities, such distinctions as ‘affected’ and ‘effected’ are irrelevant, as in (d) and (e).
(d) Virtus omnia in sese habet, omnia assunt / bona quem pene’st virtus.
(‘Courage has all goods within itself, all goods are with the man who has courage.’ Pl.
Am. 652–3)
(e) Stadium CXXV nostros efficit passus, hoc est pedes DCXXV.
(‘A stade is equivalent to 125 Roman paces, that is 625 feet.’ Plin. Nat. 2.85)
36
The best survey of effected objects can be found in Müller (1908: 4–12, 34–55).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Cognate objects (which are allowed with most one-place verbs—see § 4.10) also occur
with two-place verbs. This is the case in (a) and (b). Deponent verbs allow cognate
objects as well, as in (c).
(a) Quam magis in pectore meo foveo quas meus filius turbas turbet . . .
(‘The more I ponder in my heart what trouble my son’s stirring up . . . ’ Pl. Bac. 1076)
(b) Eorum licet iam metere messem maxumam.
(‘You can now harvest an enormous crop of it.’ Pl. Trin. 33)
(c) Metuo in commune ne quam fraudem frausus sit.
(‘I’m afraid he might have got into some mischief involving the two of us.’
Pl. As. 286)
Supplement:
. . . queror haud faciles . . . questus . . . (Stat. Silv. 4.8.32); . . . ex stipulatione quam ex-
traneus qui dotem dederit stipulatus fuerit . . . (Pap. Just. dig. 24.3.42.3); . . . ulciscar
ultionem tuam . . . (Vulg. Jerem. 51.36);
With many two-place compound verbs of motion the location involved is marked by
the accusative case. Most of these accusative-governing compounds are formed with a
preverb related to a preposition that governs the accusative, but compounds with
other preverbs may govern the accusative as well, as exirem in (c). Passivization is
possible, as in (d) and (e).
(a) Etenim quam tu domum, quam urbem adisti, quod fanum denique quod
non eversum atque extersum reliqueris?
(‘What house or town or sanctuary did you ever visit without verily sweeping and
scrubbing it?’ Cic. Ver. 2.52)
(b) Sed adire certum est hanc amatricem Africam.
(‘But I’m resolved to approach my African lover-girl.’ Pl. Poen. 1304)
(c) Iam ut limen exirem, ad genua accidit / lacrumans misera.
(‘As I was going out of the door, she fell to her knees, weeping miserably.’ Ter. Hec.
378–9—various emendations suggested by edd.)
(d) Circumiri enim sese ab aperto latere procurrentibus singulis arbitrabantur.
(‘For as the enemy kept charging singly they thought that they were being
surrounded on their exposed flank.’ Caes. Civ. 1.44.3)
(e) Per epistulam aut per nuntium, quasi regem, adiri eum aiunt.
(‘They say he’s approached by letter or by messenger, like a king.’ Pl. Mil. 1225)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The active verbs are rarely combined with a cognate object (in the accusative!). One
instance with the verb servio is (c). Servitutem is modified by the adjectives privatam
and publicam. Another example is (d), where noceo is unspecified. Note that noxam is
not modified by an attribute.
(c) Sed is privatam servitutem servit illi an publicam?
(‘But is he the slave of a private citizen or of the state?’ Pl. Capt. 334)
(d) Quandoque hisce37 homines iniussu populi Romani Quiritium foedus ictum
iri spoponderunt atque ob eam rem noxam nocuerunt, ob eam rem . . . hosce
homines vobis dedo.
(‘Whereas these men, unbidden by the Roman People of the Quirites, have guaran-
teed that a treaty should be ratified, and by so doing have committed an injury,
I therefore . . . deliver up these men to you.’ Liv. 9.10.9)
Supplement:
Si servus furtum faxit noxiamve noxit . . . (Lex XII.2);
Whether the use of a specific case with specific verbs is semantically determined is a
question that has received much attention. Although there are certain lexical classes
that have a particular case form for their second argument, attempts to generally
explain the use of a non-accusative have not been very satisfactory (see §§ 12.6–12.9).
37
For this nominative plural form, see TLL s.v. hic 2699.69ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The dative is regularly used for arguments of two-place verbs belonging to the
following four semantic classes:
(i) verbs of helping, caring, and their opposites
(ii) verbs of pleasing, flattering, and threatening
(iii) verbs of ruling, obeying, and serving
(iv) verbs of approaching and befalling.
A few verbs belonging to these classes are given in Table 4.4. Some of these verbs have
more than one frame (impero, for example, also has a three-place frame in its
meaning ‘to order’, as does minor in its meaning ‘to hold out the menace of ’) and/
or more than one case frame (ausculto, for example, governs an accusative in its
meaning ‘to listen to’). However, some verbs are found both with the dative and the
accusative, without a clear difference in meaning (curo, for example, in its meaning ‘to
care about’, in which case a prepositional construction with de exists as well). Finally,
there are diachronic developments in frames and case patterns (noceo, for example, is
found in the personal passive in Vitruvius and later, which suggests the existence of
an accusative object—see § 5.6). Instances of these verbs can be found in all periods of
Latin and in every type of text.
Other verbs belonging to the same semantic classes do not govern a dative at
all or only exceptionally. The best example is iuvo, which governs the accusative
both in its meaning ‘to help’ and in its meaning ‘to delight’.38 See below for a few
exceptions.
(a) . . . nec commodius ullo pacto ei poteris auxiliarier.
(‘ . . . and you won’t be able to help him more agreeably in any way.’ Pl. Trin. 377)
(b) . . . si segetibus aut vinetis cuiuspiam tempestas nocuerit . . .
(‘ . . . if a storm has damaged a man’s cornfields or vineyards . . . ’ Cic. N.D. 2.167)
(c) Nam illi aequom est me consulere qui causa mea / mendacium ei dixit.
(‘Yes, it’s only fair if I look after the man who told him a lie for my sake.’ Pl. Bac.
524–5)
(d) Non hic placet mi ornatus.
(‘I don’t like this outfit.’ Pl. Bac. 125)
(e) Bellus blanditur tibi.
(‘Your beau is buttering you up.’ Pl. Men. 626)
(f) Mihin’ equis iunctis minare?
(‘Are you threatening me with a span of horses?’ Pl. Men. 868)
(g) Si huic imperabo, probe tectum habebo, / malum quom impluit ceteris ne
impluat mi.
(‘If I command this one (sc. hand), I’ll have my hide protected properly, so that when
it blows rain on others they won’t rain on me.’ Pl. Mos. 870–1)
(h) Memini et praeceptis parebo.
(‘I remember it and I’ll obey your instructions.’ Pl. Mil. 1036)
(i) Meo me aequom est morigerum patri, eius studio servire addecet.
(‘It’s appropriate for me to be obedient to my father, it’s proper for me to serve his
desire.’ Pl. Am. 1004)
(j) Scripsi ad eum ut mihi Heracleam occurreret.
(‘I have written to him to meet me at Heraclea.’ Brut. ad Brut. 1.6.1)
Occasional intrusion of the dative with verbs that normally govern an accusative can
be illustrated by the following examples.
(k) Insipiens, semper tu huic verbo vitato aps tuo viro—# Quoi verbo?
(‘Stupid, always avoid this word from your husband—# Which word?’ Pl. Cas. 211)
(l) Amori haec curat (sc. Venus). Tritico curat Ceres.
(‘The latter takes care of love, but Ceres takes care of wheat.’ Pl. Rud. 146)
38
For very Late Latin instances of iuvo + dat., see TLL s.v. 747.74ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
39
The most complete discussion is Serbat (1996b). See also Pinkster (2011).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(CIL VI.8.3.40897 (Rome, 63/2 BC)); Postquam satis diu adversarios ab se ad dimi-
candum invitatos supersedere pugnae animadvertit . . . (B. Afr. 75.2);
Typically poetic instances: . . . fortis equus visae semper adhinnit equae. (Ov. Rem.
634); Aspirat primo Fortuna labori. (Verg. A. 2.385); . . . <sorsum> sapor insinuatur /
sensibus. (Lucr. 2.684);
Dative objects are also found with compounds formed with the preverbs ex- and de-,
examples of which are (f) and (g). With such compounds, the bare ablative or various
prepositional expressions are found as well (see § 4.83). The dative in (g) is a poetic
extension by Ovid instead of the ablative. Ex. (h) is morphologically ambiguous. The
dative with compounds with the preverb co-/com-/con- is relatively less common than
prepositional expressions with cum (see § 4.38).
(f) Numqui nummi exciderunt, ere, tibi / quod sic terram optuere?
(‘Did you lose the money, master, since you’re staring at the ground like this?’ Pl. Bac.
668–8a)
(g) . . . lateri cervina sinistro / vellera dependent, umero levis incubat hasta.
(‘ . . . a deer-skin hung from her left side, a light spear rested on her shoulder.’ Ov.
Met. 6.592–3)
(h) . . . laevo dependet parma lacerto.
(‘ . . . and the shield hangs from the left arm.’ Verg. A. 11.693)
Such constructions with a dative noun phrase, a form of sum, and a subject noun
phrase are usually called POSSESSIVE DATIVE (dativus possessivus) constructions. The
notion of possession, however, is problematic: the dative NP is most often an animate
being and the subject an abstract entity, but we cannot be sure that these are real
semantic restrictions on the construction. Furthermore, many of the subject
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
constituents cannot really be said to be in the possession of the dative constituent. The
semantic function of the dative constituent is much disputed. If it is taken as an
argument (that is the position taken in this Syntax), it is either a recipient or an
experiencer. As for the function of the verb sum, that is not clear either: in these
constructions sum is definitely neither a copula nor an auxiliary. Some scholars take it
as the existential use of sum (on which see § 4.94) and regard the dative as an adjunct
(for example, a beneficiary adjunct).40 We are on more solid ground when it comes to
the information structure: pragmatically, the dative NP is usually the topic and the
subject constituent the focus of the clause. As a corollary, the dative constituent is
usually definite—often a personal pronoun—and the subject constituent is most often
indefinite.
For the focal function of the subject in this construction, see already Priscian 18.11
(III.213.16–23K): Virgilius in VII (sc. 268): ‘Est mihi nata viro gentis quam iungere
nostrae’ pro ‘possideo natam’. Sed magis dativo quasi ad ignorantes utimur, ut in
supra dicto versu. Nam ad scientes esse natam, nomen autem proprium ignorantes,
dixisset ‘mea’ vel ‘mei nata Lavinia est’, subdistinctione posita post ‘natam’. Contra
autem nomine quidem cognito, ignorata vero cuius esset nata, dixisset ‘Lavinia mea
nata est’, post ‘Laviniam’ subdistinguendo . . . (‘Virgil in book 7: ‘I have a daughter; to
marry her to a man of our race . . . ’ (est mihi nata) is for ‘I possess a daughter’. But we
use the dative more to those who are in ignorance, as in the verse cited above. To
those who know that there is a daughter but do not know her name he would have
said ‘My daughter [mea or mei]—she is Lavinia’, with a break after nata. But in
contradistinction, if her name were known but if it were not known whose daughter
she was, he would have said, ‘Lavinia—she is MY daughter’ with a break after
Lavinia.’)41
The possessive dative construction is often equated with the construction of habeo ‘to
have’ or possideo ‘to possess’ with an object constituent, but the pragmatic and
semantic characteristics do not fully overlap with those mentioned above. As to the
relative frequency of the possessive dative construction and the habeo + object
construction, the information one finds in the literature varies. This is partly due to
choices of corpus, partly to the definition of the construction. Although in one
publication the possessive dative is described as decreasing after the Classical period,
it seems to be quite common in Pliny the Elder.42 The possessive dative is also often
40
For the function of the dative as experiencer, see Bolkestein (1983, 2001a). Scherer (1975: 126),
Stassen (2009: 49–50), and others regard the dative constituent as a satellite and sum as locative or
existential. Statistical, semantic, and pragmatic characteristics of the possessive dative construction can be
found in Baldi and Nuti (2010: 254–305). See also Bennett (1910/14: II.159–66), Löfstedt (1963b),
Bolkestein (1983a; 2001a), García Hernández (1993, 1994: 383–9, 1995), Magni (1999), and Cabrillana
(2003) [on the differences between genitive and dative], (2006) [a diachronic study], and (2010a: 100–2).
For individual authors, see Schenk (1892) on Cicero, Schunck (1900) on Plautus and Cicero, and
Önnerfors (2002) on Pliny the Elder.
41
The translation is by Jim Adams (p.c.).
42
On the gradual replacement of the possessive dative, see Ramos (1998). Önnerfors (2002) mentions
760 possessive dative constructions versus 232 habeo + object constructions in Plin. Nat. 7–11.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
considered interchangeable with the so-called possessive genitive, but this is certainly
wrong (see § 9.30).
The use of the dative with the verb sum can be compared to its use with the antonymic
compounds absum ‘to be lacking’ and desum ‘to be wanting’, as in (e) and (f ); and to its
use with verbs that refer to something coming into someone’s possession, like contingo
‘to fall to one’s lot’ and evenio ‘to fall by lot (to)’, as in (g). Alongside these two-place verbs
there are three-place verbs of giving and taking away that also govern a dative recipient.
Supplement:
(a) Regular instances: Pol si est animus aequos tibi, sat habes qui bene vitam colas.
(Pl. Aul. 187); Quid tibi mecum est commerci, senex? (Pl. Aul. 631); Paullisper, Lyde,
est lubido homini suo animo opsequi. (Pl. Bac. 416); Mihin’ si umquam filius erit, ne
ille facili me utetur patre. (Ter. Hau. 217); . . . <ad> Admetum, Molossum regem, cum
quo ei hospitium erat, confugit. (Nep. Them. 8.3); Belli gerendi ius Antiocho ne esto
cum illis qui insulas colunt, neve in Europam transeundi. (Liv. 38.38.3); Iter est iis
per hospitia certa. (Plin. Nat. 10.65); Leoni tantum ex feris clementia in supplices.
(Plin. Nat. 8.48); Gignit tota vita, quae est ei ad tricensimum annum. (Plin. Nat.
8.168—if it belongs here.);
(b) Less frequent instances:
Non-abstract subject: Ecquid in mentem est tibi / patrem tibi esse? (Pl. Bac. 161–2);
Sed illi patruo huius qui vivit senex / Carthaginiensi duae fuere filiae; / altera
quinquennis, altera quadrimula. (Pl. Poen. 83–5); . . . est ager sub urbe hic nobis:
eum dabo / dotem sorori. (Pl. Trin. 508–9); Non hercle <ego omnino> occidi, sunt
mi etiam fundi et aedes. (Pl. Truc. 174); Samia mihi mater fuit. Ea habitabat Rhodi.
(Ter. Eu. 106); Erant illi compti capilli et madentes cincinnorum fimbriae et fluentes
purpurissataeque buccae dignae Capua, sed illa vetere. (Cic. Pis. 25); Sunt mihi bis
septem praestanti corpore Nymphae . . . (Verg. A. 1.71); Est mihi nata viro gentis
quam iungere nostrae / non patrio ex adyto sortes, non plurima caelo / monstra
sinunt. (Verg. A. 7.268–70);
Inanimate dative constituent: Quid est enim iam non modo pudori, probitati, virtuti,
rectis studiis, bonis artibus sed omnino libertati ac saluti loci? (Cic. Fam. 5.16.4); Tum
neque nomen erat neque honos aut gloria monti. (Verg. A. 12.135); Praeterea quasdam
proprietates quibusdam locis esse . . . (Plin. Nat. 2.153); Duo sunt montes iuxta flumen
Indum: alteri natura ut ferrum omne teneat, alteri ut respuat . . . (Plin. Nat. 2. 211);
Tergori tanta duritia ut thoraces ex eo faciant. (Plin. Nat. 8.124);
Definite subject: Sunt tibi regna patris Dauni, sunt oppida capta/ multa manu . . . (Verg.
A. 12.22–3);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Finally, the dative is also used in more or less the same way with a number of copular
verbs discussed in § 4.97. Examples are (h) and (i). In addition, these verbs are used
with an existential meaning and some of them also as impersonal verbs of happening
(see § 4.12).
(h) Id quidem tibi hercle fiet, / ut vapules . . .
(‘That’ll happen to you, getting beaten . . . ’ Pl. As. 478–9)
(i) Cuius quidem tibi fatum, sicuti C. Curioni, manet . . .
(‘Whose fate awaits you, as it awaited Gaius Curio . . . ’ Cic. Phil. 2.11)
43
For instances, see TLL s.v. libet 1325.77ff., 1326.12ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
luxurior ‘to live immoderately’, scateo/scato ‘to gush’, and gemmo ‘to put out buds’.
Regular examples are (a)–(d).
(a) Amore abundas, Antipho.
(‘You’re spoiled for love, Antipho.’ Ter. Ph. 163)
(b) . . . villaque tota locuples est, abundat porco haedo agno gallina lacte caseo
melle.
(‘His entire farmhouse has an air of plenty and abounds with pork, goat’s meat, lamb,
poultry, milk, cheese, and honey.’ Cic. Sen. 56)
(c) . . . atque adeo ut frumento afluam . . .
(‘ . . . and so much that I overflow with corn . . . ’ Pl. Ps. 191)
(d) . . . mare velis florere videres.
(‘ . . . you might have seen the sea bloom with sails.’ Cato orat. 31)
Supplement (in alphabetical order by verb):
Itaque ille noster amicus, insolens infamiae, semper in laude versatus, circumfluens gloria,
deformatus corpore, fractus animo quo se conferat nescit. (Cic. Att. 2.21.3); . . . QUEM
P/RIMA·AETATE·FLORENT/EM ·MORS·DIRA·SUBRIP/UIT (CIL XII.3559.3–6 (Nîmes, AD III)); Scilicet
arte madent simulacra et docta vagantur, / nocturno facere ut possint in tempore ludos.
(Lucr. 4.792–3); . . . (Capua) optimorum civium mihique amicissimorum multitudine
redundat. (Cic. Pis. 25); . . . si vino scatat . . . (Pl. Aul. 558); Caelum caligine stat. (Sis. hist.
130);
Occasional genitives: In populo scelus est et abundant cuncta furoris. (Man. 2.600);
Ita ad satiatem terra ferarum / nunc etiam scatit . . . (Lucr. 5.39–40);
Examples of verbs of lacking are (e)–(g).44 With egeo and indigeo the genitive is
regular before Cicero (who prefers the ablative, certainly with egeo) and it is also used
in later authors. The genitive is very rare with careo. Occasional accusative alterna-
tives (other than neuter pronouns) are found mostly in Late Latin.
44
For the relative frequency of these verbs, see TLL s.v. egeo 233.48ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
4.30 The use of the ablative with fido (and its compounds) and nitor
The verb fido ‘to have confidence in’ and its compounds regularly have a dative object,
especially when referring to an animate being (see § 4.24). However, they are also
found with an ablative object denoting the source of the confidence (a so-called
ablativus causae), which is also regular with fretus ‘relying on’. In Late Latin, on the
other hand, prepositional expressions become regular. Nitor ‘to rely on’ is also
normally found with an ablative object (but also with a prepositional object with
in), and so are sto ‘to stand by’ and a few other verbs with more or less the same
meaning. Examples are (a)–(d).
(a) . . . sententiae quam ille se habiturum pecunia fidens non dubitabat.
(‘ . . . and that advice he felt sure of securing, trusting to the power of money.’ Nep. Lys. 3.5)
(b) . . . si de adiutorio Dei fideret bonus homo . . .
(‘ . . . if as a good human being he trusted in God’s help . . . ’ August. Civ. 14.27)
(c) . . . magnanimi viri freti virtute et viribus / superbe nimis ferociter legatos
nostros increpant . . .
(‘ . . . those bold men, trusting in their valour and strength, upbraided our envoys very
haughtily and aggressively.’ Pl. Am. 212–13)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Sic tribunus . . . sua virtute nixus . . . ducem hostium ferocissimum vicit . . .
(‘Thus the tribune, relying on his own excellence, conquered the most fierce leader of
the enemy.’ Quad. hist. 12)
Supplement (in alphabetical order by verb):
Ab iis qui maxime P. Clodi morte adquierunt. (Cic. Mil. 102); . . . Manlius animo
magis quam arte confisus scuto scutum percussit . . . (Quad. hist. 10b); T. Didius,
paucitate suorum (v.l. paucitati) diffidens . . . aptari iussit milites ad pugnam . . .
(Fron. Str. 1.8.5); Quod eis respondi, ea omnes stant sententia. (Pl. Cur. 250);
45
See TLL s.v. magnus 147.55ff. Torrego (2014) deals with these value expressions with sum as subject
complement.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(f) Nummi denarii decuma libella, quod libram pondo as valebat et erat ex argento parva.
(‘The tenth part of a nummus denarius (= a silver coin of ten asses) is a libella,
because the as was worth one pound by weight, and the as of silver was a small
one.’ Var. L. 5.174)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
4.33 The use of the ablative with fruor, fungor, potior, utor, and vescor
In the preceding sections, the ablative is more or less systematically used with verbs
belonging to certain semantic classes. The verbs (and a few compounds and related
expressions) in this section do not constitute one semantic class.46 What they have in
common is that they are also found with the accusative (in varying degrees) and also
with the genitive (especially potior). They are also used in the gerundive (see § 5.39).
Fruor ‘to enjoy’ is regularly found with the ablative, as in (a) (with an early
attestation of the accusative in Cato).47 Fungor ‘to perform’ is found with the
accusative in Early Latin, as in (b). The earliest undisputed attestation of the ablative
is (c). While the ablative is preferred in Classical Latin, the accusative continued to be
used in later periods alongside it. Potior ‘to obtain’48 is found with the ablative, the
accusative, and the genitive, as in (d)–(f), respectively, from Early Latin onwards.
From Cicero onwards, the ablative is the dominant case form, but the genitive
persists, especially in the expression potior rerum ‘to seize power’. With utor ‘to
use’, the ablative is dominant throughout Latinity, as in (g), although instances with
the accusative are found from Early Latin onwards, as in (h). With vescor ‘to enjoy’, ‘to
feed on’, the ablative is the regular case, as in (i), but isolated accusatives are found
from Accius onwards, as in (j).
(a) . . . ut sinat / sese alternas cum illo noctes hac frui.
(‘ . . . that he permits himself to enjoy her along with him on alternating nights.’ Pl.
As. 917–18)
(b) . . . qui . . . / tuam rem curet teque apsente hic munus fungatur tuom.
(‘ . . . who . . . happens to look after your business and who’s doing your job here in
your absence.’ Pl. Am. 826–7)
(c) . . . cum canes funguntur officiis luporum . . .
(‘ . . . when dogs act the part of wolves . . . ’ Rhet. Her. 4.46)
(d) . . . si ille hodie illa sit potitus muliere . . .
(‘ . . . in case he’s got hold of that woman today . . . ’ Pl. Ps. 1071)
(e) Nam hoc mi haud labori est laborem hunc potiri . . .
(‘Well, meeting with this suffering is no suffering to me . . . ’ Pl. Rud. 191)
(f) . . . viduli ubi sis potitus.
(‘ . . . on the same day that you get hold of your trunk.’ Pl. Rud. 1337)
(g) Qui utuntur vino vetere sapientis puto . . .
(‘I think people who drink old wine are wise . . . ’ Pl. Cas. 5)
46
The most complete survey of these verbs can be found in Nutting (1928a).
47
For instances with an accusative, see TLL s.v. fruor 1423.66ff.
48
For the use of the cases in various authors, see TLL s.v. potior 334.16ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
With vereor, the genitive is less common than the accusative; with cupio (cf. cupidus +
gen. in § 4.102) and fastidio it is almost confined to Early Latin, but it is regular with
the other verbs given above. Examples are (a)–(d). The genitive is rarely found with
49
For Greek influence, see Calboli (2009: 81–2), with references.
50
The genitive is called a ‘Genitiv des Sachbetreffs’ in German terminology. See Sz.: 82, with
references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
other two-place emotion verbs, such as gaudeo ‘to rejoice at’. In Late Latin alternative
expressions occur: for example, expressions with the accusative, especially in ecclesi-
astical texts. (For vereor + dative ‘to be worried about’ see § 4.46.)
(a) . . . sed sine argento frustra es qui me tui misereri postulas.
(‘But without money you’re deceiving yourself if you expect me to have pity on you.’
Pl. Ps. 378)
(b) Si non verear, nemo vereatur tui.
(‘If I were not in awe, no one would be in awe of you.’ Afran. com. 31)
(c) Vah, egone ut ad te ab libertina esse auderem internuntius, / qui ingenuis
satis responsare nequeas quae cupiunt tui?
(‘Bah! I dare to be a go-between for you from a freedwoman? You can hardly reply to
all the freeborn women who desire you?’ Pl. Mil. 962–3—NB: the text is a conjecture)
(d) Fastidit mei, / quia videt me suam amicitiam velle.
(‘He scorns me because he can see that I want his friendship.’ Pl. Aul. 244–6)
Supplement (in alphabetical order by verb):
Pol, quamquam domi cupio, opperiar . . . (Pl. Trin. 842); . . . verae beatitudinis . . . esurit
et sitit. (Apul. Socr. 22, p. 172); Mein’ (cj. Lipsius) fastidis, propudiose? (Pl. St.
334); . . . ipse sui miseret. (Lucr. 3.881); Miseremini familiae, iudices, miseremini
fortissimi patris, miseremini filii. (Cic. Flac. 106); . . . saepe miseritus sum generis
humani . . . (Sen. Ep. 108.13); Viduae non miserebuntur neque orphanis benefacient.
(Vulg. Bar. 6.37); . . . O superi, / . . . Arcadii . . . miserescite regis / et patrias audite preces.
(Verg. A. 8.572–4); . . . ne tui quidem testimoni . . . veritus superbum se praebuit . . . (Cic.
Att. 8.4.1); ‘Vereor’, inquam, ‘ignotae mihi feminae’ . . . (Apul. Met. 2.2.7);
Alternative case: acc.: Quid fuit boni (sc. dei) mala misereri . . . (Ruf. Adamant. 2.6);
dat.: Cui Venus postea miserta est . . . (Hyg. Fab. 58.3);
Exceptional other verbs: In amore suave est summo summaque inopia / parentem
habere avarum inlepidum, in liberos / difficilem, qui te nec amet nec studeat tui.
(Caecil. Com. 199–201); Neque ille sepositi ciceris nec longae invidit avenae . . . (Hor.
S. 2.6.83–4);51 Voti [-o cj.] gaudeo. (Apul. Met. 1.24.7—unless taken as a Graecism);
51
Certainly a Graecism, see Mayer (1999: 161–2).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
As for the variation in case mentioned above, there are both semantic and diachronic
tendencies. The accusative is more frequent with inanimate entities, as in (d). Neuter
forms of pronouns and adjectives are also in the accusative, as is often the case with
these categories. An example is (e). However, the genitive increases in frequency and
is predominant from the Augustan authors onward.52 The use of the genitive
correponds to its use with the adjective memor ‘mindful’ and its antonym immemor
(see § 4.102). The second argument may also be a clause (a prolative infinitive, an
accusative and infinitive, or—with memini—an ut clause).
(d) Mirum, ecastor, te senecta aetate officium tuom / non meminisse.
(‘It’s strange that you don’t remember your duty in your old age.’ Pl. Cas. 259–60)
(e) Em istuc si potes / memoriter meminisse, inest spes nobis in hac astutia.
(‘Well then, if you can remember this carefully, there’s hope for us in this scheme.’
Pl. Capt. 249–50)
Some scholars assume a difference in meaning between the genitive and accusative
expressions. The latter is generally regarded as more lively. This assumption is hard
to prove.
Supplement (in alphabetical order by verb):
Genitive:
Animate beings: Faciam ut mei memineris, dum vitam / vivas. (Pl. Per. 494–5); . . .
vivorum memini, nec tamen Epicuri licet oblivisci . . . (Cic. Fin. 5.3); Adhuc memi-
nimus Homeri. (Tert. Nat. 1.10.38);
Inanimate entities: Satin’ ut meminit libertatis? (Pl. Per. 658); Ipse certe agnoscet et
cum aliquo dolore flagitiorum suorum recordabitur. (Cic. Pis. 12); Iubes me bona
cogitare, oblivisci malorum. (Cic. Tusc. 3.35);
52
The most complete study is still Babcock (1901). A summary in Bauer (2009).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Accusative:
Animate beings: Cinnam memini. Vidi Sullam. (Cic. Phil. 5.17); . . . cur neminem se
retro meminit Epicurus . . . (Tert. An. 31.6); An ego Ulixem obliscar umquam aut
quemquam praeponi velim? (Acc. trag. 488);
Inanimate entities: Cum recordor eius ferocem et torvam confidentiam . . . (Pac.
trag. 36); Chaline, non sum oblitus officium meum. (Pl. Cas. 104); Obliviscor iam
iniurias tuas . . . (Cic. Cael. 50); Pronouns and adjectives: . . . quod te diligentissime
percipere et meminisse volam . . . (Cic. Fam. 11.7.2); Haecin te esse oblitum in ludo
qui fuisti tam diu. (Pl. As. 226);
53
Priscian uses the term ‘verba aequiperantia’ (III.274.3f.K).
54
For a very complete list of verbs (also including three-place verbs), see TLL s.v. cum 1370.20ff. For
com compounds, see Revuelta (forthc.).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
With some of these verbs, dative constituents are found as an alternative to the
prepositional expression with cum, if the meaning resembles one of the meaning
classes mentioned in § 4.24 (and Table 4.4). This use of the dative is typical of poetry
and post-Classical prose and regarded as a Graecism in Antiquity.55 Parallels of
(a) and (d) are (f ) and (g) (note that in (g) the dative is also accompanied by
an instrument adjunct and, in this case, a human being quo . . . viro). Compare (f)
with (b).
(f) Stoicorum quidem facilis conclusio est. Qui cum finem bonorum esse sen-
serint congruere naturae cumque ea convenienter vivere . . .
(‘For the Stoics indeed the conclusion is easy, since they hold it the sovereign good to
live according to nature and in harmony with nature . . . ’ Cic. Tusc. 5.82)
(g) . . . quo Roma viro doctis certaret Athenis—
(‘ . . . one with whom Rome could challenge learned Athens.’ Verg. Cat. 16.3—OCT
edition)
55
See Mayer (1999: 167–8). Not necessarily a Graecism according to Calboli (2009: 96–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Some of these verbs are found not only with an associative argument with cum, like
osculor in (i), but also with a patient, as in (j). Sometimes they have a plural subject
and resemble one-place verbs, as in (k).
(i) . . . quemque hic intus videro / cum Philocomasio osculantem . . .
(‘ . . . anyone I see in here kissing Philocomasium . . . ’ Pl. Mil. 460–1)
(j) Aliter homines amicam, aliter liberos osculantur.
(‘It’s one sort of kiss which a man gives his mistress, and another which he gives his
children.’ Sen. Ep. 75.3)
(k) Columbae proprio ritu osculantur ante coitum.
(‘Pigeons go through a special ceremony of kissing before mating.’ Plin. Nat. 10.158)
Supplement:
Accusative: Hic futuit multas . . . (Catul. 97.9);
Cum: ARPHOCRAS HIC CUM DRAUCA BENE FUTUIT DENARIO (CIL IV.2193 (Pompeii));
Coordination: . . . nescioquis inspectavit . . . / Philocomasium atque hospitem / oscu-
lantis. (Pl. Mil. 174–6);
Ab: . . . Romae a iudiciis forum refrixerit . . . (Cic. Att. 1.1.2) (OLD § 6b); Ei mihi
qualis erat, quantum mutatus ab illo / Hectore . . . (Verg. A. 2.274–5) (OLD
§ 2a); . . . cui mirum videbitur istum a maleficio propter acerrimam formidinem
non temperasse? (Rhet. Her. 2.29) (OLD § 2a);
Ad: . . . quae nihil attingunt ad rem . . . (Pl. Mer. 32) (OLD § 24c); Quam ad rem istuc
refert? (Pl. Epid. 276);
Cum: see § 4.38 (OLD § 13);
De: At hostes, posteaquam ex nocturno fremitu vigiliisque de profectione eorum
senserunt, . . . Caes. Gal. 5.32.1) (OLD § 12);
Ex: . . . in caede atque ex caede vivunt. (Cic. S. Rosc. 78—NB: remarkable coordin-
ation) (OLD § 14).
As with space satellites (see § 10.2) a distinction is made between three types of space
arguments: position in space (§ 4.42), direction or goal (§ 4.43), and source arguments
(§ 4.44). In the three subsections that follow, it is not always easy to decide whether
the space expression is an argument or a satellite.
With a number of verbs denoting position in space, a second argument indicating the
place where is required. Examples of position objects with the verb maneo ‘to stay’ are
(a)–(d). Note in (a) the co-occurrence of two position objects (see § 10.3 for the same
phenomenon with position adjuncts), and in (d) the co-occurrence of an abstract
position object (in servitute) and two literal position adjuncts (hic and ad suom patrem).
Sum ‘to be somewhere’, as in (e), can be considered among these verbs as well.
(a) <Ibo> atque ibi manebo apud argentarium.
(‘I’ll go and stay there at the banker’s.’ Pl. As. 126)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Position objects are found with the following verbs: habito ‘to live’, opperior ‘to wait’,
sedeo ‘to sit’, sto ‘to stand’, cubo ‘to recline’, iaceo ‘to lie’, versor ‘to be involved’, sum
‘to be somewhere’.56
The verbs consido ‘to sit’, ‘to settle’ and consisto ‘to stand still’, ‘to stay’ allow both a
position object and a direction or goal object (for three-place verbs that offer the same
option, see § 4.81). The position expression is more common than a direction or goal
expression, certainly in the Classical period, as in (f) and (g).
(f ) Sed, si videtur, considamus hic in umbra . . .
(‘But, if you please, let us sit down here in the shade . . . ’ Cic. Leg. 2.7)
(g) Cum hoc homine an cum stipite in foro constitisses nihil crederes interesse.
(‘To hold converse with such a man as this, or with a post in the forum would be all
one in your eyes.’ Cic. Red. Sen. 14)
For the various classes of constituents found as position objects (bare case forms,
prepositional phrases, adverbs, and subordinate clauses) see § 10.3.
With a number of verbs of motion, a second argument indicating the direction or goal is
required. This is the case with pervenio ‘to come (to a place or person)’ and appello ‘to
bring to shore’, as in (a) and (b), respectively.57 In (c) and (d), the reflexive expressions te
agis ‘you betake yourself ’ and me confero ‘I betake myself ’ are meaningless by them-
selves, so a direction argument is required. These direction arguments are found not only
with compound verbs of motion with the prefixes ad-, per-, sub-, but also with e.g.
56
For verbs of position in Latin, see Stengaard (1991: 29–59); also Hoffmann (1996: 203).
57
For examples of direction arguments, see Cabrillana (1997b). A number of direction expressions
(found with proficiscor and mitto) are discussed by Spevak (2010a: 164–71) and treated as arguments.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
contendo ‘to hasten’, redeo ‘to return’, revertor ‘to return’. For a remarkable use of the
dative in Late Latin, see (e).58
(a) Nunc perveni Chalcidem.
(‘Now I’ve arrived in Chalcis.’ Pl. Mer. 939)
(b) Appellitur navis Syracusas.
(‘The ship is brought into Syracuse.’ Cic. Ver. 5.64)
(c) Quo agis te? # Domum.
(‘Where are you going? # Home.’ Pl. Am. 450)
(d) Nunc ego me illac per posticum ad congerrones conferam.
(‘Now I’ll go that way to my boon companions through the back door.’ Pl. Mos. 931)
(e) . . . ubi . . . Dominus . . . clausis ostiis ingressus est discipulis . . .
(‘ . . . where the Lord came in through closed doors to his disciples . . . ’ Pereg. 39.5)
A direction object is also common with verbs like aspicio ‘to look at’, as in (f), and
specto ‘to look at’ (also in its meaning ‘face’, as in (g), indicating spatial orientation)
(NB: it is also used with a patient object). Direction of an abstract type with the same
verb is involved in (h).59
(f) Sed quaeso, hercle, agedum aspice ad me.
(‘But I ask you seriously, go on: look at me.’ Pl. Capt. 570)
(g) Huius lateris alter angulus . . . ad orientem solem, inferior ad meridiem
spectat.
(‘Of this side one angle . . . faces the east, the lower angle faces south.’ Caes. Gal. 5.13.1)
(h) Quae res ad caedem maximam spectat.
(‘This points to a large-scale massacre.’ Cic. Att. 14.13.2)
For the various classes of constituents found as direction or goal objects (bare case
forms, prepositional phrases, adverbs, and subordinate clauses), see § 10.7. Many of
these verbs also have a bare accusative case form as second argument (see § 4.22).
Source arguments are most commonly found with two-place verbs designating
‘motion away from’ or ‘separation’, as in (a)–(c). These arguments are usually
expressed by means of a prepositional phrase, but a few bare ablatives are attested
in Classical prose (even if they are often emended). Bare ablatives are preferred in
58
For the late use of the dative instead of an ad-expression under the influence of the Greek original,
see Vet. Lat. Num. 32.16 et accesserunt ei, discussed by August. Locut. Hept. 4.114: Et accesserunt ei et
dicebant. Non dicit quod usitatum est: accesserunt ad eum. See also de la Villa (1999b: 294–5).
59
For the direction verb frame of verbs of seeing, see Torrego (2003, 2007).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
poetry and are also used freely in prose from Livy onwards (the so-called ablativus
separativus), possibly because they were perceived as archaic.60
(a) . . . postea / clam patrem patria hac effugiam aut aliquid capiam consili.
(‘ . . . then I’ll flee from this country behind my father’s back or take some other
counsel.’ Pl. Mer. 659–60)
(b) Saepe etiam stellas vento inpendente videbis / praecipitis caelo labi . . .
(‘Often, too, when the wind is threatening, you will see stars shoot headlong from the
sky . . . ’ Verg. G. 1.365–6)
(c) Inde extemplo abire finibus Volcianorum iussi ab nullo deinde concilio
Hispaniae benigniora verba tulere.
(‘Being then bidden straightway to depart out of the borders of the Volciani, they
received from that day forth no kinder response from any Spanish council.’ Liv.
21.19.11)
When these verbs are used in a non-literal sense and when the source object is
more abstract, bare case expressions are quite common, as in the following examples
(d )–(f). In Classical prose such instances are sometimes emended.
(d) Eru’, quantum audio, uxore excidit.
(‘If I can believe my ears, my master has just lost a wife.’ Ter. An. 423)
(e) Hac spe lapsus Indutiomarus nihilo minus copias cogere . . . coepit.
(‘Though disappointed in this hope, Indutiomarus nonetheless set to work to raise
forces . . . ’ Caes. Gal. 5.55.3)
(f) . . . M. Catoni, qui cum multa vita excedens providit, tum quod te consulem
non vidit.
(‘ . . . Marcus Cato, who in taking leave of life displayed foresight in sparing himself
many sorry sights but none sorrier than you as consul.’ Cic. Phil. 2.12)
Source objects are found with some simple verbs (apart from the ones given, cedo ‘to
give way’, ‘to retire’ and—rarely—cado ‘to fall’, among others) and with many
compound verbs consisting of one of the preverbs ab-, de-, and ex- and verbal
stems related to the simple verbs cado, cedo, eo ‘to go’, gradior ‘to walk’, labor ‘to
slip’, sisto ‘to cause to stand’. With many of these compounds prepositional phrases
can be used as well. Compounds with the preverb dis- (e.g. discedo ‘to leave’) usually
govern a prepositional phrase (especially ab, but also ex and de), but poets and later
prose authors also use the bare ablative. (For a few other dis-compounds, see § 4.39)
For the use of the dative, especially in poetry, see (g). It is comparable with the use of
the dative with three-place verbs of taking away (see § 4.51).
60
So Adams (2013: 260–3). For Pliny the Elder and a few other instances in Cicero, see Pinkster
(2005b: 245). For the verb frame of disto, see Torrego (2014).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(g) Isque ubi se nullo iam cursu evadere pugnae / posse neque instantem
reginam avertere cernit . . .
(‘When he sees that by no fleetness can he escape combat or divert the queen from
her attack . . . ’ Verg. A. 11.702–3)
Supplement (in alphabetical order by verb):
Ablative: Sed quid currentem servom a portu conspicor, / quem navi abire vetui?
(Pl. Mer. 109–10); Quid mihi tam multas laudando, Basse, puellas / mutatum
domina cogis abire mea? (Prop. 1.4.2—NB: highly unusual); At enim, quoniam . . .
alii iracundi aut crudeles aut superbi sint, . . . alii <a> (Lambinus) talibus vitiis
abhorreant . . . (Cic. Fat. 8); Et repperi haec te qui apscedat suspicio. (Pl. Epid. 285);
‘Procul o procul este, profani’ / conclamat vates, ‘totoque absistite luco’. (Verg. A.
6.258–9); In quis notissimus quisque aut malo dependens verberabatur aut . . . (Sal.
Hist. 3.9); Si qui graviore vulnere accepto equo deciderat, circumsistebant. (Caes.
Gal. 1.48.6); . . . quae deciderant patula Iovis arbore, glandes. (Ov. Met. 1.106); Tum
regina deum caelo delapsa morantis / impulit ipsa manu portas . . . (Verg. A.
7.620–1); Numquam ante hoc tempus exercitum populi Romani Galliae provinciae
finibus egressum. (Caes. Gal. 1.44.7—NB: v.l. fines ingressum); Ubi portu eximus,
homines remigio sequi, / neque aves neque venti citius. (Pl. Bac. 289–90); . . . exit /
conditor urbe sua . . . (Ov. Met. 4.565–6); Ictum firmitas materiae sustinet et quid-
quid incidit fastigio musculi elabitur. (Caes. Civ. 2.11.1); Modo articuli suis sedibus
excidunt. (Cels. 8.11.1); Etenim prope est spelunca quaedam . . . , qua Ditem patrem
ferunt repente cum curru extitisse . . . (Cic. Ver. 4.107);
Prepositional phrase: Sed ex eo credo quibusdam usu venire ut abhorreant a Latinis,
quod . . . (Cic. Fin. 1.8); . . . uti ab signis legionibusque non absisterent. (Caes. Gal.
5.17.2); Ne balant quidem, quom a pecu cetero apsunt. (Pl. Bac. 1138a);
Dative: Huic tam pacatae profectioni ab urbe regis Etrusci abhorrens mos . . . manet
bona Porsennae regis vendendi. (Liv. 2.14.1); Si qui earum urbium cives quae regno
abscedunt cum rege Antiocho intraque fines regni eius sunt, Apameam omnes ante
diem certam redeunto. (Liv. 38.38.6); Serta procul, tantum capiti delapsa, iacebant .
. . (Verg. Ecl. 6.16);
Accusative: . . . abhorrentis quosdam cadaverum tabem detestabili voce confirmare
ausus est . . . (Suet. Vit. 10.3);
the Classical period. An example is (d). Family and clan relationships, finally, are
usually expressed with the ablatives of locus ‘place’, genus ‘family’, as in (e), familia
‘household’, or stirps ‘stock’, usually without a preposition, but see (f).
(a) Humana matre natus, humano patre, / mirari non est aequom, sibi si
praetimet.
(‘He’s born of a human mother and a human father; so it wouldn’t be fair to be
surprised if he’s afraid for himself.’ Pl. Am. 28–9)
(b) Parente P. Sestius natus est, iudices, homine, ut plerique meministis, et
sapiente et sancto et severo.
(‘Publius Sestius, gentlemen, was the son of a man, as most of you remember, who
was wise, scrupulous, and strict.’ Cic. Sest. 6)
(c) Atque ex me hic natu’ non est, sed ex fratre.
(‘In fact, he’s not my own son by birth but my brother’s.’ Ter. Ad. 40)
(d) . . . quem ait a Deucalione ortum . . .
(‘ . . . who, he says, descended from Deucalion . . . ’ Cic. Tusc. 1.21)
(e) Si me novisti minus, / genere quo sim gnatus.
(‘In case you don’t know what family I come from . . . ’ Pl. Aul. 777–8)
(f) Quo de genere natu’st illic Philocrates?
(‘What family does Philocrates belong to there?’ Pl. Capt. 277)
Supplement (in alphabetical order by verb):
. . . Telamone creatus . . . (Ov. Met. 13.22); Pandarus et Bitias, Idaeo Alcanore
creti . . . (Verg. A. 9.672); Maecenas, atavis edite regibus . . . (Hor. Carm. 1.1.1);
Haec e Tartarea tenebrica abstractum plaga / tricipitem eduxit Hydra generatum
canem? (Cic. Tusc. 2.22—NB: Cicero’s translation of Soph. Trach. 1097–8); Sic
itur ad astra, / dis genite et geniture deos. (Verg. A. 9.641–2); AROS RUFIS ATINEA
61
NATUS (CIE 3498 (Perugia)); Non enim silice nati sumus . . . (Cic. Tusc. 3.12);
Respice, quaeso, aliquando rem publicam, M. Antoni. Quibus ortus sis, non
quibuscum vivas considera. (Cic. Phil. 2.118—NB: no preposition62); Minerva
prima, quam Apollinis matrem supra diximus, secunda orta Nilo, quam Aegyptii
Saietae colunt . . . (Cic. N.D. 3.59); Denique caelesti sumus omnes semine oriundi.
(Lucr. 2.991); Tun’ meo patre es prognatus? (Pl. Men. 1079); . . . Camertem, /
magnanimo Volcente satum . . . (Verg. A. 10.562–3);
For the various classes of constituents found as source objects (bare case forms,
prepositional phrases, and adverbs) see § 10.13 (source adjuncts). For expressing
the home town or home country, the rules for topographical names apply (see
§ 10.17).
61
For possible Etruscan influence (matronymic), see Adams (2003b: 173).
62
For discussion and references, see Pasoli (1966: 69–75).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The examples of caveo below are arranged in the order ‘accusative + alternatives’ ‘to
beware of ’, as in (a)–(d) and ‘dative’ ‘to look out for’, as in (e)–(i). The second
argument in the ‘to beware of ’ cases is treated either as a patient or as a source (as in
§ 4.44). There is one odd example in Apuleius of a dative object where the meaning
must be ‘to beware of ’ (see the Supplement). Animate patients are only attested from
Cicero onwards, but that may be accidental. The ‘to look out for’ instances are much
less frequent, and it is remarkable that many datives are reflexive pronouns, as in (f).
Since reflexive pronouns also occur with ‘to beware of ’ frames in which the second
argument is a source, as in (e), or in which the second argument is an ut/ne clause (see
the Supplement), these reflexives may be also regarded as beneficiary satellites (on this
63
I thank Jeremy Brightbill and Branden Kosch for a detailed analysis of the verbs involved.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
use of reflexive pronouns, see § 10.70; see, however, also praecaveo in § 4.86,
Supplement). The first non-reflexive animate datives are attested in Cicero, as in
(g); inanimate datives are found from Velleius onwards, as in (h). There are also a few
instances of a three-place frame with an accusative object (or an ut-clause) and a
dative, as in (i) meaning ‘to arrange something for’.
(a) Cave sis malam rem.
(‘Watch out for trouble.’ Pl. As. 43)
(b) A quo admoniti diligentius et vigilantius caveamus Antonium.
(‘Warned by him, let us guard ourselves against Antonius with greater care and
vigilance.’ Cic. Phil. 11.10)
(c) Omitte, Lyde, ac cave malo.
(‘Drop it, Lydus, and watch out for trouble.’ Pl. Bac. 147)
(d) Quid est quod caveam? # Em, a crasso infortunio.
(‘What is it that I should watch out for? # There you go! For a fat thrashing.’
Pl. Rud. 833)
(e) Nunc, pater mi, proin tu ab eo ut caveas tibi . . .
(‘Now, my father, you should be careful of him . . . ’ Pl. Bac. 739)
(f) Gliscit rabies, cave tibi.
(‘His frenzy is increasing; look out for yourself.’ Pl. Capt. 558)
(g) Quid praetereo? An illud, ubi caves tamen Siculis et miseros respicis aratores?
(‘What am I leaving out? Is it the bit where you do after all safeguard the Sicilians and
take thought for these hapless farmers?’ Cic. Ver. 3.26)
(h) . . . condiciones pacis . . . discussit ac rupit, unice cavente Cicerone concor-
diae publicae.
(‘ . . . the terms of the peace . . . he broke and shattered, despite Cicero’s unparalleled
aims at preserving the harmony of the state.’ Vell. 2.48.5)
(i) Id, amabo te, huic caveas. # Quid isti caveam? # Ut revehatur domum . . .
(‘Please, do take care of this for her. # What should I take care of for her? # That she’s
taken back home . . . ’ Pl. Bac. 44)
Supplement (in alphabetical order by verb):
. . . quippe cum aviditati contumaciaeque summe cavere et utramque culpam
vitare . . . deberem. (Apul. Met. 11.21.5—NB: Oudendorp suggested emendations);
Cave sis tibi, / ne tu immutassis nomen, si hoc concreduo. (Pl. Aul. 584–5); Ne ille
edepol tergo et cruribus consuluit hau decore. (Pl. As. 409); Haruspicem, augurem,
hariolum, Chaldaeum ne quem consuluisse velit. (Cato Agr. 5.4); Non omnis aetas,
Lyde, ludo convenit. (Pl. Bac. 129); Ecquem convenisti? # Multos. (Pl. St. 342); . . . ipsi
cupio Glycerio. (Ter. An. 905); Omnia quae tu vis, ea cupio. (Pl. Per. 766); Manent
ingenia senibus, modo permaneat studium et industria . . . (Cic. Sen. 22); Quamquam
hic manere me erus sese iusserat, / certum est. (Pl. Aul. 680–1); Ego cunas recessim
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
rursum vorsum trahere et ducere, / metuens pueris, mihi formidans. (Pl. Am.
1112–13); Mihi, nisi ut erum metuam et curem, nihil est qui tergum tegam. (Pl.
Mos. 992); Nolo ego mi te tam prospicere . . . (Pl. Trin. 688); . . . ipse autem ex equo
nudatam ab se provinciam prospicit. (Cic. Ver. 2.154); Nil me curassis, inquam, ego
mi providero. (Pl. Mos. 526); Non hercle te provideram . . . (Pl. As. 450); Abi intro, ne
molestu’s, linguae tempera. Pl. Rud. 1254); Eodem modo latitudinem orbiculis ligneis
aut armillis ferreis temperato, usque dum recte temperabitur. (Cato Agr. 22.2); . . .
dicitur et Atheniensis Clisthenes Iunoni Samiae, civis egregius, quom rebus timeret
suis, filiarum dotis credidisse. (Cic. Leg. 2.41); Si illum relinquo, eius vitae timeo. Sin
opitulor, huius minas . . . (Ter. An. 210); . . . eo minus veritus navibus, quod . . . (Caes.
Gal. 5.9.1); Per . . . / Iunonem, quam me vereri et metuere est par maxume . . . (Pl. Am.
831–2); Quid agit filius? / Bene volt tibi. (Pl. Trin. 437–8); Secede huc, Libane. Te volo.
(Pl. As. 639);
In the case of cupio, metuo, timeo, and vereor in combination with a dative constitu-
ent, the existence of more or less synonymous prepositional alternatives raises the
question of whether the verbs should be regarded as one-place, combined with a
beneficiary satellite. This idea might find further support in the existence of instances
like (j) and (k), where the verb is found in its two-place frame, with its accusative
object and a dative.64
(j) Te suis matres metuunt iuvencis . . .
(‘Mothers fear you for the sake of their sons . . . ’ Hor. Carm. 2.8.21)
(k) An aliis licet, et recte licet, in meo metu sibi nihil timere . . .
(‘Are others permitted, and rightly permitted, to find no cause for personal appre-
hension in what brings fear to me . . . ’ Cic. Dom. 8)
Supplement (in alphabetical order by verb):
Quodsi interesse quippiam tantulum modo potuerit in amicitia, amicitiae nomen
iam occiderit, cuius est ea vis ut, simul atque sibi aliquid alter maluerit, nulla sit. (Cic.
Leg. 1.34); Patri quid non timeo, si eum apud piratas relinquo? (Sen. Con. 7.4.2);
Nihil est enim mali, nihil sceleris quod illa non ab initio filio voluerit optaverit
cogitaverit effecerit. (Cic. Clu. 188—NB: note the parallelism);
The verb moderor can also be used with a dative or an accusative second argument.
However, it is difficult to associate distinct meanings with the two case patterns (see
OLD). The dative is especially used by Plautus and Cicero and is preferred by them in
certain combinations (for example, orationi moderor ‘to restrain one’s language’),
where moderor ‘refer(s) to exercising self-restraint’. Moderor with the accusative
‘refer(s) to restraining or controlling an entity external to the agent’.65 But Plautus
has domino moderor ‘to restrain one’s master’ and Sallust linguam moderor ‘to keep
one’s tongue in check’. The picture is not very clear.
64
Álvarez Huerta (2010a) argues that the verb timeo has two semantically different two-place frames.
65
See Blake (2001: 39, 145).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
66
For the various constructions in which paenitet can be used, see Baños (2003). For the older
instances of these expressions, see Traina (1963, 1999: 21–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
For the verbs in the left column of Table 4.10, ‘personal’ alternatives with the
experiencer as subject exist. In this construction the cause of the emotion is still
expressed (see § 4.35). As for ‘personal’ alternatives for the verbs in the right column,
instances of ‘personal’ forms with an explicit cause constituent as subject are rare and
mostly Late.67 Examples are (h), with quod as the subject, and (i) with a clause as the
subject. The earliest attestation of personal paeniteo is (j) with the cause as the subject.
It is regarded as an intrusion of later date.68
(h) Nimio id quod pudet facilius fertur quam illud quod piget.
(‘That which causes embarrassment is much easier to handle than that which causes
regret.’ Pl. Ps. 281)
(i) . . . ut eum paeniteat non deformem esse natum.
(‘ . . . as to regret that he was born not unhandsome.’ Cic. Cael. 6)
(j) Et me quidem haec condicio nunc non paenitet . . .
(‘I don’t regret this match now . . . ’ Pl. St. 51)
Supplement (in alphabetical order by verb):
Deinde utrum id facinus sit, quod paenitere fuerit necesse . . . (Cic. Inv. 2.43);
Sequitur, ut nihil paeniteat, nihil desit, nihil obstet (Cic. Tusc. 5.53); Sapientis
est . . . proprium nihil quod (cuius cj. Lambinus) paenitere possit facere . . . (Cic.
Tusc. 5.81—see Dougan ad loc.); . . . pridie caveat ne faciat quod pigeat postridie.
(Pl. St. 122); Nilne te pudet? (Pl. As. 933); Pudet nil? (Ter. Ad. 244); Non te haec
pudent? (Ter. Ad. 754); Pudebat, credo, commemoramentum stupri. (Caecil.
com. 166 R3);
Personal constructions with the experiencer as subject are infrequent but found from
Early Latin onward, resembling in this respect verbs like gaudeo ‘to be glad’ and maereo
‘to be sad’. An example is (k). The present and perfect participles are found regularly,
67
For Late Latin examples, see TLL s.v. paeniteo 62.56 ff.
68
See Petersmann ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
especially of paeniteo and pudeo, as in (l). Personal use of paeniteo becomes frequent
from c. AD 100 onward, especially in (moralizing) texts of ecclesiastical authors.
(k) Ita nunc pudeo atque ita nunc paveo . . .
(‘So much am I ashamed now, so much am I afraid now . . . ’ Pl. Cas. 877)
(l) Optimus est portus paenitenti mutatio consili.
(‘The best recourse when experiencing regret is a change of policy.’ Cic. Phil. 12.7)
Supplement (in alphabetical order by verb):
Proloqui non paenitebunt liberi . . . (Pacuv. 31); Sapientem nihil opinari, nullius rei
paenitere, nulla in re falli, sententiam mutare numquam. (Cic. Mur. 61—NB: coord-
ination with indisputable subjects);69 Aetolos . . . si paenitere possint, posse et incolu-
mis esse. (Liv. 36.22.3); . . . et non vult paeniteri (v.l. paenitere) a fornicatione sua.
(Vet. Lat. Apoc. 2.21); . . . pertaesus, ut scribit, morum perversitatem eius . . . (Suet.
Aug. 62.2—NB: with accusative object); . . . dum pudeo, pereo (Val. com. 1.4—cj.); . . .
qui assiduis conviciis pudere didicerat . . . (Sen. Dial. 2.17.3); . . . et perhorrescat necesse
est et pudeat tacitus et paeniteat et gaudeat et admiretur . . . (Gel. 5.1.3);
Gerund and gerundive: Consilii nostri, ne si eos quidem qui id secuti sunt non
paeniteret, nobis paenitendum putarem. (Cic. Fam. 9.5.2); Ad paenitendum pro-
perat cito qui iudicat. (Pub. Sent. A 32);
Appendix 1: The verbs in the right column of Table 4.10 are also found with a genitive
constituent referring to the person in whose presence one feels a sense of shame, etc.:
Deum [= deorum] hercle me atque hominum pudet. (Pl. Trin. 912). There is one
example where both such a genitive and a genitive of cause are found together: Patris
mei meum factum [= meorum factorum] pudet . . . (Enn. scen. 59V=37J).
Appendix 2: Rarely a ‘personal’ construction is replaced by an ‘impersonal’ one with
the cause of the emotion as subject of the clause. Horreo ‘to shudder’ is found with
the cause as the object (in the accusative) in its meaning ‘to shudder at’, ‘tremble at’
from Cicero onwards, as in (m). A Late Latin instance of ‘impersonal’ horret with the
cause as the subject is (n).
69
For this example and the use of subjects with this class of verbs in general, see Fugier (2001: 345–50).
70
For discussion, see Löfstedt (1936: 130–42).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
71
For possible explanations (nominative mea res fert or ablative mea re fert), see Sz.: 84.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) Usque adeo magni refert studium atque voluntas, / et quibus in rebus
consuerint esse operati / non homines solum sed vero animalia cuncta.
(‘Of so great import are devotion and inclination, and what those things are
which not men only, but indeed all creatures, are in the habit of practising.’ Lucr.
4.984–6)
Supplement (in alphabetical order by verb):
Genitive: Quid quaeso eius intererat? (Cic. S. Rosc. 96); Et coli moverique terram
callumque summae cutis solvi aquarum interest. (Plin. Nat. 31.53); . . . ex his solis
causis quae ipsius intersint . . . (Ulp. dig. 4.4.3.4—NB: plural subject); . . . ipsorum id
plurimum referre. (Liv. 40.34.10); . . . QUOD EIUS REI QUAERUNDAI CENSEANT REFERE . . .
(CIL I2.583.32 (Lex Acilia, Rome, 122 BC)—NB: gerundival clause); Nam ut taceam de
neglegentibus, quorum nihil refert ubi litium cardo vertatur . . . (Quint. Inst. 12.8.2);
Ablative: . . . non quo mea interesset loci natura qui lucem omnino fugerem . . . (Cic.
Att. 3.19.1—NB: substantival subject); . . . ut saltem deliberare plebes Romana possit,
quid intersit sua, quid expediat . . . (Cic. Agr. 2.66); . . . videbis, ut soles, omnia quae
intelleges nostra interesse . . . (Cic. Att. 6.4.2—NB: plural subject); Sed non solum rei
publicae verum etiam mea interest hanc recipi rogationem. (Quint. Decl. 253.6);
Ecquid est quod mea referat? (Pl. Rud. 949); Quod promisisset non plus sua
referre quam si, cum auctionem venderet, domini iussu quippiam promisisset.
(Cic. Quinct. 19);
Appendix: Exceptional are instances of interest such as (f) and of rēfert such as (g),
with the entity to which something matters in the dative case, which are best
explained as beneficiary adjuncts (see § 10.70).72
72
Ex. (g) is regarded as an instance of contamination of cui rei te adsimulare profuit and quid . . . retulit
by Sz.: 84. See also TLL s.v. 2289.34 ff. and elsewhere for the variety of constructions of interest.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
It is not always easy to ascertain the valency of this class of verbs that require three
distinct arguments. Often one of the three arguments is implied on the basis of
contextual information and so remains unexpressed. Furthermore, no particular
third argument needs to be intended, as in (c). This looks like the absolute use of
two-place verbs discussed in § 4.5.
(c) Galliae . . . suum genus farris dedere . . .
(‘The Gallic provinces have . . . produced a special kind of emmer . . . ’ Plin. Nat. 18.62)
On a high level of abstraction, two or possibly three different frames can be distin-
guished. Most three-place verbs with three distinct participants implement at least
one of the following, viz.:
(i) the ‘treatment’ or ‘supplying with’ frame
(ii) the ‘transfer’ or ‘giving’ frame
(iii) the ‘spatial’ frame
Ex. (a), already quoted previously, is a good example of the ‘treatment’ or ‘supplying
with’ frame. In this example, the object (civem sceleratum) is in the accusative case,
has the semantic function of patient, and may become the subject in the passive, like
domus in (b). The third argument (pecunia), in turn, is in the ablative case and
formally resembles an instrument adjunct, but semantically not all ablative third
arguments can be described as instruments (details in § 4.53). In contrast to the
accusative object, the ablative constituent cannot become the subject of the passivized
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
clause. Prototypically, the accusative patient is an animate being, the ablative inani-
mate. Note, however, that in specific contexts the ablative constituent may be
animate, as in (c), where it occurs in a military context. The case pattern {accusative
+ ablative} is characteristic of this frame.
(a) Tu civem sceleratum . . . pecunia . . . instrues?
(‘Are you going to furnish a criminal citizen with . . . money?’ Cic. Phil. 5.6)
(b) At domus interior regali splendida luxu / instruitur . . .
(‘But the palace within is laid out with the splendour of princely pomp . . . ’ Verg.
A. 1.637–8)
(c) Itaque naves omnes . . . propugnatoribus instruxerunt . . .
(‘Accordingly they manned all the ships with combat troops . . . ’ B. Alex. 10.4)
The ‘transfer’ or ‘giving’ frame is exemplified by (d). Here, regna omnia Syrorum is
the second argument in the accusative, with the semantic function of patient. In the
passive, this constituent would be the subject of the clause. The dative constituent cui
is the indirect object, with the semantic function of recipient. It cannot become
subject of a passive clause. The case pattern of this example {dative + accusative} is
characteristic of this frame. Prototypically, the dative recipient is an animate being,
the accusative inanimate.
(d) . . . Gabini, cui regna omnia Syrorum, . . . donaras . . .
(‘ . . . Gabinius to whom you had made a present of all the realms of Syria . . . ’ Cic.
Dom. 124)
The ‘spatial’ frame is exemplified by (e). The verbs involved require a second
argument in the accusative and a position, direction, or source argument, usually
a prepositional phrase. Only the second argument can become the subject of a
passive clause.
(e) Nisi mucrones etiam nostrorum militum tremere voltis dubitantis utrum in
cive an in hoste figantur.
(‘Unless you wish the very blades of our soldiers to waver, uncertain whether they are
being plunged into a fellow citizen or into a public enemy.’ Cic. Phil. 14.6)
Whereas most verbs have only one frame, there are a number of verbs that occur both
in the first and in the second frame. The verb dono, for example, has not only the
‘transfer’ or ‘giving’ frame illustrated by (d), but also the ‘treatment’ or ‘supplying
with’ frame, as in (f). An interesting instance of frame variation can be found in (g).
There are more than 100 verbs that occur in two frames. Details are given in § 4.54.73
73
These verbs have been studied by Lemaire (1983), who gives a list (to be used with care) on
pp. 321–4.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(f) Hinc ille est anulus aureus quo tu istum in contione donasti.
(‘And hence the golden ring with which you publicly presented that man of yours.’
Cic. Ver. 3.185)
(g) Quo miraculo matris salus donata filiae pietati est, ambaeque perpetuis
alimentis . . .
(‘In consequence of this marvel the daughter’s pious affection was rewarded by the
mother’s release and both were awarded maintenance for life.’ Plin. Nat. 7.121)
From the formal point of view, the following case and preposition patterns can be
distinguished for three-place verbs with three distinct arguments (only active forms
are taken into account; in passive clauses ‘accusative’ corresponds to ‘nominative’):
Examples for the N3 constituents in the ablative and dative and with a preposition can
be found above in (a), (d), and (e), respectively. Examples for the other patterns are
(h) {acc N2 + gen N3}, (i) {acc N2 + acc N3}, and (j) {dat N2 + abl N3}.
(h) . . . atque insimulabit eam probri.
(‘ . . . and he’ll accuse her of adultery.’ Pl. Am. 477)
(i) Roga me viginti minas . . .
(‘Ask me for twenty minas . . . ’ Pl. Ps. 114)
(j) Feminis dumtaxat purpurae usu interdicemus?
(‘Shall we forbid women alone from the use of purple?’ Liv. 34.7.3)
The patterns in the first row of Table 4.11 with an accusative constituent and a non-
accusative constituent are the most frequent; those in the last row are exceptional.
Whereas most verbs in the first row of the table allow only one N3 form, quite a few
allow more than one. Verbs that belong to the same semantic class often have the
same form or a number of more or less synonymous alternative forms for the N3, but
many verbs belong to more than one semantic class and for that reason allow more
than one form for the N3 as well. The pattern with two accusatives is firmly associated
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
with a few verbs. It also occurs as an occasional alternative of all the other patterns,
although not necessarily always grammatically correct (see below § 4.72).
The remainder of this section on three-place verbs is structured along the lines of
Table 4.11. Verbs or classes of verbs that have more than one case or preposition
pattern are discussed in the section about their most common pattern.
Examples of the pattern under discussion are (a)–(d). Whereas (a)–(c) have the
prototypical combination of an inanimate object and an animate (human) recipient,
in (d) both arguments are inanimate.
(a) Ea . . . dat mihi coronas.
(‘She . . . gives me garlands.’ Pl. Aul. 23–5)
(b) Quid ait? Quid narrat? Quaeso, quid dicit tibi?
(‘What does he say? What’s his story? Please, what did he tell you?’ Pl. Ps. 1080)
(c) Toxilo has fero tabellas, tuo ero.
(‘I’m bringing these tablets to your master Toxilus.’ Pl. Per. 247)
(d) In iis tignis parietes extruito iungitoque materiae uti oneris satis habeat.
(‘On these timbers build a wall and join it to the timber to give it sufficient weight.’
Cato Agr. 18.6)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The dative recipient arguments with these verbs formally resemble dative beneficiary
adjuncts. The difference between the two becomes apparent from a comparison
between (e) recipient, and (f) beneficiary.74
(e) Ac ni C. Memmius . . . edocuisset id agi, ut per paucos factiosos Iugurthae
scelus condonaretur . . .
(‘And had not Gaius Memmius made it clear . . . that the motive of these tactics was
to pardon Jugurtha for his crime through the influence of a few of his partisans . . . ’
Sal. Jug. 27.2)
(f) Praeterita se Diviciaco fratri condonare dicit.
([Caesar has pointed out to Diviciacus’ brother Dumnorix what he has to blame in
him.]‘He [sc. Caesar] said [sc. to Dumnorix] that he excused the past in consideration
for his brother, Diviciacus.’ Caes. Gal. 1.20.6)
The four classes distinguished in Table 4.12 contain both simple verbs and compound
verbs. The latter belong to these meaning classes on the basis of either the meaning of
the preverb, the meaning of the simplex, or the combined meanings of the preverb
and the simplex.
Both simplex verbs and compounds are also found with another pattern consisting
of an accusative and a prepositional phrase denoting direction. Several of the com-
pounds are predominantly found with a prepositional third argument; the prepos-
ition often corresponds to the preverb of the compound, as with the verb accommodo
‘to adapt’ in (g) and (h). The choice between the two alternatives is to some extent
dependent on whether the third argument is inanimate or animate (in the latter case,
the dative is preferred) and on whether the verb is used in its figurative or literal
meaning (in the latter case, the preposition is used more often). The existence of two
distinct constructions such as these is an important element in the evolution of the
case and prepositional systems in Latin. See § 12.31.
(g) Haec enim nostra oratio multitudinis est auribus accommodanda, ad ob-
lectandos animos, ad impellendos . . .
(‘For this oratory of ours must be adapted to the ears of the multitude, for charming
or urging their minds to approve of proposals.’ Cic. de Orat. 2.159—NB: co-occurrence
with purpose ad-clauses)
(h) Et quoniam non ad veritatem solum, sed etiam ad opiniones eorum qui
audiunt accommodanda est oratio, hoc primum intellegamus . . .
(‘And because it is necessary to adapt one’s discourse to conform not only with the
truth but also with the opinions of one’s hearers, the first point that we must grasp is
that . . . ’ Cic. Part. 90)
The variation between the accusative + dative and accusative + prepositional phrase is
shown for verbs of giving in (i) and (j).
74
See TLL s.v. condono 157.1ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The notions of ‘bringing’ may imply the shift of ownership of a certain entity, but they
may also imply the transportation from one owner or place to the other. The verb
mitto ‘to send’, for example, may denote both transportation of an entity from one
location to another and transfer of an entity from one person to another. In the first
case, it has a direction third argument, in the second, a recipient third argument.
Examples are (k) and (l), respectively.
(k) Recita quas ad Neronem litteras misit.
(‘Read the letter which he sent to Nero.’ Cic. Ver. 1.83)
(l) . . . in litteris quas Neroni mittis, . . .
(‘ . . . in the letter you send Nero . . . ’ Cic. Ver. 1.80)
In many studies and manuals the two expressions are considered synonymous.
However, there are distributional differences both for the object constituents and
for the third arguments, which show that this is not the case.75
Supplement (in alphabetical order by verb):
Dative expressions: Nunc adeo nisi mi huc argenti affert viginti minas . . . (Pl. As.
532—NB: recipient dative and direction adjunct juxtaposed); Inde ignem in aram,
ut Ephesiae Dianae laeta laudes / gratisque agam . . . (Pl. Mil. 411–12); . . . atque
Alcumenae in tempore auxilium feram . . . (Pl. Am. 877); Megadorus iussit Euclioni
haec mittere. (Pl. Aul. 353); Redduco hanc tibi. (Pl. Per. 659);
Direction expressions: Illud facito ut memineris, / convenisse ut ne quid dotis mea
ad te afferret filia. (Pl. Aul. 257–8); . . . quamplures ad praetores et consules vinum
75
For a recent survey of the literature and the argumentation involved, see Baños (1996c, 2000) and
Adams (2013: 278–94).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
honorarium dabant. (Cato orat. 132); Tu servos iube / hunc ad me ferant. (Pl. Men.
955–6); . . . ut hanc ne quoquam mitteret nisi ad se hunc annum totum. (Pl. As. 635);
. . . Demetri Magnetis librum quem ad te misit de concordia velim mihi mittas. (Cic.
Att. 8.12.6); Ibo domum atque ad parentes redducam Selenium. (Pl. Cist. 629–30);
A number of communication verbs also have two alternative expressions: either with
a direction argument or with an addressee argument. Instances include the verbs
nuntio ‘to report’ and scribo ‘to write’ in the examples (m)–(p). The direction
expression is relatively more frequent in Cicero’s letters (which were indeed sent
and were not meant for publication) than in the letters of Seneca and Pliny.76
(m) Nunc tu illum si illo es missurus, dice, [de]monstra praecipe / quae ad
patrem vis nuntiari.
(‘Now if you’re going to send him there, tell him, show him, teach him what you want
to be reported to your father.’ Pl. Capt. 359–60)
(n) Aperite, aperite, heus, Simoni me adesse aliquis nuntiate.
(‘Open up, open up! Hey, someone announce to Simo that I’m here!’ Pl. Ps. 1284)
(o) . . . quem ad modum tute ad Neronem scripsisti . . .
(‘ . . . as you wrote to Nero yourself . . . ’ Cic. Ver. 1.84)
(p) Scripsi tibi quae hic gererentur.
(‘I am writing to tell you what is going on here.’ D. Brut. Fam. 11.11.1)
The dative can also be used independently and interpreted as the marker of the
recipient or addressee of a letter, as in (q), written on the outside of one of the
Tabulae found in the army camp in Windisch (Switzerland):
76
Baños (2000: 13) has a table that illustrates this well. Vincent (1999), following Molinelli (1996), and
Iliescu (1999) have taken these examples as exchangeable. See, however, Martín Rodríguez (1986),
Pinkster (1990), Rosén (1999: 137–49), and Baños (2000) for further discussion. The material can be
found in TLL s.v. ad 557.78ff. For Semitic influence on the use of ad with communication verbs in Latin
versions of the Old Testament, see Sznajder (2012a).
77
For the verb frame of suadeo (and persadeo), see Tarriño (2007).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The verb iungo in its meaning ‘to join’, ‘to fasten to’, ‘to attach to’ (OLD s.v. 2a) is found
with a dative recipient (see (d) above). Not surprisingly, this meaning allows a prepos-
itional directional alternative with ad as well, as in (r). A third option is a prepositional
phrase with cum, functioning as an associative, as in (s). However, iungo is also found
with the ablative (according to some scholars this was the original pattern of this verb
and of misceo ‘to mix’—see also § 4.53),78 as in (t). In this pattern, the second argument
is a patient and the third is the entity the patient is provided with by an act of joining.
Exx. (u)–(w) from Cicero, involving abstract nouns, can be explained along the same
lines: in (u) sapientia is added to eloquentia. Ex. (v) is the associative variant, whereas
(w) represents the supplying with frame. The dative is said to increase from Classical
Latin onwards, resulting in a decrease of the ablative. However, although this observation
may be correct from the statistical point of view, it is not clear whether the dative replaces
the ablative in all the frames that are typical of the ablative. It may rather be the case that
the texts of later times have more verbs that have a frame that requires the dative.
(r) Quo in certamine perditi civis erat non se ad eos iungere quibus . . .
(‘In this contest, it was characteristic of a degenerate citizen not to join himself to
those for whom . . . ’ Cic. S. Rosc. 136)
(s) Non bene cum parvis iunguntur grandia rebus.
(‘Great matters are not joined well with small ones.’ anth. 407.11)
(t) Agger is bonus, qui intrinsecus iunctus fossa aut ita arduus ut eum trans-
cendere non sit facile.
(‘Serviceable is a bank which is close to the ditch on the inside or so steep that it is not
easy to climb.’ Var. R. 1.14.3)
(u) Sive hunc oratorem, quem ego dico sapientiam iunctam habere eloquentiae,
philosophum appellare malet, non impediam.
(‘ . . . alternatively I shall raise no obstacle if he prefers to designate as a philosopher
the orator whom I on my side am now describing as possessing wisdom combined
with eloquence.’ Cic. de Orat. 3.142)
(v) Sed ut Aristoteles . . . dicere docere etiam coepit adulescentes et prudentiam
cum eloquentia iungere . . .
(‘But just as Aristotle also began to teach the young to speak and combine wisdom
with eloquence . . . ’ Cic. Tusc. 1.7)
78
For a short history of these verbs and references, see Sz.: 115. See also von Geisau (1916).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(w) Quae quo maior est vis (sc. eloquentiae), hoc est magis probitate iungenda
summaque prudentia.
(‘And the stronger this faculty is, the more necessary it is for it to be combined with
integrity and supreme wisdom . . . ’ Cic. de Orat. 3.55)
Supplement (in alphabetical order by verb):
Dative expressions: Illic homo socium ad malam rem quaerit quem adiungat sibi.
(Pl. As. 288); Ipse . . . / dona recognoscit populorum aptatque superbis / postibus.
(Verg. A. 8.720–2); Huic (sc. Palatio) Cermalum et Vel[l]ias . . . coniunxerunt. (Var. L.
5.54); . . . iners hiemi continuatur hiems. (Ov. Pont. 1.2.24); Altera ratis aeque lata,
longa pedes centum, ad traiciendum flumen apta, huic copulata est. (Liv.
21.28.7); . . . curru (NB: dative) iungit Halaesus equos . . . (Verg. A. 7.724); Demens
qui servo filiam iunxit. (Sen. Con. Exc. 7.6 tit.); . . . cum causae causa nexa rem ex se
gignat. (Cic. Div. 1.125); . . . ne cui me vinclo vellem sociare iugali . . . (Verg. A. 4.16);
Direction expressions: Quam ad probos propinquitate proxume te adiunxeris, / tam
optumum est. (Pl. Aul. 236–7); Quod ne faciat (admissarius), paulisper ad molam
iunctus amoris saevitiam labore temperat . . . (Col. 6.37);
In the above instances of iungo and its compounds, the accusative argument is an
affected patient. This frame should not be confused with the two-place frame in its
meaning ‘to make by joining’ (OLD s.v. § 5), with an effected object in the accusative and
an ablative instrument adjunct, as in (x). Here a dative is obviously excluded. In (y), the
two sides of the river Ticinus are connected, the bridge being an instrument adjunct.
mixing misceo
‘to mix’
Examples of these subclasses are (a)–(h). However, several of these verbs are found
with other case and/or preposition patterns as well, as will be shown. This use of the
ablative is usually called the ablativus instrumenti, just like the use of the ablative in
instrument adjuncts (see § 10.53). Whereas with some verbs the semantic relation of
the third argument is instrument-like, with others this is less evident. The difference
between the adjunct and the argument is proved by an example such as (f) where they
co-occur in the same clause (quī [ablative] and anima, respectively).
(a) Praedaque agroque adoriaque affecit popularis suos . . .
(‘He’s enriched his countrymen with booty, land, and fame . . . ’ Pl. Am. 193)
(b) Quin una omnes / peste hac populum hunc liberant?
(‘Why don’t they all free the people from this pest together?’ Pl. Ps. 204–4a)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Nemo hunc M. Caelium in illo aetatis flore vidit nisi . . . in M. Crassi castis-
suma domo cum artibus honestissumis erudiretur.
(‘No one ever saw this young Marcus Caelius while he was in that early youth . . .
except in the irreproachable household of Marcus Crassus while he was being
trained in the most honourable pursuits.’ Cic. Cael. 9)
(d) . . . me complevi flore Liberi . . .
(‘I’ve filled myself with the bouquet of Bacchus . . . ’ Pl. Cist. 127)
(e) De eo vino cyathum sumito et misceto aqua et bibito ante cenam.
(‘Take a cyathus of this wine, dilute it with water, and drink it before dinner.’ Cato
Agr. 114.2)
(f) . . . date ferrum qui me anima privem.
(‘ . . . give me a sword with which I may reave me of life.’ Enn. scen. 197–8V=182J)
(g) . . . Bellovacos, qui belli gloria Gallos omnes Belgasque praestabant . . .
(‘ . . . the Bellovaci, who surpassed all the Gauls and Belgae in renown of war . . . ’ Hirt.
Gal. 8.6.2)
(h) . . . qui ad perniciem innocentis fidem suam et religionem pecunia commutarit?
(‘ . . . who, to cause the ruin of an innocent man, has sold his honour and his oath for
a bribe?’ Cic. Clu. 129)
inbuerat . . . (Cic. Deiot. 28); . . . facile declaratur utrum, is qui dicat, . . . an ad dicen-
dum omnibus ingenuis artibus instructus accesserit. (Cic. de Orat. 1.73);
Verbs of filling: Quarum (sc. fossarum) interiorem . . . aqua ex flumine derivata
complevit. (Caes. Gal. 7.72.3); Quae nisi erunt semper plena, ego te implebo flagris.
(Pl. Cas. 123); Deinde pernam ponito, cutis deosum spectet, sale obruito totam. (Cato
Agr. 162.1); Meministis tum, iudices, corporibus civium Tiberim compleri, cloacas
refarciri . . . (Cic. Sest. 77); . . . ut . . . meam domum refertam viris bonis per amicos
suos complerent proscriptionis metu . . . (Cic. Dom. 55); Utraque autem ex minutis-
simis sunt instruenda, uti materia ex calce et harena crebriter parietes satiati diutius
contineantur. (Vitr. 2.8.2);
Verbs of mixing: Oh, lutum lenonium, / commixtum caeno sterculinum publicum . . .
(Pl. Per. 406–7); Matrum in liberos, virorum in uxores scelera cernitis, crudelitate
mixtas libidines videtis inmanis. (Cic. Scaur. 13); . . . permisceto lentim aceto laserpiciato
et ponito in sole. (Cato Agr. 116);
Verbs of competing: Quos (sc. Remos—subject) quod adaequare (sc. Haeduos) apud
Caesarem gratia intellegebatur . . . (Caes. Gal. 6.12.7); An ad inferiores, qui his
omnibus rebus antecelluntur. (Rhet. Her. 2.48); Simul forma, factis, eloquentia,
dignitate, acrimonia, confidentia pariter praecellebat . . . (Quad. hist. 8); . . . et coniunx
Germanici Agrippina fecunditate ac fama Liviam, uxorem Drusi, praecellebat. (Tac.
Ann. 2.43.6); . . . propterea quod tunc herba ruscida meridianam, quae est aridior,
iucunditate praestat. (Var. R. 2.2.10); . . . qui omnes homines supero [atque] antideo
cruciabilitatibus animi. (Pl. Cist. 205); . . . vicistis cocleam tarditudine. (Pl. Poen. 532);
Verbs of changing: Et tandem superis miserantibus ambo alite mutantur. (Ov. Met.
11.741–2);
With several of these verb classes alternative three-place patterns are found as well (see
§§ 4.54–4.58). Many of these verbs are also found with a two-place frame. When an
instrument or means satellite is added to such a two-place frame, the resulting utterance
contains both an accusative and an ablative (or nominative and ablative in the passive).
This is the case with et bile et pituita in (i). Vomitus is the product of ‘mixing’, and bile
and pituita are the ingredients used for that mixture. An alternative for the ablative is a
prepositional expression with ex (see § 10.55). Ablatival third arguments and ablatival
satellites are not always systematically distinguished in the scholarly literature. This has
even elicited a conjecture such as that found in (j), where the third argument iure civili
is in fact replaced by a satellite in iure civili indicating the domain within which the
training of orators (a two-place pattern) takes place.
(i) Si quis autem incidit vomitus, mixtus esse et bile et pituita debet.
(‘But should any vomiting occur, it should be of bile and phlegm mixed.’ Cels. 2.3.7)
(j) Sed tamen non fugisset hoc Graecos homines, si ita necesse esse arbitrati
essent oratorem ipsum erudire iure civili (in iure civili cj. Manutius), non ei
pragmaticum adiutorem dare.
(‘But after all, had they thought it necessary, this idea of training the orator himself in
the common law, instead of giving him an attorney to help him, would not have
failed to occur to the Greeks.’ Cic. de Orat. 1.253)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In addition to the accusative + ablative pattern, many verbs of supplying with and their
antonyms occur with an alternative accusative + dative pattern. Examples of the two
patterns with the verbs onero and fraudo are (a)–(b) and (c)–(d), respectively. (For
occasional instances—in poetry and Late Latin—of a double accusative, see § 4.75.)
Some of the verbs of depriving are also found with an accusative + genitive pattern (see also
§ 4.55 on verbs of filling). They are also regularly found with an accusative + prepositional
phrase pattern, with the separative prepositions ab, de, and ex. This is illustrated in (e)–(i)
by the verbs levo and libero. (For factors determining the choice between the ablatival
and prepositional expressions in these three-place frames, see § 12.28.)
The ablatives with verbs like libero and verbs like arceo are both called ablativus
separativus. The ‘separation’, however, occurs in opposite directions. In this Syntax
libero is treated as an antonym of e.g. instruo. For two-place verbs and adjectives of
the same meaning class governing the ablative, see § 4.29. Verbs denoting separation
as such are dealt with in § 4.83. Note that in its literal meaning (‘to set free from’)
libero can be regarded as a separation verb, as in (j).
With verbs of filling, the pattern accusative + genitive occurs as well, on the analogy of
adjectives like plenus ‘full’ (see § 4.101). Examples are (a) and (b). There are also
occasional double accusative patterns, as in (c) (further details in § 4.53).80
(a) Erroris ambo ego illos et dementiae / complebo . . .
(‘I’ll fill both of them . . . with misunderstandings and madness . . . ’ Pl. Am. 470–1)
(b) Amphoram defracto collo puram impleto aquae purae, in sole ponito.
(‘Break off the neck of a clean amphora, fill with clear water, and place in the sun.’
Cato Agr. 88.1)
(c) Implete hydrias aquam . . .
(‘Fill the pitchers with water . . . ’ Vet. Lat. Joh. 2.7)
Supplement: . . . ea omnia fecit palam / parasitus, qui me complevit flagiti et
formidinis . . . (Pl. Men. 901–2); Haec res vitae me, soror, saturant . . . (Pl. St. 18);81
With verbs of mixing, the accusative + dative pattern is found as well. It is common in
poetry from Catullus onwards (the first instance being (a)—compare (b) for the
reverse expression in Varro), but it is also used by later prose authors.82 Just like
79
See K.-St.: I.474–5 (‘sogenannter griechischer Genetiv’); Mariani (2004: 34–6); Calboli (2009: 79–81).
80 81
See Svennung (1935: 226ff.). For more instances, see Bennett (1914: II.91–2).
82
See TLL s.v. misceo 1088.73ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
iungo (§ 4.52), misceo has another frame with an associative cum-expression (ex. c)—see
§ 4.69—as well as a two-place frame with two or more coordinated accusative objects
(or passive counterparts), as in the ablative absolute clause mixta . . . aqua in (d).
(a) . . . frigori miscet calorem atque humori aritudinem . . .
(‘ . . . mingles heat with cold, dryness with moisture . . . ’ Enn. var. 46)
(b) . . . humanae quandam gentem stirpis concoquit frigus calore atque umore
aritudinem / miscet (Prometheus) . . .
(‘Prometheus cooks a certain kind of human stock, he mixes cold with heat, dryness
with wet . . . ’ Var. Men. 428)
(c) Id indito in urnam. Misceto cum musto. Id indito in doleum.
(‘Pour it into an urn. Mix with must. Pour it into a jar.’ Cato Agr. 23.3)
(d) Circuli, quod mixta farina et caseo et aqua circuitum aequabiliter
fundebant.
(‘Circuli (sc. are so called) [= ‘rings’] because they poured into the pan a regular
circuitus ‘circuit’ of a batter made of flour, cheese, and water.’ Var. L. 5.106)
With verbs of changing, the entity resulting from the change may also be expressed in
an in + accusative prepositional phrase (abstract direction), as in (a), or in a cum
prepositional phrase (associative) (see § 4.69).84
83
For these instances, see Asensio (2002: 17–18, 22).
84
For the various patterns of the verb muto, see Martín Puente and Santos Guzmán (2002).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
4.59 Factors determining the choice between the accusative + dative case
pattern and the accusative + ablative case pattern
For a number of verbs like dono ‘to present’ and circumdo ‘to surround’ there are
enough examples of both the accusative + dative pattern and the accusative + ablative
pattern to detect a number of factors that seem to be involved in the choice between
the two patterns. Typical examples of the verb dono are (a)–(c).
(a) Hinc ille est anulus aureus quo tu istum in contione donasti.
(‘And hence the gold ring with which you publicly presented that man of yours.’ Cic.
Ver. 3.185)
(b) . . . bona tui Gabini, cui regna omnia Syrorum . . . donaras, consecrasti.
(‘ . . . you consecrated the property of your friend Gabinius, to whom you had made a
present of all the realms of Syria . . . ’ Cic. Dom. 124)
(c) Multa palam domum suam auferebat, plura clam de medio removebat, non
pauca suis adiutoribus large effuseque donabat . . .
(‘He openly carried much away to his own house, secretly removed more, distributed
much with a liberal and lavish hand among those who had helped him . . . ’ Cic.
S. Rosc. 23)
The two patterns, which are not synonymous semantically, seem to allow the same
types of constituent and to be interchangeable in this respect. However, the accusative
+ dative pattern is preferred if the thing given is either a neuter pronoun or adjective
or a clause. The accusative + ablative pattern is preferred if the accusative NP
denotes a human being, and, as such, is more often found in the passive than
the other pattern. For pragmatic factors influencing the choice of the patterns,
see § 5.10. In the case of dono, there is also a marked diachronic change, with the
accusative + dative pattern strongly increasing.85 Another difference is that the
ablative in the accusative + ablative pattern can be omitted less easily than the dative
in the accusative + dative pattern.
In questions of the type quid faciam? ‘what shall I do?’ and quid fiet? ‘what is going to
happen?’ a third constituent can be used in the ablative. This constituent can be
regarded as an instrumental adjunct (see § 10.54, Appendix), but from the context it
85
For statistical data, see Martín Rodríguez (1996b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
becomes clear that it is in some way affected by what is being done to it. In (a) sua is
continued in the next sentence as subject of the passive forms dicetur and abducetur,
and in (b), te is continued as subject of confugies. The whole combination thus gives
the impression of a three-place frame. An alternative expression is the use of a
prepositional phrase with de, as in (c). Quid faciam and quid fiet may also be
combined with a beneficiary adjunct in the dative, as in (d).86
(a) Quid faciet sua? / An ea quoque dicetur huiu’, si una haec dedecori’st parum.
/ # Immo ad tuam matrem abducetur.
(‘What will he do with his own? Will she be called his as well, as if this one alone isn’t
disgrace enough? # No, we’ll take her to your mother.’ Ter. Hau. 333–5)
(b) Si inimicos, quid te futurum est? Quo confugies? Ubi nitere?
(‘If as your enemies, what becomes of you? What refuge will you find, what ground to
stand on?’ Cic. Ver. 2.155)
(c) . . . ex me quaeras quid de istis municipiis et agris futurum putem?
(‘ . . . you ask me what I think is going to happen about these towns and lands!’ Cic.
Fam. 9.17.1)
(d) Quid tu mi facies?
(‘What will you do to me?’ Pl. Cas. 117)
Supplement:
Quid facies ea? / # Intro rumpam recta in aedis. (Pl. Mil. 459–60); Quid illa faciemus
concubina quae domi est? (Pl. Mil. 973); Quid factum est eo? / # Comesum,
expotum; exunctum; elotum in balineis . . . (Pl. Trin. 405–6); Sed quid Tulliola mea
fiet? (Cic. Fam. 14.4.3);
Genitive third arguments with three-place judicial verbs and expressions denote the
matter involved in the procedure, notably the charge (or crime) of which one is
accused (or freed) or the penalty to which one is condemned. With some verbs both
the charge and the penalty patterns are possible (e.g. with the verb condemno ‘to
condemn’). This use of the genitive is often called genetivus criminis or genetivus
forensis. Charge and penalty arguments are discussed in separate sections.
It should be noted that this semantic field is highly idiomatic and that alternative
productive expressions are often available. When indicating the charge, for example,
86
A convincing analysis is Nutting (1927, 1928b—where all Classical and Silver instances can be
found). See also OLD s.v. facio § 22b; s.v. fio § 12a; TLL s.v. facio 103.73ff. See also § 4.86 on verbs of
sacrificing.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
involving adligo
in crime ‘to
implicate’
87
For a survey of the development, see Sz.: 76.
88
For a survey of examples, see Perotti (1987: 379–81).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) Sed, quod solum poteram, tacitus licet serae vindictae gratulabar.
(‘I did the only thing I could: rejoice silently in my revenge, overdue as it
was.’ Apul. Met. 7.26.2)
With some of the verbs involved, especially when the nouns are not necessarily legal
idioms, the ablative is more common, which can be explained by the fact that the
ablative is the regular marker of means and instrument adjuncts. An example is (f).
Prepositional expressions are also widely used, especially with de, as in (g). There
is no alternative for de vi, as in (h). It is not likely that the use of in ‘in’ found in
(i) is synonymous with a genitive expression. In Late Latin, however, in is used that
way, as in (j).89
89
For the various frames of accuso in Late Latin, see Roca (2003).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Appendix 1: From Lucretius onwards with certain verbs (e.g. damno ‘to condemn’),
the dative case and prepositional expressions with ad or in ‘to’ are used, as with three-
place verbs like destino ‘to designate’. Examples are (e)–(g).
Value arguments denote the value of the entity that functions as the object (or
subject in the passive) of verbs of valuing such as aestimo ‘to value’, ‘to assess’ and
verbs of opinion such as habeo, facio (fio), and duco ‘to consider’, as well as a
number of other less frequent verbs such as censeo ‘to assess’, existimo ‘to value’,
iudico ‘to judge of ’, pendo ‘to place a value upon’, as in (a) and (b). The value
constituents are usually in the genitive when the value is expressed in a general way,
such as magni ‘highly’ and parvi ‘of little value’. The genitive is also used idiomat-
ically with nouns denoting a small or insignificant entity such as flocci ‘of the
smallest value’ in (c). Additionally, it is used to mark the value argument with
sum ‘to be worth’ (see § 4.32). On the other hand, when the value is expressed in a
more specific way, the ablative is common, as viginti minis in (d) and levi momento
in (e). However, the ablative is found in general value indications as well, from the
Classical period onwards, especially with the verb aestimo, on the analogy of
adjuncts of price (see § 10.57), as in (f).90 A prepositional phrase with pro can be
used as well, as in (g).
(a) Magni enim aestimabat pecuniam non modo non contra leges, sed etiam
legibus partam.
(‘For he highly valued the money which was obtained not only not against the law,
but even by means of the law.’ Cic. Fin. 2.55)
(b) Nam non est veri simile hominem pauperem / pauxillum parvi facere quin
nummum petat.
(‘It doesn’t look natural if a poor man cares little about money, be it ever so tiny a
sum, and doesn’t ask for it.’ Pl. Aul. 111–12)
(c) Non ego te flocci facio.
(‘I don’t care a straw about you.’ Pl. Truc. 606)
(d) Nam ego te huic dedi aestumatum viginti minis . . .
(‘I’ve given you to him on a bail of twenty minas . . . ’ Pl. Capt. 364)
(e) . . . quorum salutem neque propinqui neglegere neque civitas levi momento
aestimare possit.
(‘For their kindred could not ignore their safety, nor could the state account it of
slight importance.’ Caes. Gal. 7.39.3)
90
The most detailed study is Laing (1920). Torrego (2014a) deals with value expressions with these
verbs as object complements.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apparent genitive forms are shown in (a) and (b). Most common is the combin-
ation of lucri with facio, sometimes written as one word. The genitive is usually
described as a genitive of the rubric (for which see § 11.57) but is different in
several respects. It is also unattractive to regard these expressions as object com-
plements.91 In that case one would expect subject complement counterparts with
the verb sum, but they do not seem to exist. Since these expressions lack the usual
productive characteristics of nouns in the genitive they must be regarded as idioms
outside the case system.
(a) Nam non emisti hanc, verum fecisti lucri.
(‘You didn’t buy her, you made a profit of her.’ Pl. Per. 713)
(b) Verbivelitationem fieri compendi volo.
(‘I want to cut short our word duel.’ Pl. As. 307)
Supplement:
Si nunc me suspendam, meam operam luserim / et praeter operam restim sumpti
fecerim . . . (Pl. Cas. 424–5); Iam vero ab isto omnem illam ex aerario pecuniam . . .
lucri factam videtis. (Cic. Ver. 3.176); Jesus . . . salvam facturus animam pro nomine
eius amissam, perditurus autem de contrario adversus nomen eius lucri habitam.
(Tert. Cor. 11.5);
91
As is the analysis in Burkard and Schauer (2000: 482–3).
92
For a full list of verbs (including two-place verbs), see TLL s.v. cum 1370.20ff. See also von Geisau
(1916).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples are (a)–(e). Some verbs that normally require an associative cum-
prepositional phrase also occur with a dative recipient constituent, as in (f). The
use of the dative increases in poetry and in later Latin. An example of the verb
communico ‘to share’ is (g).
(a) (sociis) . . . qui velint cum periculis nostris sua communicare.
(‘ . . . allies . . . such as are prepared to share their perils with ours.’ Cic. Balb. 24)
(b) Pol me hau paenitet, / si licet boni dimidium mihi dividere cum Iove.
(‘Well, I’m not upset if I can share half of the good with Jupiter.’ Pl. Am. 1124–5)
(c) Ut vestem cum illo mutem . . .
(‘To exchange clothes with him . . . ’ Ter. Eu. 572)
(d) . . . compara labella cum labellis.
(‘ . . . match my little lips to your little lips.’ Pl. As. 668)
(e) Ubi tecum coniunctus siem, / ubi onus nequeam ferre pariter, iaceam ego
asinus in luto . . .
(‘When I’m hitched up with you and can’t carry my burden the same way, I, the
donkey, would lie in the mud.’ Pl. Aul. 229–30)
(f) . . . quae (sc. Sicilia) et ipsa erat mihi sicut domus una coniuncta . . .
(‘Sicily, which was all united with me like one family . . . ’ Cic. Planc. 95)
(g) . . . qui uxores suas amicis communicaverunt . . .
(‘ . . . who shared their wives with their friends . . . ’ Tert. Apol. 39.12)
Supplement (in alphabetical order by verb):
Nonne compensavit cum uno versiculo tot mea volumina laudum suarum? (Cic. Pis. 75);
Veluti pecunia cum pecunia conpensatur, triticum cum tritico, vinum cum vino . . .
(Gaius Inst. 4.66); Cave siris cum filia / mea copulari hanc nec conspicere. (Pl. Epid.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
400–1); . . . mihi tecum ita dispertitum officium fuisse . . . (Cic. Fam. 5.2.1); . . . se cum
Bruto . . . iuncturum vires . . . (Vell. 2.65.1); . . . malumque ut eius cum tuo misceres malo.
(Pl. Trin. 122); . . . et cum his ansis ferreis et plumbo frontes vinctae sint. (Vitr. 2.8.4);
Appendix: Different is the use of the cum-phrase in (h), where it rather marks the
result, in a way comparable with in expressions.93
(h) Denique videbatur ea condicione tam fortis fuisse ut cum patriae caritate
constantiae gloriam commutaret.
(‘In short, there seemed to be a catch in that display of courage—he bartered
away affection for his country in exchange for the credit of steadfastness.’
Cic. Sest. 37)
With verbs like spargo ‘to sprinkle’ and lino ‘to smear’, the object may be either the
material sprinkled on a certain surface, as in (a), or the surface that is sprinkled with a
certain material, as in (b). There is also a two-place frame with the material as the
subject, as in (c). A dative is also found—especially in later periods—instead of the
prepositional expression. An example with the verb lino is (d).
(a) Stercus columbinum spargere oportet in pratum vel in hortum vel in
segetem.
(‘One ought to spread pigeon dung on the meadow, or in the garden, or in the crops.’
Cato. Agr. 36)
(b) . . . pabulum quod dabis amurca spargito.
(‘ . . . sprinkle the feed which you give with amurca.’ Cato Agr. 103)
(c) . . . sparsissent lacrimae pectora nostra piae . . .
(‘ . . . loving tears would have wet my breast . . . ’ Ov. Tr. 3.42)
(d) (medicamenta) . . . tere diligenter et exinde animali . . . linito
(‘Rub the medicine carefully and smear some of it on the animal.’ Pelag. 456)
93
For result expressions with cum, see Martín Puente and Santos Guzmán (2002).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verbs of teaching, asking, etc. governing a so-called double accusative have two
nominal arguments: one is a noun or noun phrase that usually refers to a human
being (the object), while the other usually refers to a non-human entity. An illustra-
tive selection of the verbs involved is given in Table 4.17.
hiding celo
‘to
conceal’
Examples of the double accusative verbs rogo (‘to ask for’) and doceo (‘to teach’) are (a)–(d).
Note in (d) the parallelism of the nominal argument litteras and the prolative infinitive
natare. If these verbs are in the passive, only one of the two accusative constituents—usually
the (human) object—can become subject of the clause. The second one remains in
the accusative case. Examples are (e) and (f ). However, instances such as these, in which
the passive occurs with a third argument in the accusative, are relatively rare. For a passive
instance with a clause as the third argument, see (g).
(a) Roga me viginti minas . . .
(‘Ask me for twenty minas . . . ’ Pl. Ps. 114)
(b) . . . cum illum . . . unus . . . rogaret mortem.
(‘ . . . when a man begged him to be put to death.’ Sen. Ep. 77.18)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Praeceptor tuos, qui te hanc fallaciam / docuit, ut fallaciis hinc mulierem a
me abduceres.
(‘Your instructor, the one who taught you this trick, so that you’d take the woman
away from me with tricks.’ Pl. Ps. 1193–4)
(d) Nepotes et litteras et natare aliaque rudimenta per se plerumque docuit . . .
(‘He taught his grandchildren reading, swimming, and the other elements of educa-
tion, for the most part himself . . . ’ Suet. Aug. 64.3)
(e) Qui utinam omnes ante me sententiam rogarentur . . .
(‘I wish all of them were asked their opinions before me . . . ’ Cic. Phil. 5.5)
(f) . . . potiusque illa docendi erunt minus vulgata.
(‘ . . . and it is more important that they should be taught the less hackneyed
instances.’ Quint. Inst. 1.5.11)94
(g) Rogatus est populus quem id bellum gerere placeret.
(‘The people were asked whom they wished to conduct this war.’ Cic. Phil. 11.18)
When only the thing taught or asked is expressed, this single object constituent has a
passive subject counterpart, as in (h).
(h) . . . Galliae, a qua nos tum cum consulatus petebatur, non rogabatur, petere
consulatum solebamus . . .
(‘ . . . Gaul, the region where we used to canvass for the consulship when that office
was stood for and not asked for . . . ’ Cic. Phil. 2.76)
The number of verbs with which this double accusative construction is statistically
predominant, and which are, moreover, also frequent in absolute numbers, is very
small. Rogo and doceo are the most common in this respect. Double accusative
expressions are relatively frequent in Early Latin. In Late Latin, the expression is
found with some frequency as an alternative for other three-place patterns. In the
Classical period, the second accusative is very often a neuter pronoun or adjective.
The double accusative pattern is extended to verbs with a related meaning in poetical
style, which tends to avoid prepositional expressions.95
Doceo can be regarded as the causative counterpart of disco, which has a second
argument in the accusative.
94
For more examples, see TLL s.v. doceo 1731.54ff.
95
For a historical survey of the verbs involved, see Sz.: 43.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
with the verbs perrogo ‘to ask for in turn’, consulo ‘to consult about’, and percontor ‘to
interrogate’. Examples are (a) and (b).
(a) Quin tu id me rogas?
(‘Why don’t you ask me that question?’ Pl. Bac. 258)
(b) Quod cum interrogatus Socrates esset, respondit sese meruisse ut . . .
(‘When this question was put to Socrates he replied that he had earned the
distinction . . . ’ Cic. de Orat. 1.232)
Supplement (in alphabetical order by verb):
. . . nec te id consulo. (Cic. Att. 7.20.2); Sunt quae te volumus percontari . . . (Pl. Ps.
462); Roga hoc idem Epicurum. (Cic. Tusc. 2.28); . . . te hoc beneficium rogo. (Ant.
Att. 14.13A.3); . . . factum senatus consultum ut duoviros aediles . . . dictator populum
rogaret (Liv. 6.42.14);
Passive: Hannibal nominatim interrogatus sententiam . . . (Liv. 36.7.1); Militem . . .
Graece testimonium interrogatum nisi Latine respondere vetuit. (Suet. Tib. 71);
Quisquis alimenta a mendico rogatus est . . . (Sen. Con. 1.1.10); Mirari soleo, cum
video aliquos tempus petentes et eos qui rogantur facillimos. (Sen. Dial. 10.8.1);
With other verbs of requesting the double accusative is used with varying frequency.
With flagito ‘to ask urgently’ and with oro ‘to beseech’, the number of instances in
which both the person who is asked and the entity asked for are expressed is relatively
small. With these verbs the double accusative and the alternative prepositional
expression seem to be equally frequent. The verbs posco ‘to ask for insistently’ and
postulo ‘to demand’ are frequently found with both entities expressed, but differ in the
use of the double accusative: it is the normal expression with posco (with a prepos-
itional alternative increasing in Late Latin) and the exception with postulo. The use of
the double accusative is extended to peto and a few other two-place verbs. In (very)
Late Latin, verbs of asking are treated as communication verbs with an addressee in
the dative.96
Supplement (in alphabetical order by verb):
Quod deos inmortales inter nuncupanda vota expoposci, eius me conpotem voti vos
facere potestis . . . (Liv. 7.40.5); . . . me frumentum flagitabant. (Cic. Dom. 14); . . .
idque . . . parentis suos liberi orabant . . . (Cic. Ver. 5.119); Ceterum quantum lubet
me poscitote aurum. Ego dabo. ( Pl. Bac. 703); . . . magistratum Sicyonium nummos
poposcit (Cic. Ver. 1.44); Orationes autem me duas postulas. (Cic. Att. 2.7.1—
formerly emended into a me); Aulam auri, inquam, te reposco . . . (Pl. Aul. 763);
Extension: Illud te (<a> te edd.) peto (Dolab. Fam. 9.9.2); NB: Haec eadem, Aenea,
terram, mare, sidera iuro / Latonaeque genus duplex Ianumque bifrontem / vimque
deum infernam et duri sacraria Ditis. (Verg. A. 12.197–9—see Tarrant ad loc.; for the
Graecism iuro + accusative object instead of per, see Mayer (1999: 162–3));
96
See Löfstedt 1942: I.204ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In the frame of rogo the entity that is the source, semantically speaking, is expressed in the
accusative case and behaves like a patient object, since it can become the subject in the
passive. A more transparent way of expressing this is with a source adjunct marked by
the preposition ab ‘from’ (rarely de or ex) + ablative. This alternative, which is occasion-
ally found with rogo itself, is the normal way of expressing the source with a number of
two-place verbs of requesting, such as peto ‘to request’ and quaero ‘to inquire’ and its
compounds, which are by far the most common verbs with this meaning. For two-place
rogo sententiam, with the source expressed, see (c). For the passive (without a source), see
(d). For peto and quaero with a source expression, see (e) and (f ).
(c) Aristophanes vero, cum ab eo sententia rogaretur . . .
(‘Aristophanes, however, when his opinion was asked . . . ’ Vitr. 7.pr.6)
(d) Emotis deinde curia legatis, sententiae interrogari coeptae.
(‘Then after the envoys had been ushered out of the house, opinions began to be
called for.’ Liv. 30.23.1)
(e) Nunc hinc parasitum in Cariam misi meum / petitum argentum a meo
sodali mutuom.
(‘Now I sent my hanger-on off to Caria to ask a friend of mine to lend me the money.’
Pl. Cur. 67–8)
(f) Quor hoc ego ex te quaeram? Aut quor miniter tibi . . .
(‘Why should I ask you about this? Why should I threaten you . . . ’ Pl. As. 47)
Other verbs such as erudio and instituo almost exclusively take the accusative for the
patient receiving the teaching and the ablative for the thing taught, thus behaving as
verbs of supplying with their typical case pattern (see § 4.53). However, an occasional
extension of the double accusative pattern to these and other verbs can be found in
poetry and in Late Latin. Examples are (c) and (d). Conversely, doceo is occasionally
found with the accusative + ablative pattern as well, as in (e).
(c) . . . hortaturque (sc. natum) sequi damnosasque erudit artes . . .
(‘He encourages the boy to follow, instructs him in the fatal art . . . ’ Ov. Met. 8.215)
(d) Deum nos a prophetis et a Christo, non a philosophis nec ab Epicuro
erudimur.
(‘We are educated in God by the prophets and by Christ, not by philosophers, nor by
Epicurus.’ Tert. Marc. 2.16)
(e) Socraten fidibus docuit nobilissimus fidicen.
(‘A very famous lyre player taught Socrates the lyre.’ Cic. Fam. 9.22.3)97
Supplement:
Quid nunc te . . . litteras doceam? (Cic. Pis. 73); Nunc aliam citharam me mea Musa
docet. (Prop. 2.10.10); (Catilina) iuventutem . . . mala facinora edocebat (Sal. Cat. 16.1);
Extension: . . . ignaros instruo verum. (Comm. Instr. 1.1.9);
Passive: Motus doceri gaudet Ionicos / matura virgo . . . (Hor. Carm. 3.6.21); . . . omnis
militiae artes edoctus fuerat. (Liv. 25.37.3); . . . DOCTA·ERODITA·OMNES·ARTES·VIRGO . . . (CIL
I2.1214.2 (Rome, c.65 BC));
Alternative expressions: Object in the ablative: Litteris Graecis Latinis docta . . . (Sal.
Cat. 25.2);
NB: Object in the genitive: destinavi illum artificii (artificium cj. Scheffer) docere . . .
(Petr. 46.7 (Echion speaking)); . . . iuris sui edoctus. (Ulp. dig. 11.1.11.12);98
With doceo, the entity taught can be an accusative and infinitive clause (see § 15.97).
Celo as a two-place verb means either ‘to keep (something) secret’ or ‘to keep
(someone) in ignorance’. In the latter meaning, the thing concealed may be expressed
as a respect adjunct with a prepositional phrase introduced by de ‘about’. The number
of double accusative expressions is small.
(a) Uxorem quoque eampse hanc rem uti celes face.
(‘Make sure that you conceal this business also from your wife herself.’ Pl. Trin. 800)
97
For a historical explanation of the idiomatic use of fidibus, see Ripoll (forthc., 2013). For the verbs
with which it is used, see TLL s.v. fides2, 693.16ff.
98
See TLL. s.v. doceo 1732.74ff. and edoceo 107.80ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The use of volo ‘to want’ with a personal object and a neuter pronoun as the second
object is an idiom that is almost confined to Plautus and Terence, as in (c).
(c) Quid me voltis? # Ire tecum.
(‘What do you want from me? # To go with you.’ Pl. Mer. 868)
Supplement:
Si quid te volam, / ubi eris? (Pl. As. 109–10); Numquid me vis ceterum? (Pl. Epid. 512);
Nisi quid me aliud vis, Philto, respondi tibi. (Pl. Trin. 458); Numquid, Geta, aliud me
vis? # Ut bene sit tibi. (Ter. Ph. 151); Siquid ille se velit . . . (Caes. Gal. 1.34.2);
There are a few examples of verbs of depriving with a double accusative throughout
the history of Latin (with some extension in poetry and Late Latin): eludo ‘to obtain
by trickery from’, expungo ‘to extort’, fraudo ‘to cheat’, defraudo and defrudo ‘to cheat
out of ’, spolio ‘to rob’, despolio ‘to plunder’, and possibly also exuo ‘to take off ’.
The regular pattern with these verbs is accusative + ablative (see § 4.53). Examples
are (d)–(f).100
(d) . . . nisi quid tu porro uxorem defrudaveris?
(‘ . . . unless there’s something you in turn have cheated your wife out of.’ Pl. As. 95)
(e) Princeps signa tulit Tyria Carthagine pubes, / membra levis celsique decus
fraudata superbum / corporis . . .
(‘Foremost in the ranks were the soldiers from Tyrian Carthage. Light of limb were
they and the glory of lofty stature was denied them . . . ’ Sil. 3.231–3)
(f) . . . Jesus . . . despoliatus pristinas sordes . . .
(‘Jesus . . . stripped of his former dirty clothes . . . ’ Tert. Marc. 3.7.6)
Supplement:
. . . si quid aliquem defraudavi, reddo quadruplum. (Vulg. Luc. 19.8); . . . ut mihi dicas
unde illum habeas anulum, / quem parasitus hic te elusit. (Pl. Cur. 629–30); . . .
cum . . . Pudentilla de suo quinquaginta milia nummum populum [<in> populum
edd.] expunxisset . . . (Apul. Apol. 87); . . . et exue Aaron stolam eius et indue Eleazar
filium eius . . . (Vet. Lat. Num. 20.26);
The ‘supplying with’ verbs ‘to give (something) to (someone)’ are regularly found
with the accusative + dative and the accusative + ablative patterns (see § 4.51).
99
A few other patterns are cited in TLL s.v. celo 768.3ff.
100
For these double accusatives, see Adams (2013: 323–5).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The verb condono ‘to give (something) to (someone)’ is not attested with the
accusative + ablative pattern. A few ‘mixed’ double accusative expressions occur
from Early Latin onwards, as in (g). A few other verbs of supplying with are found
as well, as in (h).
(g) Argentum quod habes condonamu’ te.
(‘We’ll let you keep the money you have.’ Ter. Ph. 947)
(h) . . . legis doctoribus vae, quia oneratis homines sarcinas importabiles . . .
(‘Fie upon the teachers of law because you burden people with unbearable loads.’ Vet.
Lat. Luc. 11.46) (NB: double accusative possibly influenced by the Greek original
[çæ Ç f I Łæ ı φορτία δυσβάστακτα]; the Vulgate has acc. + abl.: one-
ratis homines oneribus quae portari non possunt)
Supplement:
Egon pro hoc te nuntio quid [v.l. qui] donem? (Ter. Hec. 849);
Passive: Habeo alia multa quae nunc condonabitur (Ter. Eu. 17);
NB: Ablative and accusative parallel: CN·POMPEIUS SEX· F·IMPERATOR / VIRTUTIS·CAUSSA·
TURMAM / SALLUITANAM·DONAVIT . . . CORNUCULO . . . ET·FAUMENIUM (= frumentum) / DUPLEX
(CIL I2.709.col.4.1–7 (Rome, 89 BC));
Double accusative constructions are common when neuter pronouns or neuter forms
of adjectives are involved (see §§ 4.71–4.75). This preference operates on a larger scale
with a number of verbs and complex expressions of, roughly, the following meanings:
(i) three-place judicial verbs, which normally govern a genitive (see §§ 4.62–4.64), (ii)
verbs of urging or inciting, which are often accompanied by a constituent indicating
the purpose or effect of the summoning act (with nouns and noun phrases the
prepositions ad or in are normal), and (iii) three-place verbs of depriving (see exx.
(d)–(f) in the preceding section). Some of these verbs have a two-place frame in which
either a person or a thing may be the object. The double accusative construction can
be seen as a means of combining the two kinds of semantic relations to some extent.
Examples are (a) and (b).
(a) . . . si id non me accusas . . .
(‘ . . . if you don’t reproach me for this . . . ’ Pl. Trin. 96)
(b) Quod vos vis cogit, id voluntate impetret.
(‘Let her gain of your own free will what the force of law demands of you.’ Ter. Ad. 490)
Supplement (in alphabetical order by verb):
Accusative pronouns: Active: Illud accuso, non te sed illam . . . (Cic. Att. 13.22.5); Id
unum ex iis qui sibi rem aperuisset arguere se paratam esse. (Liv. 26.12.17); . . . ille
civis qui id cogit omnis . . . (Cic. Rep. 1.3); Nihil neque ante hoc tempus neque hoc
ipso turbulentissimo die criminamini Sestium. (Cic. Sest. 77); Aliud quiddam maius
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Particularly in Late Latin101 (but starting as early as the archaizing author Apuleius
and the Vetus Latina versions of the Bible), numerous compounds with circum, ex, in,
ob, and per are occasionally found with two accusatives. This occurs most frequently
with compounds meaning ‘moving an object or person to or away from another
object or person’ or ‘covering or uncovering something or someone with something’,
etc., where two perspectives are possible, with either one of the objects or persons
involved expressed as object. The utterances that we find can be seen as a form of
contamination, which in some cases may be due to ignorance, in other cases more or
less consciously executed, as in (a), and inspired by earlier examples cited in the
sections above. Simple verbs with such meanings are also occasionally used with two
accusatives. Given the existence of a roughly common semantic denominator for
these verbs, it is better not to explain the use of two accusatives as dependent on the
preverbal elements (which in the case of ex is excluded anyway), that is, one should
not analyse them in the same way as one does the verbs in the next section.
(a) . . . sciat omnes homines . . . natura insitos esse sapientiam.
(‘ . . . he knows that all men have wisdom implanted in them by nature.’ Min. Fel. 16.5)
101
For the Mulomedicina Chironis, see Ahlquist (1909: 36–7). For double accusatives with verbs like
aspergo and impleo in the Apicius text, see Milham (1959).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Muros alios novos circuit civitatem. (Anon. Vales. 12.71); Circumsperges eos aquam
purificationis. (Vet. Lat. Num. 8.7—NB: Vulg. has aqua); . . . equi sudorem frontem
[fronte det. edd. vet., fronde cj. Becichemus] curiose exfrico . . . (Apul. Met. 1.2.3);102
. . . impertite sermones non quidem curiosum . . . (Apul. Met. 1.2.6); . . . indue eos
tunicas . . . (Vet. Lat. Exod. 29.8); . . . totum caput cretam Cimoliam . . . imponito.
(Mulom. Chir. 525); . . . nisi qui passionem Christi . . . fuerit obsignatus. (Tert. adv.
Iud. 11.1); . . . hoc est quod obstricti sumus. (August. Serm. 180.14); (pullum) . . .
quem perfundis ius tale. (Apic. 6.8.13);
Greek influence may have played a role, certainly in ecclesiastical texts, in utterances
with cibo (‘to feed’) and poto (‘to make drink’) with a double accusative. These are
found from the Vetus Latina onward.103
102
For a discussion of this and the following example, see Jacquinod (1992) and Keulen ad loc.
103
For a survey of verbs, see Sz.: 45. For examples, most from didactic texts, see Löfstedt (1936:
145–53) and Svennung (1935: 226–31).
104
For these verbs, see Lehmann (1983) and Serbat (1998).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
There is one clear example of object incorporation in Latin,105 where the combination
and fusion of the object and its governing verb itself governs an argument, viz.
animum adverto or animadverto (‘to pay attention’). Synchronically, it is a two-
place compound or complex expression. The object becomes the subject in the
passive. Examples are (a) and (b). Diachronically, the person or object paid attention
to functioned as a second object, in the way we saw above with the compounds
formed with preverbs such as circum. There are a few more combinations with
animum.
(a) Nunc adeo hanc edictionem nisi animum advortetis omnes . . .
(‘Now unless you all pay attention to this edict . . . ’ Pl. Ps. 143)
(b) Cum tria sint coniuncta in origine verborum quae sint animadvertenda . . .
(‘Whereas there are three things combined which must be observed in the origin of
words . . . ’ Var. L. 7.32)
Supplement:
. . . EA·SENATUS / ANIMUM·ADVORTIT . . . (CIL I2.586.3–4 (Tivoli, c.100 BC)); . . .
2
EA·NOS·ANIMUM·NOSTRUM / NON·INDOUCEBAMUS·ITA·FACTA·ESSE . . . (CIL I .586.5–6 (Tivoli,
c.100 BC)); . . . ne hic illam (<ad> illam cj. Gronovius) me animum adiecisse aliqua
sentiat. (Pl. Mer. 334—see Enk ad loc.); . . . id quod animum induxerat . . . (Cic. Att.
7.3.8); . . . animum adverti columellam . . . (Cic. Tusc. 5.65);
105
For object incorporation, see Fugier (1994), with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
There are a few other idiomatic expressions quoted in this context, such as ludos facio
‘to make a fool of ’ in (c), cf. ludifico(r),106 and indicium facio ‘to give away a secret’
and manum inicio ‘to seize’, which govern an accusative object constituent, as is
sometimes evident from passivization, as in (d). Ex. (e) belongs here as well.107
(c) Ut scelestus, Hegio, nunc iste <te> ludos facit!
(‘How this thug is now making a fool of you, Hegio!’ Pl. Capt. 579)
(d) . . . med / . . . bis ludos . . . factum / . . . esse . . .
(‘ . . . I was fooled twice . . . ’ Pl. Bac. 1100–1)
(e) Sementim facito ocinum, viciam, f[o]enum Graecum . . .
(‘Sow clover, vetch, fenugreek . . . ’ Cato Agr. 27)
Supplement (in alphabetical order by verb):
Erum in opsidione linquet, inimicum animos auxerit. (Pl. As. 280); Rogasne, im-
probe, etiam qui ludos facis me? (Pl. Am. 571); Video hercle ego te me arbitrari,
Euclio, hominem idoneum / quem . . . ludos facias . . . (Pl. Aul. 252–3); Id anus mi
indicium fecit. (Ter. Ad. 617); Ita [<de> add. Seyffert] me uxor forma et factis faciat,
si taceam, indicium. (Caecil. com. 144); Ubi quadrupulator quempiam iniexit
manum . . . (Pl. Per. 70);
NB: ludos modified by an adjective: . . . numquam . . . quisquam me lenonem dixerit, /
si te non ludos pessumos dimisero. (Pl. Rud. 791);
Appendix: In Late Latin there are a number of expressions, such as gratias ago
‘to thank’ and mentionem facio ‘to mention’, with object constituents in the
accusative.108
106 107
For ludos facio, see Baños (2012). J. N. Adams (p.c.).
108
See Sz.: 46 for a few late instances. For incorporations in prehistoric times as in vendo (< venum do),
see Flobert (1996). For a possible instance in Palladius 3.30, see Ortoleva (2012).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Position arguments are found with the following verbs in this class: (me) contineo ‘to
keep (oneself) in a certain state/position’, contineor ‘to be kept in a certain state/
position’, (me) teneo ‘to hold (oneself) in a certain state/position’, teneor ‘to be held in
a certain state/position’, attineo ‘to hold (in a place)’ (Tac.).
Position arguments are also common with verbs indicating ‘putting an entity in a
certain place’. These verbs imply ‘movement’ and can for that reason also be com-
bined with a direction argument. Conceptually, these options are different. The
prepositions in and sub govern the ablative in the first combination, the accusative in
direction arguments. Which combination is preferred varies (in time, between authors,
for non-literal and literal use), but in some cases there is no real choice (e.g. in sole
ponere ‘to place in the sun’). There is also some variation in manuscript readings and
there are emendations, as in (j) below. For the verbs involved, see Table 4.18.
109
TLL s.v. contineo 703.23ff. interprets the verb in instances like (e) as meaning cingo ‘to surround’.
Castris would probably be regarded as instrumental.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of the verbs in the first row of Table 4.18 are (g)–(j).
(g) Postquam avem aspexit in templo Anchisa, / sacra in mensa Penatium
ordine ponuntur.
(‘After Anchises had seen a bird within the range of view, hallowed offerings were set
in a row on the table of the Household Gods.’ Naev. poet. 3.2)
(h) Quae cum in sideribus inesse videamus, non possumus ea ipsa non in
deorum numero reponere.
(‘And when we observe these attributes in the planets, we cannot fail to enrol even
them among the number of the gods.’ Cic. N.D. 2.54)
(i) . . . Curio . . . te . . . in matrimonio stabili et certo conlocavit.
(‘ . . . Curio established you in an enduring and stable wedlock.’ Cic. Phil. 2.44)
(j) Hodierno die primum, patres conscripti, longo intervallo in possessione[m]
libertatis pedem ponimus.
(‘Today for the first time, Conscript Fathers, after a long interval we set our feet in
possession of liberty.’ Cic. Phil. 3.28)
Examples of direction arguments with the same class of verbs are (k)–(m). The
variation is normally dealt with in a section on case variation with certain preposi-
tions (see § 12.26).110
(k) Stipes erat, quem, cum partus enixa iaceret / Thestias, in flammam triplices
posuere sorores . . .
(‘There was a billet of wood which, when the daughter of Thestius lay in childbirth,
the three sisters threw into the fire . . . ’ Ov. Met. 8.451–2)
(l) . . . cum vellet sororis suae filiam in matrimonium conlocare . . .
(‘ . . . who was desirous of arranging a marriage for her sister’s daughter . . . ’ Cic. Div.
1.104)
(m) . . . multas praeterea beluas in deorum numerum reponemus.
(‘ . . . we have to admit to this list of gods many beasts besides.’ Cic. N.D. 3.47)
In each of the instances above, the same verb by its meaning allows two different
options for its obligatory third argument. These must not be confused with cases
such as (n) and (o), where the position expression is a non-obligatory satellite.111
110
K.-St.: I.588–95 use the term ‘Prägnante Konstruktion der Präpositionen’.
111
Examples in K.-St.: I.593.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of the verbs in the other two rows of Table 4.18 are (p)–(s). Note the
variation in (r) and (s).
(p) . . . cum in omnis bonos ea quae ille in malis numerat cadere possint?
(‘ . . . since the things which he reckons among bad things can come upon all good
men?’ Cic. Tusc. 5.25)
(q) Ipsaque illa, quae in commentarium meum rettuli . . .
(‘And those very commonplaces, which I have set down in my notebook . . . ’ Cic. de
Orat. 1.208)
(r) Quaeque in animis inprimuntur, de quibus ante dixi, inchoatae sententiae
similiter in omnibus inprimuntur . . .
(‘And those rudimentary beginnings to which I have referred, which are imprinted
on our minds, are imprinted on all minds alike . . . ’ Cic. Leg. 1.30)
(s) . . . cum visa in animos inprimantur . . .
(‘ . . . when presentations are impressed on the mind . . . ’ Cic. Luc. 58)
To illustrate the range of constructions with verbs that do not indicate motion
explicitly but do imply it, a few examples are given below of the verb figo ‘to fasten’.
In example (t), the place where the table is to be put is expressed by an autonomous
relative clause introduced by ubi ‘where’. In (u), this is expressed in the ablative. In
(v), no specific place where the tabulae were put is mentioned, although the general
location is indicated by the satellite toto Capitolio (the tabulae are definitely not
attached to the Capitolium but are located on it). Ex. (w) demonstrates the use of a
direction argument expressed by a prepositional phrase. Just as often poets prefer a
dative expression to a prepositional one, as in (x).
(x) At lacrimans exclusus amator limina saepe / floribus et sertis operit postisque
superbos / unguit amaracino et foribus miser oscula figit.
(‘But the lover shut out, weeping, often covers the threshold with flowers and
wreaths, anoints the proud doorposts with oil of marjoram and presses his lovesick
kisses upon the door.’ Lucr. 4.1177–80)
Source arguments are found with three-place verbs designating ‘motion away from’
or ‘separation from’. Three semantic classes of verbs can be distinguished: (i) ‘moving
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
away from’; (ii) ‘keeping away from’; and (iii) compounds meaning ‘moving away
from’ on the basis of the meaning of their preverbs. A number of the verbs involved
are given in Table 4.19.
compounds ab-, de-, claudo duco fero iacio rapio terreo verto
with ex- ‘to shut’ ‘to lead’ ‘to bear’ ‘to throw’ ‘to ‘to ‘to
separative snatch’ terrify’ turn’
preverbs
Examples of verbs of moving away from are (a)–(f). The source arguments are
usually expressed by means of a prepositional phrase, as in (a), but bare ablatives do
occur in Classical prose (though sometimes emended), as in (b), and they are
preferred in poetry. They are freely used in prose from Livy onwards (the so-called
separative use of the ablative).112 With geographical names, the bare case form is
usually preferred (see § 10.17), as in (c). Furthermore, it is quite common with these
verbs when they are used in a non-literal sense and the source argument is more
abstract, as in (d) and (e). However, in general there is much individual variation.
Adverbs are used as well, as in (f). For the various classes of constituents found as
source adjuncts (bare case forms, prepositional phrases, and adverbs) see § 10.13.
(a) . . . oportet . . . / . . . hunc telo suo sibi, malitia, a foribus pellere.
(‘ . . . I should drive him away from the door with his own weapon, malice.’ Pl. Am. 268–9)
(b) . . . hunc P. Varium, fortissimum atque optimum civem, iudicem nostrum,
pellere possessionibus armis castrisque conatus est . . .
(‘ . . . he attempted by arms and soldiers to drive from his lands . . . that excellent and
gallant citizen Publius Varus, who sits upon our bench . . . ’ Cic. Mil. 74)
(c) . . . non prius exercitum Apollonia Dyrrachioque movisti quam de Antoni
fuga audisti.
(‘ . . . you did not move your army from Apollonia and Dyrrachium before you heard
of Antony’s rout . . . ’ Cic. ad Brut. 1.2.2)
112
For Pliny and a few other instances in Cicero, see Pinkster (2005b: 245).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of compound verbs with separative preverbs are (i) and (j).
(i) . . . Bibulus, qui dum unus hostis in Syria fuit pedem porta non plus extulit
quam <consul> domo sua . . .
(‘ . . . Bibulus, who so long as there was a single enemy in Syria did not stir a step from
the city gates any more than from his house when he was Consul . . . ’ Cic. Att. 6.8.5)
(j) . . . proinde uti / propere suis de finibus exercitus deducerent.
(‘ . . . so our men had better remove their troops from their territory quickly.’ Pl. Am.
214–15)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Bare ablative: Venit medio vi pontus et undis / Hesperium Siculo latus abscidit . . .
(Verg. A. 3.417–18); Qui natam possis conplexu avellere matris / conplexu matris
retinentem avellere natam . . . (Catul. 62.21–2); . . . quibus ille solebat / cantando
rigidas deducere montibus ornos. (Verg. Ecl. 6.70–1); . . . patefactis consiliis exclusus
Capua . . . (Caes. Civ. 3.21.5);
Prepositional phrase: . . . num etiam de matris hunc complexu . . . avellet atque ab-
strahet? (Cic. Font. 46); . . . C. Sulpicium praetorem, fortem virum misi, qui ex
aedibus Cethegi si quid telorum esset efferret. (Cic. Catil. 3.8); At placuero huic
Erotio, / quae me non excludet ab se . . . (Pl. Men. 670–1);
In the case of a number of these compound verbs, the meaning is such that the
‘transfer’ or ‘giving’ frame together with its case pattern {accusative + dative} is
relevant as well (see § 4.52).
With the verbs absum ‘to be (a specified distance) away’ and disto ‘to be distant (by an
amount)’,113 and also with the verb sum in combination with a source argument, an
expression indicating the distance between the two entities involved is required.
Examples are (a)–(d). In (a) the extent of space argument is in the accusative, in (b)
in the ablative. In (c) the place from which the river Anien is distant has to be inferred
from the context. Instead of source expressions, a prepositional expression with inter
is common as well, as in (d). An impersonal variant is shown in (e), with the two
entities expressed as source and direction. A few other verbs indicating relative
position are given in the Supplement.114
(a) . . . Teani Apuli, quod abest ab Larino XVIII milia passuum . . .
(‘ . . . at Teanum in Apulia, which is eighteen miles from Larinum . . . ’ Cic. Clu. 27)
(b) (sc. Epidaurus) . . . Aesculapi nobili templo, quod quinque milibus passuum
ab urbe distans nunc vestigiis revolsorum donorum, tum donis dives erat . . .
(‘(Epidaurus noted) for the famous temple of Aesculapius, which, at a distance of five
miles from the city, is now rich in the traces of gifts of which it has been robbed, but
then was rich in the gifts . . . ’ Liv. 45.28.3)
(c) Non minus XV milia Anien abest.
(‘The Anien is no less than fifteen miles away.’ Cato orat. 74)
(d) . . . turres toto opere circumdedit quae pedes LXXX inter se distarent.
(‘ . . . all around the work he set turrets which stood at a distance of eighty feet from
one another.’ Caes. Gal. 7.72.4)
113
These translations are taken from the OLD.
114
For a discussion of extent of space expressions with these verbs, see Torrego (2008) and Cabrillana
(2014: 223–5).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) Quam longe est hinc in saltum vestrum Gallicanum? Naevi, te rogo. # DCC
milia passuum. # Optime.
(‘How far is it from here to your pastures in Gaul? I ask you, Naevius. # Seven
hundred miles. # Quite right.’ Cic. Quinct. 79)
Formally these expressions are similar to extent of space adjuncts (see § 10.18). Extent
of space expressions are also used in position in space adjuncts (see § 10.3, Appendix).
It is not always easy to decide what exactly the function of an extent of space
expression in its clause is. Extent of space arguments may be in the accusative or
in the ablative. With absum and disto Caesar and Cicero use the accusative.
The ablative becomes normal from Livy onwards. The adverbs longe and procul
‘far’ are also common.
Supplement:
Accusative: . . . cum prima legio in castra venisset reliquaeque legiones magnum
spatium abessent . . . (Caes. Gal. 2.17.2); Is locus aberat <a> novis Pompei castris
circiter passus quingentos. (Gal. Civ. 3.67.1); Prima acies hastati erant, manipuli
quindecim, distantes inter se modicum spatium. (Liv. 8.8.5); Sed ubi iam haud plus
quingentos passus acies inter sese aberant . . . (Liv. 28.14.13);
Ablative: Hybanda, quondam insula Ioniae, ducentis nunc a mari abest stadiis. (Plin.
Nat. 2.204); (sc. villa) Decem septem milibus passuum ab urbe secessit . . . (Plin. Ep.
2.17.2);
Impersonal: Inde est ferme mille passum. (Cato hist. 26); . . . ad portam mille a porta
est. (Lucil. 124M=125K);115 NB: Habebat autem de eo loco ad montem Dei forsitan
quattuor milia . . . (Pereg.1.2—see § 4.16);
Cf.: personal: Polam (v.l. ad Polam) ab Ancona traiectus CXX milia p. est. (Plin. Nat.
3.129);
115
This is a quotation of Lucilius in Gellius 1.16.2, where Gellius explains that mille is singular, because
otherwise it would have been ‘mille sunt’. But in reality mille is not the subject of est but a distance
expression in the accusative.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
For adverbs functioning as extent of space arguments, see (h) and (i).
(h) Quam longe a med abest? # Lumen hoc vide.
(‘How far is he away from me? # Look at this light.’ Pl. Cur. 117)
(i) Procul enim aberam ab re ipsa et a locis.
(‘For I am so far from the circumstances and places.’ Cic. Fam. 3.5.4)
A(b) expressions with absum are dealt with as arguments and not as modifiers of
longe and procul. See § 4.44. Longe and procul can be used in various constellations.
Appendix: The ablatives intervallo ‘distance’ (also in its temporal meaning ‘interval’)
and spatio ‘gap’ (also in its temporal meaning ‘duration’) are often dealt with in this
context, but they can better be taken as means or manner adjuncts. An example is (j).
The same analysis seems preferable for (k).
116
For further examples of intervallo + genitive, see TLL s.v. 2295.15ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
latter is less common in Early Latin. Note that in (e) the patient senatus is the subject
of passive cogitur. For further details, see § 15.70.
(a) Quod vos vis cogit id voluntate impetret.
(‘Let her gain of your own free will what the force of the law demands.’ Ter. Ad. 490)
(b) Ita est, sed tu istud petisti, ego hoc cogor.
(‘True; but you sought your eminence, while mine is thrust upon me.’ Cic. Rab. Post. 17)
(c) Ad lacrumas coegi hominem castigando maleque dictis . . .
(‘I brought the chap to tears with my scolding and my harsh words . . . ’ Pl. Bac. 981–2)
(d) Non med istanc cogere aequom est meam esse matrem, si nevolt.
(‘It wouldn’t be fair of me to force her to be my mother if she doesn’t want to be.’ Pl.
Epid. 586)
(e) Senatus cum temporibus rei publicae cogitur ut decernat ut alterae decumae
exigantur, ita decernit ut . . .
(‘When the Senate is forced by the circumstances of the state to order that an
additional tithe be exacted, it decrees in the following way . . . ’ Cic. Ver. 3.42)
nulli mortalium invideo, audi quid promittam et quanti quaeque aestimem. (Sen.
Dial. 7.24.5);
With an acc.: Quotiens hoc tibi, verbero, ego interdixi, / meam ne sic volgo pollicitere
operam? (Pl. Mil. 1056–7); Interdixit histrionibus scaenam (Suet. Domit. 7.1); . . .
philosophi . . . senatusconsulto eiecti atque urbe et Italia interdicti sunt. (Gel.
15.11.4—NB: passive); Sed quae est invidia aut quid mihi nunc invideri potest?
(Cic. Fam. 9.16.5);
With ab: (Valerius Horatiusque) decemviris quoque ab ira et impetu multitudinis
praecavere iubentur. (Liv. 3.53.1);
With de: . . . P. Dolabella praetor interdixit, ut est consuetudo, ‘de vi hominibus
armatis’ . . . (Cic. Caec. 23);
117
In K.-St.: I.384–5 and Sz.: 121 the ablative is taken as an argument that is equivalent to an
accusative.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Cn. Cornelio et Q. Petillio consulibus . . . immolantibus Iovi singulis bubus uti solet . . .
(Liv. 41.14.7); Quoius quotiens sepulcrum vides, sacruficas / ilico Orco hostiis . . . (Pl.
Epid. 175–6); . . . tum (sc. Metello) venienti ture quasi deo supplicabatur. (Sal. Hist. 2.70.3);
Appendix: A rare combination of a dative and a genitive can be seen in (g).
Non-referential obligatory third arguments are found with various classes of verbs that
govern an object argument in the accusative case. The same frame is found with the verb
utor in its meaning ‘to find’. The object is usually a nominal expression (a noun, noun
phrase, or pronoun), but with certain classes of verbs it may be a clause as well. The third
argument (the object complement) can be filled by a noun (phrase) or adjective, showing
agreement in number, gender, and case with the object, but just as with subject
complements, prepositional phrases, adverbs, and nouns or noun phrases not showing
agreement are found as well (see also § 9.39).118 The various types are illustrated for the
verb habeo by (a)–(e).119 In (a), flos (cenae) agrees with the subject of the passive
sentence is. Ex. (b) has the prepositional phrase pro castris, (c) the adverb sic, and (d)
the noun odio in the dative (a so-called predicative dative—see § 9.34). Ex. (e) has a
prepositional phrase as the complement and an accusative and infinitive clause (between
curly brackets) as the object. Further examples of such clauses can be found in § 15.101.
(a) Is nunc flos cenae habetur inter istos . . .
(‘This is now held to be the flower of the feast among those . . . ’ Fav. orat. 1)
(b) Aram habete hanc / vobis pro castris . . .
(‘Have this altar as your camp . . . ’ Pl. Rud. 691–2)
(c) Atque hoc quidem omnes mortales sic habent . . .
(‘And indeed, all mortals regard this in the following way . . . ’ Cic. N.D. 3.86)
(d) . . . quod viro esse odio videas, / tute tibi odio habeas.
(‘ . . . you yourself would hate what you can see your husband hates.’ Pl. Men. 111–11a)
(e) . . . antea semper pro indignissimo habuerant {se patrio regno tutoris fraude
pulsos} . . .
(‘ . . . they had always considered it the greatest outrage that they had been ousted
from their father’s kingship by the crime of their guardian . . . ’ Liv. 1.40.2)
118
For prepositional expressions and ablative noun phrases, see Bulhart (1955).
119
For more examples, see TLL s.v. habeo 2444.19ff., 2458.83ff. and OLD s.v. habeo § 24.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In these examples habeo is clearly a three-place verb. This becomes evident if in (a)
one deletes either the subject or the subject complement. This results in (f) and (g).
The remaining utterances may be acceptable, but habeo would not have the same
meaning as in the original text.
(f) Nunc flos cenae habetur inter istos . . .
(g) Is nunc habetur inter istos . . .
Habeo occurs also in a two-place frame, in its meanings ‘to have’, ‘to possess’, and ‘to
keep’, as in (h). A clause with habeo in this meaning can be expanded with a
secondary predicate. An example is (i), with servas as secondary predicate: if it is
left out, the remaining utterance is grammatically correct, and habeo has the same
meaning as in the original sentence.
(h) Ain’, excetra, tu, quae tibi amicos tot habes tam probe oleo onustos?
(‘Do you say so, snake? Even though you have so many boyfriends so well laden with
oil . . . ’ Pl. Ps. 218)
(i) Et praedicabo . . . / quoque modo <vos> huius filias apud vos habeatis servas . . .
(‘And I’ll tell you . . . in what way you keep his daughters with you as slaves . . . ’ Pl.
Poen. 1245–6)
Whereas in the examples of habeo so far the two frames correspond to two clearly
different meanings, this issue is more complicated in instances like (j). As the
translation suggests, habeo does not really mean ‘to have’, ‘to possess’, it rather
functions more or less as a transitive copula (compare Gallia Antonio inimicissima
est).120 See also § 7.37 on habeo + the perfect passive participle.
(j) (sc. Antonius) Habet inimicissimam Galliam, eos etiam quibus confidebat
alienissimos, Transpadanos.
(‘Gaul is bitterly hostile to him; even the Transpadanes, on whom he relied,
are thoroughly estranged.’ Cic. Phil. 10.10)
Just as with copular verbs (see § 4.97), two types of verbs can be distinguished: verbs
implying no change of the properties of the object (non-dynamic) and verbs which do
imply a change (dynamic). Examples of the first type are (a) and (j) with habeo.
Examples of the second type are reddo ‘to bring into a certain condition’, as in (k), and
creo, as in (l).
(k) . . . nisi ipsos caecos redderet cupiditas . . .
(‘ . . . if greed did not make them blind . . . ’ Cic. S. Rosc. 101)
120
So already Thielmann (1885b: 383) on invisum habeo aliquem ‘I hate somebody’. See also Jacob
(1995) and Wehr (2012). For the copular characteristics of habeo, see Benveniste (1960). For more
examples, see TLL s.v. habeo 2424.42ff. For such instances in the Mulom. Chir., see Adams (2013: 628–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
121
A large number of instances can be found in Müller (1908: 155–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of object complements with verbs of giving, having, taking are (f)–(h).
Some of these verbs are used exclusively or more frequently with nouns and noun
phrases as the object complement, others (also) with adjectives. Other expressions are
sometimes found as complements as well.
(j) . . . domum eius exornatam atque instructam fere iam iste reddiderat
nudam atque inanem.
(‘ . . . he had by now reduced his well-fitted and furnished home to an almost bare and
empty state.’ Cic. Ver. 2.84)
(k) Sunt autem et aliis pluribus locis et fontes et flumina <et> lacus, qui per
salifodinas percurrentes necessario salsi perficiuntur.
(‘In many other places there are springs and rivers and lakes which run through salt
mines and necessarily are made salty.’ Vitr. 8.3.7)
Supplement (in alphabetical order by verb):
Quod eveniet, si eum benivolum, attentum, docilem confecerit. (Cic. Inv. 1.20); Ibo ad
uxorem intro, missum facio Teresiam senem. (Pl. Am. 1145); Et isti me heredem
fecit. (Pl. Cur. 639); Quodsi talis nos natura genuisset . . . (Cic. Tusc. 3.2); Senem
illum tibi dedo ulteriorem, lepide ut lenitum reddas (Pl. Bac. 1150); Multos iam
lucrum lutulentos homines reddidit. (Pl. Capt. 326); Ego vos salvas sistam, ne timete.
(Pl. Rud. 1049);
Some of the verbs in this class also have a two-place frame with the resulting state
expressed as a clause with ut, so for example facio and efficio. Especially in Early Latin
we find utterances that look like mixed expressions, with an object in the main clause
that functions also as the subject in the subordinate clause, like cenam in (l). These
objects are treated as ‘pseudo-objects’ in this Syntax (see § 9.14).
(l) . . . facito cenam mihi ut ebria sit.
(‘ . . . make sure that my dinner is sumptuous.’ Pl. Cas. 746)
Examples of object complements with verbs of calling and entitling are (r)–(t).
(r) Non patrem ego te nominem, ubi tu tuam me appelles filiam?
(‘Shouldn’t I address you as father when you call me daughter?’ Pl. Epid. 588)
(s) Mercurius, Iovis qui nuntius perhibetur . . .
(‘Mercury, who is said to be Jupiter’s messenger . . . ’ Pl. St. 274)
(t) Alter hoc Athenis nemo doctior dici potest.
(‘No second person in Athens can be said to be cleverer than him.’ Pl. Mos. 1072)122
Supplement:
Cognoscite nunc ita reum citatum esse illum ut . . . (Cic. Clu. 49); Rite vero te, Cyre,
beatum ferunt . . . (Cic. Sen. 59); . . . a Biante esse dictum crederet, qui sapiens habitus
esset unus e septem. (Cic. Amic. 59); Se miserum praedicat. (Pl. Men. 909);
Demetrium . . . qui Phalereus vocitatus est . . . (Cic. Rab. Post. 23);
122
For dico, see Ramos (2007a: 167–8). For facio, see Bodelot (2012a)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of an object complement with verbs meaning ‘showing oneself as’, ‘behav-
ing as’ that govern reflexive pronouns as the object are (w) and (x). The expressions
involved are very heterogeneous. Some of the verbs have a complement only if the
object is a reflexive pronoun, others have a frame with an object and an object
complement, but with a different meaning, as is shown by (w) and (x). Praebeo
appears in (w) in its meaning ‘to show oneself ’, while in (x) it means ‘to render’.
(w) . . . servom haud inliberalem praebes te . . .
(‘ . . . you are behaving as a slave who does not lack the properties of a free man . . . ’
Ter. Ad. 886)
(x) . . . ad praebendam opportunam consilio urbem.
(‘ . . . to make the city ready for the accomplishment of their design.’ Liv. 8.25.9)
Supplement (in alphabetical order by verb):
. . . non dubitaverim me gravissimis tempestatibus ac paene fulminibus ipsis obvium
ferre . . . (Cic. Rep. 1.7); . . . qui se intenderant adversarios in eius tribunatum . . . (Cael.
Fam. 8.4.2); . . . ut nemo tam ferus fuerit, quin eius casum lacrumarit inimicumque iis
se ostenderit . . . (Nep. Alc. 6.4); Productus est ab eo Cn. Pompeius, qui se non solum
auctorem meae salutis, sed etiam supplicem populo Romano <praebuit>. (Cic. Sest.
107); Se ad eam rem profitetur adiutorem. (Caes. Gal. 5.38.4);
Weather verbs are the most conspicuous examples, such as pluit ‘it is raining’, rorat ‘it
drizzles’, ‘the dew is falling’, tonat ‘it is thundering’. An example is given in (a).
Similar verbs of natural conditions are illucescit ‘it dawns’ (more often dies illucescit),
lucet ‘it is light’, and vesperascit ‘night is falling’. Other verbs may be used in a similar
way in one of their meanings, so for example adflat ‘it is blowing’, lapidat ‘it is raining
stones’, siccat ‘it is dry’, vibrat ‘it vibrates’. If these verbs are governed by a phasal verb
like desino ‘to stop’ or incipio ‘to begin’, the phasal verb is in the third person singular,
as in (b). The prototypical weather verbs are sometimes found with an explicit subject,
often so in a non-literal sense and in poetically inspired texts, as in (c). The substance
is occasionally expressed as an object in the accusative case, as in (d). An ablative
adjunct is more common, as in (e), and with verbs of flowing such as fluo ‘to flow’ (see
§ 4.10, Appendix). Impersonal passive forms of these verbs, as in (f), are very rare
(see also § 5.21).
(a) Tam hoc quidem tibi in proclivi quam imber est quando pluit.
(‘That requires as little effort as water when it’s raining.’ Pl. Capt. 336)
(b) Cum pluere incipiet . . .
(‘When the rains will begin . . . ’ Cato Agr. 155.1)
(c) . . . pluit . . . nimbus . . . teretis mali . . .
(‘ . . . a shower of shapely apples rains down . . . ’ Col. 10.364–5)
(d) . . . si roravit quantulumcumque imbris aut si adflavit.
(‘ . . . if the smallest sprinkle of rain has fallen or if it has been blowing.’ Plin. Nat.
17.74—NB: imbris cj. Mayhofer)
(e) Reate imbri lapidavit.
(‘At Reate there was a rain of stones.’ Liv. 43.13.4)
(f) Ubi nubilabitur et noctu sub tecto ponito. Cotidie, cum sol erit, in sole ponito.
(‘In cloudy weather or at night put it under cover, but expose it to the sun every day
when there is sunshine.’ Cato Agr. 88.2)
Gerund: . . . omnem aquam oportere arceri quae pluendo crevisset (Cic. Top. 38);
Present participle in an ablative absolute: Illinc flante ne arato . . . (Plin. Nat.
18.328—NB: flante <vento> add. Detlefsen with F2); Episcopus autem albescente
vadet semper, ut missa fiat matutina . . . (Pereg. 44.3);
Intransitive and Impersonal: . . . ubicumque obscuraverit . . . (Vulg. Sirach. 36.28);
Personal: . . . aere atque argento sternunt iter omne viarum / largifica stipe ditantes
ninguntque rosarum / floribus . . . (Lucr. 2.626–7);
Subject explicit: (with a demonstrative pronoun): Lucescit hoc iam. (Pl. Am. 543);123
Luciscit hoc iam. (Ter. Hau. 410); Nam hoc quidem edepol hau multo post luce
lucebit. (Pl. Cur. 182);
(with Iuppiter:) . . . Iove tonante . . . (Cic. Phil. 5.8);
(with other types of subject:) Prodinunt. Famuli tum candida lumina lucent (Enn.
Ann. 156V=148S—NB: the interpretation of the text is uncertain);124 . . . indignatio
(sc. domini) ningat ei dolores . . . (Ambr. Job II.1.2); Sanguinem (sanguine cj. Marsus)
pluisse senatui nuntiatum est . . . (Cic. Div. 2.58); Lapides pluere et fulmina iaci de
caelo . . . portenta esse putatis. (Liv. 28.27.16); L. autem Paulo C. Marcello cos. lana pluit
circa castellum Carissanum . . . (Plin. Nat. 2.147); . . . stridentia funda saxa pluunt . . . (Stat.
Theb. 8.416–17); . . . ut vesperascente caelo Thebas possent pervenire . . . (Nep. Pel. 2.5—
NB: ablative absolute);
Object explicit: In area Volcani et Concordiae sanguinem (v.l. sanguine) pluit. (Liv.
40.19.2—NB: see Briscoe ad loc. who reads sanguine);
Ablative satellite: Inter alia prodigia et carne pluit, quem imbrem ingens numerus
avium intervolitando rapuisse fertur. (Liv. 3.10.6); . . . lacte et sanguine pluisse . . . et
saepe alias, sicut carne . . . item ferro . . . (Plin. Nat. 2.147); Effigies quo pluit ferri
spongearum similis fuit. (Plin. Nat. 2.147);
From Antiquity onwards scholars have assumed some sort of implied subject with
impersonal expressions like pluit, for example Iuppiter. Flobert (1975: 557–9) sug-
gests pluvia ‘rain’ as an underlying subject. For weather verbs in Latin, see also de
Carvalho (1991a). The religious aspects of these expressions are discussed in Le
Bourdelles (1966). General discussion in von Seefranz-Montag (1983) and Eriksen,
Kitillä, and Kolehmainen (2010).
Weather verbs have been the subject of intense debate between linguists of various
theoretical models. See Ruwet (1989) for discussion. A thorough analysis of imper-
sonal verbs, including weather verbs, in German is Corrodi (1925). In the case of
German regnen he makes a threefold distinction (1925: 28–9): (i) es regnet as a mere
event, where regnen applies to the ‘situation’; (ii) die Wolken regnen, etc. with
reference to some cause; (iii) Blut regnet, with reference to some material coming
down in the form of rain. The latter two usages are under Latin and/or biblical
influence. It makes no sense to derive the first type from the other ones.
Two-place use of pluo is first found in Christian authors and in translations of the
Bible, as in (g). Remarkable is the combination with the verb possum in (h).
123
For this use of hoc, see TLL s.v. hic 2707.54ff.
124
See Skutsch ad loc. and Rosén (1996b: 132–3) on the ‘cognate nominative’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
As in other languages, the Latin verb sum ‘to be’ functions in various ways. It may be
used in ‘identity statements’ of the type this is my father as well as in ‘predicational
statements’ of various types. Examples of sum in identity statements are (a) and (b).
An example of a predicational statement is (c) where orator ‘orator’ indicates the class
to which the subject constituent (‘I’) belongs. A further example of sum in a predica-
tional statement is (d), where the adjective mortales (mortal) serves to indicate a
property or quality of animi ‘souls’. In all these instances the verb sum requires two
arguments, a subject and a subject complement.
(a) . . . cum ignorante rege uter Orestes <esset, Pylades> Orestem se esse diceret,
ut pro illo necaretur, Orestes autem ita ut erat, Orestem se esse perseveraret.
(‘ . . . Pylades, since the king was unsure which of the two was Orestes, declared that
he was Orestes so that he would be put to death instead of his friend, while Orestes
maintained that he was Orestes, as was the case.’ Cic. Amic. 24)
(b) Quaeso, tune is es, / Lyco trapezita? # Ego sum.
(‘Please, is it you, the banker Lyco? # Yes, that’s me.’ Pl. Cur. 419–20)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The subject complements in these examples agree with the subject in number, gender,
and case. In (a), Orestes in the clause uter Orestes esset is in the nominative (as is the
subject uter), whereas Orestem in the accusative and infinitive clause Orestem se esse is
in the accusative, as is the subject se. (For the various categories that may function as
subject complements, see § 9.20.)
The valency of the combination of sum and the subject complement depends on
the valency of the subject complement: in (d), with mortales, it requires only one
argument, animi, the subject of the sentence; in (e), the combination of the copula
with the adjective dignus ‘worth’ is two-place, the subject (‘you’) and verberibus
multis, which denotes what the subject is worth.
(e) . . . dignu’n es verberibus multis?
(‘ . . . do you deserve a lot of blows?’ Pl. Mil. 342)
The status of the verb sum in such expressions is much debated in the literature.
Whereas in the introduction to this section sum is described as a two-place verb,
other scholars125 maintain that sum in such constructions is just a means of express-
ing the person and number of the subject constituent and the tense and mood of the
clause, without contributing much meaning to the content of its clause. It is some-
times regarded as a ‘dummy’ element that only functions as a support for a complex
expression. In support of this position, scholars argue that forms of sum are often
absent (see § 4.96 on so-called nominal sentences). However, there are a number of
reasons not to adopt this analysis.126 In the first place, sum differs in meaning from
other ‘copular’ verbs that are also combined with a subject complement, such as fio
‘to become’ and maneo ‘to remain’ (see § 4.97 for details). Secondly, when a clause
has no explicit subject because it can be inferred from the preceding context, est, sunt,
or another third person form must be expressed, as in (f). Thirdly when it must be
made clear that something is indeed the case, sum cannot be omitted and is even
regularly found in the very first position of the sentence, as in (g).127
125
Including LSS § 1.1, note 4. For a discussion of the status of sum, see Hoffmann (1997a: 38–42).
126
For arguments against the ‘dummy hypothesis’, see Stassen (1997: 65–76) and Cabrillana (2010a:
22–7).
127
For other factors that favour the initial position of forms of sum, see Adams (1994b: 69–81).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The use of sum as a copula has to be distinguished from several other uses.129 Uses
discussed elsewhere in this chapter are (a) two-place sum ‘to be worth’ (see § 4.32); (b)
two-place ‘possession’ sum (see § 4.26); (c) two-place position sum ‘to be somewhere’
(§ 4.42); and (d) various uses of impersonal est, among which ‘it is possible’ (see
§ 4.15). In this section two more uses will be distinguished and perhaps a third. It is
not easy to formally distinguish these types and perhaps other types of sum, and one
may even doubt whether we should distinguish more than two verbs sum (one-place
‘existential’, two-place ‘copula’) alongside the auxiliary use.130
128
For these two properties, see Cabrillana (1998).
129
For a typological treatment of the verb ‘to be’, see Feuillet (1998).
130
Happ (1976: 560) distinguishes six types of sum. See Serbat (1983) for criticism and for the position
that only two types should be distinguished: existential sum and copulative (meaningless) sum. Adams
(1994b) distinguishes two types of sum (apart from the auxiliary use). Pezzini (2011: 340) suggests that
contraction of the form est (to st) is more common with the auxiliary use than with the existential and
other uses. The status of the verb ‘to be’ is also much debated by philosophers.
131
For statistical data, see Lease (1919: 264–6). See also Garuti (1954). For the future infinitive in
Seneca, see § 7.72.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
132
For more examples and references, see Pettersson (1930: 91–3), Sörbom (1935: 151ff.), Blomgrén
(1937: 68ff.), Sz.: 422.
133
For Varro’s usage, see Marouzeau (1910a: 288–91) and Laughton (1960: 10). For Virgil, see Görler
(1985: 274). The most complete collection of omissions of sum (in its various functions) in Virgil and
other poets is still Leo (1878: 184–93). For examples of omission of the auxiliary of perfect forms of
deponent verbs in Tacitus and a few other authors, see Sörbom (1935: ch. 5), with critical discussion in
Goodyear’s commentary (1972: I, appendix 5).
134
See Goodyear (previous footnote).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(h) Interea redit Appius et percontati ø nos ab illo et ille a nobis quid esset
dictum et factum.
(‘Meanwhile, Appius returns, and we are asked by him and he by us what
has been said and done.’ Var. R. 3.12.1)
(i) Multa suppliciter locutus ø ut de sua atque omnium salute debebat silen-
tioque ab utrisque militibus auditus ø.
(‘Much he said in the suppliant tones that he was bound to use in the
interests of his own and general safety, and he was heard in silence by both
forces.’ Caes. Civ. 3.19.3)
(j) Captum oppidum ac direptum est. Ad duo milia et trecenti occisi ø et sex
milia hominum capta ø, et miles ingenti praeda potitus ø. Quam vendere
sicut priorem coactus, Ferentinum inde, quamquam nihil quietis dabatur,
tamen summa alacritate ductus ø.
(‘The town was captured and sacked; two thousand three hundred were
slain and six thousand made prisoners, and the soldiers came into posses-
sion of huge spoils which they were obliged, as before, to sell. After that they
marched with the utmost alacrity—though they had been allowed no time to
rest—to Ferentinum.’ Liv. 10.17.8–9—NB: Punctuation as in Oakley; no
need for <est> ductus or ductus <est>)
(k) Ita victores latus hostium invecti ø.
(‘Thus, victorious, they assailed the enemy’s flank.’ Tac. Hist. 2.43.2)
(l) Per haec tempora . . . ad regendam praefecturam praetorianam ab urbe Pro-
bus accitus ø . . .
(‘Throughout this time, Probus was summoned from the city for the pur-
pose of leading the praetorian prefecture.’ Amm. 27.11.1)
Supplement:
In subordinate clause: Optas quae facta ø. (Pl. Am. 575); Postquam introgressi ø et
coram data ø copia fandi, / maximus Ilioneus placido sic pectore coepit. (Verg. A.
1.520–1);
Non-third person: Nunc adeo, ut facturus ø, dicam. (Pl. Men. 119); Ter conatus ø ibi
collo dare bracchia circum. (Verg. A. 2.792);
Non-present: (sc. Pergama) Quae neque Dardaniis campis potuere perire / nec cum
capta ø capi nec cum combusta ø cremari (Enn. Ann. 358–9V=344–5S);
Non-indicative: Quid enim? Facta necne facta largitio ø ignorari potest? (Cic. Fam.
3.11.2); Dissimulant et nube cava speculantur amicti, / quae fortuna viris ø, classem
quo litore linquant, / quid veniant. (Verg. A. 1.516–18); Neque enim ille dedisset, / si
molles habitus et tegmina foeda fateri / ausa ø seni. (Stat. Ach. 1.141–3);
For the use of a number of verbs instead of auxiliary sum in Late Latin, see Sz.: 395.
Early instances have been collected from the Augustan poets onwards (and even in
Lucretius, as in ex. (m)), but in most instances the verb has its own meaning.135
135
For maneo, see TLL s.v. 290.13ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(m) Sed ne forte putes solo spoliata colore / corpora prima manere, etiam secreta
teporis / sunt . . .
(‘But lest by chance you think that the first bodies abide bereft only of colour, they are
also sundered altogether from warmth . . . ’ Lucr. 2.842–4)
136
For more examples, see Cabrillana (2010a: 31–4).
137
On existential escit, see Fraenkel (1964: II.444–5 (‘vorhanden sein’)).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The copula sum is often absent. A sentence in which a form of copular sum is absent,
but could be added without changing the meaning and the structure of the utterance
is called a NOMINAL SENTENCE. Nominal sentences are found from the earliest period
onwards, notably in inscriptions indicating whose property the inscribed object is.
Two examples are (a) and (b). In (a), kavidios is sometimes taken as a possession
expression showing agreement with the subject eqo. Ex. (b) has a noun in the genitive
as subject complement.138
(a) EQO KAVIDIOS
(‘I belong to Gavidus’ CIL I2.474 (Ardea, first half VI BC))
(b) EKO KAISIOSIO
(‘I belong to Caesius’ CIE 8163 (Civita Castellana, 575–525 BC))139
138
For problems concerning the reading, date, and interpretation of (a), see Solin (2003) and
Hartmann (2005: 143–6, 202, 262ff.). The various ownership expressions on objects are discussed by
Agostiniani (1983).
139
See Bakkum (2009: 415).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The evident constraint on the absence of the copula is that the grammatical structure
and the context must leave no doubt about the identity of the subject and about the
temporal and modal meaning of the clause. This explains why nominal sentences are
almost absent in legal texts.140
Third person forms are more frequently absent than those of the first and second
person, because the third person subject is more often explicit. Additionally, the
present indicative forms are more often absent than other verb forms.141 There are no
restrictions on sentence type. Although nominal sentences are usually not complex,
there are no restrictions. The clauses are usually brief and without satellites. There are
a number of idioms in which absence of the copula is regular. In some of these
contexts other verbs are also often absent, especially verbs of speaking and of
doing.142 Representative examples of nominal sentences are (c)–(i).143 Ex. (c) is a
proverb-like statement; (d) is a short statement in the form of a rhetorical question;
(e) is a transitional statement; (f) is a conditional period; (g) has a relative clause.144
Ex. (h) is a dependent question where the subjunctive is regular. Ex. (i) illustrates
an idiom. In the Supplement, types of omission can be found that are less common
than these.
Variation in the frequency with which the copula is omitted is determined not only
by context (and hence also by text type) but also by the preference of individual
authors. After the Classical period the omission of forms of sum becomes a stylistic
device, exploited extensively by Tacitus.145
(c) . . . quot homines, tot sententiae.
(‘ . . . there are as many opinions as there are people.’ Ter. Ph. 454)
(d) Quid hoc tristius?
(‘What is more grim than this?’ Cic. Div. 2.79)
(e) Sed haec vetera, illud vero recens, Caesarem meo consilio interfectum.
(‘All this is old news, but there is something recent: he says I instigated Caesar’s
killing.’ Cic. Phil. 2.25)
(f) Cur hostis Spartacus, si tu civis?
(‘If you are a citizen, what makes Spartacus an enemy?’ Cic. Parad. 30)
(g) Sed nihil quod crudele utile.
(‘But no cruelty can be expedient.’ Cic. Off. 3.46)
140
For a discussion of two exceptions in Cicero, both quotations from the Lex. XII: ‘Adversus hostem
aeterna auctoritas’ (Lex XII.3.7 = Cic. Off. 1.37); ‘Ne sumptuosa respersio, ne longae coronae, ne acerrae’
(Lex XII.10.6 = Cic. Leg. 2.60), see Daube (1956: 103–11).
141
For typological observations, see Stassen (1997: 64–5).
142
See Cabrillana (2007).
143
For a classification of types of absence, see Sz.: 420–1. See also Eklund (1970: 137–9). References in
Sz.: 423.
144
For nominal relative clauses, see Touratier (1980b: 467–82).
145
For nominal sentences in Livy, see Pettersson (1930: 89–100).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
146
For the reading and similar instances, see Petersmann ad loc.
147
For more instances in Seneca, see Charney (1943).
148
See Fraenkel (1966).
149
For more instances, see Laughton (1960: 10).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(j) Non mediocris enim tenebrae in silva [sunt] ubi haec captanda [sunt] neque
eo quo pervenire volumus semitae tritae [sunt], neque non in tramitibus
quaedam obiecta [sunt] quae euntem retinere possent.
(‘For there is no slight darkness in the wood where these things are to be
caught, and there are no trodden paths to the point we wish to reach, nor do
there fail to be obstacles in the paths, which could hold back a person on his
way.’ Var. L. 5.5)
Apart from the copula sum, several verbs in at least one of their meanings occur in
equative constructions and some of them became almost equivalent to the copula or
even in the course of history replaced sum. These verbs are either non-dynamic or
dynamic. Typical instances of non-dynamic verbs are the copula sum ‘to be’ itself and
the verb maneo ‘to remain’ (not very frequent in this meaning) and its compounds
permaneo ‘to persist’ and remaneo ‘to remain’. Fio ‘to become’ is the typical dynamic
copular verb. Examples of copular verbs are (a)–(c).
(a) Numquam edepol fuit nec fiet ill’ senex insanior / ex amore quam ille
adulescens . . .
(‘That old man has never been and never will be madder on account of his love than
that young man . . . ’ Pl. Mer. 446–7)
150
See discussion in Touratier (1980b: 479–82) and Stassen (1997: ch. 3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Si tam corpus loco validum ac regione maneret / scriptoris, quam vera manet
sententia cordi.
(‘If only the writer’s body had strength to stay in its place and at its post even as the
feeling of truth stays in his heart.’ Lucil. 189–90M=192–3K)
(c) Ego sum ille Amphitruo, quoi est servos Sosia, / idem Mercurius qui fit
quando commodum est . . .
(‘I am that Amphitruo who has a slave Sosia who becomes Mercury when it’s
convenient . . . ’ Pl. Am. 861–2)
Supplement (in alphabetical order by verb):
Simul atque enim se inflexit hic rex in dominatum iniustiorem, fit continuo
tyrannus . . . (Cic. Rep. 2.48); Lenior et melior fis accedente senecta? (Hor. Ep.
2.2.211); . . . deinde, si qua gratia testes deter<r>entur, tabulae quidem certe incor-
ruptae atque integrae manent. (Cic. Font. 3); . . . quod superioris anni munitiones
integrae manebant, ut militum laborem sublevaret. (Caes. Gal. 6.32.5); . . . sive do-
minus manet sive desiit esse dominus . . . (Ulp. dig. 9.4.2.1); Tu tamen permanes
constantissimus defensor Antoni. (Cic. Phil. 8.17); Quae (lucerna) si permanserit
ardens . . . (Vitr. 8.6.13); Post obligationem vero si cesserit, nihilominus ipse heres
permanet . . . (Gaius Inst. 2.35); (sc. auctoritas) Quae tamen gravis et magna
remanebat . . . (Cic. Rep. 2.59);
In poetry, the function of this substitution is to replace the colourless verbs fio and
sum with verbs with a more ‘concrete’ meaning. As a matter of principle, these
151
See TLL s.v. devenio 580.77ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(g) Sed mihi, cum de senectute vellem aliquid scribere, tu occurrebas dignus eo
munere quo uterque nostrum communiter uteretur.
(‘But when I resolved to write something about old age, you continually
came before my mind as worthy of a gift which both of us might enjoy
together.’ Cic. Sen. 2)
Supplement (in alphabetical order by verb):
Prose instances: . . . sic nos videmus, qui oratores evadere non potuerint, eos ad iuris
studium devenire. (Cic. Mur. 29); Egregiam hanc alam equitum evasisse ferunt. (Liv.
29.1.11); Nemo dupondii evadit. (Petr. 58.14 (a freedman speaking)—NB: genetivus
pretii); (sc. pecora) Quae post tempus nascuntur, fere vitiosa atque inutilia exsistunt.
(Var. R. 2.7.7); Ergo illa falsa fuerunt. Quae certe vera extitissent, si Milo admisisset
aliquid quod non posset honeste vereque defendere. (Cic. Mil. 64); Hippocratem (et
alios) . . . non ideo quidem medicos fuisse, verum ideo quoque maiores medicos
exstitisse (Cels. 1 pr. 47); . . . antequam heres extiterit, proinde fit heres is cui in
iure cesserit . . . (Gaius Inst. 2.35); . . . fideles non exstitistis . . . (Vet. Lat. Luc. 16.11; cf.
Vulg. fuistis); Ita metopae quae proximae ad angulares triglyphos fiunt non exeunt
quadratae sed oblongiores triglyphi dimidia latitudine. (Vitr. 4.3.2);
Poetic instances: . . . non mihi sat magnus Castoris iret equus. (Prop. 2.7.16);152 . . . et
mihi ritu / Manius hic generis prope maior avunculus exit. (Pers. 6.60–1); Inde fit, ut,
quotiens existere perfida temptas, / peccatum totiens corrigat illa tuum. (Ov. Ep.
20.113–14); . . . nec Leo venator veniet nec Virgo magistra . . . (Manil. 4.382);
Alternatives to the copula sum are rare in prose, but see exsto ‘to exist’ in (h).
Lucretius already has a number of instances.
(h) Sic velut inita arbor fecundo semine fertilior exstat.
(‘The result is that the tree, being as it were impregnated with fruitful offspring,
becomes more productive.’ Col. 5.9.16)
Supplement: Primum animum dico, mentem quem saepe vocamus, / in quo con-
silium vitae regimenque locatum est, / esse hominis partem nihilo minus ac manus
et pes / atque oculei partes animantis totius exstant. (Lucr. 3.94–7—NB: note
parallelism); . . . altis urbibus ultimae / stetere causae cur perirent . . . (Hor. Carm.
1.16.18–19); Dabitur opera. # Lepidus vivis. (Pl. Trin. 390);153
152
See TLL s.v. eo 650.35ff.; s.v. exeo 1364.42ff.; s.v. exsisto 1874.35ff.; s.v. exsto 1934.25ff.
153
For instances of vivo ‘to live’ as a poetic alternative for the copula, see Heerdegen (1913) (with
critical remarks in Hofmann 1924).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Verbs indicating the physical position of human beings (iaceo ‘to lie’, sedeo ‘to sit’,
and sto ‘to stand’) have played a role in the development of the copula in Spanish and
other Romance languages. There are a few instances in which these verbs are used
without referring to a specific physical position of the subject. An Early Latin instance
of sedeo is (i), where it means ‘to be encamped’. Compare also (j) from Cicero. An
example of sto discussed in the literature is (k). The verbs in these examples refer to
‘staying somewhere’, the use discussed in § 4.42. In poetry, the use of these verbs in a
non-local meaning is attested in examples like (l).154 Instances from Late Latin in
which reference to physical position is unlikely are (m) and (n). Real copular use,
however, developed later.
(i) Ordine omne, uti quidque actum est, dum apud hostis sedimus, / edissertavit.
(‘He told me all from first to last, just how everything happened while we were in the
field with the enemy.’ Pl. Am. 599–600)
(j) Ita patent omnia quae accidere possunt ut, ea si vitem, sedendum sit cum
dedecore et dolore . . .
(‘Everything that can happen is obvious. If I avoid these risks I must accept an
ignominious and painful inactivity . . . ’ Cic. Att. 10.12a.1)
(k) . . . qui domi stare non poterant largo et liberali viatico commovebat.
(‘ . . . and those who were not able to stay at home he got moving with a liberal and
generous travel allowance.’ Cic. Flac. 14)
(l) . . . nec notis stabat contenta venenis (sc. Medea).
(‘ . . . nor abides contented with the drugs she knew.’ V. Fl. 7.354)
(m) Lectiones etiam, quecumque in ecclesia leguntur, quia necesse est Grece legi,
semper stat qui Siriste interpretatur propter populum, ut semper discant.
(‘Even the lessons that are read in church, since they must be read in Greek, there is
always standing one who interprets them into Syriac for the people’s sake, so that
they may always learn.’ Pereg. 47.4)
(n) Id est illa valle quam superius dixi, ubi sederant filii Israhel, dum Moyses
ascenderet in montem Dei . . .
(‘This is that valley, which I mentioned above, where the sons of Israel stayed while
Moses ascended the mountain of God.’ Pereg. 5.1)
Appendix: Whereas in the instances discussed above the verbs can be regarded as
‘substitutes’ of existing forms of fio and sum, there are other (Late Latin, mainly
ecclesiastical) instances of the use of exsisto in which the verb functions as a ‘suppletion’
for non-existing forms of the verb sum, notably the present participle and the gerund.155
Exx. (o) and (p) show how the Greek original is translated in different ways in the Vetus
Latina and Vulgate versions of the Bible (the Greek text has the participle æåø ).
154
See Campos (1973) and Stengaard (1991: 29–59). See also Pinkster (1987a: 215–19) on the use of sto.
155
Abundant examples in TLL s.v. exsisto 1875.33ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(o) Deus . . . caeli et terrae dominus existens, non in manufactis templis habitat.
(‘God, existing as master of sky and earth, does not inhabit temples made by
hand . . . ’ Vet. Lat. Act. 17.24 (= Iren. 3.12.9))
(p) Deus . . . caeli et terrae cum sit Dominus, non in manufactis templis habitat . . .
(‘God, since he is master of sky and earth, does not live in handmade
temples.’ Vulg. Act. 17.24)
Perception verbs in the passive can be used with subject complements. Videor, passive
form of video ‘to see’, behaves more or less as a copular verb, meaning ‘to seem’, ‘to
look’, as in (q)–(u) (but note the presence of tibi in (q)). In (q), the subject comple-
ment is an adjective that shows agreement with the subject, in (r) an idiomatic
expression, in (s) a noun phrase that agrees with the subject, in (t) a so-called
genetivus possessivus. Videor is also used with a clause as the subject and an evaluative
adjective as subject complement (its so-called impersonal use), as in (u).156 Sentio ‘to
perceive’ is also used with subject complements, likewise in the passive, as in (v).
(q) Quid nunc? Scelestus tibi videtur Chrysalus?
(‘What now? Does Chrysalus seem to be the criminal to you?’ Pl. Bac. 854)
(r) Quid tibi ego aetatis videor? # Acherunticus, / senex vetus, decrepitus.
(‘How old do you think I am? # Ripe for the Underworld, old and decrepit.’ Pl. Mer.
290–1)
(s) . . . tibi solitus sum dicere magis te fortem accusatorem mihi videri quam
sapientem candidatum.
(‘I was always telling you that you seemed to me more like a vigorous prosecutor than
an astute candidate.’ Cic. Mur. 43)
(t) . . . si hoc magni cuiusdam hominis et persapientis videtur . . .
(‘If such conduct is deemed possible only for some great man of outstanding
wisdom . . . ’ Cic. Prov. 44)
(u) Aequom videtur tibi ut ego alienum quod est / meum esse dicam?
(‘Does it seem right to you that I should say that something belonging to someone
else belongs to me?’ Pl. Rud. 1230–1)
(v) Quisquis erit ventus, si fervidus sentietur, pluribus diebus permanebit.
(‘Any wind, if it is felt to be hot, will last for several days.’ Plin. Nat. 18.339)
Supplement:
. . . principem . . . se esse mavult quam videri. (Cic. Off. 1.65); . . . cum ille . . . tardior
tibi erit visus . . . (Cic. Fam. 7.17.2); Platonicum sese philosophum dicebat et viderier
gestiebat. (Gel. 15.2.1); . . . satis nigrum <et> quasi perustum visum fuerit . . . (Larg.
156
See Joffre (1995: 132–43); Orlandini (1996b: 417–18) and Heberlein (2002: 176–8), also for other
uses of videor. See also § 4.98 for its auxiliary use.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
228); Rex vero licet securitatem praeferens vultu exsultansque specie tenus urbis
excidio videbatur . . . (Amm. 19.9.9);
. . . nec quicquam ad nos pertinere nisi quod aut leve aut asperum in corpore
sentiatur. (Cic. Tusc. 5.73); (sc. empetros) Quae longius magisque terrena, amarior
sentitur. (Plin. Nat. 27.74); . . . eaedem (sc. verae gemmae) namque in ore gelidiores
sentiuntur. (Plin. Nat. 37.199);
Another verb occurring with a subject complement is appareo ‘to be found’, ‘prove to
be’, attested from Cicero onwards. Examples are (w)–(y).
(w) . . . nisi cetera quae sunt a me in secundo libro de oratore per Antoni
personam disputata de ridiculis å Æ et arguta apparebunt . . .
(‘If the other varieties which I discussed through Antonius’ mouth in the second
volume of my treatise On the Orator shall not appear neatly pointed and secundum
artem . . . ’ Cic. Fam. 7.32.2)
(x) Si quod es appares, culpa soluta mea est.
(‘If you appear to be what you really are I am acquitted of fault.’ Ov. Tr. 4.4a.10)
(y) Cum . . . desertum apparuisset forum . . .
(‘ . . . when the forum had appeared deserted . . . ’ Liv. 3.52.5)
Supplement:
Quae, cum sint vetusta, sic apparent recentia, uti si sint modo facta. (Vitr. 2.7.4);
Promisi tibi in matrimonium filiam. Postea peregrinus adparuisti. (Sen. Ben. 4.35.1);
Appendix: Audio ‘to hear’ and—later—sono ‘to sound’ are used with a subject
complement in poetry, from Catullus onward, presumably under Greek influence,157
with the meaning ‘to be called’, ‘to be spoken of (as)’ (cf. IŒø). Examples are (z) and (aa).
(z) . . . solum / quod quondam caesis montis fodisse medullis / audit falsiparens
Amphitryoniades . . .
(‘ . . . the soil which of old the false-fathered son of Amphitryon is said to
have dug out, cutting away the heart of the hill . . . ’ Catul. 68.110–12)
(aa) . . . invictus . . . tua, Caesar, in urbe sonas.
(‘ . . . in thy city, Caesar, thou art proclaimed Unconquered.’ Mart. 7.6.8)
Supplement: . . . rexque paterque / audisti coram, nec verbo parcius absens. (Hor. Ep.
1.7.37–8); . . . ipse subtilis veterum iudex et callidus audis. (Hor. S. 2.7.101); Felix
alius magnusque sonet. (Sen. Her. O. 692);
NB: Passive: Cassusne sacerdos audior? (Stat. Theb. 4.504);
The verb clueo (from Seneca onwards also cluo) ‘to be called’, ‘to be reckoned as’ and
its passive form clueor (very rare) occur with subject complements. It is relatively
frequent in the comedies of Plautus and in Lucretius, and occurs with some frequen-
cy in some Late Latin texts. Examples are (ab) and (ac).
157
See Coleman (1975: 137). For Greek examples, see K.-G.: I.43.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) . . . quia maius est beneficium quam posse debet civis civi dare . . .
(‘ . . . since the benefaction is greater than one citizen should be able to confer
on another . . . ’ Cic. Sul. 72)
(ii) The first argument of the infinitive is unexpressed, but is the same as the
subject of the auxiliary verb. This can be seen when there is a secondary
predicate like primus ‘first’ associated with the infinitive: it agrees with the
subject of the auxiliary in number, gender, and case, as in (b).
(b) Utque iuvent alii, tu debes vincere amicos / uxor et ad partis prima
venire tuas.
(‘And to have others aid me thou shouldst win our friends, my wife, and come
foremost thyself to support thy part.’ Ov. Pont. 3.1.41–2)
(iii) With zerovalent weather verbs, the phasal auxiliaries incipio ‘to begin’ and
desino ‘to stop’ are third person singular, as in (c) and (d).
(c) Cum pluere incipiet familiam cum ferreis sarculisque exire oportet . . .
(‘When the rains begin, the whole household must turn out with shovels and
hoes . . . ’ Cato Agr. 155.1)
(d) . . . cum desierit pluere, uti tegula mutetur.
(‘ . . . so that the tile can be replaced after the rain stops.’ Cato Agr. 155.2)
158
For a discussion of the status of auxiliaries in Latin, see Scherer (1975: 124); Happ (1976: 482–4);
Touratier (1994: 560–4). For theoretical discussion, see also Heberlein (2002).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(iv) Auxiliary and full verb together form an AUXILIARY PHRASE. As far as valency is
concerned, the properties of an auxiliary phrase are that of the full verb. This
is illustrated by (e)–(h). The structural possibilities of dico and cogito are not
affected by the auxiliaries soleo and possum, respectively. Furthermore, the
same adjuncts and secondary predicates can be added as could be added to
the full verb when used as the main verb in a clause.
As example (g) shows, when verbs that are passivizable are used in combin-
ation with an auxiliary verb, passivization is still possible. When it occurs, it
is the voice of the infinitive that is changed (for the use of the passive with the
phasal auxiliary verbs incipio [coepi] and desino, see § 5.13). All other
morphosemantic properties, including tense, are carried by the auxiliaries
(for perfect infinitives with an anterior meaning, and for idiomatic uses, see
§ 7.70 and § 7.76). The clausal negator non usually precedes the auxiliary and
not the infinitive.
The infinitive may be unexpressed if it can be understood from the context.
A clear example is (i). The infinitive may also be conventionally understood, as
in a famous passage from Catullus, (j), where Maeciliam looks like an object of
solebant, but where futuere ‘to have sex with’ is understood. Similarly in (k).
(v) There are no restrictions on the infinitives auxiliaries can be used with. They
are used with animate and inanimate subjects; with dynamic and non-
dynamic, with controllable and non-controllable states of affairs; and the
infinitive may be active or passive. All three are found with neuter (especially
anaphoric, demonstrative, or relative) pronouns as their subject, referring to
some state of affairs that is needed, possible, or customary. Illustrations with
the verb possum are (l)–(p).
On the basis of these criteria, clear-cut auxiliaries are the modal verbs debeo ‘must’,159
possum ‘to be able to’, ‘to be possible’,‘to be allowed to’, queo ‘to be able’ and nequeo
‘to be unable’;160 the habitual auxiliary soleo ‘to be accustomed to’; and the phasal
verbs incipio (coepi) ‘to begin’ and desino ‘to stop’.161
The nouns and noun phrases that do occur as the second argument of auxiliaries are
like (q). Telam means ‘working on the tela’ (see texentem telam in l. 285).
159
For the modal use of debeo, see Bolkestein (1980: ch. 11, on which see also Heberlein 2002: 172–4)
and Martín Rodríguez (2002), with references.
160
For nequeo, see Moussy (2002).
161
Happ, not using criterion (iv), on the basis of his corpus also reckons audeo ‘to dare’ and conor ‘to
try’ among the auxiliaries.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apart from its use as copular verb (see § 4.97) videor, in its ‘evidential’ meaning ‘to
appear’, ‘to seem’, can be combined with infinitives, as in (r). When used in that way it
shares with auxiliaries most of the properties mentioned above, with the exception of
the first: its infinitives are not restricted to present infinitives, as is shown by (s), with
a perfect infinitive.162
(r) . . . / geminos in ventre habere videor filios.
(‘ . . . I seem to have twin sons in my belly.’ Pl. Cur. 221)
(s) Non videor vidisse lenam callidiorem ullam alteras.
(‘I don’t think I have seen a cleverer procuress anywhere.’ Pl. Mos. 270)
Appendix: Debeo is also used in the meaning ‘to owe’. In that meaning it is
commonly found with a noun as its object, which may also be the subject in a passive
clause, as in (t). The relationship between the two expressions is not clear. It is
customary to derive auxiliary meanings from full verb meanings, but then it is
remarkable that there is no trace of the dative with the auxiliary phrase. The
alternative is to assume that with the interpretation ‘owe’ an infinitive (for example,
dare ‘to give’) became conventionally understood and produced an idiom.163 Note
that debeo ‘to owe’ can be combined with an auxiliary, as in (u). Debeo in its modal
use is already attested in Plautus, though with only three instances.
(t) Quia patrem prius convenire / se non volt nec conspicari / quam id
argentum quod debetur pro illa dinumeraverit.
(‘Because he doesn’t want to meet or see his father until he’s paid the money
owed for her.’ Pl. Epid. 70–1a)
(u) Sensit enim deberi non posse huic civitati illam potestatem.
(‘For he realized that this office could not be withheld from our republic.’
Cic. Leg. 3.26)
Minor companions of auxiliary possum are queo ‘to be able to’ and nequeo ‘to be
unable to’. Apart from the uses mentioned above, possum is also found with the
meaning ‘to have power’, ‘to have influence’, etc. (OLD s.v. § 8). In this meaning,
possum is used with a so-called ‘internal’ or ‘adverbial’ accusative, expressions like
quid ‘what’, quantum ‘how much’, plurimum ‘most’, that is adjuncts of degree, as
found with the verb valeo ‘to have force or power’. Possum in this meaning therefore
is a one-place verb, which, like debeo, may have arisen of conventional omission of an
infinitive.
The phasal auxiliary verbs desino and incipio are also both used in a one-place
frame with a noun that implies a state of affairs as the subject, as in ex. (v).
162
For further examples, see OLD s.v. 20.d. See also Heberlein (2002: 176–8).
163
So Touratier (1994: 561).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Desisto ‘to leave off ’ and ingredior ‘to embark on’, ‘to begin’, which seem prima facie
to be near-synonyms, are also used with an infinitive, but—as far as the limited
number of attestations allows us to conclude—are restricted to controllable states of
affairs. There is an occasional instance of desisto with febris ‘fever’ as the subject
(Cels. 3.3.2), but this looks like an absolute use.164 So these verbs are probably of the
type ‘main verb + non-finite clause’.165 The verb pergo (literally ‘to move onward’) is
used with an infinitive in its non-literal meaning ‘to proceed’ and is described as
‘quasi verbum auxiliare’.166 Although most instances of pergo + infinitive are clearly
controllable, as one expects from its meaning, there are a few exceptions, as in (w)
and (x). In (w) ventus is an inanimate subject, although of the type that shares
properties with human beings. In (x), the state of affairs is definitely uncontrollable.
Pergo is not used with nouns that imply a state of affairs as the subject. Weak
evidence, to conclude, for auxiliary status. Cesso ‘to be slow to’ ‘to cease’, by contrast,
seems to be a better candidate—see (y)—but it is often one-place, also with animate
subjects, and (rarely) used with an ablative constituent (see § 4.29) or a prepositional
phrase denoting from what one ceases.
(w) . . . si pergerent iidem qui tum tenebant ab ortu solis flare per dies plures
venti . . .
(‘ . . . if the same winds which were then holding should continue to blow
from the east for several days . . . ’ Liv. 25.27.6)
(x) Si pro portione pergit (v.l. crescit) luxuria crescere, quid relinquitur . . . ?
(‘If luxury continues to increase in its present proportion, what remains . . . ?’
Fav. orat. 1)
(y) . . . cum . . . maesta neque assiduo tabescere lumina fletu / cessarent tristique
imbre madere genae.
(‘ . . . when my sad eyes never rested from wasting with perpetual tears, nor
my cheeks from streaming with a flood of sorrow.’ Catul. 68B.55–6)
As in other languages, the distinction between auxiliary verbs and full verbs is not
clear-cut. A large number of verbs that are normally found with nouns and pronouns
as their object are also used in combination with a present infinitive in an auxiliary-
like manner.167 One class of verbs that is often mentioned in this connection are verbs
denoting will and endeavour, especially the verbs volo ‘to wish’ and studeo ‘to concern
oneself with’.168 These verbs are found with object nouns, in the accusative and
dative, respectively, as in (z) and (aa). Instead of nouns they can also have clauses
as their object, among these a (present) infinitival clause. Further details are discussed
in § 15.121.
164 165
See TLL s.v. 731.46–7. In the terminology of Quirk et al. (1985: 137).
166
So TLL s.v. 1432.71ff., where references to more inanimate subjects can be found as well.
167
For the ‘auxiliary verb–main verb scale’ in English, see Quirk et al. (1985: 137).
168
See K.-St.: I.667 and Sz.: 346.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
When one wants to establish the valency of adjectives one encounters similar
problems as with verbs (see § 4.1).170 In the first place, second arguments can be
left unexpressed, if the context or situation contains sufficient information. Sec-
ondly, adjectives can be used ‘absolutely’, as in (d): note the parallelism with the
one-place adjectives turpis and petulans. Conversely, avarus, which on the basis of
statistical data is a clear-cut one-place adjective, is expanded on the analogy of
169
Wölfflin (1904) uses the term ‘adiectiva relativa’.
170
For a discussion of the problems one encounters when trying to determine the valency of adjectives
(on the basis of Var. R.), see Bodelot (2011).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
cupidus with a genitive argument by Tacitus in (e). Note here the parallelism with
adpetens + genitive (regular) and parcus + genitive (rare, found mainly in poetry
and poeticizing prose).
(d) . . . non esse mirandum qui in illa re turpis aut cupidus aut petulans fuerit
hac quoque in re eum deliquisse.
(‘ . . . and that it is no wonder if one who in that other affair acted basely, passionately
and wantonly should have transgressed in this case also.’ Cic. Inv. 2.33)
(e) (sc. Galba fuit) Pecuniae alienae non adpetens, suae parcus, publicae avarus.
(‘He was not greedy for another’s property; he was frugal with his own, stingy with
the state’s.’ Tac. Hist. 1.49.3)
Arguments are less frequently expressed when adjectives function as attribute at the
noun phrase level and even less when they function as head noun (see also § 11.92).
This is due to the tendency not to make the information of a clause overly complex.
The same goes for adverb phrases.
Apart from functioning as central element at the clause level, adjectives can function
as attribute at the noun phrase level and as head of a noun phrase (their so-called
substantival use). Adverbs may also be derived from them. In principle, adjectives
maintain their valency in all these situations, as in (f)–(h). In (f) cupidus is the
attribute of praetor; existimationis bonae is its second argument. In (g) cupidi is a
head noun; it governs tui as its second argument and is itself modified by the
quantifier multi (note the parallelism with the nouns amici and fautores). In (h)
cruce is the argument with the adverb dignius; the adverbial phrase functions as a
manner adjunct in its clause.
(i) . . . quid dicetis? Utrum Metellum falsum scribere an amicum laedendi esse
cupidum . . .
(‘ . . . what will you say? That what Metellus wrote is false, or that he was anxious to
damage his friend . . . ’ Cic. Ver. 3.122)
(j) Ego eius videndi cupidu’ recta consequor.
(‘I followed straight behind, eager to see her.’ Ter. Hec. 372)
(k) (sc. Phylacides) . . . cupidus falsis attingere gaudia palmis / Thessalis anti-
quam venerat umbra domum.
(‘The scion of Phylacus, yearning to touch with his unreal hands his heart’s delight,
came, a ghost of Thessaly, to his ancient home.’ Prop. 1.19.9–10)
(l) Quia enim non sum dignus prae te palum ut figam in parietem.
(‘Because compared with you I am not worthy to pound a peg into a wall.’ Pl. Mil. 1140)
(m) Is primam ante aciem digna atque indigna relatu / vociferans . . . / ibat . . .
(‘He strode ahead of the foremost line, shouting words meet and unmeet to utter . . . ’
Verg. A. 9.595–7)
Adjectives may also be expanded with various types of optional constituents, just like
satellites with verbs at the clause level. These optional constituents are discussed and
illustrated in § 11.92.
In the remainder of this section, second arguments of adjectives will be presented
in the same order as those with two-place verbs in §§ 4.23–4.45.
Adjectives governing a dative argument belong to three semantic classes, which are
presented in Table 4.21: helpfulness, interpersonal relations, and similarity and suit-
ability. They correspond to three of the classes of verbs mentioned in § 4.24. Many of
these adjectives are also used with one or more other cases. With some of them,
prepositional alternatives (especially with ad, also with in) are available or even more
With a number of adjectives, the constituent in the dative usually refers to an entity
for whom something is beneficial (or the opposite), as in (a); what that something is
beneficial for (or the opposite) is usually expressed by a prepositional phrase (with ad)
or a purpose clause, as in (b) and (c) respectively. The more exceptional use of the
dative for this relation is illustrated by (c) (note the parallelism with the ut clauses)
and (d).171 Both relations may occur with the same adjective, as in (e), so one might
consider regarding these adjectives as three-place.
(a) . . . non esse oportunissimos situs maritimos urbibus eis quae ad spem
diuturnitatis conderentur . . .
(‘ . . . that sites on the sea-coast are not the most desirable for cities founded in the
hope of long life . . . ’ Cic. Rep. 2.5)
(b) . . . si locus opportunus ad eandem rem qua de re narrabitur fuisse
ostendetur.
(‘ . . . if it can be shown that the place was suitable for the events about to be narrated.’
Cic. Inv. 1.29)
171
OLD s.v utilis § 1 makes an explicit distinction between a dative ‘of advantage’ and ‘of purpose’. See
§ 10.70 for a discussion of two types of beneficiary adjuncts.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Denique ceterae res quae expetuntur oportunae sunt singulae rebus fere
singulis, divitiae, ut utare, opes, ut colare . . .
(‘In short, all other objects of desire are each, for the most part, adapted to a single
end—riches, for spending; influence, for honour . . . ’ Cic. Amic. 22)
(d) Tum Marius . . . collis duos propinquos inter se occupat, quorum in uno
castris parum amplo fons aquae magnus erat, alter usui opportunus . . .
(‘Then Marius . . . took possession of two neighbouring hills, one of which was too
small for a camp but had a large spring of water, while the other was adapted to his
purpose . . . ’ Sal. Jug. 98.3)
(e) Neque eadem loca aestiva et hiberna idonea omnibus (sc. animalibus) ad
pascendum.
(‘Again, the same localities are not equally suited in summer and winter to the
pasturing of all species.’ Var. R. 2.1.16)
Supplement (in alphabetical order by adjective):
. . . id in quo prudentia versaretur et quod assequi vellet aptum et accommoda-
tum naturae esse oportere . . . (Cic. Fin. 5.17); Eiusmodi autem studia ad delecta-
tionem quam ad veritatem videntur adcommodatiora. (Rhet. Her. 4.32); Haec
autem, ut dixi, genera dicendi aptiora sunt adulescentibus. (Cic. Brut. 326); Terra
quae vitibus apta est eadem quoque utilis est arboribus. (Col. Arb. 18.1); Vos
quidem . . . soletis . . . quam sint omnia in hominis figura non modo ad usum
verum etiam ad venustatem apta describere. (Cic. N.D. 1.47); Cf.: O dies hospitiis
lepide discriptos et apte ad consilium reditus nostri! (Cic. Att. 16.5.4); . . .
exploratores centurionesque praemittit qui locum castris idoneum deligant.
(Caes. Gal. 2.17.2); . . . tu occupes locum quem idoneum ad vim tuam iudicaris
. . . (Cic. Agr. 2.74); . . . tris libros in disputatione ac dialogo ‘de Oratore’, quos
arbitror Lentulo tuo fore non inutilis. (Cic. Fam. 1.9.23); . . . constituunt ut ii qui
valetudine aut aetate inutiles sint bello oppido excedant . . . (Caes. Gal.
7.78.1); . . . quod et acutum genus est et ad usus civium non inutile. (Cic. Fin.
1.12); . . . ut multas me etiam simultates . . . intellegam mihi non necessarias, vobis
non inutilis suscepisse. (Cic. Man. 71); Tibi enim nota sunt, mihi ad commemo-
randum non necessaria. (Cic. Div. 1.123); Quare satis admirari nequeo . . .
ceterarum artium minus vitae necessariarum repertos antistites . . . (Col.
11.1.10); . . . ratio res ad vitam necessarias sollerter consecuta est. (Cic. Leg.
1.26); Non ego omnino lucrum omne esse utile homini existimo. (Pl. Capt.
325); . . . dant utile lignum / navigiis pinus . . . (Verg. G. 2.442–3); (sc. stercus) . . .
quod non solum ad agrum utile, sed etiam ad cibum ita bubus ac subus, ut fiant
pingues. (Var. R. 1.38.2—NB: it is not clear where the datives bubus and subus
belong);
Just like verbs of abundance and lacking, a number of adjectives with that meaning
govern ablative second arguments. Some of them are presented in Table 4.22. The
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
genitive is also used with some of these adjectives, in the Augustan period probably in
imitation of Greek.172 With some adjectives of lacking, a prepositional phrase (with
ab) may be used as well; this seems to be the preferred expression when the argument
is a human being.173
172
For possible Greek influence, see Calboli (2009: 94). For the Greek adjectives involved, see K.-G.:
I.401.
173
See K.-St.: I.374.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
signis et tabulis pictis erant vacuae. (Cic. Ver. 1.55); . . . viduus pharetra / risit Apollo.
(Hor. Carm. 1.10.11-2);
genitives: . . . soli isto praetore omnium rerum immunes fuerunt . . . (Cic. Ver. 5.58);
Indoctus quid enim saperet liberque laborum / rusticus . . . (Hor. Ars
213–14); . . . orbas auxilique opumque huc / recepit ad se Veneria haec sacerdos me
et Palaestram. (Pl. Rud. 349–50); Integer vitae scelerisque purus / non eget Mauris
iaculis . . . (Hor. Carm. 1.22.1–2); . . . ager autem aridus et frugum vacuos ea
tempestate . . . (Sal. Jug. 90.1); . . . nec viduum pectus amoris habet. (Ov. Am. 3.10.18);
prepositional phrases: Aliquod tempus immune a legibus miseriae faciunt. (Sen.
Con. 2.5.16); . . . vacui, expertes, soluti ac liberi fuerunt ab omni sumptu, molestia,
munere. (Cic. Ver. 4.23); Tam inops autem ego eram ab amicis aut tam nuda res
publica a magistratibus? (Cic. Dom. 58); Usque adeo orba fuit ab optimatibus illa
contio ut . . . (Cic. Flac. 54); . . . forum Syracusanorum, quod introitu Marcelli purum
a caede (cj. Turnebus) servatum est . . . (Cic. Ver. 4.116—NB: the prepositional
expression is the normal one); Sed quae tandem est in hac urbe tanta domus ab
ista suspicione religionis tam vacua atque pura? (Cic. Har. 11);
More or less isolated is the use of the ablative for the second arguments of dignus
‘worthy’ and indignus ‘unworthy’.174 Examples are (c) and (d). Alternative case forms
are the genitive and (later) the dative, on the analogy of other adjectives. By contrast,
some other adjectives are occasionally found with the ablative, on the analogy of
dignus. See the Supplement.
174
K.-St.: I.398 relate it to the use of the ablative with verbs of competing (see § 4.53). Sz.: 79; 128
considers it an ‘instrumental ablative’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Adjectives that govern a second argument in the genitive belong to four semantic
classes. For each class, a number of verbs are presented in Table 4.23. Some of these
adjectives are also found with other cases and/or prepositional phrases. For adjectives
of desire, prepositional phrases (ad, in) are an alternative; for adjectives of sharing and
power, the dative or the ablative; for adjectives of fullness (and especially their
opposites), the ablative (see § 4.101).175
175
For a succinct historical sketch, see Sz.: 77–81.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The genitive is also used for the second arguments of present participles in -ans and
-ens, instead of the accusative the verbs normally take, when they express a permanent
quality and function as attribute or subject (or object) complement, and are not
anchored as secondary predicate to the time of the main verb of the clause.176 While
some of the verbs involved fit in with one of the meaning classes discussed above,
there is no obvious restriction on which participles can be used with the genitive.
Examples are (e)–(g). Cupio in (e) is occasionally found with a genitive object
(compare the adjective cupidus), but that is not a condition for the use of the genitive,
as the other examples show. Note in (f) the parallelism of amantem uxoris with the
adjectives siccum, frugi, and continentem.
(e) Nam hunc anulum <ego> ab tui cupienti huic detuli, hic <ad te> porro.
(‘I brought this ring from the woman who desires you to him, and he passed it on to
you.’ Pl. Mil. 1049)
176
For the history of this use of present participles, see Wölfflin (1904: 410–12). For Cicero’s use of the
present participle, see Laughton (1964).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(f) At scelesta ego praeter alios meum virum frugi rata, / siccum, frugi, con-
tinentem, amantem uxoris maxume.
(‘But I am such an idiot! I used to think my husband was better than others, sober,
good, moderate, full of love for his wife.’ Pl. As. 856–7)
(g) Quodsi is esset Panaetius qui virtutem propterea colendam diceret quod ea
efficiens utilitatis esset . . .
(‘But if Panaetius were the sort of man to say that virtue is worth cultivating only
because it is productive of advantage . . . ’ Cic. Off. 3.12)
Supplement:
. . . Acheruntis pabulum, / flagiti persequentem / stabulum nequitiae. (Pl. Cas.
159–61); Eru’ liberalis est et fugitans litium. (Ter. Ph. 623); Educta, ut par est, expars
malitiis, metuens sui . . . (Turp. com. 157); . . . multis etiam pater, optimus vir, nimium
retinens equestris iuris et libertatis videtur. (Cic. Planc. 55); Eques Romanus locuples,
sui negoti bene gerens . . . procuratorem Romae reliquit. (Cic. Quinct. 62); Nam
consules modesti legumque metuentes impediebantur lege . . . (Cic. Red. Sen. 4);
Horum igitur exprimere mores oratione iustos, integros, religiosos, timidos, perfe-
rentis iniuriarum mirum quiddam valet. (Cic. de Orat. 2.184); Sed existimant
plerique non haec adiuvantia causarum, sed has ipsas esse omnium causas . . . (Cic.
Tim. 50); Genus hominum salubri corpore, velox, patiens laborum. (Sal. Jug. 17.6);
(sc. flumen) . . . non tamen navium patiens est . . . (Liv. 21.31.11); Mores huius pecudis
probabiles habentur qui sunt propiores placidis quam concitatis, sed non inertes, qui
sunt verentes plagarum et adclamationum . . . (Col. 6.2.14); . . . ut est (sc. populus)
novarum rerum cupiens pavidusque. (Tac. Ann. 15.46.1); . . . rhetoricus quidam
sophista utriusque linguae callens . . . (Gel. 17.5.3); Bonitas amplius delicias adiecit
homini, ut, quamquam totius orbis possidens, in amoenioribus moraretur . . . (Tert.
Marc. 2.4.4);
These participial forms resemble adjectives. They can be used in the comparative and
superlative and they have negative counterparts formed by the prefix in-. Examples
are (h)–(j) respectively.
(h) Nihil est enim appetentius similium sui nec rapacius quam natura.
(‘For there is nothing more eager and more greedy than nature for what is like itself.’
Cic. Amic. 50)
(i) Fert enim graviter homo et mei observantissimus et sui iuris dignitatisque
retinens se apud te neque amicitia nec iure valuisse.
(‘Being most attentive to me and careful of his own rights and dignity, he is greatly
put out to find that neither friendship nor justice have counted with you in his
favour.’ Cic. Q. fr. 1.2.11)
(j) Et loquimur de Alexandro nondum merso secundis rebus, quarum nemo
intolerantior fuit.
(‘And we are speaking of an Alexander not yet overwhelmed with prosperity, which
none has ever been less able to bear.’ Liv. 9.18.1)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Sed homo nec doctior nec sanctior fieri potest nec tui meique amantior. (Cic. Att.6.1.12);
Virum fortissimum, integerrimum, inimicitiarum persequentissimum . . . (Rhet. Her.
2.29); PORCIA . . . CHRESTE / AMANTISSIMA·SUIS / MONUMENTUM·FECIT·SE / VIVA·SIBI·ET . . . (CIL
VI.9133 (Rome)); . . . altaque iactat / vulneris impatiens arrecto pectore crura . . . (Verg. A.
11.638–9); Omnis tamen externi frigoris tolerantior equino armento vacca est . . . (Col.
6.22.2); Lycurgus . . . fuit severissimarum iustissimarumque legum auctor et disciplinae
convenientissimae iis . . . (Vell. 1.6.3); Nihil est tam violentum, tam incontinens sui,
tam contumax infestumque retinentibus quam magna vis undae. (Sen. Nat. 3.30.6);
There are only few instances of present participles governing a genitive in Early Latin
of verbs that do not require a genitive object. I count six instances in Bennett’s Syntax
(all attribute or subject complement, it seems). This is less than the number of
present participles that govern their regular (non-genitive) case (all secondary
predicate, it seems). Apart from these instances there are also present participles
that govern an argument clause.177 The elaborate clausal use of participles starts in
the Classical period. See § 11.41 on the use of present participles as attribute and
Chapter 21 on their use as secondary predicate.178
Second arguments in the genitive are common with adjectives in -āx, a phenomenon
that is typical of poetry and from Livy onwards also of prose. Most of these adjectives
are related to a verbal stem. An isolated early instance is (k), which, however, Plautus
may have treated as an adjective of ‘knowing and memory’.
(k) . . . si huius rei me ess’ mendacem inveneris.
(‘ . . . if you find that I’ve lied to you about this matter.’ Pl. As. 855)
(l) . . . utiliumque sagax rerum et divina futuri / sortilegis non discrepuit sen-
tentia Delphis.
(‘ . . . and the thought, full of wise words and prophetic of the future, was attuned to
the oracles of Delphi.’ Hor. Ars 218–19)
(m) Victoremque cursu omnium aetatis suae fuisse ferunt . . . cibi vinique eundem
capacissimum.
(‘It is said that he vanquished all his mates at running . . . ; that he had also the
greatest capacity for food and wine.’ Liv. 9.16.13)
Supplement (in alphabetical order by adjective):
Tempus edax rerum, tuque, invidiosa vetustas, / omnia destruitis . . . (Ov. Met.
15.234–5); (sc. boves) . . . multi cibi edaces, verum in eo conficiendo lenti. (Col.
6.2.14); . . . mandragoram illum gravedinis compertae famosum et morti simillimi
soporis efficacem. (Apul. Met. 10.11.2); Haud dubie illa aetate, qua nulla virtutum
feracior fuit, nemo unus erat vir quo magis innixa res Romana staret. (Liv. 9.16.19);
Unde isti norunt illum quondam gloriae sequentis fugacissimum . . . (Sen. Ben.
4.32.4); . . . quid iam dudum decreveritis de isto asino . . . nunc etiam mendaci fictae
177
See Bennett (1914: II.82–7); Marouzeau (1910b: 13–22).
178
For the development of the present participle, see Rosén (1999: 98–109).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
debilitatis . . . (Apul. Met. 6.31.4); . . . gener, amicus, cunctis vitae officiis aequabilis,
opum contemptor, recti pervicax, constans adversus metus. (Tac. Hist. 4.5.2);
Minus hoc quam par erat nostri celebravere, omnium utilitatium et virtutum
rapacissimi . . . (Plin. Nat. 25.4); Iustum et tenacem propositi virum / non civium
ardor prava iubentium, / non voltus instantis tyranni / mente quatit solida . . . (Hor.
Carm. 3.3.1–4);
179
For more instances, see TLL s.v. ab 11.66ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
180
For more instances, see TLL s.v. ad 505.24ff., 509.37, 541.32ff.
181
For more instances, see TLL s.v. cum 1374.33ff.
182
For more instances, see TLL s.v. de 77.21ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
With dimensional adjectives such as altus ‘high’, crassus ‘thick’, latus ‘wide’, and
longus ‘long’ a constituent indicating the extent of space covered is normally
required.184 These constituents can be expressed with more or less the same means
as are used on the clause level (for which see § 4.84 and § 10.18–21), viz. bare
accusative (from Early Latin onwards), later also ablative and genitive, case forms
and—rarely—prepositional expressions. An example of an accusative noun is (a), of
an ablative (b), of a genitive (c). A rare instance of a prepositional expression
indicating the dimension is (d).
(a) . . . totum forum longum pedes V, latum pedes II S, crassum pedes I S.
(‘ . . . the whole area 5 feet long, 21/2 feet wide, 11/2 feet thick.’ Cato Agr. 18.3)
(b) Habeant spirae earum plinthum ad circinum, altam suae crassitudinis dimi-
dia parte . . .
(‘Let the bases have their plinths circular and half the height of the thickness of the
base . . . ’ Vitr. 4.7.3)
183
For more instances, see TLL s.v. in 745.73ff., 750.6ff., 761.79ff.
184
On the status of these extent in space expressions (argument or satellite?), see Bodelot (2011:
12–14).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
185
For examples and references, see Sz.: 34.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
CHAPTER 5
In Chapter 4, it is shown that some verbs that are normally regarded as two-place are
also used as one-place verbs with roughly the same meaning (for example bibo ‘to
drink’ in § 4.5). Conversely, one-place verbs are sometimes used with an ‘externalized’
object (for example, somnium somnio ‘to dream a dream’ in § 4.10). It is also shown
that verbs participate in states of affairs of various types: actions, positions, processes,
and states. In this chapter, another type of variation will be discussed in which verbs
may be involved, viz. active/passive variation, reflexivity, and intransitivization.1
In § 3.12, a distinction is made between active and passive forms versus the meanings
that are usually attached to these forms. There is no one-to-one relation between
active forms and ‘active’ meanings, nor between passive forms and ‘passive’ mean-
ings. Verbs with active forms may refer to an action (for example, curro ‘to run’), to a
position (for example iaceo ‘to lie’), or to a process (for example, vapulo ‘to be
beaten’). Conversely, verbs with passive forms may refer to a process (for example,
morior ‘to die’) or to an action (for example, hortor ‘to urge’). The following sections
deal in detail with the use(s) of passive forms and the reasons for choosing the active
or the passive for those verbs where the choice exists (see § 5.3 for details).
The function that is most commonly ascribed to the active/passive variation is
illustrated by the textbook examples (a) and (b). In these examples the same state of
affairs (an action performed by an agent on a patient) is presented from two different
perspectives: from the perspective of the agent in (a) and from the perspective of the
patient in (b) (complications will be discussed in this section). With the verb laudo
1
A typological treatment of the passive can be found in Siewierska (1984). For historical consider-
ations, see Kiss (1998).
Different from this instance of TRUE PASSIVE use of the passive form are instances like
(c)–(f). In (c) and (d), the agent is performing an action in which it is itself involved as
the patient. In this Syntax, this use of the passive will be called AUTOCAUSATIVE (for the
term, see § 5.19). It is often called ‘middle’, ‘mediopassive’, or ‘reflexive’ in other
grammars. Related is the use of the passive for an action which the subject ‘gets done
or lets be done to him’,2 as in (e).
(c) Iam homo in mercatura vortitur.
(‘The boy’s already engaging himself in business.’ Pl. Mos. 639)
(d) . . . postridie redire iussit. Lavari se velle . . .
(‘ . . . he told them to come back the next day, as he wanted to take a bath . . . ’ Cic. Att.
10.13.1)
(e) . . . profecto numquam conquiescam neque defetigabor ante quam illorum
ancipites vias rationesque et pro omnibus et contra omnia disputandi
percepero.
(‘ . . . unquestionably, I shall never come to a standstill or give up from exhaustion
before I have a full grasp of the school’s twofold method and its system of arguing
both pro and contra about every proposition.’ Cic. de Orat. 3.145)
Instances like (f) are also of a different nature. Here, no external agent or cause is
responsible for the movements of the terra, nor is the terra itself. Note the contrast
with stare. For this use of the passive the term DECAUSATIVE is used in this Syntax.
(f ) Hicetas Syracosius, ut ait Theophrastus, caelum solem lunam stellas supera
denique omnia stare censet, neque praeter terram rem ullam in mundo
moveri.
(‘The Syracusan Hicetas, as Theophrastus asserts, holds the view that the heaven,
sun, moon, stars, and in short all of the things on high are stationary, and that
nothing in the world is in motion except the earth.’ Cic. Luc. 123)
What these three uses of the passive have in common is that in each of them an entity
is involved as patient of a process or a state. They are different, however, with respect
to the involvement of an agent or a cause and with respect to the identity of the agent
or cause, if one is involved. This is shown in Table 5.1.
2
Formulation taken from Roby (1882: I.173).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 5.1 The involvement of an agent or cause with the three types of use of the
passive form
True passives Autocausative passives Decausative passives
+ + – +/– agent or cause
– + – agent or cause = patient
In the examples of true, autocausative, and decausative passives discussed above, the
clause has a subject constituent. The common term for such passives that have a subject
is PERSONAL PASSIVE. In the following examples, by contrast, which have a third person
singular passive form, there is no subject. This type of passive is called IMPERSONAL. Ex.
(g) has a form of one-place eo ‘to go’. Ex. (h) has a form of two-place amo ‘to love’, used
absolutely (see § 4.5), and a form of egeo ‘to feel the absence of ’, also used absolutely.
The latter verb does not govern an accusative and has no personal passive forms. In the
case of amo, both the personal and the impersonal passive are allowed.
(g) Eatur.
(‘Let’s go, then.’ Ter. Hau. 743)
(h) Quid agitur, Calidore? # Amatur atque egetur acriter.
(‘What’s going on, Calidorus? # Love and a dire financial situation are what’s going
on.’ Pl. Ps. 273)
There is a fundamental difference between the personal uses of the passive and the
impersonal use: whereas the function of the personal uses is to present a process or
state (often as the result of an action, caused by an animate or inanimate entity) from
the perspective of the patient, in the impersonal passive there is no patient involved at
all. The function of the latter is rather to present an action without reference to a
specific (human) agent. One might even say that semantically the impersonal passive
is no passive, in spite of its passive form.3
Table 5.2 has data on the relative frequency of these four uses in Plautus, Cicero,
and Livy.4 The table shows considerable differences (marked in bold), which are
discussed further in later sections.
There is another use of the personal passive that is not discussed in this chapter, but
in § 15.110, which concerns the personal use of the passive in the so-called nomina-
tive and infinitive construction, as in (i).
(i) Ut demonstratae sunt mihi, hasce aedis esse oportet / Demaenetus ubi dicitur
habitare.
(‘The way it was described to me, it ought to be this house here where Demaenetus is
said to live.’ Pl. As. 381–2)
3
See Wehr (1995: 29–36).
4
Taken (with some simplification) from Joffre (1995: 86–8). She does not make a distinction between
autocausative and decausative. (Deponents are excluded.)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 5.2 The relative frequency of the uses of the passive in three Latin texts (by
percentage)
Plautus Aul. (N = 79) Cicero Ver. IV (N = 386) Livy II (N = 601)
Impersonal passives 2.5 2 14.5
Auto- + decausative 5 4 4
passives
True passives without 86 75 70.5
an agent or cause
True passives with an 6.5 19 11
agent or cause
The term ‘voice’ for the active/passive distinction does not go back to Antiquity. The
Greek term was diathesis, the Latin equivalents species verbi or genus verbi. Some
scholars use the Greek term to refer to the semantic differences mentioned above, as
opposed to ‘voice’ for the morphological forms.5
5
So Touratier (1994: 169) and Baños (coordinador) (2009: 381–2), with reference to articles of his,
among which Baños (1997). For genus verbi, see TLL s.v. genus 1901.79ff.
6 7
See Flobert (1975) and Calboli (1990). Ernout (1908–9).
8
See Calboli (1990: 105–13).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Invidetur enim commodis hominum ipsorum, studiis autem eorum ceteris
commodandi favetur.
(‘For men’s private gains breed jealousy, while their zeal for others’ service is
applauded.’ Cic. de Orat. 2.207)
Thirdly, there are a number of verbs that have active forms but function as suppletive
passive forms for verbs with active forms that cannot be passivized. The best-known
example is the verb fio. It has active forms (the active infinitive form fiere is attested
three times, e.g. Ennius Ann. 15V=11S). Apart from its copular use, it also functions
as the passive counterpart of the infectum forms of facio ‘to do’, ‘to make’, as in (b). By
contrast, in the perfectum passive forms of facio are used, as in (c). Note the agent
expression a me. Similarly, pereo ‘to perish’ and veneo ‘to be sold’ have active forms,
yet function as passive counterparts of perdo ‘to destroy’ and vendo ‘to sell’, respect-
ively. An example is (d).
(b) Fiet sedulo.
(‘It shall be done diligently.’ [Loeb translation ‘I’ll do my best.’] Ter. Ph. 228)
(c) Nunc demum a me insipienter factum esse arbitror, / quom rem cognosco.
(‘At last I realize I’ve acted foolishly, now that I understand the matter.’ Pl. Mil.
561–2)
(d) Nam si haec non nubat, lugubri fame familia pereat.
(‘Yes, if she didn’t marry, our household would perish from sorrowful hunger.’ Pl.
Cist. 45)
The combination of active forms and passive meanings resulted occasionally in
formal adjustments. The passive infinitive form fieri is the normal one from Early
Latin onwards. The passive forms fitur and fiebantur are attested for Cato. Plautus
Per. 577 has the passive infinitive veniri. Similarly, adjustment of the form to the
passive meaning can be seen in clueor instead of clueo ‘to be spoken of ’ in Pl. Ps. 918.9
From Antiquity onwards, grammarians have been interested in the fact that there
are verbs that only occur with active endings. Some grammarians used the term
‘neutra’ or ‘neutralia’ for this phenomenon. They also distinguished a few verbs that
occur in the active and as deponents with the same meaning, such as amplecto
(Pl. Rud. 816) and amplector ‘to embrace’.10
Fourthly, there are also verbs with passive and no active forms. One type concerns the
so-called DEPONENT verbs, such as nascor ‘to be born’, utor ‘to use’ (there are active
forms in Early Latin), and sequor ‘to follow’. Surprisingly, verbs from roughly the
same semantic field such as eo and sequor differ in voice: the states of affairs in which
they occur are usually actions and the active form is what one expects. Details are in
§ 5.33. A second, much smaller type is formed by the so-called passiva tantum, verbs
9
This phenomenon is called ‘innere Attraktion’ by Löfstedt (1933: II. 124). Material for veneo in
Neue-W.: III.100.
10
For Greek and Roman ideas about voice, see Flobert (1975: 1–20).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
with passive morphology and (true or autocausative) passive meaning. Examples are
crapulor ‘to get drunk’, plector ‘to be beaten’ (plecto is only found from Tertullian
onwards), and praetervehor ‘to travel past’.11
Another difficulty involved in calculating the frequency of active and passive forms
stems from the fact that verbs which do have active and passive forms cannot be used
in the passive in every state of affairs in which they occur. An example is (e). Nor can
every passive clause be converted into an active one, as in (f).
(e) Terentia magnos articulorum dolores habet.
(‘Terentia has a bad attack of rheumatism.’ Cic. Att. 1.5.8)
(f) Gallia est omnis divisa in partes tres . . .
(‘Gaul as a whole is divided into three parts . . . ’ Caes. Gal. 1.1.1)
Finally, the distinction between the various types of passive described in § 5.2 is
usually not made.
In many studies on the relative frequency of passive forms, these facts have not
been taken into consideration. If, however, one focuses on verbs that can, in theory, be
used in both voices (that is: ‘convertible’ verbs), different results emerge:12 18% of the
‘convertible’ verb forms in Plautus’ Mos. are passive. In Cicero’s Rep., 43% of such
verb forms are passive.13
Among different types of texts there is considerable variation in the frequency of
(convertible) active and passive forms. In part, this is related to the subject matter: the
difference in frequency just mentioned is closely tied to the fact that the Mostellaria is
a play about problems between young people and their parents while the de Republica
is a philosophical dialogue about the organization of human society: the former
is accordingly more concrete, the latter more abstract. Another element (partly
related to text type) that correlates with the frequency of passive forms is sentence
complexity.14 Most passive infinitives are used in accusative and infinitive clauses,
partly to avoid ambiguity between the agent and the patient; and passive perfect
participles, for which there is no active counterpart (leaving deponents out of
account), are a favourite element in the construction of periodic sentences. This is
shown in (g)–(i). In (g), the confusion of agent and patient is avoided by using the
passive infinitive and the ab phrase. In (h), the secondary predicate (perfect passive
participle) contemptus is expanded with the agent expression abs te. Ex. (i) shows a
periodic sentence with three passive participial clauses. Given the variety that is thus
created, it is difficult to measure the individual stylistic contribution to the frequency of
the passive and, consequently, difficult as well to determine diachronic developments.15
11
For a list of instances and the criteria to separate them from deponent verbs, see Flobert (1975:
409–10).
12
For the methodology, see Pinkster (1984, 1985a); Joffre (1995: 81–2).
13
In Flobert (1975), 30% of Cicero’s verb forms are counted as passive.
14
See Rosén (1999: 126, 129), with references to Nausester (1907); de Melo (2007b).
15
For Sallust, see Calboli (1990: 105).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 5.3 presents some data on sentence complexity and on the use of the active and
personal passive in a play of Terence, one book of the didactic poetry of Lucretius, and
epigraphical poems.16 The higher complexity in Lucretius correlates with the fre-
quency of the passive, as well as with the frequency of agent and cause expressions.
Table 5.3 Sentence complexity and the use of active and passive
Terentius Eu. Lucretius I Carmina epigraphica
Number of words per sentence 6.45 17.26 7.54
% of finite subordinate clauses 31.89 58.60 31.29
out of all clauses
% of passives of convertible verbs 9.80 26.25 10.88
% of agentful passives 2.91 15.87 8.33
16
Assembled from de Melo (2007b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The passive clause in (b) is often regarded as merely the reversal of the active clause
in (a). This has led to the passive being called ‘a luxury feature’ of the language.17
However, since with most true passives there is no overt agent or cause (see
Table 5.2), and since quite often with passive verb forms no agent or cause is implied
at all (see the decausative passives discussed in § 5.18), it seems better to deal with
each expression in its own right. Whereas in (a) and (b) the relationship between the
verb and the patient is essentially the same in the active and the passive, the status of
the agent or cause is different. In the case of the passive the agent or cause is rarely
expressed, whereas in the case of the active, it is obligatory. Moreover, the two expres-
sions are used under different semantic and/or pragmatic conditions (see § 5.10).18
In what follows, a distinction is made between two types of true passive clauses: (i)
clauses in which an agent or cause is expressed (these clauses are therefore AGENTFUL)
and (ii) clauses for which an agent or cause is present in the surrounding context or
the communicative situation (on the stage, for example) (these clauses are AGENTLESS).
Examples of agentful true passive clauses are (c)–(g). The common form for an agent
is a prepositional phrase with ab, as in (c) and (d); for a cause, the bare ablative case,
as in (e) and (f). There are several other ways agency can be expressed; one is
illustrated by the adverb puplicitus ‘publicly’ in (g). Details can be found in § 5.8.
(c) Numquam ego me tam sensi amari quam nunc ab illa muliere.
(‘I never realized till now that I was loved so much by that girl.’ Pl. Mil. 1202)
(d) Sic a me consul designatus defenditur ut eius nulla fraus, nulla avaritia, nulla
perfidia, nulla crudelitas, nullum petulans dictum in vita proferatur.
(‘My defence of the consul-elect is that his whole life is free from deceit, greed,
treachery, cruelty, or intemperate language.’ Cic. Mur. 14)
(e) Invitus do hanc veniam tibi, / nisi necessitate cogar.
(‘I give you this indulgence unwillingly, but I’m forced by necessity.’ Pl. Epid. 730–1)
(f) . . . sic in viris fortibus virtus virtute superatur.
(‘ . . . so, when brave men are rivals, merit is overcome by merit.’ Cic. Phil. 14.18)
(g) Ita peregre adveniens / hospitio puplicitus accipiar.
(‘So on my arrival from abroad I’d get a state reception.’ Pl. Am. 161–2)
Examples of agentless true passive clauses are (h)–(n).19 In (h), Euclio suspects that
there are thieves in his house. Their identity is unknown to him. In (i), Menaechmus
uses a passive form in order to avoid committing himself, a common feature in
17
See Wackernagel (1920: I.135): ‘Mit Recht hat man das Passiv als einen Luxus der Sprache
bezeichnet, weil der passivische Satz nichts anderes darstellt als die Umkehr des normalen aktivischen
Satzes’.
18 19
See Flobert (1975: 535–44). For agentless passive constructions, see Vester (1985).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Roman comedy. Also regarded as true passive clauses are those in which the agent or
cause is ‘the general public’ (more or less equivalent to they in English, on in French,
or man in German). Examples are (j)–(l) with dicitur, perspicitur, inventus est. The
agent or cause may also refer to ‘the usual institution’, as in (m). In didactic texts,
from Cato onwards, the agent of a passive clause is the public to which the text
addresses itself, as in (n) (often with a directive illocutionary force—see § 6.4).20
(h) Numnam ego compilor miser? . . . Aurum rapitur, aula quaeritur.
(‘Am I being robbed? Poor me! . . . My gold’s being stolen, a pot’s being looked for.’
Pl. Aul. 389–92)
(i) Sciti’n quid ego vos rogo? /# Quid? # Praeconium mi ut detis. # Dabitur.
(‘Do you know what I want of you, sirs? # What? # To let me be auctioneer. # You
shall be.’ Pl. Men. 1154–5)
(j) Ut demonstratae sunt mihi, hasce aedis esse oportet / Demaenetus ubi
dicitur habitare.
(‘The way it was described to me, it ought to be this house here where Demaenetus is
said to live.’ Pl. As. 381–2)
(k) Quamquam minus etiam perspicitur eorum virtus et integritas qui ad hanc
rem adulescentuli, quam qui iam firmata aetate descendunt.
(‘Although there is less assurance of the virtue and integrity of those who enter these
lists in extreme youth than of those who do so when they are of maturer years.’ Cic.
Ver. 3.3)
(l) Nemo inventus est tam amens qui illud argentum tam praeclarum ac tam
nobile eriperet . . .
(‘Not one of them has shown himself such a madman as to carry off those beautiful
pieces of plate . . . ’ Cic. Ver. 4.44)
(m) Apponitur iis tamen accusator Naevius Turpio quidam qui C. Sacerdote
praetore iniuriarum damnatus est, homo . . .
(‘Well, a man was at least put up to prosecute the prisoners—one Naevius Turpio,
who had been found guilty of assault when Gaius Sacerdos was governor . . . ’ Cic.
Ver. 5.108)
(n) Ideoque colubra ipsa tuto estur.
(‘Hence the snake itself may be safely eaten.’ Cels. 5.27)
Supplement:
Avoiding commitment: Ecce autem in benignitate hoc repperi, negotium! / Dabitur
opera. (Pl. Trin. 389–90); Fiat, geratur mos tibi. (Pl. As. 40);
Institutional agent: Trecentae possunt causae colligi. / Non domi est, abiit ambula-
tum, dormit, ornatur, lavat, / prandet, potat. (Pl. Mil. 250–2);21
20 21
See Calboli (1990: 111–14); Rosén (1999: 131). See Rosén (1999: 135–6).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The relative frequency of decausative and true passive clauses in Plautus’ Mostellaria
and in Cicero’s de Republica is roughly the same. There is, however, a considerable
difference in the relative frequency of agentless and agentful true passive clauses: in
Plautus’ text they are agentless much more often than they are in Cicero’s. On closer
inspection, this correlates with the semantic properties of the agents and causes: the
de Republica has more inanimate causes than the Mostellaria. This, in turn, relates to
the difference in text type: interaction between human beings in Plautus, as against a
philosophical discussion about the organization of human society in Cicero.22
Explicit agents are more frequent in prose than in poetry and more frequent in text
types with more complex sentences, as is shown in Table 5.3. In general, the frequency
of their occurrence increases over time from Plautus onwards, which may reflect,
apart from differences in text type, sentence complexity, and individual stylistic
preferences, a genuine diachronic development.23 The distribution of ab phrases is
uneven in relation to the passive forms with which they occur: they are infrequent
with finite passive forms of the infectum stem, relatively frequent with present passive
infinitives (within accusative and infinitive clauses), and most frequent with partici-
pial forms in -us, -a, -um (many of these functioning as secondary predicates or in
ablative absolute clauses). In Ovid, 32% of passive forms of the perfectum stem have
an explicit agent as against 27% of forms of the infectum stem.24
One reason for adding an ab phrase to a process or state that is itself presented
from the perspective of the patient is to identify the agent—for example, when in the
context or situation there are multiple candidates that could be interpreted as agents.
However, when there is no reasonable doubt about the identity of the agent, the
reason to add it explicitly must be to give it more prominence.25
Passivization of two- and three-place verbs is formally allowed when the object is in
the accusative case. However, passivization is difficult or non-existent if the second
argument of a two-place verb is not a patient or experiencer and (related to this) if the
first argument is not a human agent. In the terminology of this Syntax, this means
that ‘control’ is an important feature. Thus, habeo ‘to have’, in its ‘possessive’
meaning, is rarely found in the passive with a human first argument (but see a
Laide in (a)), and not at all with inanimate first arguments. There are no passive
22
See Pinkster (1985a) for details.
23
For statistical data, see Nausester (1907: 153–68, 1935: 30–1); Flobert (1975: 539–41); Rosén (1999:
128–30). For the interpretation of the development, see Calboli (1990: 110–12) and Rosén (as earlier in
this note).
24
See Hoffmann (1997a: 72–90).
25
For Cicero’s preference of agentful passive translations for active Greek ones in the Timaeus, see
Rosén (1996a: 543–8). For the factors determining the expression of the agent in Cicero’s pro Milone, see
Santos (2009).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
counterparts for (b) and (c). Habeo differs in this respect from possideo, which refers
to real (legal) possession, as in (d).
(a) Sed tamen ne Aristippus quidem ille Socraticus erubuit cum esset obiectum
habere eum Laida. ‘Habeo (sc. Laidem)’ inquit, ‘non habeor a Laide.’ (Graece
hoc melius. Tu, si voles, interpretabere.)
(‘But, after all, even the Socratic Aristippus himself did not blush when he was
taunted with having Lais as his mistress: “I have Lais as my mistress”, he said, “I’m
not had by her.” (It’s better in the Greek. Make your own rendering, if you care to.)’
Cic. Fam. 9.26.2)
(b) Sine pinnis volare hau facile est. Meae alae pennas non habent.
(‘It’s not easy to fly without feathers; my wings don’t have feathers.’ Pl. Poen. 871)
(c) . . . quod plus salis plusque leporis [hodie] / habeat.
(‘ . . . which has more wit and more charm.’ Pl. Cas. 217–18)
(d) Si res peculiaris in potestatem servi redierit . . . incipit . . . a domino possideri.
(‘If a particular thing, once taken away, returns to the power of the slave, it comes
under the jurisdiction of the master.’ Julian. dig. 47.2.57[56]2)
26
More examples in K.-St.: I.102–4.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The second type concerns coordination of a verb that governs an accusative object
and a verb with which a non-accusative object must be understood. An example is (i).
(i) . . . coliturque ea pars et habitatur frequentissime.
(‘ . . . and is also a crowded and thickly inhabited part (of the city).’ Cic. Ver. 4.119)
The third type is found in poetry and poeticizing prose. Examples are (j)–(m), where
three of the verbs normally govern a dative, whereas triumpho in (l) normally governs
a prepositional phrase (de).27 The fact that in Greek there is no restriction on
passivization of non-accusative governing verbs may have played a role.28
(j) . . . et fatis numquam concessa moveri / apparet Camerina . . .
(‘ . . . Camerina—fate forbade that she ever be disturbed—is seen . . . ’ Verg. A 3.700–1)
(k) Quid tamen hoc prodest, vetiti si lege severa / credor adulterii composuisse notas?
(‘Yet of what avail is this if men believe that I have composed directions for that
adultery which is forbidden by stern laws?’ Ov. Pont. 3.3.57–8)
(l) Addam urbes Asiae . . . / bisque triumphatas utroque ab litore gentes.
(‘I will add Asia’s cities . . . and the twice-triumphed nations from either shore.’ Verg.
G. 3.30–3)
(m) . . . ad Graeciam ire permissus est.
(‘ . . . he was allowed to go to Greece . . . ’ Amm. 15.2.15)
Passivization is very rare with idiomatic expressions like manum consero ‘to join in
close battle’ and pedem fero/refero ‘to advance’, ‘to return’, although the accusative
constituents look like normal direct objects and the states of affairs are controlled.29
Rare instances are (n) and (o).
(n) . . . non sine illa etiam spe cum prohibentibus opus Romanis manum posse
conseri . . .
(‘He was not without the hope also that he could engage the Romans if they tried to
prevent the work . . . ’ Liv. 25.11.3)
(o) Tum primum referri pedem atque inclinari rem in fugam apparuit. Tum
capi, occidi Samnis.
(‘Then for the first time were there signs of giving way and the beginning of a rout.
Then were the Samnites captured or slain.’ Liv. 7.33.15)
Perfect passive participles exist for verbs for which no passive infectum forms are attested.
This phenomenon was already observed by Latin grammarians, as in the following
quotation: ‘INACCESSAM participium sine verbi origine, “accedor” enim non dicimus;
sed tale est, ut placita nupta triumphata regnata.’ (Servius ad Verg. A. 8.195)
27
More examples in K.-St.: I.102.
28
For examples, see K.-G.: I.124.
29
Fugier (2002: 134) holds that the passive is excluded in such instances.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Restrictions on the use of the gerundive and its ‘passive’ meaning are discussed in
§ 5.37. Passivization of three-place verbs governing two objects like doceo ‘to teach’ is
dealt with in §§ 4.73–4.74.
True passive verb forms (passive participles much more often than finite forms) can
be combined with accusative constituents. Their syntactic status is much disputed. In
corresponding active expressions they usually function as object, but in the passive
they seem to be optional. The phenomenon is almost confined to poetry. It is
relatively common with verbs of hitting, wounding, binding with the (usually
human) patient as the subject and the part (of the body) in the accusative, but it
occurs with other classes of verbs as well. Imitation of semantically similar Greek
constructions or extension of the possibilities of Latin is very likely.30
The oldest example (if genuine) of a passive participle of a verb of hitting in
combination with an accusative constituent is (a). The earliest attested prose example
is (b). Note the agent expression ab hostibus in (b) and the cause expression lacte mero
in (c). In (d), the instrument is expressed (volucri sagitta). In (e), the instrument is in
the accusative (lora), whereas the body part is a prepositional satellite (per pedes
tumentis).31
(a) . . . perculsi pectora Poeni . . .
(‘ . . . the Carthaginians were stricken in their hearts . . . ’ Enn. Ann. 311V=310S)
(b) Dedecores inultique terga ab hostibus caedebantur.
(‘Disgraced and unavenged, they were struck in the back by the enemy.’ Sal. Hist. frg. 3.24)
(c) . . . nova proles / artubus infirmis teneras lasciva per herbas / ludit lacte mero
mentes perculsa novellas.
(‘Through this a new brood with tottering legs sports wanton among the soft grass,
their baby hearts thrilling with the pure milk.’ Lucr. 1.259–61—tr. Bailey)
(d) At te, Nesse ferox, eiusdem virginis ardor / perdiderat volucri traiectum terga
sagitta.
(‘But, O savage Nessus, a passion for the same maiden utterly destroyed you, pierced
through the body by a flying arrow.’ Ov. Met. 9.101–2)
(e) (sc. Hector) aterque cruento / pulvere perque pedes traiectus lora tumentis.
(‘ . . . black with gory dust, pierced with thongs through his swollen feet.’ Verg. A.
2.272–3)
The regular expression is with the body part as the object in the accusative (or subject
nominative in the passive) and a satellite in the (sympathetic) dative, as can be seen in
30
For a detailed classification, different from the one presented here, see Courtney (2004).
31
Collections of instances are Landgraf (1898/9) and Müller (1908: 127–31).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(f) (for details see § 10.96). In Late Latin, active verb forms with two accusative objects
are attested (probably also under the influence of Greek).
(f) Tum Tito Balventio . . . utrumque femur tragula traicitur.
(‘At this point, both of Titus Balventius’s thighs were pierced by a dart.’ Caes. Gal.
5.35.6)
Supplement:
In quo proelio Pacideius graviter pilo per cassidem caput ictus . . . (B. Afr. 78.10);32
. . . poenae / quam quondam silici restrictus membra catena / persolvit . . . (Catul.
64.295–7); Ecce manus iuvenem interea post terga revinctum / pastores magno ad
regem clamore trahebant / Dardanidae . . . (Verg. A. 2.57–9); . . . illa usque minatur /
et tremefacta comam concusso vertice nutat . . . (Verg. A. 2.268–9—NB: inanimate);
Aeneia puppis / prima tenet, rostro Phrygios subiuncta leones . . . (Verg. A.
10.156–7—NB: inanimate); O ego laevus, / qui purgor bilem sub verni temporis
horam! (Hor. Ars 301–2); . . . miles ait, multo iam fractus membra labore. (Hor. S.
1.1.5); . . . merdis caput inquiner albis / corvorum . . . (Hor. S. 1.8.37–8); Figitur hinc
duplici Gryneus in lumina ramo, / eruiturque oculos. (Ov. Met. 12.269–70); Ut vero
Hannibal ipse . . . adversum femur tragula graviter ictus cecidit . . . (Liv. 21.7.10);
P. Servilium magistrum equitum servaverat, ipse vulneratus umerum, femur.
(Plin. Nat. 7.103); Delmatico etiam vulnera excepit, una acie dextrum genu
lapide ictus, altera et crus et utrumque brachium ruina pontis consauciatus.
(Suet. Aug. 20.1);
However, the use of accusative constituents is not confined to the cases described
above; it also occurs with other verbs and/or with other nouns than body parts.
Examples are (g)–(i). The reason for using the passive may be the same preference for
a human subject. The cause or instrument is often expressed in the ablative.
(g) . . . Lavinia virgo, / visa . . . omnem ornatum flamma crepitante cremari, /
regalisque accensa comas, accensa coronam / insignem gemmis.
(‘ . . . the maiden Lavinia, she was seen . . . to catch fire in her long tresses, and burn
with crackling flame in all her headgear, her queenly hair ablaze, ablaze her jewelled
coronal.’ Verg. A. 7.73–5)
(h) Si curatus inaequali tonsore capillos / occurri, rides.
(‘If, when some uneven barber has cropped my hair, I come your way, you laugh.’
Hor. Ep. 1.1.94–5—NB: ablative agent expression)
(i) Hanc ego, nondum etiam sensus deperditus omnis, / molliter impresso
conor adire toro.
(‘Not yet were all my senses gone, and I tried by leaning gently on the couch to reach
her.’ Prop. 1.3.11–2)
32
On imitation of Greek in the de Bello Africo, see Adams (2005b: 93–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Quem se intorquentem lanians rostroque cruentans / iam satiata animos, iam
duros ulta dolores / abicit . . . (Cic. Mar. 5–7 = 17.5–7C); . . . ut eam devinctam lumina
somno / liquerit immemori discedens pectore coniunx. (Catul. 64.122–3); Ipse autem
caeca mentem caligine Theseus / consitus oblito dimisit pectore cuncta . . . (Catul. 64.
207–8); His animum arrecti dictis et fortis Achates / et pater Aeneas iandudum
erumpere nubem / ardebant. (Verg. A. 1.579–81); . . . multoque iacebat / membra deo
victus. (Verg. A. 9.336–7); Hinc tibi . . . saepes Hyblaeis apibus florem depasta salicti /
saepe levi somnum suadebit inire susurro. (Verg. Ecl. 1.53–5); . . . adsidua Tiphys
vultum lassatus ab Arcto. (V. Fl. 1.419—NB: agent expression); . . . ipse umerum et
sagitta praestrictus et timens, ne . . . (Amm. 24.6.13—NB: the first et is odd. Keller-
bauer reads et ipse—see den Boeft et al. ad loc.);
Expressions of decoration with either the surface or the decoration in the accusative
form a special class. Examples of the first type are (j) and (k). Ex. (l) illustrates the
second type.
There are a few instances of third arguments that in the active are in the dative case,
but become subject in the passive. An example is (m) (unless it belongs to the
decoration type).
In principle, Latin makes a distinction in the way agent and cause constituents are
marked in passive clauses. For agents in these clauses, a prepositional phrase with ab
(a, abs) + ablative case may be used, as well as the dative case (with some restrictions);
for causes, the bare ablative case is used.33 Examples are (a) and (b). Ab + ablative and
the bare ablative are also used with processes and states in non-passive clauses.
Examples of ab are (c)–(e), with verbs, and (f), with an adjective.34 Note that in (d)
and (e) ab + ablative is used for an inanimate cause, a phenomenon not uncommon in
didactic prose (see the Supplement). Poets extend the range of verbs with which ab
phrases are used. Two additional examples of causes are (g) and (h). With verbal
nouns, agent expressions with ab are very rare. An example is (i).35
(a) . . . optatus hic mi / dies datus hodie est ab dis . . .
(‘ . . . this day of my desire has been given to me today by the gods . . . ’, Pl. Per. 773b-4)
(b) Itaque nos ventisque fluctibusque / iactatae exemplis plurumis miserae
perpetuam noctem.
(‘And so we were tossed about by winds and waves in all sorts of ways the entire
night, poor us.’ Pl. Rud. 369–70)
(c) Nos hic . . . P. Sullam patrem mortuum habebamus. Alii a latronibus, alii
cruditate dicebant.
(‘Here . . . we have a death to talk about—P. Sulla senior. Some say it was bandits,
others overeating.’ Cic. Fam. 15.17.2)
(d) . . . nihil enim valentius esse a quo intereat.
(‘ . . . since nothing stronger exists to cause it to perish.’ Cic. Ac. 1.28)
(e) Atque hi tanti ignes . . . ita prosunt ut si mota loco sint conflagrare terras
necesse sit a tantis ardoribus . . .
(‘And these vast . . . fires . . . are actually beneficial, though with the qualification that
were their positions altered, the earth would inevitably be burnt up by such enor-
mous volumes of heat . . . ’, Cic. N.D. 2.92)
33
For a discussion of the range of nouns that function as cause in passive clauses, see Steele (1921).
34
For more instances, see TLL s.v. ab 29.48ff. For agent expressions with other types of expressions, see
Shalev (2001).
35
See Wiberg (1956: 155–64); Bertocchi and Maraldi (1990: 70–1, 93–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
36
For Lucretius’ ‘otiose’ use of ab, see Bailey (1949: I.106).
37 38
For more examples, see OLD s.v. unde § 2. See Tesařová-Nováková (1987: 94).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The bare ablative is also used to mark a gerund clause as the cause of the process in
the main clause. An example is (l). For more examples, see § 15.136.39
Examples of dative agents in true passive clauses are (m)–(p) (there are no dative causes).
The dative (the so-called dativus auctoris) is the common agent expression in gerundival
clauses (see § 5.40 for discussion). With the passive, it is confined to the perfectum forms
in Plautus, as in (m). In Cicero’s works infectum forms and non-pronominal agents are
relatively rare. Caesar does not use them. From Accius (and possibly Ennius) onwards,
poets in particular (and poeticizing prose authors like Livy and Tacitus) use it freely also
with infectum forms, as in (n), though less frequently than with perfectum forms. Ex. (n)
also illustrates the use of other words than pronouns as dative agents which is not
attested in Plautus. Influence from Greek authors is not unlikely.41
(m) . . . qui ubi tibi istam emptam esse scibit . . .
(‘... as soon as he knows that this girl has been bought by you . . . ’ Pl. Epid. 154)
(n) Ergo med Argos referam, nam hic sum gnobilis, / ne cui cognoscar noto.
(‘Then I’ll betake myself to Argos, lest I be recognized by one who is known to me.
For here I am notable.’ Acc. trag. 283–4)
(o) Tibi enim consulatus quaerebatur, Metello paternus honos et avitus
neglegebatur.
(‘ . . . while you were ambitious for the consulship, Metellus was indifferent to an
office held by his father and grandfather before him!’ Cic. Ver. 3.43)
(p) Est igitur ambulantibus ad hunc modum sermo ille nobis institutus . . .
(‘Well then, as we walked about, we started our conversation in this way . . . ’ Cic.
Tusc. 2.10)
39
See Tesařová-Nováková (1987: 95).
40
For a detailed discussion, see Hoffmann (1997a: 61–90).
41
The most complete statistical information can still be found in Tillmann (1881), with statistical
tables at pp. 109 and 111, and in Landgraf (1898/9). The Early Latin material can be found in Bennett
(1914: I.171–2).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Perfectum forms: Hic enim rite productu’st patri. (Pl. Bac. 457); Nam mihi /
decretum’st renumerare iam omne aurum patri. (Pl. Bac. 516–17); Ego istuc furtum
scio quoi factum est. (Pl. Rud. 958); Non sat commode / divisa sunt temporibu’ tibi,
Dave, haec. (Ter. An. 475–6); MAIORUM OPTENUI LAUDEM UT SIBEI ME ESSE CREATUM
2
LAETENTUR. (CIL I .15.3–4 (Scip. Elog., Rome, c.135 BC)); Velut ex ea disputatione,
quae mihi nuper habita est in Tusculano . . . (Cic. Tusc. 2.2); Cui non sunt auditae
Demosthenis vigiliae? (Cic. Tusc. 4.44); Rufe mihi frustra ac nequiquam credite
amice . . . (Catul. 77.1); Prima dicte mihi, summa dicende Camena . . . (Hor. Ep.
1.1.1—NB: note the parallelism with the gerundive); Silva . . . Ciminia . . . nulli, ad
eam diem ne mercatorum quidem adita. (Liv. 9.36.1); . . . Aenaria a statione navium
Aeneae, Homero Inarime dicta . . . (Plin. Nat. 3.82); . . . item natam in Scadinavia
insula nec umquam visam in hoc orbe, multis tamen narratam achlin haud dissi-
milem illi . . . (Plin. Nat. 8.39);
Infectum forms: Arcus ubi aspicitur, mortalibus quae perhibetur / <Iris> (Enn. Ann.
409V=399S—NB: textually uncertain); Quare illa nobis alio tempore atque ad aliud
institutum . . . explicabuntur. (Cic. Inv. 1.86); Sic dissimillimis bestiolis communiter
cibus quaeritur. (Cic. N.D. 2.123); Sumatur enim nobis quidam praestans vir
optumis artibus . . . (Cic. Tusc. 5.68); Infert se saeptus nebula (mirabile dictu) / per
medios, miscetque viris neque cernitur ulli. (Verg. A. 1.439–40); . . . nulla placere diu
nec vivere carmina possunt, / quae scribuntur aquae potoribus. (Hor. Carm.
1.19.2–3); . . . aura petebatur medio mihi lenis in aestu . . . (Ov. Met. 7.811); Auctores
signa relinquendi et deserendi castra non uni aut alteri militi, sed universis exerci-
tibus palam in contione audiuntur. (Liv. 5.6.14); Nam plebeios nobiles iam eisdem
initiatos esse sacris et contemnere plebem, ex quo contemni patribus desierint, coepisse.
(Liv. 22.34.8); Aex . . . (quae ita Graecis appellatur) . . . (Plin. Nat. 4.51); Feliciorem ergo
tu Maecenatem putas, cui . . . somnus per symphoniarum cantum . . . quaeritur? (Sen.
Dial. 1.3.10); Nullas Germanorum populis urbes habitari satis notum est, ne pati
quidem inter se iunctas sedes. (Tac. Ger. 16.1); Ob haec et huiusmodi multa, quae
cernebantur paucis, omnibus timeri sunt coepta. (Amm. 28.1.24);
Just as with ab agents and ablative causes (see above) this use of the dative is extended
to non-passive clauses in post-Augustan poetry (and prose). Examples are (q) and (r).
A further development is the use with adjectives, as in (s)–(u). In (s), with a -bilis
adjective, the dative may be explained in a different way (see § 5.40).
(u) . . . Mamercus corpore toto / exsolvit poenas nulli non saucius hosti.
(‘Mamercus . . . suffered in every limb and was wounded by every foe.’ Sil. 5.333–4)
There is an ongoing discussion about how to explain this use of the dative. It is likely
that the interpretation of the dative constituent as an agent was initially triggered by
its context (see § 5.40). In (most of) the instances that we have there can be no doubt
that the authors intended the dative as an agent expression.42
Appendix: Augustan and post-Augustan poets also used the bare ablative as an
agent expression, perhaps for metrical reasons (initially). Two examples in passive
clauses are (v) and (w). Then also with the verb iaceo and the adjective saucius in (x)
and (y).43
42
For a different approach, see Serbat (1996a: 500–3). See also Luraghi (2010: 44–7), with references.
Brenous (1895: 169) and others (for example, Mariani 2002) ascribe the spread of the use of the dative to
Greek influence. For discussion, see Calboli (2009: 98–100).
43
Landgraf (1898/9: 15); Kenney (1958: 59) denies the existence of such instances in Ovid.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Cicero and Caesar are (d) and (e). In Late Latin authors the use of per is quite
common.44
(a) Quid ais? Vulgo accidebantur? Per quos et a quibus?
(‘What do you say? Men were killed as an ordinary occurrence? Through whom and
by whom?’ Cic. S. Rosc. 80)
(b) . . . per monitorem appellandi sunt . . .
(‘ . . . must they be greeted by your prompter?’ Cic. Mur. 77)
(c) Quia primas partis qui aget, is erit Phormio / parasitu’, per quem res geretur
maxume . . .
(‘Since the man who plays the leading part will be the trickster Phormio, through
whom most of the action will be carried out . . . ’ Ter. Ph. 27–9)
(d) Quid est quod confecto per te formidolosissimo bello coronam illam lau-
ream tibi tanto opere decerni volueris a senatu?
(‘What can have been your motive in bringing a formidable war to its conclusion,
and showing such eagerness to have that laurel wreath decreed to you by the senate?’
Cic. Pis. 58)
(e) Nam ad eas quae factae erant Arelate per Caesarem captivae Massiliensium
accesserant sex.
(‘For six captured Massilian ships had been added to those which had been con-
structed by Caesar at Arelate.’ Caes. Civ. 2.5.1)
Supplement:
. . . quae vos per Sullam gesta dicitis . . . (Cic. S. Rosc. 143); Ipse praesidiis dispositis
Lentulum in carcerem deducit. Idem fit ceteris per praetores. (Sal. Catil. 55.2–3);
Per te ego decipior, per te deprensus inermis. / Ecce, duas uno tempore turpis amo.
(Ov. Am. 2.10.3–4); Quamquam per praetorem prope debellatum erat . . . (Liv.
31.22.3); Denique in ore atque oculis Antonianae classis per M. Agrippam Leucas
expugnata . . . (Vell. 2.84.1); Traditus posthac Mithridates vectusque Romam per
Iunium Cilonem . . . (Tac. Ann. 12.21.1); Nunc vero de agno per quem redempti ac
vivificati sumus . . . (Cypr. Ad Fort. 3); Tricesimo octavo imperii anno per Anci filios
occisus est, regis eius cui ipse successerat. (Eutrop. 1.6.2); Caesar per Crassum de
rebellione dediticiarum gentium certior factus . . . (Oros. Hist. 6.8.9—NB: cf. Quibus
de rebus Caesar a Crasso certior factus . . . (Caes. Gal. 3.9.1));
44
For instances, see TLL s.v. per 1148,11ff. For the use of per with the passive in Orosius, see Svennung
(1922: 30–1); see Dognini (2007) for the history of the expression, Donati (2004) for theoretical discussion.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
the perspective of the first argument or from that of the second argument, respect-
ively. Certain factors play a role in this decision, but none of these leads to an
inevitable conclusion one way or the other.
There is one semantic factor, viz. animateness; if, in a state of affairs, one argument
is animate and the other inanimate, the animate one is more likely to be selected as
the subject of the clause. The same goes for human/non-human: A three-year-old is
struck by lightning rather than Lightning strikes a three-year-old. In a sample of
Cicero’s Rep. the combination animate patient/inanimate agent is ten times higher
in the passive than in the active.45
In addition to the semantic factor of animateness, two discourse factors also
influence the choice between active and passive. An entity that plays (or will play) a
central role in a paragraph (the so-called DISCOURSE TOPIC) is often the entity from
whose perspective the state of affairs is presented; hence it functions as the subject of
the sentences and clauses of that paragraph. This is illustrated by (a). In this
paragraph the Helvetians have decided to prepare for a journey out of their territory,
at the instigation of one of their leaders, Orgetorix, who has been properly introduced
at the beginning of the whole story. Then follows the text in (a). The subject/agent of
sentences (i) and (ii) are the Helvetians. They are also the agent of sentence (iii), but
not the subject. The switch to the passive indicates that the perspective from which
the narrative is being presented is now shifting from that of the Helvetians to that of
the central figure of the following paragraph, Orgetorix; accordingly, in sentence (iv)
he remains the subject (the anaphoric pronoun is), now in an active sentence. In this
passage, then, the discourse function of the passive is to effect a TOPIC SHIFT from the
Helvetians to Orgetorix. Ex. (b) shows such shifts on a larger scale.46 In (c), the
contrast between the two clauses is reinforced by the shift from active to passive. In
(d), the perspective switches from the Greeks to the Trojans, and then back to the
Greeks. The shift in (e) may be purely stylistic, in order to create variation. Note that
in (c) and (e) the passive clauses are agentful.
(a) (i) Ad eas res conficiendas biennium sibi satis esse duxerunt. (ii) In tertium
annum profectionem lege confirmant. (iii) Ad eas res conficiendas Orgetorix
deligitur. (iv) Is sibi legationem ad civitates suscepit. etc.
(‘For the accomplishment of these objects they considered that two years were
sufficient, and pledged themselves by an ordinance to take the field in the third
year. For the accomplishment of these objects Orgetorix was chosen, and he took
upon himself an embassage to the communities.’ Caes. Gal. 1.3.2)
(b) Itaque hesterno die L. Flaccum et C. Pomptinum praetores . . . ad me vocavi,
rem exposui, quid fieri placeret ostendi. Illi negotium susceperunt et . . . ad
pontem . . . pervenerunt. . . . ipsi . . . fortis viros eduxerant, et ego . . . delectos
45
See Pinkster (1985a: 115–16). For the choice between active and passive in Cicero’s poetic transla-
tions, see Cerra (2009: 67–70).
46
On topic shift, see especially Panhuis (1984), from whom the abbreviated ex. (b) is taken.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
However, the passive may also be chosen to create CONTINUITY OF PERSPECTIVE. This is
illustrated by (f). The states of affairs are presented from the perspective of the trees. In the
case of scindo this implies choosing the passive. This example also demonstrates that the
language user, here Virgil, decides how long he wants to keep the perspective constant.
Continuity is especially preferred in the case of coordination, as illustrated by (g).
(f) (Itur in antiquam silvam, stabula alta ferarum.) / Procumbunt piceae. Sonat
icta securibus ilex / fraxineaeque trabes cuneis et fissile robur / scinditur.
(Advolvont ingentis montibus ornos.)
(‘(They pass into the forest primeval, the deep lairs of the beasts.) Down drop the
pitchy pines, and the ilex rings to the stroke of the axe; ashen logs and splintering
oak are cleft with wedges. (From the mountains they roll down huge ash trees.)’
Verg. A. 6.179–82)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Three-place verbs with alternative case patterns for the second and third arguments
like dono ‘to present’ allow two forms of passivization, one with the gift as subject, the
other with the recipient as subject, as in (h) and (i), respectively. In (h), the state of
affairs is presented from the perspective of the patera, about which detailed informa-
tion is given in the relative clause. In (i), the state of affairs is presented from the
perspective of te: it is not unlikely that Amphitryon, the subject of dixeras, reported
the event as something that happened to him, instead of to the bowl (for more
examples, see § 4.54).47
(h) Post ob virtutem ero Amphitruoni patera donata aurea est, / qui Pterela
potitare rex est solitus.
(‘Afterwards my master Amphitruo was presented with a golden bowl for his valour,
the one from which King Pterela used to drink.’ Pl. Am. 260–1)
(i) (auream pateram) qua te illi donatum esse dixeras?
(‘ . . . the one you’d said you were presented with there?’ Pl. Am. 761)
47
See Bolkestein (1985).
48
For ‘deagentivization’ devices in Latin, see Tesařová-Nováková (1988: 150–2). For Cicero’s use of
explicit agents with the passive in his poetic translations, see Cerra (2009: 70–87).
49
For a collection of Early Latin instances of ‘accusative-passive periphrasis’, see Rosén (1981:
131–44). See also Joffre (1995: 204–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
For the use of the third person plural active for generic agents, see § 9.11.
50
For the relative frequency of the passive forms of coepi and desino, see TLL s.v. coepi 1424.66ff.,
desino 728.36ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(f ) Optume advenis, puere, cape chlamydem atque istic sta ilico, / ut, si haec
non sint vera, inceptum hoc itiner perficere exsequar.
(‘You’ve come at the right time, boy, take my cape and stand where you are,
so that I may continue to travel on this journey I’ve begun if this isn’t true.’
Pl. Mer. 912–13)
(g) . . . illa quae temptata iam et coepta sunt ab isto . . .
(‘ . . . as for the attempts at corruption that have already been set in motion
by that man . . . ’ Cic. Ver. 5.174)
51
For passive forms of possum, see TLL s.v. 136.77ff.; for queo and neqeuo, see Neue-W.: III. 626–7.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
52
For passive forms of debeo, see TLL s.v. 95.41ff., 98.8ff.; Löfstedt (1933: II.123–4).
53
For these examples, see Joffre (1995: 329–33, 2007a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
54
See Adams (1976: 65–6, 2013: 719–21); Kiss (2012: 107–8).
55
See Petersmann (1977: 164–7).
56
For more instances, see TLL s.v. facio 119.58ff. For discussion, see Michaelis (1998); Adams (2007:
466–70, 2013: 723).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Numquam . . . fuit nec fiet ill’ senex insanior / ex amore quam ille adulescens . . .
(‘That old man . . . has never been and never will be madder on account of his love
than that young man . . . ’ Pl. Mer. 446–7)
(b) . . . gallus allatus est, quem Trimalchio iussit ut aeno coctus fieret.
(‘ . . . the cock was brought in, which Trimalchio ordered to be cooked in a saucepan.’
Petr. 74.4)
(c) Supra hanc straturam pilae laterculis argilla subacta et capillo constructae
fiant distantes a se spatio pedis unius et semissis.
(‘Above the pillars, a foot and a half apart, columns are to be constructed of bricks
with kneaded clay and hair.’ Pall. 1.39.2—tr. Adams)
Supplement:
Interpositae orationes fiunt. (Pereg. 35.3); . . . ex quo et tempora cavata fiunt et oculi
depressi. (Mulom. Chir. 148); . . . multatus fit . . . (Consentius 4.17 N.); (sc. carnes
vaccinas) Vaporatas factas et in sodinga coctas utendum . . . (Anth. 3); Isiaca de pavo
primum locum habent ita, si fricta fiunt, ut callum vincant. (Apic. 2.2.6);
Appendix: It is sometimes argued that the verb venio ‘to come’ is used in poetry and
in substandard Late Latin with a passive perfect participle meaning ‘to become’. An
example cited in this context is (d). However, in all these instances venio can be
understood as a verb of moving, in this case with irritata as a secondary predicate.57
In Late Latin devenio is occasionally used more or less in the same sense as fio, and is
then also found with a perfect passive participle, as in (e); in such cases devenio
functions not as a substitute for the passive but rather as a copular verb. For the
‘copular’ use of these verbs, see § 4.97.
57
For discussion and references concerning venio, see Adams (2013: 721–3). For passive auxiliaries
other than sum in Late Latin, see also Michaelis (1998) and Cennamo (2005).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
verbs with which the passive forms are coordinated. The subjects are normally
inanimate or abstract.
(a) Gallia est omnis divisa in partes tres . . .
(‘Gaul as a whole is divided into three parts . . . ’ Caes. Gal. 1.1.1)
(b) Relinquebatur una per Sequanos via, qua Sequanis invitis propter angustias
ire non poterant.
(‘There remained one route through the borders of the Sequani, by which they could not
march on account of its narrowness, without the consent of the Sequani.’ Caes. Gal. 1.9.1)
(c) Iactor, [crucior], agitor, / stimulor, vorsor / in amoris rota, miser exanimor,
/ feror, differor, distrahor, diripior. / Ita nubilam mentem animi habeo.
(‘I’m being thrown around, tortured, tossed around, pierced, turned on the wheel of
love; poor me, I’m being destroyed, driven, driven apart, dragged apart, torn apart: so
clouded is my mind.’ Pl. Cist. 206–10)
(d) . . . mare . . . neque redundat umquam neque effunditur.
(‘ . . . the sea . . . never floods its bounds nor overflows.’ Cic. N.D. 2.116)
(e) Quo porro modo mundus moveri carens corpore aut quo modo semper se
movens esse quietus et beatus potest?
(‘Moreover, how is motion possible for a world that is incorporeal, and how can one
that is always moving itself be at peace and blissful?’ Cic. N.D. 1.33—NB: This
passage is much debated for its philosophical content.)
(f ) Nec simplici modo quatitur umquam, sed tremit vibratque.
(‘And sometimes the earth is not shaking in a simple manner but trembles and
vibrates.’ Plin. Nat. 2.194)
Supplement:
Simulatque ego in Siciliam veni, mutatus est. (Cic. Ver. 2.63); Ardor animi non semper
adest, isque cum consedit, omnis illa vis et quasi flamma oratoris exstinguitur. (Cic. Brut.
93); Relinquitur ergo ut omnia tria genera sint causarum. (Cic. Inv. 1.12); Eademque
ratione mare . . . conglobatur undique aequabiliter . . . (Cic. N.D. 2.116); . . . pares autem
vetere proverbio cum paribus facillime congregantur . . . (Cic. Sen. 7); Omnis Italia coacta
in angustias finditur in duo promunturia, Bruttium et Sallentinum. (Sal. Hist.
4.23); . . . dum amnes ulli rumpuntur fontibus . . . (Verg. G. 3.428); . . . desuntque dies
solido anno qui solstitiali circumagitur orbe . . . (Liv. 1.19.6); Cum enim linuntur,
recipientes umorem turgescunt, deinde siccescendo contrahuntur et ita extenuati dis-
rumpunt tectoriorum soliditatem. (Vitr. 2.8.20); . . . itemque arteriae, quas carotidas
vocant, sursum procedentes ultra aures feruntur. (Cels. 4.1.2); Ergo ubi aliquis ingredi
ac moveri coepit, rumpitur illa sutura atque intestina evolvuntur. (Cels. 7.4.3B); . . .
irascentium . . . micant oculi, . . . labra quatiuntur . . . (Sen. Dial. 3.1.4); . . . ita . . . rivus aut
flumen effunditur. (Sen. Nat. 3.15.5); Varie itaque quatitur . . . (Plin Nat. 2.193); Ctesias
tradit Silan vocari stagnum in Indis, in quo nihil innatet, omnia mergantur. (Plin. Nat.
31.21); Livet Charinus, rumpitur, furit, plorat / et quaerit altos unde pendeat ramos.
(Mart. 8.61.1–2); . . . tempora eius colligantur . . . (Mulom. Chir. 270);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
58
Flobert (1975: 410–19) uses the term ‘néodéponent’. K.-St.: I.105 uses the term ‘deponentiales
Passiv’. For a list of active participles see Neue-W.: III.11–15 and Flobert (1975: 413–17).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(g) . . . hic tibi certa domus, certi, ne absiste, penates, / neu belli terrere minis.
(‘ . . . Here, thy home is sure—draw not back—and sure are thy gods! Nor be
scared by threats of war.’ Verg. A. 8.39–40)
(h) Invidia sacrae, Caesar, ne tangere famae.
(‘Be not jealous, Caesar, of those whom fame has consecrated.’ Luc. 9.982)
The number of verbs that are used in the autocausative passive has been estimated at
700 over the whole period of Latin. These verbs are also used in the true passive, some
also in the decausative passive. About half of these verbs are also used with reflexive
pronouns and about one fifth also as intransitives.61 This is illustrated below for the
verbs lavo ‘to bathe’ and moveo ‘to move’ (for moveo see also § 4.11). Varro was well
aware of the problem, as (i) shows. Propera in (j) is a proof of the active involvement
of the subject. Note that in (k) Pliny could not use lavantem alone without creating
ambiguity. Ex. (l) has autocausative moveri, (m) decausative moventur, (n) has both
59
A typological monograph on the subject is Kemmer (1993).
60
Consentius V.366.1ff. K; Priscianus II.378.10ff. K.
61
See Flobert (1975: 383 and 386). For lavo, see Traina (1999: 41–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
reflexive move te and intransitive moveo. In (o), the subject of intransitive movisse is
inanimate.
(i) Omnino et lava[n]t et lavatur dicitur separatim recte in rebus certis, quod
puerum nutrix lava<t>, puer a nutrice lavatur, nos in balneis et lavamus et
lavamur.
(‘At any rate, lavat and lavatur are used with a difference of meaning in certain
matters, because a nurse lavat “bathes” a child, the child lavatur “is bathed” by the
nurse, and in the bathing establishments we both lavamus “bathe” and lavamur “are
bathed”. ’ Var. L. 9.106)
(j) Proin tu lavare propera. # Lautus sum.
(‘So be quick to wash. # I am washed.’ Pl. St. 667)
(k) Iidem Polycles et Dionysius . . . Iovem qui est in proxima aede fecerunt, . . .
Venerem lavantem sese Daedalsas, stantem Polycharmus.
(‘Polycles and Dionysius made also the Jupiter which is in the adjacent temple, . . .
Daedalsas the Venus Bathing, and Polycharmus the Venus Standing.’ Plin. Nat. 36.35)
(l) Verum in cubiculo / . . . illam moveri gestio.
(‘But in the bedroom . . . I’m really keen on her moving.’ Pl. As. 787–8)
(m) Si dolor in gingivis est, moventurque aliqui dentes, refigi eos oportet.
(‘If it is the gums which are involved, and some teeth are loose, they should be
extracted.’ Cels. 6.15.3)
(n) Move te oro ocius, / mea nutrix. # Moveo. # Video sed nil promoves.
(‘Move faster, nurse. # I am moving. # I can see that. But you’re not making any
progress.’ Ter. Eu. 912–13)
(o) Quod C. Iulius, L. filius, pontifex, nuntiavit in sacrario <in> regia hastas
Martias movisse, de ea re ita censuerunt, uti . . .
(‘Whereas Gaius Julius, son of Lucius, the pontifex, has reported that the spears of
Mars have moved in the sanctuary in the Regia, the senate has therefore decreed with
reference to that matter, that . . . ’ S.C. apud Gell. 4.6.2 (99 BC))
The fact that the same autocausative action can be expressed in three different ways may
at first sight seem alarming. On closer inspection, however, the phenomenon is limited to
those states of affairs in which the same person (or animate being) can be easily involved
as an agent and as a patient. Verbs indicating physical treatment are a good example.
With a verb like occido ‘to kill’, the reflexive expression is the only one that is available to
mark the action as autocausative, as in (p). For passive forms to be understood as
autocausative, the identity of the agent, if unexpressed, and his active involvement
must be inferrable from the context. Similarly, for active forms to be understood as
autocausative the identity of the patient must be inferrable from the context.
Verbs that are used in the autocausative passive belong to various semantic classes.
Following Flobert (1975: 382–6), three major classes with several subclasses are
distinguished (the asterisk indicates that there is also a reflexive counterpart).
Examples are given in the Supplements. Some verbs can be used in the autocausative
passive with more than one meaning.
(i) Verbs denoting physical treatment: (a) lavor ‘to wash oneself ’*, ornor ‘to get
oneself adorned’*, polior ‘to polish oneself ’, tondeor ‘to shave’, ung(u)or ‘to
anoint oneself ’; (b) amicior ‘to dress’*, cingor ‘to gird up one’s dress’, induor
‘to put on oneself ’*, vestior ‘to dress oneself ’; (c) alor ‘to feed oneself ’*,
ebrior ‘to get drunk’, nutrior ‘to grow up’.
Supplement:
(a) Uxor eius dixit se in balneis virilibus lavari velle. (Gracch. Orat. 45); Tum
ornatam ita uti quae ornantur sibi . . . (Ter. Hau. 288); At lavamur et tonde-
mur et convivimus ex consuetudine. (Quint. Inst. 1.6.44); . . . unguor, / ut illi
placeam . . . (Pl. Cas. 226–7);
(b) Attat, cesso magnufice / patricieque amicirier . . . (Pl. Cas. 723–3a); Cingitur.
Certe expedit se. (Pl. Am. 308); Quid erat induta? (Pl. Epid. 223); Interiores
plerique frumenta non serunt, sed lacte et carne vivunt pellibusque sunt
vestiti. (Caes. Gal. 5.14.2);
(c) . . . (sc. res) quibus alimur et vivimus. (Cic. Agr. 2.95); Quibus dedit ebriati
sunt et lapidibus eum occiderunt. (Amp. 2.6); . . . <quae> sub septentrionibus
nutriuntur gentes, inmanibus corporibus . . . sunt conformati. (Vitr. 6.1.3);
(ii) Verbs denoting physical and mental changes: (a) dissolvor ‘to disintegrate’*,
exedor ‘to be consumed’, mutor ‘to suffer a change’*; (b) abdor ‘to conceal
oneself ’*, abscondor ‘to hide’*, probor ‘to pass for enlistment’*, vocor ‘to call
oneself ’*; (c) crassor ‘to thicken’, frangor ‘to abate’*, inveteror ‘to become old’,
minuor ‘to abate’*, purgor ‘to become unclouded (of the sky)’*, scindor ‘to
fork’*, stipor ‘to crowd’. (Some of the examples seem to be decausative.).
Supplement:
(a) Dissolvuntur enim tum demum membra fluuntque. (Lucr. 4.919); Exspectando
exedor miser atque exenteror, / quo modo mi Epidici blanda dicta evenant. (Pl.
Epid. 320–1); Cum vehementius in movendo ut ab se abeat foras fertur, formido.
(Var. L. 6.48); Simulatque ego in Siciliam veni, mutatus est. (Cic. Ver. 2.63); Ex
feminis mutari in mares non est fabulosum. (Plin. Nat. 7.36);
(b) (sc. delphini) Abduntur tricenis diebus circa canis ortum occultanturque
incognito modo . . . (Plin. Nat. 9.22); Ille enim dies quo primum probati
sunt veritatem ab iis originis suae exegit. (Tra. Plin. Ep. 10.30(39).2); Tale
quiddam sapiens facit. In se reconditur, secum est. (Sen. Ep. 9.17—NB:
combined with in se); Itaque nunc P. Cornelius vocatur. (Cic. Fam. 13.36.1);
(c) . . . pili mei crassantur in setas . . . (Apul. Met. 3.24.4); Quod item fit in altis /
fluminibus magnoque mari, cum frangitur aestus. (Lucr. 6.143–4); Uvae
paulum ante maturitatem decerptae siccantur acri sole, ter die versatae per
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(iii) Verbs denoting changes in physical or social position: (a) alienor ‘to become
averse to’*, appellor ‘to put in’*, feror ‘to proceed’*, vertor ‘to turn’*, volvor
‘to roll’*; (b) curvor ‘to curve’*, declinor ‘to change direction’*, efferor ‘to
rise’*, iungor ‘to join forces’*, surgor ‘to rise’*; (c) colligor ‘to gather’*, dedor
‘to surrender’*, extricor ‘to set oneself free’, implicor ‘to embrace’*, misceor
‘to share or take part in’*, socior ‘to associate oneself ’*.
Supplement:
(a) Summum erat periculum ne, si nihil impetrassent, plane alienarentur a senatu.
(Cic. Att. 1.17.9); . . . ut eos qui essent adpulsi navigiis interficere possent . . .
(Cic. Ver. 5.145); . . . iam ista serpens, quae tum hic delitiscit, tum se emergit et
fertur illuc, compressa atque inlisa morietur. (Cic. Har. 55); Idque sua causa
consuerunt scorta moveri, ne complerentur crebro gravidaeque iacerent . . .
(Lucr. 4.1274–5); Ad sepulturam inde suorum nequaquam paribus animis
vertuntur . . . (Liv. 1.25.13); Id autem cum accidit, suapte natura, quod superest,
et cylindrum volvi et versari turbinem putat. (Cic. Fat. 42);
(b) In hoc genus plerique cum declinantur et ab eo quo profecti sunt aberrarunt
. . . (Rhet. Her. 4.15); . . . turpitudo coniungendi cum tyranno. (Cic. Att.
7.20.2); Bracchia coeperunt nigris horrescere villis / curvarique manus et
aduncos crescere in ungues . . . (Ov. Met. 2.478–9); Corvus ex conspectu elatus
orientem petit. (Liv. 7.26.5); Plerique culpabant Cerialem passum iungi quos
discretos intercipere licuisset. (Tac. Hist. 4.75.2);
(c) . . . (sc. venti) quasi collecti redeunt ceduntque repulsi . . . (Lucr. 6.571); Tum
abici passim arma ac dedi hostes coepti . . . (Liv. 4.29.4); . . . aliqua ope exsol-
vam, extricabor aliqua. (Pl. Epid. 152); Ecquando, mulier, seni tuo blandius
inplicita iacuisti? ([Quint.] Decl. 2.18); . . . dum nuda palaestra / ludis et es
nudis femina mixta viris. (Ov. Ep. 16.151–2); . . . Richomeres Profuturo socia-
tur et Traiano tendentibus prope oppidum Salices. (Amm. 31.7.5);
With a verb of dressing like induo the expected three-place expression is ‘to dress
somebody (accusative) in some sort of garment (ablative)’, as in (a) (a passive clause).
Alternatively, the person may be expressed in the dative and the garment in the
accusative, as in the active clause (b) (see § 4.54 for other verbs that have these two
frames). The verb is also used in a two-place frame with the object in the accusative,
as in the active clause (c). However, from Plautus onwards the accusative for
the garment is also regularly found with the passive participle indutus, as in (d).
The passive is autocausative. Infectum passive forms of induo with the garment in the
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
accusative are found from the Augustan poets onwards, as in (e). Related verbs like
cingor ‘to gird up one’s dress’ + compounds are also found from that period onwards.
Exuor ‘to strip’ is used by Statius. Livy introduces the expression in prose.62 The
whole phenomenon is a literary extension of an originally Latin idiom on the analogy
of Greek expressions.63
(a) . . . induta fuit saeva stola . . .
(‘ . . . clothed she was in a huge gown . . . ’ Enn. scen. 410V = 396J)
(b) Cui (sc. Herculi) cum Deianira sanguine Centauri tinctam tunicam
induisset . . . ait ille: . . .
(‘When Deianira had got the shirt, steeped in the Centaur’s blood, put upon him . . . he
says: . . . ’ Cic. Tusc. 2.20)
(c) Loricam induam mi optumum esse opinor.
(‘I think it’s best for me to put on a breastplate.’ Pl. Cas. 695)
(d) Non ego te indutum foras / exeire vidi pallam?
(‘Didn’t I see you come out wearing a mantle?’ Pl. Men. 511–12)
(e) . . . ille frementis / ad iuga cogit equos clipeumque auroque trilicem / loricam
induitur fidoque accingitur ense.
(‘ . . . another couples his snorting steeds to the yoke, dons his shield and coat of mail,
triple-linked with gold, and girds on his trusty sword.’ Verg. A. 7.638–40—NB: Note
also autocausative accingitur, with an ablative.)
Supplement:
. . . inutile ferrum / cingitur ac densos fertur moriturus in hostis. (Verg. A. 2.511–12);
Tandem progreditur magna stipante caterva, / Sidoniam picto chlamydem circum-
data limbo. (Verg. A. 4.136–7); Sic fatus deinde comantem / Androgei galeam
clipeique insigne decorum / induitur . . . (Verg. A. 2.391–3); . . . terrificos umeris
Aetolus amictus / exuitur patriumque suem. (Stat. Theb. 6.835–6); Inter haec tres
pueri candidas succincti tunicas intraverunt . . . (Petr. 60.8);64
Examples like (f)–(h) are only attested once in Early Latin and are in all likelihood
literary extensions: in this type of expression the body or body part functions as the
object of an autocausative passive form (usually a perfect participle) of a verb of
dressing, covering, or adorning; the garment or the ornament is typically expressed in
the ablative.
(f ) . . . succincti corda machaeris . . .
(‘ . . . girt round their hearts with broadswords . . . ’ Enn. Ann. 400V = 519S)
62
For details see TLL s.v. induo 1266.77–1269.38; Händel and Szantyr (1954); Courtney (2004).
63
For the influence of Greek on the various types of accusative objects in this section, see Coleman
(1999: 79–81) with references to Coleman (1975).
64
Discussion in Petersmann (1977: 67–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Furthermore, accusative objects are found with all sorts of verbs that refer to actions
performed by the agent to change his/her mental or physical condition. Examples of
changes in the physical condition are (i)–(k). Exx. (l) and (m) illustrate changes in the
mental condition. There are no antecedents attested of this usage in Early Latin and
the phenomenon is entirely inspired by Greek expressions.
(i) Ille latus niveum molli fultus hyacintho / ilice sub nigra pallentis ruminat
herbas . . .
(‘He, pillowing his snowy side on soft hyacinths, under a dark ilex chews the pale
grass . . . ’ Verg. Ecl. 6.53–4)
(j) . . . ut faciem mutatus et ora Cupido / pro dulci Ascanio veniat . . .
(‘How Cupid, changed in face and form, may come in the stead of sweet Ascanius . . . ’
Verg. A. 1.658–9)
(k) . . . magni / quo pueri magnis e centurionibus orti, / laevo suspensi loculos
tabulamque lacerto, / ibant . . .
(‘ . . . to which grand boys used to go, sons of grand centurions, with slate and satchel
slung over the left arm . . . ’ Hor. S. 1.6.72–5)
(l) . . . expleri mentem nequit ardescitque tuendo / Phoenissa . . .
(‘ . . . the . . . Phoenician . . . cannot satiate her soul, but takes fire as she gazes . . . ’ Verg.
A. 1.713–14)
(m) (sc. Iuno) . . . multa movens necdum anticum saturata dolorem.
(‘ . . . pondering many a thought and with her ancient grudge still unsated . . . ’ Verg.
A. 5.608)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
. . . immani corpore pistrix / delphinum caudas utero commissa luporum. (Verg.
A. 3.427–8); Aeneas maesto defixus lumina voltu / ingreditur . . . (Verg. A. 6.156–7);
. . . nec . . . incepto voltum sermone movetur . . . (Verg. A. 6.470—NB: infectum
form); . . . non illa colo calathisve Minervae / femineas adsueta manus . . . (Verg. A.
7.805–6); . . . iuxtaque comes Lavinia virgo, / causa mali tanti, oculos deiecta decoros.
(Verg. A. 11.479–80); . . . caeruleosque implexae crinibus angues / Eumenides . . .
(Verg. G. 4.482–3); . . . nasum nidore supinor . . . (Hor. S. 2.7.38—NB: infectum
form); . . . inductaque cornibus aurum / victima vota facit. (Ov. Met. 7.161–2—NB:
the passive is sometimes regarded as a true passive, but see Bömer ad loc.); . . .
cunctisque prior Cadmeius heros / adcurrit vultum deiectus . . . (Stat. Theb. 3.366–7);
. . . adlevatur animum et quae repentina mutatio exquirit. (Tac. Ann. 6.43.2—NB:
infectum form);
For the terminology used to describe these accusative objects (accusativus Graecus,
‘retained accusative’, etc.), see § 5.7.
Two types of impersonal passive clauses can be distinguished. Examples of the first
type were provided in § 5.2, and are repeated below as (a) and (c). The verbs involved
are mainly one-place verbs that almost without exception (see below) belong to a
controlled state of affairs, as in (a) and (b). Rarely, the verbs may also be two-place
(and three-place) used absolutely, as in (c) and (d). The most common class of one-
place verbs are motion verbs. In Caesar and in the historians, the verb pugno ‘to fight’
is relatively common.
(a) Eatur. # Sequere hac.
(‘Let’s go, then. # Follow me this way.’ Ter. Hau. 743)
(b) Itur ad te, Pseudole. / Orationem tibi para advorsum senem.
(‘You’re being approached, Pseudolus. Prepare your speech against the old man.’ Pl.
Ps. 453–4)
(c) Quid agitur, Calidore? # Amatur atque egetur acriter.
(‘What’s going on, Calidorus? # Love and a dire financial situation are what’s going
on.’ Pl. Ps. 273)
(d) Quoiuis modi hic cum [mala] fama facile nubitur.
(‘It’s easy for any girl to get married here, no matter what her reputation is.’
Pl. Per. 386)
Difficult to understand are the Early Latin passive forms caletur ‘it is hot’ in Plautus:
Quasi, quom caletur, cocleae in occulto latent . . . (Pl. Capt. 80; Truc. 65) and nubi-
labitur in Cato (Agr. 88.2), imitated—as he likes to do—by Apuleius (Met. 4.1; Fl. 2),
who has also pluitur et ningitur (Fl. 2).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The second type consists of two-place (and three-place) verbs governing a non-
accusative object with that object expressed, as in (e)–(g). Similar to these verbs are
verbs with a prepositional object expressed. An example is (h), and (i) may also
illustrate this phenomenon—unless ad eum is taken as a direction adjunct. Note the
ab phrases in (h) and (i). The verbs, as those of the first type, belong almost without
exception to controlled states of affairs.
(e) Invidetur enim commodis hominum ipsorum, studiis autem eorum ceteris
commodandi favetur.
(‘For men’s private gains breed jealousy, while their zeal for others’ service is
applauded.’ Cic. de Orat. 2.207)
(f) Arma habemus non adversus eam aetatem, cui etiam captis urbibus
parcitur . . .
(‘We bear no weapons against those tender years which find mercy even in the
storming of a city . . . ’ Liv. 5.27.7)
(g) Non ideo divitiae tibi nocent, si propter divitias tibi nocetur.
(‘Riches themselves do not harm you, just because it is on account of riches that you
suffer harm.’ Sen. Ep. 87.30)
(h) . . . cum ad me aditum esset ab iis qui dicerent a se intolerabilia tributa exigi . . .
(‘ . . . when I had been approached by persons who claimed that they were being
subjected to taxation on an intolerable scale.’ Cic. Fam. 3.7.3)
(i) A qua muliere cum erat ad eum ventum et in aurem eius insusurratum, alias
revocabat eos . . . alias . . .
(‘Whenever he was approached by that woman and she whispered in his ear,
sometimes he called back those parties . . . at other times . . . ’ Cic. Ver. 1.120)
This construction of an impersonal passive with an object does not exist for verbs
governing an accusative object. A few remarkable expressions such as (j)–(m) are
explained in that way. Of these, (j) was already obscure by Gellius’ time. If construed
with vivitur, vitam would be an extraordinary instance of the cognate object (on
which see § 4.10).65 Ex. (k) is based on the authority of Servius, but may well be a
direction adjunct.66 Triclinia in (l) is in all likelihood intended as a nominative, the
subject of the very remarkable passive form faciatur.67 Ex. (m) may belong to § 5.20
(accusative patients with passive and related forms). There are a few parallel
instances with a gerundive (see § 5.39). The four cases are quite diverse and cannot
serve as proof of the existence in Latin of a rule that allowed accusative objects with
impersonal passives.68
65
See Calboli (1962a: 42–4, 1990: 118); Adams (2013: 240–2).
66
Ex. (b) is explicitly mentioned by Servius ad Verg. A. 2.10 and 11.230. Detailed discussion in Calboli
(1962a: 7–14).
67
See Calboli (1962a: 14–16).
68
For (disputed) Early Latin instances, see Bennett (1914: II.261). Also Wistrand (1941: 101–4).
For the putative existence of the construction in the Italic dialects, see Le Bourdelles (1966).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(j) Incerte errat animus, praeter propter vitam (vita cj. Salmasius) vivitur.
(‘Aimless we drift, we live but more or less.’ Enn. scen. 241V = 202J (ap. Gell.
19.10.12–14, because of praeter propter))
(k) Ita castra sine volnere introitum.
(‘Thus they entered into the camps without a wound.’ Sal. Hist. 4.10)
(l) Faciatur (faciantur cj. Goesius), si tibi videtur, et triclinia.
(‘I should also like you to make a representation of a dining-room set of
couches, if you can arrange it.’ Petr. 71.10 (Trimalchio speaking))
(m) Graeci iubent, olivam cum plantatur et legitur, . . .
(‘The Greeks order, when they plant and select the olive . . . ’ Pall. 1.6.14)69
The agent of the state of affairs in impersonal passive clauses is almost always
human71 and is very rarely expressed. There are a few instances attested in Early
Latin, like (n), which is statistically not surprising. Other examples are given above in
(h) and (i) (both ab phrases). The most common form for the agent is a prepositional
phrase with ab. The only(?) attested instance with a dative agent (dativus auctoris) is
(o), but it is usually emended. As for the presence of agents with the impersonal
passive, the reason for this formation may be that, in this way, while giving promin-
ence to the action rather than to the agent, uncertainty as to the identity of the agent
can be avoided.72 In (p), ab omnibus is contrastive focus in first position.73
(n) Ita res est, fateor. Peccatum a me maxume’st.
(‘That is so, I own it. I have been grievously in fault.’ Ter. Hau. 158—NB: cf. active
peccavi in Hau. 632 and 644)
69 70
See Svennung (1935: 473–7). See Pinkster (1992a: 164, n. 13).
71
See Pieroni (2000) and Napoli (2009) for the various types of agents with the impersonal passive and
factors determining them. See also Baños (2014a: 780–3).
72
See for this entire section Pinkster (1992a) and Napoli (2009).
73
See Napoli (2009: 174).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(o) Saepe antea <a> paucis strenuis advorsum multitudinem bene pugnatum.
(‘Often before battles had been successfully fought by a few valiant soldiers against a
multitude.’ Sal. Jug. 107.1)
(p) Cum alicuius in senatu sententiam sequimur, non potest dici: ille magis
adsentitur quam ille. Ab omnibus in eandem sententiam itur.
(‘When at a meeting of the senate we vote in favour of someone’s motion, it cannot
be said, “A. is more in accord with the motion than B.” All alike vote for the same
motion.’ Sen. Ep. 66.41)
Supplement:
Nunc tu, Cleostrata, / ne a me memores malitiose de hac re factum aut suspices, / tibi
permitto: tute sorti. (Pl. Cas. 393–5); . . . locum istum totum, ut a doctissimis Grae-
ciae quaesitum et disputatum est, explicabo . . . (Cic. Leg. 3.13); Disputatur in consilio
a Petreio atque Afranio et tempus profectionis quaeritur. (Caes. Civ. 1.67.1);
Sed nostri milites dato signo cum . . . animum advertissent non concurri a
Pompeianis . . . (Caes. Civ. 3.93.1); . . . pugnatumque ab hostibus ita acriter est ut a
viris fortibus in extrema spe salutis . . . pugnari debuit . . . (Caes. Gal. 2.33.4); Pugna-
tum est ab utrisque acriter. (Caes. Gal. 4.26.1); . . . etsi, quantum ratione provideri
poterat, ab nostris occurrebatur ut . . . (Caes. Gal. 7.16.3); Per idem tempus advorsum
Gallos ab ducibus nostris Q. Caepione et Cn. Manlio male pugnatum. (Sal. Jug.
114.1); <namque ea materies,> . . . reicitque eius (sc. ignis) vim nec patitur ab eo sibi
cito noceri . . . (Vitr. 2.9.14);
Alternative agent-like expressions: Trepidatum itaque non in agris magis quam in
urbe est. (Liv. 4.21.8); Sic cotidie utrimque eminus fundis, sagittis reliquisque telis
pugnabatur. (Caes. Civ. 1.26.1); Et Hernicis Latinisque milites imperati; utrimque
enixe oboeditum dictatori est. (Liv. 4.26.12); Neque licentia gladiorum in mediocres
saevitum . . . (Vell. 2.22.1); Effectum enim est per inimicos ut omnes eius libri
comburerentur. (Sen. Con. 10. pr. 5); Cantatur ac saltatur per omnis gentes aliquo
modo. (Quint. Inst. 2.17.10);
When the agent is not expressed, it can almost always be inferred from the context, as
in (q)–(s). In (q), the reason for not indicating the agent of itur is the same as for the
personal passive putatur ratio: it is not relevant to the story. In (r), the agent after
repeatedly refusing agrees reluctantly (see dabitur in (i) in § 5.5), so his identity is
clear from the context. In (s), the identity of the agent is not very clearly defined, but
the people coming out have been referred to in line 804 as illaec remeligines ‘those
dawdlers’. Ex. (t) is a rare instance of a clause in which a non-human agent must be
supplied from the context.74 In (u), the impersonal passive is continued by a third
person plural expression (see also below).
(q) Itur. Putatur ratio cum argentario.
(‘One goes and the accounts are reckoned up with the banker.’ Pl. Aul. 527)
74
See Napoli (2009: 167).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Exx. (x) and (y) show variation between impersonal passive and active expressions. In
(x), the agent of intenderant is the same as in the preceding context. Sallust then shifts
from using the active to the passive, although the agent remains the same. This may be
explained as a stylistic preference for variation and/or as a signal that the episode is
coming to its end. In (y), the agent of consurgitur is not the legati who have just spoken.
Continuation with a third person plural is impossible, since it was not the speakers but a
different group of people who rose (they are not further identified: the participants of the
council). After this break, it is possible to continue with the third person plural. This is
an example of the use of the passive to create topic shift (see § 5.10).
(x) Deinde paucis diebus quo ire intenderant perventum est.
(‘A few days later they came where they had intended to go.’ Sal. Jug. 107.7)
(y) (After a heated discussion between the legati in front of their council.)
Consurgitur ex consilio. Comprehendunt utrumque et orant ne . . .
(‘The council rose, and they laid hold on the two generals and prayed them not to . . . ’
Caes. Gal. 5.31.1)
Supplement:
Valuistin’ usque? # Sustentatum est sedulo. (Pl. St. 467; cf. 586); Dic hoc negoti quo
modo actum est. # Ibitur. (Pl. Trin. 578); Sed hoc agamus quoia huc ventum est
gratia. (Pl. Truc. 9); Quid istuc? Alienu’n es, amabo, mi Strabax, / qui non extemplo
<intro> ieris? # Anne oportuit? / # Ita te quidem, qui es familiaris. # Ibitur, / ne me
morari censeas. (Pl. Truc. 664–8); Sane nollem huc exitum (Ter. Ad. 776); Quid
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
agitur? # Statur. # Video. (Ter. Eu. 271); Persuasit homini; factum’st; ventum’st;
vincimur; duxit. (Ter. Ph. 135–6); Curritur ad praetorium, quo istum ex illo prae-
claro convivio reduxerant paulo ante mulieres cum cantu atque symphonia. (Cic.
Ver. 5.92); . . . nos ‘convivia’, quod tum maxime simul vivitur. (Cic. Fam. 9.24.3);
Sensit in se iri Brutus. (Liv. 2.6.8); Deinde cum aeris unda nitentes, cum perventum
ad montes, ab eorum offensa et procellis propter plenitatem et gravitatem liques-
cendo disparguntur (sc. nubes) et ita diffunditur in terras. (Vitr. 8.2.2); . . . si aequa-
liter aqua canalis summa labra tanget, scietur esse libratum. (Vitr. 8.5.2); Ideoque
aptissime Alexandriam ex Italia itur. (Cels. 3.22.8); Cum ad summum perveneris,
paria sunt, non est incremento locus, statur. (Sen. Ep. 79.8); CUM AD VADIMONIUM
VENTUM ESSET . . . (TPN 22.2.1–2 (Murecine, AD 49));
5.22 Reflexivity
The most normal use of a reflexive pronoun is to refer to the same entity to which
the subject of its clause refers, as in (a). In this example the reflexive pronoun
functions in the same way as a noun phrase referring to an entity other than the
subject of the clause, as is clear from the parallel clause with animum tuum. This
use will be called TRUE REFLEXIVE (or in other descriptions ‘semantic reflexive’).75
However, from Early Latin onwards reflexive expressions are also used in a way
that is more or less equivalent to the AUTOCAUSATIVE use of the passive (see § 5.19).
An example is (b). Ex. (c) resembles the DECAUSATIVE use of the passive since no
special agent or cause is implied in the process of scindent se. In addition to these
there are still other uses, and there are developments over time. As in the case of
the passive, the assignment of the attested instances to the various types is not
always easy.
(a) Cum igitur ‘nosce te’ dicit, hoc dicit: ‘nosce animum tuum’.
(‘So when he says “Know thyself”, he means this, “Know thy soul”. ’ Cic. Tusc. 1.52)
(b) Nam qui liberos [esse] ilico se arbitrantur, / ex conspectu eri si sui se
abdiderunt, / . . .
(‘Yes, those who immediately consider themselves free if they’ve removed themselves
from their master’s eyes . . . ’ Pl. Ps. 1105–6)
(c) . . . neque herba nascetur, neque lutamenta scindent se.
(‘ . . . weeds will not grow, nor the plaster crack.’ Cato Agr. 128)
75
On the various uses of reflexives see Geniušienė (1987). A good survey can be found in Wehr (1995:
ch. 3), on whose work the terminology used in this chapter is based. See also Flobert (1975: 382–98), Joffre
(1997), Stempel (2002), and Lühr (2011). For Priscian’s treatment of the reflexives, see Wright (2002:
79–84).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Reflexivity
This use of the reflexive is not different from the use of any other noun phrase, as was
already illustrated by (a) in the preceding section. Someone may be actively involving
either someone else or him- or herself in an action or position. Examples are (a)–(c). Note
in (a) the coordination with another noun phrase and in (b) with a personal pronoun. This
use of the reflexive pronoun is part of its normal ‘direct’ use, which is discussed in § 11.126.
(a) Mulier quae se suamque aetatem spernit, speculo ei usus est.
(‘A woman who is dissatisfied with herself and her age needs a mirror.’ Pl. Mos. 250)
(b) Res vero publica . . . ab istius inflammato atque indomito furore iam tum se
meque repetebat.
(‘As for the republic, . . . it was even then demanding me and itself back from his fiery
and irrepressible fury.’ Cic. Dom. 141)
(c) Nemo autem sibi bene facit, non magis quam sibi favet, quam suarum
partium est.
(‘But no one does good to himself any more than he befriends himself, or belongs to
his own party.’ Sen. Ben. 5.10.3)
The active and responsible involvement of the agent in a state of affairs may be made
explicit by means of the identifying pronoun ipse ‘self ’, as in (d) and (e), or by the
suffix ‐met, as in (f) and (g). These elements may be added either to the agent or to a
constituent with a different function in the clause.
(d) Qui ipsus sibi satis placet, nec probus est nec frugi bonae. / Qui ipsus se
contemnit, in eo est indoles industriae.
(‘The man who is content with himself is neither decent nor well conducted. The man
who despises himself has the possibility for industry within him.’ Pl. Trin. 321–2)
(e) . . . praeter Atheniensem Themistoclem, qui se ipsum morte multarit?
(‘ . . . save only Themistocles the Athenian, who with his own hand wrought his
death?’ Cic. Scaur. 4)
(f) Quis te verberavit? # Egomet memet, qui nunc sum domi.
(‘Who hit you? # I hit myself, the I that is at home now.’ Pl. Am. 607)
(g) Sed memet moror, quom hoc ago setius.
(‘But I’m delaying myself because I’m doing this so slowly.’ Pl. Cist. 692)
less behaves as a unit, often also with a meaning that is (slightly) different from the
meaning of the verb itself. The number of verbs involved is estimated at over 500 for
the whole period of Latin76 and many of these are also used in the autocausative
passive (see the verbs with an asterisk in § 5.19). Examples are (a)–(c). Note in (c) the
parallelism of exsurgere and erigere se.
(a) . . . ex conspectu eri si sui se abdiderunt (sc. servi) . . .
(‘ . . . if they’ve removed themselves from their master’s eyes . . . ’ Pl. Ps. 1106)
(b) Quin ego hinc me amolior?
(‘Why don’t I remove myself from here?’ Pl. Mer. 384)
(c) Haec quae procul erant a conspectu imperii non solum adflixerunt, sed
etiam, ne quando recreata exsurgere atque erigere se possent, funditus, ut
dixi, sustulerunt.
(‘These cities were far out of sight of our dominion, yet our ancestors not only
overthrew them but, to prevent their recovery and rising again with renewed strength,
as I said, they utterly destroyed them.’ Cic. Agr. 2.87)
The degree to which individual authors use the autocausative reflexive or the auto-
causative passive varies. In didactic texts, the reflexive is more frequent. Livy and
Tacitus prefer the autocausative passive expression. Poets use the alternatives to suit
their metrical needs. Diachronically speaking, in some cases the reflexive supplants
the passive, in others the converse occurs. In general, the use of the reflexive increases
in correlation with the decrease of the passive morphology of the infectum stem.
Supplement (in alphabetical order by verb):
Unde agis te . . . ? (Pl. Per. 482); Haud auspicato huc me appuli. (Ter. An. 807); Nunc
ego me illac per posticum ad congerrones conferam. (Pl. Mos. 931); . . . dummodo
cuius exemplo se consolatur, eius exitum expectet . . . (Cic. Prov. 16); . . . curvare se ad
terram maluerunt et pecudum modo humi repere. (Lact. Ir. 20.10); . . . ego [non]
declinavi paullulum me extra viam. (Pl. Aul. 711); . . . sese dedere sine fraude con-
stituunt. (Caes. Civ. 2.22.1); Hic (sc. prester) ubi se in terras demisit dissolvitque, /
turbinis immanem vim provomit atque procellae. (Lucr. 6.446–7); Expulit hanc
sanguis seque eiaculatus in altum / emicat et longe terebrata prosilit aura. (Ov.
Met. 6.259–60); (sc. Ciris) Quae simul ut sese cano de gurgite velox / cum sonitu ad
caelum stridentibus extulit alis . . . (Ciris 514–15); Deinde ut respexi ad comitem, ille
exuit se et omnia vestimenta secundum viam posuit. (Petr. 62.5 (Niceros speaking));
. . . eo se ab Tiberi ferrent aves . . . (Var. L. 5.43); . . . quemadmodum sese unusquisque
nostrum gerat in retinenda religione conservandisque legibus. (Cic. Ver. 46); Invol-
volum, quae in pampini folio intorta implicat se. (Pl. Cist. 729); Manlius post famam
adpulsae Punicae classis Caralis se recepit. Ea occasio Hampsicorae data est Poeno se
iungendi. (Liv. 23.40.7); Iidem Polycles et Dionysius . . . Iovem qui est in proxima
aede fecerunt . . . Venerem lavantem sese Daedalsas, stantem Polycharmus. (Plin.
76
See Flobert (1975: 387).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Reflexivity
Nat. 36.35—NB: for this example, see § 5.19 (k)); Infert se saeptus nebula (mirabile
dictu) / per medios miscetque viris, neque cernitur ulli. (Verg. A. 1.439–40); . . . ne
quo se numine mutet . . . (Verg. A. 1.674); . . . per eos me probavi in classe. (CEL
141.17 (Karanis, c. AD 110)); Sed eccum <video>. Nescio unde sese homo recipit
domum. (Pl. Aul. 177); . . . cui se Cicero nobilitandi causa sociaverat. (Hist. Aug. trig.
tyr. 21.1); . . . quo me vortam nescio (Pl. Cur. 69);
Verbs that in combination with a human subject and a reflexive pronoun would or at
least could be considered to refer to an action or a position also occur in combination
with non-human subjects and the reflexive pronoun. Such combinations are inter-
preted as (more or less spontaneous) processes or states, in which the subject is a
patient.79 An early example is (a), but the phenomenon is productive in all periods of
Latin. The decausative reflexive expression is synonymous with the decausative
passive. This is shown in (b), where Virgil coordinates the two.
(a) . . . commutatque (sc. brassica) se semper cum calore . . .
(‘ . . . and it (the cabbage) always changes with the heat . . . ’ Cato Agr. 157)
(b) Insula portum / efficit obiectu laterum, quibus omnis ab alto / frangitur
inque sinus scindit sese unda reductos.
(‘ . . . an island forms a harbour with the barrier of its sides, on which every wave from
the main is broken, then parts into receding ripples.’ Verg. A. 1.159–61)
77 78
See Flobert (1975: 388). See Flobert (1975: 402). See also Kiss (2014).
79
The expression is sometimes called ‘anticausative’. See Comrie (1989: 168); Siewierska (1984: 77ff.);
Michaelis (1998); Cennamo (1999 and elsewhere). Wehr (1995: 66) calls it ‘decausative’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The normal way to express reciprocity in Latin is to use a prepositional phrase with inter
‘among’, as in (a) and (b). Also common from Early Latin onwards is the use of two
forms of alius ‘other’ or alter ‘other of two’ in the same clause. Strictly speaking, these
expressions mean ‘one . . . another’, ‘one . . . the other’, but they may sometimes be inter-
preted in a reciprocal sense, as in (c) and (d).80 From Livy onwards the adverb invicem
(or in vicem) is used in the sense ‘mutually’, ‘reciprocally’ (OLD s.v. § 3), as in (e).
(a) Nam postquam porta exierunt, nil cessarunt ilico / osculari atque amplexari
inter se.
(‘After they left the gate, they didn’t hesitate at all to kiss on the spot and to embrace
each other.’ Pl. Mil. 1432–3)
80
For the frequency of these three devices in a number of authors, from Plautus to Apuleius, see
Fanelli (2010). For alius . . . alius, see TLL s.v. 1643.13ff. For alter . . . alter, see TLL s.v. 1746.20ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Reflexivity
(b) Sed ad bella externa prope supererant vires abutebanturque iis inter semet
ipsos certando.
(‘But for foreign wars there was almost a superabundance of resources, and men
misused them in quarrelling amongst themselves.’ Liv. 2.42.9)
(c) Alius alium percontamur: ‘Quoia est navis?’
(‘We ask one another: “Whose ship is it?” ’ Pl. St. 370)
(d) Condamus alter alterum ergo in nervom bracchialem.
(‘Then let’s put each other into a prison of arms.’ Pl. Poen. 1269)
(e) . . . cum inter se innexi rami vinculum in vicem praebeant . . .
(‘ . . . since the interwoven branches bind one another together . . . ’ Liv. 33.5.12)
Supplement:
Sed cum alius alium timet, et homo hominem et ordo ordinem . . . (Cic. Rep. 3.23);
Milites alius alium laeti appellant, acta edocent atque audiunt, sua quisque fortia
facta ad caelum fert. (Sal. Jug. 53.8);
Atque etiam ipsi inter se censores sua iudicia tanti esse arbitrantur ut alter alterius
iudicium non modo reprehendat sed etiam rescindat . . . (Cic. Clu. 122); Sic fortuna in
contentione et certamine utrumque versavit, ut alter alteri inimicus auxilio salutique
esset . . . (Caes. Gal. 5.44.14); . . . reos, qui noxii ambo alter in alterum causam
conferant . . . (Liv. 5.11.6); Iam Petreius et Iuba concucurrerunt iacentque alter alte-
rius manu caesi . . . (Sen. Dial. 1.2.10); Cf.: Eodemque modo de Aristotele et Isocrate
iudico, quorum uterque suo studio delectatus contempsit alterum. (Cic. Off. 1.4);
Cf.: . . . cum vir virum legisset . . . (Liv. 9.39.5.); Manus manum lavat. (Petr. 45.13
(Echion speaking));81
Appendix: Non-reciprocal combinations of alius in the same clause are instances
like: Hic ego aliud alii concedo. (Cic. Ver. 2.162); Ita duo deinceps reges, alius alia
via, ille bello hic pace, civitatem auxerunt. (Liv. 1.21.6).
81
For such combinations of nouns, see Landgraf (1888).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The often quoted first instance of the use of a reflexive expression instead of a passive
form is (a). However, among the many related expressions for naming, this may well
be an instance of normal true reflexive usage. Ex. (b) is not an early example of the
reflexive passive either.82 More convincing examples are (c) and (d). The inanimate
subjects cannot perform the action on themselves. In these instances the agents can
only be human. They are generic and cannot be expressed. Similar restrictions hold
for the so-called reflexive passive in Romance.83 There are no really convincing
instances.
(a) . . . Myrina quae Sebastopolim se vocat . . .
(‘ . . . Myrina, which styles itself Sebastopolis . . . ’ Plin. Nat. 5.121)
(b) . . . nec medici se inveniunt . . .
(‘All the doctors feel lost.’ Petr. 47.2 (Trimalchio speaking))
(c) At enim pallio nihil expeditius . . . facile se regit, facile reficit.
(‘But there is nothing so convenient as the pallium . . . it is easily arranged and easily
rearranged.’ Tert. Pall. 5.3—tr. Hunink)
(d) Mela rotunda quae orbiculata dicuntur sine cura toto anno servare se
possunt.
(‘Round apples, which are called orbiculata, can keep all year without any trouble.’
Pall. 3.25.18)
Some of the verb + reflexive pronoun combinations that are called ‘passive reflexives’
in the literature are in fact decausative reflexives (see § 5.25).84
There are a number of verb + reflexive pronoun combinations that have a meaning
which cannot be easily accounted for on the basis of the constituting parts. Two
examples are (a) and (b). In (c), habent has the same meaning as in (a) without a
reflexive pronoun.
(a) At sci’n quomodo tibi res se habet?
(‘But have you any idea what your situation is?’ Pl. Aul. 47)
82 83
See Petersmann (1977: 164). See Svennung (1935: 461–4) and Wehr (1995: 49–56).
84
See Bulhart (1957: 20) and Adams (2013: 711–17). At pp. 681–2, 712 Adams discusses (b) and (c)
and says that there is no external agency at all.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Intransitivization
(b) Invitata Psyche talium locorum oblectatione propius accessit et paulo fiden-
tior intra limen sese facit . . .
(‘Psyche, attracted by the allurement of this beautiful place, came closer and as she
gained a little more confidence, crossed the threshold.’ Apul. Met. 5.2.1)
(c) Bene habent tibi principia.
(‘So far so good.’ Ter. Ph. 429)
Whereas in the preceding types of reflexive expressions the reflexive pronoun can be
explained—at least diachronically—as the object of the verb, this is not the case in (a)
and (b), in which the reflexive is combined with a one-place verb. The meaning of the
reflexive expression is the same as that of the verb itself. For that reason this use of
the reflexive is called REDUNDANT. In cases like (a) the expression is probably triggered
by reflexive expressions like me recipio ‘I withdraw myself ’. Slightly different is (c):
the reflexive expression me stupeo means the same as stupeo itself ‘to be or become
stunned’, but this verb is found with an object in the meaning ‘to be astonished at’ in
Virgil and other authors, as in (d), which may have facilitated the use of the reflexive
in (c). Note in this case the presence of ipse. This redundant use of the reflexive is
found in Late Latin. Just like the use of reflexive beneficiaries (see § 10.70, Appendix)
it left its traces in the Romance languages.85
(a) . . . recepit se episcopus et vadent se unusquisque . . .
(‘ . . . the bishop withdrew himself and everyone departed.’ Pereg. 25.7)
(b) . . . humor sudoris . . . se desidet . . .
(‘The fluid of sweat . . . sinks down.’ Mulom. Chir. 220—Cf.: tumor desidet in Cels. 7.18)
(c) Se stupuit tunc ipse, reor, bene conscius oris . . .
(‘Then the prophet was stunned, I believe, well aware of his mouth . . . ’ Ennod. Carm.
1.9.33)
(d) Pars stupet innuptae donum exitiale Minervae.
(‘Some are amazed at the maiden Minerva’s pernicious gift . . . ’ Verg. A. 2.31)
5.30 Intransitivization
The term INTRANSITIVIZATION is used in this Syntax for two semantically different
phenomena. The common element is that the verbs involved are predominantly
85
For redundant reflexive pronouns, see Dahlén (1964, 1977); Flobert (1975: 390–2); Cennamo
(1999).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
two-place verbs, which, however, are also used as one-place verbs. One type of use is
the autocausative use of active forms (lavo ‘I wash myself ’). In this case one might
conjecture that the autocausative use arose through ellipsis of the reflexive pronoun,
but the phenomenon is so widespread that it seems better to regard it as a distinct
development. The other type concerns the use of an entity as the subject of the one-
place use of the verb that may also occur as the object of the two-place use. The term
that is sometimes applied to this use is ERGATIVE. In this Syntax the term ‘decausative’
is used.86 The number of verbs involved has been estimated at 260 over the whole
period of Latin.87
There are other remarkable one-place uses of two-place verbs that cannot be
accounted for in the way these two types are. A good example is (a).
The autocausative use of active forms was illustrated in § 5.19 for the verbs lavo ‘to
wash’ and moveo ‘to move’. The verb verto ‘to turn’ also offers three possibilities
for expressing ‘to turn oneself ’. Ex. (a) demonstrates the autocausative active, (b) the
autocausative reflexive, (c) the autocausative passive. Ex. (d) has a decausative
passive and a decausative reflexive. With moveo ‘to depart’ in (e) compare § 4.17.
The autocausative use of active forms is found in all periods of Latin. The autocau-
sative active forms could be convenient in verse. The number of verbs increases in
Late Latin, which may be related to the gradual disappearance of the synthetic passive
forms. The verbs involved are mainly verbs of movement and verbs of change.88
86
A monograph on the subject is Feltenius (1977). Flobert (1975: 408–9) uses the term ‘actif
ambivalent’ for these verbs.
87
Flobert (1975: 408).
88
See Feltenius (1977: 19–23). For Early Latin, see Bennett (1910: I.4–5). For Late Latin, see Haverling
(2010a: 324–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Intransitivization
(d) Praetereaque omnia haec tum intereunt cum in naturam aliam convertun-
tur, quod fit cum terra in aquam se vertit . . .
(‘And besides, all these elements perish when they undergo transmutation, which
occurs when earth turns to water . . . ’ Cic. N.D. 3.31)
(e) . . . gratias Deo agentes movimus inde.
(‘ . . . with thanksgivings to God we set off again.’ Pereg. 10.7)
Supplement (in alphabetical order by verb):
Accingunt omnes operi . . . (Verg. A. 2.235); Apstine maledictis et mi quod rogavi
dilue. (Pl. Rud. 1108); Hoc mi expedi, quo agis? # Quo tu? (Pl. Per. 215–16); Si ad
saxum quo capessit ea deorsum cadet, / errationis fecerit compendium. (Pl. Rud.
178–80); Vel consociare mihi quidem tecum licet. / Aequas habemus partis. (Pl.
Rud. 551–2); (sc. semina) Ubi vero iam corroboraverint, nihil dubium est, quin
caelestibus aquis plurimum iuventur. (Col. 5.5.6—NB: unless regarded as the
perfect of the hapax corroborasco, so TLL s.v.); . . . ceteris in campo exercentibus . . .
(v.l. se exercentibus) (Cic. de Orat 2.287); Inde in amicitiam insinuavit cum matre et
mecum simul / blanditiis, muneribus, donis. (Pl. Cist. 92–3); Annona (annonam
Ccorr.) haud multum laxaverat . . . (Liv. 26.20.11); . . . cape hoc flabellum, ventulum
huic sic facito, dum lavamur (v.l. lavamus). / Ubi nos laverimus, si voles, lavato.
(Ter. Eu. 594–5); Illa usque ad spinam mergens se caerula condit . . . (Cic. Arat.
416); . . . postquam movisse a Samo Romanos audivit . . . (Liv. 37.28.4); (sc. ulmus et
fraxinus) . . . non habent rigorem et celeriter pandant. (Vitr. 2.9.11); Quo deinde ruis?
Quo proripis? (Verg. A. 5.741); Rursum in portum recipimus. (Pl. Bac. 294); . . .
legiones equitesque ex navibus egressos iubet ex languore nauseaque reficere . . .
(B. Afr. 34.6); Cum spissaverit, ligas diligenter. (Apic. 5.1.1); Si ad ultores transferre
prospexit . . . (Tert. Apol. 25.4);
The decausative use of active forms is found from Early Latin onwards; it is charac-
teristic of the specialized terminology of didactic texts in particular. An early
example is (a) (which might be disputed and regarded as autocausative, because
in the Classical period and later a reflexive and a passive expression—both
autocausative—exist as well). Note in (b) coordination with the regular one-place
verb evenio.
(a) Foris aperit.
(‘The door is opening.’ Pl. Per. 300—Cf. Ecquis has aperit fores? Pl. Mos. 900)
(b) Utique tu (sc. Mars) fruges, frumenta, vineta virgultaque grandire beneque
evenire siris.
(‘ . . . and that thou permit my harvests, my grain, my vineyards, and my plantations
to flourish and to come to good issue.’ Cato Agr. 141.2)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) His rebus pace confirmata post diem quartum quam est in Britanniam
ventum naves XVIII . . . ex superiore portu leni vento solverunt.
(‘Peace was thus established. Four days after the arrival in Britain the eighteen
ships . . . weighed anchor, in a gentle breeze, from the upper port.’ Caes. Gal. 4.28.1—
Cf. Caes. Civ. 3.101.5 . . . naves conscenderunt et a terra solverunt . . . (sc. naves))
Supplement (in alphabetical order by verb):
Thessala Dardaniis tunc primum puppis harenis / adpulit et fatis Sigeo litore sedit.
(V. Fl. 2.445–6); Gaudebam sane et congelasse nostrum amicum laetabar otio. (Cic.
Fam. 2.13.3); Omnis creta coquet . . . (Plin. Nat. 18.34); Non tibi decoquet [non]
ornithon. (Var. R. 3.2.16); (sc. parietes) Molli enim et rara potestate cum sint,
exsiccant sugendo e materia sucum. (Vitr. 2.8.2); Ingeminant fluctus et primos
ultimus urget. (Aetna 323); Nam iam aliquo aufugiam et me occultabo aliquot dies,
/ dum haec consilescunt turbae atque irae leniunt. (Pl. Mil. 582–3); Varro libris quos
ad Marcellum de lingua Latina fecit: in priore verbo graves prosodiae quae fuerunt
manent, reliquae mutant; <mutant> inquit elegantissime pro ‘mutantur’. (Gel.
18.12.8); . . . ita iacuisse per aliquot dies, ut nihil odor mutaret. (Liv. 3.10.6); (sc. terra)
Quasi non . . . aquis ante segregatis superstiti limo siccaverit. (Tert. adv. Val. 24); . . .
classemque sub ipsa / Antandro et Phrygiae molimur montibus Idae / incerti quo fata
ferant, ubi sistere detur . . . (Verg. A. 3.5–7); Gangraenam vero, si . . . nondum plane
tenet, sed adhuc incipit, curare non difficillimum est. (Cels. 5.26.34.A); . . . cum mare
turbaret ac per tempestatem macellum piscinarum [mare] obsonium praeberet . . . (Var.
R. 3.17.7); (sc. oculi) Ardent, intenduntur, umectant, conivent. (Plin. Nat. 11.146);
Appendix: ‘Latin does not have a productive morphological process for the forma-
tion of causative constructions.’89 A causative relation can be expressed lexically with
the verb facio ‘to make’ and various manipulation verbs. There are also a number of
compounds of facio, such as calefacio ‘to make hot’ (compare the verbs caleo ‘to be
hot’ and calesco ‘to become hot’, and the adjective calidus ‘hot’).90
5.33 Deponents
Deponent verbs share with verbs used in the passive their passive morphology, but they
are different in several other respects. The first defining difference is that there are no
active finite forms corresponding to the passive forms and that deponents have no
passive meaning (except in certain verb forms; see § 5.34). The second characteristic of
deponent forms is that a number of non-finite forms, viz. the present and future
participles, as well as the future infinitive, the gerund, and the supines, are active.
Thirdly, usually both the simple verb and its related compound verbs have only passive
forms. Fourthly, there are syntactic differences and different conditions of use.91
89
Quotation from Lehmann (forthc.).
90
For causative constructions in Latin translations of the Bible, see Hoffmann (2008). See also Fruyt
(2001) and Fruyt and Orlandini (2008: 233–5).
91
For these criteria, see Flobert (1975: 31–5).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Deponents
Deponents resemble autocausative passive verb forms in that the latter also have
the active non-finite forms mentioned in § 5.19. The difference is, of course, that
autocausative passive verb forms have corresponding active verb forms. Deponents
may be one-, two-, or three-place, just like verbs with active verb forms. Autocausa-
tive passive verb forms may govern an object (see § 5.20) and in that respect resemble
two-place and three-place deponent verbs. Passiva tantum like plector ‘to be beaten’
differ from deponents in that they are related to simple or compound verbs that do
have both active and passive forms (and they are semantically passive).
Many deponents have active variants, with the same syntactic and (roughly) the
same semantic properties, but this phenomenon only occurs on a wide scale in the
case of a limited number of verbs. Examples are assentio/assentior ‘to agree’ and
mereo/mereor ‘to earn’ (for small differences in the distribution of the various senses
in which mereo/mereor are used, see OLD s.v.). In Late Latin there is much more
variation than in the earlier periods; indeed, in texts from the fifth century onwards
variation is the rule.
In the course of the development of Latin, some deponents were ‘activated’: that is,
they manifest active morphology (of the 270 deponents in Early Latin fifty-one were
‘activated’); conversely, twenty-nine out of 270 deponent verbs in Early Latin were
reinterpreted as passives. Verbs that became active sometimes also developed passive
forms with a passive meaning (for the passive use of deponents see § 5.34). The
general trend is an increase from 270 deponents in Early Latin to 884 at the end of the
period dealt with by Flobert (Charlemagne). The numbers for individual authors are:
Plautus 252, Cicero 291, Seneca 253, Apuleius 281, Tertullian 254. Overall 7% of the
verbs in a text are deponents (verb + reflexive pronoun combinations amount to
more or less 1%). Some of these verbs are very common, like loquor ‘to talk’, nascor
‘to be born’, patior ‘to suffer’, sequor ‘to follow’.92 The details about individual verbs
can be found in the OLD. (For semideponents, see § 3.12.)
The use of a number of verb forms of deponent verbs with a passive meaning is
widespread and is found in all periods of Latin. The number of verbs with which this
is possible increases steadily, with individual variation in creativity between authors.
The existence of passively used forms does not necessarily presuppose the existence of
morphologically active forms. The most common deponent forms with a passive
meaning are perfect participles, often coordinated with regular passive perfect parti-
ciples, as in (a). The next major category of deponent forms with a passive meaning are
present infinitives, also often coordinated with a regular passive infinitive, as in (b).
92
Further details can be found in Flobert (1975: passim, very conveniently summarized on pp. 588–93,
and with a diachronic table on p. 619).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Infectum forms are in the minority in most authors. Vitruvius’ work is mentioned as a
text with a relatively ample use of infectum forms.93
(a) Cum autem (sc. oratio) ingressa est imitata et efficta simulacra . . .
(‘When, however, the reasoning has entered the domain of imitated or created
images . . . ’ Cic. Tim. 8)
(b) . . . hac diligentia . . . nihil eorum investigari, nihil adsequi potuerit.
(‘ . . . the assiduous efforts . . . have failed to track down or lay hold of any of them . . . ’
Cic. Ver. 2.181)94
(c) Namque si duae columnae aeque crassae lineis circummetientur . . .
(‘For if two columns of equal diameter . . . have a line measured round them . . . ’ Vitr. 4.4.3)
(d) . . . et sustinetis Iovis elogia modulari.
(‘ . . . and you accept that poems in praise of Jupiter are being sung.’ Tert. Nat. 1.10.45)
Impersonal passive deponent forms are rare.95 Examples of verbs that occur more
than once are (e)–(g).
(e) Quaeso hercle propera. # Non morabitur.
(‘And, by Hercules, hurry! # There will be no delay.’ Pl. Mil. 1305)
(f) Homo . . . primo vetat sortiri, iubet extra sortem Theomnastum renuntiari.
(‘The man first tried . . . forbidding the lot, ordering Theomnastus to be returned as
appointed without it.’ Cic. Ver. 2.127)
(g) Itaque, ut huic vitio medeatur, sic erit faciundum, ut . . .
(‘To avoid this damage, therefore, we must proceed as follows . . . ’ Vitr. 6.8.6)
Gerundive forms of deponent verbs with a passive interpretation (see § 5.37 for
discussion of the passive value of the gerundive) are found throughout Latinity, both
in the impersonal and personal complex expressions with the auxiliary sum, as in
(h)–(j), and (k)–(m), respectively, and in the other uses of the gerundive, as in (n)–(p).
(h) . . . et assentandum est quicquid hic mentibitur.
(‘ . . . and I have to agree with whatever lies he dishes up.’ Pl. Mil. 35)
(i) Quo minus admirandum est eum facere illa qui immortalitatem concupiverit.
(‘So that is less a matter for wonder that he should so act who has set his heart on
immortality.’ Cic. Rab. Post. 42)
(j) . . . rescripsit gratandum esse, si . . .
(‘ . . . he wrote back that it would have been a cause for rejoicing if . . . ’ Amm. 17.3.5)
93
For details see Flobert (1975: 343–80), from which most of the examples are taken.
94
Following the reading of the most important manuscripts, with Flobert against editors who print
active investigare.
95
A list can be found in Flobert (1975: 344–5).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
96
Bolkestein (1989:11) observes that there are more limitations on the use of the gerund and gerundive
than on participles, which in themselves are already more limited than finite active and passive verb
forms.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
This section deals with instances of the gerund where it competes with the gerundive.
Most instances of the gerund of two- and three-place verbs have been described as
‘active’, in all four constructions in which the gerund is used. Ex. (a) is an instance of
the use of the gerund as an attribute at the noun phrase level (more examples in
Chapter 17). In (b), it functions as an argument at the adjective phrase level with
cupidus (more examples in Chapter 17). In (c), it is a dative object governed by the
expression operam do (more examples in § 15.136). In (d) and (e), it is a satellite
(more examples in §§ 16.99–16.104). Only (b) and (c) have an explicit object
constituent governed by the gerund (te and Epidicum, respectively).
(a) Qui modus dandi?
(‘What limit is there to giving?’ Pl. As. 167)
(b) Tum Caesar ‘Equidem’ inquit ‘Crasse, ita sum cupidus . . . te . . . audiendi . . . ’
(‘ “For my part, Crassus,” returned Caesar, “I am so desirous of hearing you . . . ” ’ Cic.
de Orat. 2.16)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
All five embedded predications are indeed actions. The agents involved are to be
supplied from the context or situation. In (a), no specific agent is referred to, whereas
in the other four examples the agent is identical to the subject of the main clause. The
semantic function of that subject constituent is of no importance: it may be an agent,
as in (c), but it may also be a patient (as in (d) and (e)). Since the agent of the
embedded actions can easily be supplied, the gerunds are interpreted as active.
However, there are also instances where it is not so easy to supply an agent for the
embedded action, and, if in such a context the verb involved is also a two- or three-
place one and the unexpressed patient can be identified with a constituent of the main
clause, this gives rise to a passive interpretation.97 Examples are (f) and (g). In (f), the
frequentia comes ‘to be registered’ by the censors. In (g), the animals are the ones ‘to
be eaten’; this example shows that such a passive interpretation is also possible with
gerunds of deponent verbs. In (h), however, no specific agent is meant either, but
there is no inclination to interpret inveniundo as passive. Note the parallelism with de
genere orationis.
(f) . . . quae (frequentia) convenit uno tempore undique comitiorum, ludorum
censendique causa.
(‘ . . . these multitudes that have simultaneously assembled, from all parts of Italy to
attend the elections and the games and to be registered by the censors.’ Cic. Ver. 54)
(g) . . . pecudes, quod . . . partim esse ad usum hominum, partim ad fructum,
partim ad vescendum procreatas.
(‘ . . . animals, because some of them have been created to be man’s slaves, some to
supply him with their products, and others to be eaten.’ Cic. Leg. 1.25)
(h) Tibi enim tantum de orationis genere quaerenti respondi etiam breviter de
inveniundo et conlocando.
(‘ . . . you asked only about the use of language, and I included a brief treatment of
invention and arrangement in my answer.’ Cic. Orat. 54)
Supplement:
Other seemingly ‘passive’ instances: (sc. iuvenci) . . . diebus paucis erunt mansueti et
ad domandum proni. (Var. R. 1.20.2); Quin etiam ceteris quae moventur hic fons,
hoc principium est movendi. (Cic. Tusc. 1.53); Eis ad ignoscendum nulla facultas est
97
For further examples and discussion, see Maraldi (1994: 151–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
data et a militibus nostris interfecti sunt. (B. Hisp. 12.2); Ulcus enim vivescit et
inveterascit alendo . . . (Lucr. 4.1068); Hostis Antonius iudicatus Italia cesserat. Spes
restituendi nulla erat. (Nep. Att. 9.2); Igitur Iugurtha . . . quom ipse ad imperandum
Tisidium vocaretur, rursus coepit flectere animum suom . . . (Sal. Jug. 62.8); Pinguis
item quae sit tellus, hoc denique pacto / discimus. Haud umquam manibus iactata
fatiscit, / sed picis in morem ad digitos lentescit habendo. (Verg. G. 2.248–50); Tum
sepeliendi causa conferri in unum corpora suorum iussit. (Liv. 22.52.6); Athenas
quoque erudiendi gratia missus. (Just. 17.3.11);98
Further remarkable instances: . . . erudiunt iuventutem venando currendo, esuriendo
sitiendo, algendo aestuando. (Cic. Tusc. 2.34); CULPAND<O> NIHILUM DIGNUS, SET
DIGNUS AMARI (Carm. Epigr. 650.2 (AD 360));
Whereas the interpretation in (f) and (g) appears to be truly passive, (i) is different.
Here, the recipient of dat is the one who receives the right to ‘defend himself ’:
defendendi is the gerund that corresponds to autocausative passive defendor. One
may compare (j).99
(i) . . . dat ipsa lex potestatem defendendi . . .
(‘ . . . the law itself authorizes self-defence . . . ’, Cic. Mil. 11)
(j) Si apud Vitellii legatum defenderer, neque facto meo venia neque dictis fides
debebatur.
(‘If I were defending myself before a legate of Vitellius, my acts would deserve no
pardon nor my words any credence.’ Tac. Hist. 5.26.2 )
Supplement:
. . . ad fugam hortatur . . . turpitudo coniungendi cum tyranno. (Cic. Att. 7.20.2);
. . . quibus ad recipiendum crates . . . impedimento fuerunt. (Caes. Civ. 3.46.5); . . . hos
autem tempore ad comparandum dato fecit robustiores. (Nep. Thras. 2.2); Is cum in
pace instituisset pueros ante urbem lusus exercendique causa producere . . . (Liv.
5.27.2); . . . ni robur legionum perexiguo ad instruendum dato tempore aciem dire-
xisset. (Liv. 28.22.13);
The gerundive is used in a wider range of constructions than the gerund and these
various uses must be examined separately. In some of its uses the gerundive refers to
something that must or (when negated) must not be done (it has a ‘deontic’ value).
This is the case in its combination with the auxiliary verb sum, which functions as a
kind of passive deontic verb form, illustrated by the impersonal expression in (a) and
the personal one in (b). The gerundive can, in principle, be combined with all finite
98
A list of occurrences that must or can be interpreted as passive can be found in Calboli-Montefusco
(1971: 473–4).
99
More instances in K.-St.: I.729 and in Flobert (1975: 412–19).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
and non-finite forms of sum.100 The passive value of these expressions is apparent
from the fact that the agent may be expressed explicitly, as in (c) (for details see
§ 5.40). This value is artistically exploited in (d) where it is contrasted with the active
present participle metuens. The attributive use in a noun phrase in (e) also has this
passive deontic value.
(a) Parcendum autem maxime est caritati hominum . . .
(‘You must be especially considerate of popular esteem . . . ’ Cic. de Orat. 2.237)
(b) . . . etiamsi navigari posset Oceanus, navigandum non esse.
(‘ . . . even if the ocean could be sailed, it should not be.’ Sen. Suas. 1.8)
(c) Mihi etsi Kalendae vitandae fuerunt . . .
(‘As for me, although I needed to avoid the Kalends . . . ’ Cic. Att. 12.5c)
(d) . . . metuendus magis quam metuens ‘Romanus sum’ inquit ‘civis’.
(‘ . . . yet as one more to be feared than fearing, “I am a Roman citizen,” he cried.’ Liv.
2.12.9)
(e) Et Hasdrubalem propediem adfore cum manu haudquaquam contemnenda.
(‘And Hasdrubal, they said, would soon arrive with a force by no means to be
despised.’ Liv. 30.7.10)
Supplement:
Mille modis, Amor, ignorandu’s, / procul adhibendu’s atque apstandu’s. (Pl. Trin.
264–5); Sed quid video? Eram atque ancillam ante aedis. Adeundae haec mihi. (Pl.
Truc. 895); Accipiunda et mussitanda iniuria adulescentium’st. (Ter. Ad. 206);
Promunt condita aut propterea quod sunt tuenda, aut quod utenda, aut quod
vendunda. (Var. R. 1.62.1); Concubitus vero neque nimis concupiscendus, neque
nimis pertimescendus est: rarus corpus excitat; frequens solvit. (Cels. 1.1.13); . . .
prohibendus bos potione per biduum . . . (Pallad. Vet. med. 7.3);
The gerundive is, however, also used in a number of constructions in which it is more
difficult to apply the notion of ‘passive’ and where the notion ‘deontic’ does not seem
to apply at all. This is illustrated by (f) and (g). In (f), the gerundive is used as a
secondary predicate, discussed in Chapter 21. In (g), the ablative gerundival clause
(included in curly brackets) can only be understood in an active sense, and no
obligation is involved. Further examples and discussion of these types follow in § 5.39.
(f) . . . patriam ipsam . . . inflammandam reliquimus.
(‘ . . . we have left our mother city herself to burning.’ Cic. Fam. 16.12.1)
(g ) . . . placet . . . contra gaudere nosmet {omittendis doloribus} . . .
(‘ . . . but on the other hand one is glad to lose a pain.’ Cic. Fin. 1.56)
100
The material can be found in Neue-W.: III.185–8.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
It is shown in § 5.6 and § 5.21 that the normal personal passive is not found with all
verbs and that the impersonal passive is in this respect more productive. The same
holds roughly for the passive/deontic personal and impersonal uses of the gerundive.
There are (almost) no restrictions with respect to the types of verbs in the case of the
impersonal gerundive: just as we find itur of the one-place verb eo we find eundum
est, as in (a) and—with an explicit agent nobis—in (b). Similarly, just as we find
impersonal parcitur of the dative governing two-place verb parco, we find parcendum
est with a dative object caritati in (c). An example, with an explicit agent and an
ablative object, is (d). Even more striking is the impersonal use of two- and three-
place verbs that govern an accusative object, as in (e), instead of the personal
gerundive construction. This phenomenon is almost limited to Lucretius and
Varro, and in later times to juridical authors, but there are isolated instances in all
sorts of texts throughout Latinity (some of them merely evidence of ignorance of the
rules of Latin).101 (For putative parallel ‘normal’ impersonal passive instances with an
accusative object, see § 5.21.)
(a) . . . tamen est eundum quo imperant ingratiis.
(‘ . . . one still has to go where they command despite oneself.’ Pl. Cur. 6)
(b) Nunc quoniam iam est Cyzicum nobis eundum . . .
(‘Now that I have to go to Cyzicus . . . ’ Cic. Att. 3.13.2)
(c) Parcendum autem maxime est caritati hominum . . .
(‘You must be especially considerate of popular esteem . . . ’ Cic. de Orat. 2.237)
(d) Constantiam probo, qua mihi quoque utendum fuit . . .
(‘Though I do approve their consistency, a consistency of which I too have been
called upon to avail myself . . . ’ Cic. Dom. 19)
(e) . . . mi advenienti hac noctu agitandum est vigilias.
(‘ . . . I do believe that on my arrival I need to keep watch this night.’ Pl. Trin. 869)
101
See Aalto (1949: 94–7), Risch (1984: 186–8), Roby (1882: lxxii–lxxiii) for material; Fugier (2002:
137–8) for discussion. Calboli (1962a: 20–6) regards agitandum as a noun.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
One-place verbs: Inambulandum est. (Pl. As. 682); Exi, inquam, age exi. Exeundum
hercle tibi hinc est foras . . . (Pl. Aul. 40); Credo ego istoc exemplo tibi esse pereun-
dum extra portam . . . (Pl. Mil. 359); . . . mihi aut cum his vivendum aut pro his esse
moriendum. (Cic. Catil. 2.27); Non faces . . . si igni, non telis pugnandum est, ereptas
ultro inferetis? (Liv. 4.33.4); Lugendum est, flendum est. In hoc me servasti? (Sen.
Con. 4.5); Semel haec mihi videnda sint, an saepe nascendum? (Sen. Ep. 65.20);
Hiems frigora adducit. Algendum est. Aestas calores refert. Aestuandum est. In-
temperies caeli valitudinem temptat. Aegrotandum est. (Sen. Ep. 107.7);
Genitive governing verbs: Consili nostri, ne si eos quidem qui id secuti non sunt non
paeniteret, nobis paenitendum putarem. (Cic. Fam. 9.5.2); Si rerum natura pateretur,
obliviscendum erat mihi patris, dum occiderem. (Sen. Con. 7.5.5);
Dative governing verbs: Linguae moderandum est mihi. (Pl. Curc. 486); Nec vero
corpori solum subveniendum est, sed menti atque animo multo magis. (Cic. Sen. 36);
Nec vero audiendi qui graviter inimicis irascendum putabunt . . . (Cic. Off. 1.88);
Omnibus modis huic rei studendum ut pabulatione aut commeatu Romani prohi-
beantur. (Caes. Gal. 7.14.2); . . . potiusque sero quam numquam obviam eundum
audaciae temeritatique. (Liv. 4.2.11); Sed studendum est concordiae viribus totis . . .
(Amm. 26.2.8);
Ablative governing verbs: . . . annum tibi illum unum domo carendum esse mere-
tricis? (Cic. Ver. 5.38); . . . vereor ne non tam virtutis fiducia nitendum nobis ad spem
beate vivendi quam vota facienda videantur. (Cic. Tusc. 5.2); . . . mihi non modo meis
sed memet ipso carendum est. (Cic. Att. 8.7.2); Eius bono fruendum est igitur, si beati
esse volumus. (Cic. Tusc. 5.67); . . . et carendum non solum crimine turpitudinis,
verum etiam suspicione. (Quint. Inst. 2.2.14);
Preposition phrase governing verbs: De quibus nobis dicendum erit. (Cic. Inv. 2.90);
Inter caelestia de terra dicendum erit. (Sen. Nat. 2.1.4);
Accusative governing verbs: Nec tamen sequendum in seminio legendo Tanagricos . . .
(Var. R. 3.9.6); . . . vasa vinaria et olearia potius faciendum (est). (Var. R. 1.13.1);
Canes potius cum dignitate et acres paucos habendum quam multos . . . (Var. R.
1.21.1); Volumus sane, nisi molestum est Cato, tamquam longam aliquam viam
confeceris quam nobis quoque ingrediundum sit, istuc quo pervenisti, videre quale
sit. (Cic. Sen. 6); . . . et istos consulatus non flocci facteon. (Cic. Att. 1.16.13); . . .
PORTAS·TURREIS MOIROS / TURREISQUE AEQUAS·QUM MOIRO / FACIENDUM COIRAVERUNT.
(CIL I2.1722.3–5 (Aeclanum, c.80 BC)); . . . et aestimato, / quot pondo est tibi mentu-
lam cacandum. (Priap. 69.3–4); Quare monendum est <te> mihi bone Egnati.
(Catul. 39.9); . . . aeternas quoniam poenas in morte timendum. (Lucr. 1.111); Mul-
taque in his rebus quaeruntur multaque nobis / clarandum’st, plane si res exponere
avemus. (Lucr. 4.776–7); . . . pacem Troiano ab rege petendum. (Verg. A.
11.225); . . . nunc pacem orandum, nunc improba foedere rupto / arma reponenda
et bellum exitiale cavendum / auctor ego. (Sil. 11.559–61); Multa dicendum fuit . . .
(Tert. Pal. 4—Greek influence assumed); Neque enim admittendum esse distinctio-
nem existimantium . . . (Papin. dig. 15.1.50.3); Iuliani sententiam sequendum est
(Tryph. dig. 49.15.12.2); . . . denegandum ei fructus petitionem Iulianus scribit (Paul.
dig. 39.2.10); . . . quod Sabinus scribit debere a capite rationem reddendum . . . (Paul.
dig. 3.5.18.1);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Praeterea sub periculo eligendum102 viarum duces idoneos (Veget. Mil. 3.6.5); XXX
psalmos expedite decantandum (Greg. T. Stell. 39);
UT . . . ROSAS (sic!) EIS DEDUCANTUR ET / CIBOS PONENDUM (CIL V.4017.10–11 (Peschiera
del Garda));
Turning now to the personal use of the gerundive, one obviously does not expect it to
occur with one-place verbs (for a few exceptions, see the Appendix). It is common with
two- and three-place verbs governing an accusative, less common with deponent verbs
(more examples in § 5.34)103 and rare with verbs governing a non-accusative object.
Two-place verbs governing a genitive object are not attested in the personal gerundive;
those usually governing a dative are very rare (and late)—an instance is (f). There are,
however, a few instances of gerundives of two-place verbs governing an ablative object,
such as (g) of the (deponent) verb utor ‘to use’. In most cases of personal gerundives of
verbs governing a non-accusative object, these verbs have either a parallel construction
with the accusative or are (mainly) found with the accusative in Early Latin texts (see
§ 4.33). This is the case with utor and other deponent verbs normally governing an
ablative object. Sometimes a verb has several verb frames: for example, the verb
suscenseo, which governs a dative object in its meaning ‘to be angry with’ but is (rarely)
also used as a three-place predicate with a pattern ‘to blame somebody (dative) for
something (accusative)’,104 which may explain (h).
102
Önnerfors now reads: eligend<ar>um.
103
Examples can be found in Flobert (1975: 347–8).
104
Psyllos re aquaria defectos eam iniuriam graviter Austro suscensuisse decretumque fecisse, uti
armis sumptis ad Austrum proinde quasi ad hostem iure belli res petitum proficiscerentur. (Gel. 16.11.6).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
etiam si non sunt dolenda. (Cic. Fam. 12.23.1); . . . tamen laetandum magis quam
dolendum puto casum tuom. (Sal. Jug. 14.22); Hortandi etiam locutionem tentare
. . . (Cael. Aurel. Chron. 2.41);
The same rules that apply to the ‘normal’ passive also apply to two-place verbs. Two
additional examples, apart from vendunda in (g), are (i) and (j). As with normal true
passives the agent may be optionally expressed.
(i) Consolandus hic mi est.
(‘He needs my consolation.’ Pl. Bac. 625)
(j) . . . excitandus nobis erit ab inferis C. Marius . . .
(‘ . . . we shall be constrained . . . to summon up from the shades of C. Marius . . . ’ Cic.
Font. 36)
Supplement:
. . . et illud autem inserviendum est consilium vernaculum. (Pl. Poen. 927); Siquis
quid reddit, magna habenda’st gratia. (Ter. Ph. 56); Omni autem in re consensio
omnium gentium lex naturae putanda est. (Cic. Tusc. 1.30); Non erat dubium,
quantum in regressu discriminis adeundum foret frumentatoribus onustis percul-
sisque. (Tac. Hist. 4.35.3);
When the gerundive is used as a secondary predicate, the same range of verbs as
mentioned above can be found, as is shown in (k)–(m). There are many more
instances of gerundives of deponent verbs used as secondary predicates than there
are of those used personally and impersonally in combination with the auxiliary,
discussed in the preceding paragraph.
(k) . . . quae utenda vasa semper vicini rogant . . .
(‘ . . . things the neighbours are all the time wanting to borrow . . . ’ Pl. Aul. 96)
(l) . . . patriam ipsam . . . inflammandam reliquimus.
(‘ . . . we have left our mother city herself to burning.’ Cic. Fam. 16.12.1)
(m) . . . Siciliae quae mihi defendenda tradita est . . .
(‘ . . . Sicily whose defence has been entrusted to me . . . ’ Cic. Ver. 5.188)
In these examples, the object constituents quae vasa and patriam in (k) and (l)
and the subject constituent quae in (m) are also engaged as patients in certain
actions that are going to take place, the agents of which are not expressed but
have to be inferred from the context. In (k), it is vicini, the subject constituent of
rogant, who are going to use the vasa. In (l) it is the unnamed recipient of
reliquimus, in (m), it is mihi, the recipient of tradita est, who are the expected
or required agents of the actions of inflammare and defendere, respectively. More
instances can be found in Chapter 21. For the (seeming absence of the) deontic
value see § 5.41.
In gerundival clauses there seem to be no formal restrictions on gerundives of two- and
three-place verbs. Both deponent verbs and verbs governing non-accusative (including
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
prepositional) arguments are found in this way.105 Two examples of deponent verbs
governing a dative are (n) and (o) (the clauses are between brackets). Of these, insidior
governs a dative object (with medeor the dative is more common, but the accusative can
be used as well). Further examples can be found in § 15.140.
(n) . . . ita dextere egit ut {medendis corporibus} animi . . . militum imperatori
reconciliarentur . . .
(‘ . . . he showed so much tact that the feelings of the men were much sooner won
over to their commander now that their bodies were being properly looked after . . . ’
Liv. 8.36.7)
(o) Magnum crimen vel {in legatis insidiandis} vel . . .
(‘It is a grave charge against a man, that he either plotted against the life of an
ambassador, or . . . ’ Cic. Cael. 51)
Supplement:
Nam illi quidem communes loci, aut qui calumniae accusatorum demonstrandae aut
misericordiae captandae aut facti indignandi aut a misericordia deterrendi causa
sumuntur, ex periculi magnitudine, non ex causae genere ducuntur. (Cic. Inv.
2.56); . . . et ingrediar in disputationem ea lege qua credo omnibus in rebus disse-
rendis utendum esse . . . (Cic. Rep. 1.38);
105
Poor knowledge of Latin may explain Callistratus’ active use of—two-place—fungendas in: Mu-
lieres quae in matrimonium se dederint non legitimum non ibi muneribus fungendas unde mariti earum
sunt sciendum est, sed unde ipsae ortae sint. (Call. dig. 50.1.37).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
106
For the position that the gerundive in this construction is voice-neutral, see Heberlein (1989) and
Hettrich (1993). The ideas of the Roman grammarians are discussed by Calboli-Montefusco (1971).
107
For adjectives of the oriundus type as vestiges of an earlier middle voice system, see Pultrová (2011).
108
Some other remarkable Late Latin cases of such ‘active’ gerundives can be found in Svennung
(1935: 426–9), Aalto (1949: 145–6), and Risch (1984: 111).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Cui testis aestas et hiems, quod in altera aer ardet et spica aret, in altera natura ad
nascenda cum imbre et frigore luctare non volt . . . (Var. L 5.61); Seclum spatium
annorum centum vocarunt, dictum a sene, quod longissimum spatium senescen-
dorum hominum id putarunt. (Var. L. 6.11); Genethliaci quidam scripserunt, inquit
(Varro) esse in renascendis hominibus quam appellant ƺتªÆ Graeci. (Varro
ap. August. Civ. 22.28); AEDIS·FLORAE QUAE·REBUS FLORESCENDIS PRAEEST DEDICATA EST-
1
PROPTER STERILITATEM·FRUGUM (CIL I , p. 236.28 (Fasti Praenestini, between 6 and
9 AD)); Semel haec mihi videnda sint, an saepe nascendum? (Sen. Ep. 65.20); . . .
in moribus inolescendis magnam fere partem ingenium altricis et natura lactis
tenet . . . (Gell. 12.1.20); . . . exque eo genitos vel nascendos filios [‘sons yet to be
born’!] similiter paternae condicioni subiacere praecipimus. (Cod. Just. 5.27.4);
In the case of both impersonal and personal use of the gerundive + sum with a passive
semantic value the agents are human beings, whether they are expressed or not. Other
animate beings (or animate by implication) are rare. This is a semantic restriction that
is responsible for the fact that utterances like (a) are not attested.109
(a) *Arbori crescendum est.
(‘The tree should grow.’)
Animate by implication is (b). A real exception is (c): there is no identifiable human
instigator for ciendum. Just as with ‘normal’ passive clauses explicit agents are
relatively rare. Much more often they have to be inferred from the context.110
(b) Nam neque is qui optime potest deserendus ullo modo est a cohortatione
nostra neque is qui aliquid potest deterrendus.
(‘For by no means must a man of the highest capacity be left without our
encouragement, or one of any ability scared away . . . ’ Cic. de Orat. 2.86)
(c) Usque adeo prius est in nobis multa ciendum / quam . . .
(‘So true is it that many particles must be moved in us before . . . ’ Lucr. 3.391–2)
The typical formal expression of the human being involved is the dative case, which is
found from Early Latin onward. Examples are (d)–(g). With the regular passive the
dative is much more exceptional (see § 5.8). The use of the dative with the gerundive
can be compared with its use with adjectives in -bilis, such as (in)habilis ‘easy/difficult
to handle’. The interpretation as an agent may originally have been dependent on the
context.111
109
The example is taken from Bolkestein (1980: 138–9).
110
For limited quantitative data (from Terence), see Shalev (2001).
111
For a discussion of the relation of the use of the dative to mark an agent with other uses, see Suárez
(2001).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The first attested instance of a gerundive expression with an ab agent is (h), where a
dative omnibus would be ambiguous (it could also go with videtur ‘it seems to
everybody’). Most of the ab-agents in Cicero are explained in this way. The jurists
also have a strong preference for the unambiguous ab expression.112 However, in
general it is much less frequent than the dative (in Cicero’s orations twenty-three
instances as against 457 datives according to Lebreton 1901a: 412–14). Just as in the
case of the dative, the agents are always human. The use of ab + ablative agents with
the gerundive instead of the dative is regarded as more emphatic by certain
scholars,113 but there are instances where the emphasis is clearly on the verb, as in
(i). There is no semantic difference between the two expressions.
(h) . . . id defendimus quod ab omnibus defendendum videtur . . .
(‘ . . . we either defend what seems to deserve defence by all men . . . ’ Rhet. Her. 1.5)
(i) . . . tamen admonendum potius te a me quam rogandum puto.
(‘ . . . I conceive that from me to you a hint is more appropriate than a request.’ Cic.
Fam. 15.4.11)
112
For the avoidance of ambiguity as a reason for using the ab expression, see Baldi (1983: 21–3).
113
So K.-St.: I.730.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Haec qui prospexerunt, maiores nostros dico, Quirites, non eos in deorum immor-
talium numero venerandos a nobis et colendos putatis? (Cic. Agr. 2.95); Aguntur
bona multorum civium quibus est a vobis et ipsorum causa et rei publicae consu-
lendum. (Cic. Man. 6); Nunc mihi tertius ille locus est relictus orationis, de ambitus
criminibus, perpurgatus ab eis qui ante me dixerunt, a me, quoniam ita Murena
voluit, retractandus. (Cic. Mur. 54—NB: parallelism); Nec, si a populo praeteritus est
quem non oportuit, a iudicibus condemnandus est qui praeteritus non est. (Cic.
Planc. 8—NB: parallelism); . . . cui senatus pro me gratias agendas putavit ei ego a me
referendam gratiam non putem? (Cic. Planc. 78); Patris autem lenitas amanda potius
ab illo quam tam crudeliter neglegenda. (Cic. Att. 10.4.6); . . . nec coniugis umquam /
busta meae videam neu sim tumulandus ab illa. (Ov. Met. 8.702–3); Ab illis enim
cultus in regem exemplum esse prodendum. (Curt. 8.5.12); Itaque quod de universo
dicebam ad haec transfer quae ab homine facienda sunt. (Sen. Ep. 65.3); . . . sed
damnum . . . ab his sarciendum erit qui . . . (Papin. dig. 31.78.2); . . . coheredem esse
defendendum ab eo qui . . . (Ulp. dig. 10.2.20.8);
This deontic meaning explains why it occurs in the same contexts as both verbs like
debeo and oportet and grammatical directive expressions with imperatives or sub-
junctives. The following passage from Vitruvius in (c) illustrates this point.115
(c) Sin autem loci natura interpellaverit, tunc convertendae sunt earum regio-
num constitutiones, uti quam plurima pars moenium e templis deorum
conspiciatur. Item si secundum flumina aedes sacrae fient, ita uti Aegypto
circa Nilum, ad fluminis ripas videntur spectare debere. Similiter si circum
vias publicas erunt aedificia deorum, ita constituantur uti praetereuntes
possint respicere et in conspectu salutationes facere.
114
The Latin expression for this deontic aspect of the meaning of the gerundive is ‘notio necessitatis’.
For an examination of the semantic properties of the gerundive in comparison to other deontic
expressions, see Bolkestein (1980: 134–45). See also Joffre (2002).
115
The overall semantic similarity also explains the so-called redundant usage of the gerundive in
combination with the governing verb oportet, for which see TLL s.v. oportet 743.37ff. and Löfstedt (1933:
II.161–2). For the use of gerundives in didactic texts, see Adams (1995b: 197–200).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(‘But if the nature of the site is such as to forbid this, then the principle of determining
the quarter should be changed, so that the widest possible view of the city may be had
from the sanctuaries of the gods. Furthermore, if a temple is to be built beside rivers, as
in Egypt on both sides of the Nile, it ought, as it seems, to face the river banks.
Similarly, houses of the gods on the sides of public roads should be arranged so that the
passers-by can have a view of them and pay their devotions face to face.’ Vitr. 4.5.2)
As for the use of the gerundive in secondary predicates as illustrated in section § 5.38
above, the notion ‘deontic value’ is less suitable there. This is shown by (d) and (g).
(d) . . . patriam ipsam . . . inflammandam reliquimus.
(‘ . . . we have left our mother city herself to burning.’ Cic. Fam. 16.12.1)
(e) Patriam ipsam (eis) reliquimus ad inflammandum.
(f ) Patriam ipsam (eis) reliquimus ut inflammarent/inflamment.
(g) Ibi agrum de nostro patre / colendum habebat.
(‘There he had some land to farm from my father.’ Ter. Ph. 364–5)
Whereas the notion ‘deontic value’ is difficult to apply to the secondary predicates
discussed above, the gerundive is found with this value in expressions with the verb
habeo (‘to have’), as in (i)–(j).
(i) (sc. sapiens) Non habet mittendos trans maria legatos . . .
(‘(The wise man) has no need to send legates across the seas.’ Sen. Ben. 7.3.2)
(j) Pugnandum habebam non imperatori sed patri.
(‘I had to fight not only for my commander but for my father.’ Sen. Con. 10.2.4)
Formally legatos in (i) resembles agrum in (g)—it looks like an accusative object—but
semantically legatos depends on mittendos. In (j) there is no object of habebam at all.
The counterpart to these expressions consisting of a gerundive with habeo is the use
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
of a gerundive with sum ‘to be’ and a dative agent, as in (k)–(l). This use of habeo
becomes more common after the Classical period, the first instance being found in
Varro (see the Supplement).
(k) Sapienti legati trans maria non sunt mittendi.
(l) Mihi pugnandum non erat imperatori sed patri.
This is one of the contexts in which sum and habeo, as the monovalent and bivalent
‘central’ stative verbs, respectively, behave in parallel ways (see § 7.37 on the devel-
opment of the active perfect paradigm of the type habeo cantatum).
Supplement:
Habeo governs an object: (sapiens) non habet mittendos trans maria legatos nec
metanda in ripis hostilibus castra, non opportunis castellis disponenda praesidia.
Non opus est legione nec equestribus turmis. (Sen. Ben. 7.3.2) (NB: parallelism with
non opus est); . . . summaque cura ne legendos (sc. libros) haberes operam dedi. (Plin.
Nat. praef. 33); . . . nec exspectandum habent oratorem, dum illi libeat de ipso
negotio dicere, sed saepe ultro admonent . . . (Tac. Dial. 19.5); . . . neque aliud excu-
sandum habeo quam quod vos Gallici foederis oblitos redisse in memoriam Romani
sacramenti temere credidi. (Tac. Hist. 4.77.3); . . . nihilque discendum haberes tem-
pore docendi. (Plin. Pan. 15.2); . . . ut consolandos eos magis imperator quam pu-
niendos habuerit. (Suet. Jul. 68.3); . . . sed intra tempora praestituta excusationem
allegandam habet et, si fuerit repulsa, tunc demum appellare debebit. (Ulp. dig.
49.4.1.1) (NB: parallelism with debebit); . . . quidquid amoris in nuptias vestras in-
pendendum habebatis, illi rependite. (August. Virg. 55.55);116
Habeo does not govern an object: Ceterum ne [et] tecum quoque habeam rixandum,
toto die me ab hoc cibo abstinebo. (Petr. 90.6); Cum patre nimirum mendicandum
habet. (Quint. Decl. 368.10); Idem, tenentibus angustias Thebanis per quas trans-
eundum habebat, flexit iter, quasi Thebas contenderet. (Fron. Str. 1.4.3); . . . cum in
aliquam invidiam aut crimen vocati sua voce respondendum haberent . . . (Tac. Dial.
36.7); . . . posse ipsam Liviam statuere, nubendum post Drusum an in penatibus
isdem tolerandum haberet . . . (Tac. Ann. 4.40.2); Finge pauperem qui, si reddere id
cogatur, laribus sepulchris avitis carendum habeat. (Cels. dig. 6.1.38);117 . . . cum
delitescendum (haec) habet . . . (Tert. Virg. vel. 17.4);
Habeo governs a communication verb: Post tantos impudicitiae quaestus si tacere
possum, confitendum habeo hac me causa afuisse ut in accessionem patrimoni
peregrinando cum uxore certarem. (Sen. Con. 2.7.1); De omnibus meis habeo
dicendum: ‘habui’. (Sen. Con. 9.5.1); In universum tamen quasi testificandum
atque saepius praedicandum habeo . . . (Col. 1.4); Nam multum interest, utrumne
de furto aut formula et interdicto dicendum habeas . . . (Tac. Dial. 37.4);
116
Material in TLL s.v. habeo 2422.80 ff. A historical survey is Thielmann (1885b). See also Pinkster
(1985b, 1987a).
117
The presence of an ablative object with carendum might be advanced as an argument for
considering carendum a gerund instead of a gerundive. This seems to be the only instance of habeo + a
putative gerund governing an argument. However, the deontic value speaks against this.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
There are a number of questions concerning the gerund and gerundive that deserve
further discussion: their origin, their chronological development, the origin of the
deontic value of the gerundive, and whether the gerund is found as subject (in the
nominative) and object (in the accusative). Some of these issues are discussed in this
section.118
The first question to be answered is whether there is a nominative gerund. There
are two ways of analysing the utterance cantandum est mihi (‘I have to sing’).
In analysis (i), cantandum is the subject of the existential verb sum ‘to be’ and is taken
as some sort of abstract noun, comparable to the noun cantatio ‘singing’ or ‘song’.
The dative case form mihi is to be explained as it is in liber est mihi ‘a book is to me’: ‘I
have a book’, in which liber is likewise the subject of est. We have, in short, some form
of ‘possessive dative’ construction with an abstract noun as the subject (more about
this construction below).119
In analysis (ii), cantandum est is a complex verb form with a passive deontic
meaning, with the third person singular form of the auxiliary verb sum ‘to be’,
which is also used in other complex verb forms (for example: carmen cantatum est
‘the song has been sung’). Mihi is the agent of the act of singing, in the dative case, as
in other passive expressions (for example: carmen mihi cantatum est ‘the song has
been sung by me’—see § 5.8 for the description of the so-called dativus auctoris).
The main argument for adopting analysis (i), and hence assuming the existence of a
nominative form of the gerund, is based on the existence of instances like (a)–(c).
118
The literature is abundant. Apart from the literature referred to by Sz.: 368–71, the matter is
discussed from a (pre)historical perspective by Drexler (1962), Strunk (1962, 1998), Hahn (1965), Blümel
(1979), Risch (1984), Hettrich (1993), Meiser (1993), Joffre (1995: 357–9), Stempel (1997), Kircher-
Durand (2008), and from a synchronic perspective by Haspelmath (1987), Heberlein (1989), and Vester
(1991). For statistical data, see Maltby (2005).
119
Advocates of analysis (i) are among others Aalto (1949), Hahn (1943), Flobert (1975: 466), and
Bolkestein (1980). Aalto’s statistical data (1949: 172–4) are affected by his taking this position.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In these examples, the words vigilias, canes, and eos depend on the -ndum expression
in the way arguments may depend on genitive, dative, and ablative gerund forms
(see § 5.38). Vigilias, canes, and eos are in the accusative case required by the verbs
from which the -ndum forms are derived. The structure of these instances might be
described as:
{verbal noun in -ndum + adnominal argument} + est
This use of the accusative case form has a parallel in the occurrence of accusative
forms with undoubtedly proper verbal nouns, as in (d) (further examples in § 11.71):
(d) Quid tibi tactio hunc fuit?
(‘What did you touch him for?’ Pl. Cas. 408)
Moreover, the distinction between personal and impersonal passive is well established in
the Latin verbal system in general. This is a sufficient argument for adopting analysis (ii).
The second problem is the origin of the deontic semantic value of the gerundive in
some of its uses—actually these uses yield most of the instances of the gerundive in extant
texts—and the absence of that value in the gerund. In the previous paragraph, arguments
have been given for regarding cantandum in mihi cantandum est as the neuter singular
form of cantandus, as it is in carmen mihi cantandum est: the only difference is that in the
first sentence cantandum is used impersonally, in the second personally. What is
problematic about this is that in gerundival clauses the gerundive lacks the deontic
value, as for example in ars carminis cantandi ‘the art of singing a song’.
It has been noted that proper verbal nouns are used in expressions with the verb sum
‘to be’ and the dative case form of a noun phrase, as in (f), which may often be
interpreted and paraphrased as an obligation or a right to do something (or, if negated,
as a prohibition from doing something). The same paraphrase is possible in (g) and (h).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Such an interpretation does not follow from an inherent feature of the lexical meaning
of the nouns involved but rather is contextually bound. Mihi cantandum est with
cantandum taken as a nominative case form of the gerund (see above) could be seen
as similar in structure to (f), and the deontic value could be regarded as a contextual
inference as well: mihi cantandum est = ‘there is singing for me’ > ‘I have to sing’. At a
later stage, the expression was reinterpreted as an impersonal gerundive expression
and the deontic value spread to other uses of this new gerundive.121 Similarly, the
deontic meaning of habeo + the accusative of the gerundive can be viewed as the
product of a diachronic development of what was at first only a contextual inference.
All this may very well be true, but in the texts preserved the deontic and non-
deontic values of the gerundive coexist from the beginning and this explanation in
diachronic terms does not contribute to a better understanding of that synchronic
relationship.
Further ‘parallels’ advanced in the literature are the following: (i) verbal abstract
nouns can be interpreted with a modal meaning when depending on the verb habeo
‘to have’, as in (i). (ii) the infinitive can also sometimes be interpreted with a modal
meaning, as in (j).
The third question to be asked is whether the gerund occurs as an object, or in other
words whether there exists a prepositionless accusative form of the gerund. The main
proponent of this idea is Aalto (1949: 82–6; further examples can be found there).
The first type of expression in which an -ndum form is taken as an accusative of the
gerund with the function object (and accordingly with an ‘active’ meaning) and not as
120
Interestingly, Donatus makes the following comment ad Ter. An. 399–400 (Puerum autem ne
resciscat mihi esse ex illa cautio’st. / Nam pollicitus sum suscepturum.): Cautione opus est, hoc est,
cavendum est.
121
A full discussion in Aalto (1949: 126–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
an accusative and infinitive clause with a passive gerundive, is with the verb censeo ‘to
give as one’s opinion’, as in (k). The second type is with the verbs curo ‘to have
something done’ and loco ‘to contract for having a thing done’, as in (l) and (m).122
The third type is with the verb habeo; this type has already been dealt with above
(§ 5.41) and hence will not be discussed further in this context. The fourth type is
constituted by a number of isolated instances in which various verbs of wanting
govern or suggest an imbedded deontic expression, as in (n). The last type is best
regarded as some form of contamination.123
(k) Ubi is homo est? # Iam devorandum censes, si conspexeris?
(‘Where is he? # You think he ought to be swallowed immediately if you set eyes on
him?’ Pl. As. 338)
(l) HAEC·OMNIA·DE·SUO / ET·VIREI· [FECIT] / FACIUNDUM·CURAVIT.
(‘The performance of all of these works she superintended out of her own and her
husband’s estate.’ CIL I2.981.7–9 (Rome))
(m) . . . <V>AL<ERIU>S L(UCI) F(ILIUS) FLACCUS / AID(ILES) D<E> STIPE AESCULAPI /
FACIUNDUM LOCAVERE . . .
(‘ . . . Valerius Flaccus son of Lucius as aediles contracted for the making of this work
out of Aesculapius’ gift-money . . . ’ CIL I2.800.2–3 (Rome))
(n) . . . cenam adferri quam optimam imperavit. Item optimis insternendum
vestimentis. Pecuniam et argentum in praesentia familiae donavit.
(‘ . . . and ordered a banquet to be served on the most lavish possible scale and the
finest tapestries likewise to be spread out; and then and there he presented his slaves
with money and silver.’ B. Hisp. 33.3)
Supplement:
Sic faciundum censeo: / Da isti cistellam et intro abi cum istac simul. (Pl. Cist.
769–70); Immo hercle etiam plus. / Nam nisi dat, domino dicundum censeo.
(Pl. Rud. 961–2); ITA·EXDEICENDUM·CENSUERE (CIL I2.581.3 (SCBac., Tiriolo, 186 BC));
Eundum in Hispaniam censui; quod si fecisset, civile bellum nullum omnino fuisset.
(Cic. Fam. 6.6.5);
2
PONTEM PEILA<S> / FACIUNDUM / COIRAVE<RE> (CIL I .1759.4–6 (Castel di Sangro,
Early)); MURUM·TURRES· POSTEICUUS / FOSSAM·FACIUNDUM·COER(AVERUNT) (CIL I2.780.3–4
(Afr. Procons., 47 BC));
As for (k), as the Supplement indicates, both the existence of personal gerundive
expressions and the existence of impersonal expressions with an overt dative agent are
strong evidence for considering devorandum in ex. (k) a gerundive and not a gerund.
As for (l), although there are other remarkable instances of the combination faciun-
dum curavit with plural objects, this seems to point to idiomatization of that expres-
sion rather than to a productive rule of Latin, all the more so since instances of these
122
There are a few related expressions with do and facio.
123
See Löfstedt (1933: II.161–2).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
verbs governing a gerundive construction are common enough (see the Supplement).
For Late bare accusative forms of the gerund instead of a prolative infinitive with
verbs like incipio ‘to begin’, see § 15.137.
Supplement:
Censeo ecastor veniam hanc dandam . . . (Pl. Cas. 1004); . . . Caesar . . . maturandum
sibi existimavit . . . (Caes. Gal. 1.37.4); Cum omnes censerent primo quoque tempore
consulibus eundum ad bellum . . . (Liv. 27.38.6);
Hoc Verrem demoliendum et asportandum nomenque omnino P. Scipionis
delendum tollendumque curasse. (Cic. Ver. 4.80); VIAS· / FACIENDAS·CURAVIT· (CIL
III.470.4–5 (Akhisar)); STATUENDOS / (TERMINOS?) LOCAVERUNT (CIL I2.400.5–6 (Portici,
early III BC)); LACONICUM·ET· DESTRICTARIUM / FACIUND(UM)·ET·PORTICUS·ET·PALAESTR<AM>
/ REFICIUNDA·LOCARUNT· (CIL I2.1635.2–4 (Pompeii, Time of Sulla));
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
CHAPTER 6
In § 2.3 four types of sentences are distinguished for this Syntax: declarative,
interrogative, imperative, and exclamatory sentences. Examples of declarative and
interrogative sentences are given in (a). Ex. (b) is an imperative sentence. Ex. (c)
has two exclamatory ones. This chapter mainly deals with finite sentences (and
not, for example, with declarative and interrogative accusative and infinitive
clauses).
(a) Quid multa? Ego amo. # An amas?
(‘To cut a long story short, I’m in love. # You’re in love?’ Pl. Bac. 1162)
(b) Ama, id quod decet, rem tuam: istum exinani.
(‘Love, as you ought to, your money; empty him.’ Pl. Truc. 712)
(c) Ut exspectatus / peregre advenisti! Quam, opsecro, cupiebat te era videre!
(‘How we longed for you to come back from abroad! Please, how keen my mistress
was to see you!’ Pl. Truc. 184–5)
These sentences fulfil different communicative functions and there are also formal
differences. In (a), the first speaker asserts that he is in love. This is the basic
function of DECLARATIVE SENTENCES. The second speaker cannot believe this and
asks for confirmation that he has understood the first speaker correctly. The basic
communicative function of INTERROGATIVE SENTENCES is a request for information.
The two sentences in (b) contain some form of instruction to do something.
That is the basic communicative function of IMPERATIVE SENTENCES. The basic
communicative function of EXCLAMATORY SENTENCES is to express joy, surprise, as
in (c), or incredulity.
Of these four sentence types the declarative sentence in (a) does not contain any
special formal marking, unlike the three others. The interrogative sentence contains
the question particle an ‘can it really be that . . . ?’ The imperative sentence (b)
contains a special imperative verb form ama, which cannot be used in other sentence
types. Ex. (c) has the exclamatory adverbs ut ‘how’ and quam ‘how much!’ Although
we have little reliable information about intonation, we must assume that the sentence
types also differed in this respect as well (see § 2.3). This explains why quam also
occurs in interrogative sentences meaning ‘how much?’ Further distinguishing prop-
erties of the sentence types are given in the separate sections.
As was observed above, the four sentence types have their own particular commu-
nicative function, or, in technical terms, their own ILLOCUTIONARY FORCE. However, in
specific situational and/or contextual conditions some of them can be interpreted in a
way that resembles that of another sentence type. This can be illustrated by the
following phenomenon. Interrogative and imperative sentences require action from
an addressee (to provide information or to perform an action). They can therefore be
combined with expressions like amabo ‘please’ to make the request more polite, as in
(d) and (e). Amabo is rarely found with a declarative sentence, but there are instances
like (f). The presence of amabo makes it clear that the sentence is not a plain assertion,
but is intended as an order or a request to Libanus to stop what he is doing. Amabo
functions as an ILLOCUTION CONVERTER.1 The terminology used to describe this phe-
nomenon is to say that sentence (f) has an INDIRECT ILLOCUTIONARY FORCE, in this case a
DIRECTIVE illocutionary force. A distinction is thus created between a BASIC and an
INDIRECT illocutionary force. In fact, interrogative sentences may be intended and
interpreted as an order instead of as a request for information too. This is shown in
(g). In this case the use of the question adverb quin ‘why not?’ (and the context, as well
as the fact that the question is addressed to a second person) suggests a directive
interpretation. Note the use of the same question word quin in the imperative
sentence (h).
(d) Quo amabo ibimus?
(‘Where will we go, please?’ Pl. Rud. 249)
(e) Da mi operam amabo.
(‘Give me your attention, please.’ Pl. Aul. 142–2a)
(f) Amabo, Libane, iam sat est.
(‘Please, Libanus, it’s enough now.’ Pl. As. 707)
(g) Iuppiter supreme, serva me. # Quid clamas? Quin taces?
(‘Jupiter above save me! # What are you shouting for? Why won’t you be quiet?’
Pl. Men. 1114–15)
(h) Quin tu tace modo.
(‘Just be quiet.’ Pl. Men. 416 )
As the examples show, the speaker/writer has, to a certain extent, the possibility of
choosing between sentence types to achieve more or less the same illocutionary goal.
The actual choice depends upon the degree of politeness or urgency he thinks is
required and appropriate in the specific communicative situation. The interpretation
of a sentence in a way that differs from its basic meaning depends partly upon the
presence of certain linguistic features in the sentence itself (for example, a specific
particle or a specific grammatical mood), partly upon the context (for example, a
1
Following Dik (1997: II.240ff.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
preceding remark by the addressee) and the extralinguistic situation (for example, the
relationship between the speaker and the addressee).2
The speaker has furthermore the possibility of mitigating or reinforcing the
illocutionary force of a sentence. Details are given in the following sections.
The organization of this chapter is according to sentence types: that is, the various
illocutionary forces with which the sentences are used are covered within the indi-
vidual sections. Since different sentence types may have more or less the same
illocutionary force, a certain amount of repetition is unavoidable.
There is discussion about how many illocutionary forces we should distinguish. An
interesting early attempt to describe sentences in terms of their communicative
functions is Brugmann (1925), who distinguishes eight ‘types of sentences according
to their basic psychological function’. In this Syntax only a few illocutionary forces
are distinguished.
(‘I solemnly swear that I will be hostile to those who I learn are hostile to Gaius
Caesar Germanicus; if anyone endangers or shall endanger him or his safety, I will
2
For the means to identify the illocutionary force of utterances, including the use of particles, see
Bolkestein (1977).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
not cease to pursue him with arms and deadly warfare on land and sea until he is
punished; I will not hold myself or my children dearer than his safety; those who shall
have a hostile intent toward him, I will consider my own enemies; if I knowingly
break or shall have broken this oath, may Jupiter Optimus Maximus, the deified
Augustus, and all the other immortal gods deprive me and my children of our
homeland, our health, and our fortunes.’ CIL II.172.5–17 (Alvéga, AD 37))
(d) Venitne homo ad te? # Non.
(‘Has he come to you? # No.’ Pl. Ps. 1067)
There are various ways for the speaker or writer to modify the force of the assertion.
There are lexical means (the use of attitudinal disjuncts, or particles, for instance) and
grammatical means (the choice between the indicative and subjunctive moods, for
instance). The speaker’s or writer’s motivation for modulating the illocutionary force
may vary: it may be for reasons of politeness (or impoliteness) or certainty (or
uncertainty).3
Examples of lexical means to strengthen the assertive illocutionary force are the
affirmative particles ne ‘indeed’ (especially in Plautus and Terence; from Cicero
onwards, only in prose, almost invariably in sentence or main clause-initial position)4
and equidem ‘indeed’, as in (a) and (b), and attitudinal adverbs like certe ‘certainly’
and profecto ‘undoubtedly’, as in (c). Ne and equidem are limited to declarative
sentences (there are a few instances of equidem in relative clauses and in clauses of
qualification—for the latter, see § 16.83).5 Attitudinal adverbs occur mainly in
declarative main sentences (they are rare in subordinate clauses other than relative
and argument clauses). (See § 10.100 on their occurrence with sentence questions.)
To weaken the illocutionary force, adverbs like fortasse ‘perhaps’ can be used. Another
lexical means is the use of parenthetical verb forms like censeo ‘I imagine’, credo ‘I
suppose’, existimo ‘I suppose’, opinor ‘I believe’, as in (d), and puto ‘I believe’, or
subordinate clauses like ut arbitror ‘(as) I think’, ut existimo ‘as I suppose’, ut opinor
‘as I believe’, as in (e), ut puto ‘as I believe’, and ut ego sentio ‘as I think’ (for more
instances, see § 16.35) and quod sciam ‘as far as I know’ (for more instances, see
§ 16.83).6
3
For politeness and formality in Cicero, see Hall (2005, 2009). For Lucretius’ usage, see Reinhardt
(2010: 223–5). For a survey of recent studies, see Unceta (2014). For expressions of gratefulness in Latin
comedy, see Unceta (2010).
4
For details, see Hand (1845: IV.22–7).
5
See TLL s.v. equidem 721.64ff. and Ludewig (1891: 32–7).
6
For the parenthetical verb forms and the ut clauses, and the differences between them, see Bolkestein
(1998).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
As for grammatical means, the subjunctive may be used as a milder way of putting
things (see § 7.41). Examples are (f )–(i). Note in (h) the combination of the attitudinal
disjunct sine ulla dubitatione and the polite perfect subjunctive form confirmaverim.
(f) Nonnulla forsitan conformare et leviter emendare possim.
(‘I may, perhaps, reshape parts of it and improve them slightly.’ Cic. Mur. 60)
(g) Sed neque verbis aptiorem cito alium dixerim neque sententiis crebriorem.
(‘Yet I could not easily name his superior in appropriateness of diction, nor in
abundance and compactness of thought.’ Cic. Brut. 264)
(h) Hoc vero sine ulla dubitatione confirmaverim . . .
(‘This, however, I would affirm without any hesitation . . . ’ Cic. Brut. 25)
(i) Cuperem equidem utrumque, si posset, sed est difficile confundere.
(‘My wish would be for both courses if it could be managed, but it is a difficult matter
to combine the two.’ Cic. Tusc. 1.23)
7 8
See Schrickx (2011: 147–51). On nempe, see Schrickx (2011: 93–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Et scilicet iam me hoc voles patrem exorare ut celet / senem vostrum?
(‘And I suppose you’ll also want me to persuade my father to keep this a secret from
your old man.’ Ter. Hau. 705–6)
(b) Illud etiam in tali consilio animadvertendum vobis censeam, patres con-
scripti, si iam duriores esse velitis, quod nullo nostro merito faciatis, cui nos
hosti relicturi sitis. Pyrrho videlicet, qui [vos] hospitum numero captivos
habuit? An barbaro ac Poeno, qui utrum avarior an crudelior sit vix existi-
mari potest?
(‘One other point I would suggest, as meriting consideration, when you deliberate
about the matter, Conscript Fathers: If haply you should incline to deal harshly
by us—which we do not in the least deserve—to what enemy would you be leaving
us? To a Pyrrhus, pray, who treated his prisoners like guests? Or to a barbarian
and Phoenician, of whom it can hardly be determined whether his avarice or cruelty
be greater?’ Liv. 22.59.14)
(c) Hoc facito, miles domum ubi advenerit, / memineris ne Philocomasium
nomines. # Quem nominem? / # Diceam. # Nempe eandem quae dudum
constituta est. # Pax! Abi.
(‘Make sure that when the soldier comes home you remember not to call her
Philocomasium. # What should I call her? # Justine. # The same we agreed on a
while ago. # Enough. Off you go.’ Pl. Mil. 806–9)
Declarative sentences with a second person singular future indicative tense form
are used with an indirect directive illocutionary force from Early Latin onwards.10
For instances and discussion, see § 7.23. A variant where the second person is not
encoded as subject in the verb form, but is otherwise expressed as an argument, is
exemplified by (a).
(a) Haec igitur tibi erunt curae . . .
(‘This then will be your concern . . . ’ Cic. Fam. 3.9.4)
9
The example is taken from Quirk et al. (1985: 814 (‘declarative questions’)).
10
For the general distinction between declarative sentence type and directive illocution, see Risselada
(1993: ch. 6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The use of ‘prescriptive’ future tense expressions is quite common, especially in legal
and didactic texts. Among legal authors, Ulpian is well known for his predilection for
the future tense. Examples are (b) and (c).11
(b) Sed et si instigatu alterius fera damnum dederit, cessabit haec actio.
(‘Nor will this action lie if the animal did harm because someone provoked it.’ Ulp.
dig. 9.1.1.6)
(c) In quod factus sit locupletior aequissimum erit in eum dari repetitionem.
(‘It will be most equitable for an action to be given against him for the amount by
which he profited.’ Ulp. dig. 2.15.8.22)
Whereas the use of the future in the above examples can be explained on the basis of
the semantic value of that tense, a similar explanation is not available for instances
of the indicative in declarative sentences that seem to be intended as directive. At
least three types must be distinguished. A first type is where the declarative sentence
contains an expression that is typical of directive utterances, such as opsecro ‘I
beseech you’ in (d), intended as a request to stop the flogging. Similarly, amabo
in (e). Expressions that may be used as illocution converters like obsecro are dealt
with in § 6.29.
(d) Oiei, satis sum verberatus, opsecro.
(‘Ow! I’ve been beaten enough, I entreat you!’ Pl. Mil. 1406)
(e) Amabo, Libane, iam sat est.
(‘Please, Libanus, it’s enough now.’ Pl. As. 707)
11
For examples and comparison with other jurists, see Honoré (1982: 59–63): ‘Ulpian is a partisan of
the future tense’. For the use of the first and second person in scientific and technical writing, see Hine
(2009).
12
Examples in Soffriti (1963) and in L. Löfstedt (1966: chs 4 and 5). Ex. (f ) is transmitted by a few
Latin grammarians precisely for its ‘imperatival’ function.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
A third type of declarative sentences that are interpreted as directive is typical of texts
describing expected behaviour of officials and the like, as in (i) probably from an old
text describing the rules of conduct of a flamen Dialis,15 and in manuals, either
literary or non-literary, addressed to a (supposed) audience of (soon to be) practising
professionals.16 With the evolution of this didactic text type, descriptive statements
concerning generally accepted practice became popular (instead of Cato’s predilec-
tion of prescriptive -to future imperative forms—for which see § 7.65). The very
fact that readers were aware of the text type, the presence of headings, and the
co-occurrence with explicit directives was sufficient to arrive at the correct interpret-
ation. An early example of the use of a third person present indicative active form
with a directive force is (j). Second person (singular) forms are very common as well,
as in (k).
(i) Super flaminem Dialem in convivio, nisi rex sacrificulus, haut quisquam
alius accumbit.
(‘No other has a place at table above the flamen Dialis, except the rex sacrificulus.’
Gel. 10.15.21)
(j) In nutricatu, cum parere coeperunt, inigunt in stabula . . .
(‘As to feeding: when they begin to bear they are driven into the pens . . . ’ Var.
R. 2.2.15)
(k) Nucleos infundes et siccas.
(‘Soak the nuts and dry them.’ Apic. 4.2.13)
Supplement:
Nodum in apice neque in cinctu neque in alia parte ullum habet. . . . Capillum Dialis
nisi qui liber homo est non detondet.17 . . . Propagines e vitibus altius praetentas non
succedit. . . . Tunica intima nisi in locis tectis non exuit se, ne sub caelo tamquam sub
oculis Iovis nudus sit. (apud Gell. 10.15.9; 11;13;20); Siqua in tonsura plagam accepit,
eum locum oblinunt pice liquida. (Var. R. 2.11.7);
13
See Adams (2007: 446). The Loeb edition has experires.
14
See Adams (1976: 68). mittes cj. Rolfe.
15
For discussion of this phenomenon, see Daube (1956: 68–72).
16
For the history of this usage in technical prose, see Adams (1995b: 204–8, 460–8). For didactic verse
and prose, see also Gibson (1997). For the use of declarative sentences in recipes, see Flobert (2001). For
medical texts, see Gaide (2002). For Apicius, see Grocock and Grainger (2006: 98–102). For the use of the
gerundive in didactic texts, see Arjoca-Ieremia (2008).
17
This is Daube’s (1956: 71) reading instead of detonset found in the majority of manuscripts.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Declarative sentences with a third person present indicative passive form with a
directive force became familiar in the course of the first century AD and continued
to be in use until Late Latin. An early example is (l). The preceding context with the
gerundive legenda and the obligation verb debet ‘must’ in combination with diligentis
makes it possible to interpret the following descriptive statements as instructions. In
such sentences, evaluating adverbs like recte ‘rightly’ and utiliter ‘profitably’ are quite
common, as in (m). The impersonal passive is likewise used in prescriptions in legal
and didactic texts and for general commands and instructions, as in (n). In addition,
there is the impersonal and personal use of the gerundive, as in (o) (for the deontic
use of the gerundive, see § 5.41).
(l) (Nam et praecox et miscella, quam vocant nigram, multo ante coquitur, quo
prior legenda (sc. est), et quae pars arbusti ac vineae magis aprica, prius debet
descendere de vite.) In vindemia diligentis uva non solum legitur <sed etiam
eligitur: legitur> ad bibendum, eligitur ad edendum.
(‘For the early grapes, and the hybrids, the so-called black, ripen much earlier and so
must be gathered sooner; and the part of the plantation and the vineyard which is
sunnier should have its vines stripped first. At the vintage the careful farmer not only
gathers but selects his grapes; he gathers for drinking and selects for eating.’ Var. R.
1.54.1–2)
(m) Ieiuno recte catapotia dantur . . .
(‘Pills . . . are given on an empty stomach.’ Cels. 3.21.6)
(n) PRIVATUM. / PRECARIO / ADEITUR.
(‘Private. Entry by request only.’ CIL I2.1606 (S. Maria Capua Vetere))
(o) Polenta, non pulenta dicendum est.
(‘One should say polenta, not pulenta.’ Caper Orth. 8)
Supplement:
Castrantur verres commodissime anniculi . . . (Var. R. 2.4.21); Quae cum ita facta
sunt, sextarius aquae calidae in eandem mensuram lentis miscetur et faucibus
infunditur, similisque medicina triduo adhibetur, ac viridibus herbis cacuminibus-
que arborum recreatur aegrotum pecus. (Col. 6.31); Ex eo paulum aut devoratur,
aut aqua diluitur, et potui datur. (Cels. 5.25); Utiliter etiam scilla cocta delingitur.
(Cels. 3.21);
Columella is one of the authors who use the present passive forms frequently. In the
veterinarian chapters of book VI he has six -to future imperative forms, forty
gerundival expressions, and 140 present indicative passive forms.18
In all declarative sentences with an indirect directive illocutionary force the negation
word is non (and not ne).19
18
Source: Adams (1995b: 464–5). See also Önnerfors (1989); Adams (1991: 88–90); Pinkster (1992b).
19
The only exception before the Merovingian period is HOC SIMULACRUM NE REVELLIS (Inscr. Not. d. Scavi
1909, 456). See L. Löfstedt (1966: 183) and Pinkster (1986a). See also Ch. 7, n. 80.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The following sections are about questions that are meant as questions, starting with those
with a request for information illocutionary force, which includes a discussion of the
various types of interrogative sentences (§§ 6.6–6.20), followed by a discussion of forms of
modulation of the request for information illocutionary force of questions (§ 6.21), and the
indirect illocutionary forces with which questions are used (§§ 6.22–6.24).
There are several types of interrogative sentences. A first distinction is that between
SIMPLE QUESTIONS and MULTIPLE QUESTIONS, also called DISJUNCTIVE questions. In the latter
20
See Dik (1997: II.258–9).
21
For the discourse functions of questions, see Risselada (2005). For the use of questions to avoid a
conversation coming to an end, see Roesch (2005). For the use of quid ais? to draw the attention of the
addressee in comedy (especially by men), see Barrios-Lech (2014).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
type, the addressee is offered two or more alternatives between which he has to
choose. They are dealt with in § 6.20.
When uttering a sentence question, a speaker may include one or more elements that
indicate what he knows or assumes already and in this way the speaker may help the
addressee to formulate his answer. The most conspicuous signals of the status of the
information of the speaker are the so-called question particles. Traditionally, the clitic
-ne is regarded as the signal that the speaker has no expectations as to the answer he is
asking for, nonne as the signal that he expects an affirmative answer, num as the signal
that he expects a negative answer, and finally an as the signal that he cannot believe or
accept the information he has. In practice, this division is not as clear as it seems.
There are other indications as well.22
22
This paragraph and the note follow Morris (1890: 178–81). For the particles an, -ne, and num, see
also Müller (1997: 77–9, 80–8).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Questions with an affirmative bias have non (sentence-initial with the verb at the
end), nonne, or -ne. It is rare for such questions not to contain a particle. Marked by
a negative bias are questions that contain adverbs like vero ‘really’ or satis ‘enough’
(plus the particle -ne), ask for someone’s belief (credi’n), or contain particles like iam
‘yet’, and more specifically etiam ‘really’, ‘actually’ and num.
Questions differ with respect to the function they fulfil in the discourse. They may
confront an addressee with something new and unexpected (‘Sir, is this the Bodleian
Library?’), continue or elaborate on preceding information, or change the subject
matter. Questions without a particle or with -ne seem to be more closely related to the
preceding context, whereas those with num or numquid serve to change the subject or
to introduce a new subject matter. Questions with an are closely related to the
preceding context due to their verifying or challenging function.23
Num and (less frequently) -ne are combined with the connector igitur ‘then’, which has
an initiating function, and much less with the interactional particle ergo ‘therefore’,
which is typically reactive.24 See the data concerning Cicero in Table 6.1. Combinations
such as videtisne igitur occur fifty-eight times, such as possumusne ergo five times.
Sentence questions are often preceded by a question Quid? ‘what?’ often in combin-
ation with an adverb or a particle. Examples are (a)–(g). Questions with quid in
combination with the interactional particle enim, which appeals to common under-
standing, or with igitur ‘what then’ and ergo ‘what in that case’, expect a positive
answer. Examples are (c)–(d). An example of quid with the epistemic adverb vero
‘indeed’ ‘conveying a sense of protest, disbelief, etc.’26 is (e). Combinations with
adverbs indicating the succession of events like deinde, postea, tandem, and tum are
also quite common. Examples are (f ) and (g).
(a) Quid? Duasne uxores habet?
(‘What? Does he have two wives?’ Ter. Ph. 754)
23
For a first exploration along these lines, see Finke (1987) and Bolkestein (1988).
24 25
See Kroon (1989b: 236–40). Source: Spevak (2006: 161).
26
OLD s.v. vero § 5e.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The information requested when using a sentence question may either regard the
sentence as a whole or a particular constituent of that sentence. In more technical
terms, sentence questions may have wider or narrower SCOPE, in this case SENTENCE
SCOPE or CONSTITUENT SCOPE. Constituent scope is more common than sentence
scope.27 Examples of sentence scope are (h) and (i). An example of constituent
scope is (j). Note in (i) the variation of nonne and non. (NB: ‘constituent scope’
must not be confused with ‘constituent question’.)
(h) Ecastor equidem te certo heri advenientem ilico / et salutavi et valuissesne
usque exquisivi simul, / mi vir, et manum prehendi et osculum tetuli tibi. / #
Tune heri hunc salutavisti? # Et te quoque etiam, Sosia.
(‘I certainly did greet you here on your arrival yesterday and asked you at the same
time if you’d been well throughout, my husband, and I took your hand and gave you
a kiss. # You greeted him yesterday, did you? # And you too, Sosia.’ Pl. Am. 714–17)
(i) Nonne hac noctu nostra navis <huc> ex portu Persico / venit, quae me
advexit? Nonne me huc erus misit meus? / Nonne ego nunc sto ante aedis
nostras? Non mi est lanterna in manu? / Non loquor? Non vigilo? Nonne hic
homo modo me pugnis contudit?
27
Risselada (p.c.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(‘Didn’t our ship come here from Port Persicus this night, the one that brought me
here? Didn’t my master send me here? Aren’t I standing in front of our house now?
Haven’t I a lamp in my hand? Aren’t I speaking, aren’t I awake? Didn’t this man here
beat me up with his fists just now?’ Pl. Am. 404–7)
(j) Num Ampelisca opsecro est? # Ten’, Palaestra, audio?
(‘Please, is it Ampelisca? # Is it you that I hear, Palaestra?’ Pl. Rud. 235)
Since questions are about missing information or certainty, attitudinal disjuncts are
hardly expected. The few exceptions are almost all either ironic or rhetorical. Real
exceptions in questions with an interrogative illocutionary force are scilicet in (k) and
fortasse in (l) (if they are questions). A remarkable instance with certe ‘certainly’ is
(m), explicitly described as a question by Suetonius. Certe reflects Augustus’ certainty
and is not part of the question proper.28 Also enim, an interactional particle that
appeals to shared knowledge between speaker and hearer (see Chapter 24), is rare in
sentence questions, unless they are understood as an assertion, as in (n).
(k) Huc quom advenio, nulla erat. / # Comites secuti scilicet sunt virginem? /
# Verum. Parasitu’ cum ancilla.
(‘When I got here, she was nowhere to be seen. # I suppose there were some
companions to escort the girl? # Yes. A sponger and a maid.’ Ter. Eu. 345–7)
(l) At nos ad te ibamu’, Phormio. # De eadem hac fortasse causa? # Ita hercle.
(‘But we were coming to see you, Phormio. # About the same business, perhaps?
# Yes indeed.’ Ter. Ph. 899–900)
(m) (sc. Augustus) . . . ita fertur interrogasse: ‘Certe patrem tuum non occidisti?’
(‘ . . . he is said to have asked: “You surely did not kill your father, did you?” Suet. Aug.
33.1)
(n) Num (non v.l.) enim alia in causa M. Cato fuit, alia ceteri qui se in Africa
Caesari tradiderunt?
(‘Did Marcus Cato find himself in one predicament, and were the others, who
surrendered to Caesar in Africa, in another?’ Cic. Off. 1.112)
Certe in (o) is not an attitudinal disjunct, as is proved by the fact that in the answer it
is modified by tam.
(o) Tuos pater— # Quid meus pater? # Tuam amicam— # Quid eam? # Vidit. . . .
Certen’ vidit? # Tam hercle certe, quam ego te aut tu me vides.
(‘Your father— # What about my father? # Your girlfriend— # What about
her? # He’s seen her. . . . Has he seen her for certain? # For as certain as I see
you or you me.’ Pl. Mer. 180–6; cf. 324; Ter. Hec. 843; in an indirect question
in Cic. Leg. 1.3)
28
On the use of ‘commitment markers’ in questions, see Schrickx (2011: 207–14).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
As for the frequency of the use of question particles in sentence questions, Bennett’s
Syntax of Early Latin contains useful information.29 His data and data based on (much
smaller) samples from Cicero and Seneca taken from Bodelot (1990: 24–5) are arranged
in Table 6.2. In Bodelot’s overall statistics half of the sentence questions have no particle.
Table 6.2 Frequency of particle use in sentence questions (in percentages rounded off
to integers)
no particle -ne num30 an(ne) nonne ecq . . .
Early Latin (N = 2750) 31 52 7 5 1 3
Cic. Orations (N = 189) 53 16 2 17 7 3
Cic. Off. 1 and 2 (N = 10) 20 – 20 20 40 –
Cic. Letters (N = 13) 68 8 8 8 8 –
Seneca Ep. (N = 40) 48 2 28 10 2 10
N = ‘total number of instances’
The overwhelming frequency of -ne in Early Latin is no doubt related to the fact that
most of these questions are found in comedy. Certain combinations with -ne are very
frequent, such as ai’n (< aisne) ‘you say?’ (almost an ‘introductory particle’31), audi’n (<
audisne) ‘you hear?’ poti’n (< potisne) ‘can you?’, sci’n/sciti’n (< scisne/scitisne) ‘you (sg./
pl.) know?’ vide’n/videti’n (< videsne/videtisne) ‘you (sg./pl.) see?’ vi’n (< visne) ‘you (sg.)
want?’ voltisne ‘you (pl.) want?’ in total c.250 instances. Also, sati’n (< satisne) ‘enough?’
(ninety-five instances), cense’n (< censesne) ‘do you think’ (eleven instances).32
Appendix: Invocations like ecastor, edepol, and hercle do not occur in questions, with
one exception, (p), unless hercle is taken to reinforce mihi, in the same way as hercle
reinforces opsecro in (q).33
(p) Nam mihi quidem hercle qui minus liceat deo minitarier / populo, ni
decedat mihi, quam servolo in comoediis?
(‘Why should I, a god, not be allowed to threaten people if they don’t get out
of my way just as much as some paltry slave in comedies?’ Pl. Am. 986–7)
(q) Quid opsecro hercle factum est? # Concrepuit foris.
(‘Please, what’s happened? # The door has creaked.’ Pl. Mos. 507)
The question particles left no trace in the Romance languages. Their disappearance
can be inferred from the way they are used by those authors who still have them, such
as Tertullian and Augustine, in combinations that are rare or non-existent in
Classical authors, like ergone, num enim, num ergo.
It is difficult to get a clear picture of sentence questions in Latin for a number of
reasons. Most of the material that is best suited to a study of sentence questions (texts
29 30 31
Bennett (1910: I.460–84). Num and compounds. Bennett (1910: I.461).
32
For the use of interrogative -ne in Terence, see Müller (1997: 83–5); for the use of cense’n in
dialogues, see Sánchez Manzano (2008).
33
See Vairel-Carron (1975: 117).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(r) Quid vero? Ille M. Cato nonne et eloquentia tanta fuit, quantam illa tempora
atque illa aetas in hac civitate efferre maximam potuit, et iuris civilis
omnium peritissimus?
(‘And what of the eminent Marcus Cato? Did he not combine eloquence as
grand as those times and that epoch could produce in this State, with an
unequalled knowledge of the common law?’ Cic. de Orat. 1.171)
34
For detailed statistical information, see Bodelot (1990: 12–13) and the sets of tables in that work.
35
See Bennett (1910: I.476–7); K.-St.: II.501–2; Sz.: 460.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Of the approx. 850 sentence questions without a particle in Early Latin (see Table 6.2),
approx. 200 contain a negation word, of which approx. 185 are non. Of these, roughly
one third are real questions, the majority are either exclamatory (see § 6.19) or—less
often—directive (see § 6.24).36 Two examples of real questions are (f ) and (g). In
these cases, the expected answer is positive. In (h), this expectation turns out to be
wrong, hence immo ‘on the contrary’. (See also § 6.12 on nonne.)
(f ) Carnufex, non ego te novi? Abin’ e conspectu meo?
(‘You good-for-nothing, don’t I know you? Won’t you get out of my sight?’
Pl. Am. 518)
(g) Non tu eum rus hinc modo / produxe aibas?
(‘Didn’t you say just now you’d seen him off to the farm?’ Ter. Ad. 560–1)
(h) Quid ego ex te audio? # Hanc Athenis esse natam liberam. / # Mea popularis,
opsecro, haec est? # Non tu Cyrenensis es? / # Immo Athenis natus . . .
(‘What do I hear from you? # That this girl was born free in Athens. # Is she my
compatriot, please? # Are you not from Cyrene? # No, I was born in Athens . . . ’ Pl.
Rud. 739–41)
36
For directive non questions, see Bennett (1910: I.25); for nonne, see p. 472.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
37
For some figures and on its disappearance, see Janson (1979: 104).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
38
For this use of iam, see Kroon and Risselada (1998, 2002).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Itane vero obturbat? (Ter. An. 926); Itane vero? Tu tua pericula communi periculo
defendes? (Cic. Ver. 5.77); Itane in geometriae pulvere haerebo? (Sen. Ep. 88.39);
Itane? (Sen. Oed. 936); Itane tu ita vivis in hoc corruptibili corpore, quod adgravat
animam, ut caro non concupiscat adversus spiritum tuum, et spiritus adversus
carnem? (August. Serm. 30.4);39
Constituent scope: Pronouns: Egone istuc dixi? # Tute istic, etiam astante hoc Sosia.
(Pl. Am. 747); Hem, Postume, tune es <C.> Curti filius, C. Rabiri iudicio et voluntate
filius, natura sororis filius? (Cic. Rab. Post. 45); Tune es? Agnosco toros / umerosque
et alto nobilem trunco manum. (Sen. Her. F. 624–5); Dic mihi, isne istic fuit quem
vendidisti meo patri, / qui mihi peculiaris datus est? (Pl. Capt. 987–8); Illicine est? #
Illic est (Pl. Ps. 954); Haecine ea’st? # Haec est. (Ter. Hec. 771); His mundus umeris
sedit? Haec moles mea est, / haecne illa cervix? (Sen. Her. O. 1242–3);
Nouns: Disne advorser? (Pl. Per. 26); Liberorumne causa? At procreare potest . . .
(Cic. Dom. 34); Quid ergo? Romanos in illis ulterioribus munitionibus animine causa
cotidie exerceri putatis? (Caes. Gal. 7.77.10);
Adjectives et al.: Cura te, amabo. Sicine immunda, opsecro, / ibis? (Pl. Cist. 113–14);
Pro lignean’ salute vis argenteam / remittere illi? (Pl. Ps. 47–8); Molestusne sum? (Pl.
Truc. 720); Opsecro, sanu’n es? # Sanus quom ted amo. (Pl. Cas. 232);
Adverbs: Sed dic mihi, / benene ambulatum’st? (Pl. Truc. 368–9); Hicne etiam sese
putat aliquid posse Chrysogonus? (Cic. S. Rosc. 141); Hodiene exoneramus navem,
frater? (Pl. St. 531); Isticin’ vos habitatis? (Pl. Rud. 110); Sed malene id factum
arbitrare? (Pl. Capt. 709); Sati’n hoc plane, sati’n diserte, ere, / nunc videor tibi
locutus / esse? (Pl. Am. 578–9); Sati’n audis quae illic loquitur? / # Satis. (Pl. Men.
602–3); NB: In a weakened sense: Sati’n abiit nec quod dixi flocci existumat? (Pl.
Mos. 76); Quid, mulier? # Mulier meretrix. # Veron’ serio? (Pl. Mer. 685);
NB: Interactional particles: Ergone et finis in caelis et passio et occisio et prima
confessio? (Tert. Sc. 10.25); Capiunt ergone te caelum et terra, quoniam tu imples ea?
(August. Conf. 1.3);
For the use of -ne in indirect questions, see § 15.52.
Appendix: Christian and other Late Latin authors occasionally use nē as a question
particle, both in direct and in indirect questions, apparently in the same way as the
clitic particle -ne.40 An example is (m). As an early example some scholars cite (n).
Here the ne clause follows a series of indirect questions and, moreover, has perfect
tense, which is difficult to explain if the clause is taken as an imperative argument
clause or as a purpose clause (see Woytek ad loc.). However, I follow Bennett’s41
interpretation. For the perfect, see § 7.100.
(m) Qua<re> nescio ne plus de vobis dei vestri quam de nobis querantur.
(‘For this reason I don’t know if your gods should complain more about you
than about us.’ Tert. Nat. 1.10.49—tr. Howe)
39
For this use of ita ‘vi debilitata’ see TLL s.v. 527.69ff. Also Bennett (1910: I.470).
40
See Blaise (1955: § 272bis); Sz.: 542; Ortoleva (2014: 331–4); TLL s.v. ne 313.54ff.
41
Bennett (1910: I.254).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(n) Nunc huc intro ibo, visam hesternas reliquias, / quierintne recte necne, num
afuerit febris, / opertaen’ fuerint, ne quis obreptaverit.
(‘Now I’ll go in here and see whether yesterday’s leftovers have slept well or
not, whether they’ve caught a fever, and whether they’ve been covered, so
that no one could have approached them stealthily.’ Pl. Per. 77–9)
6.12 Sentence questions with the particle nonne (and with the negator non)
Nonne ‘is it not the case that?’ is rare in Early Latin. Cicero uses it relatively often
(see Table 6.2 in § 6.8), as one may expect, since most of his texts in some way or
other are directed to an audience which he likes to involve. In narrative texts and
in didactic texts questions are rare, and so is nonne (and num). In Lucretius the
combination nonne vides is relatively frequent, because it is metrically convenient
and fits in well with Lucretius’ interaction with his audience. It had some success
with other poets. Cicero has non vides instead (only once nonne . . . vides) and the
same is true for Seneca. Certain authors do not or only rarely use nonne. This is
no surprise for an author like Pliny the Elder, who uses neither nonne nor num,
and an only once: it is not the type of text in which questions abound. In poetry,
apart from what is said above, nonne is rare, but so is num. In Seneca’s tragedies
there are two instances of nonne, in his prose twenty-one. Tertullian has ninety
instances of nonne, Augustine c.1730.
Nonne and non are used in negative questions and usually expect a positive
answer (for non see also the Appendix). In Early Latin non is the normal expres-
sion. Cicero’s use varies: in his orations 143 non, 120 nonne; in the philosophical
works thirty-one non, 101 nonne; in the letters there are only fourteen instances of
nonne. (Only instances of non have been counted that occur in sentences without a
question word or question particle).42 The difference between the two is often
described in psychological terms: non is said to be used in impatient, surprised,
or indignant questions, and illustrations usually come from comedy (where nonne
is rare). However, this may be a side effect: there are also many positive questions
with such emotions in comedy. There are instances of nonne in similar situations,
as in (b). Nonne and non are sometimes used in the same context when a series of
questions is involved, where -ne makes explicit that the scope is on the negation.
Examples are (a) and (b). A question with quid often precedes, as in (c). Nonne
often occupies a non-initial position in its sentence, usually after a topic constituent
(see Chapter 22).
(a) Nonne ad te L. Lentulus qui tum erat praetor, non Q. Sanga, non
L. Torquatus pater, non M. Lucullus venit?
(‘Did not Lucius Lentulus, who was at that time praetor, did not Quintus Sanga, did
not Lucius Torquatus the elder, did not Marcus Lucullus visit you?’ Cic. Pis. 77)
42
See Merguet Phil. s.v. non 723bff.; s.v nonne 748aff. Reden s.v non 332bff.; s.v nonne 364bff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(b) Nonne hunc in vincla duci, non ad mortem rapi, non summo supplicio
mactari imperabis?
(‘Are you not going to order him to be put in chains, taken off to execution, and
suffer the supreme penalty?’ Cic. Catil. 1.27)
(c) Quid? Aestus maritimi vel Hispanienses vel Brittannici eorumque certis
temporibus vel accessus vel recessus sine deo fieri nonne possunt?
(‘Cannot the tides on the coasts of Spain or Britain ebb and flow at fixed intervals of
time without a god’s intervention?’ Cic. N.D. 3.24)
Supplement:
Quis res publica et privata geritur, nonne is crederem? (Pl. Cur. 552); Eho, nonne haec
iusta tibi videntur postea? (Ter. Ad. 660); Te deiectum debeo intellegere, etiam si tactus
non fueris. Nonne? (Cic. Caec. 37); Quid? Exemplo tuo bona tua nonne L. Ninnius, vir
omnium fortissimus atque optimus, consecravit? (Cic. Dom. 125); Siqui rex, siqua civitas
exterarum gentium, siqua natio fecisset aliquid in civis Romanos eiusmodi, nonne publice
vindicaremus, nonne bello persequeremur? (Cic. Ver. 5.149); Sed nonne meministi licere
mihi ista probare quae sunt a te dicta? (Cic. Fin. 5.76); Quid? Canis nonne similis lupo . . . ?
(Cic. N.D. 1.97); Quid? Tu id pateris? Nonne defendis, non resistis? (Cic. Fam. 7.32.1);
Nonne vides etiam quanta vi tigna trabesque / respuat umor aquae? (Lucr. 2.196–7);
Nonne vides croceos ut Molus odores, / India mittit ebur, molles sua tura Sabaei . . . ?
(Verg. G. 1.56–7); Quicumque ille fuit puerum qui pinxit Amorem, / nonne putas miras
hunc habuisse manus? (Prop. 2.12.1); Nonne in mentem venit quantum piaculi commit-
tatur? (Liv. 5.52.13—NB: in a speech); Quid porro? Nonne nunc quoque, etiam si parum
sentitis, turbo quidam animos vestros rotat et involvit, fugientes petentesque eadem et
nunc in sublime adlevatos nunc in infima adlisos? (Sen. Dial. 7.28.1); Abduxi, custodivi,
[[nonne recipere non debuit qui periit quod non recepit?]] non reddidi. (Quint.
Decl. 289.4); Iam ‘itur in antiquam silvam’ nonne propriae cuiusdam rationis est?
(Quint. Inst. 1.4.28); Quid? Ipsorum magistratuum nonne plerosque variis libidinibus
obnoxios? (Tac. Ann. 3.34.3—NB: indirect speech in an accusative and infinitive); Nonne
philosophi de auctoribus suis nuncupantur Platonici, Epicurei, Pythagorici? (Tert. Apol.
3.6); Nonne posthabitis ceteris militem instantibus aerumnis eriperetis? (Amm. 25.5.3—
NB: in a speech); Sed nonne tibi videtur secundum istam rationem, nec post epitritum
secundum iambum cum silentio poni oportere? (August. Mus. lib. 4, col. 1136, l. 42);
Appendix: A question containing non may also serve as a verification of one’s
negative estimate of a situation or as a check of an answer previously given by the
addressee, as in (d)–(f ) (a so-called DUBITATIVE question).43
43
For this use, see Murphy (1990) and Orlandini (1991: 286). For num, see note 45.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(f ) Quid enim? Tu solus aperta non videbis, qui propter acumen occultissima
perspicis?
(‘For consider: Will you, whose intelligence pierces the darkest abstrusities,
alone be blind to the obvious?’ Lucc. Fam. 5.14.2)
44
The quotation is from Shackleton Bailey (1953: 121).
45
On this ‘dubitative’ use of num, see Orlandini (1991: 282–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Sed num fumus est haec mulier quam amplexare? (Pl. As. 619); Num non vis me
interrogare te? (Pl. Aul. 161); Num quisquam adire ad ostium dignum arbitratur? (Pl.
Mer. 132); Numquid nunc aliud me vis? # Ne me noveris. (Pl. Mil. 575); Num non vis
me obviam his ire, anime mi? (Pl. Mos. 336); Quid ais? Nunc tu num nevis me, voluptas
mea, / quo vocatus sum ire ad cenam? (Pl. Truc. 546–7); Num inmemores discipuli?
(Ter. An. 477); Eho, numnam hic relictu’s custos . . . (Ter. Eu. 286); Num is est Cluen-
tius? Certe non est. (Cic. Clu. 148); Num quae trepidatio, num qui tumultus, numquid
nisi modeste, nisi quiete, nisi ex hominis gravissimi et sanctissimi disciplina? (Cic. Deiot.
20); Nominis inscriptio tibi num aliud videtur esse ac meorum bonorum direptio? (Cic.
Dom. 51); Quid est? Num conturbo te? (Cic. Phil. 2.32); Quid? Municipia, colonias,
praefecturas num aliter iudicare censetis? (Cic. Phil. 4.7); Num ergo etiam armorum
interpretes quaerimus? (Cic. Phil. 5.9); Ad te ipsum, P. Servili, num misit ullas conlega
litteras de illa calamitosissima pugna Pharsalia? (Cic. Phil. 14.23); Numne si Coriolanus
habuit amicos, ferre contra patriam arma illi cum Coriolano debuerunt? (Cic. Amic. 36);
Quid vos agitis? Num sermonem vestrum aliquem diremit noster interventus? (Cic.
Rep. 1.17); Sed haec eadem num censes apud eos ipsos valere nisi admodum paucos a
quibus inventa disputata conscripta sunt? (Cic. Tusc. 2.11); Numquid ibi horribile
apparet, num triste videtur / quicquam, non omni somno securius exstat? (Lucr.
3.976–7); Num linquere castra / hortati sumus aut vitam committere ventis, / num
puero summam belli, num credere muros, / Tyrrhenamque fidem aut gentis agitare
quietas? (Verg. A. 10.68–72); Dic mihi, num teneros urit lorica lacertos? / Num gravis
imbellis atterit hasta manus? (Prop. 4.3.23–4); Periculi quod differendo bello adimus
num oblivisci nos haec tam crebra Etruriae concilia de mittendis Veios auxiliis patiun-
tur? (Liv. 5.5.8—NB: in a speech); Post tot exempla num dubium est, quin beneficium
aliquando a servo dominus accipiat? (Sen. Ben. 3.28.1); Numquid aliquis sanus filium a
prima offensa exheredat? (Sen. Clem. 1.14.1); Num igitur non haec omnia oratori
necessaria? (Quint. Inst. 1.10.22); Num enim armatis Cassio et Bruto ac Philippenses
campos obtinentibus belli civilis causa populum per contiones incendo? (Tac. Ann.
4.35.2—NB: in a speech); Num ergo aut fama mentita est Aristoteli aut daemonum ad
hoc ratio est? (Tert. An. 49.3); Num forte non pertinet ad animum? Quis hoc dixerit?
(August. Conf. 10.14); Quid? A principio ad finem num putas perveniri posse, nisi per
aliquod medium? (August. Mus. 1.col. 1095. l. 19);
From Early Latin onwards num is used in combinations with forms of the indefinite
pronoun quis/quid ‘anybody’, ‘anything’, as in (g) and (h), of the indefinite adjective
qui/quae/quod ‘any’, as in (i), and of indefinite adverbs like quando ‘at any time’, as in
(j), and qui ‘by any means?’ They are usually printed as one word. The combination
may, in turn, be combined with the particle nam, as in (k). The Early Latin instances
do not all expect a negative answer, thus (g), but later instances normally do.
(g) Numquid processit ad forum hodie novi?
(‘Has anything new come up at the forum today?’ Pl. Mos. 999)
(h) Numquis vestrum ad vim, ad facinus, ad caedem accommodatus est? Nemo.
(‘Is there any one of you inclined to violence, crime, or even murder? No one.’ Cic.
Agr. 3.16)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The accusative neuter form (num)quid is also used as a degree adjunct (a so-called
accusativus adverbialis) in (l) (similarly, one finds nil moror). Instances like this one
constitute the transition to the use of numquid as a compound question particle from
Plautus onwards, meaning ‘surely . . . not’. In (m), (num)quid is neither an adjunct, as
in (l), nor an argument as in (g) and (k). The use of numquid in indirect questions
started as early as Cicero, as in (n). From the early imperial period onwards numquid
replaces num, especially in authors writing in a less elevated style. It outnumbers num
by far in the works of Augustine, especially in his less Classical Sermones.46
(l) Numquid me morare quin ego liber, ut iusti, siem?
(‘Do you have any objections against me being free, as you said I should be?’ Pl. Men.
1146)
(m) Numquid causam dicis quin te hoc multem matrimonio?
(‘Do you have any objection to me divorcing you?’ Pl. Am. 852)
(n) Et scire sane velim numquid necesse sit comitiis esse Romae.
(‘I should be very glad to know whether I shall have to be in Rome for the elections.’
Cic. Att. 12.8)
Supplement:
Quid nunc? Numquid iratus es aut mihi aut fi- / lio propter has res, Simo? (Pl. Ps.
1329–30); Eho, numquidnam accusat virum? (Ter. Hec. 267); Numquid duas habetis
patrias? (Cic. Leg. 2.5); Numquid enim peccavi? Non sum exoratus? (Sen. Con.
2.3.16); Numquid igitur lex ad absolutos tantum patres pertinet? (Quint. Inst.
7.1.56); Quod si legeris ista cum diserto / —sed numquid sumus improbi?—
Secundo . . . (Mart. 5.80.6–7); ‘Rogo’ inquit ‘Agamemnon mihi carissime, numquid
duodecim aerumnas Herculis tenes, aut de Ulixe fabulam, quemadmodum illi
Cyclops pollicem { poricino { extorsit?’ (Petr. 48.7 (Trimalchio speaking));
46
See Raviolo (2002).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
47
See TLL s.v. an 1.11ff.
48
For an, see Bolkestein (1988). For vero in an questions, see Kroon (1995: 307–9).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(d) Tum Crassus ‘Non in hac’ inquit ‘una, Catule, re, sed in aliis etiam complu-
ribus distributione partium ac separatione magnitudines sunt artium demi-
nutae. An tu existimas, cum esset Hippocrates ille Cous, fuisse tum alios
medicos qui morbis, alios qui volneribus, alios qui oculis mederentur?’
(‘ “That is not the only loss,” Crassus rejoined, “but there are a great many others also
that have been inflicted on the wide domain of science by its being split up into
separate departments. Do you really suppose that in the time of the great Hippocrates
of Cos there were some physicians who specialized in medicine and others in surgery
and others in ophthalmic cases?” ’ Cic. de Orat. 3.132)
(e) Ego intus servem? An ne quis aedis auferat?
(‘I am to keep watch inside, am I? Presumably so that nobody takes away the house?’
Pl. Aul. 82)
(f ) Quid igitur <tu> stulte, quoniam occasio ad eam rem fuit / mea virtute parta,
ut quantum velles tantum sumeres, / sic hoc digitulis duobus sumebas
primoribus? / An nescibas, quam eius modi homini raro tempus se daret?
(‘Then why did you just take it like this, with your two fingertips, you idiot, when I’d
provided you with the opportunity to take as much as you wanted? Or didn’t you
know how rarely an opportunity of this kind presents itself to anyone?’ Pl. Bac.
673–6)
(g) An tu invenire postulas quemquam coquom / nisi milvinis aut aquilinis
ungulis?
(‘Do you expect to find any cook except one with the claws of a kite or an eagle?’ Pl.
Ps. 851–2)
Supplement:
Dialogal instances: Credam istuc, si esse te hilarum videro. # An tu [ess’] me tristem
putas? / # Putem ego quem videam aeque esse maestum, ut quasi dies si dicta sit? (Pl.
As. 837–8); An vero, quia cum frugi hominibus / ibi bibisti qui ab alieno facile
cohiberent manus? (Pl. Trin. 1018–19); Anne oportuit? (Pl. Truc. 666); Obsecro
‘non’? An illam hinc abducet, pater? (Ter. Ad. 661); Sed ad haec, nisi molestum est,
habeo quae velim. # An me, inquam, nisi te audire vellem, censes haec dicturum
fuisse? (Cic. Fin. 1.28); Alterum collegam tuum, viginti virum qui fuit ad agros
dividendos Campanos, video huc venire, Cn. Tremelium Scrofam, virum omnibus
virtutibus politum, qui de agri cultura Romanus peritissimus existimatur. # ‘An non
iure?’ inquam. (Var. R. 1.2.10);
Monologal instances: Di immortales, opsecro vostram fidem, / ubi ego perii? Ubi
immutatus sum? Ubi ego formam perdidi? / An egomet me illic reliqui, si forte
oblitus fui? (Pl. Am. 455–7); Quid erat induta? An regillam induculam? An mendi-
culam? (Pl. Epid. 223); Quid est? Num iniquom postulo? / An ne hoc quidem ego
adipiscar, quod ius publicum’st? (Ter. Ph. 411–12); Huic vos non summittetis, hunc
diutius manere patiemini? Cuius ut provinciam tetigit, sic fortuna cum improbitate
certavit ut nemo posset utrum protervior an infelicior esset iudicare. An vero in Syria
diutius est Semiramis illa retinenda? (Cic. Prov. 8–9); ‘Quidnam,’ inquit Catulus, ‘An
laudationes? Id enim video poni genus tertium.’ (Cic. de Orat. 2.43); Num igitur
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
etiam rhetorum epilogum desideramus? An hanc iam artem plane relinquimus? (Cic.
Tusc. 1.112); Unde igitur ordiar? An eadem breviter attingam quae modo dixi, quo
facilius oratio progredi possit longius? (Cic. Tusc. 2.42); Denique cur numquam caelo
iacit undique puro / Iuppiter in terras fulmen sonitusque profundit? / An simul ac
nubes successere, ipse in eas tum / descendit, prope ut hinc teli determinet ictus?
(Lucr. 6.400–3); O pater, anne aliquas ad caelum hinc ire putandum’st / sublimis
animas iterumque ad tarda reverti / corpora? (Verg. A. 6.719); Num furis? An
prudens ludis me obscura canendo? (Hor. S. 2.5.58); Tum ‘Ut omnia’ inquit ‘obli-
viscamini alia, hodiernam hanc contumeliam quo tandem animo fertis, qua per
nostram ignominiam ludos commisere? An non sensistis triumphatum hodie de
vobis esse? (Liv. 2.38.2–3); Sed esto neque melius quod invenimus esse neque par, est
certe proximis locus. An vero ipsi non bis ac saepius de eadem re dicimus et quidem
continuas nonnumquam sententias? (Quint. Inst. 10.5.6–7); Quin adhuc memorari
exempla, quae priscis moribus ad virtutem et gloriam Romana indoles prodiderit. An
parum quod Veneti et Insubres curiam inruperint, nisi coetu alienigenarum velut
captivitas inferatur? (Tac. Ann. 11.23.2–3); . . . unus ex conlibertis Trimalchionis
excanduit . . . et ‘Quid rides’ inquit ‘vervex? An tibi non placent lautitiae domini
mei? Tu enim beatior es et convivare melius soles. (Petr. 57.1–2); An vero ut per
angelum liberaret hominem? Cur ergo ipse descendit ad id quod per angelum erat
expediturus? (Tert. Carn. 14.3); Itane, domine deus meus, est quidquam in me quod
capiat te? An vero caelum et terra, quae fecisti et in quibus me fecisti, capiunt te?
(August. Conf. 1.2);
49
See Svennung (1935: 516).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
50
For a survey of the development and references, see Herman (1996a). Examples in Blaise s.v. si § 3.
51
See Hofmann in TLL s.v. ecquis 52.33ff.
52
For these interrogative indefinite expressions, see TLL s.v. ecqualis etc. 51.27ff.; references in Sz.: 464.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
From Early Latin onwards the particle ecquid is used more or less in the same way as
numquid, although it does not just invite a negative answer (in Silver Latin and later it
occasionally invites a positive answer). Examples are (f )–(i). It may be combined with
the particle -nam, as in (i). On the whole it is rare and almost absent in Christian
writers. Ecquid is also used in indirect questions.
Supplement:
Quid tu? Ecquid locutu's cum istac quam ob rem hanc ducimus? (Ter. Ph. 798);
Ecquid tandem tibi videtur, ut ad fabulas veniamus, senex ille Caecilianus minoris
facere Eutychum filium rusticum quam illum alterum Chaerestratum. (Cic. S. Rosc.
46); Ecquid ergo intellegis quantum mali de humana condicione deieceris? (Cic.
Tusc. 1.15); Et ecquid / coniugibus nostris mutua cura sumus? (Ov. Fast. 2.729–30);
Gnatos ecquid agnoscis tuos? / # Agnosco fratrem. (Sen. Thy. 1005–6); Ecquid te
satis aestimas beatum, / contingunt tibi si manus minores? (Mart. 11.1.7–8);
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Quis illaec est mulier timido / pectore peregre adveniens / quae ipsa se miseratur?
(‘Who is that woman coming from abroad with fear in her breast, who is pitying
herself?’ Pl. Epid. 533–4—subject)
(b) Ubi sunt isti scortatores qui soli inviti cubant?
(‘Where are those lechers who are lying alone against their will?’ Pl. Am. 287—
position in space adjunct)
(c) Ut matrem tuam / videas, adeas, advenienti des salutem atque osculum. /
# Quam meam matrem?
(‘So that you can see your mother, go to her, and greet and kiss her on her arrival, and
give her a kiss. # What mother of mine?’ Pl. Epid. 570–2—determiner)
(d) Quid nunc? Quam mox recreas me?
(‘What now? How soon are you restoring me to life?’ Pl. Cas. 743—degree modifier)
(vi) adverbs of degree: quam ‘to what degree?’ ‘how?’ (especially modifying
measurable adjectives and adverbs like multus ‘many’ and diu ‘long’), quan-
topere ‘how greatly?’ (especially functioning as a degree adjunct).
Many of the words in sections (ii), (iii), (v), and (vi) are also found in exclamatory
sentences. For quis and qui this is rare. Quam is much more common in exclamatory
sentences.53 This occurrence in two types of sentences is illustrated for ut ‘how?’ in (e)
(editors vary in their punctuation).
(e) Quid? Quae postea sunt in eum congesta, quae quemvis etiam mediocrium
delictorum conscientia perculissent, ut sustinuit? Di immortales, sustinuit?
Immo vero ut contempsit ac pro nihilo putavit, quae neque maximo animo
nocens, neque innocens nisi fortissimus vir neglegere potuisset!
(‘Then again, the calumnies that later were heaped upon him—calumnies which
would have stunned any man whose conscience accused him of even trivial misde-
meanours—how resolutely, ye gods, did he face those! Faced, do I say? Nay, despised
them, and set them at naught, though none that had been guilty, however high his
spirit, and none that had been innocent, unless he were a man of iron soul, could
have disregarded them.’ Cic. Mil. 64)
Question words are rarely combined with the (neutral) clitic question particle -ne (a
pleonastic expression, occasionally exploited in later times), and with the clitic
particle -nam ‘tell me’, which underlines the desire to obtain the information
requested. Examples are (f ) and (g).
(f ) O Callicles, o Callicles, o Callicles, / qualine amico mea commendavi bona?
(‘O Callicles, o Callicles, o Callicles! What sort of friend did I entrust my possessions
to?’ Pl. Trin. 1094–5)
(g) Quinam homo hic ante aedis nostras / eiulans conqueritur maerens?
(‘Who on earth is lamenting here in front of our house, with wailing and sadness?’ Pl.
Aul. 727–7a)
These question words are also used in subordinate clauses, such as accusative and
infinitive clauses, participial clauses (both secondary predicates and ablative absolute
clauses), and also finite subordinate clauses. Examples are (h)–(j), respectively.
(h) Quam multas existimatis insulas esse desertas, quam multas aut metu
relictas aut a praedonibus captas urbis esse sociorum?
(‘How many islands do you suppose were deserted, how many of your allies’ cities
either abandoned through fear or captured by the pirates?’ Cic. Man. 32)
(i) Quam vero utilitatem aut quem fructum petentes scire cupimus . . .
(‘Again, what desire for profit or advantage underlies our curiosity to learn . . . ’ Cic.
Fin. 3.37)
53
For the use of quam in interrogative and exclamatory sentences and clauses, see Bodelot (2010a).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
An interrogative sentence may contain two or more question words. An example is (k).
There are more attestations in indirect questions, as in (l).
Supplement:
Quocirca scribam tibi tres libros indices, ad quos revertare, siqua in re quaeres, quem
ad modum quidque te in colendo oporteat facere. (Var. R. 1.1.4—NB: coordination);
Quam ob rem etiam atque etiam considera, C. Piso, quis quem fraudasse dicatur.
(Cic. Q. Rosc. 21); Sed tamen videmus quibus exstinctis oratoribus quam in paucis
spes quanto in paucioribus facultas quam in multis sit audacia. (Cic. Off. 2.67); . . . ut
54
For this and other instances and its popularity in Christian texts, see Wölfflin (1887a: 617–18).
55
For the textual problems, see Sörbom (1935: 107) and Koestermann ad loc.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
non solum auribus acciperetur, sed etiam oculis cerneretur, quem ex [v.l. et ex]
quanto regno ad quam fortunam detulisset. (Nep. Timol. 2.2); . . . commemorantium
ex quantis opibus quo reccidissent Carthaginiensium res. (Liv. 30.42.18);
(iii) It may contain the particle utrum (that originated from the n. sg. pronoun
utrum ‘which of the two’), as in (f )–(g). This usage became frequent from
Cicero onwards. Utrum is occasionally combined with the clitic -ne into one
particle utrumne, as in (h). From Early Latin onwards the first part may
contain both the particle utrum and the clitic particle -ne, the latter attached
to a scope constituent of that first part further on in the sentence. Examples are
(i) and (j). This is less common than just utrum . . . an, but not at all rare.
Utrumnam is found in Late Latin.
When the second (or following) alternative is negative, an non (also written annon) is
the regular expression in direct questions, as in (k) and (l). Necne is exceptional—see
(m). Necne is, however, regular in indirect questions, as in (n) and (o). An non, as in
(p), is less frequent and absent in some authors. The verb may be omitted in the
second part, as in (l), (m), and (o).
Particles57 like iam ‘then’ and etiam58 ‘really’ convey a sense of indignation, irritation,
or surprise, as in (c) and (d). The particle tandem ‘really’, ‘then’, ‘at last’ signals
impatience, as in (e).59 For the use of vero, see especially § 6.14. The particle nam,
usually clitically attached to a question word (and the particles num and utrum),
indicates that subsidiary information is requested with respect to a preceding utter-
ance, often with an overtone of impatience or irritation, as in (f ) and (g).60 The
request for information illocutionary force may also be strengthened by the expres-
sions amabo ‘please’, quaeso ‘please’, or more urgently by obsecro ‘pray’. Examples are
(h) and (i). More details may be found in § 6.29. The adverb satin (= satisne), also in
combination with ut, makes a question more emphatic, as in (j).61
(c) Sed age responde. Iam vos rediistis in concordiam?
(‘But go on, answer me: have you returned to harmony now?’ Pl. Am. 962)
56
For the influence of the presence of an explicit subject on the interpretation of a question, see Finke
(1987: 284).
57
For particles in questions, see Risselada (2005) and Rosén (2009: 347–50).
58 59
See TLL s.v. 929.18ff. For tandem, see Müller (1997: 79-80) and Risselada (1998b).
60
For the use of nam and -nam in questions, see Kroon (1995: 165–8); Müller (1997: 76–7); Holmes
(2012); TLL s.v. nam 28.43ff.
61
For more examples, see TLL s.v. -ne.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
A way to emphasize the interrogative force in constituent questions is the use of so-
called cleft constructions (see Chapter 22),62 as in (k) and (l). In (k), the question has
an indirect assertive illocutionary force: ‘None of those Greeks would think . . . ’. Note
that this construction is also used in indirect questions, as in (l).
(k) Quis enim est istorum Graecorum, qui quemquam nostrum quicquam
intellegere arbitretur?
(‘Who is there, in fact, among those Greeks, who would credit one of us with
understanding anything?’ Cic. de Orat. 2.77)
(l) . . . quaerendum esse visum est quid esset cur plures in omnibus artibus
quam in dicendo admirabiles extitissent.
(‘ . . . it has seemed to me to be a matter for inquiry, why it was that more of them
should have gained outstanding renown in all other pursuits, than have done so in
oratory.’ Cic. de Orat. 1.6)
Interrogative sentences are often used with an indirect assertive or directive illocut-
ionary force.63
62
For these constructions, see B. Löfstedt (1966).
63
For the use of questions instead of declarative and imperative sentences, see Hofmann (1951: 66–9)
and Risselada (2005), who presents a more detailed classification of illocutionary forces.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
64
On rhetorical questions, see Hoff (1983). On the role of negation, see Orlandini (2001: ch. 9, esp. p. 278).
65
See Orlandini (1991: 143–53).
66
For enim, see Kroon (1995: 189). For further examples, see OLD s.v. enim § 3b, etenim § b.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
A special type of rhetorical questions are the so-called REPUDIATING or polemical questions,
in which a speaker reacts to the preceding words of someone else, more or less echoing the
words used, as a form of protest against the preceding utterance. Examples are (d) and (e).
Such questions may be without a question particle or with the clitic particle -ne. In the
comedies of Plautus and Terence -ne is more common. (For the use of the so-called
potential subjunctive in such utterances and for more examples, see §§ 7.42, 7.44, 7.46.)
(d) Ne fle. # Egone illum non fleam? Egon’ non defleam / talem adulescentem?
(‘Stop crying. # Should I not cry for him? Should I not weep without restraint for
such a man?’ Pl. Capt. 139–40)
(e) Bonus est hic vir. # Hic vir sit bonus?
(‘He’s a good man. # Him a good man?’ Ter. An. 915)
The distinction between real questions (percontatio) and rhetorical questions (inter-
rogatio) was already made in Antiquity and the effects of the latter well understood.
An example is the treatment in the Rhetorica ad Herennium: Interrogatio non omnis
gravis est neque concinna, sed haec quae, cum enumerata sunt ea, quae obsunt
causae adversariorum, confirmat superiorem orationem . . . (‘Not all interrogation is
impressive or elegant, but that interrogation is, which, when the points against the
adversaries’ cause have been summed up, reinforces the argument that has just been
delivered . . . ’ Rhet. Her. 4.22).67
67
For ancient views on the effect of rhetorical questions, see Calboli (1981: 135–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
68
See Risselada (1993: passim); Orlandini (1991: 291–5).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Another relatively frequent elliptical question is with ita ‘so’, as in (c) and (d), where
ita is also used in the reaction.70 An example from Cicero is (e).
(c) Ego te vehementer perire cupio, ne tu [me] nescias. / # Itane vero? # Ita
hercle vero.
(‘Just so you know, I very much wish that you come to a sticky end. # Indeed? # Yes
indeed.’ Pl. Cur. 724–5)
(d) Numquam ecastor hodie scibis prius quam ex te audivero. / # Itane est? # Itane est.
(‘You’ll never know it today until I hear it from you. # Really? # Really.’ Pl. Per. 219–20)
(e) Qui per se litem contestatur, sibi soli petit, alteri nemo potest, nisi qui
cognitor est factus. Itane vero?
(‘He who begins an action in person claims for himself only; no one, unless he has
been appointed agent, can claim for another. Do you say so?’ Cic. Q. Rosc. 53)
Appendix: The interjection hem is used in combination with questions in reaction to
a surprising utterance of an interlocutor (f), the unexpected appearance of someone
(g), etc. It may also be used alone, as in (h).71
(f ) Salve, exoptate gnate mi. # Hem, quid ‘gnate mi’?
(‘Greetings, my son; I’ve been longing for you. # Huh? What do you mean
“my son”?’ Pl. Capt. 1006)
(g) Heus, Theopropides. # Hem, quis hic nominat me?
(‘Hey, Theopropides! # Oh, who is calling me?’ Pl. Mos. 784)
(h) Geta. # Hem? # Quid egisti? # Emunxi argento senes.
(‘Geta! # Yes? # What have you done? # I’ve tricked the old men out of their
money.’ Ter. Ph. 682)
69
See Stockert (1978), who elaborates on Hofmann (1951: 103).
70
For the repetition on -ne see Sz.: 461 (Woytek ad loc. has another explanation).
71
For hem in Plautus and Terence, see Luck (1964: 13–46) and Müller (1997: 43–5).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
72
Other instances mentioned in the literature are Cic. Att. 15.4.3 (may be an elliptical purpose clause
as well).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
73
For the clausal approach, see Hofmann in his section on ‘affektische Kurzsätze’ (1951: 51); for the
independent sentence approach, see K.-St.: II.510; Woodcock (1959: 131). OLD s.v. ut § 44, cautiously
follows the independent sentence approach, while suggesting a link with the use of ut-clauses with verbs
like happening (OLD § 37b).
74
A more detailed division is given by Risselada (1989, 1993: 153).
75
See Calboli (1968: 467); Lehmann (1973: 21–34); L. Löfstedt (1966: 12–20); Prat (1975); Scherer
(1975: 76–7); Sonnenschein (1910: 7); Touratier (1977: 372). For a discussion of exceptional occurrences
of non, see L. Löfstedt (1966: 12–20) and Pinkster (1986a).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(b) Conservi conservaeque omnis, bene valete et vivite. / Bene quaeso inter vos
dicatis et med apsenti tamen.
(‘All my fellow servants, male and female, goodbye and farewell. Please speak well of
me even in my absence.’ Pl. Mil. 1340–1)
(c) At pol ego aps te concessero. # Iamne abis? Bene ambulato. / Sed recipe
quam primum potes. Cave fuas mi in quaestione.
(‘Well, I’ll get away from you. # Are you leaving already? Have a good walk. But
return as soon as you can, don’t force me to look for you.’ Pl. Per. 50–1)
(d) Ne sis plora.
(‘Please, stop crying.’ Pl. Per. 656)
(e) Proin tu ne quo abeas longius ab aedibus.
(‘So don’t go anywhere too far from the house.’ Pl. Men. 327)
(f ) Ne sis ferro parseris.
(‘Please don’t be economical with iron.’ Pl. Per. 572)
(g) Qui autem auscultare nolet, exsurgat foras . . .
(‘If anyone doesn’t want to listen, let him get up and get out . . . ’ Pl. Mil. 81)
(h) Ne quis in hanc plateam negoti conferat quicquam sui.
(‘Let no one bring any business of his into this street.’ Pl. Capt. 795)
(i) Eae (sc. litterae) te ne moverint.
(‘Don’t let it impress you.’ Cic. Att. 16.1.6)
(j) Ut voles, ut tibi lubebit, nobis legem imponito.
(‘As you wish, as you please, impose your terms on us.’ Pl. As. 239)
(k) Noli sis tu illi advorsari.
(‘Just don’t oppose him.’ Pl. Cas. 205)
(l) Quin tu hoc crimen aut obice ubi licet agere, aut iacere noli ubi non oportet.
(‘Then either launch this charge where it is lawful to start an action, or do not launch
it where it ought not to be brought.’ Cic. Q. Rosc. 25)
76
For this section, see Risselada (1993: ch. 5 ‘Imperative Directives’), Sánchez-Manzano (2002), and
Sznajder (2003: 78–80). For the use of various directive expressions in Seneca the Younger, see
Richardson-Hay (2006: 108–11). For Terence, see Núñez (1995) and Müller (1997: 93–101).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(i) A first element in this variation is the availability of different verb forms. The
meanings of the various forms are described in § 7.51 (subjunctives), § 7.64 (present
imperative), and § 7.65 (future imperative). Taking into account the diachronic
developments and the distribution over text types, for those texts in which they do
compete, one may make the following generalization. The present subjunctive is more
common in binding directives, often prohibitives.77 Imperative sentences with a
present subjunctive in Plautus are not found with expressions that invite cooperation
(sis, sultis ‘if you wish’ and sodes ‘if you don’t mind’). Conversely, imperative
sentences with an imperative form in Plautus do contain these expressions. They
also frequently contain intensifying expressions (see below, par. (v)). Generally
speaking, imperative forms can be used in a wide range of directive expressions.
Although it is difficult to make firm statements, given the much lower frequency of
second person present and perfect subjunctive forms, it is interesting that such
intensifying expressions are not found in imperative sentences with subjunctive
forms. They are also rare in declarative sentences with a future indicative form that
have an indirect directive illocutionary force and are binding as well (see § 6.4).78 As
for the difference between the present imperative and the future imperative, this can
be described as more or less immediate, respectively (see § 7.65).
(ii) A second element is ‘directive METADIRECTIVES’:79 directive expressions that refer
to the realization or non-realization of an action expressed in a subordinate con-
struction. The most common expression is fac + (ut) + second or third person
subjunctive, as in (a)–(b). Other expressions urging the addressee to bring about
the desired state of affairs are imperative forms of efficio ‘to bring about’, perficio ‘to
bring about’, operam do ‘to apply oneself ’, as in (c), curo ‘undertake’, video ‘to see
to it’, and, for non-realization, caveo ‘to beware’, mitto ‘to set aside’, as in (d), omitto
‘to leave off ’, parco ‘to refrain’. By these expressions the addressee’s responsibility and
need to participate actively is made explicit, not rarely with the addition of tu, which
makes the command or prohibition more urgent.
(a) Fac is homo ut redimatur. # Faciam.
(‘Have him ransomed. # I will.’ Pl. Capt. 337)
(b) Fac sis bonae frugi sies. Sequere istum, bella, belle.
(‘Make sure you’re a good girl, please. Follow him prettily, my pretty.’ Pl. Cur. 521)
(c) Da operam ut valeas.
(‘Take care of your health.’ Cic. Fam. 12.1.2)
77
See Risselada (1993: 151–5).
78
For ancient descriptions of the difference between subjunctive and imperative that suggest that
imperatives are more binding, see Charisius: Sed interest inter hoc (sc. ne facias) et illud quod diximus ne
fac, quod hoc imperamus, superius suademus (Charisius 228.14–15 K = 295.13–14 B). Compare Don.
Ter. An. 598 ‘Quiescas’ pro ‘quiesce’ imperativi modo, ne iniuriosum videretur. Serv. A. 11.354 ‘Adicias’
pro ‘adice’, . . . quia indecens erat imperative ad regem loqui. Magni (2010: 235) also ‘hypothesizes’ that the
manipulative strength of the imperative is higher than the subjunctive, wrongly it seems.
79
For the notion of ‘directive metadirectives’, see Risselada (1993: 266ff.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Cave and vide (ne) ‘to watch out’ function as explicit signals of warning. The fact that
they are relatively frequent does not mean that they have lost their proper meaning
and have (almost) become equivalents of the negator ne.80 Examples are (e) and (f ).
(e) Cave sis tu istuc dixeris.
(‘Please don’t say that.’ Pl. Per. 389)
(f ) Vide / ne me ludas.
(‘Make sure you aren’t making fun of me.’ Pl. Cur. 325–6)
Supplement:
Si tenetis, ducite. / Cave dirumpatis, quaeso, sinite transigi. (Pl. Poen. 116–17)—NB:
note sg. cave with pl. dirumpatis); Terram cave cariosam tractes. (Cato Agr. 34.1); Est
ita ut dicis. Cave enim putes Attici nostri Amalthio platanisque illis quicquam esse
praeclarius. (Cic. Leg. 2.7); Mihi ausculta. Vide ne tibi desis. (Cic. S. Rosc. 104);
The combination of noli or nolite ‘do not want’ + infinitive forms one of the
expressions of prohibition found from Early Latin onwards. In comparison with ne
feceris and other prohibition expressions it seems to be more formal (Cicero uses it
more frequently in court, and less frequently in his letters to Atticus and Quintus) and
less authoritative.81 Examples where it is used in the same contexts as other expres-
sions and therefore less convincingly different are (g) and (h). See also (l) in § 6.28.
The combination continued to be used until Late Latin.
(g) Noli sis tu illi advorsari. / Sine amet, sine quod lubet id faciat . . .
(‘Just don’t oppose him. Let him love, let him do what he likes . . . ’ Pl. Cas. 205–6)
(h) Sta ilico. Noli avorsari nec te occultassis mihi.
(‘Stand where you are, don’t turn away or hide from me.’ Pl. Trin. 627)
Supplement:
Noli flere (Pl. Per. 622); Noli, amabo, suscensere ero meo, causa mea. (Pl. Poen. 370);
Quid agi’? Quo abis? # Mitte me. / # Noli inquam. # Non manum abstines, mastigia?
(Ter. Ad. 780–1); Redduc uxorem ac noli advorsari mihi. (Ter. Hec. 654); Verum
80
For critical discussion of earlier ideas about these words, see Risselada (1993: 277).
81
See Risselada (1993: 300) and LSS § 10.2.1.3. For noli in non-polite situations, see also Ferri (2008,
2012: 120–2).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
enim metuo malum. / # Noli metuere. (Ter. Ph. 555–6); Hanc tollite ex civitate,
iudices, hanc pati nolite diutius in hac re publica versari. (Cic. S. Rosc. 154); Noli
mirari. Facilius est ignoscere bello quam parricidio. (Sen. Con. 10.3.15); Nolite a me
omnia exigere quae scio. (Sen. Con. 10.6.2); Ergo aut tace aut meliorem noli
molestare, qui te natum non putat. (Petr. 58.10 (a freedman speaking)); Ergo aut
eruite, si creditis, aut nolite credere qui non eruitis. (Tert. Apol. 7.1); Nolite terreri ab
eis qui vos solummodo occidere possunt . . . (Tert. Marc. 4.28.3);
(iii) Then there is the use of the subject pronoun tu ‘you’ in imperative sentences,
when it does not serve to designate the addressee or express contrast with another
possible addressee. This use of tu is typical of authoritative situations.82 In (i), the use
of tu serves to make clear which maid is meant to effectuate the command, most
probably accompanied by a gesture. But this is not the case in (j) and (k).
(iv) Latin speakers/writers use a number of expressions that make their involve-
ment more explicit, notably parenthetically used first person (singular) forms of
verbs of asking and praying like obsecro ‘I implore’, quaeso ‘I ask’, less often oro ‘I
beseech’, and (not in Early Latin) rogo ‘I ask’, usually without an explicit object te,83
as well as the expression amabo ‘please’, lit. ‘I’ll love you’. These expressions are
also used with interrogative sentences and function occasionally as illocution
converters with declarative sentences (see § 6.3). The general effect of the use of
these expressions is to make the utterance in which they are used more polite (oro
is the most insistent).84 The imperative sentences may nevertheless contain an
exhortation particle such as age, as well as curses such as hercle (typically used by
males) and ecastor (typically used by females). Examples are (l) and (m). Politeness
expressions are more frequent in female than in male speech in Roman comedy and
there is also a difference in the words used by the sexes or the frequency with which
82
Risselada (p.c.).
83
For a survey of the verbs used in this way, see Risselada (1993: 249), Rosén (2009: 419–22), and
Unceta (2009: 67–76). For the difference in the use of them by men and women in Roman comedy, see
Adams (1984) and Müller (1997: 93–101). For the differences between velim and quaeso, on the one hand
(more intimate), and oro, peto, and rogo on the other (less intimate) in Cicero’s letters, see Dickey (2012).
For rogo in private letters, see Adams (2003a: 19) and Molinelli (2008: 371–4). For the use of quaeso and
rogo as ‘interactional discourse markers’, see Molinelli (2010). For the unexpected use of parenthetical
opinor in imperative sentences, see Bolkestein (1998: 26).
84
For oro, see Risselada (1993: 254–5).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
they are used in comedy: amabo is typically female, quaeso predominently male (sis
and sodes are also male).85
(l) Opsecro ecastor, quor istuc, mi vir, ex ted audio?
(‘Good heavens, please, why must I hear that from you, my man?’ Pl. Am. 812)
(m) Age, quaeso, mi hercle translege.
(‘Go on now, please, read through it for me!’ Pl. As. 750)
(n) Atque imperce quaeso.
(‘And please don’t exert yourself too much.’ Pl. Am. 500)
(o) Sed <tu> dic, oro: pater meus tune es?
(‘But please do tell me: are you my father?’ Pl. Capt. 1021)
(p) Propera, amabo.
(‘Hurry please.’ Pl. Bac. 100)
Supplement:
Quid iam, amabo? Ne me appella. (Pl. Am. 810); Sed amabo te, mi Attice (videsne
quam blande?), omnia nostra, quoad eris Romae, ita gerito, regito, gubernato ut nihil
a me exspectes. (Cic. Att. 16.2.2—NB: note the presence of the object te, common in
Cicero, and the future imperatives);86 Urbanitatis possessionem, amabo, quibusvis
interdictis defendamus. (Cic. Fam. 7.32.2);
Nolite obsecro vos acerbiorem mihi pati reditum esse quam fuerit ille ipse
discessus. (Cic. Mil. 103); Sed, obsecro, enitamur ut aliquid ad id quod cupio
excogitemus. (Cic. Att. 12.42.2); Deinde nobis mittas, obsecro, interpretationem
tuam de septuaginta, quam te edidisse nesciebam. (August. Ep. 82.5);
Dic, oro te, clarius. Vix enim mihi exaudisse videor. (Cic. Att. 4.8a.1—NB: explicit
object is common in Cicero); At tu, oro, solare inopem et succurre relictae. (Verg. A.
9.290); Hinc, oro, participem me facias cogitationum tuarum quid in talibus vel
facere soleas vel faciendum esse reperias. (August. Ep. 80.3);
Quaeso, apsolvito hinc me extemplo, quando satis deluseris. (Pl. Am. 1097); Narra
istuc quaeso quid sit. (Ter. Eu. 562); Atque ipsum decretum, quaeso, cognoscite.
(Cic. S. Rosc. 25); Vide, quaeso, ne quid temere fiat. (Cic. Att. 15.19.1); Videte enim,
quaeso vos, quibus quatiebamur affectibus. (August. Ep. 31.3—NB: often with
explicit object); Egredere, quaeso, paulisper e corpore . . . (Hier. Ep. 22.41);
ITA ROGO QUAM PRIMUM ALIQUIT MI MITTE. (CEL appendix Vindol. Ł 14–15 (Vindo-
landa, c. AD 100));
First person singular forms of verbs of asking and praying can also be used (and most
of them more often are) with an imperative clause as their object (with or without the
subordinator ut), the so-called ‘performative’ use of these verbs. See § 15.56. It is not
always possible to decide which construction the utterance should be assigned to.
85
See Adams (1984: 55–68) and Risselada (1993), both with further references. For amabo, see Blase
(1896b) and Holford-Strevens (2010: 334–6).
86
For a possible ‘archaic and self-consciously affected air’ of amabo te in Cicero’s letters, see Hall
(2009: 80–1).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
A common idiom to make one’s wish explicit is the use of the subjunctive form velim
‘I’d like’ in combination with an imperative argument clause (without the subordi-
nator ut), as in (s) and (t) below.
(s) Vera dicas velim.
(‘I wish you would tell the truth!’ Pl. Cas. 234)
(t) Sed mihi hoc credas velim.
(‘Yet please take my word for this.’ Cic. Sul. 46)
87
See TLL s.v. inquam 1782.45ff.; Müller (1997: 180).
88
On these particles, see Risselada (1994, especially pp. 319–20). For rare instances of enim in certain
directives, see Kroon (1995: 191–5).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
nobis suades. (Pl. As. 644); I, refer. Dimidiam tecum potius partem dividam. / Tam
etsi fur mihi es, molestus non ero. I vero, refer. (Pl. Aul. 767–8); Agedum conferte
nunc cum illius vita <vitam> P. Sullae vobis populoque Romano notissimam, iudices,
et eam ante oculos vestros proponite. (Cic. Sul. 72); Quin adsidua ista numinum
dignatione laetus capesse gaudium et potius exulta ter futurus . . . (Apul. Met. 11.29.4);
In imperative sentences, the fossilized imperative forms age and agite (‘come on’)
serve as an exhortation. An example is (y), where age is combined with a first person
plural form. In reactions to a directive utterance, it indicates that the speaker accepts
(‘OK’), as in (z). The likewise fossilized hypothetic expressions sis (= si vis), sultis (< si
vultis ‘if you wish’), and sodes (< si audes ‘if you please’) insist on the cooperation of
the addressee, as in (aa) and (ab).89
(y) Age decumbamus sis, pater.
(‘Go on, let’s recline, father, if you please.’ Pl. As. 828)
(z) Quid veniam? Nil habeo. # At tamen, siquid. # Age veniam.
(‘Why should I come? I don’t have anything. # But even so, if you can— # All right,
I’ll come.’ Ter. An. 713)
(aa) Hoc agite sultis, spectatores, nunciam.
(‘Spectators, pay attention now if you please.’ Pl. As. 1)
(ab) Exi e culina sis foras, mastigia, / qui mi inter patinas exhibes argutias.
(‘Come out of the kitchen, you whipping post, will you? You’re showing me your glib
tongue among the platters.’ Pl. Mos. 1–2)
The forms sis and sultis are almost confined to Plautus. Like these two, sodes is used
in imperative sentences, mostly with an imperative verb form. Sis and sodes usually
follow the verb form, but sis may also follow other emphatic words in the sentence,
behaving in fact as a clitic particle.90 Sultis is more flexible in its position.91
Supplement:
Noli sis tu illi advorsari (Pl. Cas. 204); Ne sis me obiurga, hoc non voluntas me
impulit. (Pl. Mer. 321); Cave sis tu istuc dixeris. (Pl. Per. 389); Servate istum sultis
intus, servi . . . (Pl. Capt. 456); Tu, ut tempus est diei, vide sis nequo hinc abeas
longius. (Ter. Hau. 212); Vide sis quo loco re<s> publica siet . . . (Cato orat. 171);
Da sodes aps te. Post <ego> reddidero tibi. (Pl. Men. 545); Concede hinc a foribu’
paullum istorsum sodes, Sophrona. (Ter. Ph. 741);
89
For Plautus’ instances of sodes, sis, and sultis, see Lodge: II.631Aff.
90
See Dickey (2006).
91
For discussion, see Müller (1997: 94–5); Dickey (2006, 2012a, 2012b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
However, as was indicated above, the range of the directive illocutionary force of
imperative sentences is rather wide and this explains why we do find instances like
(a), a series of wishes for bad and good luck, and (b), a piece of advice. We even find
the imperative form of the verb vapulo (‘to be beaten’), as in (c), and of the copular
verb fio (‘to become’), as in (d). Alternatively, these instances might be regarded as
having an optative illocutionary force (see § 6.33). Unlike wishes, however, these
imperative forms are not found in combination with utinam, whereas we do find
expressions like profecto ‘undoubtedly’, as in (c), and age ‘come on’, as in (d).92
(a) Tu vel suda vel peri algu vel tu aegrota vel vale.
(‘You may sweat or die from cold or be ill or be well.’ Pl. Rud. 582)
(b) . . . nec dic quid doleas. Clam tamen usque dole . . .
(‘ . . . and do not say why you are in pain. Nevertheless secretly be in pain all the time.’
Ov. Rem. 694)
(c) Nunc profecto vapula ob mendacium.
(‘Now, indeed, get a thrashing for your lie.’ Pl. Am. 370)
(d) Age fi benignus: subveni.
(‘Go on, be generous, help me out.’ Pl. Per. 38)
Supplement:
Gaude. # Quid ego gaudeam? # Quia ego impero. Age gaude modo. (Pl. Capt. 839);
Agite bibite, festivae fores, / potate, fite mihi volentes propitiae. (Pl. Cur. 88–9); . . . fite
causa mea ludii barbari . . . (Pl. Cur. 150); Non manes? / # Vapula. (Ter. Ph. 849–50);
At tu iudicium secuta nostrum / plora, si sapis, o puella, plora. (Mart. 2.41.22–3);
Nolo ergo quemquam diligas. Vel solus peri. (August. Disc. 4); Dole illum, noli
gaudere. (August. Serm. 16A.7);
Some of these imperatives, for example vale, function also or mainly as greetings and
expressions of farewell. In (e), vale is used in both ways.
(e) Etiam atque etiam, noster Tiro, vale. Medico, Curio, Lysoni de te scripsi
diligentissime. Vale, salve.
(‘Again, dear Tiro, goodbye. I have written to the doctor, to Curius, and to Lyso
most particularly about you. Goodbye and good wishes.’ Cic. Fam. 16.9.4)
92
For the use of the imperative in non-directive speech acts, see Risselada (1993: 117–20).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The first instance from a non-didactic text of the directive use of the infinitive used to
express the action required of a specific person in a specific situation is (b), if the text
is correct.94 There are also a few instances in inscriptions. The infinitive is not used in
legal texts to express a command or a prohibition.95 Note in (b) the explicit presence
of a second person agent tu. Ex. (c) has a remarkable future infinitive fore.
(b) . . . tu (te cj. Koster) socios adhibere sacris . . .
(‘ . . . summon your comrades to the sacrifice . . . ’ V. Fl. 3.412)
(c) NE FORE STULTU(M)
(‘Don’t be stupid.’ CIL I2.2174.2 (Barbarano Vicentino, I BC (?)))
Supplement: DOMINE FRATER / ILARIS SEMPER / LUDERE TABULA (ECDS -53701340 (Rome, II/
III AD));96 UTI FELIX (CIL XIII.10017.32 (Kobern));
93 94
For discussion, see Daube (1956: 65–7). Editors now read adhibe.
95
For a discussion of the instances and a survey of the literature, see L. Löfstedt (1966: 192–206). See
also Calboli (1962b) on the Rhet. Her. 3.9 instance.
96
See Ihm (1890: 231–2).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Imperative sentences with an optative illocutionary force (wishes) share with impera-
tive sentences with a directive illocutionary force the use of the negator ne (and neve).
They may involve the speaker, the addressee, or a third person. Unlike directive
utterances, wishes do not presuppose that the subject of the sentence has control over
the event. They include curses, maledictions, and invocations of the gods and fate.
Wishes may refer to the time of speaking or to a time that is anterior to the time of
speaking. Furthermore, Latin makes a distinction between wishes that can be realized
and wishes that cannot be realized. The mood used in imperative sentences with an
optative illocutionary force is the subjunctive. The present and perfect tenses are used
for wishes that can be realized, the imperfect and pluperfect for wishes that cannot be
realized. Four examples to illustrate this are (a)–(d). More examples can be found in
§§ 7.56–7.58. (For imperative verb forms, see § 7.63.)
Wishes often contain special particles, qui (in Early Latin), ut(i) (mainly in Early
Latin), and utinam, in all periods of Latin. The adverbs ita and—especially in
poetry—sic ‘in this way’ are also used. Examples with utinam are given
above. Examples with other particles and adverbs are given in the Supplement. It is
not always clear whether we are dealing with imperative or with exclamatory
sentences.
Supplement:
Qui illum di omnes deaeque perdant. (Pl. Cas. 279); Utinam mea mi modo auferam
quae attuli, salva! (Pl. Aul. 433); Ut istunc di deaeque perdant. (Pl. Per. 298); Valeas
beneque ut tibi sit. (Pl. Poen. 912); Tantum modo Gnaeus noster ne, ut urbem
Iºªø reliquit, sic Italiam relinquat. (Cic. Att. 9.10.4); Tantum modo ut eum
intercludamus ne ad urbem possit accedere . . . (Cic. Fam. 16.12.4); Longas o utinam,
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
97
For more examples, see OLD s.v. utinam § d. Also Mesa (1998b: 542).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(d) Mea quidem hercle causa vidua vivito / vel usque dum regnum optinebit
Iuppiter.
(‘Live in divorce, for all I care, as long as Jupiter keeps his kingship if you wish.’ Pl.
Men. 727–8)
The verb is often in initial position, accompanied by the particle sane ‘admittedly’,
‘certainly’.98 Age ‘come’ and esto ‘so be it’ are also used, as in (e) and (f ).
(e) Age sit ita factum.
(‘Well, let us suppose that this was the case.’ Cic. Mil. 49)
(f ) Esto, fecerit, si ita vis, Torquatus propter suas utilitates . . .
(‘But grant your view; assume if you like that Torquatus acted for his own advantage
. . . ’ Cic. Fin. 2.61)
The concessive interpretation often follows from elements in the context. That may
be a contrasting sentence of some form following the concessive one: for example, one
containing the adverb tamen (‘nevertheless’), as in (b). Just as in modern languages,
such a sequence of sentences may have been produced as one intonation unit, which
creates a relation between the first and second sentence which resembles that between
a subordinate concessive clause and a main clause. In our texts this is often indicated
by using a comma ‘,’ or a semicolon ‘;’ between the two sentences. Concessions are
typical of the argumentative text type and, consequently, are very unevenly distrib-
uted in our corpus. There are, for example, no obvious instances in Early Latin
comedies.99
98
For the use of sane in concessions, see Risselada (1998a).
99
Bennett (1910: I.178) has only one quotation from Early Latin, viz. the example (a) quoted in
the text.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
100
For the exclamatory words and the adjectives involved, see Hoff (1983: 125–7); Vairel-Carron
(1975: passim). The material can be found in a number of studies by Flickinger. A statistical table can be
found in Flickinger and Murley (1923: 163). For quam, see Bodelot (2010a), also with statistical
information. For the question whether exclamatory sentences constitute a separate sentence type, see
Shalev (2002).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
legati ad eos <exierunt>! (Cic. Font. 16); Quam lepide se furari putat! (Cic. Ver. 3.35);
Quanto iam levior est adclamatio! (Cic. Rab. 18); Quot ego voluptates fero! (Pl. St.
657); Quotiens ego hunc Archiam vidi, iudices . . . , quotiens ego hunc vidi, cum
litteram scripsisset nullam, magnum numerum optimorum versuum de iis ipsis
rebus quae tum agerentur, dicere ex tempore, quotiens revocatum eandem rem
dicere commutatis verbis atque sententiis! (Cic. Arch. 18); Alexandria cunctaque
Aegyptus ut occulte latet, ut recondita est, ut furtim tota decemviris traditur! (Cic.
Agr. 2.41);
NB: In indirect speech: Quae Lucceium loqui, quae totam Graeciam, quae vero
Theophanem! (Cic. Att. 9.11.3);
With interrogative sentences exclamatory sentences share the use of the enclitic particle
‐ne, which is more frequent with certain noun phrases and with accusative and
infinitive constructions (see below, this section), but is perhaps also used in (h).101
(h) Oh, pergi’n mulier esse? Nullamne ego rem umquam in vita mea / volui, quin
tu in ea re mi advorsatrix fueris, Sostrata!
(‘Oh, do you have to keep behaving like a woman? Is there nothing that I have ever
wanted in my life on which you haven’t contradicted me, Sostrata?’ Ter. Hau.
1006–7—NB: v.l. ullam ego . . . ?)
Apart from these exclamatory sentences with a special exclamatory word or phrase,
there are several types and subtypes of exclamatory sentences that consist either of a
noun or noun phrase or of an accusative and infinitive clause.102 There are two main
types. One type consists of a noun (or noun phrase) or pronoun in combination with
an evaluative adjective in the accusative, which may or may not be combined with an
emotional interjection or (especially in Plautus) an invocation and cannot contain the
enclitic particle -ne (or -n). This type is called the EVALUATIVE in this Syntax. The other
(NON-EVALUATIVE) type cannot be combined with an emotional interjection (rarely an
invocation) but may contain the enclitic particle -ne. Both types have again two
subtypes.
Vairel-Carron (1975: 177–80 and passim) states that the two types had a different
intonation contour, with the first type characterized by a falling intonation, the
second type by a rising intonation.
101
For a few more instances from Terence, see Müller (1997: 86).
102
This section follows Vairel-Carron (1975). See also Álvarez Huerta (2014).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
other types, see below). The emotional interjections include heu ‘alas’ and, increas-
ingly, o ‘oh’.103 The invocations are sometimes combined with eu ‘splendid’.
(i) Mortalis malos! / Ut ingrediuntur docte in sycophantiam!
(‘What sly mortals! How cleverly they’re embarking on the trick!’ Pl. Poen. 653–4)
(j) Edepol mortalis malos!
(‘Bad people they are!’ Pl. Bac. 293)
(k) Eu, hercle mortalem catum, / malum crudumque et callidum et subdolum!
(‘Goodness, a tricky mortal, bad and unfeeling and clever and wily!’ Pl. Poen. 1107–8)
(l) O perditum latronem! <O> vix ullo otio compensandam hanc rei publicae
turpitudinem!
(‘What an unprincipled bandit! What a disgrace to the Republic, hardly to be
outweighed by any relief from war!’ Cic. Att. 7.18.2)
(m) O audaciam!
(‘The impudence!’ Ter. Ph. 360)
Supplement:
Fortunatum Nicobulum, qui illum produxit sibi! (Pl. Bac. 455—NB: proper name as
head); Edepol cor miserum meum, / quod guttatim contabescit quasi in aquam
indideris salem. (Pl. Mer. 204–5); Eu hercle odiosas res! Quotiens hoc tibi, verbero,
ego interdixi . . . ! (Pl. Mil. 1056); Pro di immortales, mulierem lepidam et pudico
ingenio! (Pl. Mos. 206); Lepidum senem! (Pl. Ps. 435); O facinus impudicum, / quam
liberam esse oporteat servire postulare. (Pl. Rud. 393–4); Hercle, rem gestam bene.
(Pl. St. 379); Nimium graphicum hunc nugatorem! (Pl. Trin. 936); Scitum hercle
hominem! (Ter. Eu. 254); O fallacem hominum spem fragilemque fortunam et
inanes nostras contentiones! (Cic. de Orat. 3.7); Occasionem mirificam, si qui
nunc, dum hi apud me sunt, emere de me fundum Formianum velit! (Cic. Att.
2.14.2); Operam (O operam cj. Lehmann) tuam multam, qui et haec cures et
mea expedias . . . ! (Cic. Att. 13.6.4); O miseras hominum mentes! O pectora caeca!
(Lucr. 2.14);
Pro fidem, Thebani cives! (Pl. Am. 376); Pro Iuppiter, / hominis stultitiam! (Ter.
Ad. 366–7); O virum! O civem! (Cic. Att. 4.15.7); O curas hominum, o quantum est in
rebus inane! (Pers. 1.1); Heu me! (Sen. Tro. 681); Heu sortem necis! (Sil. 7.628);
Delicias hominis! (Juv. 6.47);
Plautus has nugas (sc. agis, etc.) alongside nugae (sc. sunt) ‘nonsense’ and gerrae
‘nonsense’. Terence has also logi ‘fairy tales’ and fabulae ‘nonsense’ (sc. sunt). These
are idioms that cannot be compared with instances like o audaciam given above,
since the verbs can easily be supplied.104
103
For the history of the use of these interjections, see Sz.: 48. For Terence, and Early Latin in general,
see also Müller (1997: 120–38).
104
See Vairel-Carron (1975: 39–41).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Also belonging to this type, because it may be combined with emotional interjections
(apparently no invocations) and because it does not contain the enclitic particle -ne,
are expressions with a personal pronoun and an evaluative adjective, also both in the
accusative (this is the second type of expression traditionally characterized as accu-
sativus exclamationis). Examples are (n)–(q).
The first non-evaluative subtype are noun phrases in the accusative (the third type of
expression called accusativus exclamationis) consisting of a head noun and some
form of deictic expression, either the pronoun hic ‘this’ or the adjective of quantity or
degree tantus ‘such’. Such phrases may or may not contain the enclitic particle -ne.
Examples are (r) and (s). This is a marginal phenomenon and should perhaps be
regarded as an incomplete exclamatory accusative and infinitive clause.
(r) Hancin’ vitam! Hoscin mores! Hanc dementiam!
(‘What a life! What a way to behave! What madness!’ Ter. Ad. 758)
(s) Tantamne patientiam, di boni! Tantam moderationem, tantam in iniuria
tranquillitatem et modestiam!
(‘What patience, good gods, what moderation, what quiet self-effacement in the face
of injustice!’ Cic. Phil. 10.7)
Supplement:
Hoccin’ saeclum! (Ter. Ad. 304); Huncine hominem, hancine inpudentiam, iudices,
hanc audaciam! (Cic. Ver. 5.62); Hascine vires amicitiae? (V. Max. 4.7, ext. 1);
Examples are (t)–(y). They are rare or even absent in most authors, since there is
usually no context they would typically fit in with.105 They may contain the enclitic
particle -ne, attached to the first and most salient word of the sentence. This word is
usually a pronoun or an adverb. The precise status of this particle is unclear. Since -ne
is overwhelmingly used in (direct and indirect) questions, it is attractive to describe
the sentences as interrogative exclamatories.106 Editors vary in the way they punctu-
ate. The infinitive may be present or perfect, active or passive. The subject of the
accusative and infinitive sentence may be omitted if sufficiently clear in its context, as
in (u). It is not always easy to draw a borderline with rhetorical questions in the
accusative and infinitive (for which see § 15.94).107
(t) Perii, hoc servom meum non nauci facere esse ausum!
(‘I’m done for! The idea that my slave dared to hold this cheaper than rubbish!’
Pl. Bac. 1102)
(u) Sati’n hoc mandatum’st tibi? # Ah / (sc. te) rogitare, quasi difficile sit!
(‘Are my instructions quite clear? # Oh! Why keep asking, as if it were difficult?’
Ter. Eu. 208–9)
(v) Ad illum modum sublitum os esse mi hodie!
(‘Is it possible that I was fooled like that today?’ Pl. Capt. 783)
(w) Magistron’ quemquam discipulum minitarier?
(‘Is it possible that any pupil is threatening his tutor?’ Pl. Bac. 152)
(x) Perii. Numquamne etiam me illam vidisse!
(‘Damn it. To think that I have never seen her!’ Ter. Eu. 360)
(y) Sicin’ te mi hoc facere? # Quid faciam?
(‘Are you treating me like this? # What should I do?’ Pl. Per. 42)
Supplement:
Edepol senem Demaenetum lepidum fuisse nobis. (Pl. As. 580); Non mihi licere
meam rem me solum ut volo / loqui atque cogitare sine ted arbitro? (Pl. Cas. 89–90);
Hoccin’ mi ob labores tantos tantillum dari. (Pl. Truc. 537); . . . te facere ausum esse!
(Cato orat. 58); Adeon’ homines mutarier (Ter. Eu. 225); Nullane in re esse
quoiquam homini fidem! (Ter. An. 425); Tantum laborem (sc. me) capere ob
talem filium! (Ter. An. 870); Non te pro his curasse rebu’, nequid aegre esset mihi!
(Ter. Hec. 227); O spectaculum illud non modo hominibus sed undis ipsis et litoribus
luctuosum! Cedere e patria servatorem eius, manere in patria perditores! (Cic. Phil.
10.8); Tamne nullo consilio aut tam contra meum consilium gesta esse omnia!
105
For statistical data, see Melzani (1975).
106
So Melzani (1975: 97). For the reverse argumentation, starting from the affirmative particle ne
‘indeed’, see Müller (1997: 83–8).
107
In the examples quoted in the literature there is sometimes an expression in the context that
actually governs the accusative and infinitive expression. The examples in Bennett (1910: I.424–5) need to
be sorted.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(Cic. Att. 7.12.3); Hem, mea lux, meum desiderium, unde omnes opem petere
solebant, te nunc, mea Terentia, sic vexari, sic iacere in lacrimis et sordibus, idque
fieri mea culpa . . . (Cic. Fam. 14.2.2); Ergo me potius in Hispania fuisse tum quam
Formis cum tu profectus es ad Pompeium! (Cael. Fam. 8.17.1—NB: remarkable use
of ergo); ‘Mene Iliacis occumbere campis / non potuisse tuaque animam hanc
effundere dextra . . . ’ (Verg. A. 1.97–8); Huncine solem / tam nigrum surrexe mihi?
(Hor. S. 1.9.72); Haec ludibria religionum non pudere in lucem proferre et vix pueris
dignas ambages senes ac consulares fallendae fidei exquirere!’ (Liv. 9.11.12); Et cetera
sentire naturam suam, alia pernicitatem usurpare, alia praepetes volatus, alia nare:
hominem nihil scire, nihil sine doctrina, non fari, non ingredi, non vesci, breviter-
que non aliud naturae sponte quam flere! (Plin. Nat. 7.4);108 Di boni, hoc virum
audire, principem scire et usque eo licentiam pervenisse, ut, non dico consulari, non
dico amico, sed tantum marito princeps et adulterium suum narret et fastidium.
(Sen. Dial. 2.18.2);
NB: Elliptic: Generumne nostrum potissimum vel hoc vel tabulas novas? (Cic. Att.
11.23.3);
There is an extensive literature on the question how to explain the use of the
accusative in exclamations. Proposals to explain it as a normal object with a verb
of saying or a verb to be supplied from the context are unconvincing. What the two
subtypes of the evaluative type have in common is that they express an evaluation by
the speaker of the properties of an entity, a relation that can be expressed explicitly
by a verb like puto ‘to consider’, which governs an object and object complement
(a so-called double accusative). This is not really the case for the first subtype of the
non-evaluative type. The exclamatory accusative and infinitive clauses are a reflection
by the speaker about a state of affairs. Such a relation can be expressed explicitly by a
verb like miror ‘I am surprised’, which governs an accusative and infinitive clause.109
Exclamatory sentences can also consist of nouns and noun phrases in the nominative,
as in (z). Some instances may rather be taken as a nominal sentence with a subject and
a subject complement in the nominative. Examples are (aa) and (ab). Alternatively,
they might be regarded as declarative sentences with an exclamatory illocutionary
force. A one-word clause with only the subject (in the nominative) is (ac). The use of
the nominative is much less common than the use of the accusative.
(z) ‘O frustra’ inquit ‘mei (v.l. mihi) suscepti labores, o spes fallaces, o cogita-
tiones inanes meae!’
(‘Alas! that I must see my labours wasted, my hopes disappointed, my dreams
unrealized!’ Cic. Mil. 94)
(aa) O fortunata mors quae naturae debita pro patria est potissimum reddita!
(‘Death, our debt to nature, is fortunate indeed when it is paid for the sake of one’s
native land.’ Cic. Phil. 14.31)
108
For exclamatory accusative and infinitive clauses in Pliny the Elder, see Melzani (1986: 224–5).
109
For the literature, see Sz.: 48–9. A more recent discussion can be found in Vairel-Carron (1975:
49–76). The argumentation in the note is in line with K.-St.: I.272–3. See also Orlandini (1991: 295–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
6.36 Responses
The term RESPONSE is used for the reaction of a speaker to a preceding utterance by
another speaker; the preceding utterance may have a request for information or
directive illocutionary force, in which case a response from the addressee is usually
expected, or it may have an assertive illocutionary force or be an exclamation, to
110
For references, see Sz.: 85. For a discussion of the instances that are mentioned in the literature, see
Quetglas (1979). Some scholars considered these cases as Graecisms (see Calboli 2009: 95–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Responses
111
Cicero uses it as well to engage an addressee or the audience.
112
For the various functions of ita, see Thesleff (1960: 23–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Answers to questions form part of the larger set of reactive moves in discourse (for
which see Chapter 24). This section is limited to more or less cooperative answers to
questions with a request for information illocutionary force: ‘real’ answers to ‘real’
questions. Even with this limitation we can expect considerable variety: for example,
in explicitness, in politeness, and in assertiveness.115 An initial distinction must be
made between answers to constituent questions and answers to sentence (yes/no)
questions. The latter type has drawn some attention from Latinists, since Latin has no
specialized words for ‘yes’ and ‘no’. Examples are given of answers to both direct and
indirect questions, since the mechanisms are the same.
113
In Terence there is a question in every three to four lines (Müller 1997: 190).
114
A detailed study of expressions of agreement and disagreement in Latin comedy is Thesleff (1960).
This section is much indebted to that study. See also Freundlich-Amit (1971) and Müller (1997: 33–64).
115
For a classification of answers to questions, see Kienpointner (1998: 70–84).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Responses
(b) Cedo, quoium puerum hic adposisti? Dic mihi. / # Tu nescis? # Mitte id quod
scio. Dic quod rogo. / # Vostri. # Quoi<u>s nostri? # Pamphili.
(‘Come on, whose baby have you put here? Tell me. # Don’t you know? # Never mind
what I know. Tell me what I ask. # Your master’s. # Which master? # Pamphilus.’
Ter. An. 763–5)
(c) Quaesivit unde esset epistula. Respondit ‘Agrigento’.
(‘He asked where the letter came from. He answered: “From Agrigentum.” ’ Cic. Ver.
4.58)
(d) Cedo qui id credam. # Quia illas emit in Anactorio parvolas / de praedone
Siculo. # Quanti? # Duodeviginti minis / duas illas et Giddenenem nutricem
earum tertiam.
(‘Tell me how I can believe this. # Because he bought them in Anactorium from a
Sicilian pirate when they were little. # For how much? # For eighteen minas, the two
of them and Giddenis, their nurse, as a third.’ Pl. Poen. 896–8)
(e) Quis illaec est mulier? # Pro di immortales, / eri concubina est haec quidem.
(‘Who is that girl? # Immortal gods, this is master’s concubine.’ Pl. Mil. 361–2)
(f ) Quid istic scriptum? # Iuxta tecum, si tu nescis, nescio; nisi fortasse blanda
verba.
(‘What’s written there? # If you don’t know it, I don’t know it any more than you; but
perhaps flattering words.’ Pl. Per. 249–50)
The most common form of an affirmative answer echoes in some way the preceding
question by repeating either the word or phrase that constitutes the scope of the
question or the main verb (or both). Examples are (a)–(c). This ‘core’ of the answer
may be extended by additional information, as in (d).117
116
Responses to questions often contain a vocative that refers to the person who asked the question.
For the interactional function of these vocatives, see Shalev (1998).
117
In Terence ‘scope-repeating’ answers and ‘freely formulated’ answers are used with the same
frequency. See Müller (2012: 112–14).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Ipsus es? / # Ipsus, inquam, Charmides sum. # Ergo ipsusne es? # Ipsissimus.
(‘Are you the one yourself? # I’m telling you, I am Charmides myself. # Then are you
him yourself? # My selfest self.’ Pl. Trin. 987–8)
(b) Tun’ me heri advenisse dicis? # Tun’ te abiisse hodie hinc negas? / # Nego
enim vero.
(‘Are you saying that I arrived yesterday? # Are you denying that you went away from
here today? # Yes, I am denying it indeed.’ Pl. Am. 758–9)
(c) Tu[n] iubes hoc, Demea? / # Ego vero iubeo et hac re et aliis omnibus / quam
maxume unam facere nos hanc familiam . . .
(‘You told him to, Demea? # I did. In this and in every other way we should do our
best to unite this household . . . ’ Ter. Ad. 924–6)
(d) Quid nunc, mulier? Audi’n illum? # Ego vero, ac falsum dicere.
(‘What now, woman? Can you hear him? # I can indeed hear him . . . tell a lie.’
Pl. Am. 755)
Instead of repeating one or more words from the preceding utterance (more or less)
literally, the answer may contain one or more synonymous, sometimes more precise
or stronger, substitutions, as in (e) and (f ).
Proper anaphoric expressions are shown in (i)–(k). Is in (i) refers to Ballio. Istuc in (j)
refers to the preceding state of affairs. The manner adverb ita can be used in
combination with various expressions, but also alone, as in (k), equivalent to ita est
(see also the examples in § 6.36). Sic (est) ‘it is so’ is used in a similar way.
Responses
(j) Dic bona fide: tu id aurum non surripuisti? # Bona. / # Neque <eum> scis qui
abstulerit? # Istuc quoque bona.
(‘Give me your word of honour: you didn’t steal this gold? # Upon my honour, no. #
And you don’t know the man who took it away? # Upon my honour, it’s no again.’ Pl.
Aul. 772–3)
(k) Haecine tua domu’st? # Ita inquam.
(‘Is this your home? # Yes, I assure you.’ Pl. Am. 362)
Apart from the forms of expressing agreement with a preceding question that are
mentioned above there are many other ways to express (a certain degree of) affirm-
ation in an indirect way. One way is by the use of communication and cognition
verbs, such as fateor ‘I confess’, as in (l), and censeo and arbitror ‘I think so’, as in (m)
and (n). Compare also verum in (o).
(l) Est tibi nomen Menaechmo? # Fateor. # Est itidem tibi? / # Est.
(‘Is your name Menaechmus? # I admit it. # Is it also yours? # Yes.’ Pl. Men. 1107–8)
(m) Patri etiam gratulabor cum illam invenit? # Censeo.
(‘Will I also congratulate her father on finding her? # I reckon.’ Pl. Rud. 1270)
(n) Ibi tum matri parvolam / puellam dono quidam mercator dedit / ex Attica
hinc abreptam. # Civemne? # Arbitror. / Certum non scimus.
(‘While she was there, a merchant gave my mother a little girl as a present. She had
been kidnapped from here, from Attica. # An Athenian citizen? # I believe so. We
don’t know for sure.’ Ter. Eu. 108–11)
(o) Liberem ego te? # Verum, quandoquidem, ere, te servavi.
(‘I should free you? # Yes, since I’ve saved you, my master.’ Pl. Men. 1024)
118 119
For sane, see Risselada (1998a: 234–5). See, for example, Lodge: s.v. etiam § E (p. 544).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(s) ‘Audes,’ inquit, ‘servum dicere Olynthium?’ Etiam, post bellum et ante
decretum.
(‘ “Do you dare,” he says “to call an Olynthian a slave?” Yes, after the war
and before the decree.’ Sen. Con. 10.5.19)
(t) Si non, quo modo tabulas conficis? Si etiam, quam ob rem . . . hoc
nomen . . . relinquebas?
(‘If you have not, how do you make up your books? If you have, why is it
that . . . you left this item?’ Cic. Q. Rosc. 9)
Plautus and Terence also use a few Greek expressions in answer to a preceding question, as in (u).
(u) Ego amo. # An amas? # NÆd ªæ.
(‘I’m in love. # You’re in love? # Yes indeed.’ Pl. Bac. 1162)
In Late Latin, adverbs and particles became the most common form of affirmative
response to a question.120
Rhetorical questions have an assertive illocutionary force (see § 6.23) and can
therefore be used in answers to questions.121 Examples are (a)–(e).
(a) An tu me tristem putas? / # Putem ego, quem videam aeque esse maestum ut
quasi dies si dicta sit?
(‘Do you think I’m depressed? # Do I think so? I can see that you’re as sad as if you
were in for trial.’ Pl. As. 837–8)
(b) Ecquidnam meminit Mnesilochi? # Rogas? / Immo unice unum plurimi
pendit.
(‘Does she remember Mnesilochus at all? # You ask? Yes, she has the highest opinion
of him and him alone.’ Pl. Bac. 206–7)
(c) Egone osculum huic dem? # Quor non, quae ex te nata sit?
(‘I should give her a kiss? # Why not, since she was born from you?’ Pl. Epid. 574)
(d) Quid illam, meretricemne esse censes? # Quippini?
(‘What about her, do you think she’s a prostitute? # Why not?’ Pl. Bac. 839)
(e) Cape, signum nosce. nostin? # Quidni noverim?
(‘Take them, examine the seal. Do you recognize it? # Why shouldn’t I?’
Pl. Cur. 423)
120
See Müller (2012). NÆd ªæ in this example and in Pl. Ps. 484 are conjectures. For critical discussion
and an alternative, see Hausburg (2014).
121
See Kienpointner (1998: 78–80).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Responses
The most common form of a negative reaction to a preceding question is with non.
The constellations in which it is used can be schematized as follows: (i) non + the verb
and possible other constituents of the preceding utterance repeated, as in (a); (ii) non
+ a substitute verb, as in (b); (iii) non without or with (an) asseverative element(s), as
in (c) and (d), respectively. An example of non with a communication or cognition
verb is (e). In Plautus, the third alternative is rare and used by people of a low social
status. In Terence, this alternative is much more common (fifteen times as frequent as
in Plautus) and not socially marked.122
(a) Iam dedit argentum? # Non dedit.
(‘Has he given the money already? # No, he hasn’t.’ Pl. As. 638)
(b) Egon’ mea bona ut dem Bacchidi dono sciens? Non faciam.
(‘Am I to make a gift of my property to Bacchis knowingly? I won’t do it.’ Ter. Hau.
1050–1)
(c) Quid est? / Venitne homo ad te? # Non.
(‘What’s that? Has he come to you? # No.’ Pl. Ps. 1066–7)
(d) Non habes? # Non hercle vero.
(‘You don’t? # No, I don’t.’ Pl. Ps. 326)
(e) An tu ob peccatum hoc esse illum iratum putas? / # Non arbitror.
(‘Do you reckon it’s this misbehaviour of yours that has made him angry? # I don’t
think so.’ Ter. Hau. 990–1)
As for the answer non to a preceding question, it has been suggested that one should
make a distinction between questions which contain a negative element, as in (f )–
(h), and questions without such a leading negative element. It looks as if in Terence,
in other contexts than the ones below, the use of non alone is less common for old
men than for slaves and may be considered less polite.123
122 123
See Müller (1997: 201–3). See Cabrillana (2004).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Nam ubi desponsam nuntiasti filio, / continuo iniecisse verba tibi Dromonem
scilicet, / sponsae vestem aurum atque ancillas opus esse: / argentum ut dares. #
Non. # Quid? Non? # Non inquam. # Neque ipse gnatus? # Nil prorsum, Chreme.
(Ter. Hau. 892–5); Quid est, puella? Ecquid te horum lacrimae movent? ‘Non,’
inquit. ‘Ille ad magistratum veniat.’ (Sen. Con. 7.8.2);
Instead of non, neque . . . neque and minime can be used in negative answers. Haud is
very rare. Examples are (i)–(k).
(i) Obsecro, etiamne hoc negabis te auream pateram mihi / dedisse dono hodie
qua te illi donatum esse dixeras? / # Neque edepol dedi nec dixi.
(‘Please, will you also deny that you gave me the golden bowl as a gift today, the one you’d
said you were presented with there? # I didn’t give it and I didn’t say so.’ Pl. Am. 760–2)
(j) Adibon’? # Minime.
(‘Shall I approach her? # Certainly not.’ Pl. Mil. 1242)
(k) . . . nos nostras aedis postulas comburere? / # Haud postulo.
(‘ . . . do you expect us to burn down our house . . . ? # No, I don’t.’ Pl. Aul. 361–2)
Just as with answers to questions (see § 6.40) facio and fio can be used, as in (c) and (d).
(c) Cape illas scopas. # Capiam. # Hoc egomet, tu hoc convorre. # Ego fecero.
(‘Take that broom. # Yes. # I’ll sweep this, you sweep that. # I’ll do so.’ Pl. St. 351)
(d) Supsequere. # Fiat.
(‘Follow me. # Yes.’ Pl. Mos. 803)
Maxime is also used, as in (e ), but less often than minime (see § 6.45).
(e) Iace, pater, talos, ut porro nos iaciamus. # Maxime.
(‘Throw the dice, father, so that I can take my turn. # I will.’ Pl. As. 904)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Responses
The corrective adverb immo ‘rather’ is used both in reactions to a preceding question
or command and in reactions to a preceding assertion or exclamation. Examples are
(a)–(c) and (d) and (e), respectively. Without a verb is (f ).124
(a) Num non vis me interrogare te? # Immo, si quid vis, roga.
(‘You don’t mind if I ask you something, do you? # Of course not. Ask whatever you
like.’ Pl. Aul. 161)
(b) Ecquid placent? / # Ecquid placeant me rogas? Immo hercle vero perplacent.
(‘Do you like it? # Do I like it, you ask me? Yes, I absolutely love it.’ Pl. Mos. 906–7)
(c) Asserva tu istunc, medice. # Immo ibo domum . . .
(‘You watch over him, doctor. # No, I'll go home . . . ’ Pl. Men. 954)
124
For a discussion of the rare corrective use of immo in Petronius—where its most common function
is that of a coordinator or connector—(see Chapter 19) and in Apuleius, see Rosén (2003). Also Rosén
(2005: 233–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
CHAPTER 7
Locating states of affairs (for this term, see § 2.9) in time can be done by both lexical
and grammatical means. Examples of the former are adverbs like heri ‘yesterday’ and
cras ‘tomorrow’ as well as the other temporal adjuncts discussed in §§ 10.26–10.30;
the morphosemantic category of ‘tense’ constitutes the latter. Their similarity in
function is apparent from the fact that certain adverbs are compatible with certain
tense markers but not others, and vice versa. Heri, to give one obvious example,
cannot be used with a simple future tense form (*heri laudabo ‘I will praise
yesterday’).
The reference point on the timescale can be the communicative situation in which
a conversation is taking place or in which someone is addressing another person or
other people, as in (a), but may also be a point situated in the past, as in (b) and (c), or
in the future, as in (d).
In (a), the present tense forms vale’n (= vales-ne) and valeo refer to the health of the
interlocutors in the very situation in which the conversation is taking place (hereafter
to be called the ‘time of speaking’1); the perfect forms valui and valuisti, on the other
hand, refer to their health at a point in time anterior to that of speaking. In (b), in
turn, the imperfect (uxor) eram refers to the wife’s condition at the moment in the
past indicated by modo and by the temporal cum clause. In this example, the reference
point is established lexically, but it may have to be inferred from the context, or it may
even remain implicit, as something presupposed in someone’s mind. Obviously, in
the latter case we depend completely on our interpretation of the utterance(s)
involved (see § 7.19). In (c), the pluperfect fueram indicates that the speaker’s
freedom preceded his becoming a slave, an event which itself is situated before the
time of speaking, as appears from the perfect form fecit. In (d), finally, the sowing
(in the future perfect tense feceris) is obviously anterior to the harvesting; the latter, in
turn, also occurs in the future, hence the simple future tense form metes.
There are numerous instances of the present and imperfect tense forms that can
satisfactorily be described in terms of simultaneity of a state of affairs with a reference
point in the present or the past.2 Likewise, there are numerous examples where the
perfect, future perfect, and pluperfect forms can be described in terms of anteriority to
the time of speaking, to a moment in the future, and to a moment in the past,
respectively. The question, then, is whether one can describe all uses of these forms in
terms of reference points (time of speaking, past, and future) and relative position
(simultaneity, anteriority, and posteriority).
The term ‘grammatical aspect’, or ‘aspect’ for short, will be used in this Syntax to cover
the notions of ‘perfectivity’ and ‘imperfectivity’, which can be defined as a grammatical
means for presenting a state of affairs either ‘as one complete, indivisible whole’
(perfective) or ‘as being non-complete or in progress’ (imperfective) (Dik 1997:
I.221).3 It must be distinguished from ‘lexical aspect’ or ‘Aktionsart’, which refers to
related semantic properties, such as ‘non-dynamic’, ‘terminative’, and ‘momenta-
neous’. These properties are dealt with in this Syntax in the section on the typology
of states of affairs (see § 2.9).4 In order to illustrate the usefulness of the concept of
(grammatical) aspect for understanding the use of some or all Latin tense forms, one
can look for instances of compatibility or incompatibility between a certain tense and
the lexical type of states of affairs, just as was shown in § 7.2 for grammatical and lexical
temporal location (*heri laudabo). For example, it has been observed that the Latin
imperfect (which is often regarded as imperfective) when used with terminative states
of affairs can be interpreted as meaning ‘intended’ or ‘tried’, as in (a), an instance of the
1
‘Time of speaking’ is also used to cover the time of writing.
2
The future tense is slightly more complicated. See § 7.22.
3
Haverling (2010a) uses the term ‘viewpoint’.
4
Haverling (2010a) uses the term ‘actionality’.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
5
For a detailed discussion, see Haverling (2000: 3–15, 2010a: 284–340).
6
The infix -sc- has other functions as well. See Haverling (2010a: 284).
7
The most extensive study is Haverling (2000).
8
See Haverling (2010a: 321–40).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The paradigm of the Latin indicative verb forms, both active and passive, suggests a
clear division into two classes: forms based upon the ‘infectum’ stem on the one hand,
and forms based upon the ‘perfectum’ stem on the other (see Table 7.1, more details in
§ 3.11). In fact, this was already observed by Varro (116–27 BC) in his De lingua
Latina.9
Nam cum sint verba alia infecta, ut lego et legis, alia perfecta, ut legi et legisti, et
debeant sui cuius<que> generis in coniungendo copulari, et cum recte sit ideo lego ad
legebam, non recte est lego ad legi, quod legi significat quod perfectum, ut haec tutudi,
pupugi, tundo pungo, tundam pungam; item haec † catus † sum, verberatus sum,
<necor verberor, verberor> verberabor, iniuria reprehendunt, quod et infecti inter se
similia sunt et perfecti inter se, ut tundebam tundo tundam et tutuderam tutudi
tutudero; sic amabar amor amabor, et amatus eram, amatus sum, amatus ero. (‘For
since some verbs denote incomplete action, like lego [I read] and legis [you read], and
others denote completed action, like legi [I have read] and legisti [you have read], and
since in the conjoined form they ought to be connected with others of their own kind
and by this principle lego is rightly related to legebam—lego is not rightly connected
to legi, because legi denotes something completed; so that they are wrong in finding
fault with tutudi [I have pounded] and pupugi [I have pricked], tundo and pungo,
tundam and pungam, as well as necatus sum [I have been killed] and verberatus sum
[I have been beaten], necor and verberor, necabor and verberabor, because the tenses of
incomplete action are like one another. Thus we should say tundebam, tundo, tundam,
and tutuderam, tutudi, tutudero, and in the same way amabar, amor, amabor, and
amatus eram, amatus sum, amatus ero.’ Var. L. 10.48)
Table 7.1 Latin 1sg. indicative, active and passive forms of amo
(‘to love’)
Active Passive
infectum perfectum infectum perfectum
amo amavi amor amatus sum
amabam amaveram amabar amatus eram
amabo amavero amabor amatus ero
However, the description of the semantic values of the verb forms concerned has
proven to be difficult from ancient times onward. Varro describes forms like amo,
amabam, amabo as designating ‘praesens’, ‘praeteritum’, and ‘futurum’ (L. 8.20) and
assumes this distinction in all four columns of Table 7.1, which results in a matrix of
what we may call ‘aspectual’ and ‘temporal’ values. Accordingly, whereas he takes the
9
For Varro’s extraordinary treatment of the future perfect forms in -ero, considered a subjunctive by
other ancient grammarians, see Serbat (1978).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
perfect amavi as ‘praesens’, some later grammarians illustrate the temporal distinc-
tions ‘praesens’, ‘praeteritum’, ‘futurum’ with the forms amo, amavi, amabo, thus
taking amavi as ‘praeteritum’, as for example Charisius does (text written AD 362).10
Most ancient grammarians observe that the ‘praeteritum’ is split into three forms,
the ‘praeteritum imperfectum’, the ‘praeteritum perfectum’, and the ‘praeteritum
plusquamperfectum’; thus they assume that there is an aspectual distinction (imper-
fectum versus perfectum) within the ‘praeteritum’, so for instance Priscian (text
written c. AD 500).11 A final illustration of the ambivalent nature of the perfect is
the observation by Diomedes (text written c. AD 375) that the Latin perfect corres-
ponds to the Greek aorist and perfect.12 Scholars have continued to discuss
the position of the perfect in the Latin tense system and the question of whether
this system is essentially a temporal one or an aspectual one, or a mix of the two,
ever since.13
In the case of the imperfect and perfect indicative, both aspectual notions like ‘non-
completed/completed’ and temporal notions like ‘simultaneous/anterior’ are applic-
able. The concomitance of aspectual and temporal notions, especially in the domain
covered by the Latin imperfect and perfect indicative, can be found in other languages
as well.14 However, outside these two forms there is little room for the aspectual
notions, although some tense forms have sometimes been described in aspectual
terms (see, for example, certain uses of the future perfect in § 7.32). The system of
indicative forms is therefore regarded as based essentially on the combination of
relative position (anteriority, simultaneity, and posteriority) and reference point (time
of speaking, past, and future). See Table 7.2 below.
Whereas for the indicative Latin has a threefold distinction in its regular paradigm
(present, past, and future), there is only a twofold distinction (past and non-past) for
the subjunctive.15 This means that in subordinate clauses the present subjunctive
covers the area in which the indicative has a distinction between present and
simple future.
10
Tempora sunt tria, praesens, ut lego, praeteritum, ut legi, futurum, ut legam. (Charisius 214.5B). Cf.
Trifariam tamen cuncta gerimus, ideoque tria tempora esse dicimus, instans perfectum futurum; instans,
quod et praesens, cum adhuc agimus, praeteritum perfectum, cum iam fecerimus, futurum, cum acturos
nos pollicemur. (Diomedes I.335.27–30K).
11
Sed praeteritum rursus dividitur in tria, in praeteritum imperfectum, praeteritum perfectum,
praeteritum plusquamperfectum. (Prisc. 8.38 (II.405.8–10 K)).
12
At vero tempus perfectum apud nos pro Iæfiø ŒÆd ÆæÆŒØ fiø valet. (Diomedes I.336.10K).
Charisius, Diomedes, and Priscian all three worked in Constantinople and wrote for a Greek-speaking
audience.
13
For the view that the system is essentially temporal, see Pinkster (1983). Only a few of the points
discussed there are repeated here. Discussion of other views on this much debated item can be found there
as well. For a critical discussion of that article, see Moralejo (1989), Baños (1999), and Oldsjö (2001).
A survey of the literature can be found in Haverling (2000: 9–13, 2010a: 340–5). For an aspectual
approach to the Latin verb system, see Oldsjö (2001), followed by Haug (2004) and Heberlein (2011:
248–53). For a balanced view of the Latin tense system, see Kravar (1980). See also Spevak (forthc.).
14
For an illustration, see Givón (2001: I.286–300).
15
Forms like profecturus sit that function as substitutes for subjunctives of the future are left out of
account.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In comparison with the indicative forms, the number of participial and infinitival
ones is limited: six indicative versus three participial and infinitival forms. In
indirect speech, the perfect infinitive corresponds to the imperfect, perfect, and
pluperfect indicative. As a result, the distinction between the imperfect (‘simultan-
eous with the past’) and perfect (‘anterior to the time of speaking’) indicative
forms is lost in the perfect infinitive, which only marks anteriority. This is shown
in (b) and (c).
(b) Cicero Romae habitabat / habitavit.
(‘Cicero was living / lived in Rome.’)
(c) Dico Ciceronem Romae habitavisse.
(‘I say that Cicero lived in Rome.’)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The speaker is to some extent free in the choice of tense forms: that is, the same reality
can sometimes be expressed in different ways. In a situation like (a), in which the
imperfect is used, a present form could be used as well, as in (b). In situations where
alternative tense forms are available, the choice of one or the other may be motivated
by considerations that have nothing to do with time: just as in English and Dutch, a
speaker may situate a wish for the present in the past to make it more polite or less
direct. An example is the imperfect form volebam ‘I wanted’ in (c).
(a) Quo nunc ibas?
(‘Where were you going just now?’ Pl. Mer. 884)
(b) Quo nunc is? # Ad te.
(‘Where are you going now? # To you.’ Pl. St. 247)
(c) Sed quid venis? Quid quaeritas? # Demaenetum volebam.
(‘But why have you come? What are you looking for? # I wanted to see Demaenetus.’
Pl. As. 392)
The speaker is also free to some extent to determine what he or she regards as the time
of speaking and which states of affairs he or she wants to present as included within that
time frame. This is shown by the exchange in (d). The first speaker asks whether he should
tell a woman something, using a present tense form; the response (itself a question) uses
the future tense. Both speakers, however, are referring to the same future event. Compare
also examples (e)—present tense—and (f)—perfect tense—in an identical context. (Cicero
uses the present much more often than the perfect, which occurs only here.)
(d) Iamne ego huic dico? # Quid dices?
(‘Shall I tell her now? # What will you tell her?’ Pl. Cur. 132)
(e) Saepe ego doctos homines—Quid dico saepe?
(‘I have heard learned persons often—why am I saying “often”?’ Cic. de Orat. 2.365)
(f) . . . Autronium tum in campo vidimus—Sed quid dixi vidisse nos?
(‘ . . . it was Autronius whom we saw in the Campus Martius on that occasion. Why
did I say “we saw”?’ Cic. Sul. 51)
7.7 Mood
The grammatical category of ‘mood’ is one of the means by which the speaker can
convey his view of the extralinguistic reality and his communicative intentions with
regard to one or more interlocutors. Other means of doing this are lexical expressions
of various types, speech act verbs like declaro ‘to declare’ and rogo ‘to ask’; modal
verbs like possum ‘can’ and debeo ‘to be obliged’; and attitudinal disjuncts like fortasse
‘perhaps’ and certe ‘surely’ (for which see § 10.100). Furthermore, the speaker has a
choice of sentence type (declarative, interrogative, imperative, or exclamatory; see
Chapter 6) and—in part related to that—intonation and gesture.
It is evident that the moods have the semantic function discussed in the preceding
paragraph from the fact that there exist cases of (in)compatibility of (grammatical) mood
with other entities. For example, in the case of the imperative, it is very odd to order a
person to do something when he has no control over it; this explains the very low
frequency of true passive imperative forms. An exceptional instance is the passive
imperative adire in (a) (further examples in § 7.66; see also § 6.30).16 Another example
of incompatibility is the non-occurrence of disjuncts like fortasse ‘perhaps’ in commands.
(a) Ascende (sc. tribunal)! . . . Adire!
(‘Mount it (the tribunal)! . . . Let men come before you!’ Plin. Pan. 60.3)
(b) *Fortasse veni!
As for the subjunctive, on the one hand, it can be used to present a state of affairs as
‘possible/likely’ or as ‘desirable/wanted’, semantic values that resemble the meanings
of the verbs possum and debeo, respectively. This probably explains why it is often not
the subjunctive forms of these verbs that are used, but the indicative. On the other
hand, there are also cases of so-called ‘attraction’,17 where the subjunctive is used
16
Extensive discussion and conjectures; see Schuster-Hanslik’s apparatus criticus ad loc.
17
So Löfstedt (1933: II.129ff.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
instead of the indicative to mark events as less factual or real and therefore the
utterance as less pressing, as in (c) and (d) (details in §§ 7.40–7.46).
(c) . . . id quod haud sciam an tu primus ostenderis . . .
(‘ . . . a fact which I should almost think you have been the first to make plain . . . ’ Cic.
de Orat. 1.255)
(d) Peculiariter contra aconita addi parum salis oporteat.
(‘It is apt for combating aconite poisoning, and there should be added a dash of salt.’
Plin. Nat. 29.103)
It is generally assumed that Latin has three moods: the indicative, the subjunctive, and the
imperative.18 The most problematic one of these three is the subjunctive, not only for the
reasons mentioned in § 7.7, but also for several others. On the basis of example (a) and its
constructed counterpart (b), one would assume that the imperfect and pluperfect
subjunctive forms in (b) are simply the past forms corresponding to the present and
perfect subjunctives in (a). The difference would therefore seem to be temporal.
(a) . . . haud quisquam quaeret qui siem aut quid venerim.
(‘ . . . no one will ask me who I am or what I’ve come for.’ Pl. Am. 130)
(b) . . . haud quisquam quaesivit qui essem aut quid venissem.
18
A more detailed discussion can be found in LSS: ch. 10.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
However, in simple sentences like (c) and (d), the difference is better described in
terms of present ‘possibility’ or ‘likelihood’ versus present ‘counterfactuality’ or ‘pure
hypothesis’ (cf. the traditional terms coniunctivus potentialis versus coniunctivus
irrealis) than as present versus past ‘possibility’ or ‘likelihood’. That is, the difference
seems to be modal.19
(c) Ex homine hunc natum dicas?
(‘Would you say he had a human being for a father?’ Ter. Eu. 460)
(d) Canes venaticos diceres . . .
(‘ . . . you might call them hunting dogs . . . ’ Cic. Ver. 4.31)
Another problematic aspect of the subjunctive is that it can be found, on the one
hand, in both the declarative and the interrogative sentence types (just like indicative
forms), but, on the other hand, also in the imperative sentence type (just like
imperative forms). An example of the latter is (e), which can be compared with (f).
In fact, a form like dicat is given as the third person imperative in the paradigm of this
mood by Latin grammarians.20
(e) Dicat quod quisque volt.
(‘Let anyone say what he wants.’ Pl. Per. 373)
(f) . . . dic me uxorem orare ut exoret illam.
(‘ . . . say that I’m asking my wife to persuade her . . . ’ Pl. Cas. 705)
In the declarative and interrogative sentence types, the difference between the indi-
cative and subjunctive forms can be formulated in terms of ‘epistemic’ modality, with
the indicative presenting an event as ‘factive’, the subjunctive as ‘non-factive’, that is,
just giving it as a ‘possibility’ or ‘likelihood’ (in traditional terminology: it is a
coniunctivus potentialis). In the imperative sentence type, however, the subjunctive
expresses ‘deontic’ modality (it is a coniunctivus volitivus).21
The Latin grammarians not only distinguished a subordinate function (illustrated
in § 7.7) and an imperative function of the subjunctive forms but also recognized an
‘optative’ function, illustrated by expressions like utinam amarem (‘if only I loved’),
which was probably inspired by the Greek grammarians and the Greek language,
in which a separate form for this function existed. Many modern grammarians
distinguish an optative use of the subjunctive as well, often pointing out that ‘in the
Latin subjunctive both with respect to form and with respect to meaning the Indo-
European optativus, coniunctivus and iniunctivus have merged’.22
19
See Fleischman’s section on ‘Conditional as tense or mood’ (1982: 26–8).
20
. . . modo imperativo tempore praesenti numero singulari ad secundam et tertiam personam ama
amet et pluraliter amemus amate ament. (Sacerdos VI.436.3–4K. (3rd century AD)).
21
For the terms ‘epistemic’ and ‘deontic’, see Palmer (2001).
22
Sz.: 329 (my translation). For a discussion of the prehistory of the Latin subjunctive, see Magni
(2010: 204–6) with references. For recent literature on the ‘injunctive’, see Calboli (2011: 36–42).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
It is difficult to find a common denominator for the range of uses of the subjunctive
described above. In addition, this difficulty is increased by the fact that, in the period
covered by this Syntax, there is a change in the values of some of the forms involved.
Table 7.4 shows a threefold distinction for main clauses; it also shows which moods
are found in which sentence types.23 Sentence type is discussed in Chapter 6, the use
of the negators non and ne in Chapter 8.
Table 7.4 The uses of the indicative and subjunctive moods in independent clauses
Mood
Factive Non-factive
Indicative Subjunctive Subjunctive
Sentence type factual modality non-factual modality counterfactual modality
Declarative (non) + + +
Interrogative (non) + + +
Imperative (ne) – + +
+ means: does occur; – : does not occur
23
See LSS: § 10.3.
24
Cf. Orlandini (1980: 121ff., 132ff.). Examples in N.-W.: III.169–72.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Many scholars have tried to find a common denominator for the use of the subjunct-
ive with a clear semantic contribution to the meaning of the clause in which it occurs
on the one hand and its use as a subordinating device on the other.25 Such attempts
tend to produce very abstract terminology that is of little use for understanding what
the subjunctive contributes to the meaning of the clauses and sentences in which it
appears. The subjunctive mood is not the only phenomenon in Latin that shows
semantic alongside purely grammatical use: cases and prepositions are other well-
known instances (see Chapter 12). For the hearer/reader, the context in which a
subjunctive form occurs will normally provide sufficient information to determine
whether the subjunctive form must be interpreted as semantic or as purely gram-
matical. It is likely, for example, that a Roman did not feel the need to discover the
semantic contribution of a subjunctive form in an indirect question like the one in (a).
In a context where a governing cognition verb (scis ‘you know’) is used in combin-
ation with an indirect question introduced by the question word (ubi ‘where?’), there
simply was no element of choice involved. The same applies to (b), in which the
‘order’ predicate iubet is used in combination with the subordinator ut.
(a) Eho, Parmeno mi, nostin’ et scis ubi siet?
(‘Hey, my dear Parmeno, you know her? You know where she is?’ Ter. Eu. 351)
(b) Telobois iubet sententiam ut dicant suam.
(‘He ordered them to tell the Teloboians his decision.’ Pl. Am. 205)
It is very plausible that grammatical uses of a device originated from semantic ones,
but synchronically there is no common denominator. Further details about the uses
of the subjunctive in subordinate clauses will be presented in §§ 7.128–7.163.
25
A recent attempt is Müller-Wetzel (2001). For a survey of earlier studies, see Calboli (2011: 134–42).
For the methodological aspect of dealing with grammatical and semantic uses of the subjunctive, see also
H. B. Rosén (1980).
26
29 February 2004.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Before one begins a discussion of the semantic values of the individual finite verb
forms, it is important to get an idea of the frequency with which these forms are
actually used. Table 7.5 shows the distribution of finite indicative and subjunctive
forms in a comedy of Plautus.27
Table 7.5 Distribution of the tenses and moods in Pl. Am. (in absolute numbers)
Indicative Subjunctive
Active Deponent Passive Active Deponent Passive
Present 841 44 22 248 7 13
Imperfect 17 1 1 32 1 –
Simple future 158 24 5
Perfect 306 26 80 31 3 8
Pluperfect 13 – – 13 – –
Future perfect 23 1 –
‘profectus fui/fueram’ – 3 – – – –
Totals 1358 99 108 324 11 21
Total number of verb forms: 2,889, including 305 infinitives, 114 imperatives, and 107 participles
It is immediately obvious that, on the whole, indicative forms are much more frequent
than subjunctive forms, although this is not the case with the imperfect and the
pluperfect tenses. Secondly, it is also clear that the three tenses that are most appro-
priate to this type of text, which contains monologues and dialogues taking place ‘in real
time’, are immensely more frequent than the other three. These three central tenses are
the present, the perfect (indicating events that are anterior to what is actually taking
place on stage), and the simple future (referring to events that are going to happen at
some later point). Apparently, there is not much room for describing ongoing events in
the past (typically the domain of the imperfect) or for referring to events that either
occurred before other events in the past (pluperfect) or will occur before other future
events will take place (future perfect). It is also interesting to note that in the passive the
perfect is much more frequent than the present. Apparently, it is more common to
describe entities as being involved in a state rather than in a process.
The actual distribution of tenses is evidently dependent on the type of text, as can
be seen in Figure 7.1, which presents data from Seneca’s tragedies and philosophical
works (letters and treatises) and Caesar’s account of the Gallic War. From a meth-
odological point of view, in order to get an accurate picture of how the tense and
mood system functioned, one must try to start from texts that reflect communicative
27
Figures taken from the website of LASLA (29 February 2004). The figure for the future perfect
includes six sigmatic forms. The future perfect forms constitute 0.82% of all verbal forms.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
40
35
30
25
20
15
10
0
Philos. Philos. Trag. Trag. Caesar Caesar
indic. subj. indic. subj. indic. subj.
present imperfect simple future perfect
pluperfect future perfect usus fui
Figure 7.1 Distribution of the indicative and subjunctive verb forms in Seneca the
Younger (philosophical works and tragedies) and in Caesar Gal. (in percentages of
the total number of verb forms)
Data for Seneca taken from Delatte et al. (eds) (1981) and Denooz (1980), and for Caesar Gal. from the LASLA
website (29 February 2004).
interaction (plays, letters, orations, dialogues), and it is from this perspective that the
data in Table 7.5 are most illuminating. The descriptions one finds in grammars, by
contrast, are often biased towards narrative works.
Figure 7.1 shows a number of interesting things. In the first place, the expository
texts (Seneca’s philosophical works) have a slightly higher number of imperfect,
pluperfect, and future perfect forms than the tragedies, which results in a higher
proportion of perfects in the latter; this is similar to what was observed in Table 7.5.
But the two types of text are similar in the preponderance of present, perfect,
and future forms. At first sight, Caesar’s historical account shows a high proportion
of present tense forms (of course, mostly historic presents) and a low proportion
of perfects. However, if the data in the graph is combined with the information
in Table 7.6, this becomes understandable: the graph only shows the distribution of
59.4% of the verb forms in Caesar, a much lower percentage than that for the texts of
Seneca. The other verb forms in Caesar, which comprise 40.6% of the total number,
are to a great extent perfect participles (15.6% of all verb forms, almost as many as the
perfect and pluperfect indicatives combined). These participles do the same job as
perfects and pluperfects do: they mark the succession of events in the past. Conse-
quently, the relatively high frequency of present tense forms is illusory. What does
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
still stand out is the high proportions of the imperfect and—to a lesser extent—
pluperfect forms in Caesar.
Not only are subjunctive forms much less frequent than indicative ones, they are
also unevenly distributed over main and subordinate clauses. The indicative is for the
most part a main clause mood, the subjunctive a subordinate clause mood: of the
indicative forms in Plautus Am., 72% occur in main clauses, in contrast to only 18% of
the subjunctive forms. In Caesar Gal., in turn, only 12 out of 1,928 subjunctives occur
in main clauses, less than 1%.
However, these general observations about the moods can be refined for particular
combinations of tense and mood. This becomes apparent from Tables 7.728 and 7.8.29
It is interesting to note that the pluperfect and future perfect forms typically occur
in subordinate clauses (which explains the low percentages in Tables 7.7 and 7.8).
This may best be explained on the basis of their semantic values: expressing some-
thing that is anterior to a reference point in the past and future, respectively. Further,
28
Data taken from the LASLA website (29 February 2004).
29
Data taken from the LASLA website (29 February 2004).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
in Plautus the perfect turns out to be much less a main clause tense than the forms of
the infectum stem, again due to its anterior meaning. Finally, there is a comparatively
high percentage of perfect forms in main clauses in Caesar, matched by a relatively
low percentage of imperfect forms.
Tense forms are unevenly distributed if one examines which verbs are found in
which tenses; this has already been alluded to in § 7.3 and is shown below in Table 7.9
for a number of verbs in Caesar.30
Table 7.9 Distribution of indicative tense forms of sum, habeo, iubeo, and proficiscor
in Caesar (in percentages)
Pluperfect Imperfect Narr. perfect Narr. present
Verb Main Sub Main Sub Main Sub Main Sub Total
sum 0 6 45 31 15 2 – 2 330
habeo 1 9 21 54 9 2 4 – 111
iubeo 1 4 3 0 38 – 54 – 163
proficiscor 0 5 2 0 20 – 71 – 56
Total – 6 27 24 20 1 22 – 660
‘Main’ = main clause; ‘Sub’ = subordinate clause; ‘Narr.’ = narrative
The uneven distribution is a side effect of the roles these verbs play in the texts in which
they occur: action verbs like iubeo ‘to order’ and proficiscor ‘to leave’ are typically used
to narrate successive events on the main storyline, whereas stative verbs like sum ‘to be’
and habeo ‘to have’ are typically used in describing background information (see
§ 7.20). This is further confirmed by the data in Table 7.10, where it is shown that
there are apparently also verbs which Caesar prefers to use in the historic present.31
30
The table is the same as in Oldsjö (2001: 403), but here in percentages.
31
Reinhold (1956: 24) has data on these and other verbs.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
that the present tense can be substituted for all the other tense forms and that it has no
distinctive features of its own. But this is manifestly not the case.32
In using the (present) indicative form, the speaker/writer asserts the content of the
sentence, i.e. he presents it as factual or real.
Another illustration of the actual use of the present is the eyewitness report in the
Rudens of Plautus, (e). Sceparnio tells his master Daemones what he sees.
(e) S. Quod facinus video! # D. Quid vides? # S. Mulierculas
video sedentis in scapha solas duas.
Ut a fflic tantur miserae. Eugae eugae, perbene,
ab saxo avor tit fluctus ad litus scapham,
neque gubernator umquam potuit tam bene.
Non vidisse undas me maiores censeo.
Salvae sunt, si illos fluctus devitaverint.
Nunc, nunc periclum est . <Unda> e ie ci t alteram.
At in vado est , iam facile en abit . Eugepae,
32
See Pinkster (1998a: 63–6), with references. This approach is similar to that in K.-St.: I.114–15. See
also Leiwo (2010: 287–91).
33
For performative expressions in Latin, see Anscombre and Pierrot (1984, 1985) and Desbordes
(1984).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In Sceparnio’s account of the events proper (in s p aced tex t), there are seven present
tense forms (in bold) and six perfect tense forms (in italics).34 It is interesting to note
that, in this eyewitness report, the perfects are used for terminative momentaneous
states of affairs, which do not easily lend themselves to immediate reporting, whereas
the presents are used for non-momentaneous states of affairs. Statistically, it is also
noteworthy that in this eyewitness report there are more perfect tense forms than in
pseudo eyewitness accounts in the historic present (see § 7.16 and Figure 7.2).
Supplement:
Sed Philocomasium hicine etiam nunc est? # Quom exibam, hic erat. (Pl. Mil. 181);
Sponde’n ergo tuam gnatam uxorem mihi? / # Spondeo et mille auri Philippum
dotis. (Pl. Trin. 1157–8); Virum bonum eccum Parmenonem incedere / video: vide
ut otiosus it! (Ter. Eu. 918–19); Promitto hoc vobis et rei publicae spondeo, si modo
nos ipsi rei publicae satis fecimus, nunquam hunc a nostris rationibus seiunctum
fore. (Cic. Cael. 77); A. Trebonio, qui in tua provincia magna negotia et ampla et
expedita habet, multos annos utor valde familiariter. (Cic. Fam. 1.3.1); Quid meos
retines sinus / manusque matris cassa praesidia occupas? (Sen. Tro. 792–3);
It is necessary to distinguish such examples of the ‘actual’ use from the following ones,
which are sometimes called ‘timeless’ and which refer to events that are not only
pertinent to the time of speaking but are also valid in general. Exx. (f) and (g) are
proverb-like general truths; (h) is a description of a physical appearance; (i) a
definition; and (j) a geographical clarification in a narrative. In (k), finally, the present
tense is used because the existence of the text referred to is viewed as timeless.
34
Avortit in l. 165 and concidit in l. 174 are taken as perfect tense forms.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Events that a speaker asserts are going on at the time of speaking may have been going
on for some time even before then. Examples of this use of the present indicative are
(l) and (m), in which time adjuncts and the adverb iam ‘already’, neither of which is a
necessary addition, facilitate the interpretation.
35
For this very exceptional case, see Pinkster (1999: 714–15).
36
See Wisse (1996).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
7.15 The present tense used to refer to future events (praesens pro futuro)
As in many languages, the Latin present tense is used for planned events and firm
intentions or expectations that, strictly speaking, belong to the future.38 Instances of
this use of the present are often first person (singular). Although they occur in all
periods of Latin and in all types of text, they are found especially in the comedies of
Plautus and Terence (though not even there in large numbers). It is found both in
declarative and in interrogative sentences, as in (a)–(d), and (e)–(g), respectively. In
37
For Early Latin examples, see Bennett (1910: I.17–18).
38
Bybee et al. (1994: 249–51) mention a few other languages in which the present tense is used as an
‘expected’ or ‘scheduled’ future.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a), the first speaker inquires about the plans of the second, using the future tense; the
second speaker then sets out his plans in the present tense. The present tense is
normal when someone on stage tells another person that he will be back soon (c).
However, when a character tells the audience his intentions in an aside, the future
tense is used.39 Ex. (d) exemplifies a stubborn refusal. Exx. (e) and (f) illustrate the use
of the present indicative in requests for advice addressed to one or more persons
present. This use is different from the so-called ‘deliberative’ use of the present
subjunctive, for which see § 7.42. Ex. (g) is an invitation in the form of a question.
(a) Ibisne ad cenam foras? # Apud fratrem ceno in proxumo.
(‘Will you go out for dinner? # I’m dining at my brother’s next door.’ Pl. St. 612)
(b) Lentulus Spinther hodie apud me; cras mane vadit.
(‘Lentulus Spinther is staying with me today; he is leaving early tomorrow morning.’
Cic. Att. 14.11.2)
(c) Mane: / Iam redeo ad te.
(‘Wait; I’ll return to you in a second.’ Pl. Ps. 1156–7)
(d) Asside hic, pater. / # Non sedeo istic, vos sedete.
(‘Do sit down here, father. # I won’t sit there; you two should sit there.’ Pl. St. 92–3)
(e) Quid ago? # Quid agas? Rogitas etiam?
(‘What am I to do? # What are you to do? You even ask?’ Pl. Bac. 1196)
(f) Byrria, / quid tibi videtur? Adeon’ ad eum?
(‘Byrria, what do you think? Shall I approach him?’ Ter. An. 314)
(g) Imusne sessum?
(‘Shall we sit down?’ Cic. de Orat. 3.17)
Supplement:
Declarative sentences: Ego ad anum recurro rursum. (Pl. Cist. 594); Nugas! Num-
quam edepol dabit. / # Dat profecto. (Pl. Mos. 1088–9); Et is hodie apud me cenat et
frater meus. (Pl. St. 415); Ad patrem ibo, ut matris iram sibi esse sedatam sciat. / Iam
redeo. (Pl. Mer. 962–3); Quam mox mi operam das? # Tibi do equidem. (Pl. Ps.
1166); Ego hos conveniam [v.l. convenio]. Post huc redeo. (Ter. Ad. 757); Deinde me
expedio ad Drusum, inde ad Scaurum. (Cic. Att. 4.15.9); ‘Tuemini’ inquit ‘castra et
defendite diligenter siquid durius acciderit. Ego reliquas portas circumeo et castrorum
praesidia confirmo’. (Caes. Civ. 3.94.5); Sequor et qua ducitis adsum. (Verg. A. 2.701);
Mox illos sua fata manent maiore sub hoste. (Verg. A. 10.438); Adimis etiam Hispa-
nias, et, inde <si de>cessero, in Africam transcendes. (Liv. 21.44.7); Exi, defero lam-
nam. (Petr. 58.8 (freedman speaking)); SPEC<T>EMUS ILLUM / DUM VENIT ET VEN<I>O
TEQUM ALEXANDRIE ET DEDUCO TE / USQUE AD NAVE. (CEL 146.24–6 (Karanis, c. AD 115));
39
The best discussion of the difference between the uses of the future and the present is still Sjögren
(1906). See also Bennett (1914: I.18–22). For some comments on Sjögren, see Pinkster (1998a: 66–8). For
criticism by Roman grammarians of the use of the present for future events, see Adams (2013: 666–72).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Interrogative sentences: An opsono amplius? (Pl. Men. 320); Voco huc hominem?
# I, voca. (Pl. Mos. 774); Quam mox te huc recipis? (Pl. Truc. 207); . . . m<eae>
nutricem gnatae video. . . . Quid ago? / . . . Adeo, maneo dum haec quae loquitur
mage cognosco? . . . Ea’st ipsa. / Conloquar. (Ter. Ph. 736–9); Sed quid agimus?
Oblitine sumus hoc nunc nobis esse propositum . . . (Cic. Tusc. 1.81); Etsi quid mi
auctor es? Advolone an maneo? (Cic. Att. 13.40.2); Quem sequimur? Quove ire
iubes? Ubi ponere sedes? (Verg. A. 3.88);
The present indicative is also used in the apodosis of conditional periods. The protasis
may contain a present indicative as well, as in (h), but a future indicative (i)—
especially a future perfect (j)—is more common.
(h) Petito cras. # Abeo. Sat habeo, si cras fero.
(‘Ask for it tomorrow. # I’m going away; I’m satisfied if I get it tomorrow.’ Pl. Mos.
654)
(i) Si quid erit, magnum est adiumentum.
(‘If anything comes of it, it will be a great help.’ Cic. Att. 12.29.2)
(j) Salvae sunt, si illos fluctus devitaverint.
(‘They’re safe if they keep clear of those waves!’ Pl. Rud. 168)
Supplement:
Quin te quoque ipsum facio hau magni, si hoc patro. (Pl. As. 114); Si affers, tum (sc.
ianuae) patent. (Pl. As. 242); Quod si adsequor, supero Crassum divitiis atque
omnium vicos et prata contemno. (Cic. Att. 1.4.3); Hos si summoves, nihil metuo
ab externo. (Curt. 6.8.9);
Quod scribo, cum absolvero. (Cic. Att. 15.21.2); Nisi vinctos somno velut pecudes
trucidandos tradidero, non recuso eundem Ardeae rerum mearum exitum quem
Romae habui. (Liv. 5.44.7); . . . si libenter habes, quaecumque loca sunt hic grata ad
videndum Christianis ostendimus tibi. (Pereg. 19.5);40
40
‘Le présent se substitue assez souvent au futur, et, ce qui est notable, toujours en discours direct’.
(Väänänen 1987: 59).
41
See Pitkäranta (1978: 71). It is rare in Ammianus (Kroon and Rose 1996). For its relatively rare use
in Merovingian saints’ lives—where, however, it is more frequent in more elaborate narratives—see
Pinkster (1998b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) (Lycus) Quis hic est? # (Advocati) Nescimus nos quidem istum qui siet.
Nisi dudum mane ut ad portum processimus,
atque istum e navi exeuntem oneraria
videmus. Adiit ad nos extemplo exiens.
Salutat. Respondemus. # (Collybiscus) Mortalis malos,
ut ingrediuntur docte in sycophantiam. #
L. Quid deinde? # A. Sermonem ibi nobiscum copulat.
Ait se peregrinum esse huius ignarum oppidi.
Locum sibi velle liberum praeberier,
ubi nequam faciat. Nos hominem ad te adduximus.
Tu si te di amant, agere tuam rem occasio est.
(‘L. Who is he? # A. We actually don’t know who he is; but when we went towards
the harbour a while ago, early, we saw him leave a cargo ship. On leaving he
approached us at once. He greets us, we reply. # C. What sly mortals! How cleverly
they’re embarking on the trick! # L. What next? # A. He starts a conversation with us
there. He says he’s a stranger and doesn't know this city; he'd like to be given a free
space where he can do naughty things. We’ve brought him to you. If the gods love
you, you have the opportunity to do what’s to your advantage.’ Pl. Poen. 649–59)
This example illustrates some of the characteristic features of texts in which the historic
present is at home. In this particular situation the first speaker, Lycus, asks about the
identity of an unknown person, and (part of) the explanation of the stranger’s identity
may be expected to be historical, as it turns out to be. The Advocates start the historical
part of their answer off by situating it in the past (dudum, mane, ut, and processimus),
then use the particle atque to signal the change to something unexpected (see
Chapter 19). Line 653 shows the historic present in two very short sentences. After
Lycus’ question quid deinde?, which shows his eagerness, the Advocates again use a
situating adverb ibi. Ait is the first word of the next sentence, a favourite place for historic
presents. The end of the story is marked off with a perfect form (adduximus). Most of the
events in the historic present and in the perfect are ‘dynamic’ (see § 2.9); the sentences
follow each other asyndetically. Contiguity of ‘actual’ and ‘historic’ presents is carefully
avoided. All these elements are characteristic of the ‘diegetic’ narrative mode, which is
discussed in Chapter 24.42 The historic present and the features mentioned create the
impression of an eyewitness report and are especially appropriate to texts characterized
by a lot of detail. Accordingly, it comes as no surprise that the Periochae of Livy contain
no historic presents,43 while the illustration given in the Rhetorica ad Herennium (4.68)
of demonstratio (also called evidentia) is a story in which the main criminal events are
historic present forms.44 As a ‘highlighting’ device, the historic present is most common
in main clauses; however, it is also found in subordinate ones (for details see § 7.121).
42
For narrative modes, see Kroon (2002).
43
See also Oldsjö’s (2001: 281–2) treatment of the relationship between the use of the historic present
and ‘narrative pace’.
44
Demonstratio est, cum ita verbis res exprimitur ut geri negotium et res ante oculos esse vide<a>tur.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
speaking)); Dum Eumolpus cum Bargate in secreto loquitur, intrat stabulum praeco
cum servo publico aliaque sane <non> modica frequentia, facemque fumosam magis
quam lucidam quassans haec proclamavit: . . . (Petr. 97.1—NB: coordination of pre-
sent and perfect tenses); Instituuntur ab isto in provincia epulae et magnifico
apparatu exstruitur convivium; distinguuntur argenteis poculis aurea. Quid multa,
iudices? Convivii eius apparatum sensit provincia. Extrahitur quidam e carcere in
convivium praetoris, cui stupenti misero meretrix arridet. Interim virgae promuntur,
et victuma crudelitatis ante mensam ac deos trucidatur. Me miserum! Imperi Romani
terrore lusisti. (Sen. Con. 9.2.6); DICO ILLI, DA MI, DI<C>O, A<E>S PAUCUM. IBO, DICO, AD
AMICOS / PATRIS MEI. (CEL 146.10–11 (Karanis, c. AD 115)); DE<I>NDE POS<T> PAU / COS
DIES PARIT, ET NON POTERAT MIHI SUCCURRERE. (CEL 146.19–20 (Karanis, c.115)); In ipso
spatio paucorum dierum baptizati sumus, et mihi Spiritus dictavit non aliud peten-
dum ab aqua nisi sufferentiam carnis. Post paucos dies recipimur in carcerem; et
expavi, quia numquam experta eram tales tenebras. (Passio Perp. 3.5); Dehinc in illo
maerore ex semetipsa sola, nulla opera coniugii, concepit et procreat feminam. (Tert.
adv. Val. 10.1) (NB: parallelism of perfect and present tense); Tunc deinde cogitatur
quid fieret . . . Mittunt legatos . . . , etc. (Vict. Vit. 1.37);
Ex. (a) shows that, in order for a hearer/reader to be able to interpret a present tense
as ‘historic’, there must be a sufficient number of ‘past signals’ in the context or in the
communicative situation. Narratives rarely start or end with (a) historic present(s).
Rare instances are (b), a disputable case in Terence of a slave entering the stage, and
(c), the very end of Virgil’s Aeneid.
(b) Omnem rem modo seni / quo pacto haberet enarramus ordine (enarravi cj. Iov.)
(‘We’ve just told the old man the whole story from start to finish.’ Ter. Ad. 364–5)
(c) Ast illi solvuntur frigore membra / vitaque cum gemitu fugit indignata sub
umbras.
(‘His limbs grew slack and chill and with a moan his life fled resentfully to the shades
below.’ Verg. A. 12.951–2)
Supplement:
Start of a section: Prima novo principatu mors Iunii Silani, proconsulis Asiae, ignaro
Nerone per dolum Agrippinae paratur . . . (Tac. Ann. 13.1.1);45
End of a narrative episode: Cum haec diceret, venit ad nos libertus aeditumi flens et
rogat ut ignoscamus quod simus retenti et ut ei in funus postridie prodeamus.
Omnes consurgimus ac simul exclamamus ‘Quid? In funus? Quod funus?
Quid est factum?’ Ille flens narrat [follows indirect speech]. Non moleste ferentes
descendimus de aede et de casu humano magis querentes quam admirantes id
Romae factum discedimus omnes. (Var. R. 1.69.2–3); Ceterum, sive illud domesti-
cum sive publicum fuit iudicium, damnatur Servio Cornelio Q. Fabio consulibus.
(Liv. 2.41.7);
45
Cited by Sznajder (1996: 310).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
That the historic present indeed functions as a past narrative tense appears from the
fact that anterior events are often expressed in the pluperfect tense, as in (d) and (e),
and that simultaneous events are often expressed in the imperfect tense, as in (e). In
other types of subordinate clauses past tenses are also quite common, as in the
argument clause with imperat in (f). However, different authors show a great deal
of variation.
(d) Consuli primo tam novae rei ac subitae admiratio incluserat vocem. Dicere
deinde incipientem primores civitatis circumsistunt, eadem multis precibus
orant.
(‘The consul was at first prevented from uttering a word by his astonishment at this
strange and unexpected turn; then, when he tried to speak, the chief men of the state
surrounded him and with many entreaties made the same request.’ Liv. 2.2.8)
(e) Interea ea legione quam secum habebat militibusque qui ex provincia
convenerant . . . fossam . . . perducit.
(‘In the meanwhile he used the legion which he had with him and the troops which
had concentrated from the province, to construct a trench . . . ’ Caes. Gal. 1.8.1)
(f) Et simul servis suis Rubrius ut ianuam clauderent et ipsi ad fores assisterent
imperat.
(‘At the same moment Rubrius told his slaves to shut the front door and stand on
guard at the entrance.’ Cic. Ver. 1.66)
Supplement:
Ille avidus pugnae suras incluserat auro / hinc atque hinc oditque moras hastamque
coruscat. (Verg. A. 12.430–1);
The historic present is the narrative tense par excellence in Latin stories, as
becomes clear from Figure 7.2. Its frequency has been interpreted as proof of the
‘colourlessness’ of the present indicative tense (and of its ‘unmarkedness’; see § 7.13
and the introduction to this chapter), but this is unwarranted.46
Quintilian describes the historic present as a metaphor: ‘Transferuntur et tempora:
“Timarchides negat esse ei periculum a securi” (praesens enim pro praeterito posi-
tum est) . . . ’ (Inst. 9.3.11). For the ancient idea that this use of the present is typically
a stylistic feature of historians see Koller (1951: 63–4). It is interesting that there is no
historic present in the brief narrative given as an example of the simple style in Rhet.
Her. 4.14. The other side of the coin is that the perfect is not used very much in
storytelling, and this might be more interesting to explore. See § 7.30, after ex. (ad) +
Supplement.
46
Most recently and most extensively Oldsjö (2001) (see Adam’s 2002 review). My own views can be
found in Pinkster (1998a, 1999). The best publications on the historic present are Schlicher (1931) and
Kravar (1969). For quantitative data concerning the early Roman historians, see Briscoe in Cornell
(2013: I.35) .
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
90
80
70
60
50
40
30
20
10
0
Act. Pr. Hist. Pr. Hist. Pf. Auth. Impf. Pqpf. Hist.
Pf. Infin.
Plautus Cicero Caesar Hirtius
B. Afr. Vergilius Catullus
Figure 7.2 The use of verb forms in independent sentences and main clauses in seven
narrative passages (in %)
Pl. Am. 201–62 N = 45
Cic. Ver. 5.86–110 N = 139a
Caes. Gal. 1.7–19 N = 96
Hirtius Gal. 8.1–17 N = 90
B. Afr. 1–11 N = 91
Verg. A. 4.1–40 N = 138b
Cat. 63; 64.1–42 N = 45
N = Number of finite indicative verb forms in simple sentences and main clauses in a continuous narrative
passage. Needless to say, the figures include interpretative decisions on my part (distinguishing homonymous
present and perfect forms, and historic and so-called authorial (see § 7.30) perfects). The number of passages is
limited. However, this does not detract from the general conclusion that the historic present is the main narrative
tense in sustained stories in Latin.
a
The figures for Cicero are taken from Pinkster (1998a: 74), where no distinction is made between narrative and
authorial perfects.
b
The figures for Virgil are taken from Pinkster (1999). See also Adema (2008).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Speakers/writers were to some extent free to choose whether or not to use the historic
present. Whereas some authors reserve it for a ‘narrative peak’47or a ‘climax’ and the
build-up to it48 in an episode, as in (g), others use it for almost entire episodes. The
account of the battle in Pl. Am. 201–62—see Figure 7.2—is a case in point: out of
forty-five verb forms in simple sentences and main clauses, there are thirty-seven
historic presents, that is 82%. In epic narrative, the use of the historic present
throughout an entire episode is comparable; Verg. A. 4.54–90, too long to be quoted
in full, is a good example.49 Metrical considerations may have favoured the use of the
historic present as well. Other instances of this type of concentrated use of the historic
present are the narrationes in Cic. Mil. 24–9 and Cic. S. Rosc. 15–29. Different again is
(h) from Livy; here an isolated historic present (pugnant) occurs immediately after a
speech by the commander and then is immediately followed by a concluding state-
ment by Livy in the perfect (fefellit).50
(g) Non videtur esse praetermittendum de virtute militis veterani V legionis. Nam
cum in sinistro cornu elephans vulnere ictus et dolore concitatus in lixam
inermem impetum fecisset eumque sub pede subditum dein genu innixus
pondere suo proboscide erecta vibrantique stridore maximo premeret atque
enecaret, miles hic non potuit pati quin se armatus bestiae offerret. Quem
postquam elephans ad se telo infesto venire animadvertit, relicto cadavere
militem proboscide circumdat atque in sublime extollit. Armatus qui in
eiusmodi periculo constanter agendum sibi videret, gladio proboscidem qua
erat circumdatus, caedere quantum viribus poterat non destitit.
(‘I ought not, I think, to omit to mention the gallantry of a veteran soldier of the Fifth
legion. On the left wing an elephant, maddened by the pain of a wound it had received,
had attacked an unarmed camp follower, pinned him underfoot, and then knelt upon
him; and now, with its trunk erect and swaying, and trumpeting loudly, it was crushing
him to death with its weight. This was more than the soldier could bear; he could not
but confront the beast, fully armed as he was. When it observed him coming towards it
with weapon poised to strike, the elephant abandoned the corpse, encircled the soldier
with its trunk, and lifted him up in the air. The soldier, perceiving that a dangerous
crisis of this sort demanded resolute action on his part, hewed with his sword again and
again at the encircling trunk with all the strength he could muster.’ B. Afr. 84.1–3)
(h) [speech by the dictator about the strategy to follow] Sic eques, sic pedes, ut
praeceperat (sc. dictator), pugnant; nec dux legiones, nec fortuna fefellit ducem.
(‘So horse and foot fought exactly as they had been ordered; the general failed not his
legions, nor fortune the general.’ Liv. 6.12.11)
47 48
The term is taken from Longacre (1983). See Chafe (1994: 207–10).
49
Pinkster (1999) shows that there is a difference in the use of the historic presents in the narrative
narrated by Virgil and the narrative narrated by Aeneas. The phenomena described are more intense in
Aeneas’ story than in Virgil’s story.
50
For Ovid’s exploitation of the potential ambiguity of present forms (actual or historic), see Kroon
(2007: 78–90).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Even within one work the balance between the historic present and the perfect may
vary: Caesar, for example, uses 152 perfect forms in his story about Ariovistus, and
only five present forms (Gal. 1.30–54), but, conversely, in his story about Ambiorix
(Gal. 5.24–58), 181 present forms against sixty-five perfect forms.51
There is a long-standing discussion about the relationship between the historic
present and the other narrative tenses, with many scholars arguing that the historic
present equals the (narrative) perfect.52 This is not correct. The historic present is
indeed most often, as in the examples cited so far, used for successive—dynamic,
often terminative—states of affairs in an episode, thus resembling the narrative use
of the perfect (see § 7.30), but it may well be used for non-successive—including
non-dynamic—states of affairs, thus resembling the use of the imperfect (see § 7.20).
This is quite common in (epic) poetry, starting as early as Ennius—for example in his
description of Romulus and Remus observing the birds, as in (i)—but it is not restricted
to this genre. In (i), the historic present expectant is continued by the imperfect
expectabat in the author’s comment on the situation. Ex. (j) describes Dido’s state of
mind and what she is continually doing as a result. Ex. (k) is part of the description of
the sculptures on the temple in Carthage. Ex. (l) is part of a battle scene.
(i) Omnibus cura viris uter esset induperator. / Expectant veluti, consul quom
mittere signum / volt, omnes avidi spectant ad carceris oras, / quam mox
emittat pictos e faucibus currus: / sic expectabat populus atque ore timebat /
rebus, utri magni victoria sit data regni.
(‘Anxiety filled the men as to which of the two should be ruler. They waited just as,
when the consul means to give the signal, all men look eagerly at the barrier’s bounds
to see how soon he will send the painted chariots from the mouths: thus the people
were waiting, and they showed their apprehension for the future in their faces,
wondering to which of the two the victory of right royal rule should be given by
the event.’ Enn. Ann. 83–8V=78–83S)53
(j) Est mollis flamma medullas / interea et tacitum vivit sub pectore vulnus. /
Uritur infelix Dido totaque vagatur / urbe . . .
(‘All the while the flame devours her tender heartstrings, and deep in her breast lives the
silent wound. Unhappy Dido burns, and through the city wanders . . . ’ Verg. A. 4.66–9)
(k) (Troilus) fertur equis curruque haeret resupinus inani / lora tenens tamen.
Huic cervixque comaeque trahuntur / per terram et versa pulvīs54 inscribitur
hasta.
(‘(Troilus) is carried along by his horses, and fallen backwards, clings to the empty
car, still clasping the reins; his neck and hair are dragged over the ground, and the
dust is scored by his reverse spear.’ Verg. A. 1.476–8)
51
See Schlicher (1931).
52
Koller (1951) sees a connection between the use of the historic present and the meaning of the verbs:
in his view, ‘inceptive’ verbs in particular are found in the historic present.
53
The text and interpretation follow Skutsch ad loc. 54
On the long ī, see Austin ad loc.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(l) Totis trepidatur castris, atque alius ex alio causam tumultus quaerit.
(‘There was confusion throughout the camp, and one sought from another the cause
of the uproar.’ Caes. Gal. 6.37.6)
Supplement: Nec est quicquam quod ibi non est. (Apul. Met. 5.2.2);
However, there are also cases where the historic use of the present is excluded, so for
example in (m), where the parenthetical background information can only be in the
imperfect, and likewise in (n), with an authorial perfect (note also the timeless present
est). See also Chapter 24 for variation in the use of the tenses in narrative texts.
(m) Provinciae toti quam maximum potest militum numerum imperat—erat
omnino in Gallia ulteriore legio una— . . .
(‘From the whole province he requisitioned the largest possible number of troops—
there was in further Gaul no more than a single legion— . . . ’ Caes. Gal. 1.7.2)
(n) Erat isdem temporibus T. Gracchus P. f., qui bis consul et censor fuit, cuius
est oratio Graeca apud Rhodios.
(‘To the same time belonged Tiberius Gracchus, son of Publius, who was twice
consul and censor. We have his oration, written in Greek, before the people of
Rhodes.’ Cic. Brut. 79)
Whereas, in many passages in comedy and in the orations of Cicero, the special effect
of ‘nearness’ of the present when used to tell events of the past is obvious, the present
seems sometimes to be used merely to report events in the most economic way.
A good example of this so-called ‘annalistic’ use is (o). Note that not all the events are
necessarily successive.
(o) Ad eam multitudinem urbs quoque amplificanda visa est. Addit duos colles,
Quirinalem Viminalemque. Inde deinceps auget Esquilias ibique ipse, ut
loco dignitas fieret, habitat. Aggere et fossis et muro circumdat urbem. Ita
pomerium profert.
(‘To meet the wants of this population it was apparent that the city must expand. He
added two hills, the Quirinal and the Viminal, after which he proceeded to enlarge
the Esquiline, going there to live himself, that the place might obtain a good
reputation. He surrounded the city with a rampart, trenches, and a wall, and so
extended the “pomerium”.’ Liv. 1.44.3)
This passage has some of the characteristics described in the beginning of this section,
e.g. short sentences and the verb in initial position. Although the present forms as
such seem to be interchangeable with perfect forms, the very presence of these
characteristics makes real substitution impossible.
The imperfect indicative tense is used to assert that a state of affairs was taking place
at some moment in the past and had not yet finished or come to an end. This is
invariably the description of the Latin grammarians and is also nicely illustrated by
Pliny the Elder in his discussion of the use of the imperfect by the famous Greek
artists Apelles and Polyclitus.
Me non paenitet nullum festiviorem excogitasse titulum et, ne in totum videar
Graecos insectari, ex illis mox velim intellegi pingendi fingendique conditoribus,
quos in libellis his invenies absoluta opera et illa quoque, quae mirando non
satiamur, pendenti titulo inscripsisse, ut APELLES FACIEBAT aut POLYCLITUS,
tamquam inchoata semper arte et inperfecta, ut contra iudiciorum varietates super-
esset artifici regressus ad veniam velut emendaturo quicquid desideraretur, si non
esset interceptus.
(‘For myself, I am not ashamed of not having invented any livelier title. And so as not
to seem a downright adversary of the Greeks, I should like to be accepted on the lines
of those founders of painting and sculpture who, as you will find in these volumes,
used to inscribe their finished works, even the masterpieces which we can never be
tired of admiring, with a provisional title such as Worked on by Apelles or Polyclitus,
as though art was always a thing in process and not completed, so that when faced
with the vagaries of criticism the artist might have left him a line of retreat to
indulgence, by implying that he intended, if not interrupted, to correct any defect
noted.’ Plin. Nat. praef. 26)
55
Further examples in K.-St.: I.121.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Two elements of the semantic value of the imperfect will be dealt with separately:
(i) the past reference point (§ 7.19) and (ii) the element of being in progress (§ 7.20).
A third section is devoted to later developments (§ 7.21).56
56
See Gippert (1999) for theories about the prehistoric developments of the imperfect. For an approach
from the point of view of William Diver’s ‘relevance theory’, see Farron (1973).
57
For the various ways in which the imperfect locates a state of affairs in the past, see Touratier (1998).
58
K.-St.: I.124; Sz.: 317.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
petendum esse paulo ante dicebas, et . . . (Cic. de Orat. 1.230); Paulo ante dicebas:
‘corrumpere fratris uxorem ausus est, istud incestum est’. (Quint. Decl. 291.2);
Contextual indications: Sed aperiuntur aedes quo ibam. (Pl. Trin. 400); Oh, / tu
quoque aderas, Phormio? # Aderam. (Ter. Ph. 857–8); Ehem, pater mi, tu hic eras?
(Ter. Ad. 901); Demiphonem si domi’st / visam, ut quod—# At nos ad te ibamus,
Phormio. (Ter. Ph. 898–9); . . . ut exceptis Aetolis, quibus nec pro spe victoriae
praemia contigerant nec diu quies placere poterat, universa Graecia simul pacis
libertatisque perfruens bonis egregie statu suo gauderet . . . (Liv. 34.22.4); . . . ‘quid
agebas’ inquit ‘frater sanctissime?’ (Petr. 11.2);
Referring to an event in one’s memory: Ego dissuadebam, mater. (Pl. As. 931); Tu
iam quod quaerebas habes. (Pl. Mos. 210); Sed opsecro hercle, Astaphium, <i>intro . . . /
Licet. # Audi’n etiam? # Quid vis? # Di me perduint / qui te revocavi. Non tibi
dicebam ‘i’ modo? # / Quid iam revocabas? Inprobe nihilique homo? / Tute tibi mille
passum peperisti morae. (Pl. Truc. 329–34); Uno verbo . . . # Eloquere. # Mitti’n me
intro? # Mendax es. Abi. / Unum aiebas, tria iam dixti verba, atque <ea> mendacia.
(Pl. Truc. 756–7); Gaudebam. Item alio die / quaerebam. Comperibam nil ad
Pamphilum / quicquam attinere. Enimvero spectatum satis / putabam et magnum
exemplum continentiae. (Ter. An. 89–92); Non imperabat coram, quid facto esset
opus puerperae . . . (Ter. An. 490); Effudi vobis omnia quae sentiebam. (Cic. de Orat.
1.159); Ita uno die sex proeliis factis . . . cum horum omnium ratio haberetur ad duo
milia numero ex Pompeianis cecidisse reperiebamus . . . (Caes. Civ. 3.53.1);
D(IS) M(ANIBUS) / HIC IACET AUFIDIA SEVERINA SIGNO FLORENTI / BIS QUINOS DENOS QUAE
VIXIT ANNOS AETATIS / CASTA FIDE SEMPER TORU<M> MARITALE<M> DILEXIT / SOBRIA NON
MOECHA SIMPLEX ANIMOQUE BENIGNO / DEDITA CONIUGI SOLI SUO IGNARA ALIENUM / CONPARI
DULCI SUAE INCOMPARABILI SOLAE. / BASILEUS FECIT QUOD FIERI AB ILLA CUPIEBAT. (CIL
VI.12853 (Rome));
Wheeler (1906: 389–90), Bennett (1910: I.32–5), and others use the term ‘aoristic’ for
the use of the imperfect in cases like (g), noting that it is found in pre-Ciceronian
Latin especially with the verbs aio and dico and sum and its compounds.59
(g) Te velle uxorem aiebat tuo gnato dare. / Ideo aedificare hoc velle aiebat in
tuis. / # Hic aedificare volui? # Sic dixit mihi.
(‘He said you wanted to give a wife to your son, and he said that was why
you wanted to enlarge your own house here. # I wanted to enlarge my
house? # So he told me.’ Pl. Mos. 1027–9)
In many of the examples above, the perfect indicative might have been used as well,
and sometimes it is not easy to explain why the imperfect tense is used rather than the
perfect. Another example is (h). If the perfect form duxit were used instead of
ducebat, it would simply ask for the reason why he has brought his bride home,
whereas ducebat asks for the things the young man had in mind when he was taking
his bride home.
59
On these expressions, see Mellet (1987).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The imperfect is also used in letters (especially by Cicero) to indicate the circum-
stances of their composition, while in English the present is more typical (that is, ‘I am
writing this’). The writer’s viewpoint stands in contrast to that of the recipient of
the letter, for whom the letter-writing took place at a certain moment in the past. Exx.
(i)–(k) are instances of this so-called ‘epistolary’ use of the imperfect indicative. Note
in (j) the use of present tense volumus and est in a statement of general validity,
alongside the use of the imperfects erat and habebat, which locate the events Cicero is
describing in the time of writing.
The past reference point of an imperfect indicative form is often related to past
verb forms in the context, both when the imperfect is in a subordinate clause and
when it is in a main clause. Examples are (l)–(m) and (n)–(o), respectively. The
reference point of the verb form to which it is related may or may not be given
explicitly.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(l) Verum tempestas quondam, dum vixi, fuit / quom inter nos sordebamus
alter de altero.
(‘But there was a period once while I was alive when we were soiled from one
another.’ Pl. Truc. 380–1)
(m) Eademne erat haec disciplina tibi, quom tu adulescens eras?
(‘Did you have the same sort of education when you were a teenager?’ Pl. Bac. 421)
(n) Mater mihi Samia fuit. Ea habitabat Rhodi.
(‘I had a Samian mother. She lived on Rhodes.’ Ter. Eu. 107)
(o) Ut iacui, exsurgo. Ardere censui aedis. Ita tum confulgebant.
(‘Lying as I was, I stood up. The house was so bright at the time that I thought it was
on fire.’ Pl. Am. 1067)
Given the subjective nature of the category tense, the speaker sometimes has a choice.
An example of this is (c) in this section. States of affairs may be situated in the past
even when they are also pertinent to the time of speaking. Two typical examples of
this are (p) and (q): ex. (p) describes a feature of the elephant, and (q) is a geograph-
ical description. However, the fact that the imperfect and the present might both be
used does not mean that both would present the states of affairs in the same way. In
(p), the imperfect habebat refers to the time when nature provided a solution for a
then-existing problem. In (q), the imperfect erat reflects the considerations at the time
which led to the decision to depart, and makes it clear that the quod clause is not the
narrator’s responsibility.60
(p) Manus etiam data elephanto’st, quia propter magnitudinem corporis diffi-
ciles aditus habebat ad pastum.
(‘A hand (i.e. a trunk) even is given to the elephant, because, on account of the great
size of his body, he was having difficult approaches to food.’ Cic. N.D. 2.122)
(q) Ipse cum omnibus copiis in Morinos proficiscitur, quod inde erat brevissi-
mus in Britanniam traiectus.
(‘He himself, with all his forces, sets out for the territory of the Morini, because from
that place there was the shortest passage across to Britain.’ Caes. Gal. 4.21.3)
Supplement:
Namque, ut ante dixi, veteres illi usque ad Socraten omnem omnium rerum quae ad
mores hominum, quae ad vitam, quae ad virtutem, quae ad rem publicam pertine-
bant cognitionem et scientiam cum dicendi ratione iungebant. (Cic. de Orat. 3.72);
Eorum fines Nervii attingebant. (Caes. Gal. 2.15.3);
Locating a state of affairs in the past even when it also obtains at the time of speaking
may have the effect of distancing it and making it less direct. An example that is much
discussed in the literature about the Latin tense system is (r).61 It is sometimes
60 61
See Fornés (1999: 125). See Torrego (1994a), with references.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
62
For this form of ‘non-direct discourse’, see Rosén (2013: 233–6), also about the difficult passage Cic.
Att. 9.2a.3.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(u) Cum autem ex eo quaererem quid videret, quem exitum, quam rem pu-
blicam, plane fatebatur (sc. Curio) nullam spem reliquam. Pompei classem
timebat. Quae si †esset†, se de Sicilia abiturum.
(‘When I asked for his assessment—what way out, what constitution he
foresaw—he frankly admitted that the situation is quite hopeless and that he
is afraid of Pompey’s fleet. If it is put to sea (?) he said he would leave Sicily.’
Cic. Att. 10.4.9)
(v) Nuntiatur Afranio magnos commeatus, qui iter habebant ad Caesarem,
ad flumen constitisse. Venerant eo sagittarii ex Rutenis, equites ex Gallia
cum multis carris magnisque impedimentis, ut fert Gallica consuetudo.
Erant praeterea cuiusque generis hominum milia circiter VI cum servis
liberisque.
(‘Word is brought to Afranius that the great supplies on their way to Caesar
are stopped by the stream; there had come thither archers from the Ruteni
and horsemen from Gaul with a number of wagons and heavy baggage, after
the Gallic custom. There were moreover about six thousand men of every
class with their slaves and children.’ Caes. Civ. 1.51.1–2)
63
In this ‘habitual’ type of contexts the imperfect rarely occurs with frequency expressions. See
Torrego (1988, 1989a, and 1994a), who regards the co-occurrence of frequency expressions with the
imperfect as an argument against the two-dimensional approach to the Latin tense system as presented in
this Syntax.
64
See Szantyr (1970) for such instances of the imperfect.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Incipiebat enim sedens et, si quando illum produxerat calor, exsurgere aude-
bat. Illa intempestiva in declamationibus eius philosophia sine modo tunc et
sine fine evagabatur. Raro totam controversiam implebat. Non posses dicere
divisionem esse, non posses declamationem. Tamquam declamationi multum
deerat, tamquam divisioni multum supererat. Cum populo diceret, omnes
vires suas advocabat et ideo non desinebat. Saepe declamante illo ter bucina-
vit, dum cupit in omni controversia dicere non quidquid debet dici sed
quidquid potest. Argumentabatur moleste magis quam subtiliter. Argumenta
enim argumentis colligebat et, quasi nihil esset satis firmum, omnes proba-
tiones probationibus aliis confirmabat. Erat et illud in argumentatione vitium
quod quaestionem non tamquam partem controversiae sed tamquam con-
troversiam implebat. Omnis quaestio suam propositionem habebat, suam
exsecutionem, suos excessus, suas indignationes, epilogum quoque suum.
(‘He used to start off sitting down, and if his passion carried him on he would venture
to get up. He celebrated philosophical observations, which were quite out of place in
declamation, wandered on without restraint and without end. He rarely completed a
whole controversia; you couldn’t call it a division—or a declamation: for as a
declamation, it lacked much, for as a division it had much that was superfluous.
But whenever he spoke in public he used to summon up all his powers, and therefore
he would not stop: often while he was speaking the trumpet would blow three times,
since in every controversia he desired to say not what ought to be said but what was
capable of being said. He argued laboriously rather than subtly; he would accumulate
argument upon argument and, as though there were no firm ground, confirmed all
his proofs with other proofs. There was even this flaw in his argumentation, namely
that he would develop a question not as part of a controversia but as a controversia;
every question would have its own statement, its own treatment, its own digressions,
its own appeals to anger, even its own epilogue.’ Sen. Con. 7.pr.1–2)
(b) Sic enim appellabantur ii qui ostentationis aut quaestus causa philosophantur.
(‘For so were named those who pursue philosophy for the sake of display or profit.’
Cic. Ac. 2.72)
(‘Meanwhile, to the temple of unfriendly Pallas the Trojan women passed along with
streaming tresses and bore the robe, mourning in suppliant guise and beating breasts
with hands: with averted face the goddess kept her eyes fast upon the ground.’ Verg.
A. 1.479–82)
(e) Corvus . . . insilibat, obturbabat et unguibus manum laniabat et prospectum
alis arcebat atque, ubi satis saevierat, revolabat in galeam tribuni.
(‘The raven . . . jumped onto (the Gaul), harassed him, tore his hand with its claws,
obstructed his sight with its wings, and then, when it had raged enough, flew back to
the tribune’s helmet.’ Quad. Hist. 12)
(f ) (Caesar) Haeduos . . . in dicione videbat Germanorum teneri eorumque
obsides esse apud Ariovistum . . . intellegebat. Quod . . . turpissimum . . .
arbitrabatur. Paulatim autem Germanos consuescere Rhenum transire
. . . videbat neque sibi homines . . . barbaros temperaturos existimabat, quin
. . . Quibus rebus quam maturrime occurrendum putabat.
(‘(Caesar) saw that the Aedui were held in the power of the Germans and he was
aware that their hostages were with Ariovistus. He judged this to be an utter disgrace.
Moreover, he saw that the Germans were gradually becoming accustomed to cross-
ing the Rhine. And he did not suppose that the barbarians would restrain themselves
from . . . And he thought these matters must be dealt with as quickly as possible.’
Caes. Gal. 1.33.2–4)
Supplement:
Individual habits and customs: Inveniebat igitur acute Cotta, dicebat pure ac solute.
Et ut ad infirmitatem laterum perscienter contentionem omnem remiserat, sic ad
virium imbecillitatem dicendi accommodabat genus. (Cic. Brut. 202); Milonem
athletam, cum constitisset, nemo vestigio educebat, malum tenenti modo digitum
corrigebat. (Plin. Nat. 7.83); [Sed] memini Safinium. Tunc habitabat ad arcum
veterem me puero, piper, non homo. Is quacumque ibat, terram adurebat. (Petr.
44.7 (Phileros speaking)); Post solem plerumque frigida lavabatur; deinde gustabat
dormiebatque minimum. Mox quasi alio die studebat in cenae tempus. (Plin. Ep.
3.5.11); Convivabatur assidue nec umquam nisi recta, non sine magno ordinum
hominumque dilectu. (Suet. Aug. 74.1); In lectum inde transgressus non amplius
cum plurimum quam septem horas dormiebat . . . (Suet. Aug. 78.1);
Features and characteristics: Ut enim Romae consules, sic Karthagine quotannis
annui bini reges creabantur. (Nep. Han. 7.7); In summo custos Tarpeiae Manlius
arcis / stabat pro templo et Capitolia celsa tenebat . . . (Verg. A. 8.652–3); Ter circum
Iliacos raptaverat Hectora muros / exanimumque auro corpus vendebat Achilles.
(Verg. A. 1.483–4);
Repeated and parallel events: Fusis Auruncis victor tot intra paucos dies bellis
Romanus promissa consulis fidemque senatus expectabat, cum Appius et insita
superbia animo et ut collegae vanam faceret fidem, quam asperrime poterat, ius de
creditis pecuniis dicere. Deinceps et qui ante nexi fuerant creditoribus tradebantur et
nectebantur alii. Quod ubi cui militi inciderat, collegam appellabat. Concursus ad
Servilium fiebat. Illius promissa iactabant. Illi exprobrabant sua quisque belli merita
cicatricesque acceptas. Postulabant ut aut referret ad senatum, aut auxilio esset
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
consul civibus suis, imperator militibus. Movebant consulem haec, sed tergiversari
res cogebat. Adeo in alteram causam non collega solum praeceps erat sed omnis
factio nobilium. Ita medium se gerendo nec plebis vitavit odium nec apud patres
gratiam iniit. (Liv. 2.27.1–3);
There are also some instances of a series of imperfect forms that may be interpreted
as successive. An example from Caesar is his lengthy description of the building of a
particular defensive construction in Gal. 7.73.5–7. In this passage certain—
terminative—activities involved in the construction are in the imperfect for the
same reason as the other parts of the passage, but they must necessarily follow
each other in extralinguistic reality.65
Its semantic value ‘in progress in the past’ makes the imperfect indicative appropriate
for marking background information which serves as a frame for other events occurring
as incidents against that background. This background may either precede, as in (g), or
follow, as in (h). In the latter case the information in the imperfect often supplies an
explanation (note enim ‘for’; autem ‘and indeed’ is also frequent in sentences that
contain an imperfect).66 The semantic relationship between two consecutive sentences
may be made more prominent by the presence of temporal satellites like iam ‘now’,
‘already’, ‘nondum ‘not yet’, and vix ‘scarcely’ in the preceding sentences and/or repente
and subito ‘suddenly’ in the following sentences. The second sentence may then be
understood as interrupting the event expressed in the preceding sentence. An example
is (i). The pluperfect is used in the same way. More details can be found in Chapter 24.
(g) Tantos illa suo rumpebat pectore questus. / Aeneas celsa in puppi iam certus
eundi / carpebat somnos rebus iam rite paratis. / Huic se forma dei vultu
redeuntis eodem / obtulit . . .
(‘Such were the cries that kept bursting from her heart. But now that all was duly
ordered, and now that he was resolved on going, Aeneas was snatching sleep on his
vessel’s high stern. In his sleep there appeared to him a vision of the god, as he came
again with the same aspect . . . ’ Verg. A. 4.553–7)
(h) . . . tantus repente terror invasit, ut cum Lentulus consul ad aperiendum
aerarium venisset . . . , protinus . . . profugeret. Caesar enim adventare iam
iamque et adesse eius equites falso nuntiabantur.
(‘ . . . suddenly a great terror fell upon them, with the result that when the consul
Lentulus came to open the treasury he immediately fled. For it was falsely being
reported that Caesar was on the very point of arriving and that his cavalry was
present.’ Caes. Civ. 1.14.1)
(i) Iamque iugis summae surgebat Lucifer Idae / ducebatque diem, Danaique
obsessa tenebant / limina portarum, nec spes opis ulla dabatur. / Cessi et
sublato montes genitore petivi.
65
For a discussion of this passage, see Oldsjö (2001: 467–72).
66
For statistical information, see Kroon (1998: 52); also Rosén (2002b: 333–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(‘Now the Dawn Star was rising from the ridges at the top of Ida and was bringing the
day. The Danaans were holding the entrances of the gates under guard and no hope
of succour was offered. I gave way, picked up my father, and made for the mountain.’
Verg. A. 2.801–4—tr. Horsfall)
The imperfect indicative is regularly used in the main clause of a complex sentence to
mark an event that was in progress, but was then interrupted by the occurrence of
another event. The so-called ‘cum inversum’ construction is the best-known device in
this context. An example is (j). The main clause often contains expressions denoting
duration like iam ‘already’ in (k) (further instances can be found in § 16.11). Other
subordinate clauses with a similar interrupting effect are exemplified by (l) with ni/
nisi ‘if not’ (more instances in § 16.62) (for donec ‘until’, see instances in § 16.18). This
is a narrative device that is often used by historians, especially by Tacitus, and in
poetry.67 Instead of in a subordinate clause, the interrupting event may also be
expressed in a coordinate clause, as in (m) (for more instances see Chapter 19).
(j) Tu clamabas deum fidem atque hominum omnium, / quom ego accurro
teque eripio vi, pugnando, ingratiis.
(‘You were imploring the protection of gods and all men when I ran to you and
rescued you with might and main, against their will.’ Pl. Men. 1053–4)
(k) Iamque ab eo non longius bidui via aberant, cum duas venisse legiones missu
Caesaris cognoscunt.
(‘And by now they were no farther from him than a two days’ march, when they
learnt that two legions had come, as dispatched by Caesar.’ Caes. Gal. 6.7.2)
(l) Quin labebar longius, nisi me retinuissem.
(‘Indeed, I was still wavering further, if I had not prevented myself.’ Cic. Leg. 1.52)
(m) Iamque propinquabant portis, rursusque Latini / clamorem tollunt et mollia
colla reflectunt.
(‘And now they were approaching the gates when again the Latins raise their shout,
and wheel about their horses’ supple necks.’ Verg. A. 11.621–2)
In specific contexts, and with certain (in particular terminative) states of affairs,68 it
can be the case that an action was being carried out, but was never fully finished. In
such cases, the imperfect is sometimes interpreted as ‘conative’ (see also § 7.3 and
§ 7.17). Examples are (n) and (o). The fact that the actions marked by the imperfect
are interpreted as not successful is not due to their being in the imperfect tense. This
is obvious in (o): prohibebant is to be compared with the preceding propugnabant.
In both cases, the imperfects are used for the same reason: they present the
67
For the use of these patterns ‘de rupture’ in the Latin historians, see Chausserie-Laprée (1969:
597–617, 636–7). Also Mellet (1988: 230–5), Torrego (1999b), and Orlandini (2001: 178–83, 2007:
138–43). For its use in Greek and Latin epic, see Nesselrath (1992).
68
For the conative interpretation of terminative verbs, see Vester (1983: 18–20).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
background situation in which the milites legionis septimae carried out their actions.
The conative interpretation of prohibebant depends on the information given in the
following sentence, introduced by the contrastive connector at. Likewise, one could
interpret Quo nunc ibas (‘Where were you going now?’) in (c) in § 7.19 as conative,
but the fact that Charinus does not reach the destination he mentions (exile) can be
deduced from the situation on stage (adeo + direction adjunct is terminative) and
thus is not connected with the imperfect indicative itself. Exx. (p) and (q) illustrate the
same point. In (p), a perfect adii would incorrectly imply that Dido had already
reached the stars;69 in (q), a perfect didicerunt would incorrectly suggest that everyone
became a fully educated musician.
(n) Veniebatis igitur in provinciam . . . Prohibiti estis in provincia vestra pedem
ponere.
(‘Well, you were going to the province . . . You were prevented from setting foot in
your province.’ Cic. Lig. 24)
(o) Ipsi ex silvis rari propugnabant nostrosque intra munitiones ingredi prohi-
bebant. At milites legionis septimae testudine facta et aggere ad munitiones
adiecto locum ceperunt.
(‘The enemy themselves, coming out of the woods, were fighting in small groups and
were preventing our troops from entering the fortification. But the soldiers of the
Seventh Legion, with a ‘tortoise’ having been formed and a ramp thrown up against
the fortifications, took the position.’ Caes. Gal. 5.9.6–7)
(p) . . . / extinctus pudor et, qua sola sidera adibam, / fama prior.
(‘My honour has been destroyed and my former fame by which alone I was
approaching the stars.’ Verg. A. 4.322–3)
(q) Ergo in Graecia musici floruerunt discebantque id omnes, nec qui nesciebat
satis excultus doctrina putabatur.
(‘Musicians accordingly flourished in Greece. Everyone was learning music, and who-
ever was unacquainted with it was not regarded as completely educated.’ Cic. Tusc. 1.4)
Supplement:
Pallam ad Phrygionem cum corona ebrius / ferebat hodie tibi quam surrupuit domo.
(Pl. Men. 563–4); Tum adflicto P. Sulla consulatus vobis pariebatur, sicuti partus est.
(Cic. Sul. 49); Sed Iugurtha primo tantummodo belli moram redimebat existumans sese
aliquid interim Romae pretio aut gratia effecturum. (Sal. Jug. 29.3); Talibus Aeneas
ardentem et torva tuentem / lenibat dictis animum lacrimasque ciebat. (Verg. A.
6.467–8); Cornelio minus copiarum datum, quia L. Manlius praetor et ipse cum haud
invalido praesidio in Galliam mittebatur. (Liv. 21.17.7); Obstupueram ego supplicii
metu pavidus nec quid in re manifestissima dicerem inveniebam. (Petr. 108.1); . . . his
et similibus notos pariter et ignotos ad faciendum fortiter accendebat. (Amm. 16.12.29);
69
Heine (1990: 13–14) argues that hyperbolic expressions are so regular in Virgil and other authors,
that Virgil might have made Dido say ‘adii’ anyhow.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
As was shown in § 7.11, Tables 7.9 and 7.10, there is an indirect correlation between
the frequency of tense forms and the meanings of verbs. In Caesar, for example, verbs
denoting an opinion like existimo ‘to form or hold an opinion’ occur more often in
the imperfect; by contrast, verbs denoting an action like duco (legionem) ‘to lead a
legion’, and especially those involved in terminative states of affairs occur much more
frequently in the perfect. This has to do with what was mentioned above as the
distinction between ‘background’ and ‘foreground’ information in a narrative: a verb
like existimo is typically suited to refer to the considerations on the basis of which one
undertakes an action. That being said, the perfect form of existimo is certainly
possible in a suitable context such as (r), and conversely, the imperfect form of
duco in (s) is unobjectionable.
(r) His rebus cum iter agminis nostri impediretur, non omittendum sibi con-
silium Nervii existimaverunt.
(‘When the route of our army was hampered by these things, the Nervii thought that
the plan should not be abandoned.’ Caes. Gal. 2.17.5)
(s) Nam quod hosti adpropinquabat, consuetudine sua Caesar sex legiones
expeditas ducebat . . .
(‘For, because he was approaching the enemy, Caesar, according to his custom, was
leading six legions in light field order . . . ’ Caes. Gal. 2.19.2)
The imperfect indicative is rare with events that occur quickly. Examples cited in the
literature are late, notably in Ammianus. Ex. (t) is such an isolated instance.
(t) . . . amplificatis viribus ardore incohibili in excidium urbium matris Seleuciae
efferebantur . . .
(‘ . . . with increased numbers and irresistible energy they rushed on to destroy
Seleucia, the metropolis of the province . . . ’ Amm. 14.2.14)
Supplement:
Quibus ita sceleste patratis Paulus cruore perfusus reversusque ad principis castra
multos coopertos paene catenis adduxit in squalorem deiectos atque maestitiam,
quorum adventu intendebantur eculei uncosque parabat carnifex et tormenta.
(Amm. 14.5.9);70
70
See Ronconi (1959: 58–60).
71
See Haverling (1998, 2005, 2010a: 480–6) and Pinkster (1998b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The future indicative tense of the infectum stem (futurum simplex, henceforth ‘simple
future’) is used by the speaker to assert his prediction that an event will be taking place
at some moment in the future.72 As is the case with the other tenses, this need not
exclude the same event from also pertaining to either the present or the past. The only
thing the speaker does is commit himself to the assertion that the event will be taking
place in the future. The (grammatical) person is a more important factor in assessing
the communicative goal of the information presented in the simple future tense than
elsewhere. When the speaker is talking about his own future actions or states, his
statement is most likely taken as a declaration of his intention. When, in turn, the
speaker is talking about the addressee, the statement is most likely to be taken as a
prediction or an instruction. Finally, when the speaker is talking about a third person,
the statement is most likely to be taken as a prediction. In general, given the nature of
‘futurity’ itself, the simple future indicative is less assertive than the other tenses and
has some uses which resemble those of the present subjunctive. In what follows,
‘purely temporal’ instances of the simple future indicative are dealt with first; some
less temporal ones come after. It must be stressed, however, that the distinctions made
are essentially determined by the context; there are no decisive criteria by which to
assign the examples to clear-cut categories.73
72
The best discussion of the uses of the future tense in Latin is still Sjögren (1906). For a general
discussion of the category ‘future’ in language, see Fleischman (1982).
73
Bennett (1910) has the most detailed classification of the uses of the future indicative. For
discussion, see Mellet (1989). For the idiomatic use of videbimus, which almost equals ‘we shall not go
into this now’, see Nicholas (1964 and 1971).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Examples of the so-called ‘purely temporal’ use of the simple future indicative are
(a)–(c). They are found from Early Latin onwards and continue to occur throughout
the history of the language.
The future reference point may be specified in several ways, by adverbs like cras
‘tomorrow’ and postridie ‘on the following day’, as in (d), or by main or subordinate
clauses containing a verb form with future time reference or future time implication,
as in (e)–(g). Ex. (e) contains a future perfect tense form (cecidero), (f ) an imperative
form (imponito), (g) a gerundival expression (audiendi sc. sunt).
Supplement:
Non loquar nisi pace facta, quando pugnis plus vales. (Pl. Am. 390); Cyathum in
ceteram potionem indito. Alvum movebit et postridie perpurgabit sine periculo.
(Cato Agr. 115.2); Qui locus vino optimus dicetur esse et ostentus soli aminnium
minusculum et geminum, eugenaeum, heluolum minusculum conserito. (Cato Agr.
6.4); Quaeso igitur, si quod verbum absurdius aut inconsultior sensus aut infirmior
littera istic erit, id tempori adponas. (Fro. Aur. 4.2.5);
The future is also used in a context that itself refers to the past. The first instance cited is
(h) from Virgil. This authorial use of the simple future is common in poetry. It is
introduced into prose by Livy and is afterwards not uncommon in authors like Seneca.74
(h) Nescia mens hominum fati sortisque futurae / et servare modum rebus
sublata secundis! / Turno tempus erit magno cum optaverit emptum /
intactum Pallanta . . .
(‘O mind of man, knowing not fate or coming doom or how to keep bounds when
uplifted with favouring fortune! To Turnus shall come the hour when for a great
price he will long to have bought an unscathed Pallas . . . ’ Verg. A. 10.501–4)
Supplement:
Ex istis Romulus alter erit. (Ov. F. 2.386); Hic erit iuvenis, penes quem perfecti
huiusce belli laus est, Africanus ob egregiam victoriam de Hannibale Poenisque
appellatus. (Liv. 21.46.8);
Apart from its purely temporal use, the simple future is also used with all sorts of less
temporal or even non-temporal values. There are various labels for these uses, which
in practice are not always easy to distinguish and can best be regarded as contextually
determined variants. The simple future is, for example, often used in statements of a
general character, such as general truths, proven practices, etc. This use of the simple
future is sometimes called GNOMIC, as in (i) and (j). The simple future is also very
commonly used in legal texts alongside the present tense to indicate what must be or
is the proper way to proceed given a specific legal situation, although individual
authors vary in the degree to which they prefer the present or simple future tense.
Ulpian, for example, is said to use the simple future much more often than other
jurists.75 An example is (k). In a similar way, future and present tense forms are used
indiscriminately alongside each other in Late Latin.
(i) Proinde ut quisque fortuna utitur, / ita praecellet . . .
(‘In proportion as each one has Fortune, he excels . . . ’ Pl. Ps. 679–80)
(j) . . . qui utilitatem defendit enumerabit commoda pacis . . .
(‘ . . . the one who will defend expediency will relate the advantages of peace . . . ’ Cic.
de Orat. 2.335)
74
See Sz.: 310; for Seneca, see Bourgery (1922: 330).
75
For examples and discussion, see Honoré (1982: 65–7).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The simple future indicative is also used in sentences in which the speaker only
formulates an assumption about what may happen or what may be the case. Some
Latinists call this use of the simple future indicative ‘potential’.76 An example is (l). In
addition, it is sometimes used in sentences containing a (common-sense) conclusion
based on evidence mentioned in the context or on general knowledge. An example of
a sentence with such a so-called ‘deductive’ use of the future is (m).77
(l) Irata est, credo, nunc mihi. Placabit palla quam dedi, . . .
(‘She’s angry with me now, I believe; the mantle I gave her will mollify her . . . ’ Pl.
Men. 600);
(m) Haec erit bono genere nata. Nil scit nisi verum loqui.
(‘She’ll be from a good family; she knows how to speak nothing but the truth.’ Pl. Per.
645)
Supplement:
Assumptive: Quid ergo est? Quaeret fortasse quispiam displiceatne mihi legum
praesidio capitis periculum propulsare. (Cic. Clu. 144); Ex quo quamquam hoc
videbitur fortasse cuipiam durius tamen audeamus imitari Stoicos qui studiose
exquirunt unde verba sint ducta . . . (Cic. Off. 1.23); Quis tandem erit tam demens
qui dubitet utrum se esse malit? (Cic. Rep. 3.27); Ne extra rerum naturam vagari
virtus nostra videatur, et tremet sapiens et dolebit et expallescet. (Sen. Ep. 71.29);
Deductive: Sed profecto hoc sic erit: / centum doctum hominum consilia sola haec
devincit dea, / Fortuna. (Pl. Ps. 677–9); Enim Charmides? / Em istic erit: qui istum di
perdant. (Pl. Trin. 922–3); Verbum hercle hoc verum erit ‘sine Cerere et Libero friget
76
So, for example, Sz.: 311. For the term ‘assumptive’ modality, see Palmer (2001: 28–31).
77
For the term ‘deductive’ modality, see Palmer (2001: 26–31). For this use of the simple future in
Latin, see Núñez (1991: 187–93). Cf. Bolkestein (1980: 123ff.), who uses the term ‘inferential’.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Venus’. (Ter. Eu. 732); Et sunt illa sapientis. Aberit igitur a sapiente aegritudo. (Cic.
Tusc. 3.18); Huius motus succutientis terras haec erit causa. (Sen. Nat. 6.23.1); Sic et
insectis quisquis est vitalis umor hic erit sanguis. (Plin. Nat. 11.8);
The deductive modality can also be expressed with debeo (‘must’), e.g. ‘Plane’ inquam
‘hic debet servus esse nequissimus. Aliquis oblivisceretur porcum exinterare?’ (Petr.
49.7 ); Debent adcommodatae esse et equorum naturae neque ex alia causa nomen
accepisse. (Plin. Nat. 22.30)
There are a number of contexts in which the speaker can choose between the simple
future indicative and the present subjunctive. This is, of course, no proof of their
being synonymous. Two illustrations are given in (n)–(q). Whereas with the simple
future indicative in (n), the speaker expresses his assumption that somebody will
indeed say something, the speaker in (o) expresses no such assumption by using the
present subjunctive; rather, he only speculates78 that somebody might say something
or is likely to do so. Occasionally, scholars also distinguish a ‘dubitative’ use of the
simple future indicative in questions, as in (p). Here, too, subjunctive forms are found
as well (and they are much more common), as in (q). (For examples of the ‘potential’
and ‘dubitative’ use of the present subjunctive, see §§ 7.40–7.42.)
(n) ‘Dedemus ergo Hannibalem?’ dicet aliquis.
(‘ “Shall we then surrender Hannibal?” someone will ask.’ Liv. 21.10.11)
(o) ‘Quid ergo postulas?’ dicat aliquis.
(‘ “What then do you ask for?” someone may say.’ Liv. 37.53.25)
(p) Perii. Quid agam? Quo me vortam? Quid viro meo respondebo . . .
(‘I’m ruined! What will I do? Which way will I turn? What will I tell my husband . . . ?’
Ter. Hec. 516)
(q) Quam, malum? Quid machiner? Quid comminiscar?
(‘But what trick, damn it? What can I think up? What can I come up with?’ Pl. Capt.
531)
78
On ‘speculative’ epistemic modality, see Palmer (2001: 26–8).
79
See Risselada (1993: 169ff.).
80
As an exception, L. Löfstedt (1966: 12) cites Pl. Mil. 571: Ne tu hercle, si te di ament, linguam
conprimes, but ne actually is the asseverative particle, here, as so often, in combination with hercle.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
other simple future indicative forms. The speaker uses the simple future when he is
confident that what he wants will indeed happen, because of his authority, because of
previous agreement, or because of the intimacy of the relationship.
(a) Non me appellabis si sapis.
(‘You won’t call me, if you have any sense.’ Pl. Mos. 515)
(b) Tu tamen quod poteris . . . nos consiliis iuvabis.
(‘All the same you will help me with your advice so far as you will be able.’ Cic. Att.
10.2.2)
(c) Haec igitur tibi erunt curae . . .
(‘This then will be your concern . . . ’ Cic. Fam. 3.9.4)
Supplement:
Atque hercle invenies tu locum illi, si sapis. / Nullum hercle praeter hunc diem illa
apud med erit. (Pl. Mer. 584–5); Certe edepol tu me alienabis numquam quin noster
siem. (Pl. Am. 399); De Caelio tu quaeres, ut scribis. Ego nihil novi. (Cic. Att. 12.5a);
Tu interea non cessabis et ea quae habes instituta perpolies nosque diliges. (Cic.
Fam. 5.12.10); Haec igitur tibi erunt curae, quem ad modum ostendis, meque totum
et mea et meos commendatos habebis. (Cic. Fam. 3.9.4); Sed prius in eius locum
virum fortem ac strenuum novum senatorem cooptabitis quam de noxio supplicium
sumatur. (Liv. 23.3.6); Nunc auctoritate veteris imperatoris contentus eris. (Liv.
44.36.13); Tu nihil invita dices faciesve Minerva. (Hor. Ars 385); Aut dixit: ‘Venies
hodie, cessabimus una’. (Prop. 3.23.15); ‘Non ambulabis’, inquit, ‘eadem via qua ego,
non calcabis vestigia mea, non offeres delicatis oculis sordidam vestem, non flebis
invito me, non tacebis’. Perieramus, si magistratus esset. (Sen. Con. 10.1.1); Facies
(facias cj. Bücheler) et totum populum sibi suaviter facientem. (Petr. 71.10 (Trimal-
chio speaking)); DABES(= DABIS) S. CALVIS(IO) MILIT(I) L(EGIONIS) (CEL 29 (Vindonissa,
between AD 30 and 101)); . . . In genesi: et temptavit deus Abraham et dixit ad illum:
accipe filium tuum illum unicum quem dilexisti Isaac et vade in terram altam et
inpones illum hostiam in uno ex montibus, de quo tibi dixero. (Cypr. ad Quir. 3.15);
Related to this directive use of the simple future indicative in interactive communi-
cation is its use in didactic texts, where it freely alternates with other directive
expressions. An example is (d). (For the use of the simple future in relative clauses,
see § 18.21.)81
(d) . . . dissolves tubulum et circa medium fere crus, ubi maxime videbuntur
coisse, vitem serra praecidito et plagam levato terramque minutam aggerato,
ita ut tribus digitis alte plagam operiat.
(‘ . . . you will break up the pipe and cut the vine with a saw about the middle of the
stem, where it shall seem to have formed the closest mass, and smooth off the cut and
heap up fine soil, so that it covers the cut to a depth of three inches.’ Col. Arb. 9.2)
81
See Önnerfors (1989: 134) on Late Latin medical texts; Adams (1995b: 460–8) on Pelagonius and
didactic texts in general; Grocock and Grainger (2006: 98–102) on Apicius.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Conpletis sex et triginta mensibus, ad recipiendam vitem formabis, supervacuos
ramos amputabis, alterna brachia in modum scalarum relinques alternisque annis
putabis; sexto anno, si iam firma videbitur, maritabis hoc modo. (Col. Arb. 16.3);
Cum spissaverit, adicies cerebella duo cocta et selibram pulpae quasi ad isicia liatae,
cum cerebellis teres et in caccabum mittes. (Apic. 5.1.4); Si iumentum cambam
percussam habuerit . . . sic curabis. (Mulom. Chir. 47);
The simple future is common in republican laws and continues to be used in legal
texts until the Late Latin period. An example is (e).82
(e) Ideoque sublimitas tua nullum omnino faciet praesidatus praerogativa
laetari . . .
(‘Therefore, Your Sublimity shall provide that no person at all shall enjoy the
prerogative of governorship . . . ’ Cod. Theod. 12.1.4)
In the Latin translations of the Bible, the directive use of the simple future is quite
frequent, so for example in the Ten Commandments, and ecclesiastical authors,
inspired by biblical usage, continue to use it in this way. This categorical use is
different from the examples given above; in the Old Testament it represents the
continuation of the Septuagint’s attempt to render the Hebrew form of prohibition as
literally as possible.83
(f ) Non occides. Non moechaberis. Non furtum facies.
(‘You will not kill. You will not commit adultery. You will not engage in theft.’ Vulg.
Exod. 20.13–15)
(g) Cum oratis, non eritis sicut hypocritae.
(‘When you pray, you will not be in the same way as actors.’ Vulg. Mat. 6.5)
(h) Non ibis in circum . . .
(‘You will not go to the circus . . . ’ Tert. Spect. 3)
7.24 Periphrastic expressions with the future participle in -urus + a form of sum
From Early Latin onwards, combinations of the future participle in -urus and present
indicative forms of the verb sum (‘to be’) are used to assert that someone is at this very
moment about to or fated to do something and, in the case of controlled events, has
the intention and determination to do or not to do something. With respect to this
value of immediacy, fate, or intention, the -urus + sum expression differs from the
simple future, which only asserts that something will take place. Some scholars speak
of the ‘modal’ value of the periphrastic expression. In this syntax the term PROSPECTIVE
82
For the use of the ‘prescriptive’ future in laws, see Vidén (1984: 6–11).
83
See L. Löfstedt (1966: 152–62) and Vidén (1984: 9–11), who observes that the negated (with non)
future indicative becomes the counterpart of the present imperative. I owe the formulation in the text to
Roland Hoffmann (p.c.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
will be used instead.84 Exx. (a) and (b) are instances of the prospective value coming
to the fore; sometimes, as in (c), the prospective value of the periphrastic expression is
not as prominent. Note that there is no proper passive counterpart of the -urus + sum
expression (see below).
(a) Hoc, si facturu’s, face.
(‘If you’re going to do this, do it.’ Pl. Per. 146)
(b) Ne miremini. / Ipse hanc acturu’st Iuppiter comoediam.
(‘Don’t be surprised: Jupiter himself is going to act a part in this comedy.’ Pl. Am.
87–8—NB: Hanc fabulam, inquam, hic Iuppiter hodie ipse aget / et ego una cum illo.
(Pl. Am. 94–5))
(c) Atque id futurum undeunde dicam nescio, / nisi quia futurum est.
(‘And I don’t know where I should say it’ll come from, except that come it will.’ Pl. Ps.
105–6)
Supplement:
Quia id quod neque est nec fuit nec futurum est / mihi praedicas. (Pl. Am. 553–4);
Sed quid futurum est, quom hoc senex resciverit? (Pl. Bac. 358); Quiquomque ubi
sunt, qui fuerunt, quique futuri sunt posthac / stulti . . . (Pl. Bac. 1087–8); Quid
facturu’s? # Iam scies. (Pl. Cur. 75); Futura es dicto oboediens an non patri? # /
Futura. (Pl. Per. 378-9); Immortalis est. / Vivit victuraque est. (Pl. Trin. 55–6); Non
tu istuc mihi dictura aperte es quicquid est? (Ter. Eu. 819); . . . alterum, quod (sc.
apes), cum iam evolaturae sunt aut etiam inceperunt consonant vehementer, proinde
ut milites faciunt, cum castra movent. (Var. R. 3.30); Attentos habebimus . . . si
numero exponemus res quibus de rebus dicturi sumus. (Rhet. Her. 1.7); Quoquo
enim modo nos gesserimus, fiet tamen illud quod futurum est. (Cic. Div. 2.21); Hoc
enim ille exspectat videlicet, neque est facturus quicquam nisi de meo consilio. (Cic.
Att. 13.31.3); . . . tamen, si est bellum civile futurum (quod certe erit si Sextus in armis
permanebit, quem permansurum esse certo scio), quid nobis faciendum sit ignoro.
(Cic. Att. 14.13.2); Tibi enim uni credidimus et credituri sumus. (Brut. Cas. Fam.
11.2.2); Neque ego istud facturus sum et te magnopere ut de negotio desistas
adhortor. (B. Afr. 45.4); Bellum scripturus sum quod populus Romanus cum Iu-
gurtha rege Numidarum gessit . . . (Sal. Jug. 5.1); Supraque terram parietes extruantur
sub columnas dimidio crassiores quam columnae sunt futurae . . . (Vitr. 3.4.1);
Quantum possumus, causam eius apud nos agamus: ‘fortasse non potuit, fortasse
ignoravit, fortasse facturus est’. (Sen. Ben. 7.29.2); Hac quidam utendum semper
putant, quod ea fiat causa lucidior et iudex attentior ac docilior, si scierit et de quo
dicimus, et de quo dicturi postea sumus. (Quint. Inst. 4.5.1); ET SI QUID MISSURUS ES
INSCRIBE OMNIA ET / SIGNA MIHI SCRIBE IN E<P>ISTULA . . . (CEL 141.23–4 (Karanis, c. AD
115)); Petiturus sum enim, ut rursus vaces sermoni quem apud municipes meos habui
84
For this function of the -urus sum future, see Fleischman (1982: 36, 95–9). She compares it to
English ‘to be going to’ futures. Garuti (1954) has a collection of examples in Early Latin. For mostly
Classical examples, see Neue-W.: III.159ff. See also Westman (1961: 37–89) for a description of the
meanings of the finite periphrastic future forms in Seneca. For a detailed description of the ‘modal’ uses of
the -urus sum future, see Orlandini (2005a).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
bybliothecam dedicaturus. (Plin. Ep. 1.8.2); Nam cum resurrectio ipse sit, quia in illo
resurgimus, sic et regnum dei potest ipse intellegi, quia in illo regnaturi sumus. (Cypr.
Dom. orat. 13); UT DULCIA MELLE DIU DURENT: Accipies quod Graeci dicunt
cnecon et facies farinam et admisces cum melle eo tempore quo dulcia facturus es.
(Apic. 1.14); Idem dixisse fertur: ‘si habuero heredem, dabo illi tutorem qui illum haec
facere cogat quae ipse feci facturusque sum’. (Hist. Aug. Heliog. 32.3); Quid tibi
promisit deus, o homo mortalis? Quia victurus es in aeternum. (August. Psal. 148.8);
Semper dicimus et diximus et dicturi sumus. (Vict. Vit. 3.17); Cui papa Iohannes ita
respondit: ‘Quod facturus es, rex, facito citius. Ecce in conspectu tuo adsto’. (Exc.
Vales. 2.89);
NB: Qui sunt qui erunt quique fuerunt quique futuri sunt posthac, / solus ego
omnibus antideo, facile miserrumus hominum ut vivam. (Pl. Per. 777–8); Amabo,
an maritu’st? / Neque est neque erit. (Pl. Mer. 538–9); Nec fuit neque erit neque esse
quemquam hominem in terra arbitror / quoi fides fidelitasque amicum erga aequi-
peret tuam. (Pl. Trin. 1125–6);
The periphrastic expression differs from the simple future in several other ways as
well. Unlike the simple future, it is not used to express assumptive and deductive
epistemic modality,85 nor is it used in declarative sentences with a directive illocut-
ionary force. It is found, however, just like the present indicative, in interrogative
sentences with a binding directive illocutionary force, as in (d).
(d) Abi modo / . . . Etiamne abis? / . . . Abiturusn’ es? / . . . Pergi’n pergere?
(‘Just go! / . . . Won’t you go away? / . . . Will you not go away? / . . . Are you continuing
to go on?’ Pl. Poen. 430–3)
The periphrastic future is indeed especially common both in direct and in indirect
questions, which often are about the intentions of the addressee.86 Pliny the Elder uses
only the simple future form dicemus ‘I will tell’ (ninety-nine times), never dicturi
sumus: it is inappropriate to announce an actually following topic merely as one’s
intention to deal with it.
The combination of the present indicative of sum with the future participle is part
of a complete paradigm of future participle + finite (indicative and subjunctive) and
non-finite forms of sum. Examples of the other finite and of the perfect infinitive
forms of sum are given below (for the future infinitive with sum, see § 7.72). In all
these instances the element of immediacy, fate, or intention (the prospective value) is
present. The combination with the imperfect and perfect resembles the use of the
Romance conditionnel (future of the past, see § 7.33).87
Supplement:
Future indicative: Ubi sementim facturus eris, ibi oves delectato. (Cato Agr. 30);
Altera est in qua rerum earum de quibus erimus dicturi breviter expositio ponitur
distributa. (Cic. Inv. 1.31); Quod si eris facturus, velim mihi scribas, ut ego, si ullo
85
See Kooreman (1995). 86
See figures in Baños (1999: 79–80, n. 40).
87
See Orlandini (2005a: 69) and Bourova (2008: 278).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Exx. (e) and (f ) contain periphrastic future expressions in the subjunctive. Latin lacks
a synthetic subjunctive form of the future, and the use of the periphrastic expressions in
these contexts might therefore be regarded as mere suppletion of a gap in the paradigm.
However, the typical value of the periphrastic expression is present in both examples. In
(g), by contrast, it is difficult to read intention or fate in dicturi sint; in this case the
periphrastic seems to be used only because another appropriate form is lacking.
(e) Non enim debeo dubitare iudices quin . . . etiamsi is invidiosus aut multis
offensus esse videatur, etiamsi eum oderitis, etiamsi inviti absoluturi sitis,
tamen absolvatis et religioni potius vestrae quam odio pareatis.
(‘For I ought not to doubt, judges, that . . . even if he seems to be unpopular or
offensive to many, even if you hate him, even if you would unwillingly acquit him,
nevertheless, you would acquit him and you would obey your scruples rather than
your hatred.’ Cic. Clu. 158)
(f ) (adiecit se) . . . petere ut ex ea pecunia quam in triumpho latam in aerario
positurus esset id aurum secerni iuberent.
(‘(He added that) he requested that they order that this gold be separated from this
money which, once it was carried in his triumph, he would deposit in the treasury.’
Liv. 39.5.7)
(g) . . . quamquam suspicor quid dicturi sint quidam eorum . . .
(‘ . . . though I suspect what some of them will say . . . ’ Cic. Phil. 5.5)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The periphrastic expression remained in use throughout the history of the Latin
language. Given the productivity of the formation, it is surprising that it did not
survive in the Romance languages.88
Early examples cited to show that the periphrastic future gradually becomes
synonymous with the futurum simplex are found in the speech of the freedman
Echion in Petronius, for instance (h). Further examples are cited from ecclesiastical
authors like Cyprian, who, however, does not use it as an equivalent of the synthetic
future in his elevated writings. In Latin translations of the Greek Bible, the two forms
are still used distinctly, and there is much variation from one translation to another.
It is not until the fourth/fifth century that the use as synonymous with the simple
future becomes frequent in higher literature. However, even then the simple future
remained the dominant future form.89 The use of the participle in -urus with a
simple future indicative form of sum, as an equivalent of the synthetic simple
future indicative, is rare. The earliest instances are found in Tertullian. One such
example is (i).
(h) Et ecce habituri sumus munus excellente in triduo die festa.
(‘Just think, we will have an outstanding spectacle for three days.’ Petr. 45.4)
(i) . . . ipse erit Christus qui Iohannem erit subsecuturus (subsecutus cj. Kroy-
mann) ut antecursorem.
(‘ . . . it will be Christ himself who will follow John as a vanguard.’ Tert. Marc. 4.33.8)
Supplement:
Ferrum optimum daturus est, sine fuga, carnarium in medio ut amphitheater videat.
(Petr. 45.6); Sed subolfacio quia nobis epulum daturus est Mammea, binos denarios
mihi et meis. (Petr. 45.10 (Echion speaking)); Dic illi, inquit, securus sit, quia pax
ventura est. (Cypr. Ep. 11.6);
As was noted in the beginning of this section, this type of periphrastic expression
only exists in the active. From Tertullian onwards, this ‘gap’ in the paradigm was
filled by the periphrastic expression consisting of the future tense form of habeo +
present passive infinitive, as in (j). See also §§ 7.26–7.29.
(j) Nam et homo interior hic utique renovari habebit . . .
(‘For the inward man will certainly have to be renewed here . . . ’ Tert. Resurr. 40.7)
Appendix: The periphrastic construction is also found in conditional periods, as in
(k). It is also used instead of a potential subjunctive, as in (l).90
88
For suggestions how this might be explained, see Herman (1996b) and Vincent and Bentley (2001).
Bourova (2008: 279–80) suggests that the irregular future participle forms constituted an inherent
weakness.
89
See Kooreman (1996) for the specific value of the -urus + sum expression. Intentionality is still the
dominant feature of the expression in Augustine (Roca 1999: 163). The -urus expression is systematically
used for translating Greek ººø + infinitive, but also for normal Greek future forms (Herman 1996b).
90
See Bourova (2008: 276–8).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
7.25 Periphrastic future expressions with the gerundive + forms of the verb sum
The gerundive is occasionally found in Late Latin with finite forms of sum without its
common deontic meaning; this is not restricted to the present tense forms of sum. In
(a)—an early example—inscribendus est has future reference, as is clear from the
following future form loquemur. However, the deontic value is still common in
Jerome.91 (For the use of the gerundive as a passive future participle and (with esse)
as a future passive infinitive, see § 7.84 and § 7.74, respectively.)
(a) Nam et multa iam in Valeriani vita <dicta sunt, multa> in libro qui de
triginta tyrannis inscribendus est iam loquemur . . .
(‘For much has already been said in the Life of Valerianus, and we shall tell many
things in the book which is to be entitled “Concerning the Thirty Pretenders” . . . ’
Hist. Aug. Gall. 19.7)
Supplement:
. . . his qui baptizandi sunt . . . (Filastr. 82.4); Torquendus es, quia homicida es . . .
(Salv. Eccl. 1.35);
Hannibal . . . cum tradendus Romanis esset, venenum bibit et apud Libyssam in
finibus Nicomedensium sepultus est. (Eutr. 4.5.2);
Appendix: This gerundival expression is also found in the apodosis of conditional
periods, as in (b).
91 92
See Roca (1999: 166). For depo cantare, see Blasco Ferrer (1986: 123–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
7.27 Habeo+infinitive
The first instance of habeo governing an infinitive is (a). Whereas haec in (b) might be
regarded as the object of habui (‘I had this to tell you’), resembling expressions like (c)
and (d) that are found from Early Latin onward, the indirect question quem . . .
deiecerit in (a) can only be the object of dicere. Cicero restricts the combination to
communication verbs; Augustan authors widen the range of verbs to other lexical
classes and to intransitive verbs. Ex. (e) shows a use of the perfect indicative habui
that is typical of modal verbs (the combination is more or less equivalent to quid
facerem, see § 7.44).
93 94
See van Eeden (1985). Pinkster (1985b), (1987a), and (1989).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In some of the examples a paraphrase with ‘can’ is possible or likely (the so-called
‘abilitive’ meaning95), in others one with ‘must’ (the so-called ‘obligative’ or ‘deontic’
meaning). The instances which can only be interpreted as ‘deontic’ occur later than
the ‘abilitive’ ones (from Seneca the Elder onwards). A remarkable increase both in
frequency and in the range of the construction can be observed in the Vetus Latina
Bible translations and in Tertullian, clearly influenced by the use of åø in Greek (and
the Greek originals), especially in prophetic contexts. From then on, instances,
though still relatively few in number, are found where the expression refers to the
future. In some of these instances, the habeo construction serves as a suppletive form
for gaps in the Latin paradigm such as the future passive participle in (f )—note the
parallelism with the future active participle aperturus, the future passive infinitive,
and the future in the past (see § 7.33). Exx. (g) and (h) refer to the same Greek text.
The quotation from Porphyrio’s comment on Horace in (i) (dated to the early third
century AD, but the observation itself may be later) shows that the expression was
spreading outside ecclesiastical authors. The same is shown by (j) from the third-
century grammarian Sacerdos.
95
For ‘abilitive’ modality, see Palmer (2001: 76ff.), for ‘obligative’ ibidem: 71ff. Bourova (2007) uses the
term ‘modalité aléthique’ (more or less equivalent to ‘deductive’) as a third category next to ‘abilitive’, as
opposed to ‘deontic’.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The fact that there are instances of the habeo + infinitive expression that seem to be
equivalent to simple future tense forms does not imply that the habeo + infinitive
expression covers the whole range of uses of the simple future that are mentioned in
§ 7.22. In Tertullian, for example, there are no instances of assumptive and deductive
use of habeo + infinitive (note that the same holds for the -urus + sum expression).96
The synthetic form (and the periphrastic -urus expression) remained the most
common throughout the history of the Latin language. Habeo is still used as a future
auxiliary in Portuguese. In most Romance languages, however, a new synthetic future
form arose from the fusion of the present infinitive and the present indicative forms
of habeo. For example, the combination stare + habeo led to Italian starò. This
development presupposes fixation of the relative order of the infinitive and the
forms of habeo and also immediate contiguity.97 This is at first sight remarkable,
96
See Kooreman (1995: 386). For the deductive use of the simple future, see § 7.22, ex. (m); for -urus +
sum, see § 7.24.
97
In Portuguese the infinitive and the auxiliary can be separated by object pronouns.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
since in Late Latin the preferred order of modal verbs like debeo ‘to be obliged to’,
‘must’ and possum ‘to be able to’, ‘can’ + infinitive was just the reverse (see
Chapter 23). However, Tertullian prefers the order infinitive + habeo when he uses
habeo to refer to the future, and he allows only clitic words in between. This order
resembles the typical order of periphrastic expressions consisting of a participial or
gerundival verb form and a form of the verb sum ‘to be’, such as factus est ‘he was
made’ (see Chapter 23). The grammarian Pompeius (c. AD 500) reserves the order
infinitive + habeo for the obligative use of the expression. This obligative use is usually
regarded as the basis for the development into its use as an expression of futurity, but
this is not necessarily the case.98 The chronology of the development is a matter of
dispute. The earliest attested new synthetic forms in French are salvarai ‘I will
preserve’ and prindrai ‘I will take’ in the Oaths of Strassburg (AD 842).
Leumann (1962) is a succinct description and an attempt to explain the development
of habeo as an auxiliary. He assumes that Cicero introduced his extension of the use
of habeo following the use of ε῎wø in Greek, but this need not be the case, given the
availability in Latin of related expressions. A full survey of existing theories about the
origin and development of the construction can be found in Fleischman (1982: 52ff.).
See also Pinkster (1987a) and Târa (2014). The TLL s.v. habeo 2454.12ff. has a good
survey of the material.
Adams (1991) shows that the order habeo + infinitive is open to more interpret-
ations than the reverse order. Pompeius has twice as many instances of habeo +
infinitive, which is the most common order for auxiliaries in general. Nocentini
(2001) observes that the infinitive + habeo order in Pompeius correlates with the
position of the infinitive in its clause (it is in the first available position—P1, see
Chapter 23). Benveniste (1960) and Bybee et al. (1994: 258–64) draw attention to the
frequent use of other forms of habeo + infinitive than the present tense and to the
frequency of the passive infinitive; accordingly, they do not consider the deontic
instances as essential in the evolution of the new synthetic future. See also Bourova
(2008). The neutral modal meaning of habeo (as opposed to debeo and possum) as a
factor in its success is stressed by Pinkster (1989: 316). Bourova (2007) shows that,
among the 671 instances she has collected in Christian texts, the order infinitive +
habeo form has much more often a deductive sense (‘aléthique’ in her terminology),
whereas the deontic interpretation is found much more often with the order habeo
form + infinitive. More details about word order can be found in Chapter 23.
The two cases claimed to be the earliest Romance uses are pussediravit (‘shall possess’,
on a late sixth-century Merovingian buckle) and Et ille respondebat: Non dabo. Iusti-
nianus dicebat: Daras (‘you will give’ Fredeg. 85.27 (early seventh century)).99
Just as with the other auxiliaries, in principle all forms of habeo could be used with
the infinitive. In fact, in the beginning the use of the present indicative was in
the minority, and its spread may have followed the ‘success’ of other forms as
98
For Tertullian, see Raiskila (1990), Kooreman (1995), and Târa (2014: 269–98). For Pompeius, see
Adams (1991). For the ‘obligative’ origin of the habeo future, see especially Fleischman (1982: 56–9).
99
For a sceptical view on the early instances, see Posner (1996: 177–9). See also Fleischman (1982: 53–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
auxiliaries.100 Historically, the most interesting are the imperfect and perfect indica-
tives, which together with the infinitive serve as a substitute for the ‘future of the past’.
It contributed to what is known in French as the ‘conditionnel’. An example is (k). See
§ 7.33 for a few further examples. The majority of the instances until c. AD 300 have a
passive infinitive, as in (k).101
(k) Nazaraeus vocari habebat secundum prophetiam Christus creatoris.
(‘The Christ of the Creator was to be called a Nazarene according to the prophecy.’
Tert. Marc. 4.8.1)
Appendix: The combination of the imperfect form of habeo + infinitive is also used
instead of a potential subjunctive, as in (l), and therefore also used in the apodosis of
a conditional period. The perfect of habeo is used as well, most often in the apodosis
of a counterfactual conditional period, as in (m).102
In (a) we have debeo in its ‘deductive’ use (see § 7.22, note). In (b), the use of an
auxiliary verb (here: debeo) redundantly copies the governing verb (praecipiunt) (see
also below on volo).
As an instance of future use Väänänen (1981: 132) cites Eugipp. [early VI AD] Vit.
31.4 . . . oppida in quibus debeant ordinari. This example can be interpreted as
deontic in its context. For the phenomenon of ‘harmonic’ parallelism between a
governing clause and a subordinate clause see Bybee et al. (1994: 212ff.).
100 101
See Fleischman (1982: 55–6). See Bourova (2005: 306); Fruyt and Orlandini (2008).
102
See Bourova (2007) and Adams (2013: 660–5). For more examples, see TLL s.v. habeo 2458.21ff.
103
See Pinkster (1985b: 190–2) for discussion. For ‘weakened’ debeo, see Sz.: 314.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The evidence for the development of volo + infinitive into a future tense auxiliary is
not convincing either.105 Instances brought forward to support the idea that volo was
on its way to becoming a future tense auxiliary are (e)–(h). In (e (i)) the manifest
contextual differences in comparison with (e (ii)) are being overlooked. Ex. (f ) is
mock-bureaucratic language—note also the futurum exactum form voluerit; and (g)
shows the type of redundancy, velle after nititur, mentioned in regard to debere above.
Even (h), dated AD 550 and mentioned in the literature as the first ‘real’ future tense
use of velle, is not a strong case.
104
See Pinkster (1985b: 192–7) and López Fonseca (2000). Touratier (1994: 158, 557) does regard
possum + infinitive as an alternative periphrastic expression of the future.
105
See Pinkster (1985b: 187–90) for discussion.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
When using the perfect indicative tense, the speaker asserts that an event has taken
place before the moment he is uttering the assertion. The perfect is thus essentially
grounded in the time of speaking, just like the present indicative and the (simple)
future indicative. Typical instances of this use in simple sentences are (a)–(c). Exx. (d)
and (e), in turn, show the same use in subordinate clauses.
(a) Vale’n? Valuistin’? # Valeo et valui rectius.
(‘Are you well? Have you been well? # I’m well, but I’ve been better.’ Pl. Trin. 50)
(b) Ultra fuit oppidum Pirae, est colonia Minturnae . . .
(‘Beyond there was the town of Pirae; there is (now) the colony of Minturnae . . . ’
Plin. Nat. 3.59)
(c) D(IS)·M(ANIBUS). / HIC·IACET·FIRMIDONIUS. / FIRMIDONIA·FILIA·SUA / HANC SEPUL-
TURAM / PROCURAVIT.
(‘To the spirits of the dead. Here lies Firmidonius. His daughter Firmidonia attended
to this burial.’ CIL II.4365.1–5 (Tarraco))107
(d) Sed quom cogito, equidem certo idem sum qui semper fui.
(‘But when I think about it, I’m certainly the same I’ve always been.’ Pl. Am. 447)
106
References in Sz.: 313. Material in TLL s.v. incipio 919.70ff.
107
In dedicatory inscriptions the perfect is the normal tense. There are a few ‘actual’ presents. See
Horster (2001: 54).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(e) Quae illi ad legionem facta sunt memorat pater / meus Alcumenae . . .
(‘My father is telling Alcumena what happened there during the campaign.’ Pl. Am.
133–4)
Supplement:
Tune heri hunc salutavisti? # Et te quoque etiam, Sosia. (Pl. Am. 717); Quam dudum
tu advenisti? (Pl. As. 449); Nunc intellexi. (Pl. Cist. 624); Illa exclamat ‘abi, abi iam,
Aeschine: / Satis diu dedisti verba, sat adhuc tua nos frustrata’st fides’. (Ter. Ad.
620–1); Quantam optuli adventu meo laetitiam Pamphilo hodie! / Quot commodas
res attuli! Quot autem ademi curas! (Ter. Hec. 816–17); Te, Fortuna, sequor. Procul
hinc iam foedera sunto. Credidimus fatis. Utendum est iudice bello. (Lucan.
1.226–7);108 Nec tamen in triclinio ullum vetui (vetuo cj. Bücheler) facere quod se
iuvet, et medici vetant continere. (Petr. 47.5 (Trimalchio speaking));
Especially in poetry, but not only there, the perfect is often used to present an event
as anterior to another one expressed by a historic present, that is, instead of a
pluperfect, as in (f ).
Since the perfect indicative tense presents an event on the one hand as grounded in
the time of speaking, on the other hand as anterior, subordinate clauses with a main
clause in the perfect may contain both present tense forms (and perfect forms
expressing anteriority to the moment of speaking) and imperfect tense forms. In ex.
(g), suspicio belongs to the time of speaking, hence the present tense apscedat. In (h),
the intentions of the woman are anterior to the communicative situation. In (i), the
need to be careful was relevant in the past, hence imperfect caveres.
(g) Et repperi haec te qui apscedat suspicio.
(‘And I’ve found a way to make any suspicion in that matter go away from you.’ Pl.
Epid. 285)
(h) Modo intellexi quam rem mulier gesserit.
(‘I’ve just realized what the girl has been doing.’ Pl. Mil. 867)
108
For the abundant literature on credidimus, see Gärtner (2011).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
States of affairs that are presented as anterior to the time of speaking are at the same
time presented in retrospect and as one complete, indivisible whole—in other words,
they have perfective aspect (see § 7.3). That this is indeed the case appears from the
fact that expressions that can mainly or only be given in retrospect occur much more
often with the perfect than with the imperfect or the historic present. Expressions of
this kind are negative expressions (e.g. non ‘not’ in (j)); judgements evaluating the
outcome of an event (e.g. prospere ‘successfully’ in (k)); frequency expressions (e.g. bis
‘twice’ in (l)); and duration expressions (e.g. diu in (m)). Perfect dixi ‘I have come to
the end of my speech’ is the appropriate form for concluding a speech.
109
For the origins of the Latin perfect forms, see Meiser (2003). On syncretism, especially pp. 83–5.
For the -u/w- perfect forms, see Seldeslachts (2001) and de Melo (2007c). For a discussion of the uses of
the Latin perfect from a comparative perspective, see Haverling (2002).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The retrospective character of the perfect also explains why verbs meaning ‘to stop’
and ‘to complete’ (such as desino and perficio) occur much more often in the perfect
than in the historic present; the same holds true for verbs meaning ‘to have an
opinion’ like puto. Along the same lines, it is understandable that consecutive ut
clauses following expressions like tantus ‘so great’, ita ‘such’, adeo ‘to such a high
degree’ are much more frequent with main clauses in the perfect than with those in
the historic present (sixty-eight versus thirteen in Caesar). In addition, sentences with
the connectors sed ‘but’ and tamen ‘yet’, which often imply the interruption of
another event, occur more frequently in the perfect than in the historic present (in
Caesar fifty-two versus seventeen and twenty-five versus seven, respectively), and
nam ‘for’ occurs ten times more frequently in sentences with a perfect tense form than
in sentences with a historic present.110
110
These data come from Schlicher (1931). For ergo ‘therefore’ and igitur ‘consequently’ in sentences
that contain a perfect, see Kroon (1998: 52).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Depending on the type of state of affairs, perfect forms may receive special inter-
pretations. Three interpretations are discussed in the following pages: (i) the so-called
‘negative’ use of the perfect, (ii) the ‘resulting state’ interpretation of the perfect, and (iii)
the so-called ‘ingressive’ use of the perfect (for the types of state of affairs, see § 2.9).
(i) Especially with non-dynamic states of affairs, an event may be presented as anterior
to the time of speaking in such a way that the two points of time are contrasted.
Thus the event is interpreted as ‘no longer being the case now’. This is sometimes
called the NEGATIVE use of the perfect. Examples are (n) and (o).
(ii) With terminative states of affairs the perfect may be interpreted as describing a
RESULTING STATE. This is demonstrated by (p)–(q) with the verb inscribo. These
two examples refer to two books of Cicero, entitled Laelius and Hortensius. Ex.
(p) is a present state resulting from previous action, whereas (q) is a present
process. The situation they refer to is the same. The use of the perfect passive as a
present state is in fact part of an entire paradigm of forms of the perfectum stem
of sum with participles denoting a resulting state, as clausus fuit in (r) (see also
forms with fuero in § 7.32, ex. (u), and with fueram in § 7.31, ex. (l)).
(p) Nam et cohortati sumus . . . ad philosophiae studium eo libro qui est in-
scriptus ‘Hortensius’ . . .
(‘For we urged to the pursuit of philosophy in that book which has been
named “Hortensius”.’ Cic. Div. 2.1)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(q) Sed de amicitia alio libro dictum est, qui inscribitur ‘Laelius’.
(‘But friendship has been discussed in another book of mine, entitled
“Laelius”.’ Cic. Off. 2.31)
(r) Bis deinde post Numae regnum clausus fuit . . .
(‘It has been closed twice since Numa’s reign . . . ’ Liv. 1.19.3)
Cicero has eight other instances of inscribitur for book titles (all not by himself). The only
other attestation of inscriptus est (Cic. Att. 8.5.2) does not concern a book title. Inscribitur
is also the form that is used most by other authors, including Augustine. Gellius is
an exception with twenty-three instances of inscribitur vs twenty-four of inscriptus est.111
Supplement:
. . . <XVIRI QUEI EX> LEGE LIVIA FACTEI CREATEIVE SUNT FUERUNTVE . . . (CIL I2.585.77 (Lex
Agr., c.111 BC)); Itaque in iis scriptum litteris Punicis fuit regem Masinissam
imprudentem accepisse . . . (Cic. Ver. 4.103); Omnia fere quae sunt conclusa nunc
artibus, dispersa et dissipata quondam fuerunt. (Cic. de Orat. 1.187); Quo est detes-
tabilior istorum immanitas qui lacerarunt omni scelere patriam et in ea funditus
delenda occupati et sunt et fuerunt. (Cic. Off. 1.57); Gallia est omnis divisa in partes
tres, quarum unam incolunt Belgae, aliam Aquitani, tertiam qui ipsorum lingua Celtae,
nostra Galli appellantur. (Caes. Gal. 1.1.1); Idem titulus tribus signis in aede Fortunae
positis fuit subiectus. (Liv. 23.19.18); Non est ergo hic otiosus, aliud illi nomen
inponas: aeger est, immo mortuus est. (Sen. Dial. 10.12.9); Transeamus igitur id
quoque quod grammatice quondam ac musice iunctae fuerunt. (Quint. Inst. 1.10.17);
Instances like (p) above, with a passive participle in combination with sum, are easy to
find. However, perfect active forms that must be interpreted as states resulting from a
previous terminative action or process are less easy to find. The clearest ones are
found in utterances like (s), which can be compared with the passive perfect in (t).
The verbs involved are one-place verbs or two-place verbs used absolutely.
(s) Perii, interii, occidi! Quo curram? Quo non curram?
(‘I’m done for, I’m killed, I’m murdered. Where should I run? Where shouldn’t
I run?’ Pl. Aul. 713)
(t) Occisi sumus.
(‘We’re dead.’ Pl. Bac. 681)
Supplement:
Nulla sum, nulla sum: tota, tota occidi. (Pl. Cas. 621); Non hercle <ego omnino> occidi,
sunt mi etiam fundi et aedes (Pl. Truc. 174); Miles, edico tibi, / si te in platea offendero
hac post umquam, quod dicas mihi / ‘Alium quaerebam, iter hac habui’, periisti. (Ter.
Eu. 1063–5); Perii, quid faciam misera? / Non visam uxorem Pamphili, quom in
proxumo hic sit aegra? (Ter. Hec. 340–1); Si quem animum in alienae sortis exemplo
paulo ante habuistis eundem mox in aestimanda fortuna vestra habueritis, vicimus,
milites. (Liv. 21.43.1); Nisi succurritis, noverca vicit, ego victus sum. (Sen. Con. 9.6.4);
111
Haverling (2010a: 361) has different figures.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
With perfect active forms of two- and three-place verbs, the resulting state interpret-
ation is less common, although grammars suggest otherwise. As proof they refer to
memini ‘to remember’, odi ‘to hate’, and to novi ‘ to know’. The first two of these
are exceptional cases, because they are idiomatic presents and do not denote states
resulting from previous action at all. They cannot be used to refer to an anterior
event. Novi can still be found as the ‘normal’ perfect of nosco ‘to get to know’, as
in (u). Likewise, didici ‘I have learned’ may imply ‘I know’, as K.-St.: I.125 suggest,
but the difference becomes clear from (v), in which didici is used in conjunction
with scio.
(iii) Some scholars assume that with non-dynamic states of affairs the perfect
indicative can be interpreted in an ‘ingressive’, ‘inceptive’, or ‘inchoative’ sense.
Examples cited to support this idea are (y)–(ac).
(y) Sed postquam L. Sulla armis recepta re publica bonis initiis malos eventus
habuit, rapere omnes . . .
(‘But after Lucius Sulla, having taken the commonwealth by arms, had a bad
ending to his good beginning, all plundered . . . ’ Sal. Cat. 11.4)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(z) Cum acerrime comminus pugnaretur . . . subito sunt Haedui visi ab latere
nostris aperto . . .
(‘The battle was continued most fiercely at close quarters, . . . Suddenly the Aedui
were seen on the exposed flank of our troops . . . ’ Caes. Gal. 7.50.1)
(aa) Erant enim circum castra Pompei permulti editi atque asperi colles. Hos
primum praesidiis tenuit castellaque ibi communiit.
(‘Around Pompey’s camp there was a large number of steep, rough hills. These he first
seized, garrisoned and strengthened with forts.’ Caes. Civ. 3.43.1)
(ab) Molestus fuit Philerosque proclamavit . . .
(‘He was a bore, and Phileros shouted out . . . ’ Petr. 43.1)
(ac) . . . et Dorus evanuit et Verissimus ilico tacuit . . .
(‘ . . . and Dorus disappeared, and Verissimus immediately held his peace . . . ’
Amm. 16.6.3)
However, although in some cases one might be tempted to use an ingressive trans-
lation, there was no productive use of the perfect in an ingressive sense, and
convincing instances where an ingressive interpretation is likely are not attested
before Late Latin. In most instances cited from Classical and later texts, there are
contextual signals favouring an ingressive interpretation, such as the occurrence in a
temporal clause with postquam ‘after’ in (y), or the presence of an adverb like subito
‘suddenly’ in (z). This is also the case in the Late Latin ingressive (ac) with ilico
‘immediately’.112 Tenuit in (aa) is sometimes interpreted as ingressive, but it is used
for the same reason as the perfect communiit (the preferred reading) in the coordin-
ated clause. In (ab), translating fuit in an ingressive sense is quite arbitrary.
Supplement:
Sic tribunus spectante utroque exercitu et sua virtute nixus et opera alitis propugna-
tus ducem hostium ferocissimum vicit interfecitque atque ob hanc causam cognomen
habuit Corvinus. (Quad. Hist. 12 ( = Gel. 9.11)); Ibi enim adulescens habitabat cum
patre et fratribus, quod in eam pater eius Neocles agripeta venerat, sed cum agellus
eum non satis aleret ut opinor, ludi magister fuit. (Cic. N.D. 1.72); Ubi iam se ad eam
rem paratos esse arbitrati sunt, oppida sua omnia . . . incendunt . . . (Caes. Gal. 1.5.2);
Ita pugnans post paulo concidit ac suis saluti fuit. (Caes. Gal. 7.50.6); Caesar ubi luxit
omnes senatores senatorumque liberos tribunos militum equitesque Romanos ad se
produci iubet. (Caes. Civ. 1.23.1); Ita parvae res magnum in utramque partem
momentum habuerunt. (Caes. Civ. 3.70.2); Pacata provincia vectigalia magna in-
stituit ex ferrariis argentariisque, quibus tum institutis locupletior in dies provincia
fuit. (Liv. 34.21.7); Nec cibo tantum et alimentis necessarii, quippe vocis sermonis-
que regimen primores tenent, concentu quodam excipientes ictum linguae serieque
structurae atque magnitudine mutilantes mollientesve aut hebetantes verba et,
cum defuere, explanationem omnem adimentes. (Plin. Nat. 7.70); Ante aedem divi
112
See Haverling (1998, 2002: 163); Oldsjö (2001: 472–88) assumes the existence of ingressive perfects
in Classical Latin, but see Haverling (2006a).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Iulii iacuit primo ictu in poplitem, mox ab Iulio Caro legionario milite in utrumque
latus transverberatus. (Tac. Hist. 1.42); Et postquam tacuerunt respondit Iacobus
dicens . . . (Vulg. Act. 15.13); . . . videamus si floruit vinea . . . (Vulg. Cantic. 7.12); . . .
iacuit mucrone suorum. (Claud. 21.243); Matusalas . . . Lamech filium habuit, cum
esset annorum octoginta septem et centum . . . (Joseph. Antiq. 1.3.4 [86]);
The normal way of indicating the start of an event in the past is by using coepi
(perfect of the phasal verb incipio ‘to begin’) with an infinitive. Some scholars hold
that in Late Latin the combination served as an analytic counterpart of the ingressive
use of the Greek aorist, but in reality it corresponds with other tenses as well.113
The perfect indicative is also used for describing general truths and good practices.
This is relatively common in poetry, probably stimulated by the use of the aorist in
Greek in similar contexts. This is sometimes called the ‘gnomic’ use of the perfect.
An example is (ad). This gnomic use is also found in Bible translations following the
Hebrew original.114
(ad) Illius (sc. segetis) immensae ruperunt horrea messes.
(‘From that (field) boundless harvests burst the granaries.’ Verg. G. 1.49)
Supplement:
Sed simul ac cupidae mentis satiata libido est, / dicta nihil metuere, nihil periuria
curant. (Catul. 64.147–8); Qui studet optatam cursu contingere metam / multa tulit
fecitque puer, sudavit et alsit, / abstinuit Venere et vino. (Hor. Ars 412–14); Nam et
pueri os parentium feriunt et crines matris turbavit laceravitque infans et sputo
adspersit aut nudavit in conspectu suorum tegenda et verbis obscenioribus non
pepercit, et nihil horum contumeliam dicimus. (Sen. Dial. 2.11.2); Conserva me,
Domine, quoniam speravi in te. (Vet. Lat. Psalm. 16.1);
In the cases of the perfect indicative tense discussed so far, the past events referred to
in the perfects are mostly isolated events that are presented as anterior to the time of
speaking or writing. This is the most common use in text types that most directly
reflect actual speech behaviour, in particular monologues and dialogues in comedy,
speeches to an audience in court, and descriptive texts. Of the sixty perfect indicative
forms in simple sentences or main clauses in the first three hundred lines of the
Amphitruo of Plautus, fifty-one are of this type. In addition, fifteen perfect indicative
forms in this passage occur in subordinate clauses, and most of these are also
understood as anterior to the time of speaking.
However, the perfect indicative tense is also often used in more complex situations
in which a number of events are involved. In such contexts the relationship with the
time of speaking plays a minor role. The mutual relationship of the past events is
more important. A natural way of arranging such events in a text is copying the order
113
See Rosén (2012a), with discussion of earlier studies such as Reichenkron (1957); see also Rosén
(2012b, 2012c).
114
For exceptional use of the tenses in biblical Latin under Hebrew influence, see Rubio (2009:
219–20).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
of the events as they occurred one after the other (the so-called ‘iconic’ order). A good
example is (ae), the famous series of perfect indicative forms Caesar displayed on a
titulus (‘placard’) during his Pontic triumph in 47 BC. Ex. (af) has a sequence of
perfects in an angry speech in Livy; here, however, not all the events are successive.
(ae) Veni, vidi, vici.
(‘I came, I saw, I conquered.’ Suet. Jul. 37.2)
(af) In praetorio tetenderunt Albius et Atrius. Classicum apud eos cecinit.
Signum ab iis petitum est. Sederunt in tribunali P. Scipionis. Lictor apparuit.
Summoto incesserunt. Fasces cum securibus praelati sunt.
(‘Albius and Atrius camped in the headquarters. At their tent the trumpet sounded.
A signal was looked for from them. They sat on Publius Scipio’s tribunal. A lictor
appeared. They advanced on a cleared path. Fasces with their axes were carried
before them.’ Liv. 28.27.15)
Before discussing more details of this use of the perfect, it is useful to look into the
concept of ‘narration’ or ‘storytelling’.115 A distinction may be made between at least
two forms of talking or writing about the past, viz. ‘reporting’ and ‘storytelling’.
REPORTING simply relates events that have taken place before the situation in which
they are being reported, while STORYTELLING normally involves supplying background
information, showing the internal coherence of the events, indicating how they can or
should be evaluated, making them exciting for the hearer/reader, etc. In storytelling a
variety of verbal forms can be used: the imperfect and pluperfect indicative and
subjunctive forms to provide background information, the historic present and
narrative infinitive to present the events from a different angle, and participles to
supply information in a more concentrated form. Reporting, by contrast, has a much
more restricted objective. The reporting mode is common in biographies, didactic
texts, and letters, so in these text types perfect indicatives predominate while historic
presents, for example, are very uncommon (so, for instance, in Nepos, especially his
shorter lives, in Tacitus’ Agricola, and in Suetonius). In prose historiography the
situation is more complicated. In Livy, for example, there are not sufficient facts to
‘report’ in the earliest books (hence much storytelling), but there are sections where
he does do so.116 Series of perfect forms are also the norm in summaries, so for
example in the Res Gestae Divi Augusti, the Periochae of Livy, and in Trogus. Other
examples (in addition to the two examples above) of this ‘reporting’ type of text are
(ag)–(ai). Example (ag) is a rare sequence of perfects in the Amphitruo of Plautus,
who normally uses the historic present (see Figure 7.2 in § 7.16).
115
This section is heavily dependent on Schlicher (1931) and Chafe (1994). For the need to distinguish
various types of ‘narratives’ see Kroon (2000) and Adema (2008). For a demonstration of the differences
in narrating past events in an oration and a letter of Cicero see van Gils (2003, table of the tenses on p. 62).
See also Oldsjö (2001: 447–72) for a discussion of the relationship between types of events, narrative
structure, and the use of the tenses.
116
For his use of the present in an ‘annalistic’ way, see ex. (o) in § 7.16 fin.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(ag) Is amare occepit Alcumenam clam virum / usuramque eius corporis cepit
sibi / et gravidam fecit is eam compressu suo.
(‘He fell in love with Alcumena behind her husband’s back, enjoyed her body, and
made her pregnant through his embrace.’ Pl. Am. 107–9)
(ah) Quaeris ex me quid acciderit de iudicio quod tam praeter opinionem omnium
factum sit et simul vis scire quo modo ego minus quam soleam proeliatus sim.
Respondebo tibi oæ æ æ , OÅæØŒH. Ego enim, quam diu senatus
auctoritas mihi defendenda fuit, sic acriter et vehementer proeliatus sum ut
clamor concursusque maxima cum mea laude fierent. Quod si tibi umquam
sum visus in re publica fortis, certe me in illa causa admiratus esses. Cum enim
ille ad contiones confugisset in iisque meo nomine ad invidiam uteretur, di
immortales! Quas ego pugnas et quantas strages edidi! Quos impetus in
Pisonem, in Curionem, in totam illam manum feci! Quo modo sum insectatus
levitatem senum, libidinem iuventutis! Saepe, ita me di iuvent, te non solum
auctorem consiliorum meorum verum etiam spectatorem pugnarum mirifi-
carum desideravi. Postea vero quam Hortensius excogitavit ut legem de
religione Fufius, tribunus pl., ferret in qua nihil aliud a consulari rogatione
differebat nisi iudicum genus (in eo autem erant omnia) pugnavitque ut id ita
fieret quod et sibi et aliis persuaserat nullis illum iudicibus effugere posse,
contraxi vela perspiciens inopiam iudicum neque dixi quicquam pro testimo-
nio nisi quod erat ita notum atque testatum ut non possem praeterire.
Itaque si causam quaeris absolutionis, ut iam æe e æ æ revertar,
egestas iudicum fuit et turpitudo.
(‘You ask me what happened over the trial which turned out so contrary to every-
one’s expectations, and at the same time you want to know how it was that I took less
than my usual part in the fray. I shall answer you Homerically, cart before horse.
Well then, so long as I had the Senate’s authority to defend, I fought so briskly and
vigorously that crowds flocked around me shouting enthusiastic applause. If you ever
gave me credit for courage in public life, you would surely have admired me in that
affair. For when that man had betaken himself speech-making at meetings and used
my name to stir up ill feelings, ye gods, what battles, what havoc I made! What
attacks I made against Piso, against Curio, against that whole band! How I attacked
the unreliability of the old men and licentiousness of the youth! Often, god help me,
I longed for you not only as the adviser of my counsels but also as the spectator of my
amazing battles. But then Hortensius conceived the idea of getting Fufius, the
Tribune of the Plebs, to propose a law on the sacrilege differing from the consular
bill only in respect of the constitution of the jury, on which however everything
turned, and worked hard for its accceptance because he had persuaded both himself
and others that no jury on earth would acquit that man. I saw we had got a jury of
paupers and drew in my horns, saying nothing in evidence but what was so generally
known and attested that I could not leave it out.
If therefore you want to know the reason for the verdict of not guilty (to come
back from cart to horse), it was the needy and disreputable quality of the jury.’
Cic. Att. 1.16.1–2)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In describing more complex situations containing more events, the perfect indicative
tense is usually surrounded by other past tense forms, especially the imperfect and the
pluperfect, as well as by participial clauses, temporal subordinate clauses, etc. Sen-
tences in the imperfect regularly refer to accompanying circumstances, reasons,
considerations, and details. Such combinations of sentences constitute ‘stories’ or
‘narrative episodes’ sensu stricto. Instances of these are (aj) and (ak). In (aj) the
sentence with imperfect habitabat is interpreted as the temporal setting for the
following sentence with perfect dedit (as indicated in Barsby’s Loeb translation). In
(ak) there are two coordinated main clauses; the one with the imperfect censebat is
understood as causally related to the other clause with the perfect disseruit.
This kind of relationship can also be expressed overtly by using a subordinate
clause with a specific subordinator—for example, by means of quod ‘because’, as
in (al). In fact, the imperfect is found most often in subordinate clauses (see § 7.11
and Tables 7.7–7.8).
(aj) Mihi mater Samia fuit. Ea habitabat Rhodi. / # Potest taceri hoc. # Ibi tum
matri parvolam puellam dono quidam mercator dedit / . . .
(‘I had a Samian mother. She was living on Rhodes. # That can be kept secret. # While
she was there, a merchant gave my mother a little girl as a present.’ Ter. Eu. 107–8)
(ak) Ita enim censebat itaque disseruit (Socrates): . . .
(‘For (Socrates) was thinking in such a way and discussed in such a way . . . ’ Cic.
Tusc. 1.72)
117
For a different interpretation of expedivi (= expetivi), see Adams (1977: 31).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
118
For another interpretation of ex. (ap), see Oldsjö (2001: 461).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
For this autonomous use of the tenses from the perspective of the author (and not
from the perspective of the storyline) some scholars use the term ‘absolute’ (as
opposed to ‘relative’).
Supplement:
Reason clause: Ille quod eius avaritiam cives Romani ferre non potuerunt, Uticae
domi suae vivus exustus est, idque ita illi merito accidisse existimatum est ut
laetarentur omnes neque ulla animadversio constitueretur. (Cic. Ver. 1.70); Id ea
maxime ratione fecit, quod noluit eum locum, unde Helvetii discesserant, vacare.
(Caes. Gal. 1.28.4);
Concessive clause: Proditoribus extemplo in vincla coniectis de legatis paululum
addubitatum est, et quamquam visi sunt commisisse ut hostium loco essent, ius
tamen gentium valuit. (Liv. 2.4.7); Sed Apollo, quanquam Graecus et Ionicus, propter
Milesiae conditorem sic Latina sorte respondit: . . . (Apul. Met. 4.32.6);
Relative clause: Scripta iam epistula Capua litterae sunt adlatae hoc exemplo:
‘Pompeius mare transiit cum omnibus militibus quos secum habuit’. (Cic. Att.
9.6.3); Ibi retenti ventis sumus usque ad a. d. VIIII Kal. Interea qui cupide profecti
sunt multi naufragia fecerunt. (Cic. Fam. 16.9.1); Ab iis Caesar haec facta cognovit
qui sermoni interfuerunt. (Caes. Civ. 3.18.5); Arreptisque armis praetores primum
obtruncant, inde ad caedem Syracusanorum discurrunt, quosque fors obtulit, irati
interfecere atque omnia quae in promptu erant diripuerunt; (Liv. 25.29.9); Litte-
raeque quas Iulia patri Augusto cum insectatione Tiberii scripsit a Graccho compo-
sitae credebantur. (Tac. Ann. 1.53.3); Iamque proximas civitates et attiguas regiones
fama pervaserat deam quam caerulum profundum pelagi peperit et ros spumantium
fluctuum educavit iam numinis sui passim tributa venia in mediis conversari populi
coetibus . . . (Apul. Met. 4.28.4);
119
Diomedes I.336.5K. defines the pluperfect in the following way: Item praeteritum plusquamper-
fectum, cum tempus iam pridem exactum demonstramus quo quid egimus (see TLL s.v. exigo 1467.80ff.).
The best collection of examples remains Blase (1894).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
However, such explicit cases are in the minority, and often the reference point of the
pluperfect is implicit, as in (e), where the pluperfect dixeram in the subordinate clause
refers back to what was said in line 126. Similarly in (f )—an independent sentence—
dixeram refers back to several passages in the preceding discussion. In some instances
it may be quite difficult to identify the event with respect to which the pluperfect
signals anteriority.
(e) Ad fratrem quo ire dixeram mox ivero.
(‘Soon I’ll go to my brother, where I said I was going.’ Pl. Capt. 194)
(f ) His ego de causis dixeram, Scaevola, . . . iuris civilis esse cognitionem
necessariam.
(‘On these grounds, Scaevola, did I declare a knowledge of the common law to be
indispensable.’ Cic. de Orat. 1.197)
Supplement:
Non hercle te provideram. Quaeso ne vitio vortas. (Pl. As. 450); Non ipse emam, /
sed Lysimacho amico mandabo. Is se ad portum dixerat / ire dudum. (Pl. Mer.
466–8); Iamne exta cocta sunt? Quot agnis fecerat? (Pl. St. 251); Ehem Demea, haud
aspexeram te. Quid agitur? (Ter. Ad. 373); Quare, si tibi est commodum, ede illa quae
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
coeperas et Bruto et mihi. (Cic. Brut. 20); Haec tecum coram malueram; sed quia
longius fiebat, volui per litteras eadem, ut haberes quid diceres si quando in
vituperatores meos incidisses. (Cic. Fam. 7.3.6); Sed heus tu, manum de tabula!
Magister adest citius quam putaramus. (Cic. Fam. 7.25.1); Una cum his legatis
Commius Atrebas venit, quem supra demonstraveram a Caesare in Britanniam
praemissum. (Caes. Gal. 4.27.2); Istud quod modo dixeram me habere / fugit me
ratio. (Catul. 10.28–9); Plus hic invenio quam quod promiserat illa. (Ov. Her.
16.145); Ut recitata dupla in epistula pecunia est, ‘hic est’ inquam ‘pater quem
vobis laudaveram’. (Sen. Con.1.7.7); Id enim tempus fere virentium generositatem
declarat, quo sol in eandem partem signifer<i per> eosdem numeros redit per quos
cursus sui principium ceperat. (Col. 3.6.4); Non tenui ego diutius lacrimas sed ad
ultimam perductus tristitiam ‘quaeso’ inquam ‘domina, certe embasicoetan iusseras
dari’. (Petr. 24.1); Mutandus est casus, si durius is quo coeperamus feratur. (Quint.
Inst. 9.4.117);
When a sentence with a pluperfect is followed by another sentence with a past tense,
the event of the first will be understood as finished before the second event. This is not
only the case when the second sentence contains a perfect or a historic present, as in
(g) and (h), but also when it contains a pluperfect, as in (i). Various temporal satellites
may be used in either of the two sentences to underline the rapid succession of events
or to mark the second event as an incident that interrupts the first event. For similar
observations on the imperfect, see § 7.20. For similar semantic relations in the so-
called cum inversum construction and in coordinate structures, see § 16.11 and
Chapter 19, respectively.
With terminative verbs, the pluperfect may be interpreted as indicating a past state.
Examples are (j) and (k). In combination with perfect passive participles of termina-
tive verbs, the pluperfect of the verb sum is used to refer to a state that existed before
some past reference point. An example is (l).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(j) Itaque tum sic statuit, per C. Fabricium (nam L. erat mortuus) insidias
Habito comparare.
(‘And so, at this particular time, he decided to employ Gaius Fabricius—for Lucius
had died—in maturing his plot against Habitus.’ Cic. Clu. 47)
(k) Perieramus, si magistratus esset.
(‘I should be dead if this man were magistrate.’ Sen. Con. 10.1.1)
(l) Pons qui fuerat tempestate interruptus paene erat refectus.
(‘The bridge which had been broken down by the storm was almost repaired.’ Caes.
Civ. 1.41.1)
Supplement:
At eodem tempore Thebis . . . in templo Herculis valvae clausae repagulis subito se
ipsae aperuerunt armaque quae fixa in parietibus fuerant ea sunt humi inventa. (Cic.
Div 1.74); . . . pilaeque illae in quibus devotio fuerat scripta in mare praecipitatae.
(Nep. Alc. 6.5); Deinceps et qui ante nexi fuerant creditoribus tradebantur et
nectebantur alii. (Liv. 2.27.1); . . . alia supellex pretiosaque vestis et multa nobilia
signa, quibus inter primas Graeciae urbes Syracusae ornatae fuerant. (Liv. 26.21.7);
Many Latinists deny the existence of the use of the pluperfect to denote past states.
Although in some of the instances cited above the anterior interpretation may suffice,
it would be odd not to have past state interpretations given what we find for the
perfect (see § 7.30, exx. (p)–(r)). However, there is no reason to assume a stative
interpretation in instances like (m) and (n).120
(m) Post Caesaris reditum quid tibi maiori curae fuit quam ut essem ego illi
quam familiarissimus? Quod effeceras.
(‘After Caesar’s return what was a source of greater concern to you than that
I be most friendly to him? This you had accomplished.’ Cic. Fam. 11.27.5)
(n) Itaque iam feriebantur arietibus muri quassataeque multae partes erant.
Una continentibus ruinis nudaverat urbem. Tres deinceps turres quantum-
que inter eas muri erat cum fragore ingenti prociderant.
(‘And so now the walls were being battered with rams and in many places
had been severely shaken. One section, giving way continually for some
distance, had exposed the town: three towers in a row, together with the wall
connecting them, had come down with a loud crash.’ Liv. 21.8.5)
120
Examples and discussion in Wistrand (1958; appendix 1, ‘The sense of the Latin pluperfect tense’),
following Jensen (1941).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
seem to avoid it.121 However, it is during their lifetime that there occur undeniable
instances of the use of the pluperfect without any difference in semantic value from
the imperfect. It is very common in Vitruvius, and Tertullian and later authors use
it relatively frequently. Examples are (o) and (p).
(o) Quae amicis suis opus fuerant . . . omnia ex sua re familiari dedit . . .
(‘Whatever his friends needed . . . he supplied from his own means . . . ’ Nep. Att. 7.1)
(p) Pixodarus fuerat pastor.
(‘Pixodarus was a shepherd.’ Vitr. 10.2.15)
Supplement:
. . . Iccius Remus . . . qui tum oppido praefuerat . . . nuntios ad eum mittit . . . (Caes.
Gal. 2.6.4); Senatus ita uti par fuerat decernit . . . (Sal. Jug. 39.3); Si tibi non cordi
fuerant conubia nostra / at tamen in vestras potuisti ducere sedes . . . (Catul.
64.158–60); Quae naves ventum secundum nactae quarto die in portum ad Ruspi-
nam ubi Caesar castra habuerat incolumes pervenerunt. (B. Afr. 34.5); Interim
L. Titius qui eo tempore tribunus militum in legione vernacula fuerat nuntiat . . .
(B. Alex. 57.1); . . . totam urbem opplevit luctus, ut sacrum anniversarium Cereris
intermissum sit, quia nec lugentibus id facere est fas nec ulla in illa tempestate
matrona expers luctus fuerat. (Liv. 22.56.4); Qui modo Nasonis fueramus quinque
libelli, / tres sumus. (Ov. Am. 1.prol.1); Trypho Alexandrinus ibi fuerat architectus.
(Vitr. 10.16.10); Quod olim fuerat nubilum nox est . . . (Sen. Nat. 3.27.10); Nec enim
mentitus fuerat propheta . . . (Tert. Pat. 3); . . . prorumpit in patrem inquirere et
genus contrahit vitii, quod exorsum quidem fuerat in illis aliis . . . (Tert. Val. 9);
The same phenomenon can be observed for pluperfect subjunctive forms. In (q),
fuissent is the first instance of the shift of the verb sum, and habuisset in (r) is
mentioned as the first instance of the shift of habeo.
(q) Sed quom multa in eo loco sepulchra fuissent, exarata sunt.
(‘But, as there were many graves in that place, they were dug up.’ Cic. Leg. 2.58)
(r) Cum duos exercitus in provincia habuisset . . . totum prope annum Cremo-
nensibus Placentinisque cogendis redire in colonias . . . consumpsit.
(‘Although he had two armies in the province . . . he spent almost the whole year in
compelling the people of Cremona and Placentia to return to the colonies . . . ’ Liv.
32.26.2–3)
Supplement:
Quibus rebus adductus quos legatos ante exercitui praefecerat, Q. Pedium et
Q. Fabium Maximum, de suo adventu facit certiores utque sibi equitatus qui ex
provincia fuisset praesidio esset. (B. Hisp. 2.2); Ergo, uti Socrati placuit, si ita sensus
et sententiae scientiaeque disciplinis auctae perspicuae et perlucidae fuissent, non
gratia neque ambitio valeret, sed si qui veris certisque laboribus doctrinarum perve-
nissent ad scientiam summam, eis ultro opera traderentur. (Vitr. 3. pr.3);
121
The much-debated difficult cases can be found in K.-St.: I.141.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Just as the imperfect of modal verbs and complex expressions is used to refer to the
time of speaking (see § 7.19, exx. (s)–(t)), the pluperfect is used to refer to the past.
Instances of such expressions are found from Plautus onwards and increase over time;
they become popular from Tertullian onwards. Examples are (s) and (t).
A similar use of the subjunctive pluperfect forms of the modal verbs volo ‘to want’ and
possum ‘to be able’ is first attested in Vitruvius, as in (u). In most authors the shift
is restricted to conditional, temporal, and relative clauses, but in African authors,
such as Fronto, Apuleius, and Tertullian, it is found in purpose and result clauses
as well.
(u) In ea aede cum voluissent columnas conlocare . . . dimensi sunt virilis pedis
vestigium et id retulerunt in altitudinem.
(‘When they wished to place columns in that temple . . . they measured a man’s
footstep and applied it to his height.’ Vitr. 4.1.6)
Supplement:
Cuperem itaque tecum communicare tam subitam mutationem mei: tunc amicitiae
nostrae certiorem fiduciam habere coepissem, illius verae quam non spes, non timor,
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
non utilitatis suae cura divellit, illius cum qua homines moriuntur, pro qua mo-
riuntur. (Sen. Ep. 6.2);122 Neque in quoquam melius consultum rei publicae a
militibus videbatur quam quod instantibus Sarmatis creatus est imperator qui fessis
rebus mederi sua virtute potuisset. (Hist. Aug. trig. tyr. 9.1);
The pluperfect indicative, just like the imperfect indicative (§ 7.20), is found in the
apodosis with conditional subordinate clauses in the subjunctive to underline that
something had almost become reality. With a few exceptions this usage starts in the
Augustan period and then becomes very common in later Latin.123 Two examples are
(v) and (w). See § 16.54 for conditional periods in general.
122
Instances from Seneca can be found in Setaioli (2000: 29).
123
See Haverling (2010b: 149–51).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The so-called future perfect (futurum exactum, sometimes also called ‘future anter-
ior’) is certainly not a frequently used tense form. Just like the simple future, it is
infrequent both in didactic texts and in historical texts, where it is restricted to
statements of the authors and to speeches.124 It is also infrequent in comedy (see
Table 7.5 in § 7.11).
In spite of its infrequency, the future perfect has drawn the attention of scholars for
a number of reasons. One is that there is an evolution of its use. A second is that there
is an archaic—disappearing—category of so-called sigmatic future perfect forms.
These forms play a major role in the discussion about the aspectual nature of the
future perfect. The third reason is that, from roughly Cicero’s time onwards, the
future perfect indicative forms and the perfect subjunctive forms were no longer
morphologically distinct, except in the passive (but Table 7.5 shows how rare these
were) and in the first person singular (tulero versus tulerim, respectively).125 This will
be discussed in detail further on.
Appendix: The original paradigms of the indicative and subjunctive forms were
based on a short i (indicative) and a long i (subjunctive). There are traces of this in
Early Latin inscriptions and in the Early Latin comedies and tragedies. It is interest-
ing to note that ancient grammarians refer to tulero as a ‘futurum subjunctivi’ (for
example Prisc. 8.57 [II.417.28K.]).126 Compare also Charisius: subiunctiva {{instan-
tis}} cum feram, <inperfecti> ferrem, <perfecti> tulerim, <plusquamperfecti> tulis-
sem, <futuri> tulero. (Charis. 344.21f.B). Sacerdos lists it as a subjunctive of the
‘tempus futurum specie ulteriori’ cum amavero, etc. (Sacerd. VI.436.21K). But see
also Verg. Maro Epit. 7 (65.8ff.H): Veteres etiam futurum tempus velut in duas
quasdam partes dividebant ut videbo et videro, quasi aliut sit quod crastino et quod
in longo tempore sit eventurum. Late Latin authors may in turn have been influenced
by these theoretical ideas in the way they use the future perfect forms (so, for
example, the fourth-century historian Julius Valerius).127
The semantic value of the future perfect indicative is defined as follows: by using the
future perfect indicative tense, the speaker asserts his prediction that an event will
have taken place before some reference point in the future.128 This reference point
may be made explicit, for example by simple future forms in the immediate context,
notably in a main clause, or by time adverbs. Examples to illustrate this point are
(a) and (b). In (a), obviously, the inspection precedes the future knowledge, just as
124
The only example in Caesar (praestitero, in Caes. Gal. 4.25.3) is indeed in a speech by a soldier.
125
For the use of long -erīs forms (originally subjunctive forms) for the future perfect, without a
metrical need to do so, in Silver Latin poetry, see Castillo (2010).
126
See Serbat (1978); Álvarez Rodríguez (2001: 66–9); Rosellini (2009).
127
See Rosellini (2005).
128
This is possible in English as well: ‘It is possible that they (hot summers) will have become the
norm by the middle of the century.’ (The Guardian, 12 January 2004, p. 6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
in (b) the division of the land precedes the future situation in which it will no longer
be the property of the Roman citizens. However, the reference point may also be
implied by the context, as will be shown later (exx. (h) ff.).
(a) Quis id coxit coquos? / Iam sciam si quid titubatum est, ubi reliquias videro.
(‘What cook cooked it? I’ll know at once if some mistake has been made when I see
the leftovers.’ Pl. Men. 141–2)
(b) . . . is (sc. ager), cum erit divisus, <neque erit vester> neque vester esse
dicetur.
(‘ . . . the land, when it will have been divided, will neither be yours nor said to be
yours.’ Cic. Agr. 2.81)
From Early Latin onwards, the future perfect is much less common in main clauses
than the simple future, as is to be expected of a tense that locates an event in relation
to a reference point that is not the time of speaking (see also § 7.31 on the pluperfect).
This is shown in Table 7.11.129 The global quantitative differences shown here
become even more interesting, if other contextual factors are taken into account.
Simple future instances are much more common in relative clauses, which are
normally grounded in relation to the time of speaking, whereas future perfect
instances are mostly found in conditional si clauses, which usually refer to an event
that precedes the event of the apodosis. It can therefore be concluded that there is a
clear difference between the simple future and the future perfect in subordinate
clauses and that the latter indicates anteriority.
Supplement:
Hoc atque etiam: ubi erit ac<c>ubitum semel, / ne quoquam exsurgatis, donec a me
erit signum datum. (Pl. Bac. 757–8); Nam nec Bellona mi umquam nec Mars
creduat, / ni illum exanimalem faxo, si convenero, / nive exheredem fecero vitae
suae. (Pl. Bac. 847–9); Tum demum sciam / recte monuisse, si tu recte caveris. (Pl.
Men. 346–7); Venibit uxor quoque etiam, si quis emptor venerit. (Pl. Men. 1160);
. . . BACANALIA . . . / . . . IN·DIEBUS·X·QUIBUS·VOBEIS·TABELAI·DATAI / ERUNT· FACIATIS·UTEI·DIS-
2
MOTA·SIENT. (CIL I .581.28-30 (SCBac., Tiriolo, 186 BC)); Robus materies, item ridica,
129
Based on Risselada (2000), with faxo (in combination with a finite clause) and amabo ignored. For
more detailed tables, see Haverling (2010c: 134, 139).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
ubi solstitium fuerit, ad brumam semper tempestiva est. (Cato Agr. 17.1); URNA AENIA
2
PEREIT·DE·TABERNA. / SEIQUIS·RETTULERIT, DABUNTUR / HS LXV. (CIL I .1680.1–3 (Pom-
peii, II/I BC)); Nam si quotienscumque praeteritus erit is qui non debuerit praeteriri,
totiens oportebit eum qui factus erit condemnari, nihil iam est quod populo
supplicetur . . . (Cic. Planc. 14); Nisi omni tempore quod mihi lege concessum est
abusus ero, querere . . . (Cic. Ver. 1.25); (sc. Carthago) De qua vereri non ante
desinam quam illam excisam esse cognovero. (Cic. Sen. 18); Ego ad te [statim]
habebo quod scribam simul et videro Curionem. (Cic. Att. 10.4.12); Malevolentiae
hominum in me, si poteris, occurres; si non potueris, hoc consolabere quod me de
statu meo nullis contumeliis deterrere possunt. (D. Brut. Fam. 11.11.2); . . . QUEIVE
2
DEPUGNANDEI / CAUSSA AUCTORATUS EST ERIT FUIT FUERIT. (CIL I .593.112–13 (Lex Iulia
Munic., 45 BC)); Nam et prius quam incipias consulto, et ubi consulueris mature
facto opus est. (Sal. Cat. 1.6); Qui prior strinxerit ferrum, eius victoria erit. (Liv.
24.38.5); Actio recta non erit, nisi recta fuerit voluntas. (Sen. Ep. 95.57); Excanduit
Trimalchio et ‘quicumque’, inquit ‘mihi fundi empti fuerint, nisi intra sextum
mensem sciero, in rationes meas inferri vetuo’. (Petr. 53.8 (Trimalchio speaking));
Si dixero enim, inquit, vobis, non credetis. (Tert. Marc. 4.41.3 [= Vulg. Luc. 22.67]);
In Genesi: Et temptavit deus Abraham et dixit ad illum: accipe filium tuum illum
unicum quem dilexisti Isaac et vade in terram altam et inpones illum hostiam in uno
ex montibus, de quo tibi dixero. (Cypr. ad Quir. 3.15); Si autem dixero impio: morte
morieris, et egerit paenitentiam a peccato suo feceritque iudicium et iustitiam,
pignus restituerit ille impius, rapinam reddiderit, in mandatis vitae ambulaverit
nec fecerit quicquam iniustum: vita vivet, non morietur. (August. Spec. 231);
The so-called sigmatic future perfect forms, with the exception of faxo (‘I will do’,
‘I will cause’), which is discussed later in this section, are only found in subordinate
clauses, mostly conditional ones. They have the same semantic value as the ‘normal’
future perfect forms, as appears from their being coordinated, as in (c). These forms
were already archaisms in Plautus’ time and, with the exception of faxo, almost
disappeared after him.130
(c) Si hercle tu ex istoc loco / digitum transvorsum aut unguem latum excesseris
/ aut si respexis, donicum ego te iussero, / continuo hercle ego te dedam
discipulam cruci.
(‘Well, if you leave your place by just a finger’s or a nail’s breadth or if you look back
before I’ve told you, I’ll immediately put you on the cross, and that’ll teach you your
lesson.’ Pl. Aul. 56–9)
Supplement:
Uti legassit super pecunia tutelave suae rei, ita ius esto. (Lex V.3); Cum nexum faciet
mancipiumque, uti lingua nuncupassit, ita ius esto. (Lex VI.1); Peribo, si non fecero:
si faxo, vapulabo. (Pl. ap. Gel. 3.3.8 [fr. 77]); Nam si me inritassis, hodie lumbifra-
gium hinc auferes. (Pl. Am. 454); Si ecfexis hoc, soleas tibi / dabo et anulum in digito
130
For a full survey and discussion of the sigmatic future in Plautus and Terence, see de Melo (2002
and 2007a: 171–90), from whom the translation of (c) is taken. The sigmatic forms are typical of the parts
of Plautus that are written in an elevated style.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
aureum et bona pluruma. (Pl. Cas. 708–9); Qui secus faxit, deus ipse vindex erit. (Cic.
Leg. 2.19); ‘Si prior defexit publico consilio dolo malo, tum illo die, Diespiter,
populum Romanum sic ferito ut ego hunc porcum hic hodie feriam.’ (Liv. 1.24.8);
From Early Latin onwards a future perfect can be used both in the subordinate and in
the main clause of a complex sentence, especially with conditional subordinate
clauses. The general meaning of the combination is that in retrospect one event will
appear to have entailed another. Examples are (d)–(e).
(d) Pol si istuc faxis, hau sine poena feceris, / si ille huc rebitet, sicut confido
adfore.
(‘If you do that, you won’t have done it without suffering for it if that man returns
here, as I’m sure he will.’ Pl. Capt. 695–6)
(e) Respiraro, si te videro.
(‘I will have breathed again if I will have seen you.’ Cic. Att. 2.24.5)
Supplement:
Si tu argentum attuleris, cum illo perdidero fidem. (Pl. Ps. 376); Nam si semel tuom
animum ille intellexerit, / . . . / huic quantam fenestram ad nequitiem patefeceris.
(Ter. Hau. 478–82); Quare non, cum de locis dicam, si ab agro ad †agrosium†
hominem, ad agricolam pervenero, aberraro. (Var. L. 5.13); Sed me dius fidius
multo citius meam salutem pro te abiecero quam Cn. Planci salutem tradidero
contentioni tuae. (Cic. Planc. 79); De qua cum dixero, totum hoc crimen decuma-
num peroraro. (Cic. Ver. 3.154); Nam si quid ab homine ad nullam partem utili
utilitatis tuae causa detraxeris inhumane feceris contraque naturae legem. (Cic. Off.
3.30); Quam porro quis ignominiam, quam turpitudinem non pertulerit, ut effugiat
dolorem, si id summum malum esse decreverit? (Cic. Tusc. 2.16); Pergratum mihi
feceris si dederis operam ut is intellegat meas apud te litteras maximum pondus
habuisse. (Cic. Fam. 12.27); Huius de virtutibus vitaque satis erit dictum, si hoc
unum adiunxero, quod nemo ibit infitias, Thebas . . . caput fuisse totius Graeciae.
(Nep. Epam. 10.4); ‘ . . . nec ante nos hinc moverimus, quam, sicut olim Camillum a
Veis, C. Flaminium ab Arretio patres acciverint.’ (Liv. 22.3.10); Haec autem si qui
voluerit ad perfectum facile perducere, animadvertat in extremo libro diagramma
musica ratione designatum, quod Aristoxenus magno vigore et industria generatim
divisis modulationibus constitutum reliquit, de quo, si qui ratiocinationibus his
attenderit, ad naturas vocis et audientium delectationes facilius valuerit theatrorum
efficere perfectiones. (Vitr. 5.5.6); Dixeris male dicta cuncta, cum ingratum homi-
nem dixeris. (Pub. Sent. D 4);
With a number of verbs, however, the future perfect forms are not so easily inter-
preted as anterior to a reference point in the future. Examples are (f ) and (g); the
former is given by K.-St.: I.151 as an early instance of the ‘original simple future
meaning with an aoristic [= ‘perfective’; see § 7.3] aspect’. There is indeed no future
reference point in the immediate context. However, Demaenetus has asked Libanus a
few lines before: Si quid te volam, / ubi eris? ‘If I need you for something, where will
you be?’ One might therefore explain the use of fuerit in the overall context as ‘if there
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
will have arisen a need for me, you will find me in the forum’. For (g), however, such
contextual evidence is lacking, but it must be explained along the same lines: ‘if it
turns out that I have been capable of doing so’. The verbs mentioned in the literature
in this connection are the stative verbs sum ‘to be’ and habeo ‘to have’; the modal
verbs volo ‘to want’ and possum ‘can’; and the impersonal modal verbs licet ‘it is
allowed’, oportet ‘it is proper’, and placet ‘it is acceptable’. In the Classical period the
use of the future perfect forms of these verbs in subordinate clauses almost becomes
an idiomatic minor variant of the simple future. It is also found in later authors:
Augustine has quite a few instances of si ( . . . ) potuero and si ( . . . ) habuero.
(f ) Audi’n tu? Apud Archibulum ego ero argentarium. / # Nempe in foro? # Ibi,
si quid opus fuerit.
(‘Are you listening? I’ll be at the banker Archibulus’. # You mean in the market? #
Yes, there, if any need arises.’ Pl. As. 116–17)
(g) Eum morem igitur cum brevitate, si potuero, consequar.
(‘Therefore I will follow that style, with its brevity, so far as I can.’ Cic. Leg. 2.18)
Cicero has fourteen instances of si ( . . . ) potuero and si ( . . . ) potero each. There is no
observable difference in distribution over the various text types. The examples given are
taken from K.-St. Some readers may find some of them explainable in temporal terms.
Supplement:
Postid, quom lassus fueris et famelicus, / noctu ut condigne te cubes curabitur. (Pl.
Cas. 130–1); Prius audite paucis: quod quom dixero, si placuerit, / facitote. (Ter. Eu.
1067–8); Decerno igitur eorum trium nomine quinquaginta dierum supplicationes:
causas, ut honorificentissimis verbis consequi potuero, complectar ipsa sententia.
(Cic. Phil. 14.29); At vero si qui voluerit animi sui complicatam notionem evolvere
iam se ipse doceat eum virum bonum esse qui prosit quibus possit noceat nemini nisi
lacessitus iniuria. (Cic. Off. 3.76); Plura scribam si plus oti habuero. (Cic. Fam.
10.28.3); ‘Cuncta equidem tibi, rex, fuerit quodcumque, fatebor / vera’ inquit,
‘neque me Argolica de gente negabo.’ (Verg. A. 2.77–8); Itaque ut potuero quam
apertissime ex Aristoxeni scripturis interpretabor et eius diagramma subscribam
finitionesque sonituum designabo, uti qui diligentius attenderit facilius percipere
possit. (Vitr. 5.4.1); Sin autem non naturalem locum neque idoneum ad tuendas ab
tempestatibus naves habuerimus, ita videtur esse faciendum, uti . . . (Vitr. 5.12.2);
Hac ratione quotienscumque opus fuerit, in aquam poterit esse progressus. (Vitr.
5.12.4); Ambitiosus non ero, cum senex fuero. (Sen. Con. 10.2.7); Intendite. Hoc
planius dicam, si potuero. (August. Psal. 126.10);
As for the future perfect in simple sentences and in main clauses of complex
sentences, there is less agreement in the literature. Most scholars assume that already
in Plautus’ comedies there is no temporal difference between the future perfect and
the simple future. Some scholars, however, recognize traces of the assumed prehis-
toric aspectual distinction between the imperfective value of the simple future and the
perfective value of the future perfect, even in Cicero (as in ex. (l)). It is interesting to
note that in Plautus half of the future perfects are found at the end of the line, against
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
only one fifth of the simple future, which suggests that the future perfect was a
convenient metrical variant (witness (i) and (k)).131
(h) Sed huc concedam, nam concrepuerunt fores. / Mnesilochus eccum maestus
progreditur foras.
(‘But I’ll step aside here: the door has creaked. There, Mnesilochus is coming out with
a sad look on his face.’ Pl. Bac. 610–11)
(i) Eugae! Optume eccum aedium dominus foras / Simo progreditur ipsus. Huc
concessero, / dum mihi senatum consili in cor convoco. / Igitur tum accedam
hunc, quando quid agam invenero.
(‘Hurray! Look the owner of the house, Simo, is coming out himself in the nick of
time. I’ll walk over here while I’m calling a senate meeting of good counsel in my
heart. I’ll go and address him when I’ve come up with what to do.’ Pl. Mos. 686–9)
(j) [Menaechmus has been shut out by his angry wife]
Si tibi displiceo, patiundum. At placuero huic Erotio, / quae me non excludet
ab se, sed apud se occludet domi.
(‘If you dislike me, I’ll have to put up with it. But Erotium here will like me, who will
not lock me out from her place, but will lock me in at her home.’ Pl. Men. 670–1)
(k) Id utrumque, argentum quando habebo, cavero, / ne tu delinquas neve ego
irascar tibi. / # Cape atque serva. Me lubente feceris.
(‘When I have the money I’ll prevent both these things: you committing an offence and
me being angry with you. # Take it and watch over it. I’ll be delighted.’ Pl. Men. 270–2)
Contrary to what most scholars think, exx. (h) and (i) need not be regarded as
identical. Whereas concedam in (h) refers to the future, without implying that certain
events have taken place before that moment in the future, concessero in (i) is the first
step in a number of future actions and precedes the action accedam.132 Ex. (j), in turn,
is another example where the context can argue for anteriority: whatever his wife’s
future attitude will be, Erotium will have received him in her house. Finally, ex. (k)
shows that the future perfect can also be used in ‘retrospectively evaluative’133
statements (which one may paraphrase as ‘the outcome will be that something will
have happened or been brought about’). In sum, most, if not all, instances of the
future perfect in simple sentences and in main clauses in Plautus can be explained if
the larger context is taken into account.134 The effect of presenting an event in this
way is that of a definite intention or guarantee.135
131
A survey of the literature can be found in Sz.: 322–5. For possible traces of an aspectual difference in
Early Latin, see Haverling (2010a: 377–85).
132
I follow the explanations of Risselada (2000), who refers to Gaffiot (1933), and de Melo (2007a:
135–50). K.-St.: I.147 formulate this in the following way: ‘in order to indicate that a future action will
definitely lead to a certain result’ (my translation).
133
Risselada (2000).
134
For a discussion of the Plautine examples with more or less the same result, see Kümmel (2013).
135
It would be worthwhile to investigate the presence of words like certe and of subject pronouns.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Est profecto deus qui quae nos gerimus auditque et videt: / Is uti tu me hic habueris
proinde illum illic curaverit. / Bene merenti bene profuerit, male merenti par erit.
(Pl. Capt. 313–15); Si sapitis, uxor, vos tamen cenabitis, / cena ubi erit cocta. Ego ruri
cenavero. (Pl. Cas. 780–1); Quiesce, inquam: istanc rem ego recte videro. (Pl. Mer.
448); Si nequeo facere ut abeas, egomet abiero. (Pl. Poen. 442); Interibi attulerint
exta, atque eadem mulieres / iam ab re divina, credo, apparebunt domi. (Pl. Poen.
617–18); Quid opus gladio? # Qui hunc occidam atque me. / # Quin tu ted occidis
potius? Nam hunc fames iam occiderit. (Pl. Ps. 349–50); Eodem pacto quo huc
accessi apscessero. (Pl. Trin. 710); Tun’ consiliis? quicquam . . . # Ah, si pergis, abiero.
(Ter. Ad. 127); Ingeram mala multa? Atqui aliquis dicat ‘Nil promoveris’. / Multum.
Molestus certe ei fuero atque animo morem gessero. (Ter. An. 640–1); Adducti qui
illam hinc civem dicant. Viceris. (Ter. An. 892); Puerum accipias. Nam is quidem / in
culpa non est. Post de matre videro. (Ter. Hec. 699–700); Nam idem hoc argentum,
ita ut datum’st, ingratiis / ei datum erit. (Ter. Ph. 888–9);
In Cicero and later authors the future perfect decreases in frequency in main clauses,
and some expressions evolve into idioms. Most scholars consider it semantically
equivalent to the simple future, although some point to traces of an aspectual
distinction. Two examples are (l) and (m).
(l) . . . nusquam facilius hanc miserrimam vitam vel sustentabo vel, quod multo
est melius, abiecero.
(‘ . . . nowhere will I either suffer this most wretched existence more easily, or, what is
much better, discard it.’ Cic. Att. 3.19.1)
(m) Tum vixero, cum te videro.
(‘I will have lived then, when I will have seen you.’ Fro. Ep. ad M. Caes. 5.55)
The aspectual explanation given for (l)136 is not very attractive, since it seems an ad
hoc solution, using aspectual concepts that are of no use in most other instances. One
must instead either assume that the future perfect by Cicero’s time had become
synonymous with the simple future and was indeed a metrical commodity (Cicero
has abiecero three times, all three at the end of the clause) or inconvenience (in
hexameter poetry), or one must try a temporal explanation, in line with the mainly
anterior use of the future perfect in subordinate clauses. Such a temporal explanation
might run as follows: What Cicero is saying is that, before a situation might arise in
which he endures his life in Epirus, he will preferably have taken his own life (or: ‘he
will sooner discard it’). In the instances below, the element of ‘retrospective evalu-
ation’ is relevant. Similar explanations are possible in most cases. The exalted
utterance of the archaizer Fronto in (m) nicely illustrates the exploitation of the
original value of the future perfect.
136
So Sz.: 323.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
‘Nihil egero’ inquit, ‘negabunt enim omnes se coegisse.’ (Cic. Ver. 2.148); Minime id
quidem, inquam, alienum, multumque ad ea, quae quaerimus, explicatio tua ista
profecerit. (Cic. Fin. 3.14); Obieris Quinti fratris comitia, nos longo intervallo
viseris, Acutilianam controversiam transegeris. (Cic. Att. 1.4.1); Immo reliqua
exspectate. Vos invitos vincere coegero. (Cael. Fam. 8.17.2); Non Simois tibi nec
Xanthus nec Dorica castra / defuerint. (Verg. A. 6.88–9); Satis superque me benigni-
tas tua / ditavit; haud paravero / quod aut avarus ut Chremes terra premam /
discinctus aut perdat nepos. (Hor. Epod. 1.31–4); ‘Oppugnabitis enim vere moenia
unius urbis, sed in una urbe universam ceperitis Hispaniam.’ (Liv. 26.43.3); Marcel-
lus Aeserninus eundem colorem aliter induxit: Cogitabat, inquit, secum Antonius:
Quod Ciceroni excogitabo supplicium? Occidi iussero? Olim iam adversus hunc
metum emunivit animum. (Sen. Con. 7.2.10); Et bello vicerimus, si vicimus proelio.
(Curt. 4.14.15); Non tamen sine usu fuerit introspicere illa primo aspectu levia, ex
quis magnarum saepe rerum motus oriuntur. (Tac. Ann. 4.32.2); Recte, videbo te in
publicum, mus, immo terrae tuber. Nec sursum nec deorsum non cresco, nisi
dominum tuum in rutae folium non conieci, nec tibi parsero, licet mehercules
Iovem Olympium clames. (Petr. 58.4–5 (freedman speaking));
There are two types of future perfect expressions that seem to have evolved into
idioms. One is exemplified by (n). Videro can be found in Plautus and Terence as a
reassuring utterance, indicating that the speaker (or the addressee—videris—or a
third party—viderit) will take care of something (‘once it will turn out that I have
looked into this’).137 In Cicero half of the instances of videro are used in this way,
especially in the dialogues. The other half occur mostly in conditional and also in
temporal subordinate clauses; videro then has its ‘normal’ future anterior value. The
idiom is also found in later authors, including some ecclesiastical authors, as in (o),
but the ‘normal’ anterior use is much more frequent.
(n) Sed de me videro. Nunc . . .
(‘But I will see about myself. Now . . . ’ Cic. de Orat. 2.33)
(o) Deus viderit.
(‘God will see to it.’ August. Ep. 227)
Supplement:
Istanc rem ego recte videro. (Pl. Mer. 448); Cur hoc fiat, vos videritis, inquit mihi, qui
Aristotelem legitis. (Var. R. 2.5.13); Sed de te tu videris. Ego de me ipse profitebor.
(Cic. Phil. 2.118); De Capitone post viderimus si . . . (Cic. S. Rosc. 84); . . . quo modo
simul latera, os, venas, oculos, voltum occupet, viderit Democritus. (Cic. de Orat.
2.235); Quae fuerit causa, mox videro; interea hoc tenebo . . . (Cic. Fin. 1.35); Harum
sententiarum quae vera sit, deus aliqui viderit. (Cic. Tusc. 1.23); Hoc quam vobis
tutum aut honestum sit, vos videritis. Ego . . . (Liv. 42.13.11); Viderit ista deus, qui
nunc mea pectora versat! (Ov. Her. 12.211); ‘Viderit. Insanos’ inquit ‘fateamur
137
Pfister (1936: 32–3) points to the special meaning (‘to take care of ’) in this idiomatic use of
videro.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
amores!’ (Ov. Met. 9.519—NB: subject not indicated); An aliquid mali velint facere
postea videro. Interim possunt. (Quint. Decl. 225.8); Itaque hilaria mera sint, etsi
timeo istos scholasticos, ne me [de]rideant. Viderint: narrabo tamen. Quid enim
mihi aufert qui ridet? (Petr. 61.4 (Niceros speaking)); VIDERIT ·AN·NE·ALIQUIS·POST·
ME·MEA·GESTA·SEQU<A>TUR! (CIL III.3676.10 (early II AD)); Sed interim quidem
tu provinciam, quae tibi matris meae praecepto mandata est, exsequere naviter.
Cetera egomet videro. (Apul. Met. 6.21.4); Cum lignum aliquod propitiatur, viderit
habitus . . . (Tert. Apol. 16.6); Videro tamen, quid faciant Arriani; ego te, pater, si
maiorem omnibus dixero, iniuriose te tuis operibus conparavi, si maiorem filio,
ut Arrius adserit, impie iudicavi. (Ambr. De fide 5.19); Sed de aliis videro, tuae
nunc mihi respondendum epistulae est. (Ambr. Ep. 8.56.16); Sed de te videro. Stoici
certe . . . (August. Cresc. 1.13–14);
The only sigmatic form occurring in simple sentences and in main clauses of complex
sentences is faxo. In Plautus, who has seventy-five of the 114 instances attested in
Latin literature (grammarians excluded138), it is most often found parenthetically in a
sentence with a simple future or a future perfect tense form, as in (r), where it seems
to function as an emphasizer.139 There are a few certain instances of faxo governing a
clause with a present subjunctive (s), which became the norm in the rare later cases in
138
Source: BTL and CLCLT-1.
139
All instances of faxo are in TLL s.v. facio 83.18ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
poetry and archaizing prose, where some sort of literary ‘de-idiomatization’ even
produced two instances with a subordinator. It is very difficult to see in faxo anything
more than a simple future. See also (t) with ibo and faxo coordinated.
(r) Horrescet faxo lena, leges quom audiet.
(‘I’ll make sure that the madam will shudder when she hears the terms.’ Pl. As. 749)
(s) Quin venis quando vis intro? Faxo haud quicquam sit morae.
(‘Why don’t you come in as soon as you wish? I’ll make sure that there won’t be any
delay.’ Pl. Am. 972)
(t) Ibo ego ad trisviros vostraque ibi nomina / faxo erunt . . .
(‘I’ll immediately go from here to the Board of Three and make sure your names are
with them.’ Pl. As. 131–2)
Supplement:
Sequere hac me modo. Iam faxo ipsum hominem manufesto opprimas. (Pl. As. 876);
Ego te hodie faxo recte acceptum, ut dignus es. (Pl. Rud. 800); Haut sibi cum Danais
rem faxo et pube Pelasga / esse ferant, decumum quos distulit Hector in annum.
(Verg. A. 9.154–5); Audeat! Haud ultra faxo spectetur in armis. (Sil. 7.115); ‘Iam faxo
nuptias non impares, sed legitimas et iure civili congruas’, et ilico per Mercurium
arripi Psychen et in caelum perduci iubet. (Apul. Met. 6.23.4–5);
Parenthetically: Quin tu taces? / Manufesto faxo iam opprimes: sequere hac modo.
(Pl. Men. 561–2); Iam ego hoc ipsum oppidum expugnatum faxo erit lenonium.
(Pl. Ps. 766); Iam faxo hic aderit. (Ter. Ph. 308);
NB: Faxo, ne iuvet vox ista ‘veto’, qua nunc concinentes collegas nostros tam laeti
auditis. (Liv. 6.35.9 (in a speech)); cf. faxo ut in Sil. 17.235;
Just as is the case with the simple future indicative, the future perfect is found in
simple sentences and in the main clauses of complex sentences with a deductive (v) or
an assumptive (w) meaning. Instances are rare, as is to be expected for a tense form
140
See Blase (1898: 333–6), Thomas (1938: 188–98), Kiss (1982: 63–4), and Haverling (2013: 15–28).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Latin had no synthetic tense form for the future of the past. However, for the active
the old periphrastic -urus + sum expression provided a suppletive form (ex. a),
although very often the expression has its original value of immediacy, fate, or
intention (see § 7.24 for further examples). For the passive, the -ndus + sum
expression provided another suppletive form, although here too the -ndus form
often maintained its deontic value (see § 7.25). Later, from Tertullian onwards, the
infinitive + habeo construction provided a competing expression, which also supplied
an expression for the passive and was eventually to win out (ex. b) (For details, see
§ 7.27). The future in the past meaning is found when the sum and habeo forms are
imperfect or pluperfect, as is to be expected on the basis of the meanings of these
tenses. Instead of ‘future of the past’, the characterization ‘prediction of the past’ often
better suits the situation. The -urus and -ndus expressions were much more frequent
than the infinitive + habeo expressions. However, on closer inspection, the -urus
expressions and the habeo expression are not found with the same range of verbs.
In addition, some future participles were not in use.142
141
Cf. Núñez (1991: 190–1).
142
For details, see Bourova (2005, 2008) and Târa (2014: 269–98).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) . . . conclave illud ubi erat mansurus, si ire perrexisset, proxima nocte
corruit.
(‘ . . . that room collapsed the next night, where he was about to remain, if he had
continued his journey.’ Cic. Div. 1.26)
(b) Tamen et Creatori notum erat futurum. An non utique notum quod sub
caelo et in terra eius habebat revelari?
(‘But yet to the Creator also the future was known. Did he not inevitably know that
which beneath his heaven and on his earth was due to be revealed?’ Tert. Marc.
5.18.4)
Supplement: . . . ideo fecisse dicens se quod erat sine ulla dubitatione facturus.
(August. Tract. Io. 106.1); Prophetae autem maxime solent figura inprecantis futura
praedicere, sicut figura praeteriti temporis ea quae ventura erant saepe cecinerunt,
sicuti est illud. (August. Serm. Dom. 1.72);
Et ideo de Christo, quia non erat ad terrena curvandus, prophetatum est: Arun-
dinem quassatam non confringet. (Ambr. Luc. 5, 106);
Dari enim habebat circumcisio . . . (Tert. Iud. 3.4); . . . cum tamquam ovis ad victi-
mam deduci habebat . . . (Tert. Jud. 14.1); . . . eum qui nasci habebat ex virgine . . .
(Tert. Prax. 26.2); . . . ut qui eici de ecclesia et excludi habebat . . . (Cypr. Ep. 52.3);143
7.35 ‘Shift’ of deponent and passive forms of the perfectum stem (indicative forms)
In § 7.30 examples ((p)–(r)) are discussed of perfect passive forms of terminative
verbs which are not interpreted as anterior verb forms, but rather as present states:
liber inscriptus est ‘Hortensius’. As is demonstrated there, inscriptus fuit then indicates
an anterior state. Similar examples are given in § 7.31 and § 7.32 for the pluperfect
and the future perfect. In principle, with these verbs the entire paradigm of the verb
sum can be used; furthermore, from Plautus onwards the use of perfectum forms as
equivalents of infectum forms can also be observed with deponent verbs (the parti-
ciples of which often lack an anterior meaning, e.g. veritus ‘fearing’) and with non-
terminative verbs. Examples are (a)–(f ).
(a) Vel opperire: est quod domi dicere paene fui oblitus.
(‘Or wait, there’s something I almost forgot to say at home.’ Pl. Ps. 171)
(b) Coctum ego, non vapulatum dudum conductus fui.
(‘I was hired for cooking a while ago, not for getting a beating.’ Pl. Aul. 457)
143
For the material, see TLL s.v. habeo 24.55ff. Bourova (2005, 2007, 2008) discusses the degree to
which these expressions resemble the Romance conditionnel.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(c) Lucernam forte oblitus fueram exstinguere. / Atque ille exclamat derepente
maxumum.
(‘I’d accidentally forgotten to put out the lamp. And suddenly he lets out an
enormous shout.’ Pl. Mos. 487–8)
(d) Pol mihi / eo pretio empti fuerant olim. # Audi’n ‘fuerant’ dicere?
(‘I had bought them for this much long ago. # Can you hear him say “had”?’ Pl. Mos.
820–1)
(e) Quem quidem ego hominem irrigatum plagis pistori dabo, / nisi hodie prius
comparassit mihi quadraginta minas / quam argenti fuero elocutus ei post-
remam syllabam.
(‘I’ll irrigate him with blows and hand him over to the miller unless he gets me forty
minas today before I’ve pronounced the last syllable of the sum.’ Pl. Epid. 121–3)
(f ) Numquam enim quale sit illud de quo disputabitur intellegi poterit, nisi quod
sit fuerit intellectum prius.
(‘For it will never be possible to understand what sort of thing is going to be
discussed, unless it will first have been understood what the thing is.’ Cic. Rep. 1.38)
The replacement of infectum forms by perfectum forms can be traced from Early Latin
into the Romance languages. The phenomenon is known as SHIFT.144 The shift did not
take place at the same speed for all the forms involved, and in Cicero and Caesar only a
few instances are found.145 In Early Latin it is much more frequent with deponent
forms than with ‘normal’ passive forms, but in an author like Vitruvius this is no longer
the case. Shifted pluperfect and future perfect forms are much more frequent than
perfect forms, and this difference in frequency persists until Late Latin. Of the sub-
junctive forms, the perfect subjunctive is relatively common in authors like Vitruvius;
the pluperfect subjunctive, on the other hand, is less common until Late Latin, probably
because it often functioned as a past counterfactual (and not as an anterior) form.146
A few details: Plautus has more ‘regular’ combinations with the imperfect than with
the pluperfect form of sum. There are a few examples with the pluperfect in Cicero
and Caesar. Livy has several. Ovid uses them relatively often.147 They are more
frequent than the imperfect in Vitruvius (10 : 7), and that is typical of the general
trend. But Tacitus has only one shifted form, and the Classical imperfect form of sum
is absolutely predominant in Ammianus as well (26 : 1).148 Rhythmical consider-
ations also favoured the use of the pluperfect. It is sometimes observed that shifting is
relatively more frequent in relative clauses, but it is not clear whether this observation
is really based on sound statistical analysis.
144
The German term is ‘Verschiebung’. For the diachronic development and the factors involved, see
de Melo (2012). See also Dalès (1976/77), who assumes an ‘aspect aoristique’ for these expressions, and
Joffre (1995: 313–14). For another form of ‘shift’, see § 7.31.
145 146
For Cicero, see Lebreton (1901a: 203–7). So de Melo (2012: 94–5).
147
See Hoffmann (1997a: 118–214).
148
A detailed table can be found in Blase (1894: 60–1, summarized in 1903: 220–2).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Vitruvius’ manual on architecture, written in a less elevated style, has 105 forms with
fuero against thirty with ero without a noticeable systematic semantic distinction.149
In Late Latin the shifted forms are avoided by Symmachus and used sparingly by
Augustine and Ammianus, whereas they are more common in less educated writers.150
Supplement:
Perfect forms: Quod numquam opinatus fui neque alius quisquam civium . . . (Pl.
Am. 186); Quod fui iuratus feci. (Pl. Cur. 566); Is publice legatus Naupactum fuit.
(Pl. Mil. 102); Ubi sumus provecti in altum, fit quod <di> volunt, / capiunt praedones
navem illam ubi vectus fui. (Pl. Mil. 117–18); Iam pridem equidem istuc scivi et
miratus fui / neminem venire qui istas adsereret manu. (Pl. Poen. 1347–8); Sed
opportunissime nuntiis adlatis oppidum fuit defensum. (Caes. Civ. 3.101.4); Orta
suburbanis quaedam fuit Anna Bovillis . . . (Ov. Fast. 3.667); Aurea Phoebi porticus a
magno Caesare aperta fuit. (Prop. 2.31.2); Hic noster, hic plebis nostrae habitus fuit
eritque semper. (Liv. 6.26.5); Nocte ac die continuatum incendium fuit . . . (Liv.
26.27.5); Item Lacedaemone e quibusdam parietibus etiam picturae excisae inter-
sectis lateribus inclusae sunt in ligneis formis et in comitium ad ornatum aedilitatis
Varronis et Murenae fuerunt adlatae. (Vitr. 2.8.9); Etiamque Pythagorae quique
eius haeresim fuerunt secuti . . . (Vitr. 5.pr.3); Cyrene autem condita fuit ab
Aristaeo . . . (Justin. 13.7.1);151
Pluperfect forms: . . . posterius quam mercatus fueram, visus sum / in custodelam
simiae concredere. (Pl. Mer. 232–3); Detineo te: fortasse tu profectus alio fueras.
(Ter. Eu. 280); Nam abducta a vobis praegnas fuerat filia . . . (Ter. Hec. 640); Ex quo
ego veni ad ea quae fueramus ego et tu inter nos de sorore in Tusculano locuti. (Cic.
Att. 5.1.3); Huic Caesar pro eius virtute atque in se benevolentia, quod in omnibus
bellis singulari eius opera fuerat usus, maiorum locum restituerat. (Caes. Gal. 5.25.2);
Frumenta enim quae fuerant intra munitiones sata consumpserant. (Caes. Civ.
3.58.3); Causa morae nulla est, et iam puer Arcas (id ipsum / indoluit Iuno) fuerat
de paelice natus. (Ov. Met. 2.468–9); . . . matris, quam fatiloquam ante Sibyllae in
Italiam adventum miratae eae gentes fuerant. (Liv. 1.7.8); . . . foedere quod prope in
perpetuum negatum fuerat . . . (Liv. 2.22.5); Iam et Dolabella in Asia C. Trebonium,
consularem, cui succedebat, fraude deceptum Zmyrnae occiderat, virum adversus
merita Caesaris ingratissimum participemque caedis eius a quo ipse in consulare
provectus fastigium fuerat . . . (Vell. 2.69.1); . . . quoniam quidem partes sarmen-
torum quae fuerant obrutae satis multas habent radices . . . (Col. 5.5.16); An tyran-
nicida qui tyrannum a quo deprensus in adulterio fuerat occidit? (Quint. Inst.
5.10.36); Intraque lectus eburneus auro ac purpura stratus et ad caput tropaeum
cum veste in qua fuerat occisus. (Suet. Caes. 84.1); Cum alimenta per fideicommis-
sum relicta sunt non adiecta quantitate, ante omnia inspiciendum est quae defunctus
solitus fuerat ei praestare, deinde quid ceteris eiusdem ordinis reliquerit . . . (Val. dig.
34.1.22.1); Finitisque epulis Vadomarium fortiter apprehensum rectori militum arte
custodiendum apud signa commisit textu lecto iussorum comitibus eius ad sua redire
149
Livy ero 12 : fuero 16; Scribonius Largus: 26 : 3. Detailed data in Blase (1898: 322). For the shift of
ero to fuero, see also Álvarez Rodríguez (2001).
150
See Haverling (2010a: 429–32).
151
See also Fögen (1997/8: 189–90) on shifted forms in Consentius.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
compulsis super quibus nihil fuerat imperatum. (Amm. 21.4.5); Qui fodientes in eo
loco qui ostensus fuerat invenerunt speluncam . . . (Pereg. 16.6);
Future perfect forms: Si ego minam non ultus fuero probe quam lenoni dedi, tum
profecto me sibi habento scurrae ludificatui. (Pl. Poen. 1280–1); Nam et illi iam
relevabis animum, quae dolore ac miseria / tabescit et tuom officium fueris functus.
(Ter. Ad. 601–2); Verum haec et dicentur alio loco et dicta sunt. Nunc proficiscemur
ad reliqua, si pauca ante fuerimus a vobis, iudices, deprecati. (Cic. Ver. 3.10); Non, hodie
si / exclusus fuero, desistam. Tempora quaeram. / Occurram in triviis. Deducam. (Hor.
S. 1.9.57–9); Ipsa quoque infantis cum vulnere prosequar umbras / nec mater fuero dicta
nec orba diu. (Ov. Her. 11.119–20); Namque si lex perite fuerit scripta, erit ut sine captione
uterque ab utroque liberetur. (Vitr. 1.1.10); Et si quid aliud eiusmodi deinceps ausus fuerit,
corrigetur. (Amm. 28.4.16); Si ergo tibi dictum fuerit: cum dicis mihi filium hoc esse quod
pater est, profecto et filius pater est; responde: . . . (August. Enarr. in Ps. 68.1.5);
The shift of the perfect subjunctive to an alternative for the present is said to have
become common much earlier in subordinate clauses that contain generic and/or
prescriptive statements such as (c) and (d). It becomes the norm from the third
century onwards.152
152
Blase (1903: 210).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(c) Isodomum dicitur, cum omnia coria aequa crassitudine fuerint structa.
(‘It is called “isodomum” when all the layers are built of equal thickness.’ Vitr. 2.8.6)
(d) Nam . . . fere non patitur suppurationem fieri et, si facta fuerit, eadem res
persanat.
(‘For usually it prevents suppuration, and if it has developed, the same matter heals
it.’ Larg. 40)
Supplement:
Cum ergo his rationibus [erit] salubritatis [moenium conlocandorum explicatio]
regiones[que] electae fuerint fructibus ad alendam civitatem copiosae, et viarum
munitiones aut opportunitates fluminum seu per portus marinae subvectionis ha-
buerit ad moenia comportationes expeditas, tunc turrium murorumque fundamenta
sic sunt facienda . . . (Vitr. 1.5.1); Facilius enim iam patiuntur oculi, si modo exulcerati
non fuerint, iniuriam. (Larg. 23); Quae si paulum intermissa fuerit, non velocitas illa
modo tardatur, sed ipsum <os> coit atque concurrit. (Quint. Inst. 10.7.8);
Whereas instances of the shift of indicative pluperfect forms are attested from Early
Latin onwards, the first instances of shifted subjunctive pluperfect forms date from
Cicero’s time and later.153 Revocatus . . . fuisset in (e) is said to be the first instance.
The shift remained a minor variant in most authors and is absent in some (the two
Senecas and Petronius, for example). Lucifer, in the second half of the fourth century
AD, is the first author who has more shifted than regular forms.
(e) Addidit ad extremum, cum iam dimissa contione revocatus a Vatinio fuisset,
se audisse ex Curione . . .
(‘Finally, after the meeting had been dismissed and he had been called back by
Vatinius, he added that he had heard from Curio . . . ’ Cic. Att. 2.24.3)
Supplement:
Interim Numidae Gaetuli<que> diffugere cotidie ex castris Scipionis et partim in
regnum se conferre, partim quod ipsi maioresque eorum beneficio C. Mari usi
fuissent Caesaremque eius adfinem esse audiebant in eius castra perfugere caterva-
tim non intermittunt. (B. Afr. 32.3); Sat patriae Priamoque datum: si Pergama dextra
/ defendi possent, etiam hac defensa fuissent. (Verg. A. 2.291–2); Auctaque ea invidia
est ad plebem, quod tributum etiam in stipendium militum conlatum est, cum, si
spreta gloria fuisset captivae pecuniae in aerarium inlatae, et militi donum dari ex
praeda et stipendium militare praestari potuisset. (Liv. 10.46.6); Quis tibi, si sine me
fatis erepta fuisses, / nunc animus, miseranda, foret? (Ov. Met. 1.358–9); Cum ibi
magna vis aquae, luti, stercoris nocte profusa fuisset, postero die helepolis accedens,
antequam adpropinquaret ad murum, in umido voragine facta consedit nec progredi
nec egredi postea potuit. (Vitr. 10.16.7); Haec ut primum ante iudicis conspectum facta
153
A survey of the data can be found in Blase (1903: 226–32). For the rare use of the imperfect
subjunctive instead of the pluperfect subjunctive in later authors and vice versa, see Sz.: 231–2 and
Haverling (2010a: 434–7).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
est, nisu brachiorum polliceri videbatur, si fuisset deabus ceteris antelata, daturam se
nuptam Paridi forma praecipuam suique consimilem. (Apul. Met. 10.32.4);
Appendix: With the shift of the perfectum passive forms, the distinction between the
periphrastic perfectum forms (amatus sum) and the synthetic forms (amor) became
less meaningful, which contributed to the replacement of the synthetic forms in
Romance languages by successors of the Latin periphrastic form. Instances cited in
this connection are (f) and (g), but they are unconvincing: the development took
place after the period covered by this Syntax. Note in (f) the parallelism with simple
future densetur.
(f) Quae structura vectibus ligneis subinde densetur et structa sit lapide toficio . . .
(‘The concrete should be rammed periodically with wooden bars, and let it
have been made with tufa . . . ’ Pall. 9.9.2—tr. Fitch)154
(g) Qui locus ad quod lectus fuerit, tantus rugitus et mugitus totius populi est cum
fletu ut forsitan porro ad civitatem gemitus populi omnis auditus sit.
(‘By the time the passage has been read, such is the groaning and lamenting
together with weeping of the entire people that the moaning of the whole
people has possibly been heard as far as the city.’ Pereg. 36.3)155
Expressions consisting of habeo and a perfect passive participle are the basis for the
development of the periphrastic use of habeo in the Romance perfect, pluperfect, and
future perfect active forms.156 In the examples above, the combination of participle
154
For Palladius, see Svennung (1935: 456–7).
155
For the periphrastic passive in Pereg., see Väänänen (1987: 62–5).
156
Among the numerous studies on the development described in this section, see Happ (1967:
94–110); Pinkster (1987a); Hopper and Traugott (2003: 62–3); La Fauci (1991); Jacob (1995); de Acosta
(2011); Adams (2013: 615–51); Târa (2014). For an explanation in so-called pragma-historical terms, see
Detges (2000) (on Old English and Spanish). Material can be found in TLL s.v. habeo 2452.65ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
and habeo cannot be understood as one unit which governs an object constituent that
is semantically required by the participial part of that combination. Ex. (c) is usually
cited as an early instance of the periphrastic construction, with episcopum required by
invitatum, but this is incorrect: the agents of invitatum and habes are not the same.157
(c) Ecce episcopum cum duce et civibus invitatum habes et vix nobis supersunt
quattuor vini amphorae.
(‘Look, you have the bishop, together with the duke and the citizens, as an invited
guest and there are scarcely four jars of wine left.’ Greg. T. Vit. patr. 3.1)
Early examples in which the object constituent can only be understood as related to
the participial element are found with communication, cognition, and perception
verbs, especially when the object is an argument clause, as in (d). An early example
with a verb of another class governing an indirect question clause as its object is (e).
(d) Compertum ego habeo, milites, verba virtutem non addere . . .
(‘I am well aware, soldiers, that words do not supply valour . . . ’ Sal. Cat. 58.1)
(e) Quantum autem in acie tironi sit committendum nimium saepe expertum
habemus.
(‘We have seen only too often how much reliance can be placed on raw troops in
battle.’ Planc. Fam. 10.24.3)
Expressions like the ones dealt with so far are relatively common in Plautus and
Cicero, where they are not typical of a less elevated style,158 but there are fewer
attestations from Livy onwards. Scholars are not in agreement about when such
combinations of habeo + perfect passive participle first have to be regarded as truly
periphrastic forms of the perfectum. It is quite clear that by AD 400 there are enough
instances that can only be interpreted that way; an example is (f ). Disputable is (g):
one might take societatem not as required by initam but as the object of habent
(societatem habeo cum is a Classical Latin expression (Nep. Paus. 3.6)); initam is then
quite otiose. Two striking earlier examples are (h) and (i)—the latter is an impersonal
passive. The development was a gradual one, which was completed far beyond the
period covered by this Syntax.159
(f ) Nuper . . . quaesisti quid ea verba quae ex Hebraeo in Latinum non habemus
expressa apud suos sonarent . . . ut est illud: alleluia, amen . . .
(‘Recently you asked what the words which we have not translated into Latin
from Hebrew mean in their original language, for example the following: alleluia,
amen . . . ’ Hier. Ep. 26.1)
157
See Târa (2014: 181–4) and Fruyt (2011). For Gregory of Tours, see also Târa (2012) and Adams
(2013: 641–5).
158
See Happ (1967: 103): in Plautus they are more common in passages of a higher stylistic register.
159
Examples (g) and (h) are explained in another way by Adams (2013: 630, 633).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
160
For a different analysis, see Adams (2013: 630–1).
161
See Svennung (1935: 263–4) and Adams (1990: 246).
162
For this pattern of habeo + perfect passive participle and its continuation in Romance languages,
see Wehr (2012).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(j) Sati’n ego istuc habeo offirmatum? # Quod semel dixi haud mutabo.
(‘Do I have this as your firm resolution? I won’t change what I’ve said once.’
Pl. Bac. 1202)
(k) Satis iam dictum habeo.
(‘You’ve said enough, I’ve got it.’ Pl. Per. 214)
(l) . . . meridianae nationes . . . habent exsuctas ab sole animorum virtutes.
(‘ . . . southern peoples have the strength of their minds drained by the sun.’
Vitr. 6.1.10)
Before the individual subjunctive forms are discussed, it will be useful to deal with a
number of collective temporal issues first. Many of the subjunctive forms that one
encounters, as far as their tense value is concerned, do not differ essentially from the
corresponding indicative forms. A present subjunctive form, for instance, often
indicates that the state of affairs of the clause is simultaneous with the time of
speaking or the state of affairs referred to in the main clause, as in (a). A perfect
subjunctive form, in turn, signals anteriority of the state of affairs in its clause to the
time of speaking or to the state of affairs in the main clause, as in (b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) . . . cum . . . vita sine amicis insidiarum . . . plena sit, ratio ipsa monet amicitias
comparare . . .
(‘Since . . . a friendless life is beset with many secret dangers, reason itself advises the
acquisition of friends . . . ’ Cic. Fin. 1.66)
(b) De pietate autem Attici quid plura commemorem, cum hoc ipsum vere
gloriantem audierim in funere matris suae . . .
(‘Why should I say more about Atticus’ devotion to his family than this? He himself,
in my hearing, justly prided himself at the funeral of his mother . . . ’ Nep. Att. 17.1)
However, for present subjunctive forms in other types of subordinate clauses, the
notion of simultaneity is not useful at all. The events the clauses refer to are actually
posterior to the time of speaking or the event of the main clause. This is the case in (c),
an optional purpose adjunct clause, and in (d), an obligatory argument clause. The
threefold temporal distinction of the indicative does not exist for the subjunctive;
rather, the distinction is between a past and a non-past form (leaving out of account
suppletive forms, for example, future participle + sum forms, for which see § 7.24).
(c) Me a portu praemisit domum, ut haec nuntiem uxori suae . . .
(‘As for me, he’s sent me ahead home from the harbour so that I could report to his
wife . . . ’ Pl. Am. 195)
(d) Eum roga[t] ut relinquat alias res et huc veniat.
(‘Ask him to leave everything else aside and to come here.’ Pl. Rud. 1212)
The notion of anteriority is generally useful in describing the semantic values of the
perfect and pluperfect subjunctive forms. It will be shown below, however, that a
distinction between the present and perfect subjunctive in their deontic use (ne facias /
ne feceris) cannot be, or cannot simply be, described using this notion. Furthermore,
pluperfect subjunctive forms, when used counterfactually, cannot be described as
anterior to a past reference point; rather, they simply situate the counterfactual event
in the past. The imperfect subjunctive, on the other hand, situates a counterfactual
event in the time of speaking.
(For ‘shift’ of perfect and pluperfect subjunctive forms, see § 7.36.)
The main use of the subjunctive in declarative and interrogative clauses is to express
that a state of affairs may/might take place, without the speaker committing himself to
the assertion that this indeed, was, is, or will be the case. The label ‘potential’163 is
common in Latin grammars, where this use of the subjunctive goes under the name
coniunctivus potentialis. The effect of using the subjunctive is to make the statement
163
Most of this would be covered by the notion ‘abilitive dynamic modality’ in Palmer (2001).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
less direct and milder than it would be in the indicative. The negator is non. Instances
of this use are found from Early Latin onwards.
habeas, assem valeas. (Petr. 77.6 (Trimalchio speaking)); Horum denique videas
exitus turpes. (Cypr. ad Don. 11);
Third person: Apage hau nos id deceat / fugitivos imitari. (Pl. Capt. 208–9); Fors fuat
an istaec dicta sint mendacia. (Pl. Ps. 432); Atqui aliquis dicat: ‘Nil promoveris’. (Ter.
An. 640); In libris Saliorum quorum cognomen Agonensium forsitan hic dies ideo
appelletur potius Agonia. (Var. L. 6.14); Quae forsitan vobis parvae esse videantur . . .
(Cic. Ver. 4.47); ‘Quid ergo postulas?’ dicat aliquis. (Liv. 37.53.25); Dicat fortassis
aliquis: ‘Non enim invehebantur’. (Plin. Nat. 36.5); Sed fortasse aliquis opponat et
dicat . . . (Cypr. Mort. 17); . . . forte aliquis dicat . . . (August. in Ioh. 3.17); In gratiam me
quispiam putet Constantii Caesaris loqui . . . (Hist. Aug. Claud. 3.1);
The most common use of the present potential subjunctive is in clauses belonging to a
conditional period. It is found in the protasis (the si clause), but in Early Latin and still
in Cicero’s philosophical writings, it is also common in the apodosis (the main
clause). However, the apodosis may also contain a present or future indicative, or a
directive expression (imperative or subjunctive). These alternatives are found from
Cicero onwards, except in authors following the earlier practice. Examples of the
subjunctive are (e)–(g).164 For further details, see § 7.156.
(e) Non me novisti? # Non negem, si noverim.
(‘Don’t you know me? # I wouldn’t deny it if I did.’ Pl. Men. 504)
(f ) Magis dicas, si scias quod ego scio.
(‘You’d say that all the more if you knew what I know.’ Pl. Mil. 1429)
(g) Dies iam me deficiat, si . . . coner expromere.
(‘Daylight would soon fail me if I tried to expose . . . ’ Cic. Cael. 29)
Supplement:
Meam rem non cures, si recte facias. Num ego curo tuam? (Pl. Capt. 632); Me rogas?
# Pol hau rogem te, si sciam. (Pl. Men. 640); Quid nunc? Si in aedem ad cenam
veneris / atque ibi opulentus tibi par forte obvenerit / (adposita cena sit popularem
quam vocant) / si illi congestae sint epulae a cluentibus, / siquid tibi placeat quod illi
congestum siet, / edisne an incenatus cum opulento accubes? (Pl. Trin. 468–73);
Immo enim si scias quod donum huic dono contra comparet, / magis id dicas. (Ter.
Eu. 355–6); Orationes autem, quas interposuit—multae enim sunt—, eas ego laudare
soleo: imitari neque possim, si velim, nec velim fortasse, si possim. (Cic. Brut.
287); . . . (sc. patria) pro qua quis bonus dubitet mortem oppetere si ei sit profuturus?
(Cic. Off. 1.57); An hoc, si Claudiae familiae non sim . . . reticere possim . . . (Liv.
6.40.6–7); Et si exsistat hodie ab inferis Lycurgus, gaudeat ruinis eorum, et nunc se
patriam et Spartam antiquam agnoscere dicat. (Liv. 39.37.3); . . . adulteras, quarum si
censum requiras, facilius apostaticum invenias quam apostolicum, . . . (Tert. Marc.
4.5.3); Si autem divinus non sit, ne deus quidem sit . . . (Lact. Inst. 1.11.10);
164
Statistical data can be found in Blase (1896a: 25).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(ii) The next group of questions with a present subjunctive form are often labelled
‘consultative’, ‘deliberative’, or ‘dubitative’, as in (d) and (e). In these questions a
person consults his own feelings about what he can or should do in a given
situation. They are different from the instances of questions about one’s future
actions discussed in § 7.15, illustrated here by (f ).
conferam? Quo me vertam? (Cic. Ver. 5.2); Eloquar an sileam? (Verg. A. 3.39); Quid
faciamus homines miserrimi et novi generis labyrintho inclusi, quibus lavari iam
coeperat votum esse? (Petr. 73.1); Quid autem agam quove modo ad istum portum
necessarios meos congregem ut cognoscas . . . (August. Beat. 1); Heu! Quid agam?
Dubia pendens pietate laboro, / gratuler an doleam? (Paul. Nol. C. 31.7–8);
165
See Touratier (1994: 138–9). For objections, see Álvarez Huerta (2001, 2011).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The two uses of the present subjunctive discussed in (ii) and (iii) are often regarded as
deontic uses. The reason to deal with them as potential uses is that negation is by non
and not by ne.
166
Gilmartin (1975) discusses the use of the second person singular subjunctive verb form in Roman
historians from Sallust to Ammianus (putares, cerneres, etc.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(c) Cur igitur et Camillus doleret, si haec post trecentos et quinquaginta fere
annos eventura putaret et ego doleam, si ad decem milia annorum gentem
aliquam urbe nostra potituram putem?
(‘Why then should Camillus have felt pain, if he had thought that some 350 years
after his lifetime that present troubles would come and why should I feel pain if
I should think that some nation would get possession of our city at a date 10,000
years hence?’ Cic. Tusc. 1.90)
Supplement:
First person: Non ego hoc ferrem calidus iuventa / consule Planco. (Hor. Carm.
3.14.27–8);
Second person: Miserum? Quem minus crederes? (Ter. Hau. 192); At eius avuncu-
lum vix intellegeres id agere, cum ageret tamen, Africanum. (Cic. de Orat. 3.87);
Putaresne umquam accidere posse ut mihi verba deessent . . . ? (Cic. Fam. 2.11.1);
Pecuniae an famae minus parceret haud facile discerneres. (Sal. Cat. 25.3); Tum vero
in numerum faunosque feraque videres / ludere. (Verg. Ecl. 6.27–8); Iniussu signa
referunt maestique—crederes victos—, execrantes nunc imperatorem nunc navatam
ab equite operam, redeunt in castra. (Liv. 2.43.9); Occidit de lucerna equites. Putares
eos gallos gallinaceos. (Petr. 45.11 (Echion speaking)); Innocentiam iustitiamque eius
non argueres. (Tac. Hist. 3.75.1);
Third person: Nam numquam era errans mea domo efferret pedem . . . (Enn. scen.
253V=215J); Et alienis nominibus liberalitas Orestillae suis filiaeque copiis persol-
veret. (Sal. Cat. 35.3);
Hoc tantum bellum . . . quis umquam arbitraretur . . . confici posse? (Cic. Man.
31); . . . quis aequum censeret . . . ? (Liv. 21.19.5);
The potential use of the imperfect subjunctive (with a past meaning) is common in
conditional periods in Early Latin, decreases from the Classical period onwards, then
becomes more frequent once again in Late Latin. Examples with two imperfect sub-
junctives are (d)–(f ). In (e) and (f ) the conditional clause contains tum ‘then’, clearly
referring to the past. Exx. (g) and (h) contain an imperfect and a pluperfect; in (g)
advenissem is anterior with respect to diceres. Here the main clause contains tum. (For
the use of the imperfect subjunctive as counterfactual of the present see § 7.48.)
(g) Abi, stultu’s, sero post tempus venis. / # Igitur olim si advenissem, magis tu
tum istuc diceres.
(‘Go away, you’re being silly, you’re coming too late, behind schedule. # If I’d come
before, then you could have said this with better reason.’ Pl. Capt. 870–1)
(h) Quid opu’st verbis? Si invitare nos paullisper pergeret, / ibidem obdormis-
semus. Nunc vix vivos amisit domum.
(‘What need is there for words? If he’d continued to invite us for a bit longer, we’d
have fallen asleep there. Now he has sent us home, barely alive.’ Pl. Rud. 590–1)
Supplement:
Neque adeo haec faceremus, ni antehac vidissemus fieri. (Pl. Bac. 1209); Nam si esset
unde id fieret, / faceremus. Et tu illum tuom, si esses homo, / sineres nunc facere
dum per aetatem decet . . . (Ter. Ad. 106–8); Nam quem ferret, si parentem non
ferret suom? (Ter. Hau. 202); Quot res dedere, ubi possem persentiscere, / ni essem
lapis! (Ter. Hau. 916–17); Nam si quam Rubrius iniuriam suo nomine ac non
impulsu tuo et tua cupiditate fecisset, de tui comitis iniuria questum ad te potius
quam te oppugnatum venirent. (Cic. Ver. 1.80); Haec tu omnium mortalium pro-
fligatissime ac perditissime cum scires, cum audires cotidie, cum videres, si sine tuo
quaestu maximo fierent, cum tanto periculo tuo fieri paterere atque concederes?
(Cic. Ver. 3.65); Apollonium, hominem locupletissimum, qui si fugitivi bellum in
Sicilia facerent, amplissimas fortunas amitteret, belli fugitivorum nomine indicta
causa in vincla coniecit. (Cic. Ver. 5.18); Quod certe non fecisset, si suum numerum
naves haberent. (Cic. Ver. 5.133); De populo si quem ita rogavisses ‘Quis est in hac
civitate eloquentissimus?’ in Antonio et Crasso aut dubitaret aut hunc alius, illum
alius diceret. (Cic. Brut. 186); Quid gravius capta Lacedaemone serva tulissem, / si
raperet Graias barbara turba nurus? (Ov. Ep. 8.11–12); . . . siqui vidisset, avita / ex re
praeberi sumptus mihi crederet illos. (Hor. S. 1.6.79–80);
The existence of the potential use of the imperfect subjunctive is much debated
among Latinists. Often a diachronic development is assumed, in which the counter-
factual expression is thought to replace the potential one, but this is probably an
unjustified conclusion from the statistical observation that the counterfactual expres-
sion is more frequent in the Classical period.167
167
For a discussion of earlier literature, see Calboli (1966: 277–82, 2011: 142–5). Baratin (1981: 264–8)
denies that there is a difference between the potential and the counterfactual use of the imperfect
subjunctive and that there has been a diachronic development; he only recognizes the counterfactual
use. Lavency (2005) takes the opposite position. Vairel-Carron (1981a: 318–21) defends the unity of the
two expressions.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The past reference point of the polemical response tangerem in (c) becomes apparent
from the preceding perfect form tetigisti. For a second person, see (d). Negation in
these sentences is by non.
(b) [Et] credo, dixerit non decere gravitatem eius tam humiles ineptias.
(‘I suppose she told him that such low fooling was beneath his dignity.’ Petr. 52.10)
Supplement:
Nam si istuc ius sit quod memoras, piscatores perierint. (Pl. Rud. 978); Non illam
vir prior attigerit (Catul. 67.20); Forsitan addiderim . . . blanditias . . . (Ov. Met.
7.816–17); Sed tibi plane, Luci domine, soli omnium Chaldaeus ille vera dixerit . . .
(Apul. Met. 2.14.6);
From Early Latin onwards the perfect subjunctive is mainly used without an anterior
semantic value, much in the same way as the present subjunctive (§ 7.41). Among the
Early Latin instances, the archaic and disappearing sigmatic forms (such as ausim,
which is relatively persistent as an idiom throughout the history of the Latin language,
faxim, much rarer but still alive after Early Latin, and hapax negassim) predominate.
They never have an anterior meaning.168 By Cicero’s time, the regular perfect subjunct-
ive forms are more common than the present subjunctive forms and they increase in
frequency in later authors. This is certainly the case for first person singular forms. For
the other persons of the active paradigm there is no formal distinction from the future
perfect forms, and it is often difficult to tell what form one is dealing with (see § 7.32,
with the Appendix).169 Most instances belong to cognition and communication verbs.
Some expressions are set idioms, and one cannot really find parallels with a present
subjunctive. There are, for instance, no present subjunctive parallels for the expression
(non/haud) facile dixerim. The effect of using the subjunctive instead of an indicative is
often to make a statement milder and less direct, just as in the case of the present
subjunctive.170 Examples are (d) and (e).
(d) Iam hunc non ausim praeterire . . .
(‘Now I wouldn’t dare to walk past him . . . ’ Pl. Aul. 474)
(e) (sc. simulacrum) quo non facile dixerim quicquam me vidisse pulchrius . . .
(‘I cannot easily tell that I have seen a lovelier work of art than this . . . ’ Cic. Ver. 4.94)
168
For the persistence of ausim and faxim, see Happ (1967: 92) and de Melo (2005, 2007a: 190–215).
169
Still worthwhile is Roby’s discussion (and examples). He takes the dixerit + indefinite pronoun
expressions as future perfects (1882: ci–cvii).
170
For examples, see Blase (1903: 202–5).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Sigmatic forms: first person: Male faxim lubens. (Pl. Poen. 1091); Hau negassim.
(Pl. As. 503); Non mehercule, inquit, tibi repromittere istuc quidem ausim. (Cic.
Brut. 18); De comitiis consularibus, quae tum primum illo principe ac deinceps fuere,
vix quicquam firmare ausim. (Tac. Ann. 1.81.1); Atque ita ausim dicere: totius
ultimum totum est, quia, licet minus atque posterius sit, ipsius est. (Tert. An. 53);
Nam sapientiam existere a numero aut consistere in numero non ausim dicere.
(August. Lib. Arb. 2.11);
Third person: Quem si orem ut mihi nihil credat, id non ausit credere. (Pl. Bac. 697);
Regular forms: first person: Hegio, hoc te monitum, nisi forte ipse non vis,
voluerim. (Pl. Capt. 309); Si nunc me suspendam, meam operam luserim / et praeter
operam restim sumpti fecerim / et meis inimicis voluptatem creaverim. (Pl. Cas.
424–6); Ubivis facilius passus sim quam in hac re me deludier. (Ter. An. 203); Hoc
vero sine ulla dubitatione confirmaverim . . . (Cic. Brut. 25); Tum Brutus: ista vero,
inquit, quam necessaria fuerint non facile dixerim. (Cic. Brut. 52); Nos autem hos
eosdem motus concitati animi recte, ut opinor, perturbationes dixerimus. (Cic. Tusc.
3.7); Quibusdam tamen regionibus . . . capros et arietes optaverimus vel amplissimis
cornibus . . . (Col. 7.3); Quibus equidem adsenserim. (Vell. 1.7.2); Vix quicquam in
Sullae operibus clarius duxerim quam . . . (Vell. 2.24.4); Ergo non aliter utendum eo
quam contra venena censuerim. (Plin. Nat. 21.153); De regibus Romanis non facile
dixerim. (Plin. Nat. 33.9); Ne illos quidem inter otiosos numeraverim qui . . . (Sen.
Ben. 10.12.6); Nec abnuerim tegulae regulaeque et similium his rationem . . . (Quint.
Inst. 1.6.33); Sed non omiserim eorundem temporum rumorem, validum adeo ut
nondum exolescat. (Tac. Ann. 4.10.1); . . . cuius ex nutu, prope dixerim, pendebat
incidentium omnium salus. (Amm. 19.12.13); Sed illum gigantis alicuius fuisse
crediderim. (August. Civ. 15.9);
Second person: Hanc modestiam aequitatemque et altitudinem animi ubi nunc in
uno inveneris, quae tum populi universi fuit! (Liv. 4.6.12); Vix ullius gentis aetatis
ordinis hominem inveneris cuius felicitatem fortunae Metelli compares. (Vell.
1.11.5); Quamlibet enim apertum, quod modo et aliter intellegi possit, in illos
tyrannos bene dixeris quia periculum tantum, non etiam offensa vitatur. (Quint.
Inst. 9.2.67); Haec vulgo iactata super id quod nullo auctore certo firmantur prompte
refutaveris. (Tac. Ann. 4.11.1); Eoque omnium acerrimos facile dixeris bellatores,
quod . . . (Amm. 31.2.9);
Third person: Quis enim tibi hoc concesserit . . . ? (Cic. de Orat. 1.36); Tu igitur ipse
de te? Dixerit quispiam. (Cic. Phil. 14.13); Dixerit hic aliquis: Quid? (qui cj. Ald.) Tu
istaec, ianua, nosti . . . (Catul. 67.37—my punctuation); Ea peritis amnis eius vix
fidem fecerint. (Liv. 21.47.1); Quod crimen non modo latrocinium, verum etiam
parricidium quisque rectius nominarit. (Apul. Met. 7.3.2); Dixerit quispiam iam eos
Romae habitare solitos, quando expugnante fimbria cecidit ilium. (August. Civ. 3.8);
If the vast majority of potential perfect subjunctives do not differ from the present
subjunctive with respect to the temporal value, the question remains what relation-
ship exists between the two forms. One explanation is an aspectual one, but there is
no empirical evidence for this, and the fact that it only becomes popular in Cicero’s
time does not support such an explanation. Another explanation is that Cicero
introduced or expanded the use of the perfect subjunctive as a Graecism (it is also
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
found in Catullus 67.37 quoted above), but this is not likely either. The third is that
there is a difference in illocutionary force: the perfect subjunctive forms are supposed
to be milder or less direct than the present ones. There is no empirical support for
this idea, although this may have been the original motivation for using the perfect
subjunctive.171 The evidence needed to discern a connecting link of modest state-
ments between Early Latin and the Ciceronian period is simply lacking. Dixerim
probably meant roughly ‘It may turn out that I have said’, matching the future perfect
dixero ‘It will turn out that I have said’.
In Late Latin, the perfect subjunctive is used without a clearly recognizable anterior
meaning in the protasis of conditional periods and in temporal clauses. During this
period future perfect forms (which are formally often not distinct from the perfect
subjunctives—see above) are used in the same way. At the same time, the potential
meaning of the subjunctive is not evident either. Early examples of this development
are (f ) and (g). A Late one is (h). Note the parallelisms in mood and tense in these
examples.172
(f ) Non enim de harenoso neque calculoso neque sabulonoso luto sunt ducendi,
quod, ex his generibus cum sint ducti, primum fiunt graves, deinde, cum ab
imbribus in parietibus sparguntur, dilabuntur et dissolvuntur . . .
(‘For they ought not to be made from sandy nor chalky soil nor gravelly soil: because
when they are got from these formations, first they become heavy, then, when they
are moistened in the walls by rain showers, they come apart and are dissolved . . . ’
Vitr. 2.3.1)
(g) Ubi habet emplastri figuram medicamentum, tollitur ab igne <et>, cum
tepere coeperit, aphronitrum aspergitur sensim . . .
(‘When the medicament has the form of a plaster it is taken from the fire, and when it
begins to be tepid sodium carbonate is sprinkled little by little on it . . . ’ Larg. 157)
(h) Si maritima sit civitas et sales defuerint, liquor ex mari sumptus per alveos
aliaque patula vasa diffunditur . . .
(‘If the city is on the coast and salt is missing, water, once it is taken from the sea, is
poured out into troughs and other flat vessels . . . ’ Veget. Mil. 4.11.1)
171
See LSS: § 10.2.1.1, with references.
172
Further examples can be found in Blase (1898: 333–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(b) ‘Apud exercitum mihi fueris’ inquit ‘tot annos, forum non attigeris, afueris
tam diu et, cum longo intervallo veneris, cum his qui in foro habitarint de
dignitate contendas?’
(‘ “Have you been,” says he, “so many years with the army? You can never have been
near the forum. Have you been away so long? And then, when after a long interval
you arrive, will you contend in dignity with those who have been living in the
forum?” ’ Cic. Mur. 21)
Supplement:
Tun[e] mecum fueris? Quid illac inpudente audacius? (Pl. Am. 818); Ego te videre
noluerim? Immo vero me a te videri nolui. (Cic. Q. fr. 1.3.1); Unus homo et vestris, o
cives, undique saeptus / aggeribus tantas strages impune per urbem / ediderit? (Verg.
A. 9.783–5);
The counterfactual use of the subjunctive is found in all sentence types. For organ-
izational purposes the counterfactual use in imperative sentences will be dealt with in
a subsection of the deontic use of the subjunctive (§ 7.58). This section deals with
declarative and interrogative sentences. The counterfactual subjunctive is mainly
found in clauses belonging to a conditional period. The imperfect subjunctive
marks states of affairs that under certain conditions would be taking place at the
time of speaking, but in fact are not, because the conditions are not being met. The
pluperfect similarly marks past events that might have taken place, but did not,
because the conditions were not met. A common term for this use of the subjunctive
is coniunctivus irrealis.
173
The BTL CD-ROM has 646 instances of these first person forms, of which 277 are in Cicero.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
victurus esset. (Cic. Sen. 82); Disputarem etiam . . . quantae delectationi in rebus
adversis litterae esse deberent. Commemorarem non solum veterum sed . . .
gravissimos casus . . . etiam externos multos claros viros nominarem . . . Exponerem
etiam quem ad modum hic . . . viveremus . . . (Cic. Fam. 6.6.12–13); Cuperem ipse
parens spectator adesset. (Verg. A. 10.443); Quae fuit ante illam, mallem suprema
fuisset / nox mihi, dum potui Phyllis honesta mori. (Ov. Ep. 2.59–60); ‘Quam vellem,’
inquit, ‘Dareus aliquid ex hac indole hausisset!’ (Curt. 3.12.26); Si enim quis
dixerit: non esse quam miserum me esse mallem, respondebo: mentiris. (August.
L. Arb. 3.6);
The most frequent use of the imperfect subjunctive is in conditional periods. In (c)
and (d), both the protasis and the apodosis have imperfect subjunctive forms. In both
examples a certain state of affairs would be occurring at the respective times of
speaking if a certain condition were being met. Combinations of the imperfect with
a pluperfect subjunctive occur as well, as in (e) and (f ). The pluperfect in the apodosis
of (e) indicates that the state of affairs of the clause involved is anterior to the time of
speaking. In (f ), the pluperfect is used in the protasis and indicates anteriority with
respect to the state of affairs in the apodosis (which is contemporaneous with the time
of speaking). That the counterfactual imperfect subjunctive is anchored to the time of
speaking becomes apparent from so-called exceptions to the rule of the sequence of
tenses in subordinate clauses that are discussed in § 7.93.
id mallem quam Perseo scelestae fraudis praemium esse. (Liv. 40.56.5); Si vocem
haberes, nulla prior ales foret. (Phaedr. 1.13.8); Si fieri posset, quid sentiam ostendere
quam loqui mallem. (Sen. Ep. 75.2); Hoc si a principio dicerem, negaretur. (Ambr.
Spir.S. 1.3.52); Ego certe, quod intrepidus de meo corde pronuntio, si ad culmen
auctoritatis aliquid scriberem, sic mallem scribere, ut . . . (August. Conf. 12.31);
Combinations of imperfect and pluperfect forms: Mustius dixisset si viveret, sed
recenti re de Mustio auditum est. (Cic. Ver. 1.139); . . . a quo illi conatu non tanto
opere prohibendi fuissent, si ulla in Sicilia praesidia ad illorum adventum opposita
putarentur. (Cic. Ver. 5.5); Quod certe non fecisset, si suum numerum naves
haberent. (Cic. Ver. 5.133); Quae nisi essent in senibus, non summum consilium
maiores nostri appellassent senatum. (Cic. Sen. 19);
Non, si redisset, ei pater veniam daret? (Ter. Ph. 119); Quo quidem tempore si, ut
dixi, meum consilium auctoritasque valuisset, tu hodie egeres, nos liberi essemus, res
publica non tot duces et exercitus amisisset. (Cic. Phil. 2.37); Quod si id ante facere
conatus essem, nunc facere non possem. (Cic. Phil. 4.1); Quodsi ita fecisset, tamen post
illius mortem nihil de testamento illius novi iuris constitui oporteret. (Cic. Ver. 1.107);
The most common use of the pluperfect subjunctive is in conditional periods. This
usage is found from Early Latin onwards. Two examples to illustrate it are (c) and (d);
in both examples, the states of affairs in the apodosis would have occurred in the past
if certain conditions had been met.
(c) Pol si habuissem, satis cepissem miseriarum e liberis.
(‘If I’d had any, I’d have had my share of afflictions from my children.’ Pl. Mil. 718)
(d) Recepti in provinciam non sumus? # Quid si essetis?
(‘Have we not been received into the province? # What if you had been received?’
Cic. Lig. 23)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
QUIBUS SEI / IN·LONGA·LICUISET·TIBE UTIER VITA, / FACILE·FACTEIS SUPERASES·GLORIAM /
2
MAIORUM (CIL I .10.4–7 (Rome, Scip. El.—the person died c.170 BC)); Quo ex genere
mihi testium qui cum isto furati sunt, si uti voluissem, magna copia fuisset. (Cic. Ver.
1.97); Cuius ego, si potuissem, non parietem tantum, pectus ipsum perfodissem.
(Sen. Con. 10.6.2);
174
Including four instances of the fossilized imperative age and three instances of ‘metadirective’ fac +
subjunctive (see § 6.29). Data are taken from the LASLA website (17 March 2004). For a detailed survey
of Early Latin, see Vairel-Carron (1975: 304).
175
For a survey of the literature, see L. Löfstedt (1966: 126–32) and Calboli (2011: 145–7).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
First a few examples of commands and prohibitions with the present subjunctive.
From Early Latin (especially Cato) onwards commands occasionally contain the word
ut(i). The negator is normally ne. The sentences with a present subjunctive (with or
without ut) are usually closely embedded in the surrounding context, for example as
an answer to a preceding question (marked in italics, where this applies). It is not
always clear whether a ne-sentence is an independent prohibition or a (pseudo-)
purpose subordinate clause. The present subjunctive is rare in Cicero (most of his
instances are directed towards generic second person addressees) and other prose
authors, but relatively popular in Augustan and later poets. Horace, for example, uses
it more frequently than the present imperative, and especially for a milder form of
advice, as appears from the addressees to whom he directs his commands.176
Examples are (a) and (b). An example with ut is (c). A less binding example is (d).
(a) Quid nunc? # Argentum des, abducas mulierem.
(‘What now? # Give me the money and take the woman away.’ Pl. Ps. 1015)
(b) . . . da mihi hanc veniam, ignosce, irata ne sies.
(‘ . . . give me this pardon, forgive me, don’t be angry.’ Pl. Am. 924)
(c) Reddetur. Ne time. # At ut omne reddat. # Omne reddet.
(‘You’ll be paid. Don’t worry. # Make sure I’m paid in full. # You’ll be paid in full.’
Ter. Ad. 279–80)
(d) Dum ille ne sis quem ego esse nolo, sis mea causa qui lubet.
(‘So long as you’re not the one who I don’t want you to be, you can be anyone you
like for all I care.’ Pl. Trin. 979)
Supplement:
Commands: Valeas, tibi habeas res tuas, reddas meas. (Pl. Am. 928); Numquid vis? #
Abeas: celeriter facto est opus. (Pl. Bac. 604); Verum vivere hic non possum neque
durare ullo modo. / Proin tu me hinc abducas. (Pl. Men. 781–2); Hic apud nos hodie
cenes. Sic face. (Pl. Mos. 1129); Quiescas. (Ter. An. 598); Desinas. (Ter. Eu. 884);
Pater, omnia faciam. Impera. /# Uxorem ut ducas. (Ter. Hau. 1055–6); In primis
observes ne in terram nimium aridam aut variam, sed temperatam, semen demittas.
(Var. R. 1.42.1); Denique isto bono utare, dum adsit, cum absit, ne requiras . . . (Cic.
Sen. 32); Quo die venies, utique cum tuis apud me sis. (Cic. Att. 4.4) (utique <fac> cj.
Baiter; <velim> utique cj. Müller); Angar, excruciem me? Quid adsequar? Deinde,
quem ad finem? ‘Vivas’ inquis ‘in litteris’. (Cic. Fam. 9.26.1); Miser Catulle desinas
ineptire / et quod vides perisse perditum ducas. (Catul. 8.1–2); ‘Si te nulla movet
tantae pietatis imago, / at ramum hunc’ (aperit ramum, qui veste latebat) / ‘adgnos-
cas’. (Verg. A. 6.405–7); Primum ignosce patrio dolori, si quo inclementius in te sum
invectus; deinde sinas hic coram virgine nutricem percontari . . . (Liv. 3.48.4); Nec ego
ut nihil agatur <hortor>, sed ut agentem te ratio ducat, non fortuna. Tuae potestatis
semper tu tuaque omnia sint; armatus intentusque sis. Neque occasioni tuae desis
176
P. H. Schrijvers (p.c.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
neque suam occasionem hosti des. (Liv. 22.39.21); Accepta voces has reddidit:
hospes, / tutus eas! lapis iste prius tua furta loquetur. (Ov. Met. 2.695–6); Ab alio
expectes alteri quod feceris. (Pub. Sent. A 2); Dies quod donat timeas. Cito raptum
venit. (Pub. Sent. D 15); . . . ET SCIAS DOMO PER / B<E>NE OMNIA RECTE ESSE. (CEL 142.
col.2.47–8 (Karanis, c. AD 115)); ‘Sed heus tu,’ inquit Charit[at]e, ‘quam probe veste
contectus omnique comite viduatus prima vigilia tacitus fores meas accedas unoque
sibilo contentus nutricem istam meam opperiare, quae claustris adhaerens excubabit
adventui tuo.’ (Apul. Met. 8.10.5); Sis bonus, sis quietus, sis beatus . . . (August. Ep. 26
[3].1); Sumas laetus, ait, libensque carpas / quae . . . (Prud. Cath. 4.67–8);
Prohibitions: Ne exspectetis, spectatores, dum illi huc ad vos exeant: / nemo exibit,
omnes intus conficient negotium. (Pl. Cist. 782–3); Dabo. Molestus nunciam ne sis
mihi. (Pl. Ps. 118); Agedum eloquere, quid dare illi nunc vis? # Nihil quicquam, pater.
/ Tu modo ne me prohibeas accipere, siquid det mihi. (Pl. Trin. 369–70); At tu edepol
nullus creduas. (Pl. Trin. 606); Ne temere facias; neque tu haud dicas tibi non
praedictum: cave. (Ter. An. 205); Ne me attigas, / sceleste. (Ter. An. 789–90);
Praedium quom parare cogitabis, sic in animo habeto: uti ne cupide emas neve
opera tua parcas visere . . . (Cato Agr. 1.1); Tu quaeso quicquid novi (multa autem
exspecto) scribere ne pigrere . . . (Cic. Att. 14.1.2); Albi, ne doleas plus nimio memor /
immitis Glycerae, neu miserabilis / decantes elegos . . . (Hor. Carm. 1.33.1–3);177 At
tu deterius palles; ne sis mihi tutor; / Iam pridem hunc sepeli: tu restas. (Pers. 3.96–7);
Non rapias hoc nec testeris. Occasiones reddendi nihilo minus quaeras. (Sen. Ben.
7.16.4); Nec me putetis privatis simultatibus instinctum odio proprio saevire. (Apul.
Met. 3.3.4); . . . nec amara paternis admiscere velis. (Paul. Nol. Carm. 11.6–7); Non sis
incredibilis timori domini et ne accesseris ad illum duplici corde. (Vulg. Eccl. 1.36);
Non conrideas illi, ne doleas: et in novissimo obstupescent dentes tui. (Vulg. Eccl.
30.10); Ne sis sapiens apud temet ipsum. (Vulg. Prov. 3.7);
The status of ut(i) in instances like (c) above is not entirely clear. It is usually taken as
an adverb, functioning in an independent imperative sentence, but it looks very
much like an argument clause with verbs like rogo ‘to ask’ and curo ‘to take care’. In
(e), a clause with a subjunctive is coordinated with an ut-clause, which could be taken
as support for regarding ut as an adverb.178
(e) Continuo arbitretur uxor tuo gnato atque ut fidicinam / illam / . . . / ulcis-
care atque ita curetur usque ad mortem ut serviat.
(‘A wife should be decided on for your son at once, yes, sir, and as for that
lyre girl . . . you should take revenge on her and organize things so that she’s
a slave till death.’ Pl. Epid. 267–9)
Supplement:
Nutrices pueros infantis minutulos / domi ut procurent neu quae spectatum
adferant . . . (Pl. Poen. 28–9); Uti (praedium) bonum caelum habeat. (Cato Agr.
1.2); In Tusculanum nos venturos putamus aut Nonis aut postridie. Ibi ut sint
omnia parata. (Cic. Fam. 14.20.1);
177
On this instance, see Brink (1969: 4–6), who, however, takes it as a pseudo-purpose clause.
178
For discussion, see Risselada (1993: 147–9); for Cato, see Becker (2005: 17–18). Further examples in
OLD s.v. ut, § 43; Lodge: II.926–7. See also § 6.26, note.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The use of the perfect subjunctive is less frequent in Early Latin than that of the
present subjunctive. It becomes more frequent than the present subjunctive in Cicero,
Livy, and Seneca the Younger and continues to be used in modest numbers in
prohibitions until the second century AD; it disappears almost completely after that
period. Examples are (f) and (g). There is no difference in function between the
normal perfect subjunctive and the extra-paradigmatic forms (such as dixis) that are
found in Early Latin.179
(f) Ne dixis istuc. # Ne sic fueris.
(‘Don’t say that. # Don’t be like that.’ Pl. As. 840)
(g) Quam ob rem . . . solem alterum ne metueritis.
(‘For which reason don’t fear the second sun.’ Cic. Rep. 1.32)
Supplement:
Ill’ quidem hanc abducet. Tu nullus adfueris, si non lubet. (Pl. Bac. 90); Ubi voles
pater esse, ibi esto. Ubi noles, ne fueris pater. (Pl. Epid. 595); Ne occupassis, opsecro,
aram. (Pl. Mos. 1097); Ne destiteris currere. (Pl. Trin. 1012); Ne te admisce. Nemo
accusat, Syre, te. Nec tu aram tibi / nec precatorem pararis. (Ter. Hau. 975–6); Nihil
ignoveris . . . Nihil gratiae causa feceris . . . Misericordia commotus ne sis . . . In sen-
tentia permaneto. (Cic. Mur. 65); Tu vero ista ne asciveris neve fueris commenticiis
rebus adsensus. (Cic. Luc. 125); Nihil me existimaris neque usu neque a Theophrasto
didicisse, nisi . . . (Cic. Att. 2.9.2); Qua re fac ut ostendis. Ne destiteris ad me quicquid
tibi in mentem venerit scribere. (Cic. Att. 9.9.1); . . . unguinis (cj. Bentley; sanguinis
codd.) expertem non siris (cj. Lachmann; vestris codd.) esse tuam me / sed potius
largis affice muneribus. (Catul. 66.91–2); Si sine pace tua atque invito numine Troes /
Italiam petiere, luant peccata neque illos / iuveris auxilio. (Verg. A. 10.31–3);
Nullam, Vare, sacra vite prius severis arborem / circa mite solum Tiburis et moenia
Catili. (Hor. Carm. 1.18.1–2); Inde Iovem testem facit: ‘Si ego iniuste inpieque illos
homines illasque res dedier mihi exposco, tum patriae compotem me numquam siris
esse’. (Liv. 1.32.7); Nullum aciem, nullum proelium timueris. (Liv. 2.12.11); Ne aequa-
veritis Hannibali Philippum nec Carthaginiensibus Macedonas: Pyrrho certe aequabitis.
(Liv. 31.7.8); Nec tu credideris tantum cecinisse futura / Ampyciden Mopsum. (Ov. Met.
12.455–6); Non vis esse iracundus? Ne fueris curiosus. (Sen. Dial. 5.11.1); Numquam
credideris felicem quemquam ex felicitate suspensum. (Sen. Ep. 98.1); Ne, patres con-
scripti, ultimum Seiani diem, sed sedecim annos cogitaveritis. (Tac. Ann. 6.8.5); Intende
te illi et audi eum, et ne inobaudiens fueris ei. (Tert. Iud. 23.21); Haec, inquam, ne
dixeris, sed illa, precor te, videas. (Lucif. Reg. Apost. 7.12–13); Altiora te ne scrutaveris et
fortiora te ne exquisieris sed quae praecepit tibi Deus illa cogita semper et in pluribus
operibus eius ne fueris curiosus. (Vulg. Sirach 3.22); Ideo ne occideris eos, ne ipsius
gentis nomen extinxeris, ne quando obliviscantur legis tuae. (August. Ep. 149.9);
In Early Latin, reduplicated perfect forms and perfect forms built with -u/w- are rarely
used as prohibitives, whereas those formed with an -s- are regular. This is taken as an
indication that in prehistoric times aorist subjunctives were used as prohibitives.180
179 180
See de Melo (2007a: 213–15). So de Melo (2007c).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
7.54 The jussive use of third person present and perfect subjunctives
Third person present and, much less frequently, perfect subjunctive forms are used to
express orders, instructions, and prohibitions. They are used in the headings of
statutes, as in (b).181 They are also found in transition formulas—for example, in
Varro and Pliny the Elder.182 In Early Latin, but not only there, they are sometimes
combined with ut(i) (on which see § 7.53). The negator used is normally ne. Passive
forms are found as well. It is not always easy to decide whether the utterance is
in fact jussive and not a wish (discussed in § 7.57). Examples (a)–(c) illustrate
the present subjunctive, instances of which can be found throughout the history
of Latin.
(a) Fiat. Geratur mos tibi.
(‘All right, I’ll humour you.’ Pl. As. 40)
(b) PRAETOR UTEI INTERROGET.
(‘The praetor must hold an interrogation.’ CIL I2.583.35 (Lex Acilia, 122 BC))
(c) Nemo de nobis unus excellat.
(‘Let no single man among us distinguish himself above the rest. Cic. Tusc. 5.105)
Supplement:
Ne duit, si non volt. (Pl. As. 460); Inspiciat, si lubet. (Pl. Mos. 772); Sed utraque ut
dicat mihi. (Pl. St. 106); Ne quaeras. Ecferant quae secum huc attulerunt. (Ter. Hau.
745); Si quid desit in annum, uti paretur. Quae supersint, ut veneant. Quae opus sint
locato, locentur. Quae opera fieri velit et quae locari velit, uti imperet et ea scripta
relinquat. (Cato Agr. 2.6); Eam patinam in sole ponito. Arescat. (Cato Agr. 87.1);
. . . oderint, / dum metuant . . . (Acc. trag. 203); EISDEM IOUDICES UNIUS REI IN PERPETUOM
2 2
SIENT. (CIL I .583.27 (Lex Acilia, 122 BC)); IOUDEX NEIQUIS DISPUTET. (CIL I .583.39 (Lex
2
Acilia, 122 BC)); <RELIQUOM> IN AERARIO SIET. (CIL I .583.64 (Lex Acilia, 122 BC));
Donis impii ne placare audeant deos, Platonem audiant, qui . . . (Cic. Leg. 2.41);
eademque (sc. monstra) de sideribus, quae reprehensa in alio iam in hoc omittantur.
(Cic. N.D. 1.28); Suum quisque igitur noscat ingenium acremque se et bonorum et
vitiorum suorum iudicem praebeat . . . (Cic. Off. 1.114); (sc. C. Marius) . . . non
semel respondit, sed saepe: ‘moriatur’? (Cic. Tusc. 5.56); In Tusculanum nos
venturos putamus aut Nonis aut postridie. Ibi ut sint omnia parata. (Cic. Fam.
14.20); Ubi mane experrectus est, corpus eius leniter ex oleo vetere, cum
capite excepto ventre, permulceatur. Tum ambulatione quam maxime longa et
recta utatur. (Cels. 3.23); . . . Liberi parentes alant aut vinciantur. (Sen. Con. 7.4.
tit.); Per vim metumque gesta ne sint rata. Pacta conventa legibus facta rata
sint. (Sen. Con. 9.3.tit.); Nam et reliqua mediterranea eius dicantur. (Plin. Nat.
6.157);
181
The distribution of the subjunctive and the future imperative forms in statutes is discussed in
Daube (1956: 50–6).
182
Transition formulas in Pliny the Elder are discussed by Naas (2002: 209–24).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The perfect subjunctive is much less frequently used in jussive utterances, and non-
negated instances are rare (a few can be found in the Supplement).183 In some of the
instances the perfect is easily interpreted as having its normal anterior semantic value,
as in (d); in others the expression seems to have become idiomatic. Given the
infrequency of such utterances, there is no need for a detailed discussion. The old
extra-paradigmatic forms like faxint in (e) can better be regarded as equivalent to
present subjunctive forms.184
(d) Sit enim mihi tinctus litteris, audierit aliquid, legerit . . .
(‘I would have him be a man of some learning, who has done some listening and
some reading . . . ’ Cic. de Orat. 2.85)
(e) Sed te optestor, Hegio, / ne tuom animum avariorem faxint divitiae meae.
(‘But I entreat you, Hegio, do not let my riches make you greedier.’ Pl. Capt. 319–20)
Supplement:
Quod ad temporum vocabula Latina attinet, hactenus sit satis dictum. (Var. L. 6.35);
Sed de inconstantia totius illorum sententiae—si ulla sententia cuiusquam esse potest
nihil adprobantis—sit ut opinor dictum satis. (Cic. Luc. 29); Eae (sc. litterae) te ne
moverint. (Cic. Att. 16.1.6); Nec a me nunc quisquam quaesiverit, quid ita
spoponderim . . . (Liv. 9.9.9); Debita sit facto gratia nulla meo . . . (Ov. Ep. 10.142);
Et hactenus hortensia dicta sint, ciborum gratia dumtaxat. (Plin. Nat. 19.189); Quod
idem dictum sit de oratione libera . . . (Quint. Inst. 9.2.27); Sed nemo annales nostros
cum scriptura eorum contenderit qui veteres populi Romani res composuere. (Tac.
Ann. 4.32.1); . . . tuis honos sit habitus sanctissimis auribus . . . ([Quint.] Decl. 3.1);
Omnibus dictum sit: quaerite et invenietis. (Tert. Pr. Haer. 9.1); Hoc dictum sit et de
secundo, ut tempora sileantur aut quatuor, aut duo. (August. Mus. 4 [1135.14]);
183
The morphologically perfect forms meminerim, noverim, and oderim behave like present
subjunctives.
184
See de Melo (2005).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
185
For the use of the subjunctive in wishes, see Mesa (1998a).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(Liv. 6.26.6); Libidine vero ne maribus quidem cedunt, pati natae, di illas deaeque male
perdant. (Sen. Ep. 95.21); Aediles male eveniat, qui cum pistoribus colludunt ‘serva
me, servabo te’. (Petr. 44.3 (Ganymedes speaking));
A pereat, si quis lentus amare potest! (Prop. 1.6.12);
Ita me di bene ament, ut ego vix reprimo labra, . . . (Pl. Cas. 452); Poti’n ut taceas? /
Ita me di ament—# Ita non facient. Mera iam mendacia fundes. (Pl. Ps. 942); Vi’n
scire? At ita me servet Iuppiter, / ut . . . (Ter. Ph. 807–8); Saepe, ita me di iuvent, te
non solum auctorem consiliorum meorum verum etiam spectatorem pugnarum
mirificarum desideravi. (Cic. Att. 1.16.1);
Qui illum di omnes perduint! (Ter. Ph. 123);
Ut te di omnes infelicent cum male monita memoria! (Caecil. 110); Haec fient. #
Ut modo fiant! # Fient. Me vide. (Ter. Ph. 711);
Other wishes: Valeant / qui inter nos discidium volunt. (Ter. An. 129–30); Valeant,
inquit, cives mei, valeant; sint incolumes, sint florentes, sint beati; stet haec urbs
praeclara . . . (Cic. Mil. 93); Di faciant ut multa tua sit epistula dextra / scripta, sed e
multis reddita nulla mihi. (Ov. Tr. 4.7.9–10); Dominus te servet. Saluto te. Ora pro
me. (Publicola in August. Ep. 46.18); Sic pereant omnes inimici tui, domine. (Vulg.
Iud. 5.31);
. . . utinamque non peiora vincant. (Quint. Inst. 9.3.1); Sed utinam, inquit, revivis-
cat frater . . . (Gel. 10.6.2); Utinam quidem nec videre possimus quae facere nobis nefas
est. (Tert. Idol. 16); —quod utinam et ipsorum saluti prosit— (August. Cons. 1.7.10);
Invocations of a god or gods occur almost exclusively in the third person, and almost
all of them are formulaic with sigmatic forms (which function as present tense
forms—see § 7.45). Even though originally these forms may have denoted anteriority,
they stood outside the tense system already in Early Latin. After this period only the
faxint forms remained in use with some frequency. Examples are (e)–(h).
(e) Ne di sirint.
(‘May the gods forbid.’ Pl. Bac. 468)
(f ) Utinam di faxint infecta dicta re eveniant tua.
(‘May the gods take care that your words are rendered null and void by reality.’ Pl.
Am. 632)
(g) Id te Iuppiter / prohibessit.
(‘Jupiter forbid!’ Pl. Ps. 13–14)
(h) Di immortales faxint ne sit alter!
(‘God grant there may never be another!’ Cic. Ver. 3.81)
Supplement:
Ita di deaeque faxint. (Pl. Capt. 172); At vos Salus servassit. (Pl. Cist. 742); Qui illum
di deaeque magno mactassint malo. (Enn. scen. 333V=287J); ‘Iuppiter pater, si est fas
hunc Numam Pompilium, cuius ego caput teneo, regem Romae esse, uti tu signa
nobis certa adclarassis inter eos fines quos feci’. (Liv. 1.18.9); Utinam istuc ita di
faxint! (Ter. Hec. 354); Ne ille sirit Iuppiter te ea perseverare. (Cornelia Nep. fr. 2);
Quod promittis di faxint! Quid enim mihi meis iucundius? (Cic. Att. 15.29.1); Nec di
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a)–(d). In Early Latin there are a few instances in which the imperfect has its
common value of situating an event in the past, as in (e). The utterances involved
contain the particle utinam. Unrealizable wishes without utinam are very rare.186 The
negation word is ne. Instances are found in all periods.
(a) Homo hic ebrius est, ut opinor. / # Utinam ita essem.
(‘This man’s drunk, I think. # I wish I were.’ Pl. Am. 575)
(b) Utinam istuc pugni fecissent tui.
(‘Oh, how I wish your fists had done the same!’ Pl. Am. 386)
(c) Illud utinam ne vere scriberem . . .
(‘I wish it were not true to add that . . . ’ Cic. Fam. 5.17.3)
(d) Atque utinam res publica stetisset . . .
(‘I would that the government had stood fast . . . ’ Cic. Off. 2.3)
(e) Utinam te di prius perderent quam peri<i>sti e patria tua . . .
(‘I wish the gods had destroyed you before you disappeared from home . . . ’ Pl. Capt. 537)
Supplement:
Imperfect subjunctive: Utinam fortuna nunc anetina uterer, . . . (Pl. Rud. 533);
Utinam istuc verbum ex animo ac vere diceres . . . (Ter. Eu. 175); Utinam quidem
aliquid velletis esse populare! (Cic. Phil. 1.21); Utinam ego, inquit, tertius vobis
amicus adscriberer! (Cic. Tusc. 5.63); Atque utinam rex ipse noto compulsus eodem /
adforet Aeneas. (Verg. A. 1.575–6); Atque utinam pro decore tantum hoc vobis et non
pro salute esset certamen. (Liv. 21.41.13); O utinam arguerem sic, ut non vincere
possem! (Ov. Am. 2.5.7); Utinam non hoc illum liberaret metu quod iudicis suae
clementiam novit. (Sen. Con. 7.8.6); Utinam quidem sequerentur. (Sen. Ep. 104.17);
Iocari ego senem poetica levitate credebam, cum ille ‘utinam quidem sufficeret largior
scaena . . . ’ (Petr. 117.2); Utinam ne in saeculo quidem simul cum illis moraremur. (Tert.
Spect. 15.8); Utinam nec ipsa res loqueretur. (Amm. 29.2.13); Tu autem naturae
defensor—quod utinam esses, non quasi sanae defensionem falsam adhiberes, sed pro
nondum sana medicum rogares—nunc vero . . . (August. Serm. 30.5);
Pluperfect subjunctive: Utinam nasum abstulisset mordicus! (Naev. Com. 43R); Quod
utinam ne Phormioni id suadere in mentem incidisset . . . (Ter. Ph. 157); Utinam
igitur . . . sic istam calliditatem hominibus di ne dedissent, . . . (Cic. N.D. 3.75); Atque
utinam ex vobis unus vestrique fuissem / aut custos gregis aut maturae vinitor uvae!
(Verg. Ecl. 10.35–6); O utinam, Macareu, quae nos commisit in unum, / venisset leto
serior hora meo. (Ov. Ep. 11.21–2); Utinam quidem eiusmodi tempestatibus produc-
tus esses, puer! Tu potius matrem reliquisses. (Quint. Decl. 388.14); Atque utinam ego
potius filio iuveni quam ille patri seni cessisset. (Tac. A. 3.16.3); Atque utinam vesperi
de more nobis parassem corollas aliquas, ne moram talem patereris vel noctis unius.
(Apul. Met. 3.25.4); Atque utinam nullas haereses oportuisset existere, ut probabiles
186
Blase (1903: 154) mentions Catul. 2.9 possem and Ov. Ep. 8.34 posset as the only exceptions without
utinam he is aware of, but both instances are textually uncertain.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
quique emicarent. (Tert. An. 3.1); Atque utinam hoc contenta fuisset Invidia, tur-
barum acerrima concitatrix . . . (Amm. 21.13.12); Utinam non fuisset miseria, ne ista
esset misericordia necessaria. (August. Nat. et grat. 25.28);
Perfect subjunctive: Liceat esse miseros—quamquam hoc victore esse non possu-
mus; sed non loquor de nobis, de illis loquor qui occiderunt—, fuerint cupidi,
fuerint irati, fuerint pertinaces. Sceleris vero crimine furoris parricidii liceat Cn.
Pompeio mortuo, liceat multis aliis carere. (Cic. Lig. 18); Age sit ita factum. Quae
causa fuit cur Romam properaret, cur in noctem se coniceret? (Cic. Mil. 49);
‘Malus civis, improbus consul, seditiosus homo Cn. Carbo fuit’. Fuerit aliis. Tibi
quando esse coepit? (Cic. Ver. 1.37); Milia frumenti tua triverit area centum, / non
tuus hoc capiet venter plus ac meus. (Hor. S. 1.1.45–6); Sed data sim, quia danda
fui. Tot noctibus absum / nec repetor. (Ov. Ep. 3.21–2); ‘Patrocinium hoc voluit,
qui controversiam finxit.’ Fortasse enim noluit. Sed esto, voluerit. Continuone, si
ille stulte cogitavit, nobis quoque stulte dicendum est? (Quint. Inst. 9.2.84); Re-
liquerit haec sane deus humanis ingeniis eruenda, tamen fieri non potest quin
ipsius sint omnia . . . (Lact. 1.18.20);
(c) Cur enim, si rem publicam egere libertate senatoria crederet, tam levia
consectaretur? Quin de bello aut pace, de vectigalibus et legibus, quibusque
aliis Romana continentur, suaderet dissuaderetve?
(‘Why, if he thought that the public welfare required freedom of speech in the Senate,
did he pursue such trifling abuses? Why should he not speak for or against peace and
war, or on the taxes and laws and other matters involving Roman interests?’ Tac.
Ann. 13.49.2)
Supplement:
Qui si inprobasset cur ferri passus esset? (Caes. Civ. 1.32.3); Postremo quis hoc sibi
persuaderet sine certa spe Ambiorigem ad eius modi consilium descendisse? (Caes.
Gal. 5.29.5); Quam putarent continuatae militiae causam esse? Nullam profecto
aliam inventuros, quam ne quid per frequentiam iuvenum eorum in quibus vires
omnes plebis essent, agi de commodis eorum posset. (Liv. 5.2.5); Attalum vero, qui
aetate proximus sit, quis non pro rege habeat? (Liv. 45.19.11); Cur enim e rostris
fratris domum, imminentem foro et inritandis hominum oculis, quam Aventinum et
penates uxoris petisset? (Tac. Hist. 3.70.1);
The examples show that the future imperative paradigm includes third person
forms, which, unlike the second person forms, are found with explicit subjects, as
in (c) and (d).
(c) Uxor mea heres <ne> esto . . .
(‘Let my wife not be my heir . . . ’ Sen. Con. 2.7.9)
(d) Nocturna mulierum sacrificia ne sunto praeter olla quae . . .
(‘No sacrifices shall be performed by women at night except those which . . . ’ Cic. Leg.
2.21)
It has been shown that the third person future imperative forms are almost never used
((c) is exceptional) to refer to definite third persons (for which the third person
present subjunctive form is required). The present and future imperative forms are
different in other respects as well (see § 7.64).188
Imperative verb forms are typically found in imperative sentences. There they are
in competition with subjunctive forms (see §§ 7.51–7.55). The conditions for the
choice between an imperative form and a subjunctive one are described in § 6.29 (i).
The negator used in prohibitions is usually ne or a word with an incorporated
negative element. Details can be found in § 8.5.
There are a few remarkable instances of the use of aliquis with second person plural
imperative forms, as in (e).
187
I shall refer to future imperative forms as -to forms, ignoring the other forms. For the usefulness of
the word ‘passive’ in this context, see § 7.66.
188
See Risselada (1993: 130–7) with further references.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
commands and cave + subjunctive or noli + infinitive for prohibitions (see 6.29).
Thirdly, both declarative and interrogative sentences can be used with an indirect
directive illocutionary force; this is discussed in § 6.4 and § 6.21, respectively.
In Cicero imperatives are found in relative clauses. An example is (f ).189
(f ) Hic ille iam vertetur orbis, cuius naturalem motum atque circuitum a primo
discite adgnoscere.
(‘At this point begins that orbit of development. Learn to recognize the natural
motion and circular course of this from its beginning.’ Cic. Rep. 2.45)
The distributional data in Table 7.12 show that the future imperative is typical of
didactic and legal texts. Whereas in Cato, who uses the future imperative very often,
and, as far as the small numbers allow one to state anything with certainty, in Plautus,
future imperative forms are not restricted to particular verbs, Seneca and Petronius
have many instances, relatively speaking, of just a few verbs, namely scito(te) and
memento(te), for which no present imperative counterparts exist, as well as of habeto
(te) and esto(te), forms that made it into Late Latin.190 Third person future imperative
forms are extremely rare in Late Latin, and there is no trace of them in the Romance
languages.
Imperative forms are very unevenly distributed over positive and negative sentences
in Early Latin. Table 7. 13 shows that, in comparison with the subjunctive forms, the
present imperative is not often used in prohibitions in Plautus and Terence, and
the future imperative is almost absent.
In § 6.29 a distinction is made between proper directive use of imperative forms
and ‘metadirective’ use of imperative forms. A number of imperative forms are also
frequently found in idiomatic combinations with subjunctive forms, such as fac
(venias), cave ([ne] cadas), vide (ne), or with a present infinitive, such as noli.
189
For more instances, and for an exceptional imperative mementote in a tametsi subordinate clause,
see Lebreton (1901a: 306).
190
So in Augustine, especially the Psal., Ep., and Serm. (data CLCLT-4). For the historical development
of scito, esto, and habeto, see L. Löfstedt (1966: 42–59). For exceptional third person forms, see L. Löfstedt
(1966: 31–2).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Table 7.14 gives an idea of the frequency of such expressions.191 The description
below of the semantic values of the present and future imperatives ignores these uses.
191
The table is based on the material of LASLA (20-01-2006). Some of the instances are used as proper
directives.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
192
Quoted from Risselada (1993: 122).
193
Quoted from Bennett (1910: I.352), who gives a very detailed classification. See also Risselada
(1993: 153).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(Ter. Eu. 713–15); Si omnis uno ordine habetis Achivos, / . . . iamdudum sumite
poenas. (Verg. A. 2.102–3); Iamdudum194 pecca, si mea fata petis! (Ov. Ep. 19.118);
Prohibitions with a present imperative (type ne time) are quite common in Plautus
and Terence. They are typically used in situations in which someone is urged to stop
doing something in which the person is already engaged (so-called INHIBITIVE use), as
in (i). This explains its total absence from juridical and official texts and from Cato’s
de Agri Cultura—all of which contain only warnings and general advice. Further-
more, its low frequency in later stages of Latin is perhaps also determined by the type
of texts. Urging someone to stop doing something is easier with verbs referring to
externally observable behaviour than with verbs referring to internal mental activities
and processes. For the latter type of verbs Plautus uses an alternative, viz. the present
subjunctive: ne speres (‘do not hope/expect’ Pl. Mil. 1423). Prohibitions with a present
imperative fall out of use after Plautus and Terence, except in poetry, where they are
not restricted to inhibitions, as in (j). The expression becomes very rare from the
second century AD onwards.
Vairel-Carron (1975: 192–213) discusses all the relevant instances from Early Latin.
There are sixty-five in Plautus and sixteen in Terence. As it happens, there is only one
instance of ne in combination with the verb facio ‘to do’, viz. Pl. St. 20. Many instances
in Plautus refer to observable emotions and expressions of emotion. This high fre-
quency is to be explained by the fact that these are the kinds of situations Plautus likes.
(j) Tu ne cede malis, sed contra audentior ito / qua tua te Fortuna sinet.
(‘Yield not to ills, but go forth all the bolder to face them as far as your destiny will allow!’
Verg. A. 6.95–6—NB: for reading qua instead of quam of the codd., see Austin ad loc.)
Supplement:
NEI PARI MED (AE 1998, 348 (so-called Garigliano bowl, c.500 BC—if correct));195
Tace. / Ne corrumpe oculos, redibo actutum. # Id actutum diu est. (Pl. Am. 529–30);
Ne lacruma, soror, neu tuo id animo / fac quod tibi [tuos] pater facere minatur. (Pl.
St. 20–1); Per te deos oro . . . / ut me adiuves in hac re atque ita uti nuptiae / fuerant
futurae, fiant. # Ah, ne me obsecra. (Ter. An. 538–43); Ego pol hodie, si vivo, tibi /
ostendam erum quid sit pericli fallere / et illi patrem. # Ah, ne saevi tanto opere. (Ter.
An. 866–8); Vidin’ ego te modo manum in sinum huic meretrici / ingerere?
# . . . Mene? # Hisce oculis, ne nega. (Ter. Hau. 563–4); Age quaeso, ne tam offirma
te, Chremes. (Ter. Hau. 1052); Deiphobus contra ‘Ne saevi, magna sacerdos.’ (Verg.
A. 6.544); Crastino die oriente sole redite in aciem. Erit copia pugnandi. Ne timete.
(Liv. 3.2.9.); Ne metue dominam famula. (Sen. Ag. 796); Cetera parce queri neu me
meliora secutam / argue. (V. Fl. 7.225–6);
194
Instances of iamdudum with an imperative (‘at once’) can be found in TLL s.v. dudum 2180.10ff.
195
For the reading, see Vine (1998) with references.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Non-inhibitive instances: Vigila, ne somno stude. (Pl. Mil. 215); Ne parce vocem, ut
audiat. (Pl. Mil. 1220); Habe tuum negotium, nec quid res mea familiaris postulet,
quam ego non curo, sed quid velim et cur velim existima. (Cic. Att. 12.22.3); Ne
repugnate vestro bono et hanc spem, dum ad verum pervenistis, alite in animis
libentesque meliora excipite et opinione ac voto iuvate. (Sen. Dial. 2.19.4); Immo vos
non recedite, sed accedite. (August. Ep. 105.1);
Apart from a few idioms, the future imperative forms (second and third person, see
§ 7.63) occur more or less freely only in older periods of Latin and in certain text
types. Third person forms are found (alongside subjunctive forms) in legal statutes
(where they are the required form in the body of the text containing the rules)196 and
in prayers. Second person forms are quite common in Early Latin comedy and in
writings on agriculture, especially in Cato’s manual—possibly introduced for that
purpose by Cato himself197—and in other didactic texts. They remain in use in these
types of text.198 Literary authors sometimes use the -to(te) forms to achieve an effect
of solemnity. The forms are also used as a metrical commodity in poetry and
occasionally in clausulae in prose.199 Examples are (a)–(e).
(a) Si in ius vocat, ito. Ni it, antestamino. Igitur em capito.
(‘If plaintiff summons defendant to court, he shall go. If he does not go, plaintiff shall
call witness thereto. Then only shall he take defendant by force.’ Lex XII I.1)
(b) . . . ET EIDEM IN DIEBUS V APUD Q(AESTOREM) IOURANTO PER IOVEM DEOSQUE.
(‘ . . . and let the same within five days, in the presence of a quaestor, take the oath by
Jupiter and the gods.’ CIL I2.582.17 (Tab. Bant., end II BC))
(c) Si a viro tibi forte veniet nuntius, facito ut sciam.
(‘If by chance a message comes to you from your husband, let me know.’ Pl. St. 148)
(d) Abi quaeso hinc domum. # . . . # Redito huc circiter meridie.
(‘Please go home. # . . . # Return here around midday.’ Pl. Mos. 578–9)
(e) Umbram ab sulcis removeto crebroque fodito.
(‘Keep the furrows clear of shade, and cultivate frequently.’ Cato Agr. 33.3)
In Early Latin, and certainly in Plautus and Terence, the future imperative often refers
to a time after the communicative situation in which it is uttered and often only after
196
Daube (1956: 50–6) has shown that the future imperative is not used in the headings of statutes.
They are not found in praetorian and aedilician edicts either (Watson 1991).
197
For arguments in favour of this hypothesis, see Daube (1956: 93–7).
198
For its persistence as a text type characteristic in Late Latin medical texts, see Önnerfors (1989: 135,
n. 9). For Pelagonius and other veterinarian texts, see Adams (1995b: 460–8). For Apicius, see Grocock
and Grainger (2006: 98–102).
199
As an instance of solemnity, L. Löfstedt (1966: 29–30) cites Pl. Capt. 452 Bene ambulato; as an
instance of its use in a clausula, Cic. Att. 1.12.4 Si rem nullam habebis, quod in buccam venerit scribito.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
specific conditions are met (for instance applying only to certain persons). This is
shown by elements in the immediate context, such as:
(i) co-occurrence with other tense forms or temporal expressions that refer to
the future, including conditional clauses (indicated in italics in the
examples); this phenomenon is very common. Ex. (c) provides an illustra-
tion. Six out of Cicero’s ten cases of dicito (‘say’, ‘tell’) occur in combination
with a conditional clause;
(ii) the fact that it is preceded by a present imperative form which refers to the
immediate future within the communicative situation (indicated in bold
italics in the examples), see (d);
(iii) the distribution of a number of -to forms over a longer stretch of dialogue
(Ter. Ad. 575–87).
This is compatible with its use in legal texts and in manuals, where it also never refers
to the immediate future. Outside legal contexts the urgency of the request may vary,
just as with the present imperative.
The future imperative rapidly decreases in frequency after Early Latin, and the
temporal distinction between the future and present imperative disappears and is
often difficult to establish in Cicero and later authors.
Supplement:
(Pseudo-)legal texts (and oaths): QUAESTOR MORAM NEI FACITO (CIL I2.583.69
(Lex Acilia, 122 BC)); . . . NEPOTESQUE <TU>M EIEI FILIO GNATEIS CEIVIS ROMANEI IUSTEI
SUNTO . . . <SUFRAGIU>M FERUNTO INQUE EA<M> TRIBUM CENSENTO, MILITIAEQUE EIS VOCATIO
2
ESTO . . . (CIL I .583.77 (Lex Acilia, 122 BC)—NB: for censento, see § 7.66 Note);
. . . IS·EUM·AGRUM·NEI·HABETO·NIVE·FRUIMINO. (CIL I2.584.32 (Sent. Minuc., Genoa,
118 BC)); . . . IDQUE·EI·SINE·FRAUDE·SUA·FACERE·LICETO . . . (CIL I2.587.4 (Lex. Corn., 81
BC)); Si prior defexit publico consilio dolo malo, tum illo die, Diespiter, populum
Romanum sic ferito, ut ego hunc porcum hic hodie feriam; tantoque magis ferito,
quanto magis potes pollesque. (Liv. 1.24.8); Belli gerendi ius Antiocho ne esto cum
illis qui insulas colunt, neve in Europam transeundi. Excedito urbibus agris vicis
castellis cis Taurum montem usque ad Tanaim amnem, et ea valle Tauri usque ad
iuga qua in Lycaoniam vergit. Ne qua <praeter> arma efferto ex iis oppidis agris
castellisque quibus excedat. Si qua extulit, quo quaeque oportebit recte restituito.
(Liv. 38.38.3–5);
Other texts: Ergo mox auferto tecum (sc. pallam), quando abibis. (Pl. Men. 430);
Mea quidem hercle causa vidua vivito. (Pl. Men. 727); Principio, si id te mordet . . .
/ . . . quaeso hoc facito tecum cogites (Ter. Ad. 807–8); Proin tu, taceri si vis, vera
dicito. (Ter. Eu. 106); Sed pater egreditur. Cave quicquam admiratus sis, / qua causa
id fiat. Obsecundato in loco. / Quod imperabit facito. Loquitor paucula. (Ter. Hau.
826–8); Vola. / . . . Quid si non veniet? Maneamne usque ad vesperum? / # Maneto.
Curre. # Non queo: ita defessus sum . . . (Ter. Hec. 438–43); Sic dabo: age nunc,
Phormionem qui volet lacessito. (Ter. Ph. 1027); Videto quam minimi instrumenti
sumptuosusque ager ne siet. (Cato Agr. 1.5); Oblinito amphoram et post dies XXX
aperito et utitor. (Cato Agr. 127.1); In fundo dolii aut seriae sale sternito. Deinde
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
pernam ponito, cutis deosum spectet, sale obruito totam. (Cato Agr. 162.1); Tum
auctoritatem censoriam amplexato [NB: active deponent form], tum illorum iudi-
cium cum re coniunctum esse defendito. (Cic. Clu. 124); Cum ego P. Granium
testem produxero, . . . refellito, si poteris. (Cic. Ver. 5.154); Qua re cum me adflictum
et confectum luctu audies, existimato me stultitiae meae poenam ferre gravius quam
eventi . . . (Cic. Att. 3.8.4); Hoc iam sic habeto, [nec] hoc exercitu hic tanta negotia
geri potuisse. (Cic. Att. 5.20.1); Duc nigras pecudes. Ea prima piacula sunto. (Verg. A.
6.153); Brevitatem ne reformidaveris, sed qualemcunque virgam, cuius cacumen in
imum scrobem pervenit, deprimito et obruito. (Col. Arb. 7); . . . homines . . . scitote in
duas partes esse divisos. (Petr. 116.5–6); ‘Oro te,’ inquam, ‘aulaeum tragicum dimo-
veto et siparium scaenicum complicato et cedo verbis communibus.’ (Apul. Met.
1.8.5); Scito te isto nomine male velle et male abominari . . . (Tert. Apol. 34.4); Esto
fidelis ad mortem usque et dabo tibi vitae coronam. (Tert. Scorp. 12.6; cf. Vulg. Apoc.
2.10); Gestorum hic textus est, quem mente, quaeso, accipito placida. (Amm.
20.8.11); Cum videritis haec omnia, scitote quia prope est in ianuis. (August. Ep.
199.49 = Vulg. Mat. 24.33); Ibi enim dictum est: memento diem sabbati sanctificare
eum. (August. Serm. 8); Memor esto itaque, unde excideris, et age paenitentiam et
prima opera fac. (Vulg. Apoc. 2.5); Sanguinem emittito . . . (Mulom. Chir. 47);
In prohibitions the use of the future imperative is very rare, except in legal texts (and
in Cato in ‘paralegal’ instructions). An example is (f ). The aspect of ‘non-immediacy’
mentioned in § 7.64 is not relevant in Cato.200 The expressions nolito and nolitote
with an infinitive occur as an alternative form of prohibition, but are rare as well.
Originally, in respect of their meaning, they constituted an explicit appeal to the
goodwill of the addressee, as in (g). They are absent from ecclesiastical authors.201
Caveto ne and videto ne are mainly used by Cato as a metadirective (see § 6.29);
however, an example of normal directive use, without ne, is (h).
(f ) Hominem mortuum in urbe ne sepelito neve urito.
(‘You shall not bury or cremate a dead man within the city.’ Lex XII 10.1)
(g) Heus tu, si quid per iocum / dixi, nolito in serium convortere.
(‘Hey you, if I said anything in jest, don’t turn it into earnest.’ Pl. Poen. 1320–1)
(h) Caveto, cum ventus siet aut imber, effodias aut feras. Nam id maxime
cavendum est.
(‘Be careful not to dig them up or transport them when the wind is blowing or when
it is raining, for this is especially to be avoided.’ Cato Agr. 28.1)
Supplement:
Neu quisquam posthac prohibeto adulescentem filium / quin amet et scortum ducat,
quod bono fiat modo. (Pl. Mer. 1021–2); Quod granum capiat ne serito, et circum
capita addito stercus, paleas, vinaceas, aliquid horum, quo rectius valeat. (Cato Agr.
33.3); Oleam ne stringito neve verberato iniussu domini aut custodis. Si adversus ea
quis fecerit, quod ipse eo die delegerit, pro eo nemo solvet neque debebitur. (Cato
200 201
See Becker (2005: 10–13). At least those on CLCLT-1.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agr. 144.1); Socium ne quem habeto nisi quem dominus iusserit aut custos. (Cato
Agr. 145.3); Non petito bene sit . . . (Ov. Pont. 1.2.103); Ficum frigoribus ne serito.
(Col. 5.10); Sed hoc flante ne arato, frugem ne serito, semen ne iacito. (Plin. Nat.
18.334);
Nolito multum dare, ne pertaedescat, uti possit porro libenter esse. (Cato Agr.
156.6); Sin insidias fieri libertati vestrae simulatione largitionis intellegetis, nolitote
dubitare plurimo sudore et sanguine maiorum vestrorum partam vobisque traditam
libertatem nullo vestro labore consule adiutore defendere. (Cic. Agr. 2.16); Quam ob
rem nolito commoveri si audieris me regredi, si forte Caesar ad me veniet. (Pomp.
Att. 8.12C.2); A Kal. Novemb. gallinis ova supponere nolito, donec bruma conficia-
tur. (Plin. Nat. 18.231); Invidere nolito. (Mart. 2.74.4); . . . adiectis etiam petulantibus
dictis: ‘sic dividere cum fratre nolito’. (Petr. 11.4);
True present passive imperative forms are very exceptional,202 and one would not
expect them at all, because passivity and controllability do not go well together
(see § 6.30). Virgil has (a), where terrere is coordinated with absiste. There are a
few instances in the Vulgate. The picture is the same for the likewise rare decausative
passive cases, as in (b), and the slightly less rare (and also not unexpected) auto-
causative ones, as in (c) and (d). Future passive imperatives are equally rare.203 Cicero
has (e), apparently intended as a third person plural form, in the style of the
Twelve Tables.
(a) . . . hic tibi certa domus, certi, ne absiste, penates. / Neu belli terrere minis.
(‘ . . . here your home is sure—draw not back—and sure are your gods! Be not scared
by threats of war.’ Verg. A. 8.39–40)
(b) Lassare et disce sine armis / posse pati.
(‘Grow weary; learn to find life endurable without fighting.’ Luc. 5.313–14)
(c) Tute hoc intristi. Tibi omne’st exedendum. Adcingere.
(‘You mashed up this mess. You must eat it all up. Gird yourself.’ Ter. Ph. 318)
(d) Facesse hinc Tarquinios . . . Devolvere retro ad stirpem . . .
(‘Be off from here to Tarquinii! Sink back into the rank of your family . . . ’ Liv. 1.47.5)
(e) Regio imperio duo sunto, iique <a> praeeundo, iudicando, consulendo
praetores, iudices, consules appellamino.
(‘There shall be two magistrates with royal powers. Since they lead, judge, and confer,
from these functions they shall be called praetors, judges, and consuls.’ Cic. Leg. 3.8)
202
Instances of passive imperative forms are discussed by K.-St.: I.108; L. Löfstedt (1966: 83ff.); Bergh
(1975, 1990).
203
See K.-H.: 680.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
True passive: Quaecumque pestis sive quaecumque es fera, / palam timere. (Sen.
Her. O. 1260–1); Ascende . . . Adire. (Plin. Pan. 60.2); . . . venite et curamini . . . (Vulg.
Luc. 13.14);
Autocausative passive: Ergo age, care pater, cervici imponere nostrae. (Verg. A.
2.707); Desine iam tandem precibusque inflectere nostris. (Verg. A. 12.800); Proinde
in hoc discrimen, si iuvat, accingere . . . (Liv. 2.12.10); Sic ergo formare ut scias non
posse te consequi ut sis impenetrabilis. (Sen. Nat. 4a.pr.5); . . . humiliare . . . (Vulg.
Gen. 16.9 (NB: humilia te Vetus Latina));
A remarkable active form, which must be taken as a true passive, is censento in
<SUFRAGIU>M FERUNTO INQUE EA<M> TRIBUM CENSENTO, MILITIAEQUE EIS VOCATIO ESTO . . .
(CIL I2.583.77 (Lex Acilia, 122 BC)).204
In (a) the subject of dicet will claim that his action of bringing takes place at the same
time in the future as when he makes this report; in (b) the person says that he has
arrived before the time of speaking at which his arrival is reported; and in (c) the
murdering is to take place after the time of speaking. These examples represent the
overall situation for the active verb forms. However, the Latin verbal system does not
provide an equally simple picture for the passive, where a fully operational future
infinitive is lacking. The frequency with which the various infinitives are used can be
seen in Figure 7.3.
204
See Leumann (1977: 571).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
80
70
60
50
40
30
20
10
0
Pl. Am. Cic. Dom. Caes. Gal. Sen. Clem. Verg. A.
Praes. Act. Praes. P. Perf. Act. Perf. P. Fut. Act. Fut. P
Figure 7.3 clearly demonstrates that the future passive infinitive is almost non-
existent.205 In §§ 7.72–7.74 a few alternative expressions can be found. The present
infinitive is by far the most frequent; this is a consequence of its frequent use in
combination with modal and other auxiliary verbs and related expressions (its
so-called prolative use) (see § 7.70).
On closer inspection, the general picture of the infinitives is not in fact as clear as it
at first appears. The present infinitive, for example, is used as an equivalent of a finite
past verb form (the ‘historic’ infinitive). In addition, the application of the notion
‘simultaneity’ to the present infinitive is problematic in its prolative uses. Such issues
are dealt with in the sections on the individual infinitives.
205
The only ‘future passive’ forms in the data base are restitutam rem publicam fore (Cic. Dom. 146),
which is properly speaking a future perfect, and (three) forms with iri in Caesar.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Cum per eorum fines triduo iter fecisset, inveniebat ex captivis Sabim flumen
a castris suis non amplius milibus passuum X abesse.
(‘After he had made three days’ march through their territories, he discovered from
some prisoners that the river Sambre was not more than ten miles from his camp.’
Caes. Gal. 2.16.1)
(b) Haec Casina huius reperietur filia esse ex proxumo . . .
(‘This Casina will be discovered to be the daughter of this man from next door . . . ’ Pl.
Cas. 1013)
(c) Patrem vita privare si per se scelus est, Saguntini, qui parentes suos liberos
emori quam servos vivere maluerunt, parricidae fuerunt.
(‘If to rob a father of life is in itself a crime, the people of Saguntum who chose that
their own parents should all die free men rather than live as slaves were guilty of
parricide.’ Cic. Parad. 24)
(d) Nihil erat latum de me. Non adesse eram iussus. Non citatus afueram.
(‘No law had been passed respecting me. I had not been ordered to appear in court;
I had not failed to answer any summons.’ Cic. Dom. 62)
Supplement:
Obliviscor enim Roscium et Cluvium viros esse primarios. Improbos temporis causa
esse fingo. (Cic. Q. Rosc. 50); Fac, quaeso, qui ego sum esse te. (Cic. Fam. 7.23.1);
Querebatur igitur se tum, cum illa videre coepisset, extingui. (Cic. Tusc. 3.69); Itaque
illi qui Graeciae formam rerum publicarum dederunt corpora iuvenum firmari
labore voluerunt. (Cic. Tusc. 2.36); Num igitur ignobilitas aut humilitas aut etiam
popularis offensio sapientem beatum esse prohibebit? (Cic. Tusc. 5.103);
However, the reference point of the present infinitive need not be that of the finite
verb of the clause in which it occurs. In (e) filium suom esse is definitely valid at the
time of speaking and not only at the time of dixit in the main clause. Note the perfect
subjunctive vicerit in the relative clause, which is also related to the time of speaking.
Note also the immediately following tuum esse . . . puerum. A similar example with a
future tense in the governing clause is (f ). The reference point of esse is the time of
speaking. There are also instances in which the present infinitive has a past reference
point in an accusative and infinitive clause that depends on a cognition verb in the
present tense (especially one of remembering). Examples are (g)–(i). Note in (h) the
imperfect subjunctive form requireret and in (i) the cum clause.
(e) Is (sc. Juppiter) se dixit cum Alcumena clam consuetum cubitibus, / eumque
filium suom esse qui illas anguis vicerit. / Alterum tuom esse dixit puerum.
(‘He said that he’s slept with Alcumena in secret, and that the son who’d crushed
those snakes was his; he said that the other was your boy.’ Pl. Am. 1122–4)
(f ) Magnum esse solem philosophus probabit, quantus sit mathematicus . . .
(‘The philosopher will demonstrate that the sun is a large body, while the astronomer
will compute just how large . . . ’ Sen. Ep. 88.27)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Non commeministi semul / te hodie mecum exire ex navi? (Pl. Men. 1074–5);
Memini et scio / et te me orare et mihi non esse quod darem. (Pl. Per. 118–19);
Meministin’ tibi me dudum dicere / eam veniisse militi Macedonio? (Pl. Ps.
1089–90); Memoria tenetis Artemidorum Aetnensem, legationis eius principem,
publice dicere Apronium venisse Aetnam cum Veneriis . . . (Cic. Ver. 3.105); Sed
ego idem . . . recordor longe omnibus unum me anteferre Demosthenem . . . (Cic.
Orat. 23); Nam a primo tempore aetatis iuri studere te memini, quom ipse etiam
ad Scaevolam ventitarem . . . (Cic. Leg. 1.13); Atque equidem memini (fama est
obscurior annis) / Auruncos ita ferre senes . . . (Verg. A. 7.205–6); Audire me memini
ex senioribus visum saepius inter manus Pisonis libellum . . . (Tac. Ann. 3.16.1);
This use of the present infinitive to refer to past events was already noted in Antiquity
(ordo est ‘memini vidisse’, dicimus autem et ‘memini videre’ (Serv. ad Verg. G.
4.127)). The explanation sometimes given for this is the original perfect character of
the form memini. This explanation, however, is not very helpful since at the time of
Plautus memini otherwise functions as a present form.
The present and perfect infinitives in accusative and infinitive clauses depending
on a verb of remembering are of course also used in their common way as simul-
taneous with or anterior to the reference point of the governing verb. Two examples
are (j) and (k), respectively.
The independence of the infinitive in the accusative and infinitive clause as far as the
reference point is concerned is the same as in finite subordinate clauses (see § 14.8).
For the use of the tenses in subordinate clauses depending on an accusative and
infinitive clause, see § 7.111.
There are two types of context in which the present infinitive cannot be interpreted as
denoting simultaneity. The first is in certain accusative and infinitive constructions,
notably those in which the infinitive depends upon a verb of promising or hoping
where a future infinitive would be possible as well. Secondly, the present infinitive is
used as a prolative infinitive with two-place modal and phasal auxiliary verbs such as
volo ‘to want’ and desino ‘to stop’ and with three-place manipulation verbs such
as impero ‘to order’.
In Early Latin (in Plautus more than in Terence) cognition and communication verbs
are regularly used with a present infinitive, when strictly speaking the state of affairs
referred to by the infinitive belongs to the future. The states of affairs involved are
almost always terminative, as opposed to accusative and infinitive expressions with a
future infinitive, which are mostly non-terminative.206 Cicero and Caesar use the
future infinitive for both terminative and non-terminative states of affairs, but poets
and other authors continue using the present infinitive, expanding the range of
governing verbs a bit. The verb spero ‘to hope’ is slightly different, because one may
hope that something has happened, is happening, or will happen. It is not restricted to
the future as some of the other verbs are. (For its occurrence with a prolative
infinitive, see § 15.122). A good illustration of the coexistence of the present and
future infinitive with one and the same verb is (a). Notice that, in the case of the
present infinitive dare, the agent of dare is not expressed and that it is identical to the
agent of the governing verb. This is often the case, certainly in Early Latin. One might
compare the use of the present infinitive in such cases with its prolative use with
modal and phasal verbs (see § 7.70). Notice in (b) the presence of cras ‘tomorrow’,
which is also quite common with the present indicative (see § 7.15). Passive present
infinitives are infrequent, as is to be expected, but later authors seem to prefer them to
the -um iri future passive infinitives.207 Relatively many instances can be found in
legal texts with verbs like iuro ‘to swear’, promitto ‘to promise’, and stipulor ‘to exact’.
In ecclesiastical authors it is also found with praedico ‘to predict’ and related verbs.
Tacitus uses the present passive infinitive regularly instead of the -um iri future
passive infinitive in other contexts as well.
206
See de Melo (2007d).
207
Bennett (1910: I.426–7) has a list of the governing verbs in Early Latin. Sz.: 357–8 gives the
following proportions of future infinitives and present infinitives used to refer to the future: Plautus
144 : 50; Terence 218 : 15.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Infinitives governed by modal and phasal verbs are normally present infinitives (the
so-called PROLATIVE use of the infinitive, see § 15.113), but perfect infinitives are used
occasionally. The present infinitive is shown in (a)–(c), the perfect in (d)–(e).
Whereas the perfect infinitive has its normal anterior meaning, it makes no sense
to apply the notion of simultaneity to the present infinitives.
(a) Qui potuit videre? # Oculis.
(‘How could he see her? # With his eyes.’ Pl. Mer. 183)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The use of the present infinitive as a main verb with an almost exclusively past
temporal value (the so-called HISTORIC INFINITIVE—infinitivus historicus) is found in
narrative texts from Plautus onward.208 That the historic infinitive functions as a past
tense appears from the use of past tense forms in subordinate clauses (see § 7.90 for
more details). Examples are (a)–(c).
(a) Obiurigare pater haec noctes et dies.
(‘My father scolded me about this night and day.’ Pl. Mer. 46)
(b) Ille contra haec omnia. / Ruri agere vitam.
(‘He’s been exactly the opposite. He’s lived in the country.’ Ter. Ad. 44–5)
(c) Interim cotidie Caesar Haeduos frumentum quod essent publice polliciti
flagitare.
(‘Meanwhile every day Caesar was asking the Aedui for the corn which they had
promised in the name of the state.’ Caes. Gal. 1.16.1)
208
See Viljamaa (1983), Hessen (1984), Orlandini (2002), Rosén (1995), Beltrán (1996), Sznajder
(1996), Schwartz (2002: 157–64), Adema (2007), and Jurado (2010). For the material and the develop-
ment, see Schlicher (1914a,b; 1915), summarized in Sz: 367–8. Quintilian deals with the historic infinitive
as a rhetorical figure and is the first to state that the phasal verb coepit must be understood (Inst. 9.3.58).
Beltrán (1994) discusses his and later views.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Sallust and Virgil). It was avoided by the historian Ammianus Marcellinus,209 but
Claudianus and Sidonius used it freely.210
Historic infinitives are often found in clusters. The subject of the infinitive is
usually already known from the preceding context. Almost all instances of the historic
infinitive are found in declarative sentences (see for an exceptional—rhetorical—
question (d); for its rare use in subordinate clauses, see § 7.122). The subject is almost
always a human agent. Passive infinitives are rare. The verbs are most often dynamic
verbs, sometimes governed by a modal or phasal verb, as in (e). They often describe
the start of a new (series of) event(s), indicated by tum (vero) ‘then (indeed)’ or
hinc ‘next’.
(d) Qui mori timore nisi ego?
(‘Who was dying from fear but me?’ Petr. 62.9 (Niceros speaking))
(e) Verum ingenium eius (sc. Semproniae) haud absurdum: posse versus facere,
iocum movere, sermone uti vel modesto vel molli vel procaci.
(‘Nevertheless, she was a woman of no mean endowments; she could write verses,
bandy jests, and use language which was modest, or tender, or wanton.’ Sal. Cat. 25.5)
There is much discussion in the literature about the position of the historic infinitive
among the tenses that are used in narrative texts. It is found parallel with sentences
containing an imperfect, as in (f ). However, it is also found parallel with sentences
containing a (historic) present, as in (g). In general, it occurs both in contexts where a
perfect (and historic present) might be used and in contexts where an imperfect might
be used. It more often resembles an imperfect, which one would expect on the basis of
the fact that it normally expresses simultaneity; however, unlike a series of imperfects,
a series of historic infinitives often denotes successive stages in the narrative. Fur-
thermore, the historic infinitive is also distinct from the historic present in that the
former is found in authorial statements in a past narrative. An example is (h). It is
often described as more ‘emotional’, ‘intense’, or ‘expressive’ than the other narrative
forms, but there is not sufficient evidence for that.211 Its precise interpretation
(‘imperfect-like’ or ‘perfect-like’) depends to a large extent on the lexical meanings
of the verb and the subject.
(f ) (Antipho) Noster quid ageret nescire. Et illam ducere / cupiebat et metuebat
absentem patrem.
(‘Our Antipho didn’t know what to do; on the one hand he was eager to marry the
girl, on the other he was afraid of his absent father.’ Ter. Ph. 117–18)
209
See den Boeft (1992: 16–17) for the dubious instances in Ammianus and the suggestion that
Ammianus avoids the expression because it is unknown in Greek. See also Kroon and Rose (1996), with
references.
210
For Sidonius, see Köhler (1999).
211
Dressler (1965, 1968) holds that the historic infinitive has imperfective aspect. He and Orlandini
(2002, with further references) argue for the emotional, etc. aspect of the historic infinitive. For counter-
arguments, see especially Rosén (1995); also Oldsjö (2001: 378–91) (on Caesar).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(g) Interea dum sedemus illi, intervenit / adulescens quidam lacrumans. Nos
mirarier. Rogamus quid sit.
(‘Meanwhile, while we were sitting there, a young man came up in tears. We
wondered at this and asked him what the matter was.’ Ter. Ph. 91–3);
(h) His freti occursant portis, ingerunt probra, aegre abstinent quin castra
oppugnent. Enimvero non ultra contumeliam pati Romanus posse. Totis
castris undique ad consules curritur.
(‘Relying on these convictions, they charged up to the gates, flung gibes at their
defenders, and scarcely refrained from assaulting the camp. At this, the Romans
could no longer brook their insults; from all over the camp, they came running to the
consuls.’ Liv. 2.45.10–11)
Supplement:
Imperator utrimque, hinc et illinc, Iovi / vota suscipere, <utrimque> hortari exerci-
tum. (Pl. Am. 229–30); Obiurigare pater haec noctes et dies, / perfidiam, iniustitiam
lenonum expromere. / Lacerari valide suam rem, illius augerier. / Summo haec
clamore. Interdum mussans conloqui. / Abnuere, negitare adeo me natum suom. /
Conclamitare tota urbe et praedicere, / omnes tenerent mutuitanti credere. (Pl. Mer.
46–52); Ruri agere vitam; semper parce ac duriter / se habere; uxorem duxit. (Ter. Ad.
43–6); Sic vita erat: facile omnes perferre ac pati. / Cum quibus erat quomque una, eis
sese dedere, / . . . eorum studiis obsequi. (Ter. An. 62–4); Venit Chremes postridie ad
me clamitans: / indignum facinus; comperisse Pamphilum / pro uxore habere hanc
peregrinam. Ego illud sedulo / negare factum. Ille instat factum. (Ter. An. 144–7); Ille
suam egit semper vitam in otio, in conviviis, / clemens, placidus, nulli laedere os,
adridere omnibus. (Ter. Ad. 863–4); Magnas vero agere gratias Thais mihi? (Ter. Eu.
391);212 Tum vero iste clamare voce ista quae perfacile cuivis rubores eicere potest.
(Rhet. Her. 4.14—NB: in the illustration of the simple style); Hic ego postulare coepi
ut mihi tabulas obsignare ac deportare liceret. Ille contra dicere, negare esse illud
senatus consultum in quo praetor appellatus esset, negare id mihi tradi oportere. (Cic.
Ver. 4.149); Tum ille fidenter, homo peritissimus, confirmare ita se rem habere ut
respondisset nec dubium esse posse. (Cic. de Orat. 1.240); Brevi spatio interiecto vix
ut iis rebus quas constituissent conlocandis atque administrandis tempus daretur,
hostes ex omnibus partibus signo dato decurrere, lapides gaesaque in vallum con-
icere. (Caes. Gal. 3.4.1); Ita partim precibus, partim vi bene magna comparata manu
provinciam vastare. (B. Hisp. 1.2); Interea Catilina cum expeditis in prima acie
vorsari, laborantibus succurrere, integros pro sauciis arcessere, omnia providere,
multum ipse pugnare, saepe hostem ferire. Strenui militis et boni imperatoris officia
simul exequebatur. (Sal. Cat. 60.4); Hinc mihi prima mali labes, hinc semper Ulixes /
criminibus terrere novis, hinc spargere voces / in vulgum ambiguas et quaerere
conscius arma. (Verg. A. 2.97–9); Male salsus / ridens dissimulare, meum iecur urere
bilis. (Hor. S. 1.9.65–6); Interim legati alia moliri, aperte bona repetentes clam
212
On the use of the historic infinitive in this interrogative context, see Calboli (1991).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The infinitive is also found (much less frequently) as a main verb in imperative
sentences (the so-called infinitivus pro imperativo), where the infinitive necessarily
refers to the future. Examples are discussed in § 6.31.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Appendix: There are a few attestations of the use of the present infinitive instead of a
finite verb form in an indirect question or in an autonomous relative clause. One
example is (i).213
213
For discussion, see Adams (2013: 770–2).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Ilion igni. (Ov. Ep. 16.49); Nam quemadmodum scio omnia accidere posse, sic scio et
non utique casura. (Sen. Ep. 88.17); . . . rumor incesserat pontem cessurum oneri dolo
fabricantium. (Tac. Ann. 15.15.3);
The irregular future infinitive forms include fore (of the verb sum) and perhaps the
rare sigmatic infinitives of the type impetrassere, which may, however, just as well be
taken as present infinitive forms.214
The form esse is absent in eleven out of twelve instances in Pl. Am.; in Cic. Dom. in ten
instances out of fifteen. Caesar has 150 without, against fifteen with esse, according to
Dittenberger (1869). Westman (1983: 163–4) gives the following statistics for the use of
the future infinitive of sum in accusative and infinitive clauses in Seneca the Younger:
with esse one, without esse forty-two (fore six). For future infinitives of other verbs in
this author the figures are 7 : 117. There are four instances of -urus + fuisse.
Marouzeau (1910a: 189–92) suggests that there is a semantic difference between the
infinitive without esse (‘future infinitive’) and with esse (‘prospective infinitive’).
However, Garuti (1954: 60–6) shows that both forms are used for the future, although
certain instances can be interpreted as prospective; additionally, Westman (1961: 88)
finds no confirmation for a difference in meaning between the two forms in Seneca.
For the future passive infinitive Latin grammarians cite the periphrastic expression
consisting of the supine and the passive infinitive iri as in (e)–(g). Except for in Cicero
(fifty-eight instances) and, given the size of his body of work, in Terence (seven instances),
this expression is very infrequent: there are fewer than 200 instances in the BTL and
CLCLT-I corpora altogether (disregarding grammarians). Indeed, it is completely absent
from Augustan and post-Augustan poetry and left no traces in the Romance languages.
(e) . . . nisi se sciat vilico non datum iri.
(‘ . . . only . . . if she knows she won’t be given to the overseer.’ Pl. Cas. 699)
(f ) . . . absolutum iri Sopatrum videbat.
(‘ . . . he saw that . . . Sopater would be acquitted.’ Cic. Ver. 2.74)
(g) Quem defendis sperat se absolutum iri.
(‘The man whom you defend hopes to be acquitted.’ Cic. Sul. 21)
Supplement:
Scilicet / datum iri. (Ter. Hau. 856–7); Sciebas tibi crimini datum iri? (Cic. Ver.
5.74); Ipsi vero nihil nocitum iri inque eam rem se suam fidem interponere. (Caes.
Gal. 5.36.2); Puto Id. Sext. de ea re actum iri. (Cael. Fam. 8.4.4); Hac habita oratione
Scipio cum existimasset pro suo beneficio sine dubio ab his gratias sibi actum iri,
potestatem iis dicundi fecit. (B. Afr. 45.1); . . . te mihi mater, veridica interpres deum,
aucturum caelestium numerum cecinit tibique aram hic dicatum iri . . . (Liv. 1.7.10);
An iustitiam auctum iri et decurias equitum egregium iudicandi munus <melius>
expleturos, si fractos sonos et dulcedinem vocum perite audissent? (Tac. Ann.
14.20.5); Nam et iuniorem incoram sui funerari videbat et alterum ob incestum
214
See de Melo (2007a: 236–8).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
parricidiumque capitis scilicet damnatum iri certo sciebat. (Apul. Met. 10.5.6); Porro
qui <ab> his ait devoratum iri mortale, ab aliis negavit. (Tert. Res. 42.6); Inde est,
quod deorum multorum falsorumque simulacra et praedixit eversum iri et praecepit
everti. (August. Ep. 91.3);
To refer to the future, the present infinitive is much more frequent, as in (h) (see also
above § 7.69). As its alternatives the expressions with futurum esse ut and fore ut are
also available, as in (i). These expressions are not restricted to verbs that lack a supine
form, and were slightly more popular in most authors (see below § 7.73). From
Tertullian onwards the gerundive in -ndus + esse comes into use as a future passive
infinitive, as in (j) (see also § 7.74). A further Late Latin alternative for the use of the
future passive infinitive in the accusative and infinitive construction is the use of
subordinate clauses introduced by quod or quia with a finite future passive indicative
form, as in (k) (see also § 15.9 and § 15.19). In addition, by that time the expression
with habere was available (see above § 7.27).215
(h) . . . magnitudine poenae reliquos terreri sperans.
(‘ . . . hoping for the rest to be deterred by the greatness of the punishment.’ Caes.
Civ. 3.8.3)
(i) Sperant fore ut patris litteris nuntiisque filius ab illo furore revocetur.
(‘They expect that the father’s letters and messengers would make the son give up his
insane purpose.’ Cic. Ver. 2.97)
(j) . . . spero utique aliud esse inveniendum . . .
(‘ . . . I hope that in any case something else will be found . . . ’ Tert. Praescr. 11.2)
(k) . . . contemplans quod alter . . . forsitan libens declarabitur princeps . . .
(‘ . . . thinking that another would perhaps willingly be proclaimed emperor . . . ’
Amm. 20.8.10)
For verbs lacking a supine form the supine + iri expression naturally was not
available. Apart from expressions with the infinitive of a modal verb (possum ‘can’,
volo ‘want’, debeo ‘must’), which by implication refer to the future, the expressions
fore ut and—less frequently—futurum (esse) ut + present or imperfect subjunctive
served as a substitute. Examples are (a)–(c). The same expression came into use with
verbs which had a supine form, as is illustrated by (d) with passive revocetur and (e)
with active deflagret in the ut clauses. The same is shown by examples (f )–(g) with
futurum. A related expression is shown in (h), with fore aliquem + qui. Originally, in
215
For the (pre)history of the supine + iri expression, see Pinkster (1985b, 1989). For statistical data
and the development, see Baños (1996a,b). He also draws attention to the relatively frequent use of the
auxiliary possum ‘can’ + present passive infinitive (1999: 64–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
accordance with its semantic structure, fore/futurum ut was used, in contrast to the
future active infinitive, to present a state of affairs as happening unintentionally, as
being beyond the control of the subject of the governing verb, and as someone’s
opinion or assumption; however, after Livy its use became less constrained.216
(a) Hoc ego numquam ratus sum / fore me ut tibi fierem supplex.
(‘I never thought it would come about that I’d be your suppliant.’ Pl. Ps. 1318–19)
(b) Spero fore ut contingat id nobis.
(‘I hope that this will be our lot.’ Cic. Tusc. 1.82)
(c) Deinde fore ut angeretur cum a fratre familiaritate et omni gratia vinceretur.
(‘Thereupon it would be the case that he would be mortified whenever he
was outdone by his cousin in friendly contacts and general considerations.’
Cic. Att. 12.7.1)
(d) Sperant fore ut patris litteris nuntiisque filius ab illo furore revocetur.
(‘They expect that the father’s letters and messengers would make the son give up his
insane purpose.’ Cic. Ver. 2.97)
(e) Fore tamen aliquando ut omnis hic mundus ardore deflagret.
(‘But that nevertheless a time will come when all this world will be burnt out with
heat.’ Cic. Luc. 119)
(f ) Futurum esse, nisi provisum esset, ut Roma caperetur.
(‘That unless this was seen to, Rome would be captured.’ Cic. Div. 1.101—tr. Wardle)
(g) Futurum esse paucis annis uti omnes ex Galliae finibus pellerentur atque
omnes Germani Rhenum transirent.
(‘In a few years all the natives will have been driven forth from the borders of Gaul,
and all the Germans will have crossed the Rhine.’ Caes. Gal. 1.31.11)
(h) . . . non despero fore aliquem aliquando, qui . . . existat talis orator, qualem
quaerimus . . .
(‘ . . . I do not despair that some day someone . . . will stand forth as an orator such as
we are seeking . . . ’ Cic. de Orat. 1.95)
Supplement:
Quod ego numquam credidi / fore, ut ille hac viva posset animum inducere / uxorem
habere. (Ter. Hec. 98–100); . . . fore putabam ut exemplum rei publicae conservandae
mecum simul interiret. (Cic. Sest. 49); Ei visum in quiete egregia facie iuvenem dicere
fore ut perbrevi convalesceret . . . (Cic. Div. 1.53); . . . puto fore ut me praesentem non
sustineat. (Cic. Att. 14.18.1); Etsi iam sperabam, cum has litteras accepisses, fore ut
ea quae superioribus litteris a te petissemus impetrata essent tamen . . . (Cic. Att.
216
For syntactic, semantic, and pragmatic constraints on the fore/futurum ut-expression and the
diachronic development, see Stephens (1989, 1990 (summarizing the main points of 1989 and 1991),
1991). The data can be found in Sjöstrand (1892).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
16.16E.2); Hoc per aliquot dies fecerunt sperantes fore ut Romani equites abeuntium
novissimum agmen adgrederentur. Inde certamine orto . . . se, qui equitatu et levi
armatura plus possent, conversuros aciem. (Liv. 42.57.11);
Simul illud quod indignissimum est futurum arbitrati sunt, ut in hac causa non de
inprobitate Sex. Aebuti, sed de iure civili iudicium fieri videretur. (Cic. Caec. 4);
Vivus enim sibi cum proponit quisque futurum, / corpus uti volucres lacerent in
morte feraeque . . . (Lucr. 3.879–80); . . . id quod evenit futurum, credo, etiam rati, ut
mox opes eorum qui praeessent ipsi honori ius maiestatemque adicerent. (Liv. 4.8.5);
Excepta vox est, cum teporem incusaret, statim futurum ut in<c>alescerent.
(Tac. Hist. 3.32.3);
Most authors agree that the use of the gerundive with the infinitive esse without a
deontic meaning (on which see § 5.41) as a substitute for the future passive infinitive
starts in Tertullian (see ex. c), although Livy’s uncommon use of sistendam in (a)
looks quite similar and Tacitus’ (b) has also been suggested.217
(a) . . . puellam sistendam in adventum eius . . . promittat.
(‘ . . . he should promise that the girl will be produced upon his arrival.’ Liv. 3.45.3)
(b) (Tiberius) nuntianti centurioni, ut mos militiae, factum esse quod impe-
rasset, neque imperasse sese et rationem facti reddendam apud senatum
respondit.
(‘When the centurion reported, according to military custom, that his instructions
had been carried out, the emperor replied that he had given no instructions and the
deed would have to be accounted for in the senate.’ Tac. Ann. 1.6.3)
(c) . . . nec intellexissent Christum in tempore suae visitationis inveniendum.
(‘ . . . and they had not understood that Christ would be found in the time of his visit
to them.’ Tert. Iud. 13.26)
Supplement:
Ideo si navem a se fabricandam quis promiserit . . . (Ulp. dig. 46.3.31); In quo
magistratu ut a Sura conperit adoptandum se a Traiano esse, ab amicis Traiani
contemni desiit ac neglegi. (Hist. Aug. Hadr. 3.10); Perpendebat enim ad relationem
suam, quam olim putabatur misisse, abstrahendos e Galliarum defensione pugnaces
numeros barbarisque iam formidatos. (Amm. 20.4.7); Quosdam non esse morituros
et de praesenti vita rapiendos in futuram . . . (Hier. Ep. 119.4); . . . exaudiebas et
faciebas ordine quo praedestinaveras esse faciendum. (August. Conf. 5.9.17); . . .
promittens multis eum divitiis cumulandum . . . (Vict. Vit. 1.47);
217
See Vidén (1984: 13), who compares redditurum in Suet. Tib. 22.1.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Just like finite forms of the perfect stem, with terminative states of affairs the perfect
infinitive may denote a state resulting from a preceding action or process. This is
quite common in the passive. Two examples to illustrate this are (e) and (f ). In order
to express the anteriority of such a resulting state the perfect infinitive fuisse is used,
as in (g).
(e) Melius fuit perisse . . . quam haec videre.
(‘It would have been better to be dead than to see these things.’ Cic. Att. 14.10.2)
(f ) ‘Equidem’ inquit ‘maluissem vos togatos esse. <Si> pigitum est, cinctos
saltem esse et paenulatos’.
(‘“For my part,” he said, “I should have preferred for you to be wearing togas. If that was
too much trouble, at least girded and wearing a cloak.”’ Castricius ap. Gel. 13.22.1)
(g) (sc. duumvir) . . . <EI EUM / AGRUM LO>CUM NEIVE EMPTUM NEIVE ADSIGNATUM ESSE
NEIVE FUISSE IUDICATO.
(‘ . . . the member of the Board of Two shall deliver judgement to the effect that the
said land or ground is neither under the ownership nor the assignation of the said
person nor has it been.’ CIL I2.585.56–7 (Lex Agr., c.111 BC))
218
I follow Rizzo’s edition (1991). See her comments ad loc.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Quod quidem nuper in columna ahenea meminimus post rostra incisum et per-
scriptum fuisse. (Cic. Balb. 53); . . . dico eius adventu maximas Mithridati copias
omnibus rebus ornatas atque instructas fuisse, urbemque Asiae clarissimam nobis-
que amicissimam Cyzicenorum oppressam esse ab ipso rege . . . (Cic. Man. 20);
Sopitum fuisse regem subito ictu; ferrum haud alte in corpus descendisse; iam ad
se redisse; inspectum volnus absterso cruore . . . (Liv. 1.41.5);219
However, there are certain contexts in which the perfect infinitive is or is seemingly
used without an anterior value. This is the case from Early Latin onwards until the
Augustan period in prohibitions of the type ne quis fecisse velit. Two examples of this
use are (h) and (i). These expressions are characteristically found in decrees, especially
those issued by magistrates; they denote a severe threat, meaning ‘No one should wish
to be found guilty of, to be sentenced for, this crime’.220 The perfect infinitive is
entirely in place here and still denotes anteriority: the expression refers to (deliberate,
controllable) actions and threatens that ‘the result [of these actions] is going to be
undesirable’. It is sometimes found outside of decrees, either in imitation of the value
just described or because it had become idiomatic. The original explanation of the use
of the perfect is similar to the one provided for the use of the perfect subjunctive in
prohibitions (see § 7.53). In both cases the use of the perfect became idiomatic. Note
that there are no positive counterparts with a perfect.221
(h) NEIQUIS·EORUM·SACANAL·HABUISE·VELET.
(‘Let none of them wish to have a Bacchanal.’ CIL I2.581.3 (SCBac., Tiriolo, 186 BC)—
NB: error S instead of B)
(i) Nolito edepol devellisse.
(‘Don’t pluck them!’ Pl. Poen. 872)
Supplement:
BACAS·VIR·NEQUIS·ADIESE·VELET·CEIVIS·ROMANUS . . . (CIL I2.581.7 (SCBac., Tiriolo, 186 BC));
. . . NEI QUIS INTRA / TERMINOS PROPIUS / URBEM USTRINAM / FECISSE VELIT NEIVE / STERCUS
CADAVER / INIECISSE VELIT. (CIL VI.31515.4–7 (Rome, 89 BC));
Outside decrees: (sc. vilicus) Chaldaeum ne quem consuluisse velit. (Cato Agr. 5.4);
(sc. vilicus) Ne quid emisse velit insciente domino, neu quid dominum celavisse velit.
(Cato Agr. 5.4); Interdico ne extulisse extra aedis puerum usquam velis (Ter. Hec.
563—NB: second person); Ne quis, qui Bacchis initiatus esset, coisse aut convenisse
sacrorum causa velit . . . (Liv. 39.14.8); Ne quis humasse velit Aiacem, Atrida, vetas.
(Hor. S. 2.3.187);
219
For a few more instances, see Neue-W.: III.151–2.
220
This paraphrase is taken from Daube (1956: 40), whose chapter 3 is very relevant to this section in
general. The explanation is the same as in K.-St.: I.133 and in Barbelenet (1913: 21). Similarly Becker
(2005: 21–2). For the development of the expression, see Bortolussi (2002).
221
See Daube (1956: 44–5).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
A similar parallelism between the use of the perfect infinitive depending on a modal
verb in the subjunctive and the perfect subjunctive can be seen in the expression
dixisse liceat, as in (k). Compare the expression pace tua dixerim in § 7.57.
(k) Pace, novem, vestra liceat dixisse, sorores.
(‘With due respect to you, nine sisters, would I say it.’ Ov. Tr. 5.12.45)
Supplement:
Sed, pace Theophrasti dixisse liceat, non est hoc simile veri . . . (Sen. Nat. 3.11.4); Sed
et de Tigri ipso dixisse conveniat. (Plin. Nat. 6.127); Atque haec omnia medici—
quod pace eorum dixisse liceat—ignorant. (Plin. Nat. 34.108);
Outside of prohibitions and in non-legal contexts, the non-anterior use of the perfect
infinitive depending on forms of volo ‘to want’ and nolo ‘not to want’ starts with the
poets Lucretius, as in (l), and Catullus as a metrical commodity223 and possibly
also in imitation of Greek poets.224 It became especially popular from the Augustan
period onwards, where its use expands to other governing modal verbs, as in (m)
(note the parallelism with a present infinitive) and other expressions (n), and is no
longer restricted to poetry, as in (o) and (p). It is even found in combination with
adjectives, as in (q).
(l) Unde homines dum se . . . / effugisse volunt longe longeque remosse . . .
(‘While men desire to flee far from them and to remove themselves far away . . . ’
Lucr. 3.68–9)
(m) Tum certare odiis, tum res rapuisse licebit.
(‘Then shall it be lawful to vie in hate, then to ravage.’ Verg. A. 10.14)
(n) . . . vesanum tetigisse timent fugiuntque poetam / qui sapiunt.
(‘ . . . so men who are wise fear to touch a crazy poet and run away.’ Hor. Ars 455–6)
222
See Adams (2003b: 545–6) and Halla-aho (2010: 182) on this instance.
223
See Howard (1890: 120–2).
224
See Coleman (1999: 83–4). For possible Greek influence on the spread of the perfect infinitive in
poetry, see Calboli (2009: 134–6).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
A final group of instances of the perfect infinitive without an anterior meaning are
explained as cases of ‘attraction’, with the infinitive copying the past value of the
governing verb. They are said to occur with past forms of expressions like oportet ‘it is
proper’, convenit ‘it befits’, melius est ‘it is better’ from Plautus onwards, but the forms
found in Plautus are of the type exemplified above by examples (a)–(d), that is they
have an anterior meaning. However, more convincing instances are cited from
Tertullian and others.226 Examples are (r)–(t).
(r) Tunc flesse decuit cum adempta nobis sunt arma . . .
(‘It was right to weep then, when our arms were taken from us . . . ’ Liv. 30.44.7)
225
For Tibullus’ use of the perfect infinitive, see Perotti (1989).
226
See Hoppe (1903: 53ff. (=1985: 106–9)), whose examples need closer examination. For the reading
of (s), see Hoppe (1932: 40).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Participles mark the relative position of a state of affairs in relation to some reference
point, usually that of a finite verb form in the sentence in which they occur. They
present this state of affairs as simultaneous (present participle), anterior (perfect
participle), or posterior (future participle) with respect to that reference point. The
perfect (passive or deponent) participle in a terminative state of affairs may refer to a
state resulting from a preceding action or process, which is then simultaneous with
the participle’s reference point. The semantic values of the three types of participle are
illustrated by (a) and (b).
(a) (‘considering also that men like you two are going to meet him’) . . . quid
agenti, quid acturo?
(‘ . . . doing what and about to do what?’ Cic. Att. 8.9.2)
(b) Ignoratur enim quae sit natura animai, / nata sit an contra nascentibus
insinuetur . . .
(‘For they know not what is the nature of the soul, whether it is born or on the
contrary finds its way into men at their birth . . . ’ Lucr. 1.112–13)
As is shown in Figure 7.4, the frequency of each of the participles varies from author
to author and from text type to text type. The future participle is relatively uncom-
mon, except for in Seneca’s treatises.227
The system of participles in Latin is deficient in that there are no pairs of participles
for the active and the passive (see § 3.19). It is shown below that this especially
influenced the use of the present participle as a secondary predicate (on which see
Chapter 20).
227
Source: LASLA website (17 November 2004). Some perfect forms may be part of periphrastic
forms. The overall picture remains the same.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
100
80
60
40
20
0
Pl. Am. Cic. Dom. Caes. Gal. Sen. Clem. Verg. A.
228
For the equivalence of the participial expression and the temporal cum clause, see Lavency (2003a)
and already Hale (1891: 223–9).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
This is the normal use of the present participle. As with the finite present tense, it is
important not to conceive of ‘simultaneity’ in a strict chronological sense. By doing so
Latinists have sometimes described the present participles in (g)–(i) as anterior, but
this is not necessary. The cum-paraphrase mentioned above fully applies to these
examples as well.
(g) Aio, adveniensque ilico / me salutavisti . . .
(‘I do, and on your arrival you immediately greeted me . . . ’ Pl. Am. 799–800)
(h) . . . quo accusante P. Sextius, praetor designatus, damnatus est ambitus.
(‘It was through his prosecution that the praetor elect Publius Sextius was convicted
of bribery.’ Cic. Brut. 180)
(i) . . . Tullum Hostilium populus regem interrege rogante comitiis curiatis
creavit . . .
(‘ . . . the people in their curiate assembly presided over by an interrex chose Tullius
Hostilius as their king . . . ’ Cic. Rep. 2.31)
From Augustan times onward more remarkable instances of the anterior use of the
present participle are found, such as (j) and (k), where the participle is followed by
another subordinate clause referring to a state of affairs that comes between those
referred to by the present participle and the main verb (the ut clause and ablative
absolute clause, respectively). This becomes quite common in Late Latin authors,
where the present participle comes to function as a substitute for the non-existent
perfect active participle. This also occurs in translations of Greek.229 Note that the
229
Some Latinists refer to this use of the present participle as ‘perfective’ (Adams 1976: 60–2) or
‘aoristic’ (Sz.: 386). References in Sz.: 387, Adams loc. cit., and Pitkäranta (1978: 78–9). In Late Latin
authors the order is often Participle – Subject – Main Verb (Adams loc. cit.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
order of the participial clause and the main verb is usually in accordance with the
course of events (the iconic order).
(j) At pius Aeneas, per noctem plurima volvens, / ut primum lux alma data’st,
exire . . . constituit . . .
(‘But loyal Aeneas, through the night revolving many a care, so soon as kindly light
was given, determines to issue forth . . . ’ Verg. A. 1.305–9)
(k) Sostratus . . . pauca . . . respondens petito secreto futura aperit.
(‘Sostratus made a brief reply, but asked for a private interview in which he disclosed
the future.’ Tac. Hist. 2.4.2)
Supplement:
Hic situs est vitae iampridem lumina linquens, / qui quondam Hectoreo perculsus
concidit ense. (Cic. Fragm. 25.1–2 B. (= Gel. 15.6.3, translates ŒÆÆŁ ÅH, a
perfect participle, in Hom. Il. 7.89–90)); Haec Maurus secum ipse diu volvens
tandem promisit. (Sal. Jug. 113.1); Eum primo incertis implicantes responsis, ut
metus tormentorum admotus fateri vera coegit, edocuerunt litteras se ab Hasdrubale
ad Hannibalem ferre. (Liv. 27.43.3); Apud quos praemonente Narcisso pauca
verba fecit. (Tac. Ann. 11.35.2); . . . Symplegades . . . cedentes . . . retrorsus acri assultu
ad ea reverti quae pulsarant. (Amm. 22.8.14); Unde denuo proficiscens pervenimus
in nomine Christi nostri Edessam. (Pereg. 19.2); Incidens difficiliter resurgit . . .
(Mulom. Chir. 384); . . . Nepos . . . ascendens navem fugam petit . . . (Anon. Vales.
7.36);
Grammars also draw attention to instances of the present participle that have to be
interpreted as posterior to the reference point of the main verb. This occurs especially
with movement verbs, with the participial clause referring to the purpose of the event
of the main verb. Livy has quite a few examples, but it occurs in later texts as well. An
example is (l).230
(l) . . . legati ab Ardea veniunt . . . auxilium prope eversae urbi implorantes.
(‘ . . . there came envoys from Ardea begging to send help to their city which was on
the brink of ruin.’ Liv. 4.9.1)
Supplement:
Quibus territi malis Colophonii oratores Samum ad L. Aemilium fidem praetoris
populique Romani implorantes miserunt. (Liv. 37.26.9); . . . legati a Saguntinis
Romam missi auxilium ad bellum iam haud dubie imminens orantes. (Liv. 21.6.2);
From Plautus onwards there are found in small numbers expressions consisting
of a form of the verb sum and the present participle that cannot be described as a
combination of the copula sum and an adjective-like participle functioning as a
230
See Adams (1976: 60–1). The participle often follows the main verb.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In translations of the Bible, especially in the Vetus Latina versions, quite a few
instances can be found. They correspond to similar periphrastic forms in the Koine
original of the New Testament and serve the same aspectual (imperfective) function.
An example is (d). Instances are also found in ecclesiastical authors, as in (e).232 In
other Latin texts they become more frequent from Gregory of Tours onwards,
without continuation in the Romance languages.233 For a similar use of the verb sto
+ participle see (f ).234
(d) Erat autem docens in synagoga eorum sabbatis.
(‘He was teaching, moreover, in their synagogue on the Sabbaths.’ Vulg. Luc. 13.10—
˙ b ØŒø )
231
See Hoffmann (1997a: 238–46), Arias (1999), and Amenta (2003).
232
The most complete survey of the periphrastic use of sum + present participles in these texts can be
found in Dietrich (1973).
233
See Blase (1903: 256–7); Eklund (1970); Hoffmann (1997a); Haverling (2005).
234
See Haverling (2010a: 496–7).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Further examples of the future participle can be found in Chapter 21. For the Late
Latin development of the gerundive into a future passive participle, see § 7.84.
235 236
See Eklund (1970) and Arias (1999). See Vidén (1984: 12ff.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The perfect participle of a (semi)deponent verb locates a state of affairs before the
reference point of the main verb of its clause, as hortatus and orsus, both secondary
predicates, do in (h) and (i), respectively. As with perfect passive participles, the
perfect participle of a (semi)deponent verb used in a terminative state of affairs may
be interpreted as referring to a state resulting from a previous (action) process, so that
it is simultaneous with the main verb in its clause. An example is amplexae in (j)
which can be compared to the present participle amplexantes in (k). However, the
non-anterior use of the perfect participle is also found in non-terminative states of
affairs, as exemplified by (l) and (m). Remarkable is the non-anterior use of the
autocausative perfect passive participle tunsae in (n).
Supplement:
. . . Caesar isdem ducibus usus . . . Numidas . . . sagittarios . . . subsidio oppidanis mit-
tit. (Caes. Gal. 2.7.1); . . . quos simul a Troia ventosa per aequora vectos / obruit
auster . . . (Verg. A. 6.335–6); Secutis omnibus imi / convivae lecti nihilum nocuere
lagoenis. (Hor. S. 2.8.40–1); . . . avertit ab consciis in insontes indicium, Thrasonem
esse auctorem consilii mentitus . . . (Liv. 24.5.11);
Autocausative: . . . e speculis percussae pectora matres / femineum clamorem ad caeli
sidera tollunt. (Verg. A. 11.877–8);
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
A number of the perfect participle forms that are regularly used in a non-anterior way
are given in Table 7.15. Most of the verbs involved are deponent.
237
For the gradual replacement of the present participle by the gerund, see Chapter 20.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In later Latin the gerundive develops into a passive participle.238 Early instances are
found in Tertullian, as in (i) and (j). Among non-ecclesiastical authors Ammianus has
quite a few instances, such as (k). Note the coordination with a future active participle
in (i) and (k). This development of the gerundive is also apparent in its occurrence
with finite and non-finite forms of the verb esse, for which see § 7.25 and § 7.74.
(i) . . . (Iesus) talis etiam descensurus, ut angeli adfirmant, agnoscendus scilicet
et eis qui illum convulneraverunt.
(‘Jesus shall descend in like manner, as the angels affirmed, so as even to be
recognized by those who pierced him.’ Tert. Res. 51)
(j) Qua poena timorem puniat, nisi quam negator relaturus est cum corpore et
anima occidendus in gehenna?
(‘What penalty will he appoint as the punishment of fear but that which the apostate
will pay who is going to die with body and soul in hell?’ Tert. Scorp. 12.5)
(k) . . . quam profligato morem gererent nebuloni destituendo iam et casuro.
(‘rather than to gratify a profligate wretch who was already on the point of being
deserted and was approaching his fall.’ Amm. 26.9.5)
Supplement:
Res monuit super hoc eodem Eutherio pauca subserere forsitan non credenda ea re,
quod, si Numa Pompilius vel Socrates bona quaedam dicerent de spadone dictisque
238
The fullest discussion is in Odelstierna (1926). The instances in Ammianus can be found in Langen
(1870: 482–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The ‘relative’ strategy is illustrated below with two types of subordinate clause, those
depending on the verb scio ‘to know’ and those depending on the verb impero ‘to
order’. Table 7.16 shows a number of indirect questions depending on a clause with
present tense scio and imperfect tense sciebam.241 Examples illustrating the simplified
examples in Table 7.16 are given in (b)–(g).
239
See de la Villa (1999a) for a comparative approach to Latin tense usage in complex sentences.
240
This is rightly stressed by Moralejo (1998, 1999).
241
The table follows K.-St.’s illustration of the sequence of tenses (II.175).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Table 7.16 Tenses in indirect questions depending on the verb scio ‘to know’
Tense in the Tense in the subordinate clause
main clause
Simultaneous Anterior Posterior
present present perfect present periphrastic future
scio quid agas quid egeris quid acturus sis
‘I know’ ‘what you are doing’ ‘what you did’ ‘what you are planning to do’
imperfect imperfect pluperfect past periphrastic future
sciebam quid ageres quid egisses quid acturus esses
‘I knew’ ‘what you were doing’ ‘what you had done’ ‘what you were planning to do’
It should be noted that the fact that the periphrastic future can be used instead of a
non-existent simple future subjunctive to indicate posteriority does not necessarily
imply that it lacks its usual prospective value (immediacy, fate, or intention—see
§ 7.24). This option is not available at all for the passive: the Latin verbal paradigm
lacks a passive future subjunctive expression. Latin authors might have used the
expression futurum sit (esset) ut + a passive present or imperfect subjunctive form,
but they did not.242 Present and imperfect subjunctive forms are used instead.
(b) Non edepol nunc ubi terrarum sim scio, si quis roget . . .
(‘If anyone were to ask, I don’t know where on earth I am . . . ’ Pl. Am. 336)
242
Discussion and data in Baños (1999: 73–80, 58–66).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The alternative strategy of selecting a reference point independently from the tense in
the main clause (or: absolute use) would result in another set of combinations than
the ones given in Table 7.16 and would produce combinations like scio quid ageres
‘I know what you were doing’, sciebam quid agas ‘I knew (all the time) what you are
doing’, sciebam quid egeris ‘I knew what you did’, and sciebam quid acturus sis ‘I knew
what you are planning to do’. An example of such a combination is (h). (NB: The
indicative praedicabant is not uncommon in Early Latin.)
(h) Non tu scis, mulier, Hecubam quapropter canem / Graii esse praedicabant?
(‘Don’t you know, woman, why the Greeks said that Hecuba was a dog?’ Pl. Men. 714–15)
In Table 7.16 there are six combinations with the verb scio. If the combinations of the
absolute strategy are added, there are in fact even more. In the case of impero, on the
other hand, the situation is entirely different; for this verb only two combinations are
possible. This is shown in Table 7.17 and illustrated by (i)–(k).243
Table 7.17 Sequence of tenses with the verb impero ‘to order’
Tense in the main clause Tense in the subordinate clause
primary tense present
dux imperat / imperavit / imperabit / imperaverit ut milites procedant
‘the commander orders / has ordered / will order / ‘that the soldiers move forward’
will have ordered’
historic tense imperfect
dux imperabat / imperavit / imperaverat ut milites procederent
‘the commander was ordering / ordered / had ordered’ ‘that the soldiers move forward’
(i) Ecce, Appollo mihi ex oraclo imperat, / ut ego illic oculos exuram lampadi-
bus ardentibus.
(‘Look, Apollo tells me through a divine utterance to burn out that woman’s eyes
with flaming torches.’ Pl. Men. 840–1)
(j) Esto, putabas. Quamobrem imperabas ut adderent?
(‘Oh, well, you did think so: but why did you order that payment?’ Cic. Ver. 3.71)
(k) Pseudolus mi ita imperavit ut aliquem hominem strenuom, / benevolentem
adducerem ad se.
(‘Pseudolus commanded me to bring him some energetic chap well disposed to me.’
Pl. Ps. 697–8)
The tenses in the first row on the left are traditionally called PRIMARY TENSES, those in
the second HISTORIC TENSES.244 With manipulation verbs like impero there is little
243
The table is taken from Woodcock (1959: 101–2).
244
‘Haupttempora’ and ‘Nebentempora’ in German.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
choice in the subordinate clause when the tense of the main clause is taken as the
reference point. Firstly, one cannot order someone to ‘have done’ something, which
excludes a perfect or pluperfect. Secondly, the semantic value of the -urus + sum
future (prospective—see § 7.24) is incompatible with use in a clause governed by a
manipulation verb. What one is ordered to do is posterior to the act of ordering, but
Latin uses the present or imperfect subjunctive form in this situation.245 In the case of
manipulation verbs there is therefore a simple rule: after primary tenses use the
present subjunctive, after historic tenses use the imperfect subjunctive. This limitation
of the choice of tenses follows from the meaning of the verb in the main clause.
However, in this case too one can imagine a commander giving an order in the past
that still has to be followed in the present. This is illustrated by (l).
(l) Durum videtur et grave quod Dominus imperavit ut si quis eum vult sequi
abneget se ipsum.
(‘Hard and grievous does that appear which the Lord has enjoined, that whosoever
will come after Him, must deny himself.’ August. Serm. 96)
The strategy of determining the tense of a subordinate clause on the basis of its
governing (main or other subordinate) clause is the statistically predominant one in
the Classical authors Cicero and Caesar (although there are quite a few instances of
the other strategy); for that reason it has played an important role in the manuals used
to teach students to write Latin. The rules formulated in this context are known as the
CONSECUTIO TEMPORUM ‘sequence of tenses’. This term is used especially for the rules
that govern the use of subjunctive forms in subordinate clauses (for an explanation
see § 7.117). However, similar observations can be made about indicative forms in
subordinate clauses.
Already the grammarian Charisius (3.8 [I.263-4K = 347–9B]) while teaching his son
(and other non-Latin-speaking readers) the rules of Latin, observed that only a
limited number of combinations of tenses in main and subordinate clauses are
actually used. A combination of a perfect tense in the main clause and a present
tense in the subordinate occurs nonnumquam ‘sometimes’ and the reverse combin-
ation of an imperfect tense in the subordinate clause and a present tense in the main
clause occurs raro ‘rarely’. Charisius adds that if his readers find this complicated,
they should not blame him but the tenuitas rei ‘the subtlety of the matter’. The matter
has indeed intrigued grammarians (and students of Latin) ever since. It must be
stressed here that Charisius’ description concerns the frequency of combinations of
tense forms in main and subordinate clauses, not the grammaticality or ungram-
maticality of these combinations.
Charisius was followed by Diomedes in his section de coniunctione temporum
(I.388–91K). Both illustrate the ‘rare’ phenomenon with a reference to the auctoritas
‘authority’ of Cicero: Cum nulla iam proscriptionis mentio fieret, cum etiam qui
245
For the ‘prospective’ use of the present and imperfect subjunctives, see Fridh (1971). See also § 7.69
on the use of the present and future infinitives.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In order to illustrate the strategy of taking the governing verb as the reference point
for the subordinate clause, a number of examples from Cicero are given below; they
belong to various classes of subordinate clauses. What they have in common is that
the state of affairs referred to in the subordinate clause also applies to the moment of
speaking.
The imperfect and pluperfect tenses in the subordinate clauses in (m)–(p) mark the
states of affairs as past and situated in or before the point in time at which the states of
affairs in the governing clauses occur. In (m) the fact that the two Metelli ‘are’
brothers is given as the motive for Cicero’s behaviour in the past. In (n) the artes
are still relevant to the correct way of life and still belong to the realm of philosophy,
but they are here presented as Cicero’s considerations for his past decision to write his
Tusculanae. In a similar way Menedemus’ opinions in (o) are relevant to orators and
audiences at any time, but the imperfect tenses of the relative clauses locate the
actions in Menedemus’ time. The pluperfect in this example marks the statement in
the relative clause as Menedemus’ responsibility. A perfect subjunctive would be
grammatical, but then the statement would be Antonius’, who tells this story. In
(p), finally, the use of the imperfect esset does not suggest that command over foreign
peoples is no longer prestigious; rather, it is merely set in the past.
(m) Addam illud etiam, quod iam ego curare non debui sed tamen fieri non
moleste tuli atque etiam ut ita fieret pro mea parte adiuvi, ut senati consulto
meus inimicus, quia tuus frater erat, sublevaretur.
(‘I will add something else: although I could not be expected, after what had passed,
to be active in the matter, I was not sorry to see my enemy relieved (because he was
your brother) by a senatorial decree, and I even assisted, so far as in me lay, to bring
this about.’ Cic. Fam. 5.2.9)
(n) . . . et cum omnium artium quae ad rectam vivendi viam pertinerent ratio et
disciplina studio sapientiae, quae philosophia dicitur, contineretur, hoc mihi
Latinis litteris inlustrandum putavi . . .
(‘ . . . and since the system and method of instruction in all the arts which have a
bearing upon the right conduct of life is bound up with the study of wisdom which
goes by the name of philosophy, I thought this must be made clear by me in a Latin
work . . . ’ Cic. Tusc. 1.1)
(o) Caput enim esse arbitrabatur (sc. Menedemus) oratoris ut et ipsis apud quos
ageret talis qualem se ipse optaret videretur; id fieri vitae dignitate, de qua
nihil rhetorici isti doctores in praeceptis suis reliquissent; et uti ei qui
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
audirent sic adficerentur animis ut eos adfici vellet orator; quod item fieri
nullo modo posse, nisi cognosset is qui diceret quot modis hominum mentes
et quibus et quo genere orationis in quamque partem moverentur; haec
autem esse penitus in media philosophia retrusa atque abdita, quae isti
rhetores ne primoribus quidem labris attigissent.
(‘For [Menedemus] was of the opinion that the main object of the orator was that he
should both appear himself, to those before whom he was pleading, to be such a man
as he would desire to seem; an end to be attained by a reputable mode of life, as to
which those teachers of rhetoric had left no hint among their instructions; and that
the hearts of his hearers should be touched in such fashion as the orator would have
them touched; another purpose only to be achieved by a speaker who had investigated
all the ways wherein, and all the allurements and kind of diction whereby, the judgement
of men might be inclined to this side or to that. But according to him such knowledge lay
thrust away and buried deep in the very heart of philosophy, and those rhetoricians had
not so much as tasted it with the tip of the tongue.’ Cic. de Orat. 1.87)
(p) (Sicilia) . . . prima docuit maiores nostros quam praeclarum esset exteris
gentibus imperare . . .
(‘She first taught our ancestors how splendid it is to rule over foreign peoples.’ Cic.
Ver. 2.2)
The choice of this strategy by a given speaker was apparently, judging from our
sources, not entirely free. It seems to have been the ‘default’ option; the only time
it was not chosen was when the relevance of the state of affairs to the time of speaking
was at stake. In such a situation a tense with the time of speaking as its reference point
was obligatory. However, using the default option did not mean ‘mechanically copy-
ing’ the tense of the main clause to the subordinate clause in the way agreement in
noun phrases works; instead, it was in the end a semantically based decision. This can
be shown by the following set of examples.
In (q) Varro opts for the past reference point implied by perfect institui and thus
chooses the pluperfect subjunctive form essent imposita, but he could also have taken
the time of speaking as his reference point, which would have given perfect subjunct-
ive sint imposita, as in (r). This would make the event referred to by the subordinate
clause less a part of a historical narrative and more a statement in retrospect from the
point of view of the author. In (s) Cicero switches from perfect subjunctive dixerim to
imperfect subjunctive monerem. The perfect subjunctive presents the state of affairs
pro me dixerim from the perspective of the time of speaking and as finished before
that time; the imperfect subjunctive, by contrast, presents the state of affairs quosdam
nimis ieiuno animo et angusto monerem as Cicero’s intention when he was addressing
the Senate (and as an ongoing reminder).
(q) Quemadmodum vocabula essent imposita rebus in lingua Latina, sex libris
exponere institui.
(‘In what way names were applied to things in Latin, I have undertaken to expound,
in six books.’ Var. L. 5.1)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(r) In his (libris) ad te scribam, a quibus rebus vocabula imposita sint in lingua
Latina.
(‘I will write you in these books from what things names have been applied in Latin.’
Var. L. 5.1)
(s) Haec interposui, patres conscripti, non tam ut pro me dixerim . . . quam ut
quosdam nimis ieiuno animo et angusto monerem . . .
(‘I have interposed these remarks, Conscript Fathers, not so much as an apology
for myself . . . as that I might advise . . . certain persons of too puny and narrow a
spirit . . . ’ Cic. Phil. 14.17)
Tables 7.16 and 7.17 illustrate that the expression of the temporal relationship
between a main clause and a subordinate clause is not the same for all lexical classes
of governing verbs. But there are other factors which make it impossible to mechan-
ically choose the tense of a subordinate clause on the basis of the tense of its governing
clause. Four factors will be dealt with in the following sections:
• the tense of the verb form of the governing clause
• the relative order of subordinate and main clause
• the type of subordinate clause
• semantic properties of the subordinate clause.
246
K.-St.: II.186, n.1 refer to Wetzel’s (1884) mentioning c.100 instances that were emended in Cicero.
247
See Torrego (1999a) for similar reservations.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
From Table 7.17 in § 7.85 it can be inferred that the perfect tense counts both as a
primary and as a historic tense, and this is understandable on the basis of its semantic
value as described in § 7.30: it presents a state of affairs as anterior to the time of
speaking, but since the state of affairs itself belongs to the past, the perfect, by
implication, functions as a past tense as well. Exx. (a)–(c) illustrate this point. Ex.
(a) demonstrates that the perfect form imperavit functions as a past verb form, just
like the imperfect form imperabas in (b). The ut clauses are both in the imperfect tense,
and the states of affairs referred to are both presented as the fulfilment of commands of
the past. However, in (c), with a present subjunctive form in the subordinate clause, the
fulfilment of the order of God is relevant to the time of speaking. Accordingly, the use of
different tenses in (a) and (c) corresponds to a difference in meaning.
(a) . . . Antiochum ad se vocavit et voce clara imperavit ut eum qui illum
olearum ordinem intrasset occideret.
(‘ . . . and he called Antiochus to himself and ordered him in a loud voice to kill
anyone who came within the row of olive trees.’ Cic. Caec. 22)
(b) Esto, putabas. Quamobrem imperabas ut adderent?
(‘Oh, well, you did think so: but why did you order that payment?’ Cic. Ver. 3.71)
(c) Durum videtur et grave quod Dominus imperavit ut si quis eum vult sequi
abneget se ipsum.
(‘Hard and grievous does that appear which the Lord has enjoined, that whosoever
will come after Him, must deny himself.’ August. Serm. 96.1)
Many ‘exceptions’ to the rule that a historic tense in the main clause should entail a
historic tense in the subordinate clause are in fact instances of perfect : present
combinations, several of which can be explained from the double nature of the
perfect, as illustrated above.
The historic present is treated both as a primary and as a historic tense. Plautus,
Caesar, and poets like Ovid prefer to use the present in the main and the subordinate
clauses (that is, they transpose the entire complex of past events to the present), but
Cicero and the historians have a clear preference for using a past tense in the
subordinate clause (they only transpose the state of affairs in the main clause to the
present). Primary sequence is more common in argument and purpose clauses—
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
types of clauses that are more closely related to the governing clause—and also in
restrictive relative clauses. Two examples of a historic present in the main clause from
Cicero are (a) with present subjunctives in the subordinate clause, and (b) with an
imperfect subjunctive in the subordinate clause. In contrast to what was observed
above about tense variation in subjunctive clauses after the perfect form imperavit, the
difference in tense in (a) and (b) does not correspond to a difference in meaning: both
expressions are used to describe the same sets of states of affairs, albeit in different
ways. Example (c) illustrates the difference in the choice of tense between different
types of subordinate clauses, with primary sequence in the argument clause and
secondary sequence in the conditional clause.
(a) Imperat Agyrinensibus ut decumas ipsi publice accipiant, Apronio lucrum dent.
(‘He orders the people of Agyrium to take over, as a community, the collection of the
tithe and to pay Apronius a bonus as well.’ Cic. Ver. 3.71)
(b) Ei palam imperat ut omne argentum quod apud quemque esset Catinae
conquirendum curaret et ad se adferendum.
(‘He openly ordered him to see that all the silver plate in all the houses in Catina was
sought for and brought to him.’ Cic. Ver. 4.50)
(c) Rogat et orat Dolabellam ut de sua provincia decedat, ad Neronem profi-
ciscatur. Se demonstrat incolumem esse non posse, si Philodamo vivere
atque aliquando Romam venire licuisset.
(‘He begged and implored Dolabella to leave his province and pay a visit to Nero. He
pointed out that he could not be safe if Philodamus were allowed to live and later on
to come to Rome.’ Cic. Ver. 1.72)
The relative order of main and subordinate clauses plays a role as well. When a
subordinate clause precedes a main clause in which the verb is in the historic present,
the tense in the former is more often historic, especially when there is no historic
present in the preceding sentence, as in (d). The use of the past tense in the preceding
subordinate clause probably serves as a signal that facilitates the interpretation of the
present tense in the main clause as historic (see § 7.16). For similar reasons a second
degree subordinate clause depending on another subordinate clause with a historic
present is more often in a historic tense, at least in Cicero; examples are (e) and (f ). In
Livy, however, other combinations are found as well, as in (g).
(d) Homo ut aliquo modo in provinciam illum revocaret, hanc excogitat ratio-
nem, si haec ratio potius quam amentia nominanda est.
(‘To get him back somehow to the province, [Verres] devised the plan, if this must be
called a plan rather than madness.’ Cic. Ver. 4.40)
(e) Adversarii postulant ut in eam rem iudices dentur ex iis civitatibus quae in id
forum convenirent.
(‘His adversaries demanded that judges for that matter be given from those cities
which assembled in that forum.’ Cic. Ver. 2.38)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The historic infinitive is treated primarily as a historic tense in Cicero and Caesar, but
other authors seem to use it as a primary tense. In (a) the infinitive counts as past, as
appears from the use of the imperfect tense dicerentur in the indirect question. In (b),
by contrast, a perfect tense form ambederit occurs in the explicative subject ut
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
clause, so the infinitive is treated as primary. The perfect may either be explained as
determined independently on the basis of the time of speaking or as related to the
interpretation of the infinitive in its main clause as primary. The same variation is
found in non-argument clauses (see examples in the Supplement). There is no
difference in meaning involved.
(a) Graecus primo distinguere atque dividere illa quemadmodum dicerentur.
(‘The Greek at first drew distinctions and split hairs about the meaning of those
doctrines.’ Cic. Pis. 69)
(b) Mi illud videri mirum ut una illaec capra / uxoris simiai dotem ambederit.
(‘It seemed strange to me that that one goat should have eaten up the monkey’s wife’s
dowry.’ Pl. Mer. 240–1)
Supplement:
Primary sequence: Nam postquam iste advenit Chremes adulescens, frater virginis, /
militem rogat ut eum admitti iubeat; continuo ille irasci, / neque negare audere.
Thais porro instare ut hominem invitet. (Ter. Eu. 617–19); Hunc ubi contiguum
missae fore credidit hastae, / ire prior Pallas, siqua fors adiuvet ausum / viribus
imparibus . . . (Verg. A. 10.457–9); Miles aegre teneri, clamare et poscere ut perculsis
instare liceat. (Liv. 2.65.2);
Historic sequence: Ita loqui homines: ‘Quod iudicia tam diu facta non essent,
condemnari eum oportere qui primus in iudicium adductus esset.’ (Cic. S. Rosc.
28); Neque Hasdrubal alium quemquam praeficere malle, ubi quid fortiter ac strenue
agendum esset . . . (Liv. 21.4.4);
(a) Nam ni ita esset, tecum orarem ut ei quod posses mali / facere faceres.
(‘If it weren’t like this, I’d ask you to do him every bad turn you could.’ Pl. Bac.
554–5)
(b) . . . vellem scriberes cur ita putares.
(‘ . . . would that you were writing why you were thinking in that way.’ Cic. Att. 11.24.5)
(c) Hisce ego rebus exempla adiungerem, nisi apud quos haec haberetur oratio
cernerem.
(‘I would attach examples to these matters, if I were not seeing among whom this
oration is held.’ Cic. de Orat. 1.190)
(d) Quae si bis bina quot essent didicisset Epicurus, certe non diceret.
(‘If Epicurus had learnt how much twice two was, he certainly would not speak in this
way.’ Cic. N.D. 2.49)
However, there are also instances in which the tense of the subordinate clause is
relative to the time of speaking. This is relatively common in the types of subordinate
clauses in which deviation from the sequence of tenses is also found with some
frequency in other constellations (see §7.95ff.), but there are also instances in indirect
questions following their main clause and in temporal clauses. In (e) present sub-
junctive macerent is in an indirect question embedded in a protasis with counterfac-
tual imperfect subjunctive possem. An example of a preceding indirect question, with
present subjunctive valeant is (f ).
(e) Si . . . fieri possem certior, / ere, quae miseriae te tam misere macerent, /
duorum labori ego hominum parsissem lubens . . .
(‘Master, if I could get the information . . . as to what misery is vexing you so
miserably, I’d have been happy to spare two people from trouble . . . ’ Pl. Ps. 3–5)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(f ) Quae quantum in provincia valeant, vellem expertus essem, sed tamen suspicor.
(‘Would that I knew from experience how much some things are strong in the
province, but, nonetheless, I suspect.’ Cic. Fam. 13.6.4)
Supplement:
Indirect question: Si fieri posset, quid sentiam ostendere quam loqui mallem. (Sen.
Ep. 75.2);
Temporal clause: Eundem equidem mallem audire Cottam, dum qua eloquentia
falsos deos sustulit eadem veros inducat. (Cic. N.D. 2.2);
Reason clause: Qui cum tantum ausus sit ustor pro mortuo, quid signifer pro vivo
non esset ausurus? (Cic. Mil. 90);
Relative clause: Eorum aliquid, in quibus aut causa sit honestissima aut necessitudo
certissima tamenne accusaretis? (Cic. Inv. 2.140); Qui enim haec fugiens fecerit quid
faceret insequens? (Cic. Phil. 3.27);
Result clause: Adduci tamen non possem ut quemquam mortuum coniungerem cum
deorum immortalium religione, ut, cuius sepulcrum [n]usquam exstet ubi parentetur,
ei publice supplicetur. (Cic. Phil. 1.13);
Purpose clause: Sublimen medium arriperem et capite pronum in terra statuerem, /
ut cerebro dispergat viam. (Ter. Ad. 316–7);
Manuscript variation is found in the following texts: Iam videretis quibus hominibus
omnium rerum et vendendarum et emendarum potestatem permiseritis. (Cic. Agr.
2.63—v.l. permitteretis); Memorare possum, quibus in locis maxumas hostium copias
populus Romanus parva manu fuderit . . . ni ea res longius nos ab incepto traheret.
(Sal. Cat. 7.7—v.l. possem).
248
The relevant instances in Cicero can be found in Lebreton (1901a: 248–50, 260–1). For discussion,
see Wisse (2013).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
of this type. (For the reverse situation when the main clause has a historic present, see
§ 7.89.) In this case the presence of deictic hesterno die may have contributed as well.
(a) Quae fuerit hesterno die Cn. Pompei gravitas in dicendo, iudices, . . .
perspicua admiratione declarari videbatur.
(‘What weight of words was shown by Gnaeus Pompeius in his speech yesterday,
gentlemen, . . . was clearly manifested . . . by your evident admiration.’ Cic. Balb. 2)
249
The need to deal with different types of clauses in different ways is very clearly expressed by
Lebreton (1901a: 277–8).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Table 7.18 Constraints on the use of the tenses in a number of subordinate clauses
Sentence level Noun phrase level
(relative clauses)
Arguments Satellites ‘Final’ ++
impero, volo ut/ne ++ adjuncts disjuncts descriptive +
timeo ne + purpose ++ illocutionary – other 0
non dubito quin 0 result + attitudinal – parenthetical –
scio, dico quod 0 other 0
++ = strong restrictions; + = moderate restrictions; 0 = no restrictions; – = sequence of tenses is excluded
Supplement:
Ego patrem exoravi. # Nempe ergo hoc ut faceret quod loquor? / # Immo tibi ne noceat
neu quid ob eam rem suscenseat. (Pl. Bac. 689–90—NB: variation in tense between
faceret and noceat, suscenseat); Nunc hoc me orare a vobis iussit Iuppiter, / ut
conquistores singula in subsellia / eant per totam caveam spectatoribus . . . (Pl. Am.
64–6); HAICE·UTEI·IN·COVENTIONID·EXDEICATIS·NE· MINUS·TRINUM / NOUNDINUM·, SENATUOS-
QUE·SENTENTIAM·UTEI·SCIENTES·ESETIS·EORUM / SENTENTIA·ITA·FUIT·: SEI·QUES·ESENT·QUEI·ARVOR-
SUM·EAD·FECISENT·QUAM·SUPRAD / SCRIPTUM·EST·EEIS·REM·CAPUTALEM·FACIENDAM·CENSUERE·
ATQUE·UTEI / HOCE·IN·TABOLAM· INCEIDER·ETIS·ITA·SENATUS·AIQUOM·CENSUIT / UTEIQUE·EAM·FI-
2
GIER·IOUBEATIS·UBEI·FACILUMED·GNOSCIER·POTISIT. (CIL I .581.22–7 (SCBac., Tiriolo, 186
BC)—note the variation between present and imperfect); Nunc se obsequentem atque
hilare dixi praebeat. (Afran. com. 316);
Whereas manipulation and related verbs normally imply the posteriority of the state
of affairs in the subordinate clause with respect to the one in the governing clause,
with some other verbs there is no reason why the state of affairs in the subordinate
clause cannot be positioned independently. Exx. (c)–(f ) are deviations from the
sequence of tenses, which in their contexts are fully justified.
make the narration of the order stick more closely to its actual phrasing—that is, it is a
form of repraesentatio (see § 7.113).
(g) Adhortatus milites ne necessario tempore itineris labore permoveantur
cupidissimis omnibus progressus milia passuum XXV agmen Haeduorum
conspicatur.
(‘He urged the troops not to be disturbed by the fatigue of a march which the
emergency rendered necessary, and then amid the greatest eagerness of all ranks,
he advanced for five-and-twenty miles, when he caught sight of the column of the
Aedui.’ Caes. Gal. 7.40.4)
(h) Ibi dictator . . . tribunis militum imperavit ut sarcinas in unum conici
iubeant . . .
(‘Then the dictator . . . directed the military tribunes to make the soldiers throw down
their packs in one place.’ Liv. 3.28.1)
Supplement:
Ad haec Q. Marcius respondit, si quid ab senatu petere vellent (NB: remarkable
imperfect), ab armis discedant, Romam supplices proficiscantur. (Sal. Cat. 34.1);
Nepesinis inde edictum ut arma ponant, parcique iussum inermi. (Liv.
6.10.5); . . . nixi genibus ab senatu petierunt ne se socios foedius spoliari vexarique
quam hostes patiantur. (Liv. 43.2.2); . . . identidem monuit ac saepe terruit ne quando
sororum pernicioso consilio suasa de forma mariti quaerat neve se sacrilega curio-
sitate de tanto fortunarum suggestu pessum deiciat nec suum postea contingat
amplexum. (Apul. Met. 5.6.6);
Four Caesarian instances (Gal. 3.18.8; 7.40.4 (ex. (g) in the running text); 7.80.1; Civ.
3.20.4) were emended by Meusel or others to fit in with the sequence of tenses.
Instances of a primary tense in the main clause with a historic tense in the argument
clause are rare, although with some of the classes of governing verbs there is no
semantic constraint. Instances from Gaius and Tertullian are (k) and (l).250
250
The examples are taken from Kalb (1912: 76) and Hoppe (1903: 68 (= 1985: 134)), respectively.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
251
So Grassi (1967: 53).
252
Ex. (h) is taken from K.-St.: II.175–6, where other types of subordinate clauses with the same
phenomenon are quoted.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(h) (Sicilia) . . . prima docuit maiores nostros quam praeclarum esset exteris
gentibus imperare.
(‘Sicily was the first who made our forefathers perceive how splendid it is to rule over
foreign nations.’ Cic. Ver. 2.2)
(i) (Antonius) Doceret me vel potius vos, patres conscripti, quem ad modum
ipse Caesaris acta defenderet.
(‘He (Antony) would explain to me, or rather to you, Conscript Fathers, how he
himself defended Caesar’s acts.’ Cic. Phil. 1.16)
7.99 The use of the tenses in declarative quod and quia argument clauses
The following discussion of quod and quia clauses deals with indicative and sub-
junctive clauses. The tense in declarative subordinate clauses governed by cognition
and communication verbs with subordinators like quod and quia in Late Latin (where
the accusative and infinitive construction is still the norm) is sometimes related to the
tense of the main clause, thus conforming with the sequence of tenses, but independ-
ent tense selection is found as well.253 This is illustrated by the following examples
from Augustine. In (a)–(c) the state of affairs of the subordinate clause is related to
the time of speaking. Ex. (a) has a simple future in the subordinate clause, (b) a
periphrastic future expression with the present indicative form es, (c) a periphrastic
future expression with the present subjunctive form simus. In (d)–(e), by contrast, the
tense forms of the subordinate clauses (periphrastic future expressions with an
imperfect subjunctive esset and imperfect indicative erant) are related to the tenses
in the main clause (perfect and imperfect, respectively). In (f ), the tense of the
subordinate clause (imperfect subjunctive dares) is also related to the past tense
(promiseras) of the main clause.254
(a) Quare intellegendum est et illud recte dictum esse quod Deus retribuet
parentum peccata filiis qui eum oderunt.
(‘And therefore it must be understood that also this has been said correctly, namely
that God repays the parents’ sins on their children, who hate him.’ August. Adim. 7)
(b) Quid tibi promisit Deus, o homo mortalis? Quia victurus es in aeternum.
(‘What did God promise to you, mortal man? That you would live forever.’ August.
Psal. 148.8)
(c) Promisit tamen nobis quod ibi futuri simus unde venit.
(‘Nevertheless, he promised to us that we would be there from whence he came . . . ’
August. Psal. 148.8)
(d) Quando promisit filium, credidi quod mihi daturus esset Deus posteritatem.
(‘When he promised his son, I believed that God would give the future to me.’
August. Serm. 2.1)
253
Subordinate clauses with quod and quia are dealt with here, although the mood is often indicative.
254
The examples are taken from Roca (1999: 159, 164).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(e) Erant ibi alii qui . . . promittebant sibi quia ad dexteram futuri erant quibus
dicturus erat Christus.
(‘There were others there, who were promising themselves that they, to whom Christ
was about to speak, were to be on the right-hand side.’ August. Psal. 90.10)
(f ) Promiseras enim quod non dares sanctum tuum videre corruptionem, et
nunc in sepulcro tegitur.
(‘For you had promised that you would not allow your holy object to see corruption,
and now it is covered in the tomb.’ Hier. in Psal. II.88.272)
7.101 The use of the subjunctive tense forms in purpose (final) adjunct clauses
The relationship between the state of affairs referred to in the main clause and that
referred to in a purpose adjunct clause resembles the one between argument clauses
and the manipulation verbs that govern them: the state of affairs of the subordinate
clause is posterior to that of the main clause.255 The use of the tenses is very similar. In
(a), the main clause has a present tense form sumus, and the final ut clause is in the
present as well. Ex. (b) has the imperfect proficiscebar in the main clause, and the ut
clause is likewise in the imperfect. In (c), the perfect form locutus sum is treated as a
past (anterior to the time of speaking) tense, and the purpose clause has imperfect
excitarem. In (d), however, it is the time of speaking element of the perfect praemisit
that determines present nuntiem in the ut clause. A remarkable example is (e), where
the perfect form must be explained as ‘retrospective’ (see § 7.30): ‘to obtain a situation
in which I appear to have refuted these claims’.256 Cases with a secondary tense in the
subordinate clause and a primary one in the main clause are rare. An example from
Tertullian is (f ).
(a) Legum ministri magistratus, legum interpretes iudices, legum (v.l. legibus)
denique idcirco omnes servi sumus (v.l. servimus), ut liberi esse possimus.
(‘The magistrates who administer the law, the jurors who interpret it—all of us in
short—obey the law to the end that we may be free.’ Cic. Clu. 146)
(b) A. d. VI Id. Mai., cum has dabam litteras, ex Pompeiano proficiscebar ut eo
die manerem in Trebulano apud Pontium.
(‘When I was dispatching this letter on 10 May, I was just about to leave Pompeii for
the Trebula country where I am to stay the night with Pontius.’ Cic. Att. 5.2.1)
(c) Atque haec non ut vos . . . excitarem locutus sum . . .
(‘And I said these things not to excite you . . . ’ Cic. Catil. 4.19)
255
For a treatment in the framework of Functional Grammar, see Vester (1994: 274–5).
256
For a few instances of the perfect in purpose clauses in Early Latin, see Bennett (1910: I. 254, 257).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(g) Pinnas autem ideo illi fabulosa vetustas aptavit, ut adesse velocitate volucri
cunctis existimetur . . .
(‘Moreover the storied past has given her wings in order that she might be
thought to come to all with swift speed . . . ’ Amm. 14.11.26)257
7.102 The use of the subjunctive tense forms in result (consecutive) adjunct clauses
In result (consecutive) adjunct clauses with a main clause that contains a past tense,
imperfect subjunctive forms predominate statistically, which seems to indicate ‘obedi-
ence’ to the rules for the sequence of tenses. An example of this is (a).258 However,
here too the so-called deviations from these rules often correspond to the author’s
desire to avoid ambiguity. An example of a present subjunctive in the subordinate
clause is (b). The present tense indicates that, at the time of Cicero’s speech,
reconstruction is still impossible. Ex. (c) has a perfect subjunctive in the subordinate
clause. There are also instances of the coordination of two result clauses with different
257
For repraesentatio in Ammianus, see Ehrismann (1886: 16ff.).
258
More parallels can be found in K.-St.: II.187–9, who refer to the abundant material in Lebreton
(1901a: 227ff.). Recent studies are Sánchez-Lafuente and García de la Calera (1997), Cabrillana (1999),
Gallego (2007, 2010), and Cuzzolin (2008). See also the discussion and references in Calboli (2012: 48–56,
59).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
tense forms. An example of this is (d), where a clause with a perfect subjunctive
(caruerint) is coordinated with one with an imperfect subjunctive (sustentarent—
often emended). Ex. (e) shows embedding of one result clause in another, with two
imperfect subjunctives in the coordinated higher result clauses (cingeretur and nite-
retur) and a perfect subjunctive (permanserit) in the one that is embedded within ut
ita munita arx . . . saxo niteretur. Exx. (f ) and (g) are almost a minimal pair, describ-
ing more or less the same situation, the first with an imperfect subjunctive in the ut
clause, the second with a perfect subjunctive. In sum, examples (d)–(g) demonstrate
that there is a choice between the perfect and the imperfect, and that the choice
correlates with a difference in meaning. The perfect subjunctives refer to the states of
affairs as a conclusion in retrospect, just as the perfect indicative does in simple
sentences. Evaluative expressions like certe in (h) are therefore not uncommon in
result clauses in the perfect.
(a) (Viriatus) . . . quem C. Laelius . . . fregit et comminuit ferocitatemque eius ita
repressit ut facile bellum reliquis traderet.
(‘But Gaius Laelius . . . crushed his power, reduced him to terms, and so checked his
intrepid daring that he left to his successors an easy conquest.’ Cic. Off. 2.40)
(b) . . . Sicilia, quam iste per triennium ita vexavit ac perdidit ut ea restitui in
antiquum statum nullo modo possit . . .
(‘ . . . (Sicily), which island for the space of three years he devastated and ruined so
effectively that nothing can restore it to its former condition . . . ’ Cic. Ver. 1.12)
(c) Quamquam enim adeo excellebat Aristides abstinentia ut unus post homi-
num memoriam, quem quidem nos audierimus, cognomine Iustus sit
appellatus, tamen a Themistocle collabefactus testula illa exilio decem
annorum multatus est.
(‘For although Aristides so excelled in honesty that he is the only one within the
memory of man—at least so far as we have heard—who was given the title “the Just”,
yet his influence was undermined by Themistocles and he was sentenced to exile for
ten years by that process known as the shard-vote.’ Nep. Ar. 1.2)
(d) Summa difficultate rei frumentariae adfecto exercitu . . . usque eo ut complures
dies frumento milites caruerint et . . . extremam famem sustentarent . . .
(‘So the army suffered from the utmost difficulty in its corn supply . . . so much so
that for several days the troops were without corn, and staved off the extremity of
famine . . . ’ Caes. Gal. 7.17.3)
(e) (Urbs) Cuius is est tractus ductusque muri . . . definitus . . . ut unus aditus . . . fossa
cingeretur vastissima, atque ut ita munita arx . . . saxo niteretur ut etiam in illa
tempestate horribili Gallici adventus incolumis atque intacta permanserit.
(‘The line and course of its walls . . . were planned . . . that the single approach . . . was
girt about by . . . a mighty trench; and our citadel was so well fortified by . . . the rock
. . . so that it remained safe; impregnable even at the terrible time of the advent of the
Gauls.’ Cic. Rep. 2.11)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The previous examples illustrate the independent positioning of the state of affairs in
the result clause. The following examples (i) and (j) show another form of deviation
from the sequence of tenses: while imperfect anteponeret and essem locate in the past
the states of affairs in the result clauses, the main clauses are in the present tense.
(i) Ac si nos . . . nostra patria delectat, cuius rei tanta est vis ac tanta natura ut
Ithacam illam . . . sapientissimus vir immortalitati anteponeret . . .
(‘And if our own native land is our joy, the force and nature of which reality is so great
that the wisest man preferred that Ithaca of his to immortality . . . ’ Cic. de Orat. 1.196)
(j) Tanta est enim caritas patriae ut vestris etiam legionibus sanctus essem, quod
eam a me servatam esse meminissent.
(‘For so great is love of country that even to your legions I was sacred because they
remembered that it had been saved by me.’ Cic. Phil. 2.60)
The use of periphrastic -urus + sum forms is rare.260 When they are used, they keep
their specific prospective value (see § 7.24). Two examples are (k) and (l), the latter a
typical Livian idiom. See also exx. (i) and (j) in § 7.97.
(k) Tantum fortunam mutasse ut qui modo ipsi exercitum ante moenia
Romana habuissent victores . . . ii verso Marte Africae populationes et ob-
sidionem Carthaginis visuri forent.
(‘So far had fortune changed that those who as victors had but lately had
their own army before the walls of Rome with the shifting fortune of war
were now about to see the devastation of Africa and a siege of Carthage.’ Liv.
29.3.10)
(l) Id imperium adeo superbum et indignum Lacedaemoniis visum est ut, si
antiqua civitatis fortuna esset, haud dubie arma extemplo capturi fuerint.
(‘This demand seemed to the Lacedaemonians so insolent and unmerited
that without doubt they would immediately have been ready to take up
arms, if the fortune of the state had been as it was in olden days.’ Liv.
38.31.3)
259
See Cabrillana (1999: 104, table 5: twenty-three out of twenty-four instances), also for a review of
recent publications and statistical data.
260
For the use of the periphrastic future in result clauses in the historians, see Gallego (2010).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Quae cum esset civitas aequissimo iure ac foedere, ascribi se in eam civitatem
voluit . . .
(‘Full civic privileges had been accorded to this town by the terms of its treaty
with Rome, and Archias expressed a wish to be enrolled among its citizens . . . ’ Cic.
Arch. 6)
(b) Nam Drusus, quamquam arduum sit eodem loci potentiam et concordiam
esse, aequus adulescentibus aut certe non adversus habebatur.
(‘For Drusus, difficult as it is for power and concord to dwell together, had the reputation
of being well disposed or at least not inimical, to the youths.’ Tac. Ann. 4.4.1)
(c) Laudantur (laudabantur cj. Ernesti) oratores veteres, Crassi illi et Antonii,
quod crimina diluere dilucide, quod copiose reorum causas defendere
solerent.
(‘Famous orators of the past, like Crassus and Antonius, are celebrated for their
brilliant way of undermining the prosecutor’s case and bringing up a mass of
arguments to support the accused.’ Cic. Ver. 2.191)
Supplement:
Reason clauses: Non pol quo quemquam plus amem aut plus diligam, / eo feci.
(Ter. Eu. 96–7); . . . quod hunc laborem alteri delegavi, non quin mihi suavissimum sit
et occupato et ad litteras scribendas, ut tu nosti, pigerrimo tuae memoriae dare
operam . . . (Cic. Fam. 8.1.1);
Concessive clauses: Tu ipse iam statue quam verum sit, cum paucas tegulas
deicere inpune familia tua non potuerit, maximam caedem sine fraude facere
potuisse[t]. (Cic. Tul. 53); Quae quamquam ita sint in promptu ut res disputa-
tione non egeat, tamen sunt a nobis alio loco disputata. (Cic. Off. 1.5); Addunt
etiam quicquid valde utile sit id fieri honestum, etiam si antea non videretur.
(Cic. Off. 3.103); Nam hoc toto proelio, cum ab hora septima ad vesperum
pugnatum sit, aversum hostem videre nemo potuit. (Caes. Gal. 1.26.2); Neque
tamen temperamenti egebat, cum aequabili auctoritate et gratia apud Tiberium
viguerit. (Tac. Ann. 4.20.2);
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Disjuncts indicate the way the speaker or author wants the message of the main clause to
be understood and are for that reason very much connected to the time of speaking.
‘Deviations’ from the sequence of tenses are therefore to be expected. Ex. (a) is an example
of an illocutionary purpose disjunct (see § 16.50), in (b) quod sciam restricts the validity of
the content of the main clause (see § 16.83). Loosely attached ironic or sarcastic compari-
son clauses also often convey the speaker’s or author’s point of view. An example is (c).
(a) Quodsi Catilina in urbe ad hanc diem remansisset . . . tamen, ut levissime
dicam, dimicandum nobis cum illo fuisset . . .
(‘Had Catiline remained in the city to this day . . . we should have had to fight him to
the finish—I cannot put it less strongly . . . ’ Cic. Catil. 3.17)
(b) Me quidem praesente numquam factum est, quod sciam.
(‘This never happened in my presence, as far as I know.’ Pl. Am. 749)
(c) Ita facti sunt repente Attici, quasi vero Trallianus fuerit Demosthenes.
(‘And so they suddenly turned into “Attici” as if Demosthenes came from Tralles!’
Cic. Orat. 234)
Supplement:
Illocutionary purpose disjuncts: Tantus igitur te stupor oppressit vel, ut verius
dicam, tantus furor ut . . . (Cic. Phil. 2.65); Fuit aliquis fatalis casus, ut ita dicam,
quem tulimus, quoquo modo ferendus fuit. (Cic. Phil. 6.19);
Clauses restricting the validity: Non edepol ego te, quod sciam, umquam ante hunc
diem / vidi nec novi. (Pl. Men. 500–1); Sestius VI Id. exspectabatur sed non venerat,
quod sciam (Cic. Att. 16.2.4); Neminem proconsulum, quod sciam, provincia Africa
magis reverita est . . . (Apul. Flor. 9);
The use of the perfect subjunctive without an anterior meaning in an illocutionary
purpose disjunct (see § 16.49) is rare and only found in Silver and Late Latin, as (d).
The use of the tenses in subjunctival relative clauses is dealt with in §§ 18.20–18.22.
The ‘exceptions’ to the sequence of tenses discussed in the preceding sections can
actually be understood as semantically justified because a certain tense is required to
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
convey the proper meaning. Relating the tense of a subordinate clause to the tense of
the main clause is in a similar way avoided when the corresponding simple sentence
would have a past potential (either in an interrogative or in a declarative sentence—
see § 7.43) or a counterfactual subjunctive. In these cases the past tense seems to
be in conflict with the sequence of tense rules, when the main clause has a present
or future tense: the sequence of tense rules do not predict a historic sequence with
such a main clause. Exx. (a)–(c) are indirect questions with a past potential
subjunctive, each with a present tense in the main clause. Ex. (d), in turn, is
also past potential and corresponds with the use of these forms in declarative
sentences; the main clause contains a future tense form. Exx. (e)–(h) have a
counterfactual subjunctive, just as they would have had in a simple sentence. For
a few conditional clauses, see the Supplement. In all these cases ‘obedience’ to the
sequence of tense rules would have resulted in sentences with another meaning
than was intended.
(‘ . . . but because it is of such a nature that even if men were unaware of its existence,
or never spoke of it, it would still be worthy of praise for its own beauty and
loveliness.’ Cic. Fin. 2.49)
Supplement:
Deinde (sc. comitia) habere coepit (sc. consul) subito praeter opinionem omnium, ut,
ne si cogitasset quidem largiri quispiam, daretur spatium comparandi. (Cic. Planc.
49); Quod ille si repudiasset, dubitatis quin ei vis esset adlata . . . (Cic. Sest. 62); Tanta
mihi cum eo necessitudo est familiaritasque ut, si mea res esset, non magis labo-
rarem. (Cic. Fam. 13.44.1); Neque enim verisimile est ut (sc. Paulus) avertens a
Iudaeismo non pariter ostenderet, in cuius dei fidem averteret. (Tert. Marc. 5.1.8);
With verbs like spero (ut) ‘I hope, expect that’ and (non) dubito (quin) ‘I have no
doubt that’, the state of affairs of the subordinate clause is often posterior to the time
of speaking. Although Latin may use the periphrastic -urus + sum expression as a
substitute for the missing (simple) future subjunctive if the prospective meaning of
this expression is applicable to the situation at hand, the normal way to express
posteriority is with the present and imperfect subjunctive forms that are related to a
primary or a secondary tense in the governing clause, just as with manipulation verbs.
Illustrations of this are (a)–(d). In (e) the -urus + sum expression with non dubitare,
quin has its usual prospective value.
(a) Non speramus ut illum iudex probet, sed ut dimittat.
(‘We do not hope that the judge will approve of him but that he will acquit him.’ Sen.
Con. 9.2.18)
(b) Qui si sensit quo se scelere devinxerit, non dubito quin sit miserrimus.
(‘If he only understood to what wickedness he has bound himself, I do not doubt that
he would be most wretched.’ Cic. Har. 5)
(c) Speraverat ut occidere me posset meo veneno.
(‘He had hoped that he would be able to kill me with my poison.’ [Quint.] Decl. 17.6)
(d) Non enim dubitabat Xeno quin ab Ariopagitis invito Memmio impetrari non
posset.
(‘For Xeno did not doubt that it would not be able to be obtained from the
Ariopagates with Memmius unwilling.’ Cic. Att. 5.11.6)
(e) Non enim dubitabam quin eas libenter lecturus esses. Verebar ut
redderentur.
(‘For I did not doubt that you would be glad to read it, but I was uneasy about its safe
delivery.’ Cic. Fam. 12.19.1)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
At one extreme, the tense in each clause is determined independently (type I). At the
other extreme the three clauses all have the same reference point (that of the main clause)
(type V). In between these two extremes there are three possibilities: the reference point
of the tense of the second degree clause may be the same as that of the immediately
governing first degree clause (type II); it may also be the same as that of the highest
governing (main) clause (type III); or it may differ from the reference point of both
(type IV). In terms of the sequence of tenses rules, the sequence is fully observed in type V.
In fact there is a strong correlation between the tense of the main clause and the
tense of the second degree clause, when the tense of the main clause is a historic one.
Since in such a situation the tense of the first degree clause is also often relative to the
tense of the main clause, type V is much more common than type III. When, however,
the verb of the main clause is present or future and the first degree clause has a perfect
(subjunctive) tense or a perfect infinitive, there is usually no such correlation between
the tense of the main verb and that of the second degree subordinate clause—that is,
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
one usually finds type II or type I. However, there are many exceptions to this general
‘rule’ (which is also mainly formulated on the basis of Cicero’s work), just as there are
many exceptions for first degree subordinate clauses, for which the factors involved
are largely the same.
There is usually no correlation between the tense in the lowest subordinate clause and
the tense in the main clause if the latter clause has a present or future tense. If that
subordinate clause depends on a subordinate clause with a perfect tense (usually in
the subjunctive) there are two possibilities, in accordance with the double nature of
the perfect (see § 7.88). This is illustrated by (a) and (b). In (a) the quod clause has a
perfect that is related to the time of speaking; in (b), by contrast, videremus in the ut
clause has a past reference point. Note that the third degree relative clause has viderit,
which is anchored to the time of speaking.
(a) . . . nemo in Sicilia dubitat quin eo sit occisus quod habere clausa non
potuerit sua consilia de Verre. (cf. occisus est quod non potuit)
(‘ . . . no one in Sicily has any doubt that he was murdered because he could not keep
his purpose hidden from Verres.’ Cic. Ver. 3.63)
(b) . . . qui motus hominum, qui ferarum non ita expictus est ut, quae ipse non
viderit, nos ut videremus effecerit? (cf. ut videremus effecit)
(‘ . . . what movements of men, of animals has he not depicted so vividly that he has
made us see the things which he himself did not see?’ Cic. Tusc. 5.114)
Supplement:
Sin ille a senatu notandus non fuit, quid potest dicere quin, cum de illo tacuerit, se
hostem confessus sit? (Cic. Phil. 3.21); Restat, iudices, ut hoc dubitemus, uter potius
Sex. Roscium occiderit, is ad quem morte eius divitiae venerint an is ad quem
mendicitas . . . (Cic. S. Rosc. 88);
Volo uti mihi respondeas . . . quae tanta in te fuerit audacia, quae tanta vis ut, quod
novem tui conlegae sibi timendum esse duxerint, id unus tu . . . inridendum putares.
(Cic. Vat. 17); Quod ad me scribis de sorore tua, testis erit tibi ipsa quantae mihi
curae fuerit ut Quinti fratris animus in eam esset is qui esse deberet. (Cic. Att. 1.5.2);
Non-finite verb forms such as infinitives, participles, and supines have no reference
point of their own. The tense of subordinate clauses that depend on clauses with such
verb forms is therefore usually either related to the tense of the finite form of the main
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
clause (type V in Table 7.19) or determined independently (type IV). For accusative
and infinitive clauses see § 7.111.
For participles and the supine the principle of taking the tense of the main clause as
the basis for the selection of the tense of the embedded clause (type V in Table 7.19) is
illustrated by the following examples. In (a) the present participle takes its reference
point from the present tense of the main verb, and in (b) the perfect participle in an
ablative absolute construction functions as a future perfect with respect to the simple
future accedam. This explains primary tense sit in the subordinate clause. Ex. (c) has a
supine consultum that takes its reference point from the perfect tense of the main
verb; likewise, in (d) the gerund takes its reference point from the perfect tense of the
main verb.
(a) Lacrimae meorum me interdum molliunt precantium ut de Hispaniis
exspectemus.
(‘The tears of those I love, who entreat me to wait for the outcome in Spain,
sometimes sap my resolution.’ Cic. Att. 10.9.2)
(b) Itaque eo quale sit breviter, ut tempus postulat, constituto accedam ad omnia
tua, Torquate, nisi memoria forte defecerit.
(‘Accordingly, I will begin by defining, with such brevity as the occasion demands,
the nature of Moral Worth; and then, Torquatus, I will proceed to deal with each of
your points, unless my memory should happen to fail me.’ Cic. Fin. 2.44)
(c) . . . miserunt Delphos consultum quidnam facerent de rebus suis.
(‘ . . . they sent them to ask the advice of Delphi, as to what they should do about their
matters.’ Nep. Them. 2.6)
(d) . . . cupido incessit animos iuvenum sciscitandi ad quem eorum regnum
Romanum esset venturum.
(‘ . . . a desire came upon the minds of the young men for ascertaining to which of
them the kingdom of Rome would come.’ Liv. 1.56.10)
Supplement:
. . . Artaxerxes reminiscens a quanto bello ad quam parvam rem principem ducum
misisset se ipse reprehendit . . . (Nep. Dat. 5.1); Tunc legatos ad urbem misit Licinium
et Probum, per conloquium petens ut gener apud socerum, id est Maxentius apud
Galerium, precibus magis quam armis optata mercaretur. (Exc. Vales. 3.7.2);
Adiuncto vero ut idem etiam prudentes haberentur nihil erat quod homines iis
auctoribus non posse consequi se arbitrarentur. (Cic. Off. 2.42);
Difficile est dictu, Quirites, quanto in odio simus apud exteras nationes . . .
(Cic. Man. 65);
The same principle holds for non-verbal categories, such as adjectives used as
secondary predicates or in apposition. In (e) ignari takes its reference point from
the perfect tense of the main verb. Even nouns can function as a reference point for
subordinate clauses, as in (f ) and (g). The subordinate clauses given below belong to
various types.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Hunc exitum habuit vir, nisi in libera civitate natus esset, memorabilis. (Liv. 6.20.14);
Quam possumus reperire ex omnibus rebus civilibus causam contentionemque
maiorem quam de ordine, de civitate, de libertate, de capite hominis consularis,
praesertim cum haec non in crimine aliquo, quod ille posset infitiari, sed in civili
iure consisteret? (Cic. de Orat. 1.181); Quod est igitur meum triste consilium? Ut
discederem fortasse in aliquas solitudines. (Cic. Fam. 2.16.2);
Appendix: Infinitives governed by auxiliary verbs are irrelevant in this context
because they form a phrase together with the governing verb and are not a clause
in and of themselves. It is the governing verb that determines the tense of the
subordinate clause, as is shown in (h) and (i). In (h) the phrase animum advertere
soleo behaves in the same way as experior, with which it is coordinated.
non valde moveri, non desinam te uti teste, ut hoc quod dico probem. (Cic. Rep. 1.67);
Vetere in nova coepi uti consuetudine, / in experiundo ut essem. (Ter. Hec. 37–8); Ex
quo leges quoque incipiunt neglegere, ut plane sine ullo domino sint. (Cic. Rep. 1.67);
Nam cum sciam spiritum illius in meo verti, incipio, ut illi consulam, mihi consulere.
(Sen. Ep. 104.2); Tris minas accudere etiam possum, ut triginta sient. (Pl. Mer. 432);
Ipse fugi adhuc omne munus, eo magis quod ita nihil poterat agi ut mihi et meis rebus
aptum esset. (Cic. Att. 11.4); Ipsa vero praecepta sic inludere solebat ut ostenderet non
modo eos illius expertes esse prudentiae quam sibi adsciscerent, sed ne hanc quidem
ipsam dicendi rationem ac viam nosse. (Cic. de Orat. 1.87);
In (a)–(f ) the reference point of the second degree subordinate clauses belonging to
an accusative and infinitive clause is the same as that of the main clause. Within these
examples there are three instances (a)–(c) of a main clause in the present tense and a
primary tense in the subordinate clause and three (d)–(f ) of an imperfect tense in the
main clause and a secondary tense in the subordinate clause. The present and future
infinitives of the accusative and infinitive clauses have the same reference point as the
main clauses as well. All of these are therefore clear examples of type V (A, a, Æ) in
Table 7.19. In terms of the sequence of tenses rules, the sequence is fully observed.
(a) Quod scribis te velle scire qui sit rei publicae status, summa dissensio est, sed
contentio dispar. [indirect question]
(‘You write that you wish to know what is the political situation. There is bitter
conflict, but the two sides are unequally matched.’ Cic. Fam. 1.7.10)
(b) Vatinium autem scire te velle ostendis quibus rebus adductus defenderim et
laudarim. [indirect question]
(‘But you intimate that you would like to know my reasons for defending Vatinius
and speaking to his character.’ Cic. Fam. 1.9.4)
(c) C. Arrius . . . etiam se idcirco Romam ire negat ut hic mecum totos dies
philosophetur. [purpose clause]
(‘C. Arrius . . . actually says he won’t go to Rome in order that he may philosophize
with me here all day long.’ Cic. Att. 2.14.2)
(d) Nullum in urbe vicum, nullum angiportum esse dicebant in quo non Miloni
conducta esset domus. [consecutive relative clause]
(‘They were saying that there was no quarter, no alley in the city, in which a house
had not been hired for Milo’s use.’ Cic. Mil. 64)
(e) Nam tantum abesse dicebat ut id consentaneum esset, ut maxime etiam
repugnaret. [argument clause + result clause]
(‘He used to declare that this was so far from being consistent that it is actually
grossly inconsistent.’ Cic. Luc. 28)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(f ) Quid minitabas te facturum, si istas pepulissem fores? (cf. quid facturus eras,
si pepulissem fores?)
(‘What did you threaten to do if I knocked at that door?’ Pl. Am. fr. 5) (cf. ‘what were
you going to do if I knocked at the door?’)
Supplement:
Additional examples with a present or future accusative and infinitive clause:
Primary sequence: Sic enim dico: si Campanus ager dividatur, exturbari et expelli
plebem ex agris, non constitui et collocari. (Cic. Agr. 2.84); Quemquam igitur putas
dubitaturum quin is quem tibi inimicissimum esse oporteat, qui aps te gravissimas
iniurias acceperit, pecuniam statuae nomine dederit . . . (Cic. Ver. 2.153); (sc. Pom-
peius) dicitque se perfecturum ut in illum Caesar incumbat. (Cic. Q. fr. 3.6.6); Ego
istum iuvenem domi tenendum sub legibus, sub magistratibus, docendum vivere
aequo iure cum ceteris censeo, ne quandoque parvus hic ignis incendium ingens
exsuscitet. (Liv. 21.3.6);
Historic sequence: Etenim minime conveniebat ei deorum honores haberi qui
simulacra deorum abstulisset. (Cic. Ver. 4.151); Accepi tuas litteras, quibus intellexi
te vereri ne superiores mihi redditae non essent. (Cic. Fam. 14.5.1); Nam cum in
concilio dixisset tuenda mediterranea Acarnaniae esse et omnibus qui arma ferrent
exeundum ad Medionem et Thyrreum, ne ab Antiocho aut Aetolis occuparentur,
fuere qui dicerent . . . (Liv. 36.11.10–1); Adiecit non esse credibile ut, quem tantopere
amaret, ab eo invicem non diligeretur. (V. Max. 4.1.2);
However, the reference point of the accusative and infinitive clause may also be
different from that of the main clause. In that case the reference point of the second
degree subordinate clause is usually the same as that of the accusative and infinitive
clause, that is, it belongs to type II of Table 7.19 (A, b, ). Examples are (g)–(i).
(g) De Britannicis rebus cognovi ex tuis litteris nihil esse nec quod metuamus
nec quod gaudeamus.
(‘As to British matters, I see from your letter that we have no cause for alarm nor for
rejoicing either.’ Cic. Q. fr. 3.1.10)
(h) . . . te interrogante augures responderunt cum de caelo servatum sit, cum
populo agi non posse . . .
(‘ . . . you questioned the augurs, and they replied that when any magistrate was
observing the heavens, no business could be transacted in the assembly of the people.’
Cic. Dom. 40)
(i) Omnino plura me scribere, cum tuum tantum consilium iudiciumque sit,
non ita necesse arbitrabar.
(‘Speaking generally, considering your own consummate sagacity and judgement,
I do not deem it so very necessary to write at greater length.’ Cic. Fam. 10.25.3)
Suppplement:
Audistis profecto dici philosophos Epicureos omnis res quae sint homini expetendae
voluptate metiri. (Cic. Pis. 68); Ex quibus hoc iis qui aderant—nam magno conventu
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In the following examples, the accusative and infinitive clauses have a perfect infini-
tive. Just as with finite clauses (see § 7.109), there are two possibilities for the tense in
the second or lower degree subordinate clause, depending on how the perfect
infinitive is interpreted. This is illustrated by (j) and (k). In (j) the sentence in direct
speech in the right-hand column has an imperfect tense in the subordinate clause,
which is then retained when it depends on the accusative and infinitive clause in the
left-hand column. In (k), however, the perfect of the subordinate clause is related
independently to the time of speaking.
(k) Eius porro civitatis sapientissimum Sapientissimus Solon fuit, is qui leges
Solonem dicunt fuisse, eum qui leges scripsit.
quibus hodie quoque utuntur scripserit.
(‘They say the wisest of that State was Solon,
the man who drew up the laws which are still
in force among them at the present day.’ Cic.
S. Rosc. 70)
This independence of the second degree subordinate clauses holds for all sorts of
subordinate clauses, as is shown by the following examples. In (l)–(n) the main clause
has a present tense. In (l) and (m) the tense of the subordinate clause is related to the
time of speaking; in (n) it is related to the past, just as in (o) and (p) (which have an
imperfect in the main clause). The correspondence between the present tense in the
main clauses in (l) and (m) and the primary tense in the subordinate clauses is purely
accidental.
(l) Ecquem scis in Sicilia antea captum archipiratam qui non securi percussus
sit? [adnominal relative clause]
(‘Do you know of any pirate-captain who was captured in Sicily before now who was
not beheaded?’ Cic. Ver. 5.67)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(m) Hanc pecuniam . . . abs te sic laceratam esse dico ut possim illud probare, si
velim, omnem te hanc pecuniam domum tuam avertisse. [result clause]
(‘Upon this sum of money, I assert, such inroads were made by you that I could, if
I chose, make my hearers believe that you diverted the whole of it into your coffers.’
Cic. Ver. 3.164)
(n) Rem navalem primum ita dico esse administratam, non uti provincia defen-
deretur, sed uti classis nomine pecunia quaereretur. [two result clauses]
(‘In the first place, I say that the naval affairs were managed, not with a view of
defending the province, but of acquiring money under pretence of providing a fleet.’
Cic. Ver. 5.43)
(o) Cominii dicebant idem quod ego dico, Staienum ab Oppianico pecuniam
accepisse ut iudicium corrumperet. [purpose clause]
(‘The Cominii said the same thing that I say now, that Staienus received money from
Oppianicus to induce him to corrupt the tribunal.’ Cic. Clu. 100)
(p) Brutum autem, hominem sapientem, quod filii nequitiam videret, quae
praedia ei relinqueret testificari voluisse dicebat. [reason clause + indirect
question]
(‘He said that Brutus, a wise man, because he saw his son’s profligacy, wished to leave
a record behind him of what farms he left him.’ Cic. Clu. 141)
Just as with first degree subordinate clauses, the choice between a primary and a
historic tense in one of the second degree with a present or future main clause seems
to be determined—at least in Cicero—by the type of clause. Only in argument clauses
and purpose clauses, and in temporal clauses that would have a historic tense in less
complex sentences, is the historic sequence predominant. A few illustrative instances
are given below. Most instructive is the variation in tense between the cum clause and
the simul ac clause in (t). A historic tense is regular in a temporal cum clause of this
type (see § 7.125). With simul ac a perfect indicative is normal. Further examples with
a governing present or future tense (or a verb form interpreted as such) are given in
the Supplement. They are almost all taken from Cicero.261
(q) Laudationem Porciae gaudeo me ante dedisse Leptae tabellario quam tuas
acceperim litteras.
(‘I’m glad I gave Porcia’s eulogy to Lepta’s courier before I received your letter.’ Cic.
Att. 13.37.3)
(r) Profecto intelligetis illinc ab initio cupiditatem pugnasse et audaciam, hinc
veritatem et pudorem quoad potuerit restitisse.
(‘You will see, in truth, that on the one side there were engaged from the very
beginning covetousness and audacity, that on the other side truth and modesty
resisted as long as they could.’ Cic. Quinct. 79)
261
The material for Cicero can be found in Lebreton (1901a: 161–73).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(s) Tu vero quid habes, quare putes, si paulo ante cor fuerit in tauro opimo,
subito id in ipsa immolatione interisse?
(‘But what reason do you have why you think, if only a little while before the heart was in
the sacrificial bull, it suddenly disappeared at the moment of immolation?’ Cic. Div. 2.37)
(t) Memini enim Antiochum ipsum, cum annos multos alia sensisset, simul ac
visum sit sententia destitisse.
(‘For I remember that Antiochus himself, after he had held other views for many
years, abandoned his opinion as soon as he saw fit.’ Cic. Luc. 2.63)
Supplement:
Primary sequence:
Argument clause: Quamquam intellego in nostra civitate inveterasse iam bonis
temporibus ut splendor aedilitatum ab optimis viris postuletur. (Cic. Off. 2.57);
Illud vehementer admiror non scripsisse te mihi manendo in provincia an ducendo
exercitum in Italiam rei publicae magis satis facere possim. (Pol. Fam. 10.31.6); Unde
existimant accidisse ut qui diligentissimi artium scriptores extiterint ab eloquentia
longissime afuerint. (Quint. Inst. 8.pr.3);
Autonomous relative clause: Id erit profecto, credamus hoc: Sex. Naevium cuius
caput oppugnet, eius auribus pepercisse. (Cic. Quinct. 40);
Temporal clause: . . . confitere te, cum ‘alio die’ dixeris, sobrium non fuisse. (Cic. Phil.
2.84); Hoc video, dum breviter voluerim dicere, dictum a me esse paulo obscurius,
sed experiar et dicam, si potero, planius. (Cic. de Orat. 1.187); Quem enim ardorem
studii censetis fuisse in Archimede, qui dum in pulvere quaedam describit attentius,
ne patriam <quidem> captam esse senserit? (Cic. Fin. 5.50); Nam deos hominesque
amicitiamque nostram testificor me tibi praedixisse neque temere monuisse sed,
postquam Caesarem convenerim sententiamque eius qualis futura esset parta vic-
toria cognorim, te certiorem fecisse. (Cael Fam. 8.16.1);
Result clause: Ecquem tu horum qui adsunt . . . aut tam sceleratum statuis fuisse ut
haec omnia perire voluerit, aut tam miserum ut et se perire cuperet et nihil haberet,
quod salvum esse vellet? (Cic. Sul. 32) (NB: coordination of primary and historic
sequence); Verum ex ista reperietis hereditate ita istum praedatum ut perpauca
occulte fecerit . . . (Cic. Ver. 2.45);
Purpose clause: . . . intellego tempus hoc vobis divinitus datum esse, ut odio invidia
infamia turpitudine totum ordinem liberetis. (Cic. Ver. 43);
Reason clause: Ego vero fateor me, quod viderim mihi auxilium non deesse, idcirco
me illi auxilio pepercisse. (Cic. Planc. 86);
Concessive clause: Itaque huic loco primum respondeo me, quamquam iustis de
causis rationes referre properarim, tamen te exspectaturum fuisse nisi in provincia
relictas rationes pro <re>latis haberem. (Cic. Fam. 5.20.2);
Comparative clause: Vides tamen omnia fere contra ac dicta sint evenisse. (Cic. Div. 2.53);
Conditional clause: Hic tibi ego de testimoniis meis hoc respondeo, si falsum
dixerim, te in eosdem dixisse. (Cic. Sul. 21);
Historic sequence:
Argument clause: (sc. plebem) . . . eam quo tutior esset vestra vita, se fecisse
commemorat ut non modo virtute flecteret, sed etiam tribus suis patrimoniis
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
deleniret . . . (Cic. Mil. 95); Iam illud mihi quidem mirum videri solet tot homines tam
ingeniosos post tot annos etiam nunc statuere non potuisse utrum ‘diem tertium’ an
‘perendinum’, ‘iudicem’ an ‘arbitrum’, ‘rem’ an ‘litem’ dici oporteret. (Cic. Mur. 27);
Adnominal relative clause: Eandemque diem intellego . . . unoque tempore in hac
re publica duos civis exstitisse, quorum alter finis vestri imperi non terrae, sed
caeli regionibus terminaret, alter eiusdem imperi domicilium sedisque servaret.
(Cic. Catil. 3.26);
Autonomous relative clause: Qui ad vos ab exteris nationibus venirent captos querar,
cum legati populi Romani redempti sint? (Cic. Man. 32); Ut verum videretur in hoc
illud quod Demosthenem ferunt ei qui quaesivisset quid primum esset in dicendo,
actionem, quid secundum, idem et idem tertium respondisse. (Cic. Brut. 142);
Temporal clause: Memoria teneo Milesiam quandam mulierem, cum essem in
Asia, . . . rei capitalis esse damnatam. (Cic. Clu. 32); Exempli causa ponam unum
illud, C. Getam, cum a L. Metello et Cn. Domitio censoribus ex senatu eiectus esset,
censorem esse ipsum postea factum . . . (Cic. Clu. 119);
Reason clause: . . . confitere huc ea spe venisse quod putares hic latrocinium, non
iudicium futurum. (Cic. S. Rosc. 61);
Concessive clause: Legem Liciniam et Muciam de civibus regundis video constare
inter omnes, quamquam duo consules omnium quos nos vidimus sapientissimi
tulissent, non modo inutilem sed perniciosam rei publicae fuisse. (Cic. fr. 10);
This section deals with the use of the tenses in subordinate clauses depending on
accusative and infinitive clauses in indirect speech and with the use of the tenses in
imperative and interrogative sentences in indirect speech. The most conspicuous
phenomenon in indirect speech is the use of the present tense for past events.
(b) ( . . . ) Futurum esse paucis annis uti omnes ex Galliae finibus pellerentur
atque omnes Germani Rhenum transirent.
(‘The consequence would be, that in a few years they would all be driven from
the territories of Gaul, and all the Germans would cross the Rhine.’ Caes. Gal.
1.31.11)
(c) ( . . . ) Nullam exoriri moram posse quominus cum venisset, si imperata non
facerent ad nutum e vestigio diriperentur.
(‘Nothing is able to arise to prevent their being plundered immediately upon his
arrival if they do not carry out orders at his command.’ Caes. Civ. 2.12.3)
Ex. (d) is completely different from (b), although both have been taken from the same
speech of Diviciacus in Caes. Gal. 1.31; instead of imperfect and pluperfect forms,
Caesar uses present and perfect ones. The same goes for (e) and (f ) from Livy and
Tacitus.
(d) Nisi si quid in Caesare populoque Romano sit auxilii, omnibus Gallis idem
esse faciendum quod Helvetii fecerint, ut domo emigrent, aliud domicilium,
alias sedes, remotas a Germanis, petant fortunamque, quaecumque accidat,
experiantur.
(‘Unless there was some aid in Caesar and the Roman people, the Gauls must all do
the same thing that the Helvetii have done, emigrate from their country, and seek
another dwelling place, other settlements remote from the Germans, and try what-
ever fortune may fall to their lot.’ Caes. Gal. 1.31.14)
(e) Cum . . . patres censuissent, praetor Samnitibus respondit. Nec quo minus
perpetua cum eis amicitia esset per populum Romanum stetisse, nec contra-
dici quin, quoniam ipsos belli culpa sua contracti taedium ceperit, amicitia
de integro reconcilietur. Quod ad Sidicinos attineat, nihil intercedi quo
minus Samniti populo pacis bellique liberum arbitrium sit.
(‘After . . . the senate voted, the praetor returned this answer to the Samnites: It
neither had been the fault of the Roman people that their friendship with them
was not perpetual; nor was any objection made to that friendship being once more
re-established, since they themselves were now become tired of a war entered into
through their own fault. With respect to what regarded the Sidicinians, they did not
interfere with the Samnite nation having the free decision of peace and war.’ Liv.
8.2.1–3)
(f ) Praetendebat lascivire militem diductum. Si quid subitum ingruat, maiore
auxilio pariter subveniri.
(‘His pretext was that discipline suffered when troops were dispersed. If a sudden
peril threatened the support would be more effective if the soldiers supported all
together.’ Tac. Ann. 4.2.1—using Martin and Woodman ad loc.)
This use of the primary tenses is essentially different from the use of the present
and perfect tenses that was discussed in the preceding section: the events described
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
in the subordinate clauses definitely belong to the past. The phenomenon therefore
resembles the use of the historic present to refer to past events. This stylistic feature is
sometimes called ‘represented speech’ or repraesentatio. When used in this way the
tenses in the subordinate clauses reflect more precisely the tenses as they were actually
used in direct speech. This usage is attested only once in a reported speech in Plautus
(see ex. (a) in § 7.116) and is absent from Cicero, but it is common in Caesar, Livy,
Tacitus, and Ammianus and is also found outside historical prose.262 It is found both
in argument clauses and in satellite clauses (see also § 7.97 above for repraesentatio in
less complex sentences).
Whereas Caesar shifts between historic and primary sequence without a clear
reason, Livy maintains the primary sequence wherever a subjunctive form with the
same tense value is available. Obviously this is excluded in the case of imperative
forms and simple and future perfect forms (there are no future subjunctive forms).
Accordingly, these forms are most often found as past forms in indirect speech.
Primary sequence is also excluded when a historic tense is required for semantic
reasons. At the beginning of a passage of indirect speech, when the subordinate
clauses are still directly depending on the main clause, Livy seems to prefer past
forms; this indicates that these forms function as signals for the interpretation of the
present and perfect ones to follow. A similar strategy seems to apply when historic
presents are being used (see § 7.16).263 Tacitus’ usage, in turn, is different again; he
seems above all to try to achieve variation. Three longer passages are given below
from Caesar, Livy, and Tacitus to illustrate the phenomenon. The tenses in the finite
subordinate clauses are marked, the historic ones in bold, the primary ones in italics.
In the left-hand column is the transmitted text. The right-hand columns of the
passages of Caesar and Livy have a transposition into direct speech. It is interesting
to note where and how often primary tenses in direct speech are put into historic
sequence in the indirect speech as we have it. Note that historic sequence in direct
speech is maintained in indirect speech.
(g)
His Caesar ita respondit. ‘Eo sibi minus His Caesar ita respondit. ‘Eo mihi minus
dubitationis dari, quod eas res quas legati dubitationis datur, quod eas res quas
Helvetii commemorassent memoria commemoravistis memoria teneo, atque eo
teneret, atque eo gravius ferre quo minus gravius fero quo minus merito populi
merito populi Romani accidissent. 2 Qui si Romani acciderunt. 2 Si alicuius iniuriae
alicuius iniuriae sibi conscius fuisset, non mihi conscius fuissem, non fuisset difficile
fuisse difficile cavere. Sed eo deceptum, cavere. Sed eo deceptus sum, quod neque
quod neque commissum a se intellegeret commissum a me intellego quare timeam
quare timeret neque sine causa timendum neque sine causa timendum puto. 3 Quod si
262
For instances in Ovid, see Bömer ad Ov. Met. 10.30.
263
See Andrewes (1937, 1951) on Caesar, Livy, and Tacitus; Conway’s appendix to his commentary on
book II of Livy (1901) and Briscoe’s refinements of Conway’s observations (2008: appendix 1, 2012:
appendix 1); Hyart (1960) on Caesar; Lambert (1946) on Livy.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(‘To these remarks Caesar replied as follows: As he remembered well the events which the
Helvetian deputies had mentioned, he had therefore the less need to hesitate; and his
indignation was the more vehement in proportion as the Roman people had not deserved
the misfortune. If the Romans had been conscious of some outrage done, it would not have
been hard to take precaution; but they had been misled, because they did not understand that
they had done anything to cause them apprehension, and they thought that they should not
feel apprehension without cause. And even if he were willing to forget an old affront, could
he banish the memory of recent outrages—their attempts to march by force against his will
through the Province, their ill-treatment of the Aedui, the Ambarri, the Allobroges? Their
insolent boast of their own victory, their surprise that their outrages had gone on so long
with impunity, pointed the same way; for it was the wont of the immortal gods to grant
a temporary prosperity and a longer impunity to make men whom they purposed to
punish for their crime smart the more severely from a change of fortune. Yet, for all this,
he would make peace with the Helvetii, if they would offer him hostages to show him that
they would perform their promises, and if they would give satisfaction to the Aedui in
respect of the outrages inflicted on them and their allies, and likewise to the Allobroges.
Divico replied: It was the ancestral practice and the regular custom of the Helvetii to receive,
not to offer, hostages; the Roman people was witness thereof. With this reply he departed.’
Caes. Gal. 1.14)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(h) 264
Ab hominibus nobilibus, per multos annos Ab hominibus nobilibus, per multos annos
bellum quaerentibus, Hannibalem in bellum quaerentibus, Hannibal in Italiam
Italiam adductum. Ab iisdem, cum debellari adductus est. Ab iisdem, cum debellari
possit, fraude bellum trahi. Cum quattuor possit, fraude bellum trahitur. Cum
legionibus universis pugnari posse quattuor legionibus universis pugnari posse
apparuisset eo quod M. Minucius absente apparuisset eo quod M. Minucius absente
Fabio prospere pugnasset, duas legiones Fabio prospere pugnasset, duae legiones
hosti ad caedem obiectas, deinde ex ipsa hosti ad caedem obiectae sunt, deinde ex
caede ereptas ut pater patronusque ipsa clade ereptae ut pater patronusque
appellaretur qui prius vincere prohibuisset appellaretur qui prius vincere prohibuisset
Romanos quam vinci. Consules deinde Romanos quam vinci. Consules deinde
Fabianis artibus, cum debellare possent, Fabianis artibus, cum debellare possent,
bellum traxisse. Id foedus inter omnes bellum traxere. Id foedus inter omnes
nobiles ictum nec finem ante belli habituros nobiles ictum est nec finem ante belli
quam consulem vere plebeium, id est, habebitis quam consulem vere plebeium, id
hominem novum fecissent. Nam plebeios est, hominem novum feceritis. Nam plebeii
nobiles iam eisdem initiatos esse sacris et nobiles iam eisdem initiati sunt sacris et
contemnere plebem, ex quo contemni contemnere plebem, ex quo contemni
patribus desierint, coepisse. Cui non patribus desierunt, coepere. Cui non
apparere id actum et quaesitum esse ut apparet id actum et quaesitum esse ut
interregnum iniretur, ut in patrum interregnum iniretur, ut in patrum
potestate comitia essent? Id consules ambos potestate comitia essent? Id consules ambos
ad exercitum morando quaesisse. Id postea, ad exercitum morando quaesiere. Id postea,
quia invitis iis dictator esset dictus quia invitis iis dictator erat dictus
comitiorum causa, expugnatum esse ut comitiorum causa, expugnatum est ut
vitiosus dictator per augures fieret. Habere vitiosus dictator per augures fieret. Habent
igitur interregnum eos. Consulatum unum igitur interregnum ei. Consulatus unus certe
certe plebis Romanae esse. Populum plebis Romanae est. Populus liberum
liberum habiturum ac daturum ei qui habebit ac dabit ei qui [magis] vere vincere
[magis] vere vincere quam diu imperare quam diu imperare mavolt.
malit.
(‘The nobles, he said, had been seeking war for many years, and it was they who had brought
Hannibal into Italy. It was their machinations, too, that were spinning out the war, when it
might be brought to a victorious conclusion. That four legions if united were able to hold their
own in a general engagement had been shown in a successful battle fought by Marcus
Minucius, when Fabius was absent. Notwithstanding this, two legions had just been exposed
to be massacred by the enemy and subsequently rescued from the massacre, to the end that the
names of Father and Patron might be conferred on one who had kept the Romans from
conquering before keeping them from being conquered. After that the consuls had employed
the arts of Fabius to prolong the war, when they were able to have ended it. The nobles had all
made a compact to this effect; nor would his hearers see an end of the war until they had
264
This and the following parallel texts are taken from Conway’s commentary (1901: 191–3).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
elected a true plebeian, a new man, to the consulship; for the plebeian nobles had already been
admitted to the same rites as the others and had begun to look down on the plebs from the
moment when they themselves had ceased to be looked down on by the patricians. Who could
fail to see that their end and purpose in resorting to an interregnum had been to keep the
election in the hands of the patricians? To this end both consuls had remained with their army
in the field; to this end, later on, because a dictator had been named, against their wishes, for
the purpose of holding an election, they had succeeded in having the augurs declare that there
had been a flaw in his appointment. They had therefore the interregnum they desired. But at
least one consulship belonged to the Roman plebs; and the people meant to keep it free, and
bestow it on him who would rather win an early victory than remain long in command.’ Liv.
22.34.4–11)
The differences between the two texts are quite remarkable, as Table 7.20 shows.
Historic sequence is used throughout in the first part of Caesar’s speech; then there is
a shift from historic to primary sequence in § 5. It is difficult to tell what triggers it.
Maybe it is the general truth of the statements about the gods. In Livy, on the other
hand, the situation is entirely different. Historic tenses in the indirect speech corres-
pond to historic ones in the parallel direct speech, and primary tenses correspond to
one another in the same way. The only shift from a primary to a secondary tense is the
pluperfect subjunctive fecissent, which in direct speech is in the future perfect
(feceritis). In this case a simple shift to the subjunctive was impossible, because that
form does not exist. In Table 7.20 a distinction is made between ‘convertible’ and
‘non-convertible’ primary tenses. The former are the tenses used in direct speech for
which a subjunctive parallel exists; the latter comprise the simple future, the future
perfect, and the imperative forms.
Table 7.20 Comparison of primary and historic tenses in direct and indirect speech
in Caesar and Livy
Tense Direct speech Indirect speech
Caesar Livy Caesar Livy
Primary Historic Primary Historic
Conv. primary 20 3 13 7 3
Non-conv. primary 1 1
Historic 2 9 2 9
detraxerit in direct speech, but it would be strange for Tiberius to use an anterior
indicative form. Erumpere and convelli towards the end are used instead of eruptum
iri and convulsum iri (see for this Tacitean usage § 7.69). The passage is continued in
direct speech (for this type of variation see Chapter 24).
(i)
(‘In reply Tiberius . . . wrote again: “With other men, the standpoint for their deci-
sions was what was in their own interests: the lot of princes was very different, as
their weightiest affairs had to be regulated with an eye upon public opinion. There-
fore he did not take refuge in the answer which came most readily to the pen—that
Livia could determine for herself whether she ought to marry after Drusus or rest
content with her old home, and that she had a mother and grandmother who
were more natural advisers. He would deal more openly: and first with regard to
Agrippina’s enmity, which would blaze out far more fiercely if Livia’s marriage
divided, as it were, the Caesarian house into two camps. Even as matters stood,
there were outbreaks of feminine jealousy, and the feud was unsettling his grand-
children. What then if the strife was accentuated by the proposed union?” ’ Tac. Ann.
4.40.1–3)
the corresponding direct speech. The negation word is ne, as normal. See (c).
If there is a series of commands/wishes, the first may be introduced by ut(i), as
in (d).265
(a)
‘Eos inter se, quia nemo unus satis dignus Ei inter se, quia nemo unus satis dignus
regno visus sit, partes regni rapuisse. Bona regno visus est, partes regni rapuerunt. Bona
sua diripienda populo dedisse, ne quis mea diripienda populo dederunt, ne quis
expers sceleris esset. Patriam se regnumque expers sceleris esset. Patriam regnumque
suum repetere et persequi ingratos cives meum repetere et persequi ingratos cives
velle. Ferrent opem, adiuvarent. Suas volo. Ferte opem, adiuvate. Vestras quoque
quoque veteres iniurias ultum irent, totiens veteres iniurias ultum ite, totiens caesas
caesas legiones, agrum ademptum.’ legiones, agrum ademptum.
(‘His enemies, perceiving that no single claimant was fit to be king, had seized and usurped the
power amongst themselves, and had given up his goods to be plundered by the people, that
none might be without a share in the guilt. He wished to regain his country and his
sovereignty, and to punish the ungrateful Romans. Let them succour and support him, and
avenge, as well, their own long-standing grievances, the oft-repeated destruction of their
armies, and seizure of their lands.’ Liv. 2.6.3)
(c)
However, the stylistic option of using the present subjunctive for commands and
wishes of the past is also quite frequent from Caesar onwards. This is illustrated by
(d) and (e) (these examples are also used in § 7.62).
265
For the use of ut in such imperative sentences, see Sznajder (2001).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(d) . . . inviti et coacti Scipionis sententiam sequuntur: uti ante certam diem
Caesar exercitum dimittat; si non faciat eum adversus rem publicam fac-
turum videri.
(‘ . . . against their will and yielding to pressure, they adopt the proposal of Scipio that
Caesar should disband his army before a fixed date, and that, if he failed to do so, he
would be judged to be committing an act hostile to the republic.’ Caes. Civ. 1.2.6)
(e) Cicero ad haec unum modo respondit: ‘Non esse consuetudinem populi
Romani ullam accipere ab hoste armato condicionem. Si ab armis discedere
velint, se adiutore utantur legatosque ad Caesarem mittant. Sperare pro eius
iustitia quae petierint impetraturos.’
(‘To this Cicero made one remark only in reply: that it was not the practice of the
Roman people to accept terms from an enemy in arms; if they would lay down their
arms, they should use his good offices and send deputies to Caesar; he hoped that,
having regard to the justice of Caesar, they would obtain their petition.’ Caes. Gal.
5.41.7–8)
Supplement:
Present subjunctive: (Vercingetorix said:) . . . perfacile esse factu frumentationibus
pabulationibusque Romanos prohibere. Aequo modo animo sua ipsi frumenta cor-
rumpant aedificiaque incendant . . . (Caes. Gal. 7.64.2–3);
Imperfect subjunctive: Responsum ex decreto est optare pacem Rhodios. Si bellum
esset, ne quid ab Rhodiis speraret aut peteret rex quod veterem amicitiam . . .
diiungeret sibi ac Romanis. (Liv. 42.46.6);
(d) Iam de deorum immortalium templis spoliatis in capta urbe qualem calum-
niam ad pontifices attulerit?
(‘What sort of malicious story did he tell the pontiffs about the spoliation of
the temples of the gods in the captured city?’ Liv. 39.4.11)
(e) Qua sapientia, quibus philosophorum praeceptis intra quadriennium regiae
amicitiae ter milies sestertium paravisset?
(‘By what kind of wisdom or maxims of philosophy had he within four
years of royal favour amassed three hundred million sesterces?’ Tac. Ann.
13.42.4)
266
See K.-St.: II.541. More instances are mentioned in Koestermann’s commentary ad Tac. Ann.
13.42.4 (ex. e), including Caes. Civ. 1.32.1, quoted above in the Supplement.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
267
Lebreton (1901a: 273–7).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
considerations are taken into account. The indicative is primarily a main clause
mood, the subjunctive a subordinate clause mood (see Tables 7.7 and 7.8 in § 7.11).
Moreover, in the domain of subordinate clauses there are large areas where the
indicative and subjunctive do not compete. An area where they are real competitors
are relative clauses (but not all types of relative clauses—see § 18.26 on relative clauses
of purpose). An area where they are not real competitors are argument clauses.
Although there are quod and quia argument clauses with an indicative, they are—at
least in the Classical period268—limited to a few small classes of verbs such as emotion
verbs, e.g. miror quod ‘to be surprised that’ (see § 15.8). Compared with the enormous
number of verbs governing ut, ne, quin, and other subjunctive clauses, e.g. impero ut
‘to order that’, timeo ne ‘to be afraid that’, and non dubito quin ‘to have no doubt that’,
the number of those governing an indicative clause is almost negligible.269 In add-
ition, as is shown in § 7.95 above, some of the subjunctive subordinate clauses are
much more integrated into their matrix clause than the indicative ones. In particular,
the regular mood of subordinate clauses in indirect speech is the subjunctive, and this
may be another factor why the sequence of tense rules are particularly relevant to the
subjunctive. All in all, it seems understandable that subjunctive subordinate clauses
whose tense is relative to that of the main clause are much more frequent than
indicative ones.
268
For the expansion of quod and quia to other classes in Christian authors, see Blaise (1955: 147–52).
269
At a rough estimate the proportion between indicative and subjunctive argument clauses in
Classical texts is about 1 : 25.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Itaque quae erant prudentiae propria suo loco dicta sunt. Quae autem harum
virtutum (sc. sunt) de quibus iam diu loquimur, quae pertinent ad verecun-
diam et ad eorum approbationem quibuscum vivimus, nunc dicenda sunt.
(‘Accordingly, the properties which, as we found, were peculiar to prudence were
discussed in their proper place. Now those properties must be discussed which are
peculiar to these virtues of which we have for some time been speaking, which relate
to considerateness and to approbation of our fellow-men.’ Cic. Off. 1.143)
(b) (sc. naturam) Quam si sequemur ducem nunquam aberrabimus sequemur-
que et id quod acutum et perspicax natura est . . .
(‘If we follow nature as our guide, we shall never go astray, but we shall be pursuing
that which is in its nature clear-sighted and penetrating . . . ’ Cic. Off. 1.100)
(c) Nam si quotienscumque praeteritus erit is qui non debuerit praeteriri,
totiens oportebit eum qui factus erit condemnari . . .
(‘For if a person who has been elected is to be condemned upon all occasions when
someone who should not have been passed over is passed over . . . ’ Cic. Planc. 14)
(d) Is in divitias homo adoptavit hunc quom diem obiit suom.
(‘He adopted him and made him his heir when he passed away.’ Pl. Poen. 904)
(e) De te autem, Catilina, cum quiescunt, probant, cum patiuntur, decernunt,
cum tacent, clamant . . .
(‘In your case, however, Catiline, while they are quiet they approve, while they permit
me to speak they decide, when they are silent they are loud and eloquent . . . ’ Cic.
Catil. 1.21)
(f ) Dum enim in una virtute . . . omnia esse voluerunt . . . virtutem ipsam, quam
amplexabantur, sustulerunt.
(‘While they wished all things to reside in one virtue, they destroyed the very virtue
which they embraced.’ Cic. Fin. 2.43)
However, these are subjective choices, and the above holds true especially for Cicero
(whose usage has also been studied most intensively). There are many ‘exceptions’,
and one must bear this in mind when considering some of the further illustrations in
the Supplement.270
Supplement:
Simple future: Ut voles, ut tibi lubebit, nobis legem inponito. (Pl. As. 239); Mane,
mane. Iam ut voles med esse ita ero. (Pl. Ps. 240); Qui ager frigidior et macrior erit,
ibi oleam licinianam seri oportet. (Cato Agr. 6.2); Ager oleto conserundo, qui in
ventum favonium spectabit et soli ostentus erit. Alius bonus nullus erit. (Cato Agr.
8.2); Per ver serito in loco ubi terra tenerrima erit (quam ‘pullam’ vocant), ubi aqua
propter siet. (Cato Agr. 151.2); Qui ad dignitatem impellit, maiorum exempla quae
erunt vel cum periculo gloriosa colliget, posteritatis immortalem memoriam augebit.
270
Further examples and references in K.-St.: I.151–6.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(Cic. de Orat. 2.335); Sit modo is, qui dicet aut scribet, et institutus liberaliter
educatione doctrinaque puerili . . . (Cic. de Orat. 3.125); De qua quoniam in Catone
maiore satis multa diximus illinc assumes quae ad hunc locum pertinebunt. (Cic. Off.
1.151); Qui igitur adipisci veram gloriam volet iustitiae fungatur officiis. (Cic. Off.
2.43); Si Lentuli navis non erit, quo tibi placebit imponito. (Cic. Att. 1.8.2);
Imperfect: . . . hoc melius ea fortasse quae erant vera cernebant. (Cic. Tusc. 1.26);
Bene igitur nostri, cum omnia essent in moribus vitia, quod nullum erat iracundia
foedius, iracundos solos morosos nominaverunt. (Cic. Tusc. 4.54); Ipse cum omnibus
copiis in Morinos proficiscitur, quod inde erat brevissimus in Britanniam traiectus.
(Caes. Gal. 4.21.3);
Iterative anteriority: Messanam ut quisque nostrum venerat, haec visere solebat.
(Cic. Ver. 4.5); Cum autem ver esse coeperat . . . dabat se labori atque itineribus. (Cic.
Ver. 5.27); In qua permaneo, Catule, sententia meque, cum huc veni, hoc ipsum nihil
agere et plane cessare delectat. (Cic. de Orat. 2.24); Quae (sc. terra) cum gremio
mollito ac subacto sparsum semen excepit, primum id occaecatum cohibet . . . (Cic.
Sen. 51); Nam ut quisque hostem inpigre occiderat, primum capite aegre inter
turbam tumultumque abscidendo terebat tempus. (Liv. 24.15.4); Talia eduntur
tonitrua cum conglobata nubes dissolvitur, et eum quo distenta fuerat spiritum
emisit. (Sen. Nat. 2.27.3) (NB: variation in tense in parallel clauses);
Coinciding events: Venus multipotens, bona multa mihi / dedisti, huius quom
copiam mi dedisti. (Pl. Cas. 841–2); Nam quid emebat cum te emebat? Ut ad se
transires? (Cic. Flac. 83); Is enim profecto mortem attulit qui causa mortis fuit.
(Cic. Phil. 9.7); Exspectationem nobis non parvam attuleras cum scripseras Var-
ronem tibi pro amicitia confirmasse causam nostram Pompeium certe
suscepturum . . . (Cic. Att. 3.18.1); Coinciding events with modal verbs: Ac si,
pontifices, neque is cui licuit, neque id quod fas fuit dedicavit, quid me attinet . . .
(Cic. Dom. 138); . . . quos voluit expulit, quos potuit occidit. (Cic. Phil. 14.23); Eum,
ut potui, per litteras cohortatus sum gratiasque egi, ut debui. (Cic. Fam.
14.2.3); . . . et eo magis quod Hermia, quem eodem die venire oportuerat, non
venerat. (Cic. Fam. 16.15.1);
271
See Sz.: 552.
272
See Hoppe (1903: 67–70 (= 1985: 133–8)) and Blaise (1955: 181–2).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
273
Sznajder (1996: 310) reports that in the corpus she has examined slightly more than 2% of the
historic presents are in subordinate clauses. For the exceptional use of the present in an ut-argument
clause, see § 7.97, (g) and (h).
274
Sz.: 621.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Ceterum ubi deserta munimenta nec facile se tantum praegressos adsecuturum videt,
ad mare ac naves rediit. (Liv. 21.32.2);
Discubitum noctu ut imus, ecce ad me advenit / mulier, qua mulier alia nulla est
pulchrior. (Pl. Mer. 100–1);
The historic present is also used in other types of subordinate clauses, notably reason,
concessive, and relative ones, especially in the Early Latin comedies, in Livy and in the
historical narrative of poets like Virgil. Examples are (g) reason, (h) concessive, and
(i) relative.
(g) <De> navi timidae ambae in scapham insiluimus, quia videmus / ad saxa
navem ferrier.
(‘In our terror, we both jumped down from the ship onto its boat because we saw that
the ship was being carried to the cliffs.’ Pl. Rud. 366–7)
(h) Quamquam probe scit . . . infamem suam cunctationem esse, obstinatus
tamen . . . aestatis reliquum extraxit . . .
(‘Though he knew full well that his policy of waiting was in bad repute . . . he
doggedly . . . dragged out the remainder of the summer.’ Liv. 22.15.1)
(i) Sub tecto miles dextra ac sinistra muro tectus adversus plutei obiectu operi
quaecumque sunt usui sine periculo subportat.
(‘Under this cover the soldiers, sheltered to right and left by the wall, in front by the defence
of a screen, bring up without danger whatever is of use for the work.’ Caes. Civ. 2.15.3)
Supplement:
Reason clauses: Quoniam videmus auro insidias fieri, / capimus consilium continuo.
(Pl. Bac. 299–300); Ind’ nunc aufugit, quoniam capitur oppidum. (Pl. Poen. 665);
Concessive clauses: . . . etsi probe scit cui parti Charopo principe excepto Epirotae
favissent, tamen quia ab satisfaciendi quoque cura imperata enixe facere videt, ex
praesenti eos potius quam ex praeterito aestimat habitu . . . (Liv. 32.14.5–6);
Result clauses: Quam rem ita graviter tulit Alexander ut secunda legatione denuo
bellum deprecantibus ita demum remiserit ut oratores et duces, quorum fiducia
totiens rebellent, sibi dedantur. (Justin. 11.4.10);
Relative clauses: . . . castella expugnat, magnas praedas capit, quarum partim suis
dispertit, partim ad Datamen mittit. (Nep. Dat. 10.2); Ei mihi qualis erat, quantum
mutatus ab illo / Hectore, qui redit exuvias indutus Achilli . . . (Verg. A. 2.274–5);
Euryalus phaleras Rhamnetis et aurea bullis / cingula, Tiburti Remulo ditissimus
olim / quae mittit dona . . . (Verg. A. 9.359–61); Iam perventum ad suos . . . erat, qui
adventu exercitus imperatorisque pugnam renovant . . . . (Liv. 33.8.6);
Indirect questions: Ea quemadmodum ad istum postea per pseudothyrum rever-
tantur, tabulis vobis et testibus, iudices, planum faciam. (Cic. Ver. 2.50);275
Argument clauses: Accessit huc quod postridie eius absolutionem in theatrum Curionis
Hortensius introit, u<t> puto, ut suum gaudium gauderemus. (Cael. Fam. 8.2.1);
275
K.-St.: I.115 quote this example as a ‘curious instance of a historic present in a subjunctive
subordinate clause’.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Nam postquam haec aedis ita erant ut dixi tibi, / continuo est alias aedis
mercatus sibi.
(‘Well, after our house was the way I told you, he immediately bought himself
another.’ Pl. Mos. 647–8)
(b) Postquam structi utrimque stabant, cum paucis procerum in medium duces
procedunt.
(‘When heaps (of arms) were standing on both sides, the leaders, attended by a few of
their nobles, advanced to the middle of the field.’ Liv. 1.23.6)
Supplement:
Postea quam omnis actio huius anni confecta nobis videbatur, in Asiam ire
nolui . . . (Cic. Att. 3.19.1); Labienus postquam neque aggeres neque fossae vim
hostium sustinere poterant . . . Caesarem per nuntios facit certiorem quid faciendum
existimet. (Caes. Gal. 7.87.3); Postquam una nocte et <parta> die ad oppidum
consumpta neque responsum ullum a Considio dabatur neque . . . non est visa ratio
ad oppugnandum oppidum commorandi . . . (B. Afr. 5.1); (sc. Samnitium exercitus)
Deinde, postquam nemo obvius ibat, infestis signis ad castra hostium succedit. (Liv.
7.37.7); Quibus paulum aegreque vitatis, postquam mutabat aestus eodemque quo
ventus ferebat, non adhaerere ancoris, non exhaurire inrumpentes undas poterant.
(Tac. Ann. 2.23.4); Nihil ex eo tempore praetermissum est ad maiestatem eius
augendam ac multo magis, postquam Agrippa abdicato atque seposito certum erat
uni spem successionis incumbere. (Suet. Tib. 15.2);
Quod ubi saepius eius equitatus facere non intermittebat . . . reliqui perterriti fuga
se in castra recipiunt. (B. Afr. 29.3); . . . et, ubi nemo obvius ibat, pleno gradu ad castra
hostium tendunt. (Liv. 9.45.14); Igitur tribuni, ut inpediendae rei nulla spes erat, de
proferendo exercitus exitu agere . . . (Liv. 3.20.6);
Of the examples of ubi and ut quoted in the literature (e.g. K.St.: II.361), many may
be interpreted as simultaneous. So, for example, Cic. Ver. 18 Nam ut Q. Hortensius
consul designatus domum reducebatur e campo . . . , fit obviam casu ei multitudini
C. Curio. Woodcock (1959: 176) translates as follows: ‘As Hortensius was being
escorted home from the campus, he was met by C. Curio.’ Further examples can be
found in §§ 16.23–16.24.
Some instances of the pluperfect in postquam clauses and clauses with the other
subordinators just mentioned are quite similar to the imperfect instances in that they
denote resulting states from previous actions, as is noted for terminative states of
affairs in § 7.31 in other contexts as well. An illustration of this is (c), where a
pluperfect and an imperfect are coordinated. This use of the pluperfect is different
from its anterior use, which is discussed in § 16.21.
(c) Postquam lege hac fabularum ab risu ac soluto ioco res avocabatur et ludus in
artem paulatim verterat, iuventus . . . ridicula intexta versibus iactitare coepit.
(‘When, by adopting this method in the presentation of pieces, the old farce and
loose jesting was given up and the play gradually became a work of art, the young
people . . . began to improvise comic verses.’ Liv. 7.2.11)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The same use of the imperfect (and pluperfect) can be observed in clauses with
priusquam and antequam ‘before’, especially when the main clause is negated or
negation is implied (and the temporal sequence is therefore the same as with post-
quam). Two of the rare instances are (d) and (e).
(d) Quid illi ex utero exitio est prius quam poterat ire in proelium?
(‘Why did he leave the womb before he could go into battle?’ Pl. Truc. 511)
(e) . . . nec prius quam recipi tuto signa non poterant, inminentem capiti hostem
vidit.
(‘ . . . nor, until they were unable to withdraw to safety, did he perceive that the enemy
was posted on the heights above him.’ Liv. 7.34.2)
Supplement:
Imperfect: . . . nec ante quam vires ad standum in muris ferendaque arma deerant,
expugnati sunt. (Liv. 23.30.4);
Pluperfect denoting a resulting state: Prius multo ante aedis stabam quam illo
adveneram. (Pl. Am. 603); . . . nec ante ad curandum corpus recessit, quam praeter-
ierat omne agmen . . . (Curt. 7.5.16);
Time clauses with the subordinator cum (quom) ‘when’ indicate at what time the
event of the main clause took place (so-called cum temporale or cum narrativum), as
is shown in (a).
(a) Quid ergo? Non aliquando in ira quoque . . . dimittunt incolumes . . . ? #
Faciunt. # Quando? # Cum adfectus repercussit adfectum . . .
(‘What then? Do not men sometimes even in the midst of anger send them away safe
. . . ? # They do. # But when? # When passion has beaten back passion . . . ’ Sen. Dial. 3.8.7)
Cum in this ‘locating’ sense can be combined with all indicative tenses. It sometimes
correlates with a temporal adverb in the main clause, as it does with nunc in (b) and
(c). Sometimes the states of affairs in the main and subordinate clauses more or less
coincide, as in (d) and (e) (so-called cum identicum). In some contexts the event
referred to in the cum clause is interpreted as habitual or iterative (‘whenever’), as in
(f ) and (g). (In these three contexts the use of the indicative remained regular until
the time of Livy, but gradually the subjunctive took its place—see § 7.142.)
(b) . . . qui modo nusquam comparebas, nunc, quom compares, peris.
(‘A moment ago you were nowhere to be seen, but now that you are to be seen you’re
finished!’ Pl. Aul. 629)
(c) Nunc quom vivum nescioquem istum producis, tamen te derideri vides.
(‘Even as it is, when you do produce your man, whoever he is, alive, you see that we
laugh at you.’ Cic. Ver. 5.79)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
276
The examples from Cicero can be found in Lebreton (1901a: 327–35).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
obsolescerent renovabam cum licebat legendo. (Cic. Ac. 1.11); Itaque cum
sumus necessariis negotiis curisque vacui tum avemus aliquid videre, audire,
addiscere . . . (Cic. Off. 1.13); Sed nescio quo modo, dum lego, adsentior, cum
posui librum et mecum ipse de inmortalitate animorum coepi cogitare, adsensio
omnis illa elabitur. (Cic. Tusc. 1.24); Nam et laqueis falces avertebant, quas cum
destinaverant, tormentis introrsus reducebant . . . (Caes. Gal. 7.22.2); Alii ruentes
in hostem undique configebantur, et cum comminus venerant, gladiis a velitibus
trucidabantur. (Liv. 38.21.11–12); . . . proceres civitatis in agris morabantur et,
cum consilium publicum desiderabatur, a villis arcessiebantur in senatum.
(Col. 1.pr.18);
Miscellaneous: . . . eademne erat haec disciplina tibi, quom tu adulescens eras? (Pl.
Bac. 421); Horrescet faxo lena leges quom audiet. (Pl. As. 749); Nam tu cum quaestor
es factus, etiam qui te numquam viderant, tamen illum honorem nomini mandabant
tuo. (Cic. Pis. 2); Et praeter eos quamvis enumeres multos licet, cum deni creantur,
non nullos in omni memoria reperies perniciosos tribunos . . . (Cic. Leg. 3.24); For-
mam igitur mihi totius rei publicae, si iam es Romae aut cum eris, velim mittas . . .
(Cic. Att. 6.3.4); Quid? Cum dabas iis litteras . . . non eos ad me venturos arbitrabare?
(Cic. Fam. 3.7.3);
Cum clauses with a tense form of the perfectum stem normally locate the state of
affairs of the subordinate clause at a time anterior to the state of affairs of the main
clause. This is clearly the case with the pluperfect and the future perfect in (h) and (i),
respectively. Relatively rare are sentences in which the perfect tense occurs in both the
cum clause and the following main clause and where, just as in a sequence of
independent sentences with the perfect tense, the states of affairs must be interpreted
as successive, that is, where the cum clauses must be interpreted as anterior with
respect to their main clause. This is sometimes called the ‘narrative’ use of cum (cum
narrativum). Two examples are (j) and (k). Much more typical in such a situation are
temporal ubi-, simulac-, and postquam clauses with a perfect indicative tense (see
§ 16.19–16.26) or cum clauses with a pluperfect subjunctive (see § 7.140). More often,
however, when both clauses have a perfect tense, the two events referred to are not
successive but coincide, as in (e).
(h) Cum iste . . . multa digna dederat quamobrem responsurus non videretur,
mentio de lege nulla fiebat.
(‘There was no talk of the bill at the time when Verres . . . had told us, by many
indications, why he meant to offer no defence.’ Cic. Ver. 5.178)
(i) Cum ego P. Granium testem produxero . . . refellito, si poteris.
(‘When I call for the evidence of Publius Granius . . . you shall prove him a liar if you
can.’ Cic. Ver. 5.154)
(j) Cum ad Clodium ventum est, cupiit diem consumere, neque ei finis est
factus.
(‘When it came to Clodius, he desired to finish the day, but the end was not made for
him.’ Cic. Att. 4.2.4)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(k) Haec cum facta sunt in consilio, magna spe et laetitia omnium discessum
est . . .
(‘When these things were done in council, a departure was made with the great hope
and happiness of all . . . ’ Caes. Civ. 3.87.7)
Supplement:
Sed cum ductus sum in ludum saltatorium, plus medius fidius in eo ludo vidi pueris
virginibusque quinquaginta . . . (Scip. Min. ap. Macrob. Sat. 3.14.7); Ita cum ad eum
locum venerunt iubet binos equites conscendere . . . (B. Hisp. 3.6—NB: historic
present in the main clause); Quos cum et a patribus conlaudari et a militari aetate
tamquam bonos cives conspici vulgus hominum vidit, repente spreto tribunicio
auxilio certamen conferendi est ortum. (Liv. 4.60.8); In questus enim flebiles, cum
sibimet ipsi consulere iussi sunt, sese in vestibulo curiae profuderunt. (Liv. 23.20.5);
Nam cum secum servilis animus praemia perfidiae reputavit simulque immensa
pecunia et potentia obversabantur, cessit fas et salus patroni et acceptae libertatis
memoria. (Tac. Ann. 15.54.4);277
Due to the scarcity of instances, editors have tried to eliminate them completely. In
(k) Meusel reads facta essent; he also reads postea quam instead of postea cum in
Caes. Civ. 2.17.3, as well as caperet instead of pluperfect ceperat in Caes. Gal. 7.35.4.
In Liv. 42.5.8 nuntiatum est, edd. read esset.
Cum clauses with a perfect indicative tense form can also be combined with main
clauses that have an imperfect or pluperfect indicative verb form, as in (l)–(n) and
(o)–(p), respectively. A main clause with an imperfect tense form indicates a situation
or an ongoing event in which the state of affairs in the cum clause intervenes. A main clause
with a pluperfect tense form describes a state of affairs that precedes the state of affairs in the
cum clause. As before, a cum clause may either precede or follow its main clause.
(l) Etiamne in ara tunc sedebant mulieres, / quom ad me profectu’s ire?
(‘Were the women still sitting by the altar when you set out to come to me?’ Pl. Rud.
846–7)
(m) Cum Caesar in Galliam venit, alterius factionis principes erant Haedui,
alterius Sequani.
(‘When Caesar arrived in Gaul, the leaders of one party were the Aedui, of the other
the Sequani.’ Caes. Gal. 6.12.1)
(n) Sed quom eae litterae adlatae, forte Nabdalsa exercito corpore fessus in lecto
quiescebat . . .
(‘Now when this letter arrived, it chanced that Nabdalsa, fatigued by his bodily
exercise, was resting on his couch . . . ’ Sal. Jug. 71.1)
(o) Bis consul fuerat P. Africanus . . . cum accusavit L. Cottam.
(‘Publius Africanus had been consul twice . . . when he accused Lucius Cotta.’ Cic.
Mur. 58)
277
Further examples can be found in K.-St.: II.337.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(p) Sed cum Placentiam consul venit, iam ex stativis moverat Hannibal . . .
(‘But when the consul reached Placentia, Hannibal had already broken camp . . . ’ Liv.
21.39.4)
Supplement:
CERTE CUM HOC SCRIBSI VALEBAM RECTE. (Tab. Vindol. III. 664 (c. AD 110));
7.126 The use of the present indicative in certain temporal dum clauses
In dum clauses denoting a state of affairs that is contemporaneous with that of the main
clause but which lasts longer (‘while (yet)’, ‘during the time that’)—see § 16.17—the
present indicative is regular even if the state of affairs the dum clause refers to clearly
belongs to the past, as appears for example from perfect tense audivimus in the main
clause of (a). This is common in Early Latin and in Cicero and Caesar. However, the
imperfect indicative came to be used in the dum clause as well, as in (b) (the first attested
instance), and also the imperfect subjunctive instead of the indicative, as in (c) (the first
surely attested instance). There are also instances from Early Latin onwards in which the
perfect is found in dum clauses. By using this tense, the speaker represents the state of
affairs referred to in the clause from the point of view of the time of speaking. There is often
an implicit contrast between the events in the main and subordinate clauses, as in (d) and
(e). The ‘actual’ present is also found in dum clauses, for instance, when the main clause
refers to the time of speaking or to the future. Examples are (f ) and (g), respectively.
(a) Dum haec aguntur, interea uxorem tuam / nec gementem nec plorantem
nostrum quisquam audivimus.
(‘Meanwhile, while this was going on, none of us heard your wife groaning or
weeping.’ Pl. Am. 1098–9)
(b) Dum is in aliis rebus erat occupatus qui summam rerum administrabat,
erant interea qui suis vulneribus mederentur.
(‘While he who wielded the supreme power was occupied with other matters, there
were some who in the meantime were attending to their own wounds.’ Cic. S. Rosc. 91)
(c) Hic dum in opere nostri distenti essent, complures ex superiori loco adver-
sariorum decucurrerunt . . .
(‘At this point, while our men were busily engaged in the operation, a fair number of
the enemy swooped down upon them from higher ground . . . ’ B. Hisp. 23.2)
(d) Dum te fidelem facere ero voluisti, absumptu’s paene.
(‘While you wanted to show yourself a slave faithful to his master, you were almost
ruined.’ Pl. Mil. 409)
(e) Dum enim in una virtute sic omnia esse voluerunt . . . virtutem ipsam quam
amplexabantur sustulerunt.
(‘For at the very moment they wanted all things to be contained in a single virtue, . . . they
removed the very virtue they were embracing.’ Cic. Fin. 2.43)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In historical narrative prose, time clauses with the subordinators postquam, simul,
ubi, ut, and rarely cum often contain two coordinated events, one of them simultan-
eous with the event in the main clause, the other in a successive relationship. Livy and
Tacitus in particular exploit this form of coordination. Ex. (a) has a sequence of
perfect visum est and imperfect dabatur. Ex. (b) has a pluperfect contigerat and an
imperfect cernebat. Even more manneristic are instances like (c), with a historic
infinitive and an imperfect.
(a) Postquam id difficilius visum est neque facultas perficiendi dabatur . . . ad
Pompeium transierunt . . .
(‘But when this seemed rather difficult and no opportunity of carrying it out was
afforded . . . they crossed over to Pompey . . . ’ Caes. Civ. 3.60.4)
(b) . . . postquam ea (sc. occasio) nulla contigerat tutumque ad Syracusas et
munimento et viribus hostem cernebat . . . castra inde movit . . .
(‘ . . . after no such opportunity fell to him, and he saw the enemy safe near Syracuse
thanks to their fortifications and military strength . . . he moved his camp away . . . ’
Liv. 24.36.8)
(c) Ubi crudescere seditio et a conviciis ac probris ad tela et manus transibant,
inici catenas Flaviano iubet.
(‘As the mutiny grew fiercer, and the soldiers went on from abuse and taunts
to use their hands and their weapons, he ordered that fetters be put on Flavianus.’
Tac. Hist. 3.10.3)
278
Sz.: 613 explicitly rejects the interpretation of the present in dum clauses as a historic present. So
also Heberlein (2011: 299) with references.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The first instance of this type of parallelism quoted in the literature is (d). However,
simul . . . neglegentiam may be taken as an independent parenthetical sentence.
(d) Postquam videt me eius mater esse hic diutius, / simul autem non manebat
aetas virginis / meam neglegentiam, ipsam cum omni familia / ad me
profectam esse aibant.
(‘I was told that her mother felt I was staying here too long, and the girl was
of an age where she couldn’t wait for me to act; so she set out with her whole
household to come and find me.’ Ter. Ph. 569–72)
Supplement:
Postquam divitiae honori esse coepere et eas gloria imperium potentia sequebatur,
hebescere, virtus, paupertas probro haberi, innocentia pro malevolentia duci coepit.
(Sal. Cat. 12.1); Deinde, postquam quietae res ex Volscis adferebantur et apparuit
nescire eos victoria et tempore uti, revocati etiam inde exercitus ac duces otiumque
inde, quantum a Volscis, fuit. (Liv. 6.30.7); At postquam exui aequalitas et pro
modestia ac pudore ambitio et vis incedebat, provenere dominationes multosque
apud populos aeternum mansere. (Tac. Ann. 3.26.2);
Simul in silvam ventum est, ubi plures diversae semitae erant, et nox adpropinqua-
bat, cum perpaucis maxime fidis via devertit. (Liv. 44.43.2);
Quod ubi auditum et languescere ira, redire amor ac, si cunctarentur, propinqua nox
et uxorii cubiculi memoria timebantur, prorumpit Narcissus . . . (Tac. Ann. 11.37.2);
Inde, ut circumagi signa obvertique aciem viderunt in hostem et dux . . . se offerebat,
increpare singuli se quisque et alios, et adhortatio in vicem totam alacri clamore
pervasit aciem. (Liv. 6.24.7);
279
For a survey and critical discussion of recent studies on the moods in subordinate clauses, see
Calboli (2011: 9–142). For a discussion of the types of clauses in which the subjunctive is used and the
value of the subjunctive, see Burkard (2006).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(i) In the first place there are contexts in which the subjunctive is optional and
manifests either one of the semantic uses found in simple sentences or one that
can be derived from such uses in an unartificial way. Examples are the use of the
subjunctive in relative clauses, as in (a) and (b), and its use in temporal clauses
referring to future events, as in (c) with priusquam ‘before’. Here the subjunctive
resembles the use of this mood in a potential sense in simple sentences. Other
examples of the meaningful use of the subjunctive in subordinate clauses can be
found below.
(ii) Secondly, there are contexts in which the subjunctive and the indicative are both
possible, each with their regular semantic value, but where the use of the
subjunctive appears to be generalized. An instance of this is the increasing use
of the subjunctive in indirect questions from Early Latin onwards. The subjunctive
is understandable in a ‘deliberative’ indirect question in the same way as it is in
direct questions, as in (d), but it is not when the question is about a fact, as in
(e)—the preceding direct question is in the indicative. Variation in the use of the
moods is not uncommon in such a situation, as in (f) with indicative venio and
subjunctive venerim.
(iii) Thirdly, there are contexts in which the subjunctive is obligatory and therefore
seems to be grammatically determined, but in which it is also ‘harmonious’ in
the sense that it is used in a subordinate clause governed by a verb with a lexical
meaning that resembles the semantic value of the subjunctive in simple sen-
tences. For example, the subjunctive in an ut clause depending on the volition
verb volo ‘to want’ resembles the deontic use of the subjunctive in imperative
(independent) sentences. This harmonic use of the subjunctive is reinforced by
the fact that the negation words are the same in both cases (ne). Many Latinists
explain the use of the subjunctive in such argument clauses diachronically, by
assuming that paratactic structures in which the subjunctive had its regular
semantic value preceded the subordinate ones in which it became a purely
grammatical device. Following this line of argument, the subordinate structure
in (g) could be regarded as derived from the sequence of two simple sentences
in (h). (See § 14.1 for further discussion of this explanation.)
(g) Tu, quod tuo commodo fiat, quam primum velim venias . . .
(‘May you come as soon as possible, which happens at your convenience . . . ’ Cic.
Fam. 4.2.4)
(h) Numquid vis? # Venias temperi.
(‘Is there anything you want? # Come early.’ Pl. Capt. 191)
(iv) The most difficult cases are those where the use of a subjunctive is due to
so-called modal assimilation, some sort of copying of the mood in clause
A from the mood in a governing clause B, without any recognizable semantic
or grammatical justification. Such instances will be dealt with in § 7.162.
The development of the subjunctive from a semantically justified form into a gram-
matical device is difficult to trace and took place mainly before the period covered by
this Syntax. In some cases we can follow the replacement of the indicative by the
subjunctive, without having a full explanation of why the replacement took place
(temporal cum clauses are a good example). There are also reverse developments of
the indicative replacing the subjunctive (as in certain result clauses). Given the
different motivations for the subjunctive discussed in (i)–(iv) above, it is no surprise
that it is not always easy to decide on what has come down to us in the manuscript
tradition. For language users as well, especially language learners, it must have been
complicated to discover a rationale behind the various uses of the subjunctive in
subordinate clauses.
Apart from the four types of use of the subjunctive discussed above, there is
another type that is meaningful but not in one of the two ways (deontic and potential)
that the subjunctive is used in simple sentences: the subjunctive may be used in
subordinate clauses to indicate that the speaker or writer of the clause does not
present it as a statement of fact, but as the thought or opinion of some person,
especially of the subject of the clause. That person is usually someone other than the
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
speaker or writer, but it may be the speaker himself. This use of the subjunctive is
called OBLIQUE; it represents in fact the most basic meaning of the subjunctive, that of
‘non-factive’ (see § 7.8). Examples are (i)–(k). A special variety of this oblique use is in
sentences with a cognition verb that governs an accusative and infinitive clause.
Normally the content of an accusative and infinitive clause is asserted (is presented
as real or factual) and there is no non-factual alternative. The solution is the use of the
subjunctive for the governing verb. Two examples are (l) and (m). In (l) ‘the object of
wonder is Cato’s bearing in old age, not Scipio’s perception of it’.280 For relative
clauses see § 18.24–18.26.
(i) Paetus, ut antea ad te scripsi, omnis libros quos frater suus reliquisset mihi
donavit.
(‘Paetus, as I told you in my last letter, has presented me with all the books which his
cousin left.’ Cic. Att. 2.1.12—so he said)
(j) Nonne ob eam causam expulsus est patria, quod praeter modum iustus esset?
(‘Was he not banished from his country because he was too just?’ Cic. Tusc. 5.105—
so his fellow-citizens thought)
(k) Praeclare in epistula quadam Alexandrum filium Philippus accusat quod
largitione benivolentiam Macedonum consectetur.
(‘In one of his letters Philip takes his son Alexander sharply to task for trying by gifts of
money to secure the goodwill of the Macedonians.’ Cic. Off. 2.53—as Philip thought)
(l) Saepe numero admirari soleo . . . vel maxime quod numquam tibi senectutem
gravem esse senserim . . .
(‘I am frequently wont to marvel . . . especially at the fact that, so far as I have been
able to see, old age is never burdensome to you . . . ’ Cic. Sen. 4)
(m) At etiam litteras, quas me sibi misisse diceret, recitavit homo et humanitatis
expers et vitae communis ignarus.
(‘But he even quoted a letter which he said I had written him—this fellow devoid of
good breeding and ignorant of the usages of life!’ Cic. Phil. 2.7)
Supplement:
Cognition verb in the subjunctive governing an accusative and infinitive: Quae est
ergo apud Caesarem querela, cum eum accusetis, a quo queramini prohibitos vos
contra Caesarem gerere bellum? (Cic. Lig. 25); Nominat iste servum quem magis-
trum pecoris esse diceret. (Cic. Ver. 5.17); Cum enim permissu Hannibalis exisset e
castris rediit paulo post quod se oblitum nescio quid diceret . . . (Cic. Off. 1.40); Ille
omnibus primo precibus petere contendit ut in Gallia relinqueretur, partim quod
insuetus navigandi mare timeret, partim quod religionibus impediri sese diceret.
(Caes. Gal. 5.6.3); . . . Romanis indignantibus quod victoribus victi ultro inferrent
arma, Poenis quod superbe avareque crederent imperitatum victis esse. (Liv. 21.1.3);
280
So Powell ad loc.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In the sections that follow, the use of the moods in finite subordinate clauses is dealt
with according to the type of clause in which they occur: argument clauses, satellite
clauses, relative clauses, and comparative clauses.
In this Syntax several types of finite argument clauses are distinguished. For the
purpose of the present discussion a distinction is made between ut clauses, quod/quia
clauses, and si clauses. A fourth subsection is devoted to indirect questions. A detailed
treatment of these types can be found in Chapter 15.
281
See Nelson (1981: 400).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Nunc quid nos vis facere? # Enim nihil est nisi ut ametis impero.
(‘What do you want us to do now? # There’s absolutely nothing for you to do,
except that I command you to make love to the girls.’ Pl. Bac. 702)
(b) . . . oculis mihi signum dedit / ne se appellarem.
(‘ . . . she gives me a sign with her eyes not to address her.’ Pl. Mil. 123–4)
K.-St.: I.236–7 deal with the impersonal expressions necesse est, opus est, usus est, and
oportet as expressions governing a ‘consecutive’ subordinate clause. However, one
reason to regard these clauses as imperative argument clauses is that the subordi-
nator ut is regularly absent; further, there seem to be no examples with ut non, and
just one with ne: Plin. Ep. 7.6.3 nihil enim tam contrarium, quam si advocatus a
senatu datus defenderem ut reum, cui opus esset, ne reus videretur. The main reason
to deal with them here is that the content of the clauses (perhaps with the exception
of the few governed by usus est) is deontic.
(ii) The subjunctive in argument clauses with verbs and expressions of hinder-
ing, opposing, and preventing introduced by the negative subordinators ne
and quominus (and by quominus and quin when negated) can be regarded as
harmonic as well. The verbs involved are more or less antonyms of the verbs
under (i) above. Examples are (c) and (d).282
(c) Quoi male faciundi est potestas, quae ne id faciat temperat.
(‘If a woman has the opportunity to act badly but refrains from doing so.’ Pl. St. 117)
(f ) Quid ergo dubitas quin lubenter tuo ero meus . . . / faciat male eius
merito?
(‘Then why do you doubt that my master would love to do harm . . . to your
master, which is what he deserves?’ Pl. Poen. 881–2)
282
For the subjunctive with verbs of hindering and refusing, see Orlandini (1996a).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(iv) The subjunctive in argument clauses with verbs and expressions of fearing
cannot be related to the deontic use of the subjunctive in simple sentences,
although the subordinator used is ne (negated ne non—predominant in
Cicero—or ut—predominant in Plautus). However, the subjunctive can be
understood in another way: fearing that something might be the case or will
be the case is less factive than knowing or asserting that something is the
case. This resembles the potential use of the subjunctive in simple sentences.
Examples are (g) and (h). Note in (g) the perfect frausus sit, expressing
anteriority. It should be noted that with these verbs (and the related nouns)
the accusative and infinitive that is normal with communication and cogni-
tion verbs is found as well (for these, see § 15.96). An example is (i).283
(g) Non placet. Metuo in commune ne quam fraudem frausus sit.
(‘I’m not happy about it: I’m afraid he might have got into some mischief
involving the two of us.’ Pl. As. 286)
(h) Ei misero mihi. / Metuo ne non sit surda atque haec audiverit.
(‘Poor me! I’m afraid that she isn’t deaf and has heard all this.’ Pl. Cas. 574–5)
(i) Ego me, illum acerrimum regum hostem, ipsum cupiditatis regni crimen
subiturum timerem?
(‘Could I possibly have feared that I, well known as the bitterest enemy of kings,
should myself incur the charge of seeking kingly power?’ Liv. 2.7.9)
The usual diachronic explanation for the use of the subjunctive with these verbs and
expressions starts from the assumption that the complex sentence is the product of
the combination of two independent sentences: Timeo ne cadas ‘I am afraid that you
fall’ < Ne cadas. Timeo. The use of the subjunctive in this line of reasoning is
deontic.284
(v) The use of the subjunctive is difficult to explain in a large variety of argument
clauses that are commonly called ‘consecutive’ noun clauses. The content of
many of these clauses is factive, and the regular negation word is non, which
is typical of declarative sentences. With many of the governing verbs and
expressions the accusative and infinitive is found as well without a noticeable
difference in meaning and indeed is sometimes more common. In addition,
with some of the governing verbs and expressions subordinate clauses
with quod + indicative are also used. In one subtype (verbs of causation)
the negation word in the subordinate clause is often ne, just as in the
imperative clauses mentioned in paragraph (i) above. The following subtypes
can be distinguished (for a detailed treatment of the ut-clauses involved, see
§§ 15.25–15.36):
283
For instances of metuo + accusative and infinitive, see TLL s.v. 903.76ff. (also a rare quod clause
in Greg. T.).
284
Orlandini (1996a) regards the use of the subjunctive with verbs of fearing as deontic.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The only element all these clauses have in common is the subordinator ut. It is very
difficult to see how the subjunctives can be related to the deontic or potential use of
the subjunctive in simple sentences.
(j) Alterum facio, ut caveam, alterum, ut non credam, facere non possum.
(‘I follow the one precept, to be on my guard, but the other, not to trust, is beyond
me.’ Cic. Att. 2.20.1)
(k) . . . non committam ut, si defugerim, tibi causam aliquam dem recusandi.
(‘ . . . I will not run away and so give you any occasion for refusing.’ Cic. de Orat.
2.233)
(l) . . . eumque ita / faciemus ut quod viderit ne viderit.
(‘ . . . we’ll bring about that he didn’t see what he did see.’ Pl. Mil. 148–9)
(m) Ex quo efficitur non ut voluptas ne sit voluptas, sed ut voluptas non sit
summum bonum.
(‘The conclusion is not that pleasure is not pleasure but that pleasure is not the Chief
Good.’ Cic. Fin. 2.24—NB: variation in negation words)
(n) Sequitur igitur ut etiam vitia sint paria . . .
(‘It follows therefore that vices also are equal . . . ’ Cic. Parad. 22)
(o) . . . consequens esse videtur ut scribas tu idem de legibus.
(‘ . . . I consider it a logical thing that . . . you should also write one on its laws.’ Cic.
Leg. 1.15)
(p) Si est ut velit redducere uxorem, licet.
(‘If in fact he wants to take his wife back, he can.’ Ter. Hec. 501)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(ac) . . . consequens est ut utilis mihi actio adversum te dari debet (debeat eds.).
(‘It is logical that a case useful to me ought to be given against you.’ Gaius
Inst. 2.78)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
7.131 The use of the moods in quod and quia argument clauses
In declarative argument clauses with quod and quia, both moods are used.
With emotion verbs like miror ‘to be surprised at a circumstance’ and suscenseo ‘to
be angry at a circumstance’ and impersonal expressions like mirum est ‘it is surpris-
ing’, the mood is more often the indicative, which is in accordance with their often
factive content, but the subjunctive is found as well. With verbs like conqueror and
queror ‘to complain’ and glorior ‘to boast’, the subjunctive is more common. The
subjunctive marks the content of the clause as non-factive, as something for which the
speaker/writer does not accept commitment, which is ‘harmonic’ with the subjective
meaning of the governing verbs. Examples are (a) and (b). Conversely, in (c) and (d)
the statistically abnormal indicative forms seem to be the only possibility, given the
undoubtedly factive content of the quod clauses. Although often it is possible to give a
semantic explanation for the use of the moods, quite a few instances must be
explained as due to modal assimilation (attraction) (on which see § 7.162) or accepted
as arbitrary choices. The Supplement has a few additional verbs and expressions of
emotion with a quod + subjunctive clause.
(a) Mirabar quod Apollonius, homo pecuniosus, tam diu ab isto maneret
integer.
(‘I was wondering how a well-to-do man like Apollonius had escaped our friend
yonder so long.’ Cic. Ver. 5.16)
(b) Quod ubi fieri milites sensere, alii gaudere palam quod fessos viae labore
flagrantissimo aestu non coegisset pugnare . . .
(‘When the soldiers perceived that this was being done, some rejoiced openly that he
had not compelled men wearied by the toil of marching to fight a battle in scorching
heat.’ Liv. 44.36.7)
(c) Quam multi quod nati sunt queruntur.
(‘How many complain because they have been born!’ Sen. Ben. 1.11)
(d) Adulteram dimittam, patiar adulterum, qui non tam glorior quod filius sum
Miltiadis quam quod vicarius?
(‘Am I to let an adulteress go? Shall I wink at a wife’s deceptions?—I who do not
boast so much because I am the son of Miltiades as because I am his deputy.’ Sen.
Con. 9.1.4)
Supplement (in alphabetical order by governing verb):
. . . cum indignatione conquerimur, quod ab iis a quibus minime conveniat male
tractemur . . . (Cic. Inv. 1.109); Non pudet philosophum in eo gloriari quod haec non
timeat . . . (Cic. Tusc. 1.48); Quod tamen ignavo cadat et sine sanguine leto / maeret
. . . (Ov. Met. 8.518–19); Maximeque hoc in hominum doctorum oratione mihi
mirum videri solet, quod . . . ad gubernacula se accessuros profiteantur excitatis
maximis fluctibus. (Cic. Rep. 1.11); Quem quidem ego hostem, Quirites, quam
vehementer foris esse timendum putem, licet hinc intellegatis, quod etiam illud
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
moleste fero quod ex urbe parum comitatus exierit. (Cic. Catil. 2.4); ‘Nam cum
obiceretur’, inquit, ‘Rodiensibus, quod bellum populo Romano facere voluissent,
negavit poena esse dignos . . . ’ (Tiro gram. 6); Is mihi etiam queritur quod ab nobis
novem solis diebus prima actio sui iudicii transacta sit . . . (Cic. Ver. 1.156);
In quod and quia clauses with verbs and expressions of perception, cognition, and
communication, the subjunctive is also often semantically justified before the Late
Latin period, possibly also for the earliest instance, (e). Other examples are (f ) and (g).
However, after the Classical period, the choice of the indicative or the subjunctive
is often difficult to explain, and some authors seem to use the indicative and
subjunctive at random (or for the sake of stylistic variation). On the whole, quod is
found more often with the subjunctive, quia more often with the indicative. (See
§ 15.9 for additional examples.)
(e) Equidem scio iam filius quod amet meus / istanc meretricem e proxumo,
Philaenium.
(‘Well, I already know that according to some people my son is in love with that
prostitute from next door, Philaenium.’ Pl. As. 52–3)
(f ) Dum haec geruntur, legati Carteienses renuntiaverunt quod Pompeium in
potestatem haberent.
(‘In the course of these proceedings envoys from Carteia duly reported that they had
Pompeius in their hands.’ B. Hisp. 36.1)
(g) Scio enim quia valde me bene ames.
(‘For I know that you love me very well.’ Scaev. dig. 44.7.61.1)
Stylistic variation is the reason for Ammianus’ choice of the indicative or the
subjunctive in quod clauses. They are found coordinated in Amm. 29.2.12 . . .
reputante . . . quod . . . convenit . . . nihilque sit tam iniquum . . . (Hagendahl 1921:
128–9; Calboli 2003: 481–90). In the Peregrinatio Egeriae the use of the indicative
and subjunctive seems to be semantically justified (Wirth-Poelchau 1977: 23;
Bejarano 1983). The same goes for Augustine (Bejanaro 1994; Roca 1999: 161–3).
Gregory of Tours uses the subjunctive more often than the indicative, but even in
this late author the indicative cases are all semantically justified (Raible 1992:
318–26).
(f) Edepol minime miror, si te fugitat aut oculos tuos / aut si te odit, qui istum
appelles Tyndarum pro Philocrate.
(‘Seriously, I’m not surprised at all if he’s avoiding you, or eye-contact with
you, or if he hates you, since you address him as Tyndarus instead of
Philocrates.’ Pl. Capt. 545–6)
(g ) Quid mirum igitur in senibus, si infirmi sint aliquando . . .
(‘What wonder, then, in the case of the aged, if they are sometimes weak . . . ’
Cic. Sen. 35)
285
See Nutting (1925: 72–6).
286
For a discussion of publications of the last fifty years, see Calboli (2012: 97–110).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
along the same lines as subjunctives in simple sentences. Examples are (a)–(c).
They resemble the deontic and potential use of subjunctives in simple sentences.
However, most instances of the subjunctive in indirect questions cannot be
explained on the basis of its uses in simple sentences; rather, its use in this
construction is a purely grammatical phenomenon. A few examples (d)–(g) will
suffice. More instances of the subjunctive in indirect questions can be found in
§§ 15.45–15.62.
(a) Sci’n quid nunc facias? # Loquere.
(‘Do you know what you should do now? # Tell me.’ Pl. Cas. 490—see § 7.41)
(b) Nunc ego inter sacrum saxumque sum nec quo fugiam scio.
(‘Now I’m between the altar and the knife and don’t know where to flee.’ Pl. Cas. 970)
(c) Sed quaero a te cur C. Cornelium non defenderem.
(‘But I ask you why I should not have defended Gaius Cornelius.’ Cic. Vat. 5—see § 7.55)
(d) Vosmet videte quam mi valide placuerit.
(‘You yourselves can see how much I liked her.’ Pl. Mer. 103)
(e) Loquere porro quid sit actum.
(‘Keep telling me what happened.’ Pl. Mer. 198)
(f ) Roga numquid opus sit.
(‘Ask him if he needs anything.’ Pl. Poen. 1008)
(g) Sed ut minuam controversiam, videte quaeso quam in parvo lis sit.
(‘But so that I may diminish the controversy, look, I ask, at how the lawsuit is of little
value.’ Cic. Luc. 83)
There are, throughout the history of Latin, instances of indirect questions (and
exclamations) in the indicative mood, as in (a) and (b), and there are even a few
instances of coordination of two clauses, one in the subjunctive, the other in the
indicative, as in (c).
(a) Non edepol tu scis, mulier, / quantum ego honorem nunc illi habeo.
(‘Woman, you don’t know how great an honour I’m doing her now.’ Pl. Mil. 1074–5)
(b) Vide num moratur.
(‘See if he keeps you waiting.’ Pl. Mos. 614)
(c) Nescis quid te instet boni / nec quam tibi Fortuna faculam lucrifera[m]
adlucere volt.
(‘You don’t know what harvest awaits you and what torch profitable Good Fortune
wants to light for you.’ Pl. Per. 514–15)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The indicative mood in indirect questions is regular, certainly in Plautus, with the
following expressions:287
(i) first person present tense performative expressions like quaero ‘to ask’, rogo
‘to ask’, volo scire ‘I’d like to know’;
(ii) second person questions, often with a directive illocutionary force, like vide’n
= videsne ‘do you see’ and audi’n = audisne ‘do you hear’;
(iii) imperative verb forms like dic (mihi) ‘tell me’, loquere ‘tell me’, responde mihi
‘answer me’, vide ‘look’, and the expression cedo mihi ‘tell me’, ‘describe to
me’, ‘explain to me’.
What these expressions have in common is that they make the desire of the speaker to
obtain information explicit (one might call them ‘metarogatives’—see the discussion
of ‘metadirectives’ in § 6.29). They are typical of interactive text types, such as Plautus’
comedies, Cicero’s letters, and his dialogues (often emended, as in (h)288), and are also
found in poetry. It is sometimes difficult to distinguish the indirect questions follow-
ing these expressions from direct ones (see also § 15.45). Accordingly, Latinists have
tried to explain the indicative in such questions by dealing with these instances as
direct questions; in fact, editors often punctuate them as such. However, it is certainly
not possible nor is it necessary to explain all indicative instances in this way. Examples
of first person expressions in combination with an indirect question are (d)–(f ).
Examples of indirect questions in combination with a directive expression are (g)–(j).
The indicatives in (d)–(e) and (g)–(i) have their own semantic value. However, the
subjunctive can be found as well, also with its own semantic value, as in (f ) and (j).
(d) Ecquem in angiporto hoc hominem tu novisti te rogo. (NB: Some editors
read two sentences.)
(‘Do you know any person in this lane, I ask you?’ Pl. Ps. 971)
(e) Sed volo scire quid debetur hic tibi nostrae domi?
(‘But I want to know, what are you owed here in our house?’ Pl. Truc. 261)
(f ) Sicut ego adventu patris nunc quaero quid faciam miser.
(‘In just the same way I’m asking what to do, now that my father has arrived, wretch
that I am.’ Pl. Mos. 381)
(g) Videte, quaeso, quid potest pecunia.
(‘Please see what money can do.’ Pl. St. 410)
287
The most detailed study of the use of the indicative in indirect questions is Bräunlich (1920).
Calboli (1968: 410–16, 2012: 104) gives a survey of the literature until that period. Stephens (1985) offers a
pragmatic analysis of indirect questions as far as mood is concerned (see below). Bodelot (1987) provides
a study of the moods in a well-defined corpus (Plautus, Terence, Cicero, Seneca). Bodelot (1990) covers
the period from Plautus to Juvenal. See also Bodelot (2003: 264–326). For a discussion of the various
contexts in which the indicative is normal, and the change over time, see Bolkestein (1995).
288
More instances in K.-St.: II.490. See also the following note in the text.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Sentences containing idioms formed with nescio ‘I don’t know’, such as nescio quis/
quid, nescio qui, nescio quo modo, and also those with mirum, such as mirum
quantum, and similar idioms cannot be described as indirect questions at all.289
The nescio idioms function as indefinite expressions and are often printed as one
word; the mirum idioms, in turn, function as quantity adjuncts (see § 10.65). They
have no influence on the mood of the sentence, so that the indicatives in (a) and (b)
need no special explanation.
(a) Certe enim hic nescioquis loquitur.
(‘Certainly someone’s speaking here.’ Pl. Am. 331)
(b) Id mirum quantum profuit ad concordiam civitatis iungendosque patribus
plebis animos.
(‘It was amazing how much it helped for the harmony of the State and for bringing
the patricians and plebeians together.’ Liv. 2.1.11)
Ammianus consistently uses the indicative in this type of expression. Other Late
authors vary.290
289
A survey of these expressions can be found in Bodelot (1990: 58–69).
290
See Bodelot (1999: 216–17).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Due to the formal overlap of some interrogative pronouns, adverbs, and adjectives on
the one hand and relative pronouns, adverbs, and adjectives on the other, it is not
always easy to determine whether the subordinate clause is an indirect question or an
autonomous relative clause. Two formally ambiguous examples are (a) and (b). Ex.
(a) has a relative or interrogative pronoun, (b) a relative or interrogative phrase (for
these two types of autonomous relative clauses, see § 18.15).291
(a) Huc adhibete auris quae ego loquor (v.l. loquar) . . .
(‘Apply your ears here to what I’m saying . . . ’ Pl. Ps. 153)
(b) . . . dum patefacio vobis quas isti penitus abstrusas insidias se posuisse
arbitrantur contra Cn. Pompei dignitatem.
(‘ . . . while I reveal the snares which they think they have laid with complete secrecy
against the honour of Gnaeus Pompeius.’ Cic. Agr. 2.49)
There has been a tendency to systematically regard formally ambiguous clauses with
an indicative as (autonomous) relative clauses, especially in Cicero, or to choose the
subjunctive if the manuscripts vary and take the clause as an indirect question, or
even to emend the text. There are, however, instances in Plautus and other authors
where it is odd to infer from the difference between the indicative in (c) and the
subjunctive in (d), for example, that the former contains a relative clause and
the latter an indirect question. It is more reasonable to assume that Plautus allows
the indicative in this type of indirect question, just as he does in other indirect
questions, for example those with num (see ex. (b) in § 7.134). In the same way it
seems not unreasonable to assume that the same variation occurs in other authors as
well, even in Cicero. Some instances are given below in the Supplement.
(c) Nunc quam rem oratum huc veni, primum proloquar.
(‘Now I’ll first tell you what I’ve come here to ask you for.’ Pl. Am. 50)
(d) Observabo quam rem agat.
(‘I’ll observe what he’s up to.’ Pl. Am. 270)
Supplement:
At sci’n quo modo tibi res se habet? (Pl. Aul. 47); Quae hic monstra fiunt, anno vix
possum eloqui. (Pl. Mos. 505); Dicam prius qui deorum causa, tum qui hominum
sunt instituti. (Var. L. 6.12); Quibuscum bellum gerimus, iudices, videtis. (Rhet.
Her. 4.13); Quae res mihi non mediocrem consolationem attulit volo tibi
commemorare . . . (Sulp. Ruf. Fam. 4.5.4); Scis quare non possumus ista? Quia nos
posse non credimus. (Sen. Ep. 116.8); . . . tenes . . . de Ulixe fabulam, quemadmodum
291
The second type, illustrated by (b), is taken as an instance of a relative clause with an ‘incorporated’
antecedent by K.-St.: II.493 (for terminology, see § 18.15). Similarly in Pl. Men. 685: Video quam rem agis.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The same type of formal ambiguity arises from the overlap between a number of
relative adverbs and adjectives on the one hand and interrogative (or exclamatory)
adverbs and adjectives on the other, such as quam ‘as much as’ / ‘how much?’,
quantus ‘as much as’ / ‘how much?, and ut ‘just as’/ ‘how?’ The use of the indicative
in indirect questions in Early Latin, as in (e)–(g), is often explained as being due to
this formal ambiguity of the introductory words. However, the subjunctive is the
more normal expression, as in (h)–(j).
(e) Non potest / dici quam indignum facinus fecisti et malum.
(‘It can’t be said what an evil and wicked deed you’ve done.’ Pl. Mos. 458–9)
(f ) Non edepol tu scis, mulier, / quantum ego honorem nunc illi habeo.
(‘Woman, you don’t know how great an honour I’m doing her now.’ Pl. Mil.1074–5)
(g) Cor dolet, quom scio ut nunc sum atque ut fui.
(‘My heart aches since I know how I am now and how I used to be.’ Pl. Mos. 149)
(h) Sci’n quam bono animo sim?
(‘Do you know how cheerful I am?’ Pl. Am. 671)
(i) Scis tu quidem hercle, mea si commovi sacra, / quo pacto et quantas soleam
turbellas dare.
(‘You know what fine little uproars I usually cause and how I do so, if I get ready for
the sacrifice.’ Pl. Ps. 109–10)
(j) Iam scis ut convenerit.
(‘You already know how it was agreed.’ Pl. Cur. 435)
In Cicero, clauses introduced by these words, when they contain an indicative, are regarded
as comparative clauses, and not as indirect questions. There are, however, a few instances
of such indirect questions with an indicative in the Classical and later periods as well.
Supplement:
. . . huius tyranni . . . interitus declarat quantum odium hominum valet (v.l. valeat
edd.) ad pestem . . . (Cic. Off. 2.23); Enumeratione utemur, cum dicemus numero
quot de rebus dicturi sumus. (Rhet. Her. 1.17); Si scias quam sollicitus sum, tum hanc
meam gloriam, quae ad me nihil pertinet, derideas. (Cael. Fam. 8.15.1—Lambinus
read sim); Adice quot nuptas prius, / quot virgines dilexit. (Sen. Her. O. 364–5);
292
More instances in K.-St.: II.492–3; Bräunlich (1920: 96–104).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The use of the moods in indirect questions shows a development over time. Excluding
the indirect questions discussed in § 7.134 (questions with an information-requesting
illocutionary force or exclamations), which are common in Plautus and probably
largely determined by his comical style, about 10% of the other forms have an
indicative.293 In Cicero, Caesar, Sallust, Livy, and Tacitus the subjunctive is the
norm, but in later authors the indicative is more common, though generally less
frequent than the subjunctive.294 A number of factors are involved, among them
sociolinguistic and stylistic ones: the indicative is relatively common in Vitruvius and
in the speech of the freedmen in Petronius (although the subjunctive prevails), as well
as in (non-didactic) poetry.295 In Late Latin authors, the degree to which they use the
subjunctive or the indicative also depends on their attitude towards the Classical
models. Examples of indicatives are given in the Supplement.
Supplement:
Prose: . . . non debere extrinsecus adsum<i> cur similia sunt. (Var. L. 8.69); Nulla lex
satis commoda omnibus est: id modo quaeritur, si maiori parti et in summam
prodest. (Liv. 34.3.5); Relinquetur desideratio . . . quid ita non etiam ibi nascitur
pulvis . . . (Vitr. 2.6.4); De singulis generibus quibus proprietatibus e natura rerum
videantur esse comparatae quibusque procreantur rationibus exposui. (Vitr. 2.9.17—
NB: variation of moods); Nunc ut in singulis mensibus sol signa pervadens auget et
minuit dierum et horarum spatia dicam. (Vitr. 9.2.4); Narratis quod nec ad caelum
nec ad terram pertinet, cum interim nemo curat, quid annona mordet. . . . Et quo-
modo siccitas perseverat. (Petr. 44.1–2 (Ganymedes speaking)); Tantum quod mihi
non dixerat quid pridie cenaveram. (Petr. 76.11 (Trimalchio speaking)—NB: apart
from the presence of quid this looks like an autonomous relative clause296); Multa
sunt quae esse concedimus, qualia sunt ignoramus. (Sen. Nat. 7.25.1—editors read
sint); Qualem mihi animum esse existimas, quom cogito quam diu te non vidi et
quamobrem non vidi? (Aur. Fro. 3.20); Nam si cognoverimus quae istae personae
sunt (sint edd. nonn.), simul intellegemus quae sui iuris sint. (Gai. Inst. 1.50—NB:
variation); Accipe igitur quem ad modum homo curiosus iumenti faciem sustinens
cuncta quae in perniciem pistoris mei gesta sunt cognovi. (Apul. Met. 9.30.5); . . . ut
possit ei dici cur non multo ante profectus est quam . . . (Ulp. dig. 2.11.2.8); . . . ne
cui lateat in qua principalis quaestio dimicatura est. (Tert. Marc. 1.1.7); Vide quam
ex aequo habetur qui viduae benefecerit . . . (Tert. Ux. 1.8.1); . . . quendam Vitalianum,
quem si norat ambigitur, Latine salute data blande produxit . . . (Amm. 26.7.15);
Unde factum est ut cruenta quaestione vexati cervicibus perirent abscisis, quid acturi
293
This figure is based on calculations by Bodelot (1987).
294
See Adams (2013: 762–9).
295
A survey of the diachronic development can be found in Sz.: 538–9 and, for the period from Plautus
through Juvenalis in Bodelot (1990: 119–21). For the use of the indicative in Late Latin, see Bodelot
(1999). See also Adams (2013: 747–70).
296
See Adams (2013: 763–4).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The use of the moods in satellite clauses is treated separately for the various semantic
types that can be distinguished. In some of these there is a choice between the
indicative and the subjunctive, with a semantic justification for the choice. Examples
are quia and quod reason clauses. In these clauses the indicative is used for factive
information, whereas the subjunctive marks the content of the clause as non-factive
(see § 7.145). In principle, in these clauses the same choices are available as in simple
sentences, with the usual semantic values. There are also types of satellite clauses in
which the subjunctive is regularly used without an evident (synchronic) justification
or where the indicative and subjunctive are both used without an evident difference in
meaning. What makes the description of the use of the moods in the various types
even more complicated is the fact that there are only very few types in which one
297
According to Sz.: 538, Augustine has the indicative regularly in his Confessiones. With accipio and
audio the subjunctive is regular (see SLA s.v.).
298
See Wanner (1990: 261–7). For a sociolinguistic explanation, see Bodelot (1987: 128–30), for a
structural one (the indicative is the unmarked, neutral form, that may be used instead of the subjunctive),
see Suárez (1994b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
consistently finds either an indicative or a subjunctive, and within the same class the
use of the mood sometimes varies with the specific subordinator used. Finally, there
are developments in the course of time. Table 7.21 presents a survey of the use of
the moods in the various types of satellite clauses with some initial comments. Details
can be found in the sections that follow. Oblique use of the subjunctive due to the
particular syntactic position of the clause within a complex sentence will be left out of
consideration.
Table 7.21 The distribution of the indicative and subjunctive moods in satellite
clauses
Type of Indicative Subjunctive Comments More
satellite instances
clause in §
Space + + selection of the mood as in simple 16.6
sentences
Time ++/– +/– selection of the mood depends on 16.7–16.30
the individual subordinator; with
some subordinators (meaning
‘until’ or ‘before’) both moods are
possible, with a semantic difference
Reason ++ + with most reason subordinators 16.39–16.48
both moods are possible with a
semantic difference; the causal
interpretation of cum is only found
with the subjunctive
Purpose – + the subjunctive is semantically 16.49–16.52
justified
Stipulative – + the subjunctive is semantically 16.53
unjustified
Result – + the subjunctive is semantically 16.54
unjustified
Condition + + selection of the mood as in simple 16.55–16.66
sentences
Concession +/– +/– selection of the mood depends on 16.68–16.82
the individual subordinator; there
is a diachronic development
Manner + the indicative is normal 16.33–16.36
‘+’ means that the type of clause is used with the mood of the column
‘–’ means that the type of clause is not used with the mood of the column
‘+/–’ means that the use of the mood varies with the subordinator used in that type of clauses
‘++’ means that there are more subordinators in that type of clauses with which the indicative is regular
NB: ‘Oblique’ use of the subjunctive and use due to attraction are ignored.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
299
Statistical data concerning the use of the moods in temporal clauses can be found in Heberlein
(2011: 261).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(c) <Non hercle> hoc longe destiti / instare usque adeo donec se adiurat anus /
iam mihi monstrare.
(‘I didn’t stop insisting on this until the old woman swore that she’d show her to me
right away.’ Pl. Cist. 582–4)
(d) Rationes ad aerarium, antequam Dolabella condemnatus est, non audet referre.
(‘Verres dared not submit his accounts to the Treasury till after Dolabella’s condem-
nation.’ Cic. Ver. 1.98)
(e) Prius respondes quam rogo.
(‘You answer before I ask.’ Pl. Merc. 456)
Supplement (in alphabetical order by subordinator):
. . . coquito (sc. brassicam) usque donec commadebit bene, aquam defundito. (Cato
Agr. 156.5); . . . tamen usque eo timui nequis de mea fide atque integritate dubitaret,
donec ad reiciundos iudices venimus. (Cic. Ver. 1.17); Sed neque credes tu mihi,
donec compleris sanguine campum . . . (Liv. 25.12.6);
Opperire dum effero ad te argentum. (Pl. Epid. 633); Ego in Arcano opperior dum
ista cognosco. (Cic. Att. 10.3); Tityre, dum redeo (brevis est via) pasce capellas . . .
(Verg. Ecl. 9.23);
Ab eo praeco dicitur locare, quod usque † id † emit, quoad in aliquo constitit
pretium. (Var. L. 5.2.15); Atque hoc scitis omnes, usque adeo hominem in periculo
fuisse quoad scitum est Sestium vivere. (Cic. Sest. 82); Hasdrubal . . . procul ab hoste
intervallo ac locis tutus tenebat se, quoad multum ac diu obtestanti quattuor milia
peditum et <quingenti> equites in supplementum missi ex Africa sunt. (Liv. 23.26.2);
Id fit tum, antequam gemmare aut florere quid incipit. (Var. R. 1.40.4); Nunc,
antequam ad sententiam redeo, de me pauca dicam. (Cic. Catil. 4.20); Sed antequam
opprimit lux maioraque hostium agmina obsaepiunt iter, per hos qui inordinati
atque incompositi obstrepunt portis erumpamus. (Liv. 22.50.8);
Nam semper occant priusquam sariunt rustici. (Pl. Capt. 663); Quem ad modum
igitur membris utimur prius quam didicimus, cuius ea causa utilitatis habeamus,
sic . . . (Cic. Fin. 3.66); ‘Sine, priusquam conplexum accipio, sciam’ inquit, ‘ad hostem
an ad filium venerim, captiva mater ne in castris tuis sim. (Liv. 2.40.5);
Already in Cicero’s time, and regularly in Silver Latin, the subjunctive is also found
when the temporal clause contains a factual time indication. Two clear examples
where a semantic justification is lacking are (k) and (l).300
(k) Nam Rhenus . . . in duos amnes dividitur servatque nomen . . . donec Oceano
misceatur.
(‘For the Rhine . . . divides itself into two rivers, retaining its name . . . till it mingles
with the Ocean.’ Tac. Ann. 2.6.4)
(l) . . . mansere infensi ac minitantes donec magistratu abirent.
(‘ . . . they maintained their hostile and threatening attitude till they went out of
office.’ Tac. Ann. 5.11.2)
The same variation in mood is found in comparative quam clauses with pridie and
(rare) postridie, as in (m) indicative and (n) subjunctive, below. See also § 7.161 on
potius quam.
300
Taken from Woodcock (1959: 183). See also Heberlein (2011: 322–7, 337–51).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(m) Si hic pridie natus foret quam ille est, hic haberet regnum in caelo.
(‘If he’d been born a day earlier than Jupiter, he would have the kingship in
heaven.’ Pl. Mil. 1083)
(n) Nam cum pridie quam legati ad Tarquinios proficiscerentur cenatum forte
apud Vitellios esset . . . sermonem eorum ex servis unus excepit . . .
(‘For when on the day before the envoys were leaving for the Tarquins, they
had by chance dined at the house of the Vitelli . . . their conversation one of
the slaves overheard . . . ’ Liv. 2.4.5)
With antequam and priusquam the subjunctive prevails when the main clause is
positive, the indicative when it is negative or when negation is implied. This has to do
with the fact that the antequam/priusquam clause is more likely to be interpreted as
an intention with a positive main clause. See (o) and (p), taken from K.-St.: II.367.
The second situation in which the subjunctive is used in a temporal satellite clause is in
certain clauses with the subordinator cum (quom) ‘when’. Cum is used both as a relative
temporal adverb301 in an adnominal relative clause (related to the relative pronoun qui)
and as a temporal subordinator. These temporal clauses can be and sometimes are indeed
taken as autonomous adverbial relative clauses (see § 16.10). When cum was used
adnominally there was initially a clear difference between the use of the indicative
(‘factive’) and the subjunctive (‘potential’). Whereas cum clauses with an indicative
indicate at what time or in what period the state of affairs of the main clause took place
(they are ‘identifying’ clauses—see § 18.5), those with a subjunctive indicate in what kind
of time or period the state of affairs of the main clause took place (they are ‘descriptive’
clauses). This is still valid in Cicero’s works. Examples are (a) and (b). Something similar is
the case in (c) and (d) with the correlative adverb tum in the main clause.
(a) Ne tu, Eruci, accusator esses ridiculus, si illis temporibus natus esses cum ab
aratro arcessebantur qui consules fierent.
(‘In truth, Erucius, you would have made an absurd accuser, if you had been born in the
days when men were summoned from the plough to become consul.’ Cic. S. Rosc. 50)
(b) Accepit enim agrum temporibus iis cum iacerent pretia praediorum.
(‘For he got the farm at a time when the prices of estates were down.’ Cic. Q. Rosc. 33)
301
For discussion and references, see Heberlein (2011: 263–5). For the use of the subjunctive in cum
clauses, see also Calboli (2012: 56–77).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(c) Quidnam homines putarent, si tum occisus esset cum tu illum in foro
spectante populo Romano gladio insecutus es . . .
(‘What would men think if he had been killed at the time when you pursued him in
the Forum with a sword . . . ’ Cic. Phil. 2.21)
(d) Brundisio quae tibi epistulae redditae sunt sine mea tum videlicet datas cum
ego me non belle haberem.
(‘The letters which were brought to you from Brundisium without one from me must
have been dispatched when I was not so well.’ Cic. Att. 5.11.7)
The same difference is also discernible in Cicero’s time for the temporal subordinator
cum. Whereas cum clauses with an indicative indicate at what time or in what period
the state of affairs of the main clause took place, those with a subjunctive indicate in
what kind of time or period the state of affairs of the main clause took place. The latter
often allow a causal or a concessive interpretation and thus resemble quoniam
disjunct clauses and concessive and conditional clauses. In many contexts a speaker
could choose between an ‘identifying’ and a ‘descriptive’ expression, which means
that, at least in Cicero’s time, the two expressions were not synonymous. The use of
the subjunctive in temporal cum clauses is rare in Early Latin, and the earliest
instances in Ennius and Plautus are often emended (e.g. ex. (e)).302 In the Classical
period, however, the subjunctive is the regular mood. In Late Latin, in turn, the
indicative became more common once again.303 Examples with the subjunctive
are (e)–(h).
(e) Quem quidem hercle ego, in exilium quom iret (ibat [?] cj. Leo), redduxi
domum.
(‘A son whom I brought back when he was going into exile.’ Pl. Mer. 980)
(f ) Cum essem in provincia legatus, quamplures ad praetores et consules vinum
honorarium dabant.
(‘When I was legate in the province, very many people used to give a voluntary
contribution of wine to the praetors and consuls.’ Cato orat. 132)
(g) Magistratus quom ibi adesset, occepta’st agi.
(‘When the official arrived, the performance began.’ Ter. Eu. 22)
302
The instances of cum + subjunctive for which there is no evident semantic justification can be
found in Lodge: s.v. I.529.
303
A survey of the literature and the theories about the emergence of the subjunctive in cum clauses
can be found in Sz.: 620; Calboli (1968: 416–23, 2012: 56–77) has an excellent survey of the literature on
the development in temporal clauses in general. See also Wanner (1990: 268–73). The explanation given
in the text follows Hale (1891). The difference in meaning is expressed in the following way in the OLD:
‘at the time that, when’ (OLD s.v. §1) ‘at a time or in circumstances when’ (OLD s.v. §3). Examples (e)–(h)
are taken from Hale (1891: 192–3). The translations are partly taken from the original English edition
(p. 159). The overall argumentation follows Lavency (2003a), which summarizes and expands several
earlier publications of his. See also Heberlein (2011: 273–86, 314–18).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(h) Sed tamen, cum horas tris fere dixisset, odio et strepitu senatus coactus est
aliquando perorare.
(‘Yet when he had spoken for nearly three hours, he was compelled finally to finish
his speech by the hatred and outcry of the senate.’ Cic. Att. 4.2.4)
Supplement:
Cumque caput caderet, carmen tuba sola peregit / et pereunte viro raucum sonus
aere cucurrit. (Enn. Ann. 519–20V = 485–6S); Cum Sulla conatus esset tempore
magno, eduxit copias, ut Archelai turrim unam, quam ille interposuit, ligneam
incenderet. (Quad. hist. 81); Cum enim rex Pyrrhus populo Romano bellum ultro
intulisset cumque de imperio certamen esset cum rege generoso ac potente, perfuga
ab eo venit . . . (Cic. Off. 3.86); Caesari cum id nuntiatum esset eos per provinciam
nostram iter facere conari, maturat ab urbe proficisci . . . (Caes. Gal. 1.7.1); Tum
quoque cum fugerem quaedam placitura cremavi . . . (Ov. Tr. 4.10.63);
Temporal cum clauses are sometimes said to be interchangeable with present or
perfect participles that function as a secondary predicate (see § 7.78). Grammars tend
to distinguish a habitual or iterative use of such cum clauses (‘on any or every
occasion on which’—OLD s.v. § 2), but this distinction is entirely contextual.
304
More of such almost minimal pairs can be found in K.-St.: II.345.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(k) Num etiam eius collega P. Decius . . . cum se devoveret (devoverat cj. Davi-
sius) et equo admisso in mediam aciem Latinorum inruebat . . .
(‘Yet what about his colleague Publius Decius . . . When Decius vowed him-
self to death, and setting spurs to his horse was charging into the thickest of
the Latin ranks . . . ’ Cic. Fin. 2.61)
305
References in Sz.: 624; for Plautus, see Lodge: II.529A §2; for Symmachus, see Haverling (1988: 232).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Classical and later examples with the indicative: HS <DC> exprimes ab Hermogene,
cum praesertim necesse erit. (Cic. Att. 12.25.1); . . . nec veterum dulci scriptorum
carmine Musae / oblectant cum mens anxia pervigilat . . . (Catul. 68.7-8); . . . non
tamen omnino Teucros delere paratis, / cum talis animos iuvenum et tam certa
tulistis / pectora. (Verg. A. 9.248–50); Ita cum est levis (sc. terra), aere solidata non
recipit in se nec combibit liquorem. (Vitr. 2.3.4); Sed differentiae ratio manifesta est,
cum a parentibus et a coemptionatoribus isdem verbis mancipio accipiuntur (acci-
piantur edd. nonn.) quibus servi (Gai. Inst. 1.123); Quantum detrahant fidei . . .
disciplina ecclesiae et praescriptio apostoli declarat, cum digamos non sinit praesi-
dere, cum viduam adlegi in ordinem nisi univiram non concedat. (Tert. Ux. 1.7.4—
NB: variation in the moods of the two cum clauses); Quantum Deo sanctitas placet,
cum illam etiam inimicus affectat? (Tert. Ux. 1.7.5);306 . . . quae superfluum est
explicare, cum neque operae pretium aliquid eorum habuere proventus nec
historiam producere per minutias ignobiles decet. (Amm. 27.2.11); Nam cum eum
argumenta deficiunt quibus probet sabbato ieiunandum, in luxurias epularum . . .
invehitur . . . (August. Ep. 36.2.3);
The few Plautine subjunctive instances are explained away in various ways, just as
Ciceronian instances of the indicative have been emended. In (e) Piderit suggested
quoniam for cum. A summary can be found in Sz.: 620.
In dum clauses with dum meaning ‘for the duration of the time in which’ (see
§ 16.17), the subjunctive is sometimes used to express intention or purpose (‘for as
long as is needed’—OLD s.v. § 1d); this is the case in (a) and (b). However, there are
also instances where this explanation does not hold. Instances of this final use are also
occasionally found in Late Latin.
(a) Inde huc exii, crapulam dum amoverem.
(‘I went out here so as to get rid of my headache.’ Pl. Ps. 1282)
(b) Videto ut bene et otiose percoquas. Aperito, dum inspicias, bis aut ter.
(‘See that it bakes thoroughly and slowly; uncover it two or three times in order to
examine it.’ Cato Agr. 76.4)
Supplement:
. . . horam sibi octavam, dum in foro bovario inquireret, postulavit. (Cic. Scaur. 23);
Multa Antonio, dum (tunc mss.) interfectores patris ulcisceretur, multa Lepido
concessisse. (Tac. Ann. 1.9.4);307
No intention or purpose implied: Dum habeat, tum amet. Ubi nil habeat, alium
quaestum coepiat. (Pl. Truc. 232); Ne tam diu quidem dominus erit in suos dum ex
iis de patris morte quaeratur? (Cic. S. Rosc. 78); Denique isto bono utare, dum adsit,
306
For more instances in Tertullian, see Hoppe (1903: 80 (= 1985: 153–5)).
307
For more instances, see TLL s.v. dum 2230.36ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
cum absit, ne requiras . . . (Cic. Sen. 33); . . . archontes usque ad Charopem, dum
viverent, eum honorem usurpabant. (Vell. 1.2.2); . . . dum superesset, ille timens ne
proderetur tractabilis erat et mollior. (Amm. 28.1.7);
The subjunctive is, however, sometimes used without a semantic justification in Silver
and Late Latin. Instances can be found in the following sections.
308
See Ghiselli (1961).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
7.146 The use of the moods in quia and quod reason clauses
In quia- and quod reason clauses the indicative is the regular mood in all periods of
Latin. Examples are (a) and (b) below. The subjunctive is used when the speaker of
the sentence does not commit himself to the content of the reason clause. Examples
are (c) and (d).309
(a) Hoc adeo hoc commemini magis, quia illo die inpransus fui. /
(‘I remember this all the more clearly because that day I went without lunch.’
Pl. Am. 254)
(b) Mater dicta quod sum, eo magis studeo vitae.
(‘I’m all the keener on his survival because I’ve been called his mother.’ Pl. Truc. 457)
(c) Nunc mea mater irata est mihi, / quia non redierim domum ad se . . .
(‘Now my mother’s angry with me, on the grounds that I didn’t return home to
her . . . ’ Pl. Cist. 101–2)
(d) Laudat Africanum Panaetius quod fuerit abstinens.
(‘Panaetius praises Africanus for his self-restraint.’ Cic. Off. 2.76)
Supplement:
Chrysalus mihi usque quaque loquitur nec recte, pater, / quia tibi aurum reddidi et
quia non te defrudaverim. (Pl. Bac. 735–6—NB: variation); Diana dicta quia noctu
quasi diem efficeret. (Cic. N.D. 2.69);
Iuro per Iovem et Mavortem me nociturum nemini, / quod ego hic hodie vapu-
larim. (Pl. Mil. 1414–15);
The subjunctive is also regular when the content of the reason clause is an imaginary
and not a real reason for the main clause and is then corrected in a coordinate clause
with sed (usually in the indicative) or in a following contrastive sentence, usually
immediately following. Sometimes the correction is not expressed at all. Examples
of such imaginary reasons are (e) and (f ). The same holds for a variety of
expressions: for example, non quo . . . sed, as in (g). These expressions are often
combined with a cataphoric expression like eo, ideo, idcirco, or ob eam causam.
Subjunctive quia clauses are less often used in this type of expression than quod
clauses. The indicative is normal when two facts are simply opposed by means of a
negation and sed. Examples are (h) and (i).310 However, in poetry and in prose
from Livy onwards, the indicative is sometimes used in quia clauses where one
would expect a subjunctive. The subjunctive is more common in quod clauses than
in quia clauses generally.311
309
Further examples in OLD s.v. quia § 2a,b; quod § 10b.
310
Further examples in OLD s.v. quia § 3; quod § 11c.
311
See Sz.: 588 for references.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(e) Ea res me male habet, at non eo quia tibi non cupiam quae velis: / verum
istam amo.
(‘That does make me feel down; but not because I wouldn’t wish you to have what
you want; but I love her.’ Pl. As. 844–5)
(f ) Pugiles vero . . . in iactandis caestibus ingemescunt, non quod doleant ani-
move succumbant, sed quia profundenda voce omne corpus intenditur . . .
(‘Boxers in fact . . . groan in the swinging of their gauntlets, not because they feel pain
or are losing heart, but because by the burst of sound the whole body is made more
tense . . . ’ Cic. Tusc. 2.56)
(g) Non edepol quo te esse inpudicam crederem: / verum periclitatus sum
animum tuom . . .
(‘Well, not because I believed you’re unchaste. But I was testing your attitude . . . ’ Pl.
Am. 913–14)
(h) Neque vero hoc, quia sum ipse augur, ita sentio, sed quia sic existimare nos
est necesse.
(‘But it is not because I myself am an augur that I have this opinion, but because the
facts compel us to think so.’ Cic. Leg. 2.31)
(i) Quod si respondet, non quia mane est, sed quia remissio est, dari debet.
(‘But if it works, the reason for giving food is not the fact that it is morning, but the
fact that the patient has more of a remission.’ Cels. 3.5.4)
Supplement:
Imaginary reasons in the subjunctive: Non eo haec dico, quin quae tu vis ego velim
et faciam lubens: / sed ego hoc verbum quom illi quoidam dico, praemonstro tibi, /
ut . . . (Pl. Trin. 341–2); Non quo dissimilis res sit, sed quo is qui facit. (Ter. Ad. 825);
Non eo dico, C. Aquili, quo mihi veniat in dubium tua fides et constantia aut quo non
in eis quos tibi advocavisti . . . spem summam habere P. Quinctius debeat. (Cic.
Quinct. 5); Non quo mea quidem iam intersit . . . sed tamen ista tua nullum ad
usum meum, tantum cognoscendi studio adductus requiro. (Cic. de Orat. 2.74);
Nemo enim ipsam voluptatem, quia voluptas sit, aspernatur aut odit aut fugit, sed
quia consequuntur magni dolores eos qui . . . (Cic. Fin. 1.32); . . . hoc mihi Latinis
litteris inlustrandum putavi, non quia philosophia Graecis et litteris et doctoribus
percipi non posset, sed meum semper iudicium fuit . . . (Cic. Tusc. 1.1); Ego me
ducem in civili bello quoad de pace ageretur n<eg>avi esse, non quin rectum
esset sed quia, quod multo rectius fuit, id mihi fraudem tulit. (Cic. Att. 7.26.2);
Etsi non idcirco eorum usum dimiseram quod iis suscenserem sed quod eorum me
subpudebat. (Cic. Fam. 9.1.2); Tertiam ad te hanc epistulam scripsi eodem die
magis instituti mei tenendi causa, quia nactus eram cui darem, quam quo haberem
quid scriberem. (Cic. Fam. 16.6.1); Iactatum in condicionibus nequiquam de
Tarquiniis in regnum restituendis, magis quia id negare ipse nequiverat Tarquiniis,
quam quod negatum iri sibi ab Romanis ignoraret. (Liv. 2.13.3); Non enim . . . ista
accidunt, quod dii vestri a nobis non colantur, sed quod a vobis non colatur Deus.
(Cypr. Ad Demetr. 5);
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(l) (sc. interemit) Salvium Cocceianum, quod Othonis imperatoris, patrui sui,
diem natalem celebraverat; Mettium Pompusianum, quod habere impera-
toriam genesim vulgo ferebatur et quod depictum orbem terrae in mem-
brana[s] contionesque regum ac ducum ex Tito Livio circumferret quodque
servis nomina Magonis et Hannibalis indidisset.
(‘He put to death Salvius Cocceianus, because he had kept the birthday of
the emperor Otho, his paternal uncle; Mettius Pompusianus, because it was
commonly reported that he had an imperial nativity and carried about a
map of the world on a parchment and speeches of the kings and generals
312
See den Hengst (2007). For diachronic data, see Baños (2014: 103–8). See also § 7.131.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
from Titus Livius, besides giving two of his slaves the names of Mago and
Hannibal.’ Suet. Dom. 10.3)
(m) Pius cognominatus est a senatu, vel quod soceri fessi iam aetatem manu
praesente senatu levaret (quod quidem non satis magna<e> pietatis est
argumentum, cum impius sit magis, qui ista non faciat, quam pius qui
debitum reddat) vel quod eos, quos Hadrianus per malam valetudinem
occidi iusserat, reservavit . . .
(‘He was given the name of Pius by the senate, either because, when his
father-in-law was old and weak, he lent him a supporting hand in his
attendance at the senate (which act, indeed, is not sufficient as a token
of great dutifulness, since a man were rather undutiful who did not
perform this service than dutiful if he did), or because he spared those
men whom Hadrian in his ill-health had condemned to death . . . ’ Hist.
Aug. Pius 2.3)
In quando(quidem) reason clauses the regular mood is the indicative. The subjunctive
is used, as with the other causal subordinators, when the speaker does not commit
himself to the truth of the content of the clause. An instance of such a subjunctive is
(a). Subjunctives without a semantic justification occur in Late Latin and are in some
authors normal (Tertullian) or predominating (Cyprian)—(b) and (c). In later
authors the indicative becomes normal once again.
(a) Neque (sc. terra) fluminibus adgesta semper laudabilis, quando senescant
sata quaedam aqua.
(‘Nor is alluvial soil deposited by rivers always to be recommended, seeing that some
plants become weak in a damp environment.’ Plin. Nat. 17.27)
(b) . . . non . . . haec quoque distributio animae ad animae sectiones pertinebit,
quando ne ipsum quidem corpus ita dividatur in membra . . .
(‘ . . . this distribution of the soul will not also be suitable to the sections of the soul,
since not even the body itself is divided thus into limbs . . . ’ Tert. An. 14.3)
(c) Quando aliud baptisma praeter unum esse non possit, baptizare se
opinantur.
(‘Since there can be no other baptism than just one, they believe that they baptize.’
Cypr. Unit. eccl. 11)
In quatenus reason clauses the regular mood is the indicative. For examples, see § 16.45.
In siquidem reason clauses the indicative is the regular mood. Later authors also use
the subjunctive. For examples, see § 16.63.
313
A few can be found in Sz.: 642–3.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
corresponds to the use of the subjunctive in imperative sentences. This also resembles
the use of the subjunctive in imperative argument clauses (see § 7.130 (i)). Two
examples are (a) and (b). Further examples of purpose clauses can be found in
§§ 16.49–16.52.
(a) Ego, nisi quid me vis, eo lavatum, ut sacruficem.
(‘Unless you want anything from me, I’m going to bathe so I can sacrifice.’
Pl. Aul. 579 )
(b) Ego servi sumpsi Sosiae mi imaginem, / . . . ut praeservire amanti meo
possem patri / atque ut ne qui essem familiares quaererent . . .
(‘I’ve taken on the slave Sosia’s image, . . . so that I can be in attendance on my
father during his love affair and so that the family servants won’t ask who I am . . . ’
Pl. Am. 124–7)
The subjunctive is also regular in si clauses meaning ‘to see if ’, as in (c). However, the
indicative is used as well, as in (d). Further examples can be found in § 16.58.
314
See Calboli (2003; 2011: 130) on rare instances of the indicative in (Late) Latin. A discussion
of explanations of the subjunctive can be found in Gallego (2007: 182–3).
315
A summary can be found in Sz.: 639. Instances from Gregory of Tours can be found in Gallego
(2007). See also Cabrillana (2011: 37–41).
316
For the need to change the text, see Woodman ad loc.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(‘It was clear that the trenches wandered from a straight line to such an extent that
the upper trench made towards the south.’ CIL VIII.2728.24–9 (Lambaesis, Numidia,
c. AD 152))
(f) Inter tot urguentia Hermogene defuncto ad praefecturam promovetur Hel-
pidius ortus in Paphlagonia, aspectu vilis et lingua, sed simplicioris ingenii,
incruentus et mitis adeo, ut, cum ei coram innocentem quendam torquere
Constantius praecepisset, aequo animo abrogari sibi potestatem orabat
haecque potioribus aliis ex sententia principis agenda permitti.
(‘In the midst of such urgent affairs Hermogenes died and Helpidius was promoted
to the prefecture, a man born in Paphlagonia, ordinary in appearance and speech, but
of a simple nature, so averse to bloodshed and so mild that once when Constantius
had ordered him to torture an innocent man in his presence, he quietly asked that his
office might be taken from him and these matters left to more suitable men, to be
carried out according to the sovereign’s mind.’ Amm. 21.6.9)317
(g) Ecclesia autem, ibi que (= quae) est, ingens et valde pulchra et nova dis-
positione, ut vere digna est domus Dei.
(‘Moreover the church, which is there, is great and truly beautiful and has a new
arrangement, as is a house truly worthy of God.’ Pereg. 19.3)
When the content of the si clause is presented as a fact, the indicative is used. There
are no restrictions on the sentence type and the mood of the apodosis. The apodosis
317
See Blomgrén (1937: 56ff.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
of (a), for example, contains a future indicative form, and that of (b) a present
subjunctive in a question.
(a) Dixi equidem: sed si parum intellexti, dicam denuo.
(‘I’ve told you, but if you didn’t understand it fully I’ll say it again.’ Pl. Rud. 1103)
(b) Ergo ad cenam petitionis causa si quis vocat, condemnetur?
(‘Are we then to condemn everyone who gives an invitation to dinner for this
purpose?’ Cic. Mur. 74)
The choice between presenting a situation as potential (in the present or the past) or
as counterfactual (in the present or in the past) is a subjective one. The use of the
counterfactual option is in the minority in the comedies of Plautus (the potential is used
three times as often), but is predominant in Terence and even more so in Cicero.
However, what manifests itself as a diachronic evolution may also reflect a difference in
text type. In poetry the potential expression is preferred because ‘the present and perfect
subjunctive represent a hypothetical condition more vividly by not excluding the idea of
fulfilment’.318 The choice between the potential and the counterfactual expression has
nothing to do with whether the events referred to actually take or took place: both
expressions are assumptions. An interesting pair often cited in the literature are (a)
potential and (b) counterfactual. In (a) ‘Cicero is suggesting the possibility that one’s
country may appeal to one, and is not excluding it.’ In the context of (b) ‘he goes on to
say that, since the whole province cannot speak for itself, it is necessary for it to choose a
representative. Therefore he is careful to use the tense which does not admit the
possibility.’319 Both situations are equally impossible in reality, and it is the speaker
who decides which expression is the most suitable in a given situation.320
(a) Haec si tecum ita . . . patria loquatur, nonne impetrare debeat . . . ?
(‘If our country speaks to you thus, ought she not to obtain her request . . . ?’ Cic.
Catil. 1.19)
(b) Sicilia tota si una voce loqueretur, hoc diceret . . .
(‘If all Sicily spoke with a single voice, this is what she would say . . . ’ Cic. Div. Caec. 19)
Since the potential and the counterfactual expressions are fictitious, the correctness of
the assumption can be denied in the following context, as is shown in (c) and (d).321
318 319
Woodcock (1959: 155); cf. Lavency (1999: 387). Woodcock (1959: 154).
320
See Lavency (1999, 2005). Baratin (1981) is a representative of an approach that takes the
extralinguistic reality too much into account.
321
See Lavency (1999: 380).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(c) Summa in utroque est honestas, summa dignitas. Quam ego, si mihi per
Servium liceat, pari atque in eadem laude ponam. Sed non licet.
(‘In each of the two men there is the greatest honour, the greatest merit; and,
if Servius were to allow me, I would hold that of each man in exactly the
same esteem. But he does not allow me.’ Cic. Mur. 21)
(d) . . . sapientia, quae ars vivendi putanda est, non expeteretur, si nihil efficeret.
Nunc expetitur . . .
(‘ . . . wisdom, which must be considered as the art of living, if it effected no
result would not be desired; but as it is, it is desired . . . ’ Cic. Fin. 1.42)
There is overlap in the forms of the potential and the counterfactual. Both the past
potential and the present counterfactual are expressed by the imperfect subjunctive.
The four possibilities of the potential and counterfactual use of the subjunctive are
given in Table 7.22, with the illustrative examples (e) and (f )—potential, and (g) and
(h)—counterfactual. It is not always easy to distinguish the past potential from the
present counterfactual. Sometimes contextual information helps, as tum in (i) (past
potential) and nunc in (j), which locates the counterfactual situation at the time of
speaking. (Note that the preceding sentence is past potential.) In (k) the si clause is
past counterfactual, the main clause past potential. The past potential is relatively
frequent in Early Latin, rare in later periods, but then is being used more often once
again in Late Latin. Whether this reflects a diachronic development is unclear.322
322
See Lavency (1999: 384, 387).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
323
Manuwald ad loc. takes esses as equivalent to fuisses.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(‘If, in a case for the recovery of goods which have been sold under contract, a man
who has assumed the contractual obligation should incur the rest of a court decision,
then it is surely even more appropriate in the trial of a consul-designate that the
consul above all who has declared his election should be the guarantor of the office
conferred by the Roman people and defend it from all danger.’ Cic. Mur. 3)
(d) Lippi illic oculi servos est simillimus. / Si non est, nolis esse nec desideres. /
Si est, apstinere quin attingas non queas.
(‘That slave is very similar to a bleary eye; if you don’t have one, you don’t want or desire
to have one. If you do have one, you can’t refrain from touching it.’ Pl. Bac. 913–15)
(e) Intro abi ergo et si isti est mulier, eam iube / cito domum transire . . .
(‘Then go in and if she’s there, have her return home quickly . . . ’ Pl. Mil. 255–6)
(f ) Si volebas participari, auferres dimidium domum.
(‘If you wanted to have your share, you should have taken half home.’ Pl. Truc. 748)
(g) Si sciens fallo, tum me, Iuppiter Optime Maxime, domum, familiam remque
meam pessimo leto adficias.
(‘If I wittingly speak false, may Jupiter Optimus Maximus utterly destroy me, my
house, my family, and my estate.’ Liv. 22.53.11)
(h) Tum id si falsum fuerat, filius / quor non refellit?
(‘If it had been false, why didn’t your son refute it then?’ Ter. Ph. 400–1)
(i) Etenim si ille tali ingenio exitum non reperiebat, quis nunc reperiet?
(‘For if a man of Caesar’s genius could find no way out, who will find one now?’ Cic.
Att. 14.1.1)
(j) Quid igitur timeam, si aut non miser post mortem aut beatus etiam futurus
sum?
(‘What, then, am I to fear, if after death I am going to be either not miserable, or even
happy?’ Cic. Sen. 67)
Main clauses combined with a conditional clause in the subjunctive are often also in
the subjunctive (and mostly have the same tense), but there are no restrictions, just as
with the indicative.324 Examples of subjunctives in both clauses can be found in § 7.40
(potential) and § 7.47 (counterfactual). With si clauses that contain a potential
subjunctive, the main clause may be of any type (declarative, imperative, or inter-
rogative), and there are no restrictions on mood and tense. There is considerable
variation among authors and among text types in the relative use of the patterns ‘si
sit . . . sit’ (both subjunctive) and ‘si sit . . . est/erit’ (subjunctive . . . indicative). In prin-
ciple these patterns have different meanings (the usual semantic difference between
subjunctive and indicative). The latter type, however, seems to become more common
324
A very detailed survey of the combinations attested in Early Latin can be found in Bennett (1910:
I.273ff.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Examples of the potential conditional clause referring to the past with an indicative
main clause are (r)–(t). Ex. (t) is interpreted as habitual or iterative.
(r) Tetigistin’ foris? / # Quo modo pultare potui, si non tangerem?
(‘Did you touch the door? # How could I have knocked without touching it?’
Pl. Mos. 461–2)
325
The figures can be found in Blase (1896a: 25).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(s) Quid enim poterat Heius respondere, si esset improbus, si sui dissimilis?
(‘For what could Heius have answered if he were a wicked man, unlike the man he
is?’ Cic. Ver. 4.16)
(t) Fugitivis omnibus nostris certus erat Alexandriae receptus certaque vitae
condicio . . . Quorum siquis a domino prehenderetur, consensu militum
eripiebatur . . .
(‘All our own fugitive slaves had a sure place of refuge at Alexandria and
assurance of their lives on the condition . . . And if any one of them was arrested
by his owner, he would be rescued by the common consent of the soldiery . . . ’
Caes. Civ. 3.110.4)
Supplement:
Inimicum habebas neminem. Si haberes, tamen non ita vixeras ut metum iudici
propositum habere deberes. (Cic. Ver. 5.74);
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(aa) Sed quid ego haec interrogo, cum, si ego tacitus abissem, tamen tibi ad
voluntatis interpretationem meae dirigenda tua sententia fuerit?
(‘But why do I put these questions, since, had I gone off without a word,
nevertheless your thoughts should have been directed to the interpretation
of my will.’ Liv. 8.32.5)
(ab) . . . ostendis qualis tu, si ita forte accidisset, fueris illo tempore consul
futurus.
(‘ . . . you demonstrate what sort of consul you, had it so happened, would
have been at that crisis.’ Cic. Pis. 14)
Supplement:
Ceterum haud dubium fuit quin, nisi ea mora intervenisset, castra eo die Punica capi
potuerint. (Liv. 24.42.3); Quis enim dubitat quin, si Saguntinis obsessis fidemque
nostram implorantibus impigre tulissemus opem . . . totum in Hispaniam aversuri
bellum fuerimus (Liv. 31.7.5);
. . . adeoque inopia est coactus Hannibal ut, nisi cum fugae specie abeundum
timuisset, Galliam repetiturus fuerit . . . (Liv. 22.32.3); . . . fecerunt ut, si admotus
extemplo exercitus foret, capi castra potuerint. (Liv. 43.4.1); Adeoque parata apud
malos seditio . . . ut postero iduum die redeuntem a cena Othonem rapturi fuerint, ni
incerta noctis et tota urbe sparsa militum castra nec facilem inter temulentos con-
sensum timuissent . . . (Tac. Hist. 1.26.1);
Cf.: . . . ut vix decem dierum . . . frumentum superesset Hispanorumque ob inopiam
transitio parata fuerit, si maturitas temporum exspectata foret. (Liv. 22.40.9);
Scis quid hinc porro dicturus fuerim, ni linguae moderari queam. (Pl. Per.
296–7—NB: unique instance in Early Latin; note present subjunctive queam); . . .
quaero, si . . . hanc quoque mortiferam plagam inflixisses auguratus tui, utrum de-
creturus fueris . . . (Cic. Vat. 20); . . . ‘dic agedum’, inquit, ‘Appi Claudi, quidnam
facturus fueris, si eo tempore quo C. Furius et M. Geganius censores fuerunt censor
fuisses.’ (Liv. 9.33.7);
Licet was originally a modal verb governing a subjunctive. When its use as a
subordinator had developed, the clause in which it occurred was in the subjunctive
mood, but gradually the indicative came to be used as well (for a variety of reasons,
similar to those with quamvis, as mentioned already). Quamquam is mostly found
with the indicative, but the subjunctive is found when it is semantically or syntactic-
ally justified. Only in later periods is the subjunctive found without a semantic or
syntactic justification. With the si compounds both moods are found, essentially with
the same values as in conditional si clauses. Here too the subjunctive became more
common as the mood of subordination as such, without a specific semantic or
syntactic justification. Details can be found in §§ 16.68–16.82.326
326
For the use of the moods in concessive clauses, see Spevak (2002).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
However, the moods in the two clauses may also be different, with the mood in each
clause having its own semantic justification. Very few examples with the subjunctive
in the quam clause are cited in the literature for Early and Classical Latin, a few more
for later periods. An example is (e) with counterfactual vellem. Comparable is (f ),
where the comparison is on the noun phrase level and the subjunctive represents
Messius’ thoughts (‘oblique’ use of the subjunctive—see § 7.128.init.(iv)). Sometimes
in such cases it is difficult to establish the semantic value of the subjunctive.
(e) . . . serius misi litteras quam vellem . . .
(‘I sent my dispatch later than I should have wished . . . ’ Cic. Fam. 3.9.4)
(f ) . . . alteram (sc. legem) Messius, qui omnis pecuniae dat potestatem et
adiungit . . . maius imperium in provinciis quam sit eorum qui eas obtineant.
(‘Messius proposed an alternative bill which gives him control over all moneys
and in addition . . . authority in the provinces superior to that of their governors.’
Cic. Att. 4.1.7)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
. . . Segestanis praeter ceteros imponebat aliquanto amplius quam ferre possent. (Cic.
Ver. 4.76); Ut vero Pansae consulis accessit cohortatio gravior quam aures Ser. Sulpici
ferre didicissent, tum vero denique filium meque seduxit . . . (Cic. Phil. 9.9); Quin pro
eo quod pluribus verbis vos quam vellem fatigavi veniam a vobis petitam impetra-
tamque velim, patres conscripti. (Liv. 38.49.13); Est tamen aliquis minor, quam in
sinu eius condenda sit civitas. (Sen. Ben. 2.16.2); Nam efficacius citius ardentius
natura mortalium culpat quam laudet. (Sidon. Ep. 4.18.6);327
In the examples above with potius quam, the uses of the subjunctive are the same as
the ones that are found in simple sentences. However, the subjunctive is also
regularly found in comparative clauses, especially with potius quam, without such
a semantic justification. This use of the subjunctive resembles its use in temporal
clauses with future oriented subordinators such as priusquam (see § 7.141). The
quam clause describes a state of affairs that is to be avoided. The main clause may
be declarative, as with the indicative mood in (g) and (h); or imperative, with an
imperative mood in (i) and a subjunctive in (j). The same use of the subjunctive in
the quam clause is found when the main clause is an accusative and infinitive (k) or
a prolative infinitive (l).328
(g) Et, si hunc videbo non dare argentum tibi / quod dixit, potius quam id non
fiat ego dabo.
(‘And if I see that he doesn’t give you the money he said he would, I’ll give it to you
myself rather than that it shouldn’t happen at all.’ Pl. Ps. 553–4)
(h) Zeno proponatur Eleates, qui perpessus est omnia potius quam conscios
delendae tyrannidis indicaret.
(‘Let him call up the image of Zeno of Elea who endured every torment rather than be
brought to divulge his accomplices in the plot to overthrow tyranny.’ Cic. Tusc. 2.52)
(i) . . . potius quam litis secter aut quam te audiam, / . . . / abduce hanc, minas
quinque accipe.
(‘ . . . rather than take you to court or listen to your pleading, . . . take her away. Here
are five minas.’ Ter. Ph. 408–10)
(j) . . . optemus potius ut eat in exsilium quam queramur.
(‘ . . . let us hope, then, that he is going into exile, and not complain about it.’ Cic.
Catil. 2.16)
(k) . . . mansurum (sc. se) potius quam . . . vitae parceret.
(‘ . . . he would rather stand his ground than . . . spare his life.’ Sal. Jug. 106.3)
327
For more instances from Sidonius and other later authors, see Mohr (1889).
328
For more instances of potius see TLL s.v. potis 351.33ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(l) . . . inopemque optavit potius eum relinquere / quam eum thensaurum com-
monstraret filio.
(‘He wished to leave him penniless rather than show this treasure to his son.’ Pl. Aul.
11–12)
Supplement:
Atque occidi quoque / potius quam cibum praehiberem. (Pl. Ps. 367–8); Itaque in
hoc genus coloni potius anniversarios habent vicinos quibus imperent, medicos,
fullones, fabros, quam in villa suos habeant . . . (Var. R. 1.16.4); ‘Depugna’ inquis
‘potius quam servias’. (Cic. Att. 7.7.1); . . . cur non in proelio cecidisti potius quam in
potestatem inimici venires? (Nep. Eum. 11.4); . . . nos potius nostro delicto plectemur,
quam res publica tanto suo damno nostra peccata luat. (Liv. 8.7.15);
‘Prius undis flamma[m]’, ut ait poeta nescio quis, prius denique omnia quam
aut cum Antoniis res publica aut cum re publica Antonii redeant in gratiam. (Cic.
Phil. 13.49);
With an (accusative and) infinitive: Thermitani miserunt qui decumas emerent agri
sui magni sua putabant interesse publice potius quamvis magno emi quam in
aliquem istius emissarium inciderent. (Cic. Ver. 3.99); Quaero etiam, ille vir bonus,
qui statuit omnem cruciatum perferre, intolerabili dolore lacerari potius quam aut
officium prodat aut fidem. (Cic. Luc 23); . . . ideo potius delectos patrum ad eum
missos quam legatis eius Romae daretur responsum . . . (Liv. 2.15.2);
With other comparative expressions: Moriri sese misere mavolet / quam non
perfectum reddat quod promiserit. (Pl. As. 121–2); Hiemi fluctibusque sese commit-
tere maluit quam non istam communem Siculorum tempestatem calamitatemque
vitaret. (Cic. Ver. 2.91); Praestare omnes perferre acerbitates quam non civibus
Romanis qui Cenabi perfidia Gallorum interissent parentarent. (Caes. Gal. 7.17.7);
Different are instances in which the subjunctive in a quam clause is due to assimi-
lation to the subjunctive in a governing clause, as in the following examples: Cum
vultus Domiti cum oratione non consentiret atque omnia trepidantius timidiusque
ageret quam superioribus diebus consuesset . . . (Caes. Civ. 1.19.3); Sed etiam, cum
esset in urbe et propter infinitas suas occupationes minus saepe quam vellet Attico
frueretur, nullus dies temere intercessit quo non ad eum scriberet . . . (Nep. Att.. 20.2);
Cum magis vellet credere quam auderet . . . (Liv. 32.11.5); His diverso vitiorum
genere grassantibus adeo se abutendum permisit et tradidit, ut vix sibi ipse constaret,
modo acerbior parciorque, modo remissior ac neglegentior quam conveniret prin-
cipi electo atque illud aetatis. (Suet. Gal. 14.2).
Difficult to explain is the use of the subjunctive in a clause B that is itself embedded in
a clause with a subjunctive A, when there is no obvious semantic justification for the
use of the subjunctive in clause B. Instances of this phenomenon are almost restricted
to Cicero and to a lesser extent Caesar, with only a few instances cited from Plautus
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
and from later authors. This phenomenon is often called MODAL ASSIMILATION or
‘MODAL ATTRACTION’. Whereas some scholars consider it a form of mechanical copying
of the mood of the governing clause that is allowed when there is no obvious reason
for using the indicative, others consider it a stylistic preference.329 Examples are (a),
(b), (c), and (e). Note the use of the indicative in (d).
(a) Numquam hercle deterrebor / quin viderim id quod viderim.
(‘I’ll never be deterred from having seen what I have seen.’ Pl. Mil. 369–70)
(b) Utinam te priusquam oculis vidissem meis / malo cruciatu in Sicilia
perbiteres . . .
(‘If only you’d perished in torments in Sicily before I’d set eyes on you!’ Pl. Rud.
494–5)
(c) De re publica valde timeo, nec adhuc fere inveni qui non concedendum
putaret Caesari quod postularet potius quam depugnandum.
(‘The political situation alarms me deeply, and so far I have found scarcely anybody
who is not for giving Caesar what he demands rather than fighting it out.’ Cic. Att.
7.6.2)
(d) Nam ego is sum qui illi concedi putem utilius esse quod postulat quam signa
conferri.
(‘For I am one of those who hold it more expedient to concede his demands than to
join battle.’ Cic. Att. 7.5.5)
(e) His rebus gestis omni Gallia pacata tanta huius belli ad barbaros opinio
perlata est uti ab iis nationibus quae trans Rhenum incolerent legati ad
Caesarem mitterentur . . .
(‘These achievements brought peace throughout Gaul, and so mighty a report of this
campaign was carried to the natives that deputies were sent to Caesar from the tribes
dwelling across the Rhine . . . ’ Caes. Gal. 2.35.1)
Supplement:
Haec curata sint fac sis . . . / . . . atque ut ministres mi, mihi cum sacruficem. (Pl. Am.
981–3); Tu ut subservias / orationi, utquomque opus sit, verbis vide. (Ter. An.
735–6); Tum mittit, si videatur ut reddat. (Cic. Ver. 4.66); Primum (sc. fuit in
Hortensio) memoria tanta quantam in nullo cognovisse me arbitror ut quae secum
commentatus esset, ea sine scripto verbis eisdem redderet quibus cogitavisset.
(Cic. Brut. 301); Cumque (sc. Demosthenes) ita balbus esset ut eius ipsius artis cui
studeret primam litteram non posset dicere, perfecit meditando, ut nemo planius esse
locutus putaretur. (Cic. de Orat. 1.260); . . . legati venerunt, qui se de superioris
temporis consilio excusarent . . . seque ea quae imperasset facturos pollicerentur.
(Caes. Gal. 4.22.1);
329
A survey of the various explanations can be found in Bertrand-Dagenbach (1995). For another use
of the term ‘attraction’, see § 7.7.init.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The indicative is obligatory if doubt might arise as to the factual status of the content
of the clause. A few examples will suffice.
(f ) Nam illud quidem statim curatur, ut quicquid caelati argenti fuit in illis
bonis ad istum deferatur, quicquid Corinthiorum vasorum, stragulae vestis.
(‘For one thing was promptly seen to—namely, that whatever engraved silver there
was among their goods be brought to Verres, along with all the engraved silver plate
and Corinthian brass and tapestries.’ Cic. Ver. 2.46)
(g) Quotus enim quisque philosophorum invenitur qui sit ita moratus, ita animo
ac vita constitutus ut ratio postulat?
(‘How few philosophers are found to be so constituted and to have principles and a
rule of life so firmly settled as reason requires?’ Cic. Tusc. 2.11)
(h) Cur paupertas aliarum sub hac legis specie latet, ut quod habere non possunt
habiturae, si liceret, fuisse videantur?
(‘Why does the poverty of other women lie concealed under cover of this law, that
they may seem, if it were legal, to have owned what they are unable to own?’ Liv.
34.4.14)
330
The examples are taken from Handford (1947: 152).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(d) Is (sc. Divico) ita cum Caesare egit: Si pacem populus Romanus cum
Helvetiis faceret, in eam partem ituros atque ibi futuros Helvetios ubi eos
Caesar constituisset atque esse voluisset.
(‘Divico treated with Caesar as follows: If the Roman people would make peace with
the Helvetii, they would go whither and abide where Caesar should determine and
desire.’ Caes. Gal. 1.13.3)
The explanation that is commonly given for this use of the subjunctive is that it
reflects the thoughts of the speaker/writer whose words are reported. In that view it
resembles the oblique use of the subjunctive dealt with in § 7.128.init.(iv).331 This
explanation sometimes seems plausible (so, for example, in (d) above), but it does not
work everywhere. As for a diachronic explanation, Bennett (1910: I. 312, 315–16)
suggests that mood assimilation (see § 7.162) is at the basis of the phenomenon: apud
me sis, miles quom veniat, volo. > . . . apud me te esse ob eam rem, miles quom veniat,
volo (Pl. Bac. 85). The subjunctive then spread to other verbs that govern an
accusative and infinitive clause.
331
See K.-St.: II.542; Sz.: 547.
332
Examples from Cicero can be found in Lebreton (1901a: 365–72). See also Wiesthaler (1956: 84–9).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Dixeritne L. Vettius in contione tua rogatus a te sese auctores et inpulsores et socios
habuisse sceleris illius eos viros quibus e civitate sublatis, quod tu eo tempore
moliebare, civitas stare non posset? (Cic. Vat. 24); Nisi forte existimatis, posteaquam
iudicia severa Romae fieri desierunt, Graecos homines haec venditare coepisse . . .
(Cic. Ver. 4.133);
Another type of clauses in which the indicative is regular are restrictive relative
clauses, as in (c), and autonomous relative clauses determined by a preceding or
following form of is, as in (d) and (e).
(c) Inveteravit enim iam opinio . . . his iudiciis quae nunc sunt pecuniosum
hominem, quamvis sit nocens, neminem posse damnari.
(‘For a belief has become widespread . . . that, in these courts as they are currently
constituted, it is impossible for a man with money, no matter how guilty he may be,
to be convicted.’ Cic. Ver. 1)
(d) Quas ob res ita censeo: Eorum qui cum M. Antonio sunt, qui ab armis
discesserint . . . eis fraudi ne sit quod cum M. Antonio fuerint.
(‘Therefore, I give my vote, “That of those men who are with Marcus Antonius, those
who abandon his army . . . shall not be liable to prosecution for having been with
Antonius.” ’ Cic. Phil. 8.33)
(e) Statuit tamen nihil sibi . . . gravius esse faciendum quam ut illa ne uteretur, ne
quae videre sine summo animi dolore non poterat, ea si matre uteretur non
solum videre, verum etiam probare suo iudicio putaretur.
(‘However, he came to the conclusion that . . . he ought to take no stronger steps
than merely to refrain from all contact with such a mother; lest the very things
he could not look upon without anguish he might be thought, if he kept in
contact with her, not merely to look upon but even to stamp with his approval.’
Cic. Clu. 16)
Supplement:
. . . homini Romae in foro denuntiat fundum illum de quo ante dixi, cuius istum
emptorem demonstravi fuisse mandatu Caesenniae, suum esse seseque sibi
emisse. (Cic. Caec. 19); Sequitur illud quod a Milonis inimicis saepissime
dicitur, caedem in qua P. Clodius occisus est senatum iudicasse contra rem
publicam esse factam. (Cic. Mil. 12); Quem egomet dicere audivi <non> tum se
fuisse miserum . . . cum parva navicula pervectus in Africam, quibus regna ipse
dederat, ad eos inops supplexque venisset. (Cic. Red. Pop. 20); Nolite enim
putare, quemadmodum in fabulis saepenumero videtis, eos qui aliquid impie
scelerateque commiserunt agitari et perterreri Furiarum taedis ardentibus. (Cic.
S. Rosc. 67);
CHAPTER 8
Negation
Latin has three devices to form negative clauses. One device consists of NEGATION
ADVERBS such as non and haud ‘not’. A second device consists of ZERO QUANTIFIERS, also
called negative indefinite pronouns, adjectives, and adverbs, such as nemo ‘nobody’,
nullus ‘no’, and nusquam ‘nowhere’, respectively.1 NEGATIVE VERBS are a third device,
such as nescio ‘not know’ and nolo ‘not wish’. Through its imperative forms nolo
also plays a role in negative imperative sentences. Examples of these three types are
(a)–(c).
(a) Iam dudum est intus. # Hac quidem non venit. # Angiporto / illac per
hortum circum iit clam . . .
(‘He’s already been inside for a long time. # Well, he didn’t come this way. # He
secretly went around by the alley that way through the garden.’ Pl. As. 741–2)
(b) Nam nemo ad te venit / nisi cupiens tui.
(‘Nobody comes to you who does not desire you.’ Ter. Hec. 141–2)
(c) Nescio pol quae illunc hominem intemperiae tenent.
(‘I just don’t know what sort of delusions hold that man in their grip.’ Pl. Aul. 71)
In the second sentence of (a), the negation adverb non ‘not’ signals that the event of
hac quidem venit has not taken place. In (b) and (c), the negation element is
incorporated in a specific constituent, a pronoun and a verb, respectively. In (b), it
is asserted that ‘it is not the case that anyone comes to you who does not like you’. In
(c), it is asserted that ‘it is not the case that I know what sort of delusions hold that
man in their grip’. Whereas (a)–(c) exemplify CLAUSE NEGATION, the negation adverbs
non and haud are also used for so-called LOCAL NEGATION. This is demonstrated by (d).
(d) Fuit igitur in Catulo sermo Latinus. Quae laus dicendi non mediocris ab
oratoribus plerisque neglecta est.
(‘As I was saying then, Catulus had a pure Latin diction, no small merit of style, to
which most orators pay little attention.’ Cic. Brut. 133)
1
So Bertocchi et al. (2010: 72–3).
Negation
In (d), non is part of the adjective phrase non mediocris, which itself functions as the
attribute of laus dicendi. The whole noun phrase non mediocris laus dicendi is in turn
determined by the relative determiner quae. Here, non and mediocris together form a
complex adjectival expression with a meaning that is more or less comparable to the
simple adjective magna ‘great’ and with which it is syntactically interchangeable. Non
is a modifier of mediocris, comparable to some extent with a degree adverb like valde
in valde mediocris. Unlike (a)–(c), the clause as a whole is not negative.
A further distinction that must be made is that between SEMANTIC NEGATION and
2
PRAGMATIC NEGATION. Semantic negation occurs when a sentence (or clause) contains
an explicit marker of negation, pragmatic negation when a sentence (or clause) is
interpreted as negative by implication. Examples of semantic negation are (a)–(c).
Another example is (e): the negation adverb non is responsible for the negative
interpretation. Ex. (f) is a so-called rhetorical question (see § 6.23), implying that
nobody is equally miserable. Note that both in semantically and in pragmatically
negative sentences, the non-specific indefinite pronoun quisquam ‘anyone’, ‘any’ and
the non-specific indefinite adverb usquam ‘anywhere’ are used. Pragmatic negation is
discussed in § 8.50.
(e) Non esse servos peior hoc quisquam potest / nec magis vorsutus nec quo ab
caveas aegrius.
(‘There can’t be a worse slave than this one, or one cleverer and more difficult to be
on one’s guard against.’ Pl. As. 118)
(f) An quisquam usquam gentium est aeque miser?
(‘Is there anyone in the whole world as miserable as I am?’ Ter. Hec. 293)
2
For the distinction and its application to Latin, see Orlandini (2001).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(j) Nam Polybius . . . ex decem nobilissimis qui tum erant missi novem revertisse
dicit re a senatu non impetrata . . .
(‘Polybius . . . states that of the ten eminent nobles who were sent at that time, nine
returned when their mission had not been obtained from the senate.’ Cic. Off. 3.113)
3
For the use of the term ‘scope’ in the discussion of negation, see Ramat (1999: 242).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation adverbs are not evenly distributed over sentence (and clause) types. The
most common negation adverbs in declarative sentences and clauses are non and
haud. In imperative sentences and clauses nē is the normal negation word. This is also
used in certain subordinate clauses, both in argument clauses and in satellite clauses.
In interrogative sentences and clauses non is the most common negation adverb.
Haud is not used as a clause negator in interrogative sentences and clauses. Nī is
found in Early Latin in a special context (see § 8.15). There are moreover a number of
(interrogative and relative) negative reason adverbs. Nē, quin, and quominus (also
quo . . . minus, cf. si . . . minus) are negative subordinators. However, in negative con-
texts they are also used without a negative implication. Details are given in the
following sections about the three sentence types.
4
For the diversity of communicative functions of negative sentences, see Touratier (2007: 13–15),
where also references can be found to studies in which more communicative functions of negative
sentences are discussed.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
is the adverb non. It may also be used without further sentential support, as in (g) and
(h). In comedy, curses are a frequent companion as an asseverative device. It may be
assumed that this use of non was marked by intonation and/or word order (sentence-
initial). The negation functions in this case at the level of interaction between
participants in a communicative situation.5
(d) (After a long exchange) . . . Quid tibi surrupui? # Redde huc sis! # Quid tibi
vis reddam? . . . Non hercle equidem quicquam sumpsi nec tetigi.
(‘What did I steal from you? # Give it back, will you! # What do you want me to give
back to you? . . . By Hercules. I didn’t take or touch anything.’ Pl. Aul. 634–40)
(e) Si mihi / alia mulier istoc pacto dicat, dicam esse ebriam. / # Non ecastor
falsa memoro.
(‘If any other woman were to speak to me like this, I’d say she’s drunk. # Honestly,
I’m not telling lies.’ Pl. Cist. 666–8)
(f) Asside hic, pater. / # Non sedeo istic. Vos sedete. Ego sedero in subsellio.
(‘Do sit down here, father. # I won’t sit there; you two should sit there. I’ll sit on the
bench.’ Pl. St. 92–3)
(g) Quid iam? # Quia enim non venalem iam habeo Phoenicium. / # Non habes?
# Non hercle vero.
(‘How so? # Because I don’t intend to sell Phoenicium anymore. # You don’t? # By
Hercules, no, I don’t.’ Pl. Ps. 325–6)
(h) Non edepol istaec tua dicta nunc in auris recipio. / # Non? Hem. Quid agis
igitur?
(‘By Pollux, I’m not letting those words of yours get into my ears now. # You aren’t?
What’s that? What are you doing then?’ Pl. Cist. 510–11)
Most often the negation concerns only one or more words in the clause to which it
belongs. The constituent in question is either a contrastive topic or a focus (contrast-
ive or completive). In (i), there is a contrast between ego and ille mentioned in the
preceding sentence. Whereas the latter has seen the twins that this text is about, ego
has not. In its context, then, ego is a contrastive topic. The focal information is non
vidi, with non immediately preceding the verb vidi. The scope of the negation is ego: it
is not the case that ‘I have seen’. In (j), the scope of haud is huc (note quidem). In (k),
the ablative absolute clause me . . . auctore is the scope of non (note quidem). Ex. (l)
looks like a refusal, but the following reaction shows that this is still an exchange of
opinions. Both here and in (k) non is in initial position for the sake of emphasis. In
(m), the scope of non in the ut argument clause is the phrase et in hoc magistratu et in
5
Latin has no lexical positive counterpart of this use of non (as e.g. wel in Dutch Hij is wel gekomen). In
Functional Grammar denials and refusals are described as ‘propositional negation’, whereas the type of
negation described below and in the remainder of this section is called ‘predicational negation’. See Dik
(1997: II.ch. 8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
omni vita, marked as focal by the parallel coordination et . . . et. Ex. (n) is part of a
discussion about the presence of water in mountainous regions. Pliny states that there
is a widespread opinion that there is plenty of water on the north sides of mountains.
His argument is that there are indeed cases in which this is true, but that nature has
more variation. The example (the first in Pliny’s argumentation) starts with a setting
constituent In Hyrcanis montibus, followed by the contrastive topic a meridiano
latere. The scope of non is a meridiano latere. Again, non immediately precedes the
verb pluit. As the examples show, the scope may be on a nuclear constituent (ego is an
argument, and huc may be one as well) or on an adjunct (me auctore, et in hoc
magistratu et in omni vita, and a meridiano latere). The scope may concern one word
or a subordinate clause. Ex. (k) illustrates an ablative absolute clause. Ex. (o) has a
finite reason clause, with non immediately preceding the subordinator quia.
(i) Ut quidem ille dixit mihi qui pueros viderat. Ego illos non vidi, nequis
vostrum censeat.
(‘At least someone who’d seen the boys told me so. I haven’t seen them, in case any of
you supposes that I did.’ Pl. Men. 22–3)
(j) Huc quidem hercle haud ibis intro, ne quid –æÆ feceris.
(‘By Hercules, here you won’t go in, don’t play the harpy.’ Pl. Ps. 654)
(k) Non me quidem / faciet auctore hodie ut illum decipiat.
(‘He won’t be acting on my advice today in deceiving him.’ Pl. St. 602–3)
(l) Potare ego hodie, Euclio, tecum volo. / # Non potem ego quidem hercle. # At
ego iussero / cadum unum vini veteris a me afferrier.
(‘I want to drink with you today, Euclio. # I certainly won’t drink. # But I’ll have a jug
of old wine brought here from my place.’ Pl. Aul. 569–71)
(m) Neque enim ullo modo facere possum ut . . . non et in hoc magistratu et in
omni vita essem popularis.
(‘Nor in any way can I . . . help acting as the people’s friend while I hold this office and
throughout my life.’ Cic. Agr. 2.7)
(n) In Hyrcanis montibus a meridiano latere non pluit. Ideo silvigeri ab Aqui-
lonis tantum parte sunt.
(‘In the Hyrcanian mountains it does not rain on the southern slope, and so only on
the north side are there woods.’ Plin. Nat. 31.43)
(o) . . . alienum mea natura videbatur quicquam de existimatione P. Fabi dicere,
non quia res postulare non videretur.
(‘ . . . it appeared foreign to my character to say anything of the reputation of Quintus
Fabius; not because the case did not open the door to such statements.’ Cic. Tul. 5)
Negation
are no attestations of haud being used in this way)—or of a negation adverb and the
constituent to which the negation applies, as in (r)–(t). Note quidem in (r) and (t).
In (u), at introduces an elliptical correction of a conclusion that the speaker presumes
the addressee may draw. The eo quia clause has contrastive focus.
(p) Quid iam? # Quia enim non venalem iam habeo Phoenicium. / # Non habes?
# Non hercle vero.
(‘How so? # Because I don’t intend to sell Phoenicium anymore. # You don’t? # No,
by Hercules, I don’t.’ Pl. Ps. 325–6)
(q) Non edepol istaec tua dicta nunc in auris recipio. / # Non? Hem. Quid agis igitur?
(‘By Pollux, I’m not letting those words of yours get into my ears now. # You aren’t?
What’s that? What are you doing then?’ Pl. Cist. 510–11)
(r) Eho, an dormit Sceledrus intus? # Non naso quidem.
(‘Tell me, is Sceledrus sleeping inside? # Well, not with his nose.’ Pl. Mil. 822)
(s) Quomodo igitur hoc assequar iudices ut sciam quantum lucri factum sit?
Non ex Aproni tabulis quas ego cum conquirerem non inveni . . .
(‘How, therefore, gentlemen, am I to succeed in finding out the amount of the profit
made? Not from the accounts of Apronius, which I did not find when I looked for
them . . . ’ Cic. Ver. 3.112)
(t) . . . animum induxerunt socrus / omnis esse iniquas. Haud pol me quidem.
(‘ . . . they’re so convinced that all mothers-in-law are unreasonable. But not me, for
heaven’s sake.’ Ter. Hec. 277–8)
(u) . . . ea res me male habet, at non eo quia tibi non cupiam quae velis.
(‘ . . . that does make me feel down; but not because I wouldn’t wish you to have what
you want.’ Pl. As. 844)
When the negation adverb does not concern one of the arguments or an adjunct of its
clause, it is difficult to decide whether it has the verb or the entire clause as its scope.
The position of non (the number of instances of haud is too small to allow a detailed
discussion) does not offer a clue since it is usually in front of the finite verb, even if the
scope of the negation is elsewhere. In the examples given above, quidem was found
with constituents in the scope of negation, but it is not often used with verbs.
Examples of non given by Lodge as ‘pertinet ad verbum’ are (v)–(x). For a disputed
instance of quidem see (y).
(v) Istuc facinus quod tu insimulas nostro generi non decet.
(‘The deed you’re accusing me of doesn’t become our family.’ Pl. Am. 820)
(w) Illud mihi verbum non placet, ‘quod nunc habes’.
(‘This phrase, “what you have now”, doesn’t please me.’ Pl. Aul. 547)
(x) Non licet colloqui te? / # At mi non lubet.
(‘Can’t we speak to you? # But I don’t want to.’ Pl. Ps. 252–3)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(y) Non opinor quidem (equidem cj. Halm), quia, si vi deieci M. Claudium, dolo
malo deieci.
(‘I don’t suppose so, because, if I drove out Marcus Claudius by violence, I drove him
out by malice.’ Cic. Tul. 29)
Many instances that are given in the literature to show that the negation concerns the
whole sentence or clause (and which are not of the interactional type mentioned at
the beginning of this section) are not conclusive (as most authors also indicate).6
However, sometimes the entire communication is so emphatic that the negation does
indeed seem to concern the whole sentence or clause. An illustration of this is (z). Non
is in initial position; ego is combined with the focusing particle quidem; usquam and
quicquam are both emphatic; the coordinated pair (non) nuto neque nicto is emphatic
as well, and tibi seems to be focal both because it is in final position and because of the
sentence-internal contrast with ego. Ex. (aa) is a generic statement which is valid in its
entirety, unless one would hold that the words concerned by the negation are istanc
and/or ullum. In (ab), non indeed concerns the sentence as a whole. Ex. (ac), too, has
non in initial position, and it seems to concern the entire following sentence (‘it is not
the case that’);7 however, one might hold that it only concerns idcirco (referring to the
content of the si clause) and that its initial position, away from putabunt, is due to
contrast. An example of a negated question with sentence scope is (ad). In negated
questions preceded by quid? the negation also concerns the entire sentence or clause,
as in (ae).8 Ex. (af) may be an example of haud with an entire clause in its scope.
(z) Non hercle ego quidem usquam quicquam nuto nec nicto tibi.
(‘Not I, not a bit of it, by Hercules! I’m not shaking my head at all, or winking at you,
either.’ Pl. Men. 613)
(aa) Non istanc aetatem oportet pigmentum ullum attingere . . .
(‘Girls of your age shouldn’t touch any colour . . . ’ Pl. Mos. 263)
(ab) ‘Non et natus est quis oriente Canicula et is in mari morietur.’
(‘It is not the case both that some person was born at the rising of the dogstar and
that that person will die at sea.’ Cic. Fat. 15—NB: a formulation ascribed to the
philosopher Chrysippus)
(ac) Non si Opimium defendisti, Carbo, idcirco te isti bonum civem putabunt.
(‘This tribunal, Carbo, is not going to deem you a patriotic citizen just because you
defended Opimius.’ Cic. de Orat. 2.170)
(ad) Fides non reddis? # Nec fides nec tibias.
(‘Won’t you return my lyre to me? # Neither your lyre nor your flute!’ Pl. Epid. 514)
6
So, for instance, Lodge: s.v. non § III.A.6 (1933: II.189A–190A). See also Merguet (Reden) s.v. non
330B–334B ‘bei Verben und Sätzen’.
7
So Touratier (2007: 24–5). A comparable example (Cic. Phil. 8.12) is given in the OLD s.v. non § 1.a,
to illustrate the use of initial word order ‘where negation is concentrated on a particular word or idea’.
8
For more examples, see Merguet (Reden) s.v. non 333A/B.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
Negative declarative sentences and clauses may contain epistemic attitudinal dis-
juncts. In (ag), certe expresses Cicero’s certainty that his opponent could not have said
that Cicero had been condemned. In (ah), certe and fortasse non balance each other.
The adverbs are both part of a declarative argument clause. Haud is very rarely
attested in this way.
(ag) Quid enim diceres? Damnatum? Certe non.
(‘For what could you say? That I had been condemned? Assuredly not that.’ Cic.
Dom. 51)
(ah) Ita fit ut Demosthenes certe possit summisse dicere, elate Lysias fortasse non
possit.
(‘In other words, Demosthenes could certainly speak calmly, but Lysias perhaps not
with passion.’ Cic. Opt. Gen. 10)
Supplement:
Est enim aliquid quod non oporteat, etiamsi licet. Quicquid vero non licet, certe non
oportet. (Cic. Balb. 8); Num is est Cluentius? Certe non est. (Cic. Clu. 148); Profecto
non negabo debere et dabo. (Pl. Mos. 1024); Agros vero Vaticanum et Pupiniam cum
suis opimis atque uberrimis campis conferendos scilicet non putabunt. (Cic. Agr.
2.96); Tu eum videlicet non potuisti videre. (Cic. Fam. 16.17.2);
. . . scilicet haud nobis quicquam, qui non erimus tum, / accidere omnino poterit
sensumque movere . . . (Lucr. 3.840–1);
The negation adverb nē in combination with the focusing particle quidem functions
as a strong negator ‘not even’, signalling by its position the focus function of the
constituents it surrounds. It is used with individual nouns, pronouns, adjectives,
adverbs, and verbs, but also with noun phrases, as in (ai), prepositional phrases,
and rarely also with clauses, as in (aj) and (ak). It usually surrounds the words it
modifies, occasionally with another word inserted, as in (al). In (am), ne . . . quidem
surrounds the conditional subordinator si and the focal constituent of the conditional
clauses.9
(ai) Ne bonum verbum quidem unum dixit.
(‘She didn’t even say one single kind word to me.’ Pl. Truc. 543)
9
For the various positions of ne and quidem, see Devine and Stephens (2006: 266–72).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(aj) Dixi antea. Ne si argueret quidem tum denique, cum esset interrogatus, id
mihi criminosum videretur.
(‘I had already said that not even if he had in the end accused Sulla in reply to a direct
question, would I think it proof of his guilt.’ Cic. Sul. 38)
(ak) Si quidem nos, ne cum volueritis quidem, creare interdum poteritis, istos,
etiam si nolueritis, necesse erit.
(‘To elect us will sometimes be beyond your power, even though you wish it; but
those persons you would be compelled to elect, even against your inclinations.’ Liv.
6.41.3)
(al) Ne hercle operae pretium quidem est / mihi te narrare, tuas virtutes qui
sciam.
(‘By Hercules, it isn’t even worth your while to tell me about it: I know your exploits.’
Pl. Mil. 31–2)
(am) Atque hac voce Q. Fufi capti sumus: ‘Ne si a Mutina quidem recesserit,
audiemus Antonium, ne si in senatus quidem potestatem futurum se
dixerit?’
(‘Moreover, we were taken in by these words of Quintus Fufius: “Shall we not listen
to Antonius even if he withdraws from Mutina, even if he promises to be under the
senate’s control?” ’ Cic. Phil. 12.4)
Negation
(c) Quid modi flendo quaeso hodie facies? # Quid ego ni fleam?
(‘Please, what end will you put to your crying today? # Why shouldn’t I cry?’ Pl. Mil.
1311)
(d) Nimis tandem ego aps te contemnor. / # Quippe ego te ni contemnam,
/ stratioticus homo qui cluear?
(‘I’m really being despised by you. # Why shouldn’t I despise you, I who am a famous
military man?’ Pl. Ps. 916–18)
Apart from the few instances in Plautus where ni is separated from quid, quidni
functions as a negative interrogative adverb of reason, commonly in polemical
questions in the subjunctive with an indirect assertive illocutionary force (for the
subjunctive, see §§ 7.40–7.46). The verb often repeats the verb of the preceding
sentence, as in (e). When used independently it is equivalent to ‘naturally’, ‘of course’.
Another negative interrogative adverb of reason is quin, which often occurs in second
person present tense questions with an indirect directive illocutionary force, as in (f)
(see § 6.24). Both quidni and quin differ from cur non.10
(e) Nostin’? # Quidni noverim?
(‘Do you recognize it? # Why shouldn’t I?’ Pl. Cur. 423)
(f) Quin fles? # Pumiceos oculos habeo.
(‘Why aren’t you crying? # I have eyes as dry as flint.’ Pl. Ps. 75)
The example of quippe . . . ni in (d) above is unique. Elsewhere quippini is one word. It
functions as an interactional particle, indicating that something is a self-evident
certainty. Illustrative is the sequence in (g). It is limited to Plautus, once followed
by Apuleius Met. 9.26.3, but wrongly.11
(g) Impudice. # Ita est. # Sceleste. # Dicis vera. # Verbero. / # Quippini? #
Bustirape. # Certo. # Furcifer. # Factum optume.
(‘You shameless creature! # Indeed. # You criminal! # You speak the truth. #
You whipping stock! # Why not? # You grave-robber! # Certainly. # You
thug! # Perfect.’ Pl. Ps. 360–1)
10
For a discussion of the differences between quidni and cur non, see Hoff (1983: 124). For the
differences between quidni and quin, see Fleck (2008: 82–6).
11
See Schrickx (2011: 99–101).
12
For the historical development of negation particles in prohibitions, see L. Löfstedt (1966: ch. 1).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) and Classical Latin exceptions with non several can be explained as emphatic
negation (so probably (e)), although this is not always easy to prove. However, in the
course of time non simply replaced ne. That this was already well under way in
Quintilian’s time appears from his statement: ‘Someone who says non feceris instead
of ne feceris makes the same type of mistake, because the former is the form of
negation, the latter of prohibition’ (Quint. Inst. 1.5.50). Another means of negating
imperatives is the use of the imperative forms of the negative verb nolo, noli and
nolite, as in (f). Details on prohibitions are given in § 6.28 and § 7.63. NB: Haud is not
used in imperative sentences.
(a) Ne male loquere apsenti amico.
(‘Stop insulting a friend in his absence.’ Pl. Trin. 926)
(b) Proin tu ne quo abeas longius ab aedibus.
(‘So don’t go anywhere too far from the house.’ Pl. Men. 327)
(c) Verum postremo impetravi ut ne quid ei suscenseat.
(‘Well, in the end I got him to agree not to be angry with him.’ Pl. Bac. 533)
(d) Hic est ille. Non te credas Davom ludere.
(‘This is the man. Don’t imagine you’re fooling Davos.’ Ter. An. 787—NB: Priscianus
has ne, and in that case the clause could be pseudofinal—tr. Brown)
(e) Interea quidem, per deos immortales, quoniam omnia commoda nostra, iura
libertatem salutem denique legibus obtinemus, a legibus non recedamus . . .
(‘But in the meanwhile, in Heaven’s name, since it is the laws that give us all our
advantages, our rights, our freedom, and our security—let us not abandon laws.’ Cic.
Clu. 155)
(f) Noli sis tu illi advorsari.
(‘Just don’t oppose him.’ Pl. Cas. 204–5)
The earliest example with non mentioned in the literature is found in a fragment of
Plautus—(g) below. Obviously, it is not a prohibition and it need not even be
emphatic. (Instead of non, some editors read ne or nolo.)
(g) Exi tu, Dave, age sparge. Mundum hoc esse vestibulum volo. / Venus
ventura est nostra. Non hoc pulveret.
(‘You, Davus, come out and sprinkle it. I want this hallway to be neat. Our
Venus is coming. This should not be full of dust.’ Pl. frg. 164–5 = Gel.
18.12.4)
Negation
example is (a): Regulus was of the opinion that the prisoners (including himself)
should not be restored. Ex. (b) is similar. In (c), the speaker expects that those who
have weapons will not allow it.13 In (d), non relates to the content of the preceding
question. The effect of negator climbing may be to make the negation less direct.14 See
also § 14.2.
(a) Regulus . . . captivos reddendos in senatu non censuit . . .
(‘Regulus . . . made the motion in the Senate that the prisoners should not be
restored . . . ’ Cic. Off. 1.39)
(b) Clodiae sane placent, sed non puto esse venalis.
(‘Clodia’s gardens would be admirable, but I don’t think they’re for sale.’ Cic. Att.
12.38a.2)
(c) Passurine haec isti sint, nescio. Non spero esse passuros illos qui arma
habent.
(‘Whether these people intend to tolerate such conduct, I do not know: I believe that
those who have arms will not endure it.’ Liv. 3.47.7)
(d) Etiamne adveniens complectar eius patrem? # Non censeo.
(‘Should I also embrace her father on my arrival? # I reckon not.’ Pl. Rud. 1277)
The same phenomenon can be observed with negative verbs, as in (e), with the verb
nego, and in (f), with the verb nolo. As the latter example shows, negator climbing also
concerns finite subordinate clauses.
(e) Negat haec filiam me suam esse: non ergo haec mater mea est.
(‘She denies that I’m her daughter: then she isn’t my mother.’ Pl. Epid. 590)
(f) Nolo ames. # Non potes impetrare.
(‘I don’t want you to love me. # You can’t succeed.’ Pl. Cas. 233)
Examples of LOCAL NEGATION are (a)–(d). In (a), non modifies the adjective magnum
and belongs together with magnum to the noun phrase non magnum modum agri.
Non magnum can be replaced by a syntactically equivalent simple adjective like
parvum ‘small’. Similarly, in (b), non goes with eiusdem; In (c), non omnino is more
or less equivalent to paene. In none of these examples could one say that the negation
adverb concerns the clause as a whole or the verb. The negation adverb forms a
13
For more instances, see OLD s.v. non § 2. On negator climbing, see Orlandini (2001: 20–1); Sznajder
(2003: 41–5); Devine and Stephens (2013: 354–8). J. N. Adams (p.c.) has his doubts about (a), (c), and (d).
For (a), compare the more detailed version of the story in Off. 3.99–101.
14
So Devine and Stephens (2013: 358).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
syntactic unit with the word it modifies (and which it usually precedes immediately).
This is the preferred environment for haud. Compare (d). When an adjective
functions as a subject complement, it is not always easy to decide whether we are
dealing with clause negation or with local negation. Ex. (e) is such a case. Ne is rarely
used as a local negator, but when it is so used, it is especially found with quantifying
expressions, as in (f), and, occasionally, with attributive adjectives, as in (g).15 (For the
exceptional use of nec, see § 8.12.)
(a) Agri reliquit ei non magnum modum / quo cum labore magno et misere
viveret.
(‘He did leave him a piece of land, not a big one, though, so that he could live on it
with great toil and miserably.’ Pl. Aul. 13–14)
(b) . . . quod Siculus cum Siculo non eiusdem civitatis (sc. agat), ut de eo praeter
iudices . . . sortiatur.
(‘ . . . a case between one Sicilian and another not from the same city, for that the
praetor must choose judges by lot.’ Cic. Ver. 2.32)
(c) Non omnino iam perii, est relicuom quo peream magis.
(‘I am not yet completely ruined, there’s still something left that allows me to be
ruined even more.’ Pl. As. 233)
(d) Statura hau magna, corpore aquilo est. # Ipsa ea est.
(‘She’s not of great height, and has darkish skin. # It’s her very self.’ Pl. Poen. 1112)
(e) Hic homo non sanu’st.
(‘This man isn’t in his right mind.’ Pl. Mer. 951)
(f) Vitibus sulcos et propagines ne minus p. IIS quoquoversus facito.
(‘Dig furrows and trenches for vines not less than two-and-a-half feet deep and the
same distance wide.’ Cato Agr. 43.2)
(g) (sc. equus si est) . . . genibus rutundis ne magnis, ungulis duris.
(‘ . . . if it has the knees round but not large, and hard hoofs.’ Var. R. 2.7.5)
There are a few combinations in which non modifies a noun. In some of the instances
another explanation is possible, so for instance in (h): nec and non form a sort of
double negation (for which see § 8.43). In (i), non voce is triggered by preceding non
humana voce. More convincing is the much-cited (j). For the rare use of nec = non in
the same construction, see (k). In this context (l) is also mentioned, with negation of a
substantivally used infinitive.16 The entire phenomenon is marginal, but grammatical
(as in modern languages) (for a few instances of litotes, see § 8.8).
15
For further instances, see TLL s.v. ne 312.1ff.
16
For a few more instances, see K.-St.: I.821–2 and Sz: 452. For its use in ecclesiastical authors and in
the Vulgate, see Löfstedt (1976: 151, 1985b: 106).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(h) . . . nec vero (sc. arbitrabatur) aut quod efficeret aliquid aut quod efficeretur
posse esse non corpus.
(‘ . . . that, however, anything capable of acting, or being acted upon, cannot be non-
corporeal.’ Cic. Ac. 1.39)
(i) Nam et illa quae est sexta nostrarum (sc. litterarum) paene non humana voce
vel omnino non voce potius inter discrimina dentium efflanda est.
(‘The sixth letter of our alphabet [the f] has to be blown out through the teeth, and
the voice is hardly human, or rather not a voice at all.’ Quint. Inst. 12.10.29)
(j) . . . scire licet gigni posse ex non sensibus sensus.
(‘ . . . we may be sure that sensation can be produced from not-sensation.’ Lucr. 2.930)
(k) Magna autem differentia est inter mancipi res et nec mancipi.
(‘A great difference exists between things which are saleable by mancipation and
things which are not.’ Gai. Inst. 2.18)
(l) . . . hoc non dolere solum voluptatis nomine appellaret . . .
(‘ . . . he might have confined the name of pleasure to this state of freedom from
pain . . . ’ Cic. Fin. 2.18)
17
For these notions, see Hoffmann (1987: 98–118).
18
For the use of the term in this sense and a discussion of the phenomenon in general, with many
illustrations, see Hoffmann (1987). Orlandini (2001: 203–5) uses the concept also for expressions like non
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) P. Memmius secutus est, qui suum non parvum beneficium commemoravit
in amicos Caecinae.
(‘There followed P. Memmius who recorded the considerable kindness which he had
done to Caecina’s friends.’ Cic. Caec. 26)
(b) Habet enim ille tanquam hiatus et concursus vocalium molle quiddam et
quod indicet non ingratam neglegentiam de re hominis magis quam de
verbis laborantis.
(‘For the hiatus and clash of vowels have something agreeable about it and show a not
unpleasant carelessness on the part of a man who is paying more attention to thought
than to words.’ Cic. Orat. 77)
(c) Tam etsi dominus non invitus patitur, servi murmurant.
(‘Even if the master puts up with it not unwillingly, the slaves mutter.’ Pl. Mil. 744)
(d) Pol profecto haud est dissimilis, meam quom formam noscito.
(‘By Pollux! He’s indeed not dissimilar when I look at myself.’ Pl. Men. 1064)
Words that belong to opposite pairs but are ungradable, such as vivus ‘alive’ and
mortuus ‘dead’ do not occur in litotes. In addition, relational words like ante ‘before’
and post ‘after’, maior ‘greater’ and minor ‘smaller’, augeo ‘to increase’ and minuo ‘to
decrease’ are not normally found in litotes.19
In (a)–(d), litotes is illustrated for adjectives. More examples are given in the
Supplement. Examples of the other word classes that occur in litotes are (e)–(h),
with a verb, two adverbs, and a preposition, respectively. Instances of nouns are rare,
and most in reality function as adverbs (non iniuria ‘not unjustly’, for example). In (i),
one may consider haud labori an instance of local negation,20 with the phrase
functioning as a whole as subject complement, but this is difficult to prove. For litotes
with quantifiers and verbs and adjectives with a negative prefix, see § 8.42.
(e) Non sum aspernatus.
(‘I did not underrate him.’ Cic. de Orat. 2.88)
(f) . . . animadverteram posse pro re nata te non incommode ad me in Albanum
venire III Non. Ian. sed, amabo te, nihil incommodo valetudinis feceris.
(‘ . . . I had observed that it would not be inconvenient, relatively speaking, for you to
visit me at Alba on 3 January. But don’t, I beg you, do anything to the detriment of
your health.’ Cic. Att. 7.8.2)
(g) Hau male illanc amovi . . .
(‘Not badly did I remove that female!’ Pl. Men. 853)
potest fieri quin. For the difference in meaning between a negative expression and its positive counterpart,
see also Fruyt (2002b: 41).
19 20
See Hoffmann (1987: 101–5). So Hoffmann (1987: 81).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(h) . . . iunctae copiae saltusque haud sine clade, maiore tamen iumentorum
quam hominum pernicie, superatus.
(‘ . . . his forces were reunited and surmounted the pass; and though they suffered
some casualties, still they lost more baggage-animals than men.’ Liv. 21.35.1)
(i) Nam hoc mi haud labori est laborem hunc potiri . . .
(‘Well, meeting with this suffering is no suffering to me . . . ’ Pl. Rud. 191—NB: sat
labori’st cj. Müller, followed by OCT)
Supplement:
Dum Libys haud laetus Rhegina ad litora tendit . . . (Sil. 13.94); Campani . . . haud
parva, ut arbitror, accessio bonis rebus vestris in amicitiam venimus vestram. (Liv.
7.30.6); Nec longum tempus, et ingens / exiit ad caelum ramis felicibus arbos /
miraturque novas frondes et non sua poma. (Verg. G. 2.80–2);
Non me indicente haec fiunt. (Ter. Ad. 507—NB: see § 8.51); Nec Praenestinae
fundator defuit urbis . . . (Verg. A. 7.678);
Haud aliter Rutulo muros et castra tuenti / ignescunt irae, duris dolor ossibus
ardet. (Verg. A. 9.65–6); Itaque Iulius Candidus non invenuste solet dicere aliud esse
eloquentiam, aliud loquentiam. (Plin. Ep. 5.20.5); At non Thucydides, ne ille quidem
haud paulo maior scriptor Plato . . . (Cic. Orat. 151);
. . . ut intellegatis primum ex omni numero quam non multi ausi sint dicere,
deinde ex iis ipsis quam pauci fuerint laude digni. (Cic. Brut. 270); Non semel et
satyros eluserat illa sequentes . . . (Ov. Met. 1.692); . . . Fulvius excepit non unum
pectore vulnus. (Sil. 17.304);
Haud mora, constituunt diversis partibus ambae / et gracili geminas intendunt
stamine telas. (Ov. Met. 6.53–4);
Litotes is also possible with adjectives and adverbs representing an extreme (high or
low) degree of the quality they refer to, such as superlative forms and diminutives.
Examples are (j) and (k).
(j) . . . alterum non doctissimum, alterum plane indoctum fuisse.
(‘ . . . that one of them had no great learning and the other none at all.’ Cic. de Orat. 2.7)
(k) Novi non inscitulam / ancillulam . . .
(‘I know a little serving-maid, who is not at all naive . . . ’ Afran. com. 386)
(b) Iucunda certe mihi fuerunt neque solum non longa, sed etiam breviora quam
vellem.
(‘Certainly I found it interesting, and far from being too much, it was even briefer
than I could have wished.’ Cic. Brut. 52)
(c) Nam cum de tribus unum est optandum, aut facere iniuriam nec accipere,
aut et facere et accipere aut neutrum . . .
(‘For we must choose one of three things—to do injustice and not to suffer it, or both
to do it and to suffer it, or else neither . . . . .’ Cic. Rep. 3.23)
The pair neque . . . neque is regularly used in corresponding negative phrases (see
Chapter 19). Occasionally a preceding phrase has to be understood as negative on the
basis of the following negatively coordinated phrase. The first instance is (j) followed
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
There are several forms of adversative local coordination of a negative member. In the
first place the adversative coordinators may be combined with non (haud is not
attested), as in (l) and (m). Secondly, the members may be coordinated asyndetically
(indicated by Ø), as in (m)–(o). The contrast may be made explicit by the use of
quidem, as in (p).
(l) Quis est qui isti aetati atque isti dignitati non possit quam velit petulanter,
etiam si sine ulla suspicione, at non sine argumento male dicere?
(‘Who is there who cannot make some scandalous attack as impudently as he pleases
against that age and that handsomeness you referred to, even if with no ground for
suspicion, yet not without some basis of accusation?’ Cic. Cael. 8)22
(m) Ut petulantia, ut lubido magis est adulescentium quam senum nec tamen
omnium adulescentium, sed non proborum, sic ista senilis stultitia quae
deliratio appellari solet senum levium est, Ø non omnium.
(‘Just as waywardness and lust are more often found in the young man than in the
old, yet not in all who are young, but only in those naturally base; so that senile
debility, usually called “dotage”, is a characteristic, not of all old men, but only of
those who are weak in mind and will.’ Cic. Sen. 36)
(n) . . . supplicio dignus esset, Ø non, quia fecerit, praemio.
(‘ . . . he would have deserved punishment . . . rather than reward for its accomplish-
ment.’ Cic. Inv. 2.113)
(o) . . . hic vos / dormitis interea domi atque erus in hara, Ø haud aedibus,
habitat.
(‘ . . . here you’ve been drowsing all the time at home, and your master living in a pig-
pen, not a house.’ Pl. As. 429–30)
21
See Sz.: 517, with references. Ex. (j) I owe to N. Holmes (p.c.).
22
A much disputed text (on account of the first isti). See Austin and Dyck ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In Lodge s.v. c.40% of instances of haud are categorized as ‘cum verbis’. These cannot
always be distinguished from clause negations (and this is also the case for some of
the combinations with adjectives). The same holds for McGlynn’s treatment of haud
in Terence. Of the less than 100 instances in Cicero, not more than a handful are
clause negations and fifty-five are part of the idiom haud scio/sciam an (which
means, roughly, ‘perhaps’—nescio an occurs eighteen times). There are ten instances
of haud sane ‘not very’. Caesar has only one example (haud scio). Historians from
Sallust onwards like it. Livy is responsible for one third of all instances of haud on the
BTL disk (more than 800 instances) with a marked preference for combinations such
as with words indicating doubt (dubi*) (e.g. haud dubie 88, non dubie 0 instances)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
and with procul ‘far’ (haud 110, non 0) and sane (haud 20, non 0). In poetry it occurs
occasionally. In ecclesiastical authors it is rare (429 instances on the CLCT-4 disk),
with Jerome using no less than 109 instances of haud with a word expressing doubt
(dubi*).23
It is not difficult to find instances of non in combinations that resemble haud
combinations (e.g. non procul: eighteen instances in Pliny the Elder against haud
procul four), but in view of the overall proportion of the two adverbs, the figures for
non are small. Of the first 135 instances of non on the BTL disk for Cicero, twenty-six
are of the non . . . sed type.
Supplement:
A few more examples of clausal haud: Pol hau rogem te, si sciam. (Pl. Men. 640); His
enim rebus inbutae mentes haud sane abhorrebunt ab utili aut a vera sententia. (Cic.
Leg. 2.16); Haud equidem adsentior, tu tamen ad reliqua pergas velim. (Cic. Leg.
3.26); Haud amo vel hos designatos, qui etiam declamare me coegerunt, ut ne apud
aquas quidem acquiescere liceret. (Cic. Att. 14.12.2); Haud equidem credo. (Ov. Met.
15.359);
Appendix: Instead of nōn (<*ne-oinom (= unum)),24 a few older forms are found in
Early Latin, Varro, and Lucretius (not Ennius ap. Cic Sen. 10).25 An example is (a).
(a) Noenum mecastor quid ego ero dicam meo / malae rei evenisse quamve
insaniam / queo comminisci.
(‘I simply cannot imagine what misfortune or what madness I should say
has come over my master.’ Pl. Aul. 67–9)
23
For historical information on the use of haud, see also Adams (1994b: 14–15).
24 25
For the etymology, see de Vaan (2008: 403). See Powell ad loc.
26
See Pascucci (1968: 22–9). For the uses of nec, see Orlandini and Poccetti (2008).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) . . . pedibus me ascendere necesse erat, quia prorsus nec in sella ascendi
poterat . . .
(‘I had to ascend on foot, because it was quite impossible to ascend in a chair.’ Pereg. 3.2)27
Supplement:
Tu dis nec recte dicis. Non aequom facis. (Pl. Bac. 119); Si minus in omnis litabit, sic
verba concipito: ‘Mars pater, si quid tibi in illisce suovitaurilibus lactentibus neque
satisfactum est, te hisce suovitaurilibus piaculo’. (Cato Agr. 141.4); Nunc victi tristes,
quoniam Fors omnia versat, / hos illi (quod nec vertat bene) mittimus haedos. (Verg.
Ecl. 9.5–6); Et (del. Bentley) nec destiterit per continuos dies quicquam cuiquam nisi
sub exceptione polliceri. (Suet. Jul. 78.2); Furtorum autem genera Ser. Sulpicius et
Masurius Sabinus IIII esse dixerunt, manifestum et nec manifestum, conceptum et
oblatum. (Gaius Inst. 3.183); Include eum loco tenebroso, ubi nec lumen viderit
(Mulom. Chir. 380—Vegetius 2.113.1 has non);28
Cf.: Nolo ego cum improbis te viris, gnate mi, / neque in via neque in foro / necullum
sermonem exsequi. (Pl. Trin. 281-2a); . . . et neque utrum superavit eorum nec
superatum’st. (Lucr. 4.1217);
Nec without a coordinating meaning is also found in the compound adjective
necopinatus ‘unexpected’ (= later inopinatus). Necdum ‘not yet’ is also used as a
temporal adverb without a coordinating meaning (= nondum) from the Elder Seneca
onwards. (For necne ‘or not’ in direct and indirect questions, see § 6.20.)
27
For nec in the Peregrinatio, see Väänänen (1987: 96).
28
See Löfstedt (1911: 88–9).
29
For more examples, see OLD s.v. neque § 2. Also Orlandini and Poccetti (2008: 7–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(h) . . . non modo Sunium superare sed ne<c> extra fretum Euripi committere
aperto mari se audebant.
(‘(sc. the ships of the pirates from Chalcis) . . . not only would not venture past
Sunium, but did not even dare to enter the open sea beyond the strait of Euripus.’
Liv. 31.22.7)
As to the distribution of neque and nec, some authors avoid nec before vowels. This
is a strong tendency already in Plautus,30 valid in all thirty-six instances of Caesar
(who strongly prefers neque), but Cicero ignores this distribution. After the
Classical period, poets preferred nec; neque was preferred by some authors (in
some works) as the more elevated form (so in Tacitus Ann. I–VI; Apul. Apol.). Nec
is the form that survived in the Romance languages. Nec and neque are used in
different contexts in Gaius and other jurists. Neque . . . neque in corresponding
members is more persistent, notably in the jurists.31 Also in certain combinations
it remained more popular, e.g. neque enim. In Early Latin the situation is quite
diverse, as (i)–(k) show (local negation in italics). More precise information is
obscured by the fact that the reports of what the manuscripts have are not very
reliable.32
(i) Nam nec se Septentriones quoquam in caelo commovent, / nec se luna
quoquam mutat atque uti exorta est semel, / nec Iugulae neque Vesperugo
neque Vergiliae occidunt.
(‘The Great Bear isn’t moving anywhere in the sky, the Moon isn’t going to
any place different from where it was when it first rose, and Orion, the
Evening Star, and the Pleiades aren’t setting either.’ Pl. Am. 273–5)
(j) Caput dolet neque audio neque oculis prospicio satis.
(‘My head’s in pain, I can’t hear, I can’t see well with my eyes.’ Pl. Am. 1059)
(k) Nec me miserior femina est neque ulla videatur magis.
(‘And there isn’t a more wretched woman than me, nor could there seem to
be one.’ Pl. Am. 1060)
The negation ne- inherent in neque (with a short ‘e’) is also present in words
like nescio, neuter, nullus, non (see § 8.11, Appendix), nemo, nihil, and
quin.33
8.13 Ne
Nē is part of the focus subjunct ne . . . quidem (see § 8.14). It also is a prefix in the
clausal negative compound adverbs nequaquam ‘not at all’ and nequiquam ‘vainly’.
30
For Plautus, see Lodge: s.v. neque/nec § I; for Silver Latin, see Lease (1902); for the history of the two
words in general, see Löfstedt (1942: I2.331–8; 1933: II.286–8); also Sz.: 451–2. For nec in poetry, see
Axelson (1945: 115–18); for Tacitus, see Eriksson (1934: 78–80).
31 32
See Nelson (1981: 408–9). N. Holmes (p.c.).
33
For the etymology and its relation to nē, see de Vaan (2008: 403–4). Also for the relationship
between nec and neque (is nec an apocopated form of neque or not?).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
8.14 Ne . . . quidem
Etymologically, nē . . . quidem is the combination of the negation adverb nē and the
focusing particle quidem. However, unlike ne, the combination ne . . . quidem is not
limited to imperative sentences and certain subordinate clauses. It actually functions
as a focus subjunct (see Chapter 22). Although it is found from Early Latin onwards,
it only becomes common in prose afterwards, especially in Cicero, and then also in
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
Livy, Seneca, Pliny, and Tacitus.34 If later authors use it, it is a literary or learned
phenomenon. Gaius and others use ne quidem next to each other, as in (b) below.35
There are instances of ne alone, functioning as ne . . . quidem, from Petronius
onwards, as in (c). Instances in authors of the Classical period are usually emended.36
(a) . . . ne epistula quidem ulla sit in aedibus / nec cerata adeo tabula.
(‘ . . . there shall not even be any letter in the house, nor as much as a wax tablet.’ Pl.
As. 762–3)
(b) . . . quia ne quidem civis Romanus est . . .
(‘ . . . because he is not even a Roman citizen . . . ’ Gaius Inst. 1.67)
(c) Hoc solum vetare ne Iovis potest.
(‘This one thing even Jupiter himself cannot forbid.’ Petr. 47.4 (Trimalchio
speaking))
Supplement:
Ne istuc quidem edepol postulo. (Pl. Mos. 1008); Omnino irasci amicis non temere
soleo, ne si merentur quidem. (Cic. Phil. 8.16); Ne tam diu quidem dominus erit in
suos dum ex eis de patris morte quaeratur? (Cic. S. Rosc. 78); . . . neminem eloquentia
non modo sine dicendi doctrina, sed ne sine omni quidem sapientia florere umquam
et praestare potuisse. (Cic. de Orat. 2.5); Ingemescere non numquam viro concessum
est, idque raro, eiulatus ne mulieri quidem. (Cic. Tusc. 2.55);
HOMO INPIENTISSI/ME QUI MIHI NE UNAM E/PISTULAM MISISTI (CEL appendix Vindol. Æ
(Vindolanda, AD 97–103)); Nam prae metu latronum nulla <s>es[c]sibula ac ne
sufficientem supellectilem parare nobis licet. (Apul. Met. 1.23.2);
NB: . . . more praepotentium qui possident finis gentium quos ne circumire quoque
valent . . . (Col. 1.3.12); (sc. lupi) . . . nulli contra nos aditum tulerunt ac ne procul
saltem ulli comparuerant. (Apul. Met. 8.16.7);
8.15 Ni
Nī is used in Early Latin in the way nē is used. In (a) it is in an imperative sentence; in
(b) in an argument clause; in (c) in a purpose satellite clause.
(a) Rem divinam ni faciat neve mandet (sc. vilica) . . .
(‘She (the housekeeper) must not engage in religious worship nor get others to
engage in it . . . ’ Cato Agr. 143.1)
(b) Caveat ni labet columella.
(‘He must prevent the pivot from shaking.’ Cato Agr. 20.1—NB: text uncertain)
34
For details about the distribution of ne . . . quidem, see TLL s.v. ne . . . quidem 321.3ff.
35
For Gaius’ use of ne quidem, see Nelson (1981: 409–10). More examples in OLD s.v. ne1 § 6.a fin. Ne
quidem in Cic. Att. 2.16.4 was emended by Corradus.
36
For examples in the Vindolanda tablets and in Petronius, see Adams (1995a: 131–2, 2003b: 11). See
Quint. Inst. 1.5.39 for its ‘vulgar’ character. For the history of ne = ne . . . quidem, see Sz.: 447–50. See also
TLL s.v. ne 313.9ff., for a number of emendations.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Ni is mainly used in conditional clauses (in the sense of nisi)37—see § 16.56. For its
use with quid (quidni) see § 8.4. It is furthermore probably incorporated in nimirum
‘undoubtedly’.38 The forms nei and neive are found in Early Latin inscriptions and in
manuscripts of Plautus.
8.16 Neve/neu
Nēve is the combination of nē and the disjunctive coordinating suffix -ve (for which
see Chapter 19). The use of -ve is remarkable, since neve functions mainly as a pure
conjunctive coordinator. Some legal instances obviously coordinate actions which can
or need not be performed simultaneously, so disjunctive -ve is in place there. As to the
distribution of neve and neu, the former is more common than the latter, in some
prose authors much more common. Both forms are relatively uncommon in poetry,
which has to do with its preference for simple sentences.39
From Early Latin onwards nihil (or nil) ‘nothing’ is used at the clause level as a degree
adjunct and, as such, functions more or less as a stronger negation than non. The
traditional term for this use is accusativus adverbialis. Examples are (a)–(d). Note the
presence of the object constituents.
(a) Nilne adiuvare me audes?
(‘Don’t you want to support me at all?’ Pl. Ps. 78)
(b) . . . apage te, Amor, non places, nil te utor.
(‘ . . . away with you, Love, I don’t like you, I’m not having any dealings with you.’ Pl.
Trin. 258)
(c) . . . Pompeius, cum mihi nihil ostendisset se esse offensum, in Sardiniam et in
Africam profectus est . . .
(‘Pompey (without giving me any indication of his displeasure) left for Sardinia and
Africa . . . ’ Cic. Fam. 1.9.9)
37
For explanations of how ni came to be interpreted as nisi, see Bertocchi and Maraldi (2011b: 121–2).
38
For the various etymologies that have been proposed, see Schrickx (2011: 163–5).
39
For the use of -ve, see Sz.: 535–6; de Vaan (2008) s.v. For the frequency of neu in poetry, see Lease
(1913: 431–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
8.19 Male
40
See TLL s.v. malus 243.18ff.; OLD s.v. male, §§ 5–6.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
8.20 Minus
Minus ‘less’ is interpreted as a mild negation from Early Latin onwards.41 It is found
both at the clause level and with adjectives and adverbs. Examples at the clause
level are (a) and (b); in (c) it is used with an adjective, in (d) with an adverb.
It is part of quominus (see §§ 8.30–8.32), sometimes also separated quo . . . minus, as
in (e). It is also used as a negation in conditional clauses, as in (f), sometimes in
elliptical clauses, as in (g).42
(a) Id ego primo minus animadverti.
(‘At first I paid less attention to it.’ Cato orat. 46)
(b) Suspicor illud tibi minus probari quod a tuis familiaribus . . . non
probabatur . . .
(‘I suspect that another proposal of mine is less to your liking—your friends . . . did
not like it either . . . ’ Cic. ad Brut. 1.15.9)
(c) . . . propter quam in summa infamia sum et meo patri minus [sum] obse-
quens, / quoi<u>s nunc pudet me et miseret.
(‘ . . . for your sake I’m in utter disgrace and have disobeyed my father. I now feel
shame and regret towards him.’ Ter. Hau. 259–60)
(d) . . . nihil est quod ceterorum res minus commode gestas proferas . . .
(‘ . . . it is useless for you to mention the unsuccessful performances of other men . . . ’
Cic. Ver. 5.131)
(e) Qui sibi mandasset delegati ut plauderent / quiv’ quo placeret alter fecisset
minus, / eius ornamenta et corium uti conciderent.
(‘Should anyone have given instructions that claqueurs should applaud him, or
should anyone have caused another to be unsuccessful, they should beat his costume
and his skin to pieces.’ Pl. Am. 83–5)
(f) Si minus credis, respice.
(‘If you distrust me, look around here.’ Pl. Truc. 900)
(g) Educ tecum etiam omnis tuos, si minus, quam plurimos.
(‘Take all your men with you or, if you cannot take them all, take as many as you can.’
Cic. Catil. 1.10)
The degree adverb minime ‘least’ in combination with an ungradable verb comes
close to expressing a high degree of negation. Examples are (a) and (b). Ex. (b) is a
41
See TLL s.v parvus 580.35ff.; OLD s.v. minus, § 4. For later developments, see Sz.: 454–5.
42
For si minus, see Bertocchi and Maraldi (2011b: 122–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
prohibition where minime functions as a strong ne. Other strong negative expres-
sions, which need no further illustration, are nequaquam ‘by no means’, neutiquam
‘in no wise’, nullatenus ‘in no wise’, nullo modo ‘in no way’, in nullo ‘in nothing’.43
(a) Tamen si faciat, minime irasci / decet neque id immerito eveniet.
(‘Still, should he do it, we ought not to be angry with him, and it won’t happen
without good reason.’ Pl. St. 27–8)
(b) Etiamne unguentis unguendam censes? # Minime feceris.
(‘Do you think I should also apply ointments? # Most certainly not.’ Pl. Mos. 272)
Latin has a number of negative subordinators: ne, quin, quominus. The distribution of
these subordinators varies. They can also be used without a negative meaning, if the
governing verb has a negative implication. The use of the negative subordinator can
be regarded as a reflex of that negative implication. This resembles the EXPLETIVE use of
the negator, familiar from French and other languages.44
Examples of the various classes of argument clauses with which ne is used with a
negative meaning are (a)–(j). The directive force of these clauses is obvious when a
person is engaged in some way to get another person to do something. In other cases
this is not so obvious: for example, when the subjects of the main and the argument
clause are identical. It is therefore no surprise that sometimes instead of ne, ut non is
used. (Ut non is required in the case of local negation.)
(a) . . . opsecro / ne indicium ero facias meo.
(‘ . . . I’m imploring you not to betray us to my master.’ Pl. Mos. 743–5)
43
For details about these words, see Hofmann (1951: § 77).
44
For the use of ‘ne explétif ’ in French, see Grevisse (1993: § 983). See also Orlandini (1993, 2005b) on
French and Italian, with references to Latin, for which she rejects expletiveness (see especially 2005b: 164).
For the use of expletive negators in contexts like these in other languages, see Hentschel (1998: 28-33). For
the use of these three subordinators in argument clauses, see Orlandini (1996a, 2003—on ‘explétif ’,
2005b: 485–6).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Nunc ibo intro atque edicam familiaribus, / profecto ne quis quicquam
credat Pseudolo.
(‘Now I’ll go in and tell my household members that absolutely no one should trust
Pseudolus in anything.’ Pl. Ps. 903–4)
(c) Non it, non it, quia tanto opere suades ne ebitat.
(‘He isn’t going, he isn’t going, because you advise him so strongly not to go out.’
Pl. St. 608)
(g) Tacuisti, sedasti etiam litis illorum et sponsio illa ne fieret laborasti.
(‘You held your tongue; you allayed the contention between the two men; you did
your utmost to stay the challenge from going forward.’ Cic. Ver. 3.132)
With many of the governing verbs involved, the argument clause may lack a
subordinator (for examples, see § 15.82). With positive subordinate clauses there is
the choice between Ø and ut, with negative clauses between ne and ut ne. Ut and ne
are the regular expressions in Classical Latin.
Ne is used without a negative meaning in imperative argument clauses with verbs and
expressions of forbidding, hindering, and opposing, just like quominus (see § 8.31),
which occurs more often, and quin (see § 8.27), which is rare and almost only used
when the governing clause contains a negation. Examples are (k)–(m). For further
examples, see also § 15.71.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
45
For prohibeo, see TLL s.v. 1789.71ff.
46
For this argument, see Orlandini (2003: 500–1).
47
See K.-St.: II.252. For a discussion of the various explanations of the use of ut = ne non, see Ripoll
(2014). For the strange use of ut with vereor in the sense of ne in Hor. 1.3.120–1 (often emended), see
Victor (2013).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of purpose satellite clauses with the negative subordinator ne are (a) and
(b). Ne is also used in stipulative clauses. Further examples can be found in § 16.50
and § 16.53, respectively.
(a) Nam novom maritum et novam nuptam volo / rus prosequi . . . ne quis eam
abripiat.
(‘I want to accompany the new groom and the new bride to our estate . . . so that no
one abducts her.’ Pl. Cas. 782–4)
(b) Ne cadam, amabo, tene me.
(‘Please hold me so that I don’t fall.’ Pl. Cas. 634)
48
For the contexts in which quin is used, see Moussy (1987, 1998, 2012b), Fleck (2008: 203–413).
49
See Sz.: 679.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
Various classes of verbs can be distinguished with which quin is used as a subordi-
nator without a negative meaning.
(i) Quin is used in imperative argument clauses with main clauses containing a
verb or expression of forbidding, hindering, and opposing. In this use it does
not have a negative meaning, just like ne (more frequent—see § 8.24) and
quominus (the most frequent—see § 8.31) in such constructions. This use is
infrequent with non-negated main clauses—examples are (a) and (b)—but
quite common with negated ones (although not as frequent as quominus), as
in (c) and (d). Further examples can be found in § 15.71.
(a) . . . tantus incessit timor ut . . . alii morari Caesarem dicerent quin naves
conscendere iuberet.
(‘ . . . so great was the panic that took hold upon them that . . . some asserted
that Caesar was being slow in giving orders to embark.’ B. Alex. 7.1—NB:
variously emended)
(b) . . . Veranius . . . quin ultra bellum proferret morte prohibitus est . . .
(‘Veranius . . . was prevented by death from carrying his arms further . . . ’ Tac.
Ann. 14.29.1)
(c) . . . nec tuis depellar dictis quin rumori serviam.
(‘ . . . and I won’t be deterred by your words from having regard for public
opinion.’ Pl. Trin. 640)
(d) . . . non cunctandum existimavit quin pugna decertaret.
(‘ . . . he therefore considered that he must not delay to fight a decisive battle.’
Caes. Gal. 3.23.7)
(iii) Quin is used without a negative meaning and is the regular subordinator in
declarative argument clauses governed by a negative main clause expressing
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(iv) Quin is also used without a negative meaning in imperative argument clauses
with these verbs and expressions in their meaning ‘to hesitate’, as in (k), but
the prolative infinitive is much more common (see § 15.122).
(k) Quare nolite dubitare quin huic uni credatis omnia . . .
(‘Then hesitate no longer to entrust supreme command to this one man . . . ’
Cic. Man. 68)
(v) Quin is used without a negative meaning in declarative argument clauses with
negated main clauses containing a verb or expression of cognition or commu-
nication, with which an accusative and infinitive clause is normal (see § 15.97).
Two examples are (l) and (m). Further examples can be found in § 15.38.
(l) Nec pol me multum fallit quin quod suspicor sit quod velit.
(‘And, unless I’m much mistaken, I’ve a good idea of what he’s after.’ Ter. Hec. 728)
(m) Atqui alterum dici non potest quin i qui nihil metuant . . . beati sint . . .
(‘Well at any rate there can be no question of the one point—that those who
have no fear . . . are happy . . . ’ Cic. Tusc. 5.17)
Quin is used with a negative meaning in declarative argument clauses with negated
main clauses containing a verb or expression that in combination with the negator
indicates necessity or unavoidability. Examples are (a)–(d). Further examples can be
found in § 15.38. Instead of quin, ut non is used as well, as a stronger form of
negation.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(a) . . . nullo modo / potest fieri prorsus quin dos detur virgini.
(‘ . . . in absolutely no way is it possible for it to happen that a dowry not be given to
the girl.’ Pl. Trin. 729–30)
(b) Nam ill’ non potuit quin sermone suo aliquem familiarium / participaverit
de amica eri . . .
(‘For he couldn’t have helped talking and letting someone of the household share his
news about master’s mistress . . . ’ Pl. Mil. 262–3)
(c) Eheu, nequeo quin fleam . . .
(‘Dear me! I can’t refrain from crying . . . ’ Pl. Mil. 1342)
(d) Fieri nullo modo poterat quin Cleomeni parceretur.
(‘It was absolutely impossible that Cleomenes should not be spared.’ Cic. Ver. 5.104)
Quin is used with a negative meaning in result satellite clauses in combination with
negated main clauses (see § 16.54), both those containing an expression of the
‘expansion’ type, as with tam in (a), and those containing a zero quantifier, like the
combination nec . . . quicquam in (b). An example of a rhetorical question is (c).
Instead of quin, ut non is used occasionally.
(a) Qui homo culpam ammisit in se, nullu’st tam parvi preti / quin pudeat, quin
purget sese.
(‘No man who’s stained himself with guilt is so worthless that he wouldn’t be
ashamed and wouldn’t apologize.’ Pl. Aul. 790–1)
(b) . . . nec nisi disciplina apud te fuit quicquam ibi quin mihi placeret.
(‘ . . . and there wasn’t anything I didn’t like, apart from the way your servant is
trained.’ Pl. Cist. 17)
(c) Quis umquam templum illud adspexit quin avaritiae tuae . . . testis esset?
(‘Has anyone looked at that temple without becoming a witness to your rapacity . . . ?’
Cic. Ver. 1.154)
Quominus (= quo + minus (see § 8.20), sometimes written and printed as two words)50
and—rarely—quo setius are used in imperative argument clauses with verbs and
expressions of forbidding, hindering, and opposing, without a negative meaning, just
50
For the development of quominus, see Fruyt (2012).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
like ne (see § 8.24), which is less frequent, and quin (see § 8.27). Examples are (a)–(c).
Sometimes it is a negative element in the context that licenses quominus, as in (d).51
(a) IN·EO·AGRO QUO·MINUS·PECUS <P>ASCERE·GENUATES . . . LICEAT . . . NI QUIS·PROHIBETO·,
NIVE·QUIS·VIM·FACITO·NEIVE·PROHIBETO·QUO·MINUS·EX·EO·AGRO·LIGNA . . . SUMANT.
(‘Let no one prevent that the Genuates be allowed . . . to pasture the flocks in that
land . . . no one may hinder them or oppose it by force or hinder them to take . . .
wood from that land.’ CIL I2.584.33–5 (Sent. Minuc., Genoa, 117 BC))
(b) Si tabulas C. Fannius accepti et expensi profert . . . , quo minus secundum
illum iudicetis non recuso.
(‘If Gaius Fannius produces his accounts of money received and expended . . . , I have
no objection to your giving judgement in his favour.’ Cic. Q. Rosc. 2)
(c) Et vendidisset, si tantulum morae fuisset quominus ei pecunia illa numeraretur.
(‘And he would have sold them, had there been the smallest delay in paying him that
money.’ Cic. Ver. 2.93)
(d) Sed tamen, si nihil de me tulisti quo minus essem non modo in civium numero,
sed etiam in eo loco in quo me honores populi Romani conlocarunt . . .
(‘Supposing that, in spite of your claim, you carried no motion to oust me, not merely
from my place as a citizen, but from the position in which I have been placed by the
distinctions which the Roman people have conferred upon me . . . ’ Cic. Dom. 82)
In Tacitus and other Silver and Late Latin authors, quominus is also used without a
negative meaning in satellite clauses, usually with main clauses containing a negation,
in which quin is the normal expression, as in (e). Quominus is also used in imperative
argument clauses with expressions of hesitating, where quin is the normal expression
(see § 8.27). An example is (f).
(e) . . . nihil reliqui faciunt quo minus invidiam, misericordiam, metum et iras
permoverent.
(‘ . . . they left out nothing which would arouse resentment, pity, dread, and anger.’
Tac. Ann. 1.21.2)
(f) Nec apud duces Vitellianos dubitatum quo minus pacem concederent.
(‘And the generals of Vitellius did not long hesitate to grant terms.’ Tac. Hist. 2.45.2)
However, quominus is used with its negative meaning ‘that not’ in declarative
argument clauses with a number of main clauses that do not contain a verb with a
51
For the contexts that favour the use of quominus, see Moussy (1987, 1998, 2012b). For its frequency
in various authors, see Sz.: 680–1. For more examples from Tacitus, see OLD s.v. quominus, § 2.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
52
For more instances, see K.-St.: II.260.
53
For ‘adjektivische und substantivische Negation’, see Sz.: 204–5.
54
For a discussion of ‘two strategies to talk about nothing’, see Dik (1997: II.183–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Mihi quidem edepol insignite facta est magna iniuria. / Duobus nupsi, neuter
fecit quod novae nuptae solet.
(‘But by Pollux, clearly a great injustice has been done to me: I married two men, but
neither did what one normally does to a newly-wed woman.’ Pl. Cas. 1010–1)
(e) Numquam ego illi possum gratiam referre, ut meritu’st de me.
(‘I can never repay him the way he’s deserved of me.’ Pl. Mos. 214)
(f) . . . nusquam invenio Naucratem.
(‘ . . . I can’t find Naucrates anywhere.’ Pl. Am. 1014)
(g) Nam quod fortasse non nemo vestrum audierit, istum a Syracusanis publice
laudari . . .
(‘As to the statement, which perhaps some of you heard, that the Syracusans are
giving him an official eulogy . . . ’ Cic. Ver. 2.15)
(h) Aperte enim adulantem nemo non videt, nisi qui admodum est excors.
(‘No one, to be sure, unless he is an utter fool, fails to detect the open flatterer.’ Cic.
Amic. 99)
Of these quantifiers, nemo and nihil (also nil) function mainly the way nouns and
independent pronouns do, but there are a few further uses that are noteworthy,
as illustrated below for nemo, in particular. In (i), neminem is the determiner of
hominem iracundiorem, a phrase which is found, especially in this order and with
discontinuity, also in Cicero, infrequently also later. In (j), neminem is the
determiner of substantival alium (also found in later authors). In (k), nemo is
combined with the indefinite pronoun quisquam (the precise syntactic relation is
difficult to define). This use is limited to Early Latin and archaizing later authors.
In the case of nemo and nihil the set to which they refer may be indicated (the
so-called genetivus partitivus), as in (l). Parallel expressions for these examples
with nihil are easy to find.55 Nihil also functions as an adjunct of degree and is for
that reason considered a stronger alternative to non, as in (m) (further instances
in § 8.18). Neuter may also be combined with an expression that indicates the
domain, as in (n).
55
For more examples, see OLD s.v. nemo § 1.c,d; 5; nihil § 1.d; 2; 4.a.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(l) Ea signa nemo homo horum / videre poterit, verum vos videbitis.
(‘No one of the household here will be able to see these marks, yet you will see them.’
Pl. Am. 146–7)
(m) Nilne adiuvare me audes? # Quid faciam tibi?
(‘Don’t you want to support me at all? # What should I do for you?’ Pl. Ps. 78)
(n) Certumne est neutram vostrarum persequi imperium patris?
(‘Is it your decision that neither of you will follow her father’s command?’ Pl. St. 141)
The main function of nullus is as a determiner, as in (o), but especially in the plural it
is often used substantively, as in (p). Its genitive and ablative singular forms also serve
as a substitute for lacking (or very rare) neminis and nemine. The dative singular form
is also used as an alternative for nemini. In Late Latin other case forms also replace
nemo, as in (q). For missing forms of nihil, those of nulla res ‘no thing’ can be used as
an alternative. Remarkable is the use of nullus as a so-called FLOATING QUANTIFIER (see
Chapter 21) to make a sentence or clause negative, as in (r)–(t).56
(o) Homo nullu’st te scelestior.
(‘There’s no greater criminal than you . . . ’ Pl. Aul. 419)
(p) Qua in oratione nonnulli aliquando digni maiorum loco reperiuntur,
plerique autem hoc perficiunt ut . . .
(‘And in such speeches some men are sometimes found to be worthy of the rank
which their ancestors attained, but the majority only make it seem . . . ’ Cic. Agr. 2.1)
(q) . . . specialiter illa die nullus recedit a vigiliis usque in mane.
(‘ . . . on that special day not a soul withdraws from the vigils until morning.’ Pereg.
36.3)
(r) Libanum in tonstrinam ut iusseram venire, is nullus venit.
(‘When I had ordered Libanus to come to the barber’s, he didn’t come at all.’ Pl. As. 408)
(s) Ill’ quidem hanc abducet. Tu nullus affueris, si non lubet.
(‘He will take her away. Don’t help me if you don’t want to.’ Pl. Bac. 90)
(t) Quod qui dubitet haud sane intellego cur non idem sol sit an nullus sit
dubitare possit.
(‘If a man doubts this, I really cannot see why he should not also be capable of
doubting whether the sun exists or not.’ Cic. N.D. 2.4)
This use of nullus as a floating quantifier is common in Early Latin comedy (and so
also in Apuleius), but also not infrequent in Cicero. Livy has quite a few instances as
well. It is rarely used in Late Latin. It has to be distinguished from the use of nullus as
a subject complement, as in (u) (see § 9.24).57
56
For more examples, see OLD s.v. nullus § 6; Hofmann (1951: 80).
57
For a discussion of the distinction between the expressions, see Bagordo (2001: 87–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(u) Surge, amator, I domum. / # Nullus sum. # Immo es, ne nega, omnium
<hominum> pol nequissumus.
(‘Get up, lover, go home. # I am no more. # O yes, by Pollux, you are, you are
the most wicked of men.’ Pl. As. 921–2)
Zero quantifiers can be used in all sentence types. See, for instance, their use in the
imperative sentences (v)–(y).
(v) Alienum hominem intro mittat neminem.
(‘She shall not let any male outsider in.’ Pl. As. 756)
(w) Numquam istuc dixis neque animum induxis tuom.
(‘Never say that, and never believe that.’ Pl. Capt. 148)
(x) ‘Nihil ignoveris’. ‘Nihil gratiae causa feceris’. ‘Misericordia commotus ne sis’.
(‘ “Never forgive!” “Never show favour!” “Do not be moved by pity!” ’ Cic. Mur. 65)
(y) Nullam aciem, nullum proelium timueris.
(‘Fear no serried ranks, no battle.’ Liv. 2.12.11)
These examples can be paralleled with examples of haud without too much difficulty. For
ullus and umquam, haud is first encountered in Lucretius. Examples are (ae)–(ai).
(ae) Nunc quom esse credent servom et conservom suom, / hau quisquam
quaeret qui siem aut quid venerim.
(‘Now when they believe that I’m a slave, their fellow slave in fact, no one will ask me
who I am or what I’ve come for.’ Pl. Am. 129–30)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
58
Etymologically this is not the case for nequeo, but the authors of our texts do not seem to be aware of
this. See de Vaan (2008: 506).
59
For the difference between non scio and nescio, see: Non scire quidem barbarum iam videtur, nescire
dulcius. (‘Non scire seems barbarous nowadays: nescire is pleasanter’, Cic. Orat. 157—see also Orat. 154
on nequeo and nolo.)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
60
For the relationship between negation and coordination, see Torrego (2009: 477–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(d) Tum ille clare omnibus audientibus sese id non esse facturum neque se usque
eo Sthenio esse inimicum ut eum rei capitalis affinem esse diceret.
(‘Agathinus replied, in a loud voice which all could hear, that he would not do so, that
he was not so bitter an enemy to Sthenius as to allege him to be connected with a
capital offence.’ Cic. Ver. 2.94)
(e) . . . hau materiae reparcunt, / nec sumptus ibi sumptui esse ducunt.
(‘ . . . they do not spare their building materials, and they don’t consider an expense to
be an expense in that case.’ Pl. Mos. 124–5)
(f) Nemo etiam me accusavit / merito meo nec me Athenis alter est hodie
quisquam quoi credi recte aeque putent.
(‘No one’s accused me deservedly yet and there isn’t any other man in Athens these
days who people believe can be trusted equally well.’ Pl. As. 491–3)
(g) Aut quod illa dicat peregre allatam epistulam, / ne epistula quidem ulla sit in
aedibus / nec cerata adeo tabula.
(‘Or in case she say a letter has been delivered from abroad, there shall not even be
any letter in the house, nor as much as a wax tablet.’ Pl. As. 761–3)
(h) Is scit adulescens quae sit quam compresserit, / illa illum nescit neque
compressam autem pater.
(‘This young man knows who the girl he raped is, but she doesn’t know him, and her
father doesn’t even know that she’s been raped.’ Pl. Aul. 29–30)
Supplement:
Non ego te hic lubens relinquo neque abeo aps te. (Pl. Am. 531); Quamquam non
multum fuit molesta servitus / nec mihi secus erat quam si essem familiaris filius. (Pl.
Capt. 272–3); Ego te non novi neque novisse adeo volo. (Pl. Men. 296); Ne tu opinere,
hau quisquam hodie nostrum curret per vias / neque nos populus pro cerritis
insectabit lapidibus. (Pl. Poen. 527–8); Hau placet neque id viri officium arbitror.
(Pl. St. 297);
Nil moror neque scire volo. (Pl. Bac. 989a); Amicior mi nullus vivit atque is est /
qui illam habet neque est quoi magis me melius velle aequom siet. (Pl. Mer. 897–8);
Nam equidem Athenas antidhac / numquam adveni nec te vidi ante hunc diem
umquam oculis meis. (Pl. Ps. 620–1);
Quippe illi iniqui ius ignorant neque tenent. (Pl. Am. 37); Nolo ego, soror, me credi
esse immemorem viri / neque ille eos honores mihi quos habuit perdidit. (Pl. St. 48–9);
A preceding positive clause may be explicitly marked as positive by et followed by the
neque clause, as in (i). The same is possible with local negation, as in (j). For the
various combinations of neque, see Chapter 19.
(i) Nam et in recentibus pomariis . . . alii conserunt hortos, alii quid aliud,
neque cum convaluerunt arbores idem faciunt . . .
(‘Thus, on the one hand, in young orchards . . . some plant garden crops, and
others plant other crops; but they do not do this after the trees have gained
strength . . . ’ Var. R. 1.23.6)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Neque is also used when the scope of the negation is on a particular constituent of the
following clause and in cases of local negation. Examples are (k)–(m).61 In (k), the
negation element modifies immerito, an instance of litotes (see § 8.8). In (l) the scope
of the negation is aequom. In (m) the negation belongs to the ablative absolute clause.
(k) Tamen si faciat, minime irasci / decet neque id immerito eveniet.
(‘Still, should he do it, we ought not to be angry with him, and it won’t happen
without good reason.’ Pl. St. 27–8)
(l) Illa enim se negat (sc. copiam facturam esse eius) / neque eum aequom facere ait.
(‘She said no, and told him he was behaving improperly.’ Ter. Ph. 113–14)
(m) . . . cum C. Sulpicius et C. Licinius Calvus consules in Hernicos exercitum
duxissent neque inventis in agro hostibus Ferentinum urbem eorum vi
cepissent, revertentibus inde eis Tiburtes portas clausere.
(‘ . . . the consuls Gaius Sulpicius and Gaius Licinius Calvus led an army against the
Hernici, and not finding the enemy abroad, captured their city of Ferentinum by
assault. As they were returning thence, the men of Tibur closed their gates against
them.’ Liv. 7.9.1)
Rather than by means of neque, a negative clause may also be coordinated by one of
the conjunctive coordinators ac, atque, and et in combination with non (not haud—
only used locally), as in (n)–(p). This is relatively rare and usually contrastive, as in
(o), or correcting, as in (p). In the case of corresponding coordination it is more
common, as in (q) and in (r), an instance of contrastive negation. (For instances of
local negation by means of ac non and et non, also atque non, see § 8.9.) The
combination -que non is also used, as in (s).
(n) Patior iudices et non moleste fero me laboris mei, vos virtutis vestrae fructum
esse laturos.
(‘Very well, gentlemen. I am not sorry that I am to reap the reward of my hard work,
and you the reward of your courage.’ Cic. Ver. 1.2)
(o) Id si relinquo ac non peto, omnes ilico / me suspicentur, credo, habere
aurum domi.
(‘If I let this pass and don’t demand my share, I think everybody would immediately
suspect that I have gold at home.’ Pl. Aul. 109–10)
(p) Ita iocaris, quasi ego dicam eos miseros qui nati non sint et non eos miseros
qui mortui sunt.
(‘You are poking fun at me as if my position, instead of being that those who are dead
are wretched, were that those who are unborn are wretched.’ Cic. Tusc. 1.13)
61
For more examples, see OLD s.v. neque, § 4.a; K.-St.: II.39–40.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
Supplement:
. . . decipiam ac non veniam, Syrus mihi tergo poenas pendet. (Ter. Hau. 728); Quasi
quicquam redivivi ex opere illo tolleretur ac non totum opus ex redivivis constitu-
eretur. (Cic. Ver. 1.148); Amo te et non neglexisse habeo gratiam. (Ter. Ph. 54); Cur
igitur decidit et non restipulatur neminem amplius petiturum? (Cic. Q. Rosc.
36); . . . ipse solus reum absolutum, a quo pecuniam acceperit, condemnet pecuniam-
que non reddat? (Cic. Ver. 2.79);
Exx. (t) and (u) illustrate the adversative coordination of a negative clause by means
of sed and at. Exx. (v) and (w) illustrate disjunctive coordination by means of aut and
vel. Note that haud is not used with either adversative or disjunctive coordination; it
is, however, used in local negation.
(t) Audio, sed non abire possum ab his regionibus.
(‘I can hear you, but I can’t leave these regions.’ Pl. Men. 837)
(u) Fuisset tum illos mi aegre aliquot dies, / at non cotidiana cura haec angeret
animum.
(‘I’d have been upset for those few days; but I wouldn’t have this anxiety tormenting
my mind day after day.’ Ter. Ph. 159–60)
(v) Sin autem temere aliquid alicuius praeterisse aut non satis eleganter
secuti videbimur, docti ab aliquo facile et libenter sententiam com-
mutabimus.
(‘But if it shall prove that I have been too rash in passing over some point in an
author or have not followed him with sufficient discrimination, I shall, when
someone points out my error, readily and gladly change my opinion.’ Cic. Inv. 2.9)
(w) Vel tibi paruimus vel non occurrimus vel hoc fuit rectius.
(‘Perhaps it was that I followed your advice, perhaps that I did not fall in with
Pompey, perhaps that it was the better thing to do.’ Cic. Att. 8.11.4)
Negation of the second or later member of multiple questions is dealt with in § 6.20.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
62
The most detailed (not very accurate) information (until Apuleius inclusively) can be found in Lease
(1913).
63
In the formulation of Dyck in his commentary ad loc. (2004: 437).
64
For a survey of combinations, also in subordinate clauses, see Lease (1913: 256–60).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(d) Omnium primum divorsae state—em sic, istuc volo. / Ne<ve> inter vos
significetis, ego ero paries.
(‘First of all, stand apart from each other—there, like this, that’s what I want; and so
that you won’t make signs to each other, I’ll be a wall.’ Pl. Truc. 787–8—NB: cj.
Camerarius)
Supplement:
Suspiciones omnis ab se segreget / neque illaec ulli pede pedem homini premat, /
quom surgat. (Pl. As. 774–6); . . . teneamus eum cursum qui semper fuit optimi
cuiusque neque ea signa audiamus quae receptui canunt . . . (Cic. Rep. 1.3); Atque
utinam ante vidisses neque totum animum tuum maerori mecum simul dedisses!
(Cic. Att. 3.8.4); . . . adnitimini mecum et capessite rem publicam, neque quemquam
ex calamitate aliorum aut imperatorum superbia metus ceperit. (Sal. Jug. 85.47);
Concedite atque apscedite omnes, de via decedite, / nec quisquam tam av<i>dax
fuat homo qui obviam opsistat mihi. (Pl. Am. 984–5); Recordare enim quibus
laudationem ex ultimis terris miseris; nec hoc pertimueris, nam a me ipso laudantur
et laudabuntur idem. (Cic. Fam. 1.9.19); . . . cum suis vivat valeatque moechis / . . . /
nec meum respectet ut ante amorem . . . (Catul. 11.17–21); Pone graves curas man-
dataque falle mariti; / nec dubita, cum te partu Lucina levarit, / tollere quidquid erit.
(Ov. Met. 9.697–9); Clausos omnis in curiam accipite, solos, inermis. Nec quicquam
raptim aut forte temere egeritis. (Liv. 23.3.2–3); Ceteri abolete memoriam foedissi-
mae noctis. Nec illas adversus senatum voces ullus usquam exercitus audiat. (Tac.
Hist. 1.84.3);
Vicinas aliasque mulieres quam minimum utatur neve domum neve ad sese
recipiat. (Cato Agr. 143.1—NB: unless quam minimum is taken as a negation
expression); Nunc, o nunc liceat crudelem abrumpere vitam, / . . . / . . . gravior neu
nuntius auris / vulneret. (Verg. A. 8.579–603); Serus in caelum redeas diuque / laetus
intersis populo Quirini, / neve te nostris vitiis iniquum / ocior aura / tollat. (Hor.
Carm. 1.2.45–9); . . . genetrici corpus inemptum / reddite, neve auro redimat ius triste
sepulcri, / sed lacrimis. (Ov. Met. 13.471–3);
Consulite atque fugae medium ne temnite tempus. (V. Fl. 4.154); (after a lacuna)
ne volucri ritu flammarum moenia mundi / diffugiant subito magnum per inane
soluta / et ne cetera consimili ratione sequantur / neve ruant caeli tonitralia templa
superne . . . (Lucr. 1.1102–5); Morere et fratrem ne desere frater. (Verg. A. 10.600);
(g) . . . suadebit tibi ut hinc discedas neque mihi verbum ullum respondeas.
( . . . he will insist on your leaving the court without attempting a single word in reply
to me.’ Cic. Div. Caec. 52)
(h) Adiungam etiam de memoria, ut labore Crassum levem neque ei quicquam
aliud de quo disserat relinquam nisi ea quibus haec exornentur.
(‘I will also add something on the subject of memory, in order to lighten the task of
Crassus and leave him nothing else to discuss except the method of elaborating these
subjects.’ Cic. de Orat. 2.350)
(i) . . . aut, ut vetera exempla, quorum est copia digna huius imperi gloria, relin-
quam, neve eorum aliquem qui vivunt nominem, qualis nuper Q. Catulus
fuit . . .
(‘ . . . or—not to recall so many ancient examples, the number of which is worthy of
the glory of this Empire, and without mentioning by name any of the great men still
alive—such a man as Quintus Catulus lately was . . . ’ Cic. Sest. 101)
(j) . . . legatos ad Caesarem miserunt oratum ne se in hostium numero duceret
neve omnium Germanorum qui essent citra Rhenum unam esse causam
iudicaret.
(‘ . . . he sent envoys to Caesar to beg him not to count them among his enemies, nor
to consider that there was common cause among all the Germans on the Roman side
of the Rhine.’ Caes. Gal. 6.32.1)
(k) Te atque senatum obtestamur, consulatis miseris civibus, legis praesidium
quod iniquitas praetoris eripuit restituatis, neve nobis eam necessitudinem
inponatis . . .
(‘We implore you and the senate to take thought for your unhappy countrymen, to
restore the bulwark of the law, of which the praetor’s injustice has deprived us, and
not to impose upon us the necessity . . . ’ Sal. Cat. 33.5)
Negation
(c) De sepulchris autem nihil est apud Solonem amplius quam ‘ne quis ea deleat
neve alienum inferat’ . . .
(‘But Solon had no more rules about graves except one to the effect that no one is to
destroy them or place the body of a stranger in them . . . ’ Cic. Leg. 2.64)
(d) Ne lacruma, soror, neu tuo id animo / fac quod tibi [tuos] pater facere
minatur.
(‘Stop crying, my sister, and don’t inflict on your heart that distress which our father
is threatening to cause you.’ Pl. St. 20–1)
(e) . . . IS·MAGISTRATUM·INPERIUMVE·NEI·PETITO·NEIVE·GERITO·NEIVE·HABETO . . .
(‘ . . . let him not seek or take up or hold office or authority . . . ’ CIL I2.582.20 (Tab.
Bant., Banzi, end II BC))
(f) Nunc iam illa non volt—tu quoque inpotens <noli> / nec quae fugit sectare
nec miser vive / sed obstinata mente perfer obdura.
(‘Now she desires no more—no more should you desire, poor madman, nor follow
her who flees, nor live in misery, but with resolved mind endure, be firm.’ Catul.
8.9–11)
(g) Sta ilico, noli avorsari nec te occultassis mihi.
(‘Stand where you are, don’t turn away or hide from me.’ Pl. Trin. 627)
Supplement:
Ne duas neu te advexisse dixeris. (Pl. Mer. 401); Ad cenam ne quo eat neve
ambulatrix siet. (Cato Agr. 143.1); Hominem mortuum, inquit lex in XII, in urbe
ne sepelito neve urito. (Cic. Leg. 2.58); At tu, dum corpore non es / passa nefas,
animo ne concipe, neve potentis / concubitu vetito naturae pollue foedus. (Ov. Met.
10.351–3);
‘Ne cape’ de populo quem terra creaverat unus / exclamat ‘nec te civilibus insere
bellis.’ (Ov. Met. 3.116–17);
Appendix: Different is the use of neu in (h), which is typical of Ovid.65 The negation
element ne belongs to the purpose subordinate clause gratia . . . rependatur. The
element -ve coordinates the two clauses containing the imperatives nega and cape.
(h) ‘Quisquis es, hospes,’ ait ‘si forte armenta requiret / haec aliquis, vidisse
nega, neu gratia facto / nulla rependatur, nitidam cape praemia vaccam’, / et
dedit.
(‘ “Whoever you are, my man, if anyone should chance to ask you if you
have seen any cattle going by here, say that you have not; and, that your
kindness may not go unrewarded, you may choose out a sleek heifer for your
pay”; and he gave him the heifer forthwith.’ Ov. Met. 2.692–5)
65
Also rarely in Lucan; see Lease (1913: 434).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
66
For a discussion of the possible combinations, see Orlandini (1993).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
Negation adverbs are also used in combination with negative verbs (for which see
§ 8.34) and related adjectives. For combinations with verbs, there are relatively many
attestations in Cicero. Haud is rare. With adjectives, the full spectrum of negation
adverbs is attested. Examples with negative verbs are (h)–(j), with an adjective (k).
(‘Now you are not unaware, gentlemen, that even brute beasts, when prompted by
hunger, generally return to the place where they have at some time previously found
food.’ Cic. Clu. 67)
(j) Plurima vero mutatione figurarum: scio ‘non ignoro’ et ‘non me fugit’ et ‘non
me praeterit’ et ‘quis nescit?’.
(‘Very often we use a change of figure: for “I know” we say “I am not unaware”,
“It does not escape me”, “It does not pass me by”, “Who does not know?”.’ Quint.
Inst. 10.1.12)
(k) . . . vir bonus et sapiens et legibus parens et civilis offici non ignarus utilitati
omnium plus quam unius alicuius aut suae consulit.
(‘ . . . a good, wise and law-abiding man, conscious of his duty to the state, looks
after the advantage of all more than that of himself or of any single individual.’
Cic. Fin. 3.64)
Supplement (in alphabetical order by verb and adjective):
Quamquam haud ignoro quae bona sint sinistra nos dicere, etiamsi dextra sint. (Cic.
Div. 2.82); Materiam rei non ignoras. (Cic. Q. fr. 2.1.1); Equidem ego non ignoro, si
iam mihi respondere velint, abunde illis facundam et conpositam orationem fore.
(Sal. Jug. 85.26); Nec ignoro quosdam veteres auctores praecepisse ne seminarentur
agri, nisi cum terra pluviis permaduisset. (Col. 2.8.4); Hic quoque non nescit quid sit
amare chorus. (Prop. 2.30.34); Cumque is se non nolle dixisset, locutus esse dicitur
homo copiosus aliquot horas de imperatoris officio et de omni re militari. (Cic. de Orat.
2.75); Proximis enim tuis litteris primum te id non nolle cognovi. (Cic. Att. 13.22.1);
Fuere item ea tempestate qui crederent M. Licinium Crassum non ignarum eius
consili fuisse. (Sal. Cat. 17.7); Ac fuere qui tum Albinum haud ignarum consili regis
existumarent . . . (Sal. Jug. 36.3); Non ignara mali miseris succurrere disco. (Verg. A.
1.630); Neque segnius ad hostes bellum apparatur, qui et parti ante decoris memores
neque ignari auctarum virium hostis suas quoque vires augent. (Liv. 7.7.4); Causam
Sopatri defendebat Q. Minucius eques Romanus in primis splendidus atque honestus
vobisque iudices non ignotus. (Cic. Ver. 2.69); . . . quem quondam dicitur Orphne, /
inter Avernales haud ignotissima nymphas, / ex Acheronte suo silvis peperisse sub
atris. (Ov. Met. 5.539–41); Nec ignotum quid Glyconi, cui Spiridion fuit cognomen,
acciderit. (Quint. Inst. 6.1.41); Carthaginis expugnationem in hunc annum contuli
multis auctoribus, haud nescius quosdam esse qui anno insequenti captam
tradiderint . . . (Liv. 27.7.5); Pleraque eorum quae rettuli quaeque referam parva
forsitan et levia memoratu videri non nescius sum. (Tac. Ann. 4.32.1); Neque nescium
habebat Anteium caritate Agrippinae invisum Neroni . . . (Tac. Ann. 16.14.1);
The interpretation of a negated adjective with a negative prefix is positive. A very
strange combination, usually emended, is haud impigre in (l).67
67
For discussion, see Löfstedt (1933: II.215–16) and Briscoe ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
In (m) and (n), a single clause contains two negation adverbs, one of which negates
the clause as a whole while the other belongs to a prolative infinitive dependent on a
modal verb that is the main verb of the clause. The resulting interpretation is more or
less that of the opposite modal verb: non possum non X = ‘I must X’.
(m) Iuvenem nostrum non possum non amare, sed ab eo nos non amari plane
intellego.
(‘As for our young man, I cannot help feeling affection for him, but I plainly see that
he has none for us.’ Cic. Att. 10.10.6)
(n) Ne licet quidem tibi iam tantis rebus gestis non tui similem esse.
(‘After such exploits you no longer have the right to fall below your own standard.’
Cic. Att. 14.17a.7)
Supplement:
Nam ceteri dolores mitigantur vetustate, hic non potest non et sensu praesentis
miseriae et recordatione praeteritae vitae cottidie augeri. (Cic. Att. 3.15.2); Non enim
quae (sc. structurae) sunt e molli caemento subtili facie venustatis, non eae possunt
esse in vetustate non ruinosae. (Vitr. 2.8.8); Vir bonus non potest non facere quod
facit; non enim erit bonus, nisi fecerit. Ergo nec bonus vir beneficium dat, quia facit
quod debet, non potest autem non facere quod debet. (Sen. Ben. 6.21.2);
Cf. Qui profecto de tanta pecunia si esset data nihil audisse non possent.
(Cic. Flac. 93);
Two types of combinations can be distinguished, one in which the effect of the second
negator is cancelled out and another one in which the clause remains negative.
In the combination of neque or nec with non in one clause the resulting interpretation
is positive. Neque or nec may belong to a pair of coordinators or be a connector. An
example of the former is (d), of the latter (e) and (f). Our grammars observe that
Cicero prefers not to juxtapose non to the connector, as in (e), whereas other authors
do so regularly, as in (f). The combination nec non (less frequently neque) developed
into a strong positive expression (‘[and] also’, ‘[and] furthermore ’) which is some-
times written as one word. It is sometimes further reinforced by a focusing particle
(et, etiam, or quoque). An example is (g). Nec non is also used in the case of local
coordination, as in (h), but not in Cicero.68
(d) At neque non diligunt nos . . . neque ignorant ea quae ab ipsis constituta et
designata sunt. Neque nostra nihil interest scire ea quae eventura sint . . .
neque hoc alienum ducunt maiestate sua . . . neque non possunt futura
praenoscere.
(‘But it is not true that the gods do not love us . . . nor is it true that they do not know
their own decrees and their own plans. Nor is it true that it is of no advantage to us to
know what is going to happen . . . nor is it true that the gods think it inconsistent with
their dignity . . . nor is it true that they have not the power to know the future.’ Cic.
Div. 1.82)
(e) Nec vero illud non eruditorum temporum argumentum est quod . . . fides
praecinunt . . .
(‘And it is by no means a proof of an uninstructed age that . . . stringed instruments
play a prelude . . . ’ Cic. Tusc. 4.4)
(f) Neque non quaedam nationes harum pellibus sunt vestitae . . .
(‘Some nations, too, use their skins for clothing . . . ’ Var. R. 2.11.11)
(g) Nec non etiam precor Lympham ac Bonum Eventum . . .
(‘And I shall not fail to pray also to Lympha and Bonus Eventus . . . ’ Var. R. 1.1.6)
68
For distributional data, see Sz.: 524–5 and 778–9. The fullest collection of instances can be found in
Önnerfors (1956: 99–108). For nec non et, see TLL s.v. et 915.49ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(h) Nam ibi vidi greges magnos . . . gruum, pavonum, nec non glirium, piscium,
aprorum, ceterae venationis.
(‘For I have seen there large flocks of . . . cranes, and peafowl, not to speak of numbers
of dormice, fish, boars, and other game.’ Var. R. 3.2.14)
Supplement:
Coordination of clauses: Neque enim tristius dicere quicquam debeo ea de re quam
tu ipse om<i>nibus optimis prosequeris neque non me tamen mordet aliquid. (Cic.
Fam. 3.12.2); Sunt tibi regna patris Dauni, sunt oppida capta / multa manu, nec non
aurumque animusque Latino’st. (Verg. A. 12.22–3);
Sentence connection: Neque haec tu non intellegis . . . (Cic. S. Rosc. 44); Nec non et
Teucri socia simul urbe fruuntur. (Verg. A. 3.352); Nec non olus quoque silvestre
est . . . (Plin. Nat. 19.144);
Local coordination: Fere morborum causae erunt, quod laborant propter aestus aut
propter frigora nec non etiam propter nimium laborem <a>ut contra[riam] nullam
exercitationem . . . (Var. R. 2.1.22); . . . minor ocius et melius deflorescit, habilis arbori
nec non iugo . . . (Col. 3.2.8); Et ec<c>e superveniunt Hypat[ri]a quos ibi reliqueram
famulos . . . nec non et equum quoque illum meum reducentes . . . (Apul. Met. 11.20.6);
Cf. Nec minus et miseri diversa in parte Latini / innumeras struxere pyras . . . (Verg.
A. 11.203–4); Fiunt etiam purpurei colores infecta creta rubiae radice et hysgino, non
minus et ex floribus alii colores. (Vitr. 7.14.1);
Examples of epexegetic negation (also called ‘resumptive’ negation) are (a)–(d). Exx.
(a)–(c) demonstrate the three types of negative expressions followed by an epexegetic
elaboration. Neque . . . neque is especially common. Ex. (d) shows the rare inverse,
pragmatically marked, order.69
(a) Verum hercle vero nos parasiti planius (sc. invocati sumus), / quos num-
quam quisquam nec vocat nec invocat.
(‘But when it comes to us hangers-on, people shout “out” even more, clearly; no one
ever calls on us or calls out our names.’ Pl. Capt. 75–6)
69
For the history of this type of negation, see Murphy (1954/5, 1956). Ter. An. 563–4 is much
disputed. See Levy (1958) and Murphy (1959).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Unde is? # Egone? Nescio hercle / neque unde eam neque quorsum eam.
(‘Where’ve you been? # Me? I’ve no idea where I’ve been or where I’m going.’ Ter. Eu.
305–6)
(c) Non enim praetereundum est ne id quidem.
(‘And the following occurrence should not be passed unnoticed.’ Cic. Ver. 1.155)
(d) Ne aegri quidem quia non omnes convalescunt idcirco ars nulla medicina est.
(‘Nor do all sick persons get well, but that does not prove that there is no art of
medicine.’ Cic. N.D. 2.12)
Supplement:
At pudicitiae eius numquam nec vim nec vitium attuli. (Pl. Epid. 110); Non
redderes, / nec de illo quicquam neque emeres nec venderes / nec qui deterior
esset faceres copiam. (Pl. Trin. 133–5); Habeo hic neminem / neque amicum neque
cognatum. (Ter. Eu. 147–8); Nulla enim vitae pars neque publicis neque privatis
neque forensibus neque domesticis in rebus . . . vacare officio potest . . . (Cic. Off.
1.4); Nihil nec imperium nec maiestas valebat. (Liv. 4.38.1); . . . Romanae societatis
causam . . . quam nemo neque temptat neque oppugnat . . . (Liv. 41.24.2);
(sc. Caesar) . . . qui negat neque honestius neque tutius mihi quicquam esse quam
ab omni contentione abesse. (Cic. Att. 10.9.1); Nolle inultos hostes, nolle successum
non patribus, non consulibus. (Liv. 2.45.5);
Hoc intellexti[n] an nondum etiam ne hoc quidem? (Ter. An. 201); Non fugio ne hos
quidem mores . . . (Cic. Ver. 3.210); . . . neget se posse ne suspicari quidem. (Cic. Fin. 2.30);
. . . ut ego tua magnufica verba neque istas tuas magnas minas / non pluris facio
quam . . . (Pl. Cur. 579–80); Nec sursum nec deorsum non cresco . . . (Petr. 58.5
(Hermeros speaking));
Appendix: Another form of epexegetic negation is by means of nedum ‘let alone’, as in (e).
(e) Satrapes si siet / amator, numquam sufferre eiius sumptus queat / nedum tu
possis.
(‘If she had a satrap for a lover, he’d never be able to sustain her extrava-
gance; you’d be even less able to.’ Ter. Hau. 452–4)
The term ‘double negation’ (also called ‘pleonastic’ negation) covers diverse phe-
nomena.70 In some of the attested instances, the intended effect of repeating
negators is to create a strong negation. In other instances the second negation
is indeed superfluous. The latter type became so normal in later Latin that the
fourth-century grammarian Diomedes explicitly marks it as wrong71 and indeed, it
70
For double negation in general, see Löfstedt (1933: II.209–14), Molinelli (1989), Orlandini (2001:
117–49), Torrego (2009: 478–9). For double negation in Pompeian inscriptions, see Perl (1978). For
Medieval Latin, see Norberg (1943: 103–16). Bagordo (2001: 73–5) refutes Hofmann’s idea (1951: 97–8)
that this type of (emphatic) double negation is colloquial.
71
I.455.11ff.K. Cf. Sz.: 803.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
(a) . . . neque ego hau committam ut, si quid peccatum siet, / fecisse dicas de
[me] mea sententia.
(‘ . . . and I won’t take the risk that if anything goes wrong you might say you acted
according to my verdict.’ Pl. Bac. 1037–8)
(b) Neque ego homines magis asinos numquam vidi.
(‘I've never seen human beings who were such asses.’ Pl. Ps. 136)
(c) Iura te non nociturum esse homini de hac re nemini . . .
(‘Swear that you will not harm a living soul for all this . . . ’ Pl. Mil. 1411)
(d) Debebat Epicrates nummum nullum nemini.
(‘Epicrates owed no one a penny.’ Cic. Ver. 2.60)
(e) Absenti nemo non nocuisse velit.
(‘Let no one wish to injure me in my absence.’ Prop. 2.19.32)
Supplement:
Nec te aleator nullus est sapientior . . . (Pl. Rud. 359); Lapideo sunt corde multi quos
non miseret neminis (Enn. scen. 139V=140J); Ne temere facias. Neque tu haud dicas
tibi non praedictum: cave. (Ter. An. 205); Proprium vero nil neminem habere. (Lucil.
551); Vidi ipse Formiis universos neque mehercule numquam (v.l. unquam; del.
Shackleton Bailey) homines putavi. (Cic. Att. 9.19.1); Sed vides nullam esse rem
publicam, nullum senatum, nulla iudicia, nullam in nullo (v.l. ullo) nostrum digni-
tatem. (Cic. Q. fr. 3.4.1); Et quem ad modum nemini illorum molestus nulla (ulla cj.
Lambinus) in re umquam fuisti, sic cura . . . (Q. Cic. Pet. 20); Neminem nihil boni
facere oportet. Aeque est enim ac si in puteum conicias. (Petr. 42.7 (Seleucus
speaking)); . . . Q(UAESTORI)·AER(ARII)·QUI NU(M)Q(UAM)·REI P(UBLICAE)·NIL·DEBUIT / BAEBIA
ATERONIA C(ONIUGI)·K(ARISSIMO) / V(IVA)·F(ECIT) (CIL V.6520.3–5 (Novara, early II AD)); . . .
dilacerare . . . eum velle per viscera qui non modo nullum intulerit malum nulli,
sed . . . (Arn. Nat. 1.65.7);
NB: The following text is much discussed. However, it is not an instance of double
negation: non goes with sine. Succurrendum est D. Bruto, omnes undique copiae
72
For the development from Latin to Italian, see Molinelli (1988).
73
For the material in Plautus, see Lodge: s.v. neque 159, § C.
74
For a discussion of various proposals, see Fedeli ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
conligendae. Moram (cj. Schöll, horam mss.) exhibere nullam (ullam cj. Ursinus)
i<n> tali cive liberando sine scelere non possumus. (Cic. Phil. 6.7);75
Truly pleonastic instances of double negation are (f)–(h). Note in (f) the combination
haud immerito. Ex. (g) is usually emended. In (h), the double negation occurs in a
subordinate clause.
(f) Quia pol te unum ex omnibus / Venus me voluit magnuficare neque id haud
inmerito tuo.
(‘Because out of all men Venus wanted me to hold you alone in esteem, and not
undeservedly so.’ Pl. Men. 370–1)
(g) Non sum nescius ab eadem neglegentia qua nihil deos portendere volgo nunc
credant neque nuntiari admodum nulla (ulla cj. Gronovius)76 prodigia in
publicum <ne>que in annales referri.
(‘I am not unaware that, as a result of the same disregard that leads men generally to
suppose nowadays that the gods foretell nothing, no portents at all are reported
officially, or recorded in our histories.’ Liv. 43.13.1)
(h) Non ego signatis quicquam mandare tabellis, / ne legat id nemo quam meus
ante, velim . . .
(‘Never would I choose to entrust my messages to tablets under seal, that none might
read them before my lover . . . ’ [Tib.] 4.7.7–8)
Supplement:
Nec haec non minima causa, quod oliveta dicant alternis annis non ferre fructus aut
non aeque magnos. (Var. R. 1.55.3); Si qua recordanti benefacta priora voluptas / est
homini cum se cogitat esse pium / nec sanctam violasse fidem nec foedere nullo (v.l.
ullo) / divum ad fallendos numine abusum homines . . . (Catul. 76.1–4); Nec tamen
non tantum architecti non possunt in omnibus rebus habere summum effectum, sed
etiam ipsi qui privatim proprietates tenent artium non efficiunt ut habeant omnes
summum laudis principatum. (Vitr. 1.1.14); Nec non nihil esse etiam parvo vehi . . .
(Col. 1.3.4); . . . NEQUE EPISTULAE COMMANDATICIAE NIHIL VALUNT . . . (CEL 142.39–40
(Karanis, AD 100–25)); . . . NEQUE HABERE NULLUM NEGOTIUM CUM·FILIAS . . . (CPL 220.I.26
(Arsinoë, AD 131)); . . . NO<N>·FRUITUS·NE/MINE·FUNERAVIT·NATOS . . . (CIL VIII.5370.3–4
(Guelma)); Nam postea nunquam nec qualiscumque humor ibi apparuit usque
in hodie. (Pereg. 19.12); Nec hi admittendi non erunt ad admissuram. (Mulom.
Chir. 753);
. . . circa nomen Latinum a praetore litterae missae . . . captivi ne minus decem
pondo compedibus vincti in nulla alia quam in carceris publici custodia essent. (Liv.
32.26.18); . . . curare voles ne nunquam (sc. iumentum) suffundet sic curabis.
(Mulom. Chir. 100);
75
For a slightly different explanation, see Manuwald ad loc.
76
For a discussion of this and other doubtful cases of two negations in Livy, see Briscoe ad loc.; also
Oakley ad Liv. 9.34.25.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
Sometimes double negation may be a simple mistake (or ignorance) by the author, as
maybe in (i) and (j) below.
(i) Neque ideo non in quo agro idoneae possunt esse non exercendae . . .
(‘This is not to say that they are not to be worked on land where it is
suitable . . . ’ Var. R. 1.2.23)
(j) . . . haec cura non per longum tempus non prodest.
(‘This treatment is not salutary for very long.’ Mulom. Chir. 104)
77
See Sz.: 805; Löfstedt (1976: 136) for Early Medieval Latin instances in Spanish texts.
78
For more instances and references, see Sz.: 805.
79
He also rejects the instance in Gellius 16.2.10, where he follows Gruppe in deleting non (in the
sequence si non).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
are (a)–(c). (For the position of forms of sum with respect to non and haud, see also
Chapter 23.)
(a) . . . quod aliud medicamentum facere non potest neque purgare.
(‘ . . . a thing which no other medicine can do or cleanse.’ Cato Agr. 157.5)
(b) . . . tamen illo ipso in furore suo non est ausus meam domum . . . possidere.
(‘ . . . in spite of all, in the midst of his reckless career, he had not the courage to enter
upon possession of my house.’ Cic. Dom. 107)
(c) Postea L. Sulla . . . populum Romanum . . . adligare novo quaestionis genere
ausus non est.
(‘Subsequently L. Sulla . . . did not dare to inflict this new kind of court upon the
Roman people.’ Cic. Clu. 151)
In the case of emphatic denial or refusal, non is often placed in initial position (see
also § 8.3), as in (d) and (e).
(d) Non edepol tu illum magis amas quam ego, mea, si per te liceat.
(‘By Pollux, there is no way you love him more than I do, my dear, if you allowed me
to do so.’ Pl. Mil. 1263)
(e) Non, si tibi antea profuit, semper proderit.
(‘Even if it may have worked to your advantage previously, it will not always work to
your advantage.’ Cic. Phil. 8.13)
In addition, non and haud may stand before the word(s) they have as their scope,
sometimes separated by one or two words. The combination may stand in sentence-
initial position when carrying focus. Examples are (f)–(i).
(f) Non usque eo est improbus. Non omnia sunt in uno vitia. Numquam fuit
crudelis.
(‘He was not such a scoundrel as that. No one is bad all through, and Verres was
never cruel.’ Cic. Ver. 4.86)
(g) Non iam tibi sic respondebo ut ceteris . . .
(‘I will not answer to you in the same way I would to the others . . . ’ Cic. Dom. 18)
(h) Sed haec iam non ex te, Quinte, quaero, verum ex ipso poeta . . .
(‘But my next question must be addressed not to you, Quintus, but to the poet
himself.’ Cic. Leg. 1.3)
(i) Quae quidem mihi virtus non interdum (interdum non cj. R. Klotz) minor
videtur quam omnino non irasci.
(‘To do this, I sometimes think, is as great a moral accomplishment as not to be angry
at all.’ Cic. Q. fr. 1.1.38)
When functioning as local negators, non and haud regularly stand immediately before
the constituent they modify. In Plautus non and nec are not found at the end of the
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
80
See Fortson (2008: 132).
81
For more examples, see K.-St.: I.638–9; Orlandini (2001: ch. 1, 2003: 512–13); Bertocchi et al. (2010:
83–95); Orlandini and Poccetti (2012).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
When expressions of doubt are used in rhetorical questions, the subordinator quin is
used in the argument clause, as in the case when the main clause is negated (see
§ 8.27). Examples are (e) and (f ).
(g) Superas facile, ut superior sis mihi quam quisquam qui imperant.
(‘You easily gain the upper hand so that for me you are above any of those who
command me.’ Pl. Men. 192)
(h) Ita fustibus sum mollior magis quam ullus cinaedus.
(‘Thanks to your clubs I’m softer than any catamite.’ Pl. Aul. 422)
(i) Quasi vero . . . tum quisquam fuerit qui dubitaret quin emeretur
Caesenniae . . .
(‘Or as if . . . anyone doubted at the time that it was being purchased for Caesennia . . . ’
Cic. Caec. 16)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Negation
82
For negative prefixation in Latin, see Baldi (1989). For many lexical negative expressions no positive
counterpart exists (and vice versa). See Fruyt (2002b).
83
See Spevak (2005b: 77) on quamvis.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Quia, ruri dum sum ego unos sex dies, / me apsente atque insciente, inconsultu
meo, / aedis venalis hasce inscripsit litteris. (Pl. Trin. 166–8); Feci hodie ut fierent
insperante hoc atque invito Pamphilo. (Ter. An. 603); Sunt ea quidem innume-
rabilia quae a tuis emebantur non insciente te . . . (Cic. Phil. 2.93); Nam et duo boni
consules etiam me indicente omnia e re publica fide vestra faceretis . . . (Liv.
22.39.2);
Lucretius and later poets use in- separated from the participle/adjective by the
coordinator -que, as in (c).
(c) . . . saepe nocentes / praeterit exanimatque indignos inque merentes?
(‘ . . . often passes the guilty by and slays the innocent and undeserving?’
Lucr. 2.1103–4)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
CHAPTER 9
1
For this ‘negative’ aspect of the third person, see Vairel-Carron (1980: 280–3).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In the case of first and second person subjects, the regular lexical category the subjects
belong to are personal pronouns2 (or relative pronouns that refer to them; see exx. (b)
and (d) in § 18.24). Exceptional are instances like (a), in which philosophi functions as
an appositive, for which see § 11.86.
(a) Philosophi sumus exorti . . . qui amori auctoritatem tribueremus.
(‘We philosophers have come forward . . . to attribute authority to love.’ Cic.
Tusc. 4.71)
With second person pronouns in combination with a second person verb form it is
not always clear whether the pronoun is (part of) a form of address (an extraclausal
constituent; see § 12.22 and Chapter 22) or the subject of the verb.3 There are
instances of tu (or vos) in combination with a vocative, where tu (vos) is probably a
form of address; ex. (b) is an illustration of this usage. The status of tu with the
restrictive relative clause in (c) is less clear in view of the object te . . . caudicali in the
following sentence. In contrast to the pronouns in (b) and (c), those in (d)–(e) do not
serve to address the addressees and so may well be regarded as subjects.4 They are
2
For the special position of the personal pronouns ego, tu, nos, vos, see §§ 11.121–11.123.
3
For a discussion of the status of tu (vocative or nominative?) in expressions like tu negas and tu nega,
see Vairel-Carron (1986).
4
For more instances, see Lodge: s.v. tu 793 § e,f,g; 796 § e.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
pragmatically important in their context. In (d), vos and ego are in parallel focus. In
(e), there is a change of topic. In (f ), the focality of tu derives from the situation on
stage, where apparently Sosia enters first. The pragmatic motivation for the use of
explicit subjects in these imperative sentences is the same as in other sentence types
(see below).
(b) Gripe, accede huc, tua res agitur. Tu, puella, istinc procul / dicito quid insit et
qua facie.
(‘Gripus, come here; it’s your case that’s on. You, girl, say from there from a distance
what’s inside and what it looks like.’ Pl. Rud. 1148–9)
(c) Tu qui urnam habes aquam ingere. . . . / Te cum securi caudicali praeficio
provinciae.
(‘You there, you who have a bucket, put water into it. . . . You with the axe I make head
of the woodcutting department.’ Pl. Ps. 157–8)
(d) Ubi estis vos? Ite actutum, Tyndarum huc arcessite. / Vos ite intro. Interibi
ego . . .
(‘Where are you? Go immediately, fetch Tyndarus here. You two go in. In the
meantime, I . . . ’ Pl. Capt. 950–1)
(e) Tu vero ascribe me talem in numerum.
(‘Yes, indeed, add my name to such a roll.’ Cic. Phil. 2.33)
(f ) (IUP) Ibo et Mercurium supsequar. (exit Juppiter) / # (enter Sosia and
Amphitruo) (AMP) Age i tu secundum. # (SOS) Sequor, supsequor te.
(‘I’ll go and follow Mercury. # Come on, walk behind me. # I’m following you, I’m
following you closely.’ Pl. Am. 550–1)
In interactive texts, first and second person pronouns are required with first and
second person verb forms for the purpose of identification or if the subject has the
pragmatic function of focus. Typical examples are (g)–(l). Ex. (g) shows REPLACING
FOCUS, (h) EXPANDING FOCUS, (i) COMPLETIVE FOCUS, (j) PARALLEL FOCUS. In (k), tu serves to
single out a person among a group of people: it has CONTRASTIVE FOCUS.5 In (l), a new
person is introduced. (For these types of focus, see Chapter 22.)
(g) Ego sum, non tu, Sosia.
(‘I am Sosia, not you.’ Pl. Am. 379)
(h) Quid me? Num quid vis? # Vale! / # Et tu, frater.
(‘What about me? Is there anything else I can do for you? # Be well. # You too, dear
brother.’ Pl. Aul. 175–6)
(i) Quis ad fores est? # Ego sum.
(‘Who’s at the door? # It’s me.’ Pl. Am. 1021)
5
For more examples, see Lodge: s.v. tu 791 § .
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The use of explicit first and second person subjects is favoured (but not required) in
special semantic configurations. Their presence underlines the subject’s personal
involvement (EMPHASIS). First person pronouns are used with verbs expressing a
subjective opinion like existimo ‘to think’, as in (m), with threats, as in (n), and
with promises. The use of tu (and vos) is favoured in imperative sentences (examples
are given above) and in certain questions, for example, with an in a verifying, possibly
surprised, question, as in (o).6 It is also commonly used in appeals to shared
knowledge, as in (p).
(m) Quod ego longe secus existimo, iudices, deque eo pauca disseram.
(‘I disagree entirely, gentlemen, but I shall speak only briefly on the matter.’ Cic. Mur. 31)
(n) Iam quidem hercle ego tibi istam / scelestam, scelus, linguam apscidam.
(‘I’ll cut out this villainous tongue of yours this instant, you villain.’ Pl. Am. 556–7)
(o) Credam istuc si esse te hilarum videro. # An tu ess’ me tristem putas?
(‘I’ll believe this if I see you cheerful. # Do you think I’m depressed?’ Pl. As. 836)
(p) TU OPTIME SCIS QUOMODO ABS TE EXIVERIM.
(‘You know best how I departed from you.’ CEL 147.3 (Karanis, c. AD 115))
Supplement:
Ego istuc scio / ita fore illi, dum quidem cum illo nupta eris. # Ego censeo. (Pl. As.
869–70); Fiam, ut ego opinor, Hercules, tu autem Linus. (Pl. Bac. 155); Quid agimus?
# Praeterea hoc etiam, quod ego vel primum puto. (Ter. Eu. 1081); Non apud
indoctos, sed ut ego arbitror in hominum eruditissimorum et humanissimorum
coetu loquor. (Cic. Pis. 68); Sed tamen est eorum doctrina, quantum ego iudicare
possum, perridicula. (Cic. de Orat. 2.77);
Quoius ego hodie in tergum istaec faxo expetant mendacia. (Pl. Am. 589); Hercle
vero tu cavebis ne me attingas, si sapis . . . (Pl. As. 373); Interimam hercle ego <te>,
si vivo. (Pl. Mos. 1168); Pedicabo ego vos et irrumabo / Aureli pathice et cinaede
Furi . . . (Catul. 16.1–2); Ego flumen Muluccham, quod inter me et Micipsam fuit, non
egrediar neque id intrare Iugurtham sinam. (Sal. Jug. 110.8);
6
For these contexts, see Pinkster (1986b, 1987b), Adams (1999), Fruyt (2008), and Lücht (2011).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Dormiunt. Ego pol istos commovebo. (Ter. Hau. 730); Tum cum ille dubitaret,
quod ad fratrem promiserat, ‘ego’ inquit Iulius ‘pro utroque respondeo’. (Cic. de Orat.
2.27);
In other text types, especially in narrative texts, the use of subject pronouns does not
have the same pragmatic justification as it does in the examples above, but plays a role
in the structure of the discourse. In his Brutus, Cicero reports a discussion in which
he himself participates. In the course of this discussion, he uses ego to indicate that
there is a change of speaker from one of the others to himself, as in (q). It functions in
the same way as the proper name Brutus in (r) and ille in (s). Another example is (t).
Note the presence of autem. All four of these are instances of topic change.7
(q) Hic ego ‘noli’ inquam, ‘Brute, existimare his duobus quicquam fuisse in
nostra civitate praestantius.
(‘Then I replied: “Don’t think, Brutus, that there has ever been in our state a more
distinguished talent than these two men.” ’ Cic. Brut. 148)
(r) Hic Brutus ‘ai’n tu?’ inquit.
(‘At this Brutus said: “Do you mean to say that . . . ?’ Cic. Brut. 152)
(s) Hic ille ‘primum, quis Antonio permisit’ inquit ‘ut et partes faceret et utram
vellet prior ipse sumeret?’
(‘Hereupon, “In the first place,” said Crassus, “who gave Antonius leave to divide the
subject up into shares and to be the first to choose which of the two parts he
wanted?” ’ Cic. de Orat. 2.366)
(t) Haec sunt, ut opinor, in re publica. Ego autem, ut semel . . . non destiti eadem
animi magnitudine in re publica versari et illam institutam ac susceptam
dignitatem tueri.
(‘That is all, I think, in the way of public affairs. As for me, ever since . . . , I continue to
play my part in politics with the same disregard of self and to maintain the position
and responsibilities I then assumed.’ Cic. Att. 1.19.5–6)
Furthermore, the use of the first person singular pronoun ego as subject is favoured in
the shadow of a word that for some reason or other is entitled to occupy the first or an
early position in its clause. This is shown in (u)–(y). In (u), the relative pronoun has
the first position, in (v) the interrogative pronoun in an exclamative sentence, in (w)
the demonstrative pronoun hic. In (x), coquom has replacing focus; in (y), numquam
signals an emphatic denial. There is no obvious pragmatic explanation for the
occurrence of these pronouns.8 There are similar instances of tu.
(u) Etiam Epidicum, quam ego fabulam aeque ac me ipsum amo, / nullam
aeque invitus specto, si agit Pellio.
7
For the need to take the type of text into account, see Pinkster (1986b: 318–19).
8
For these ‘structural’ factors, see Adams (1999: 100–33), who pays special attention to Republican
and Augustan poetry.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(‘Even the Epidicus, a play I love as much as myself—well, there’s no play I enjoy
watching less if Pellio’s acting in it.’ Pl. Bac. 214–15)
(v) Quas ego hic turbas dabo.
(‘What trouble I’ll stir up here!’ Pl. Bac. 357)
(w) Hic mihi corrumpit filium, scelerum caput. / Hic dux, hic illi est paedagogus.
Hunc ego / cupio excruciari.
(‘This man is spoiling my son, the fount of iniquity; he is the leader, he is his tutor,
him I wish to be tortured.’ Pl. Ps. 446–8)
(x) Coquom ego, non furem rogo.
(‘I’m asking for a cook, not a thief.’ Pl. Aul. 322)
(y) Numquam ego hanc viduam cubare sivi.
(‘I've never let her sleep alone.’ Pl. Cist. 44)
Supplement:
Cui ego crimini respondere sine mea maxima laude non possum. (Cic. Dom. 95);
Quibus ego rebus, iudices, ita flectebar animo atque frangebar ut . . . (Cic. Sul.
18); . . . atque id durius est faba et lapillis. / Quod tu si manibus teras fricesque /
non umquam digitum inquinare posses. (Catul. 23.21–3);
Sed quid ego haec memoro? (Enn. Ann. 314V=314S); Quos ego homines effugi, cum in
hos incidi? (Cic. Att. 2.15.3); Quem tu Pseudolum, quas tu mihi praedicas fallacias? (Pl. Ps.
1195); ‘Quem tu mihi’ inquit Mucius ‘Staseam, quem Peripateticum narras?’ (Cic. de Orat.
1.105); Sed quid ego ignaris nequiquam conqueror auris / . . . ? (Catul. 64.164);
Huic ego homini iam ante denuntio . . . (Cic. Div. Caec. 25); Haec tu commoda tam
beata Furi / noli spernere nec putare parvi. (Catul. 23.34–5); His ego non horam
possum durare querelis. (Prop. 1.6.11); Nympha sub hoc ego sum Cereri gratissima
ligno . . . (Ov. Met. 8.771);
Rerum ego vitia collegi, non hominum. (Cic. Att. 14.14.2); Nam quid feci ego
quive sum locutus / cur me tot male perderes poetis? (Catul. 14.4–5); Possum ego in
alterius positam spectare lacerto? (Prop. 2.8.5);
Quibus pro tantis rebus, Quirites, nullum ego a vobis praemium virtutis, nullum
insigne honoris, nullum monumentum laudis postulabo praeterquam huius
diei memoriam sempiternam. (Cic. Catil. 3.26); Avus vero tuus et P. Nasicae tibi
aediliciam praedicaret repulsam, quo cive neminem ego statuo in hac re publica
fortiorem . . . (Cic. Planc. 51); Sed ut ad prima illa redeam, nihil ego illa impudentius,
astutius, lentius vidi. (Cic. Att. 1.12.1); . . . / numquam ego te vita frater amabilior /
aspiciam posthac? (Catul. 65.10–11);
Obviously, first and second person subjects have to be expressed in configurations in
which there is no explicit first or second person verb form, as is often the case in answers, as
in (z), in comparative clauses, as in (aa), with a historic infinitive, as in (ab), in coordinated
structures, as in (ac), etc. Such subjects are not necessarily pragmatically motivated.
With third person subjects there is a wide variety of constituents that may function as
such. Three major groups can be distinguished: (a) noun phrases (nouns, substantival
adjectives, several types of pronouns, numerals) and autonomous relative clauses; (b)
clauses (infinitives, accusative and infinitive clauses, finite clauses of various types
(declarative, interrogative, and imperative)); (c) quotations.
9
For the generic use of the first and second person in the New Testament, see Beneš (1971).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Table 9.1 Noun phrases functioning as subject (in active clauses—first argument)
Type of noun phrase Examples For more
examples,
see §
Noun without Pater nunc intus suo animo morem gerit. passim
or with (‘So now my father is inside indulging his heart’s desire.’
a modifier Pl. Am. 131)
Quae illi ad legionem facta sunt, memorat pater / meus
Alcumenae.
(‘My father is telling Alcumena what happened there
during the campaign.’ Pl. Am. 133–4)
‘Quasi- Credo edepol ego illic inesse argenti et auri largiter. 11.22
substantival’10 (‘By Pollux, I believe there’s all sorts of gold and silver in
adverb it.’ Pl. Rud. 1188)
Bonorum autem partim necessaria sunt ut vita pudicitia
libertas, ut liberi coniuges germani parentes, partim non
necessaria.
(‘Among good things, some of them are necessary, for
instance life, self-respect, freedom, children, wives,
relations, parents, and some not necessary.’ Cic. Part. 86)
10
The term ‘quasi-substantival’ is taken from the OLD (e.g. s.v. largiter § 1). For discussion, see Ripoll
(2010a).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Quantifier . . . coniectura facile fit, / duo quom idem faciunt, saepe ut 11.11
possis dicere / ‘hoc licet inpune facere huic, illi non licet’ . . . ’
(‘ . . . you can readily assess the character of a person, so that,
when two people do the same thing, you can say, “This
person can safely be allowed to do this, that one can’t” . . . ’
Ter. Ad. 822–4)
Finite clauses
(selection)
(i) ut clause Si verum est, quod nemo dubitat, ut populus 15.34, 15.77–15.78
Romanus omnes gentes virtute superarit, non est
infitiandum . . .
(‘If it is true, as no one doubts, that the Roman
people have surpassed all other nations in valour,
it must be admitted . . . ’ Nep. Han. 1.1)
(ii) quod clause Hoc loco percommode accidit quod non adest is 15.11–15.12
qui paulo ante adfuit . . .
(‘Now it happens most conveniently at this point
that there is absent from the court one who was
here but recently . . . ’ Cic. Caec. 77)
(iii) Ø clause Habeas licet. 15.82
(without a (‘You can have her.’ Pl. Epid. 471)
subordinator)
Infinitives and infinitive clauses functioning as subject and object (sometimes called
‘nominal’ or ‘substantival’ infinitives) are found from Early Latin onwards. Examples are
(a)–(d). The infinitive may be accompanied by its second or third argument (not by the
first argument) and by satellites, as in (a), (c), and (d). In (c), the secondary predicates
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
totum and insanum are in the accusative, which is related to the understood generic first
argument. Infinitives in subject (and object) function may be modified in various ways in
conformance with their nominal properties, as in (d), where the infinitive is modified by
the anaphorically used pronoun hoc (see § 11.8 for details).
(a) Petere honorem pro flagitio more fit.
(‘To seek honour thereby in place of disgrace is the usage.’ Pl. Trin. 1035)
(b) Non cadit autem invidere in sapientem. Ergo ne misereri quidem.
(‘Envy, however, does not befall the wise man. Therefore neither does compassion.’
Cic. Tusc. 3.21)
(c) Nam bonum est pauxillum amare sane, insane non bonum est. / Verum
totum insanum amare, hoc est quod meus erus facit.
(‘Yes, it is good to be a little bit in love sanely, but being insanely in love is not good.
But being completely insanely in love, that’s what my master is.’ Pl. Cur. 176–7)
(d) . . . meque . . . hoc ipsum nihil agere et plane cessare delectat.
(‘ . . . and it is just this inaction and utter idleness that charm me . . . ’ Cic. de Orat. 2.24)
Supplement:
Certumne est tibi istuc? # Non moriri certius. (Pl. Capt. 732); Non hercle verbis, Parmeno,
dici potest / tantum quam re ipsa navigare incommodum’st. (Ter. Hec. 416–17); Nam
quibusdam . . . totum hoc displicet, philosophari. (Cic. Fin. 1.1); Carere igitur hoc signifi-
cat: egere eo quod habere velis. Inest enim velle in carendo . . . (Cic. Tusc. 1.88); Quam enim
turpis est adsentatio, cum vivere ipsum turpe sit nobis. (Cic. Att. 13.28.2); Nam fovea atque
igni prius est venarier ortum quam saepire plagis saltum . . . (Lucr. 5.1251–2);11 Tempus
vero conloquio non dare neque accessurum polliceri magnam pacis desperationem
adferebat. (Caes. Civ. 1.11); Res gerere et captos ostendere civibus hostis attingit solium
Iovis et caelestia temptat. (Hor. Ep. 1.17.33–4); . . . pars est meminisse doloris. (Ov. Met.
9.291); Adde quod ingenuas didicisse fideliter artes emollit mores nec sinit esse feros. (Ov.
Pont. 2.9.47–8); Habere eripitur, habuisse numquam. (Sen. Ep. 98.11); Multis ipsum
metuisse nocet. (Sen. Oed. 992); Maledici enim eam sequebatur . . . (Tert. Pud. 14.17);
Cum ergo Christum videre gaudere sit . . . (Cypr. Mort. 5);
11
For Lucretius’ use of the infinitive, see Reinhardt (2010: 208–11).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(b) Nihil gravius, nihil philosophia dignius, nisi idem hoc ipsum ‘honeste
sapienter iuste’ ad voluptatem referret.
(‘Nothing could be more dignified, nothing more worthy of philosophy, if he did not
go on to make pleasure the standard of this selfsame “honourably, wisely, justly”.’ Cic.
Tusc. 5.26)
Table 9.3 Types of constituents functioning as third person subject in four different
texts (only finite clauses; absolute numbers)
Text Proper name or Ø is hic/ille Relative Other Total
noun ( mod.) pronoun
or subst. adj.
Pl. Aul. 178–267 30 17(1*) 3/2 4 10 66
Cic. Ver. 5.86–91 40 13(5*) 1 1/– 3 3 61
Cic. Att. 1.1 24 13(3*) 2 5/– 12 8 66
Liv. 6.11–12 27 15 –/3 9 7 61
NB: The column ‘Other’ includes various expressions, notably indefinite pronouns and clauses.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Included in the Ø column (implicit subject) are instances in which the subject is
lacking in the second member of a coordination pair, as with fecit in (a), which can
better be regarded as due to the rule of conjunction reduction (on which see
Chapter 19). If these instances (marked with *) are ignored, the number of implicit
subjects is small in Cicero. In Plautus, by contrast, the number is relatively large
and is due to the fact that Euclio is speaking in asides about Megadorus who is next
to him on the stage. Livy, in turn, has the most instances of implicit subjects that
refer to an entity that was previously introduced, including clauses with a historic
infinitive, as in (b). In Cicero, implicit subjects occur particularly in subordinate
clauses, also present in the Livian example (censebat, animadvertit). What is
remarkable is the small number of anaphoric and demonstrative pronouns. Hic
and ille in Plautus are mainly deictic, and Cicero’s instances of hic are anaphoric or
cataphoric sg. or pl. neuter hoc and haec. Note the variation in the use of relative
pronouns. (For details on the choice between Ø, is, and the demonstrative pro-
nouns, see Chapter 24.) In Cicero’s narrative there is much internal variation of
protagonists, hence the highest number of nouns and proper names (and the low
number of implicit subjects).12
(a) Sed ubi hic est homo? Ø Abiit nec me certiorem Ø fecit.
(‘But where is this man? He went away and didn’t tell me his decision.’ Pl. Aul. 243–5)
(b) (sc. Manlius) His opinionibus inflato animo, ad hoc vitio quoque ingenii
vehemens et inpotens, postquam inter patres non, quantum aecum Ø
censebat, excellere suas opes Ø animadvertit, primus omnium ex patribus
popularis factus cum plebeis magistratibus consilia Ø communicare, cri-
minando patres, adliciendo ad se plebem iam aura, non consilio Ø ferri
famaeque magnae Ø malle quam bonae esse.
(‘Puffed up with these opinions, and being besides, through a defect of nature,
impetuous and passionate, when he perceived that his abilities did not bring him
that leadership amongst the nobles which he thought they merited, he was the first of
all the patricians to turn demagogue and to cast in his lot with the plebeian
magistrates. He abused the nobles, he courted the favour of the plebs; and swept
along by the breath of popularity and not by good counsel chose rather to be reputed
great than virtuous.’ Liv. 6.11.6–7)
Example (b) shows that more factors may contribute to the correct identification of
an implicit subject. The connection between the preceding context and the new
sentence is made via animo, which usually is understood to be the animus of some
person. The singular forms vehemens and inpotens narrow down the possible number
of referents, and so does sg. masc. factus.
12
Implicit subjects are exceptional in the Romance languages. For the spread of pronominal subjects,
see Sornicola (2002, 2005).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
9.10 Identity of the third person subject inferred from the context
Sometimes the entity to which a third person subject refers has not been introduced as
such into the discourse but has to be inferred from the preceding context or on the
basis of general knowledge.13 Specific words in the context may contain a clue to the
identity of the subject, as in (a) and (b), but sometimes it is the content as a whole that
provides the necessary information, as in (c). In (a), the subject of concupierit is
something like aemulans, to be inferred from the abstract noun aemulatio. In (b), the
subject of nolunt has to be inferred from the two gerundival clauses in the preceding
sentence. In (c), something like spectatores has to be inferred from clamores tota cavea
in the preceding sentence. In (d), the reflexive adjective suo refers to something like
alicuius gaudentis, to be inferred from voluptas.14
(a) . . . est aemulatio aegritudo, si eo quod concupierit alius potiatur, ipse careat.
(‘ . . . rivalry is distress, if that which one desires another man possesses while he
himself lacks it.’ Cic. Tusc. 4.17)
(b) Praetermittendae autem defensionis deserendique officii plures solent esse
causae. Nam aut inimicitias aut laborem aut sumptus suscipere nolunt . . .
(‘The motives for failure to prevent injury and so for slighting duty are likely to be various:
people either are reluctant to incur enmity or trouble or expense . . . ’ Cic. Off. 1.28)
(c) Stantes plaudebant in re ficta. Quid arbitramur in vera facturos fuisse?
(‘The audience stood up to applaud this imagined incident: what are we to think they
would have done in a real case of this sort?’ Cic. Amic. 24—tr. Powell)
(d) Voluptatis autem partes hoc modo describunt, ut malevolentia sit voluptas ex
malo alterius sine emolumento suo, delectatio voluptas suavitate auditus
animum deleniens.
(‘Further, the divisions of pleasure are described in this way, that malice is pleasure
derived from a neighbour’s evil which brings no advantage to oneself; that rapture is
pleasure soothing the soul by charm of the sense of hearing.’ Cic. Tusc. 4.20)
In texts that are intended for a specific audience, the third person subject is regularly
not mentioned explicitly. Examples are (e)–(h). In Cato’s Agr. the subject is often ‘the
farmer’ or ‘the owner’, in medical texts ‘the patient’, or even ‘the most relevant
person’. Sometimes the subject is plural, as in (h). Unspecified third person subjects
are also common in legal texts, as in (i). Here, too, plural subjects occur, as in (j).
In (k), a protocol of a Church conference of AD 411, the plural is used for ‘the
opponents’.15 A harsh example of a third person singular is (l), where in Classical
Latin something like (ali)quis as the subject in the quom clause would be normal.
13
For the types of unexpressed subjects discussed in this section, see Löfstedt (1965/6), with refer-
ences. Herman (1991: 420–5) deals with cases where the reader has to infer the identity of the subject on
the basis of general knowledge. The number of such cases decreases in Late Latin texts.
14 15
For this explanation, see Dougan ad Cic. Tusc. 4.17. See Herman (1991).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Sometimes this use of the third person is difficult to distinguish from the generic use
dealt with in the next section.
(e) Dolia olearia C, labra XII, dolia quo vinacios condat X, amurcaria X . . .
(‘100 oil jars, 12 pots, 10 jars in which he can put grape pulp, 10 for holding amurca . . . ’
Cato Agr. 10.4)
(f ) Si imbecillitas occupavit, pro exercitatione gestatio est. Si ne hanc quidem
sustinet, adhibenda tamen frictio est.
(‘If weakness has supervened, rocking should replace exercise; if the patient cannot
even endure this, nevertheless rubbing should be applied.’ Cels. 3.15.6)
(g) Amicitia voluntas erga aliquem rerum bonarum illius ipsius causa quem
diligit, cum eius pari voluntate.
(‘Friendship is a desire to do good to someone simply for the benefit of the person
whom one loves, with a requital of the feeling on his part.’ Cic. Inv. 2.166)
(h) Admissuras cum faciunt (sc. sues), [et] prodigunt (sc. porcarii) in lutosos limites . . .
(‘At the time of breeding the swineherds drive them into muddy lanes . . . ’ Var. R. 2.4.8)
(i) Si in ius vocat, <ito>. Ni it, antestamino. Igitur em capito.
(‘If plaintiff summons defendant to court, he shall go. If he does not go, plaintiff shall
call witness thereto. Then only shall he take defendant by force.’ Lex XII 1.1—NB:
emendation dubious)
(j) Rem ubi pacunt, orato. Ni pacunt, in comitio aut in foro ante meridiem
causam coiciunto.
(‘When parties compromise the matter, official shall announce it. If they do not
compromise, they shall state outline of case in Meeting-place or Market before noon.’
Lex XII 1.6–7)
(k) Petilianus, episcopus partis Donati, dixit: ‘Profiteantur ad quaesita. Professionem
petivimus, non conflictum’ . . . Augustinus, episcopus ecclesiae catholicae, dixit:
‘Si horum criminum iacturam faciunt, de chartis publicis nihil proferimus.’
(‘Petilianus, bishop of the party of Donatus said: “They must formulate a statement
on what was asked. We have asked for a statement, not a conflict.” Augustinus, bishop
of the catholic church said: “If they drop their accusations, we will publish nothing
from the public archives.” ’ Acta Carth. 3.186–7)
(l) Senex quom extemplo est, iam nec sentit nec sapit, / aiunt solere eum rursum
repuerascere.
(‘As soon as someone is an old man and no longer has his senses or wits about him,
they say that he enters his second childhood.’ Pl. Mer. 295–6)
Supplement:
( . . . faber haec faciat) . . . paullulam pilam ubi triticum pinsat (sc. pistor) . . . (Cato. Agr.
14.2); Columellam ferream . . . cuneis salignis circumfigi oportet bene. Eo plumbum
effundere. Caveat ni labet columella. (Cato Agr. 20.1); Villa[m] aedificanda[m]
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
potissimum ut intra saepta villae habeat aquam . . . (Var. R. 1.11.2); Si neque censu nec
vindicta nec testamento liber factus est, non est liber. (Cic. Top. 10); Capitis dolori
medetur panacis omnium generum radix in oleo contrita, aristolochia, hiberis adalli-
gata hora vel diutius, si pati possit, comitante balinei usu. (Plin. Nat. 25.134); Turbam
enim tantum modo congregat (sc. discipulus), ex qua sine discrimine occupet pro-
ximum quodque. (Quint. Inst. 10.1.7);
There are also situations in which the information provided by the third person finite
verb form itself is sufficient. This is the case when it is not necessary or possible to
specify which entity is involved in the state of affairs the verb refers to, or in other
words when the entity is ‘generic’ (for this term, see § 11.101). Very common is the
use of the third person plural with verbs of saying. Examples are (a)–(c). This verb
form is also used with other verbs with a related meaning, as in (d). This use of a
‘subjectless’ third person plural verb form resembles the use of one in English (‘people
in general’), man in German, and on in French.
(a) . . . nec mutam profecto repertam nullam esse / <aut> hodie dicunt mulierem
<aut> ullo in saeclo.
(‘In fact, people say a silent woman has never been found now or in any generation.’
Pl. Aul. 125–6)
(b) Id quod audivi iam audies. / Stratippoclem aiunt, Periphanei filium, / absen-
tem curavisse ut fieret libera.
(‘You shall hear what I have heard. People say that Stratippocles, the son of Periph-
anes, took care in his absence that she should be freed.’ Pl. Epid. 507–9)
(c) Etenim prope est spelunca quaedam . . . qua Ditem patrem ferunt repente
cum curru extitisse abreptamque ex eo loco virginem secum asportasse . . .
(‘There is indeed in the neighbourhood a cave . . . from which people say that father
Dis suddenly issued with his chariot; he seized the maiden, carried her away thence
with him . . . ’ Cic. Ver. 4.107)
(d) . . . pars cordis parte quadam quam ventriculum cordis appellant . . .
(‘ . . . part is received by a certain part of the heart they call the cardiac ventricle . . . ’
Cic. N.D. 2.138)
Supplement:
‘Actum’ aiunt ‘ne agas’. (Ter. Ph. 419); Quot ego tuas petitiones ita coniectas, ut vitari
posse non viderentur, parva quadam declinatione et, ut aiunt, corpore effugi! (Cic.
Catil. 1.15); Saepe illam perhibent ardenti corde furentem / clarisonas imo fudisse e
pectore voces . . . (Catul. 64.124–5); Namque ferunt olim classi cum moenia divae /
linquentem gnatum ventis concrederet Aegeus / talia complexum iuveni mandata
dedisse: . . . (Catul. 64.212–14); Veritatem laborare nimis saepe aiunt, extingui num-
quam. (Liv. 22.39.19); Adice quod genus ultionis est eripere ei qui fecit factae
contumeliae voluptatem. Solent dicere: ‘O miserum me! Puto, non intellexit’. (Sen.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Dial. 2.17.4); Narrant et in Ponto caecian in se trahere nubes. (Plin. Nat. 2.126); Nam
hoc eius ut tradunt sepulc[h]rum. (Var. L. 5.4.30);
Nomine Burrus uti memorant a stirpe supremo . . . (Enn. A. 178V=166S); Per ver
serito in loco ubi terra tenerrima erit (quam ‘pullam’ vocant), ubi aqua propter siet.
(Cato Agr. 151.2); Eodem die vulgo loquebantur Antonium mansurum esse Casilini.
(Cic. Att. 16.10.1); Adiciunt miracula huic pugnae. (Liv. 2.7.2); ‘Romanus sum’
inquit ‘civis. C. Mucium vocant.’ (Liv. 2.12.8); Aeque notus est morbus quem
interdum arquatum, interdum regium nominant. (Cels. 3.24.1–2); A meridie auster
et ab occasu brumali Africus. Notum et Liba nominant. (Plin. Nat. 2.119);
The third person singular seems to be limited to inquit (and occasionally ait), used
for an imaginary opponent, as in (e).
(e) ‘Nondum gustaverat’, inquit, ‘vitae suavitatem; hic autem iam sperabat
magna, quibus frui coeperat.’
(‘ “The infant had not yet tasted the sweetness of life,” someone objects: “but
the other was already forming high hopes, which he was beginning to
enjoy.” ’ Cic. Tusc. 1.93)
This use of the third person plural in a generic sense is also found outside the range of
verbs mentioned above. Examples are (f )–(h).16 In (i) the subject of the infinitive is a
generic third person plural (possibly a unique attestation). A remarkable instance of
coordination of a generic and a specific third person plural is (j).
(f ) Aedes quom extemplo sunt paratae, expolitae, / factae probe examussim, /
laudant fabrum atque aedes probant. Sibi quisque inde exemplum expetunt . . .
(‘The moment a house has been finished, polished, built to perfection, people praise the
builder and commend the house; everyone wants one like it for himself . . . ’ Pl. Mos. 101–3)
(g) Fugit forum, fug<it>at suos cognatos, / fugat ipsus se ab suo contutu, / neque
eum sibi amicum volunt dici.
(‘The lover flees the forum, flies from his own kin, puts himself to flight from his own
self-scrutiny, and men dislike to have him called their friend.’ Pl. Trin. 261–3)
(h) Quod sequitur vero non solum ad religionem pertinet, sed etiam ad civitatis
statum, ut sine iis qui sacris publice praesint religioni privatae satis facere non
possint.
(‘The provision which follows really has to do with the condition of the State as well
as with religion, so that people could not satisfactorily perform private worship
without those who are in charge of the public rites.’ Cic. Leg. 2.30)
(i) In Tmoli montis cacumine quod vocant Tempsin CL annis vivere Mucianus
auctor est . . .
(‘On the peak of Mount Tmolus which is called Tempsis, Mucianus is the authority
for people living to the age of 150 . . . ’ Plin. Nat. 7.159)
16
The examples are taken from Löfstedt (1965/6: 97–9). In certain grammatical traditions this generic
use of the third person is called ‘impersonal’. For the New Testament, see Beneš (1971).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
For objects of two-place verbs and accusative objects of three-place verbs (and their
subject equivalents in the passive) the same constituents can be found as in Table 9.1
for subjects. There is no need for further detailed illustration, except perhaps for the
relatively rare category of so-called substantival expressions that do not show agree-
ment. In addition to the categories functioning as subject, prepositional phrases
function as second or third argument as well. Examples are (a)–(c) (for further
examples, see §§ 4.37–4.40).17 Adverbs are rarely used for second arguments in
idiomatic combinations like Graece scio, as in (d).18 For autonomous relative clauses
functioning as object, see § 18.16.
(a) Iam tanta religio est sepulchrorum, ut extra sacra et gentem inferri fas
negent esse . . .
(‘Now graves are the objects of so much religious veneration that it is considered
sinful that corpses not belonging to the clan or participating in the rites be buried in
them . . . ’ Cic. Leg. 2.55—NB: If it is an argument, as K.St.: I.2 have it. Inferri may be
taken as impersonal.)
(b) . . . ubi de magna virtute atque gloria bonorum memores, quae sibi quisque
facilia factu putat aequo animo accipit, supra ea veluti ficta pro falsis ducit.
(‘ . . . when you commemorate the distinguished merit and fame of good men, while
every one is quite ready to believe you when you tell of things which he thinks he
could easily do himself, everything beyond that, like made-up stories, he regards as
false.’ Sal. Cat. 3.2)
17
For further examples of prepositional phrases, see von Nägelsbach and von Müller (1905: 43, 311).
18
For the use of adverbs denoting a language that function as arguments, see Ripoll (2007: 159–80).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
As for clauses, the same range is found with objects as with subjects. A few examples
are given in Table 9.4.
Table 9.4 Clauses functioning as object (or as subject in a passive clause)
Type of clause Examples For more
examples, see §
Infinitive Istic est thensaurus stultis in lingua situs, / ut quaestui 9.15
habeant male loqui melioribus.
(‘A fool’s funds being in his tongue, he makes his
investments by abusing his betters.’ Pl. Poen. 625–6)
Et simul illud videtote, aliud habitum esse sepelire et urere.
(‘Note that in this case also burial and cremation are treated
as different things.’ Cic. Leg. 2.60)
Accusative and . . . id ego hoc anno desisse dubitari certum habeo. 15.95–15.99
infinitive (‘ . . . I am certain that the doubt has this year been
resolved.’ Liv. 5.3.2)
Lautissimum quippe habetur e nardi folio eas (sc. coronas)
dari aut veste Serica versicolori, unguentis madida.
(‘In fact it is deemed most elegant that chaplets made of
nard leaves be given or ones made of multicoloured silk
steeped in perfumes.’ Plin. Nat. 21.11.3)
Finite clauses
(i) indirect Quid negoti sit rogas? 15.48–15.61
questions (‘You’re asking what’s the matter?’ Pl. Aul. 296)
Roga numquid opus sit.
(‘Ask if he has need of anything.’ Pl. Poen. 1008)
(ii) ut clauses, Argenti rogo ut faciat copiam. 15.65–15.76
etc. (‘I asked him to provide you with the money.’ Pl. Cur. 330)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In § 4.5 the problem is discussed of verbs that have a two-place frame, but whose
second argument is absent either because the verb frame is conventionally reduced in
specialized contexts or because the state of affairs is unspecified (the so-called
absolute use). The discussion in this section concerns the situation in which there is
a second entity involved in the state of affairs that is implicit and has to be inferred
from the situation (on the stage, for example) or from the preceding or (less
19
On Pelagonius’ use of the infinitive, see Adams (1995b: 479–80).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
frequently) following context.20 The general condition for the object not to be
expressed is that the entity it refers to is well-established in the communicative
situation. That entity may be a first-order entity, as in (a) (a noun), or a third-order
entity, as in (b), where the entire preceding clause is understood. Omission of the
object may occur within a complex sentence, as in (c), or across sentence boundaries,
as in (a). Such omission usually follows an entity mentioned explicitly in the preced-
ing context, as in (a)–(c), but it may also precede explicit mention, as in (d).
A secondary predicate showing agreement with the absent object may help facilitate
correct identification, as in (e).21
(a) Edepol vel elephanto in India / quo pacto ei pugno praefregisti bracchium. / #
Quid, bracchium? # Illud dicere volui: femur. / # At indiligenter Ø iceram.
(‘Or take the elephant in India, how you broke its arm with your fist. # What? Its
“arm”? # I meant to say its “leg”. # But I only hit it casually.’ Pl. Mil. 25–8)
(b) Non hercle te provideram. Quaeso ne vitio Ø vortas.
(‘I really hadn’t noticed you before—please don’t take any offence.’ Pl. As. 450)
(c) Iam se ad prohibenda circumdari opera Aequi parabant, cum ab interiore
hoste proelio coepto, ne per media sua castra fieret eruptio, a Ø munientibus
ad pugnantes introrsum versi vacuam noctem operi dedere.
(‘The Aequi were already preparing to resist the work of circumvallation, when the
attack was begun upon their inner line. Lest a sortie should be made through the
midst of their camp, they turned their backs on those who were entrenching, faced
the attacking forces, and left the others free to work all night.’ Liv. 3.28.7)
(d) . . . ne[c] Ø optantibus quidem nobis, quia non posse fieri videbatur, firmis-
simum exercitum . . . comparavit . . .
(‘ . . . without our . . . even praying for such a thing because it seemed impossible of
accomplishment, . . . he raised a very strong army . . . ’ Cic. Phil. 3.3)
(e) . . . declararique maluerim quanta vis esse potuisset in consensu bonorum, si
iis pro me stante pugnare licuisset, cum adflictum Ø excitare potuissent.
(‘I . . . chose rather that it should be made manifest to all the world, how powerful the
unanimity of loyal citizens might have proved had they been allowed to fight for me
in the days when I stood upright, seeing that they had been able to set me on my feet
again when I lay prostrate.’ Cic. Fam. 1.9.13)
When two coordinated clauses share the same object, omission of one of the objects is
what one expects. This will usually, but not always, be the object in the second
member. Examples are (f ) and (g). Such instances are handled as ‘conjunction
reduction’ in this Syntax (for further examples, see Chapter 19). (In the examples
20
For a collection and classification of instances in Cicero, see Lebreton (1901a: 150–70) and Happ
(1976: 239–61). More recent studies are Mulder (1991); Johnson (1991); Luraghi (1997, 2004); Sznajder
(1998); Pieroni (2001); Spevak (2010a: 99–106).
21
See Luraghi (1997: 246–7).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
below, where the verbs also have the same subjects, one might even hold that they do
not exemplify coordination of clauses, each with their own object, but of verbs.)
(f ) Nam rex Seleucus me opere oravit maxumo, / ut sibi latrones cogerem et Ø
conscriberem.
(‘Well, King Seleucus asked me as a matter of great importance to muster and enrol
mercenaries for him.’ Pl. Mil. 75–6)
(g) Sed inscientes sua sibi fallacia / ita Ø compararunt et confinxerunt dolum . . .
(‘But in their scheme they’ve unknowingly prepared and contrived their trick . . . ’ Pl.
Capt. 46–7)
As for the frequency of Ø objects, this varies from author to author and from text type
to text type. From the two studies in which quantitative data are supplied, it is clear
that Ø objects are more frequent than pronominal objects and that the percentage of
Ø objects presumed to be known on the basis of the preceding context or the situation
varies from 8% to 26% in Classical prose.22
Supplement:
Complex sentences: Di tibi omnes omnia optata <of>ferant, / quom me tanto
honore honestas quomque ex vinclis Ø eximis. (Pl. Capt. 355–6); . . . cepi tabellas,
Ø consignavi clanculum, / Ø dedi mercatori quoidam, qui ad illum Ø deferat / meum
erum . . . (Pl. Mil. 130–2); Quas tu incredibile est quam brevi tempore quanto dete-
riores offensurus sis quam Ø reliquisti. (Cic. Att. 1.11.3); Caesar . . . ad hostes con-
tendit, eo minus veritus navibus, quod in litore molli atque aperto deligatas ad
ancoras Ø relinquebat . . . (Caes. Gal. 5.9.1); (sc. Vercingetorix) . . . convocatis suis
clientibus facile Ø incendit. (Caes. Gal. 7.4.1); Verginius sordidatus filiam secum
obsoleta veste Ø comitantibus aliquot matronis cum ingenti advocatione in forum
deducit. (Liv. 3.47.1); Haec quo die feceris, necessaria, eadem, si cottidie Ø fecisse te
reputes, inania videntur, multo magis, cum secesseris. (Plin. Ep. 1.9.3);
Across sentence boundaries: Vi’n proferri pateram? # Ø Proferri volo. (Pl. Am. 769);
Sci’n me tuom esse erum Amphitruonem? # Ø Scio. (Pl. Am. 1082); Ecquid matrem
amas? # Egone illam? Nunc Ø amo, quia non adest. (Pl. As. 900); Quid tam novum
quam adulescentulum privatum exercitum difficili rei publicae tempore conficere? Ø
Confecit. (Cic. Man. 61); Oriens incendium qui restinguerent summos viros mi-
simus. Ø Repudiasti. (Cic. Phil. 13.48); Quid? Hoc dasne aut manere animos post
mortem aut morte ipsa interire? # Ø Do vero. (Cic. Tusc. 1.25);
In the Romance languages zero objects are very uncommon. A pronominal object is
normally obligatory to refer to known entities. Research of diachronic developments
from Latin to Romance is sparse and inconclusive. The most detailed study suggests a
slight increase of object pronouns from Ammianus to Fredegar.23
22
Mulder (1991: 18) has data on Plautus, Celsus, and Ammianus; Spevak (2010a: 104) on Cicero,
Caesar, and Sallust. The data cannot easily be compared.
23
See Johnson (1991). No increase was found between Plautus and Ammianus (Mulder 1991: 18).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
24
The German term is ‘Verschränkung’, as with relative clauses (see § 18.10).
25
For the properties of the pseudo-object expression, and its pragmatic motivation, see Touratier
(1980a); Bolkestein (1981: 74–89); Maraldi (1986); Rosén (1992); Calboli (1997: 49–66); Bodelot (2003:
205–10); Álvarez Huerta (2005, 2007); Halla-aho (2012), with references.
26
For the history of the proleptic accusative and for references, see Sz.: 471–2 and Halla-aho (2012);
for more examples, see K.-St.: II.579–81 (who used Lindskog 1896) and Bennett (1914: II.223–4). For
pseudo-objects in the Vulgate translation, see Sznajder (2012b).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Cognition and communication verbs (for Plautus in alphabetical order by verb):
Agedum, contempla aurum et pallam sati’n haec <me> deceat, Scapha. (Pl. Mos.
282—NB: pallam repeated in haec); Quid tu me curas quid rerum geram? (Pl. Rud.
1068); Ecastor te experior quanti facias uxorem tuam. (Pl. Am. 508); . . . meminerunt
sese unde oriundi sient. (Pl. Aul. 542); Eos nunc homines metuo mihi ne opsint . . .
(Pl. Mil. 996a); Celabat metuebatque te ne tu sibi persuaderes / ut . . . (Pl. Truc.
200–1—NB: te followed by tu); Sed servom meum / { Strobilum { miror ubi sit . . .
(Pl. Aul. 696–7); Verum meam uxorem, Libane, nescis qualis siet? (Pl. As. 60); Nunc
ego lenonem ita hodie intricatum dabo, / ut ipsus sese qua se expediat nesciat. (Pl. Per.
457–8); At pol ego eum qua sit facie nescio. (Pl. Poen. 592); . . . quom illum rescisces
criminatorem meum / quanto in periclo et quanta in pernicie siet. (Pl. Bac. 826–7);
Vide’n tu ignavom, ut sese infert? (Pl. Mil. 1045); Nunc lenonem quid agit intus visam,
convivam meum. (Pl. Rud. 592); Nunc ego Simonem mihi obviam veniat velim . . . (Pl.
Ps. 1061); Prius villam videat clausa uti siet et uti suo quisque loco cubet . . . (Cato Agr.
5.5); Metuo fratrem / ne intus sit; porro autem pater ne rure redierit iam. (Ter. Eu.
610–1—NB: variation); Illum ut vivat optant, meam autem mortem exspectant scilicet.
(Ter. Ad. 874); Nunc, Parmeno, ostendes te qui vir sies. (Ter. Eu. 307); Ille flens narrat
ab nescio quo percussum cultello concidisse, quem qui esset animadvertere in turba
non potuisse . . . (Var. R. 1.69.2); Quis tuum patrem ante quis esset quam cuius gener
esset audivit? (Cic. Deiot. 30); Veteranos non veremur—nam timeri se ne ipsi quidem
volunt—quonam modo accipiant severitatem meam? (Cic. Phil. 12.29); . . . Sophocles
ob oculos versabatur, quem scis quam admirer quamque eo delecter. (Cic. Fin. 5.3—
NB: continuation with eo); Et nosti virum quam tectus. (Cic. Att. 14.21.2); . . . rem
frumentariam ut satis commode supportari posset timere dicebant. (Caes. Gal. 1.39.6);
Tum ait miles: Iam me quis sim intelleges . . . (B. Afr. 16.3); Nosti Marcellum, quam tardus
et parum efficax sit, itemque Servius (Servium cj. Manutius), quam cunctator. (Cael. Fam.
8.10.3—NB: variation); Quem legis expertes Latinae / Vindelici didicere nuper / quid
Marte posses. (Hor. Carm. 4.14.7–9); Aspice me, cui parva domi fortuna relicta est, /
nullus et antiquo Marte triumphus avi, / ut regnem mixtas inter conviva puellas . . .
(Prop. 2.34.55–7); Eum postulare ut sibi dedatur . . . (Liv. 23.10.3); Hunc Aristonem
Carthagine obversantem non prius amici quam inimici Hannibalis qua de causa venisset
cognoverunt. (Liv. 34.61.4); . . . (sc. cibos) eosque nemo dubitat quin optimi sint . . . (Col.
6.3.3); Si . . . (sc. ossilaginem) eam facere voles ne crescat . . . (Mulom. Chir. 642);
Nam et patriam et patrem commemorant pariter qui fuerint sibi. (Pl. Men.
1083); . . . tuae / uxori rem omnem iam uti sit gesta eloquar. (Pl. Men. 518–19); At
ego illam quaero quae fuit. (Pl. Per. 636); Responde, quo leto censes me ut peream
potissumum? (Pl. Mer. 483);
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Causation verbs: Sati’n est si hanc hodie mulierem efficio tibi / tua ut sit . . . (Pl. Ps.
112–13); Tu facis me quidem ut vivere nunc velim. (Pl. Rud. 244); Quor eiulas, /
quem ego avom feci iam ut esses filiai nuptiis? (Pl. Aul. 796–7—NB: with object
complement); . . . eumque ita / faciemus ut quod viderit ne viderit. (Pl. Mil. 148–9—
NB: with object complement); . . . ego omnis hilaros, ludentis, laetificantis faciam ut
fiant . . . (Pl. Per. 760—with object complement); . . . nec potui tamen / propitiam
Venerem facere uti esset mihi. (Pl. Poen. 453–4—NB: with object complement);
ATQUE / UTEI·EA·BACANALIA . . . FACIATIS·UTEI·DISMOTA·SIENT (CIL I2.581.27–30 (SCBac.,
Tiriolo, 186 BC)); Vectes iligneos, acrufolios, laureos, ulmeos facito uti sient parati.
(Cato Agr. 31.1); Quod, si lupus aliusve quis his vulneratus est, reliquas quoque canes
facit quae id non habent ut sint in tuto. (Var. R. 2.9.15); . . . easque cellas provident ne
habeant in solo umorem . . . (Var. R. 3.10.4); Alii aquam mulsam in vasculis prope ut
si[n]t curant . . . (Var. R. 3.16.28);27 Quae est igitur natura, quae volucris huc et illuc
passim vagantis efficiat ut significent aliquid . . . (Cic. Div. 2.80); Omnia perfecit quae
senatus salva republica ne fieri possent perfecerat. (Cic. Phil. 2.55); Quod Caesaris,
quod Pompei gratiam tenemus, haec me ut confidam faciunt. (Cic. Q. fr. 2.15.2); . . .
sagittas telaque uti fierent complura curare (B. Afr. 20.3); Philosophia vero perficit
architectum animo magno et uti non sit adrogans . . . (Vitr. 1.1.7); . . . lotionibus . . . et
cocturis . . . efficiuntur, ut adveniant, colores. (Vitr. 7.9.1—NB: remarkable word
order); Fac me . . . ut sim certus . . . (Gel. 19.13.2);
Other: Interea unamquamque turmam, manipulum, cohortem temptabam quid
facere possent. (Cato orat. 35);
Difficult to judge are such cases with the verb iubeo ‘to order’. Whereas in (i) famulos
is often regarded as a real object addressee of iube, this is impossible with quem in (j).
Ex. (k) is also better taken as a pseudo-object (see § 15.99 for a discussion of iubeo
and related verbs).28
(‘Also command that the said camels be given water and that they should
return(?)’ M 1107.6–8 (Maximianon (El Zerqa), II AD))
As a parallel of the pseudo-object expressions dealt with above there are also PSEUDO-
SUBJECT expressions with a personal passive construction. There is one early instance
in Terence, but the phenomenon is common enough in Cicero, and also in Vitruvius.
27
For Varro’s usage, see Laughton (1960: 6).
28
For ex. (j), see Bendz (1941: 31); for (k), see Halla-aho (2009: 113–14).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Examples are (l) and (n)–(p). In (l), res is the subject of est videnda and of the
subordinate clause in tuto ut conlocetur. With video in this meaning ‘to watch out’, the
object (subject in the passive) is normally a clause. Ex. (l) corresponds precisely to
the pseudo-object example in (m), where villam is both the object of videat and the
subject of ut clausa sit.
(l) Nam amici . . . res est videnda in tuto ut conlocetur . . .
(‘For we need to make sure your friend’s affairs too are put on a safe footing . . . ’ Ter.
Hau. 689–90)
(m) Prius villam videat clausa uti siet et uti suo quisque loco cubet et uti iumenta
pabulum habeant.
(‘Before then he must see that the farmstead is closed, that each one is asleep in his
proper place, and that the stock have fodder.’ Cato Agr. 5.5—NB: variation)
(n) Nec mirum, cum . . . etiam Teucer Livii post XV annos ab suis qui sit ignoretur.
(‘Nor is this astonishing: . . . when even Teucer’s own family, in the play of Livius
Andronicus, do not know who he is after his absence of fifteen years.’ Var. L. 7.3)
(o) . . . non crederes leones et pantheras esse, cum tibi quales essent Ø dicerentur . . .
(‘ . . . you would refuse to believe that lions and panthers exist, when they were
described to you . . . ’ Cic. N.D. 1.88)
(p) Quae nobis designatis timebatis, ea ne accidere possent consilio meo ac
ratione provisa sunt.
(‘By resolution and judgement I have taken steps to prevent what you feared might
happen when we were elected consuls.’ Cic. Agr. 2.102)
Supplement:
Nam superiore parte legis quem ad modum Pompeium Ø oppugnarent, a me indicati
sunt. Nunc iam se ipsi indicabunt. (Cic. Agr. 1.5); Seiunctus igitur orator a philoso-
phorum eloquentia a sophistarum ab historicorum a poetarum explicandus est nobis
qualis futurus sit. (Cic. Orat. 68); Quidam saepe in parva pecunia perspiciuntur quam
sint leves . . . (Cic. Amic. 63); Atqui multa quaeruntur in Mario fictane an vera sint . . .
(Cic. Leg. 1.4); Sed haec paucis diebus ex illius ad nostra responsa responsis intelle-
gentur quorsum evasura sint. (Cic. Att. 7.17.4); . . . nisi ista pecunia gravissimis esset
certissimisque monumentis testata, cui data, quo senatus consulto, quibus tuis, quibus
meis litteris P. Sestio tradita esset. (Cic. Fam. 5.20.5); . . . aspiciantur animoque adver-
tantur, qua membratura sint qui circa eos fontes habitant homines. (Vitr. 8.4.1);
This use of the accusative and nominative is related to the use of the accusative in the
accusative and infinitive clause and the nominative and infinitive clause, respectively.
An alternative way of promoting a constituent of a subordinate clause to a prominent
position in the main clause is the use of a theme constituent with the preposition de,
as in (q) (for further examples, see Chapter 22).29
29
For the differences between the pseudo-object and theme expressions, see Bolkestein (1981).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Proper name Quae illi ad legionem facta sunt memorat pater / meus
Alcumenae.
(‘My father is telling Alcumena what happened there
during the campaign.’ Pl. Am. 133–4)
(Continued)
30
For this type of explanation, see Sz.: 471–2.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
that in the course of time came to serve as substitutes of the dative. The autonomous
relative clause is also included. More examples are found elsewhere in this Syntax.
Various types of clauses are used as third argument as well. This is illustrated for the
verb admoneo in (b)–(e). An example of an infinitive clause is (b), of an accusative
and infinitive clause (c), of a finite ne and ut clause (d), and of an indirect question (e).
For more examples with other verbs, see Chapter 15.
(b) ‘Nonne te . . . Quinta illa Claudia aemulam domesticae laudis in gloria mu-
liebri esse admonebat . . . ’
(‘ . . . did my descendant, the famous Quinta Claudia, not give you counsel either, to
rival her family renown in the sphere of womanly honour?’ Cic. Cael. 34)
(c) Tantum te admonebo, si illi absenti salutem dederis, praesentibus te his
daturum.
(‘I will merely remind you that in granting life to the absent Ligarius you will grant it
to all these here present.’ Cic. Lig. 38)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Adjective (including Sed Bacchis etiam fortis tibi visa est? 9.23
participial adjective and (‘But did Bacchis seem attractive to you?’ Pl. Bac. 216)
gerundive) . . . quia illa foeda atque intoleranda viris videbantur,
haec sequi decrevistis.
(‘ . . . but since such conditions seemed base and
intolerable to true men, you decided upon this
course.’ Sal. Cat. 58.14)
Quantifier Sed si parum multi sunt qui nobilitatem ament, num 9.24
ista est nostra culpa?
(‘If, then, the lovers of birth are less numerous than
you would have them, is that my client’s fault?’ Cic.
Planc. 18)
Numeral Quoi, malum, erae? # Duae sunt istae (sc. erae). 9.25
(‘What mistress, damn it? # They are two in number.’
Pl. Truc. 801)
Quae usque ad Carneadem perducta, qui quartus ab
Arcesila fuit, in eadem Arcesilae ratione permansit.
(‘And right down to Carneades, who was fourth in
succession from Arcesilas, it continued to remain true
to the same theory of Arcesilas.’ Cic. Ac. 1.46)
Pronoun 9.27
anaphoric/demonstrative Eho, quaeso an tu is es? / # Is enim vero sum.
(‘Dear me! Can it be you’re that man? # I am him,
indeed.’ Pl. Trin. 986–7)
personal Hoc memorabile est. Ego tu sum, tu es ego. Unianimi
sumus.
(‘This is remarkable: I am you, you are me, we’re in
full accord.’ Pl. St. 731)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Autonomous relative Pro Iuppiter, / hic est quem ego tibi misi natali die. 18.16
clause (‘Great heavens! This is what I send you on your
birthday!’ Pl. Cur. 655–6)
Infinitive clause Quid est agrum bene colere? Bene arare. Quid 9.38
secundum? Arare. Tertio? Stercorare.
(‘What is good cultivation? Good ploughing.
What next? Ploughing. What third? Manuring.’ Cato
Agr. 61.1)
(a) Haec urbs Epidamnus est, dum haec agitur fabula. / Quando alia agetur (sc.
fabula), aliud fiet oppidum.
(‘This city is Epidamnus as long as this play is being staged. When another is staged
it’ll become another town.’ Pl. Men. 72–3)
(b) Sed num fumus est haec mulier quam amplexare?
(‘But is the girl you’re embracing smoke?’ Pl. As. 619)
(c) Formido male / ne ego hic nomen meum commutem et Quintus fiam e
Sosia.
(‘I’m terribly afraid that I might change my name here and turn from Sosia into Sosia
the Fifth.’ Pl. Am. 304–5)
(d) Homines enim sumus et occupati officiis . . .
(‘For we are but human and beset with duties . . . ’ Plin. Nat. pr. 18)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Certain abstract nouns that occur as subject complement in the nominative are also,
and often more frequently, found in the dative. Examples with the noun impedimen-
tum are (e) (nominative), (f ) and (g) (datives). With the dative expression another
dative constituent is normal. For other differences between the two expressions,
see § 9.34.31
Instead of nouns, substantival neuter sg. adjectives may be used as subject comple-
ment, as in (h). Less common are instances like (i) and (j). Exceptional, and maybe
due to Greek influence, are instances like (k).32
(h) Ut enim non efficias quod vis, tamen mors ut malum non sit efficies.
(‘For though you may not succeed in your wish, still you will succeed in showing that
death is not an evil.’ Cic. Tusc 1.16)
(i) Pax est tranquilla libertas, servitus postremum malorum omnium . . .
(‘Peace is tranquil liberty, servitude is the worst of all evils . . . ’ Cic. Phil. 2.113)
(j) Gravius autem tumultum esse quam bellum hinc intellegi potest quod bello
[Italico] vacationes valent, tumultu non valent.
(‘And that a public emergency is something more serious than a war can be inferred
from the fact that exemptions from military service are valid in a war but are not valid
in a public emergency.’ Cic. Phil. 8.3)
(k) Triste lupus stabulis, maturis frugibus imbres, / arboribus venti, nobis Ama-
ryllidis irae.
(‘Terrible is the wolf to the folds, the rains to the ripened crop, to the trees the gales,
and to me the anger of Amaryllis!’ Verg. Ecl. 3.80–1)
31
Editors sometimes emend nominatives to datives. See Axelson (1933: 90–1) for Seneca.
32
For further examples, see K.-St.: I. 32; discussion in Orlandini (1994: 172–4); Pieroni (forthc.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) Atque etiam nunc satis boni sunt (sc. postis), si sunt inducti pice.
(‘And even now they’re good enough if they’re coated with tar.’ Pl. Mos. 827)
(b) Agite bibite, festivae fores, / potate, fite mihi volentes propitiae.
(‘Go on, drink, dearest door; imbibe and be favourable and well-disposed towards
me.’ Pl. Cur. 88–9)
(c) Fit vasta Troia, scindunt proceres Pergamum.
(‘Troy is being levelled, the chiefs are sacking the city.’ Pl. Bac. 1053)
(d) Est tuum nomen utraque familia consulare.
(‘You are of consular rank both on your father’s and your mother’s side.’ Cic.
Planc. 18)
(e) Nunc te (sc. Priapum) marmoreum pro tempore fecimus. At tu, / si fetura
gregem suppleverit, aureus esto.
(‘Now we have made you of marble for the time; but if births make full the flock, then
you shall be of gold.’ Verg. Ecl. 7.35–6)
(f ) Ac si quis est talis, qualis esse omnis oportebat, qui . . .
(‘But if there is anyone of such a sort (such as was fitting for everyone to be), who . . . ’
Cic. Catil. 2.3)
Supplement:
Cecidi ego, cadet qui sequitur, laus est publica. (Laber. com. 170);33 Nam illum
agrum publicum esse fatentur (Cic. Agr. 2.57); Peloponnesias civitates omnis
maritimas esse . . . Dicaearchi tabulis credidi. (Cic. Att. 6.2.3); Ex iis omnibus longe
sunt humanissimi qui Cantium incolunt, quae regio est maritima omnis . . . (Caes.
Gal. 5.14.1);
33
Equivalent to laudare populi est (Rosén 1981: 65).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(g) . . . ille populus Romanus, cuius usque ad nostram memoriam nomen invic-
tum in navalibus pugnis permanserit . . .
(‘ . . . the great People of Rome, whose name, till within our own memory, remained
invincible in naval battles . . . ’ Cic. Man. 54)
(h) . . . quia illa foeda atque intoleranda viris videbantur, haec sequi decrevistis.
(‘ . . . but since such conditions seemed base and intolerable to true men, you decided
upon this course.’ Sal. Cat. 58.14)
(i) NAM·QUID / <TAM> EGREGIUM·QUIDVE·CUPIENDUM·EST·MAGIS·QUAM . . . ?
(‘For what is so distinguished or what should be more desired than . . . ?’ CIL
VI.9632.6–7 (Rome, probably Early))
Adjectives of the second declension and of the third declension in -is can be used as
subject complement with clauses as their subject, as in (j) and (k) (see § 15.32 and
§ 15.101 for more examples).
(j) Nam in metu esse hunc illi’st utile.
(‘It is in his interest for his son to be afraid of him.’ Ter. Hau. 199)
(k) Nam, si nomine appellari abs te civis tuos honestum est, turpe est eos
notiores esse servo tuo quam tibi.
(‘For if it is honourable for you to call your fellow-citizens by name, it is a disgrace
that they should be better known to your slave than to you!’ Cic. Mur. 77)
34
For more instances of is, see TLL s.v. is 462.60ff. See also Bortolussi and Sznajder (2008).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
The label ‘possessive genitive’ is used in a strict sense for a possessor : possessee
relation between two first-order entities.36 Typical examples are (a) and (b), with a
human possessor and an inanimate (concrete) possessee. Such a relationship involv-
ing two human beings is shown in (c). There are examples of a genitive expression
coordinated or parallel with a possessive pronoun, as in (d). The semantic equivalence
of the sum + genitive of the possessor expression to the habeo expression is suggested
by (e). It is not difficult to find corresponding parallels for these examples at the noun
phrase level (see § 11.47). The use of the possessive genitive as a subject complement
35
For such tests, see Lavency (19972: 147ff.) and Cabrillana (2002).
36
For the linguistic category of possession and the ways it is expressed in various languages, see Heine
(1997). For Latin, see Baldi and Nuti (2010).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
37
For statistical data concerning the various possession expressions, see Baldi and Nuti (2010: 306–7).
For the semantic and pragmatic differences between the possessive dative expression and the genitive
subject complement expression, see Baldi and Nuti (2010: 313–22), Bolkestein (1983), and Cabrillana
(2002).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In (f ) and (g) the possessor and possessee are both inanimate. In (f ) the possessor is
abstract, but a human possessor is understood. The relationship between possessor
and possessee can still be understood as ‘possession’.
(f ) . . . non ut esset sui iuris ac mancipi res publica.
(‘ . . . not so that the rights of ownership over the commonwealth should belong to the
commonwealth itself.’ Brut. ad Brut. 1.16.4)
(g) Animos, qui nostrae mentis sunt, eosdem in omni fortuna gessimus . . .
(‘But our aspirations, which are subject to the will of our minds, we have kept
unchanged in every kind of fortune . . . ’ Liv. 37.45.12)
Supplement:
Ars enim earum rerum est quae sciuntur. (Cic. de Orat. 2.30); Idem de servo dicam.
Mei mancipii res est . . . (Sen. Ben. 5.19.1);
However, in (h) and (i) the semantic relation is rather that something is ‘typical of ’,
‘suitable for’, ‘the duty of ’ somebody or something else (sometimes called the
genetivus proprietatis). Here, parallels at the noun phrase level are less easy to find.
(h) Signa nostra . . . velim imponas, et si quid aliud . . . et maxime quae tibi pa-
laestrae gymnasique videbuntur esse.
(‘Yes, I should be grateful if you would ship . . . my statues . . . and anything else . . .
especially things you think suitable to a palaestra and lecture hall.’ Cic. Att. 1.10.3)
(i) Magni pectoris est inter secunda moderatio.
(‘It is the sign of a great spirit to be moderate in prosperity.’ Sen. Suas. 1.3)
Supplement:
Haec mage sunt hominis. (Ter. Ad. 736); Temeritas est videlicet florentis aetatis,
prudentia senescentis. (Cic. Sen. 20);
The same semantic relation is present in instances such as (j)–(n), where the possessee
is a clause, an accusative and infinitive clause in (j), an infinitive in (k), or a finite
clause introduced by a subordinator such as ut in (l) and quod in (m). In (j)–(l) the
possessor is a human being, in (m) and (n) an abstract entity (but note vestrum). In
(n), quod refers to the preceding context. More or less parallel with such genitives is
the use of meum ‘mine’ in (o). Note that for the genitive abstract entities there is no
equivalent with the possessive dative expression (for which see § 4.26).
(j) Sed nunc ea me exquirere / iniqui patris est.
(‘Well, for me to look into that now is to behave like an unreasonable father.’ Ter. An.
186–7—tr. Brown)
(k) Nam . . . sapientis est consilium explicare suum . . .
(‘It belongs to the wise man to expound his advice . . . ’ Cic. de Orat. 2.333)
(l) Est miserorum ut malevolentes sint atque invideant bonis.
(‘It’s typical of wretched people to be spiteful and jealous of respectable ones.’ Pl. Capt. 583)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(m) Quod sordidatus fui luctus est. Quod flevi pietatis est. Quod non accusavi
timoris est. Quod repulsus est vestrum est.
(‘That I was in mourning is due to grief; that I wept, to affection; that I did not accuse
you, to fear; that he was rejected is your doing.’ Sen. Con. 10.1.1)
(n) . . . confidens mala consuetudine loquendi in vitio ponitur, ductum verbum a
confidendo, quod laudis est.
(‘ . . . “confident” is by a mistaken usage of speech used in a bad sense, though the
word is derived from confidere, “to have trust”, which implies praise.’ Cic. Tusc. 3.14)
(o) Nam praefinire non est meum.
(‘It is not my place to lay down the law.’ Cic. Tusc. 5.22)
Supplement:
Infinitive clause as subject: . . . simulare certe est hominis. (Ter. Ad. 734); Ergo prae-
torum est ante et prae ire. (Lucil. 1160M=1178K); (sc. est) . . . viri fortis ne suppliciis
quidem moveri. (Cic. Mil. 82); . . . tamen et tardi ingenii est rivolos consectari, fontis
rerum non videre, et iam aetatis est ususque nostri a capite quod velimus arcessere et
unde omnia manent videre. (Cic. de Orat. 2.117); Duri enim hominis vel potius vix
hominis videtur periculum capitis inferre multis. (Cic. Off. 2.50); Tantae molis erat
Romanam condere gentem (Verg. A. 1.33); Pauperis est numerare pecus (Ov. Met.
13.824); . . . non liberalis, sed levis arbitrabatur polliceri quod praestare non posset.
(Nep. Att. 15.1); Non operae est persequi ut quaeque acta in his locis sint . . . (Liv.
33.20.13); . . . periti medici est non protinus ut venit adprehendere manu brachium
(Cels. 3.6.6); Incedere magno comitatu, splendido cultu, non est fortunae meae. (Sen.
Con. 10.1.3); Spectacula enim edita effusasque opes aut operum magnificentiam in hac
parte enumerare luxuriae faventis est. (Plin. Nat. 7.94); Primum ergo miserationis fuit
non expavescere . . . (Plin. Nat. 8.60); Inpudentissimae igitur stultitiae est vitam aeter-
nam a talibus diis petere vel sperare . . . (August. Civ. 6.1.4); Nunc vero summae
stultitiae est renuntiare saeculo, dimittere patriam . . . (Hier. Ep. 58.4.4);
Indirect question: Tui consili est, si tempus, si senatus coget, si honeste a nobis
recusari non poterit, velisne perseverare. (Cael. Fam. 8.10.5);
Pronominal subject: Si hoc magni cuiusdam hominis et persapientis videtur . . .
(Cic. Prov. 44); Est enim hoc Gallicae consuetudinis, uti et viatores etiam invitos
consistere cogant . . . (Caes. Gal. 4.5.2);
Subject to be inferred from preceding context: Intellego, / et bene monitum duco,
atque esse existumo humani ingeni. (Pl. Mos. 813–14);
Noun phrases in the genitive are used as subject complement to indicate a property of
the subject, the so-called genitive of description or quality (genetivus qualitatis).
Examples are (a)–(c). They can be coordinated with adjectives in the same function,
as in (b) and (c). This use is comparable with the use of genitive noun phrases at the
noun phrase level (see § 11.48 for further details and for a discussion of the
relationship between these expressions and ablative expressions).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
38
See Löfstedt (1933: II.408, n.2), Moure Casas (1998), and Haverling (2006b: 354–6).
39
For the various subject complement possibilities of onus, see TLL s.v. onus 647.19ff.
40
For other instances in the Peregrinatio, see Väänänen (1987: 31).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Noun phrases in the genitive that denote a whole to which another entity belongs or is
part of can be used as subject complement. Examples are (a)–(d). The same relation-
ship is also found at the noun phrase level (see §§ 11.50–11.55).
(a) Cedo signum, si harunc Baccharum es.
(‘Give me the password, if you’re one of these bacchantes.’ Pl. Mil. 1016)
(b) Cuius erat ordinis? Senatorii.
(‘To what order did he belong? The senatorial.’ Cic. Clu. 104)
(c) Fies nobilium tu quoque fontium . . .
(‘Thou, too, shalt be numbered among the far-famed fountains . . . ’ Hor. Carm.
3.13.13)
(d) . . . EUM QUOQ(UE) QUI ORDINIS SENATORI NONDUM ESSET . . .
(‘ . . . even that man, who was not yet a member of the senate . . . ’ S.C. de Cn. Pisone
patre 21, AD 20)
The label ‘partitive genitive’ is also used for entirely different expressions, where, at
least synchronically, no part/whole relation is at stake. The genitives are either
abstract nouns—(e) and (f )—or neuter adjectives, as in (g) (see also § 11.58, from
which these instances cannot always be distinguished).42
(e) Hoc signi erit, ubi calx cocta erit.
(‘This will be the indicator when the whole has calcined.’ Cato Agr. 38.4)
(f ) Hoc commodi est quod ita vixerunt ut testes omnes, si cuperent, interficere
non possent.
(‘But it is of benefit they led such a life that they could not have destroyed all the
witnesses, even if they had so desired.’ Cic. S. Rosc. 91)
41
See TLL s.v. mos 1528.84ff. Also Haverling (2006b: 354–6, 2012: 95–6).
42
So, for example, Sz.: 59. Adjectives of the third declension seem to be excluded from this construc-
tion. See Baldi (1977, 1983).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
In the Supplement follow residual instances of nouns and noun phrases in the
genitive which are difficult to classify on semantic and formal grounds.
Supplement:
Is (sc. a statue of Hercules) dicebatur esse Myronis, ut opinor, et certe. (Cic. Ver. 4.5);
Latro hanc controversiam, quasi tota offici esset, declamavit. (Sen. Con. 7.4.3);
Quisquis est deus, si modo est alius, et quacumque in parte, totus est sensus, totus
visus, totus auditus, totus animae, totus animi, totus sui. (Plin. Nat. 2.14);43 Nemo
dupondii evadit. (Petr. 58.14); Quod si rei interest, minoris fit pecuniae condemna-
tio quam . . . (Ulp. dig. 13.4.2.pr.);
43
For a discussion of this difficult passage, see Luque (2014).
44
Roby (1882: xxvff.) and Bennett (1914: II.171ff.) have extensive collections of instances.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Examples of modified nouns are (d) and (e). Ex. (e) has two copulative clauses, one
with an adjective, the other with a dative noun in the function subject complement.
The subject of the sentence may be a clause, as in (f ), with quod (for more examples,
see § 15.101). The combination of the verb and the predicative dative regularly
governs another dative constituent, as is in (a), thus behaving like a support verb
combination (see § 4.4).
(a) Etiam histriones anno quom in proscaenio hic / Iovem invocarunt, venit.
Auxilio is (= iis) fuit.
(‘Last year, when the actors called upon Jupiter here on stage, he also came and was a
source of aid to them.’ Pl. Am. 91–2)
(b) . . . idque eo mihi magis est cordi, quod . . .
(‘ . . . and this thought is the more pleasing to me because . . . ’ Cic. Amic. 15)
(c) Multa, cibo quae sunt, vitalia multaque morbos / incutere et mortem quae
possint adcelerare.
(‘Many (which serve as food) being useful to life, and many such as can strike us with
disease and make death come quickly.’ Lucr. 6.771–2)
(d) Nil me paenitet iam quanto sumptui fuerim tibi.
(‘I have put you to enough expense already.’ Pl. Mil. 740)
(e) M. Antonius ad me tantum de Cloelio rescripsit, meam lenitatem et clem-
entiam et sibi esse gratam et mihi voluptati magnae fore.
(‘Mark Antony has replied to me about Cloelius, simply expressing his gratitude for
my leniency and clemency and assuring me that it will turn out a source of great
pleasure to myself.’ Cic. Att. 14.19.2)
(f ) Quo antiquissumos augures non esse usos argumento est quod decretum
collegii vetus habemus omnem avem tripudium facere posse.
(‘That such a practice did not prevail with the augurs of ancient times is proved by an
old ruling of our college which says, “Any bird may make a tripudium”.’ Cic. Div. 2.73)
Supplement:
With the verb sum (understood): Is apud scortum corruptelae est liberis . . . (Pl. As.
867); . . . quod viro esse odio videas, / tute tibi odio habeas. (Pl. Men. 111–11a);
Gaudio ero vobeis— # At edepol nos voluptati tibi. (Pl. Poen. 1217); Ita nunc per
urbem solus sermoni omnibu’st . . . (Pl. Ps. 418); Quapropter haec res ne utiquam
neclectu’st mihi. (Ter. Hau. 357); Sed dissimulabam. Ita magno desiderio fuit ei
filius. (Ter. Hau. 753—NB: attribute); (sc. id) Et rei et virtuti et gloriae erit (sc. patri
familias). (Cato Agr. 3.2); . . . greges . . . quibus rupes ac silvae ad pabulandum cordi.
(Var. R. 2.10.3); . . . quae domus erat ipsa indicio crudelissimi tui dominatus . . . (Cic.
Dom. 110—NB: attribute, the ms. reading is usually changed to indicium
(Niländer)); . . . eisque eam rem magnae laudi magnoque honori fore? (Cic. Phil.
6.6—NB: attribute); Potestne bonum cuiquam malo esse, aut potest quisquam in
abundantia bonorum ipse esse non bonus? (Cic. Parad. 7); Valitudinem respondeo
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
causae, non maestitiam fuisse. (Cic. Amic. 8);45 Mihi autem non minori curae est
qualis res publica post mortem meam futura, quam qualis hodie sit. (Cic. Amic. 43—
NB: attribute); Aliter ampla domus dedecori . . . domino est. (Cic. Off. 1.39); . . . te ad
me rescripsisse eam rem summae tibi voluptati esse. (Cic. Att. 2.25.1—NB: attri-
bute); Sin timore defugiant illis se oneri non futurum et per se rem publicam
administraturum. (Caes. Civ. 1.32.7); Praeterea decernit uti . . . Cicero urbi praesidio
sit. (Sal. Cat. 36.3); . . . nostris vehementer carus, Numantinis maxumo terrori esset.
(Sal. Jug. 7.4—NB: parallelism with carus; attribute maxumo); Sed quoniam eo natus
sum, ut Iugurthae scelerum ostentui essem . . . (Sal. Jug. 24.10—NB: attribute); Cibus
illis advorsum famem atque sitim, non lubidini neque luxuriae erat. (Sal. Jug. 89.8);
Res secundae mire sunt vitiis optentui. (Sal. apud Sen. Con. 9.1.13); Id quantae saluti
fuerit universae Graeciae bello cognitum est Persico. (Nep. Them. 2.4—NB: attri-
bute); Et huic proelium cum Tuscis ad Ianiculum erat crimini. (Liv. 2.52.7); Invidiae
tamen res ad exercitum fuit (Liv. 3.31.4); . . . ipsa res ad levandam publica cura
annonam impedimento fuerat . . . (Liv. 4.13.2); Quae (parietes) si non erunt, vox ibi
disputantium elata in altitudinem intellectui non poterit esse audientibus. (Vitr.
5.2.2); Ulceribus gallae tritae remedio sunt nec minus sucus marrubii cum fuligine.
(Col. 6.13.2); Contra rabiosi quoque canis morsum nuclei a ieiuno homine com-
manducati inlitique praesenti remedio esse dicuntur. (Plin. Nat. 23.149—NB:
attribute); . . . ipsis principibus prosperum vel exitio fuit. (Tac. Hist. 2.1.1—NB:
coordination); . . . non quae esui et potui forent . . . (Gel. 4.1.20); . . . proprietas linguae
Latinae iudiciis otiosorum maximo spretui est . . . (Sidon. Apol. Ep. 3.14.2—NB:
attribute);
With other verbs: . . . scilicet haec aliis praedae lucroque iacebant . . . (Lucr. 5.875);
Lepus omnium praedae nascens solus praeter dasypodem superfetat . . . (Plin. Nat.
8.219—NB: genitive attribute); (sc. Meherdatem) . . . auribus decisis vivere iubet os-
tentui clementiae suae et in nos dehonestamento. (Tac. Ann. 12.14.3) . . . ne defor-
mitas curae fiat. (Mulom. Chir. 650);
When the predicative dative is used with the verb sum, the presence of another
constituent in the dative is very common, as in (a), (b), (d), and (e), above, and
even when absent, it is often understood. The function of this dative constituent
in its clause is disputed. It is sometimes regarded as a beneficiary adjunct (that is:
an optional constituent). See the parallel use in (g) with the dative turpitudini and
(h) with the adjective turpe, below. However, some of the nouns functioning as
predicative dative are bivalent, so the other dative might be explained as being
governed by the noun. This applies to auxilio in (a), above. Auxilio esse can be
regarded as a support verb expression and be compared to auxiliari, which
governs the dative as well. In (i), usui est is paralleled by prodest, which also
governs a dative.46
45
Recent mss. have causam. The reading is defended by Freundlich (1967). Cf. . . . id quod maximo
malo illi causae fuit . . . (Cic. Clu. 54) and TLL s.v. caussa 695.11ff., to which add Caes. Civ. 3.72.2; Liv.
38.52.3, 38.52.10.
46
For this line of reasoning, see Baños (1995).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
47
For the diachronic and individual variation, see Sz.: 99–100.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Appendix: Adopted from Greek is the use of sum with a dative constituent and a
participle or adjective indicating volition, also in the dative, as in (n)–(p).48 Sallust
has the first attested instances. Note that the participle or adjective cannot be omitted
without making the resulting utterance ungrammatical. In that respect it looks like a
subject complement.
(n) . . . neque plebi militia volenti (sc. esse) putabatur . . .
(‘ . . . it was thought that the commons were disinclined to military service .
. . ’ Sal. Jug. 84.3)
(o) Grande periculum Lilybaeo maritumisque civitatibus esse et quibusdam
volentibus novas res fore.
(‘[he said that] Lilybaeum and the cities of the coast were in great danger,
and revolution would be to the liking of some of them.’ Liv. 21.50.10)
(p) Fama dediti benigneque excepti Segestis vulgata, ut quibusque bellum
invitis aut cupientibus erat, spe vel dolore accipitur.
(‘The report of Segestes’ surrender and his gracious reception, once it
became generally known, was heard with hope or sorrow depending on
whether war was against the will or to the liking of each.’ Tac. Ann. 1.59.1)
Supplement:
. . . non tam diffidentia futurum quae imperavisset, quam uti militibus exaequatus
cum imperatore labor volentibus esset. (Sal. Jug. 100.4); Maturo ceterisque remanere
et in verba Vespasiani adigi volentibus fuit. (Tac. Hist. 3.43.2); Nunc quoque, si tibi
fabulam brevem libenti est audire, audi. (Fro. De feriis Alsiensibus 3.8 (vdH 231.15));
48
See K.-G.: I.425–6.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
49
See Bagordo (2001: 77–80), commenting on Sz.: 171 and others.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(c) Iam hisce ambo, et servos et era, frustra sunt duo, / qui me Amphitruonem
rentur esse.
(‘Now both of them, slave and mistress, are fooled; they think I’m Amphitruo.’ Pl.
Am. 974–5)
(d) Nam si vir scortum duxit clam uxorem suam, / id si rescivit uxor, impune est
viro.
(‘If a man has brought home a prostitute behind his wife’s back and the wife finds out
about it, the husband goes unpunished.’ Pl. Mer. 819–20)
(e) . . . ni tua pax propitia foret praesto.
(‘ . . . had not your favouring kindness been nigh at hand.’ Pl. Trin. 837)
Supplement:
Necesse est, quo tu me modo voles esse, ita esse, mater. (Pl. Cist. 46); Ita est enim
interpretatio illa pontificum ut eidem potestatem habeant iudicum. (Cic. Har. 13);
Quae cum ita sint, quaero abs te quam ob causam recusaris. (Cic. S. Rosc. 120);
Ne dixis istuc. # Ne sic fueris. (Pl. As. 839–40); Sic vita erat. (Ter. An. 62); Sic est
volgus. (Cic. Q. Rosc. 29);
Nunc rem ut est eloquamur. (Pl. As. 731); Graeci autem homines, non satis
animosi, prudentes, ut est captus hominum, satis, hostem aspicere non possunt
(Cic. Tusc. 2.65); . . . milites disponit, non certis spatiis intermissis, ut erat superiorum
dierum consuetudo . . . (Caes. Civ. 1.21.3);
Vae misero mihi, / mea nunc facinora aperiuntur, clam quae speravi fore.
(Pl. Truc. 794–5); Haec commemoro quae sunt palam. (Cic. Pis. 11—NB: facta
add. Halm); . . . perpetuo tamen id fore clam diffidere debet. (Lucr. 5.1157); . . .
quorum nihil posse fieri tam palam est ut huius rei causa nullius philosophi schola
intranda sit. (Sen. Nat. 4b.7.3);
Postquam accepere ea homines, quibus mala abunde omnia erant . . . (Sal. Cat. 21.1);
Nec nil hodie nec multo plus tu hic edes, ne frustra sis. (Pl. Capt. 584); . . . quippe
quoius neque consilium neque inceptum ullum frustra erat. (Sal. Jug. 7.6); Si quis his
personis quae agere non potuerunt fideiussorem iudicio sistendi causa dederit,
frust<r>a erit datio. (Ulp. dig. 2.8.2.2);
Atque nos omnia plura habere volumus et id nobis impoene est. (Cato orat 167);
Novi est in lege hoc ut qui nummos in tribu pronuntiarit, si non dederit, impune
sit . . . (Cic. Att. 1.16.13);
Ea vero quem ad modum inlustrentur, praesto est qui omnis docere possit . . . (Cic.
de Orat. 2.121); Portus (sc. laborum) enim praesto est . . . (Cic. Tusc. 5.117);
Neque enim multo secus est parens liberis. (Cic. Planc. 29);
Adverbs that are (derivationally) related to adjectives, such as bene ‘well’ and iucunde
‘pleasantly’, are less commonly used as subject complements and several instances
that are cited as subject complement should not be analysed as such. In the examples
(f )–(h), for example, ‘behaviour’ is meant, so the adverbs look like manner adjuncts.50
50
See von Nägelsbach and von Müller (1905: 618).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
Cum in convivio comiter et iucunde fuisses, tum illuc isti, ut dixeras. (Cic. Deiot. 19);
Ibi cum ventum exspectarem (erat enim villa Valeri nostri, ut familiariter essem et
libenter) . . . (Cic. Att. 16.7.1); . . . Octavius langu<i>de et incuriose fuit . . . (Sal. Hist.
2.42M=1.7–8K); . . . uti prospectus nisusque per saxa facilius foret. (Sal. Jug.
94.1); . . . se sub imperio populi Romani fideliter atque oboedienter futuros.
(Liv. 8.19.2); UBI CABALLI BELLE SUNT (Tab. Vindol. III.632, second II AD—NB: if the
final e is correct);
The adverb in expressions like bene est is not a subject complement. For such
impersonal expressions, see § 4.15.
Some adverbs that are related to an adjective or an adjectival form such as alius and
may be taken as subject complements are aliter ‘otherwise’, longe ‘far and wide’, recte
‘OK’ and tuto ‘safely’, as in exx. (i)–(l). Aliter in (i) and recte in (j) are clear subject
complements; no ‘adjectival’ alternatives exist (? alius,? rectae). In (j), res erunt is
more or less equivalent to erit (see § 4.15 at the end). The other two examples are
more complicated: in (k), erit need not be the copula. In (l), tuto may be interpreted
as ‘in safety’.
(i) Tui benevolentis, si ita es ut ego te volo, / sin aliter es, inimici atque irati tibi.
(‘[That voice] is a well-wisher’s, if you’re the sort I wish you to be; but if
you’re otherwise, an enemy’s, and an angry one.’ Pl. Trin. 46–7)
(j) Uni nimirum recte tibi semper erunt res / . . .
(‘For you alone, of course, things will always go well . . . ’ Hor. S. 2.2.106)
(k) Longe illi dea mater erit . . .
(‘His goddess-mother will not be at his side . . . ’ Verg. A. 12.52)
(l) Ut tuto sim quod laboras, id mihi nunc facillimum est . . .
(‘As to your anxiety for my personal safety, that is now the easiest thing in
the world for me . . . ’ Cic. Fam. 14.3.3)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
51
For more examples, see TLL s.v. cum 1354.71; s.v. in 788.46ff.; s.v. pro 1430.41ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
. . . tum vero sine metu, sine cura omnes erant qui Sopatrum defendebant. (Cic.
Ver. 2.70); Aut ubi quidam morbi, qualis comitialis, qualis insania est, sine febre,
quamvis diu, manent. (Cels. 2.15.4); . . . sanguinis emissione exhaustus sine viribus fit
(sc. equus). (Pelag. 34.3);
The six classes of verbs that require an object complement as their third argument are
treated in detail in §§ 4.87–4.89. Table 9.7 contains a survey of the categories that may
function as object complement. They are essentially the same as for subject comple-
ments, although the entire range demonstrated for subject complements is not
actually attested.52 The objects to which the complements are related are usually in
the accusative (or nominative in the passive), but there are also ablative objects (see
§ 4.89). The examples within each section of the table are arranged according to the
six classes of verbs involved; not all categories are found with all the classes. Further
examples of these classes are given in § 4.87.
52
For a good survey of the options with facio, see TLL s.v. 112.63ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Adverb Si forte pure velle habere dixerit, / tot noctes reddat spurcas 9.45
quot pure (puras mss.) habuerit.
(‘If by chance she says she wants to spend some nights
chastely, she shall give you as many unchaste nights as she
had chaste ones.’ Pl. As. 806–7)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Infinitive clause . . . mirumque quid hoc est, / aut aliquid certe simile huic, 9.47
quod amare vocatur.
(‘ . . . and some miracle is this thing which is called love, or at
least something like this.’ Ov. Met. 7.12–13)
. . . te pugna Cannensis accusatorem sat bonum fecit. (Cic. S. Rosc. 89); . . . socer . . .
non orandus erat mihi, sed faciendus Erechtheus. (Ov. Met. 6.701);
Laudo malum quom amici tuom ducis malum. (Pl. Capt. 151); Mos habebatur
principum liberos cum ceteris idem aetatis nobilibus sedentes vesci in adspectu
propinquorum propria et parciore mensa. (Tac. Ann. 13.16.1); . . . eos et habemus et
dicimus . . . fratres . . . (Lact. Inst. 5.15.3); Talem . . . solemus et sentire bonum civem
et dicere. (Cic. Off. 1.124);
Me videlicet latronem ac sicarium abiecti homines et perditi describebant. (Cic.
Mil. 47); Caelum dicunt Graeci Olympum . . . (Var. L. 7.20); . . . eadem epistula illum
et dixisti amicum et negasti. (Sen. Ep. 3.1); (sc. Thessalus) . . . cum monumento suo,
quod est Appia via, iatronicen se inscripserit. (Plin. Nat. 29.9); . . . Capsa, quoius
conditor Hercules Libys memorabatur. (Sal. Jug. 89.4); Demetrium qui Phalereus
vocitatus est . . . (Cic. Rab. Post. 23); Quis tu homo es? / Libertus illius, quem omnes
Summanum vocant. (Pl. Cur. 412–13);
Quam diu tu voles, vitiosus consul Dolabella. Rursus, cum voles, salvis auspiciis
creatus. (Cic. Phil. 2.84); . . . consules declarantur M. Tullius et C. Antonius. (Sal. Cat.
24.1); Ipse ob id meritum ingenti patrum studio creatus consul collegam Ap.
Claudium Crassum dixit. (Liv. 7.24.11); Roma patrem patriae Ciceronem dixit
(Juv. 8.244); . . . propter eloquentiam legatus a suis civibus electus est . . . (Cic. Ver.
2.156); . . . non ut ille filius instituatur, sed ut e patriciis exeat <et> tribunus plebis fieri
possit, idcirco adoptari, neque id obscure. (Cic. Dom. 37); . . . persuasi, ut scribam
iurati legerent eum quem . . . (Cic. Clu. 126); Caput imperii sui Corfinium legerunt . . .
(Vell. 2.16.4); . . . me augurem a toto conlegio expetitum . . . nominaverunt . . . (Cic. Phil.
2.4); Eo modo sacerdos Climachias renuntiatus est. (Cic. Ver. 2.130);
. . . legum . . . suasorem se haud dubium ferebat. (Liv. 6.36.7); Se ad eam rem
profitetur adiutorem. (Caes. Gal. 5.38.4); . . . qui se architectum profiteatur. (Vitr.
1.1.3); Nec tacui demens et me, fors siqua tulisset, / . . . / promisi ultorem et verbis
odia aspera movi. (Verg. A. 2.94–6);
Ecquod in Sicilia bellum gessimus quin Centuripinis sociis, Syracusanis hostibus
uteremur? (Cic. Ver. 5.84);
53
For Varro’s use of facio + adjectival object complement, see Bodelot (2014).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
doctior dici potest. (Pl. Mos. 1072); . . . ut meae quoque essent orationes quae con-
sulares nominarentur. (Cic. Att. 2.1.3); . . . (sc. caules Sabellici) dulcissimi perhiben-
tur ex omnibus. (Plin. Nat. 19.141); Se miserum praedicat. (Pl. Men. 909);
Te quoque magno opere, M. Fanni, quaeso ut, qualem te iam antea populo
Romano praebuisti, cum huic eidem quaestioni iudex praeesses, talem te et nobis
et rei publicae hoc tempore impertias. (Cic. S. Rosc. 11); . . . se fortibus factis dignum
tant<a>e maiestatis infulis declaravit . . . (Hist. Aug. Gall. 10.1); . . . si . . . metuendos
vos praebituri estis . . . (Liv. 3.53.10);
Hic vide quam me sis usurus aequo. (Cic. Ver. 5.154);
54
For habeo, see TLL s.v. 2430.12ff.; 2447.62ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
55
For prepositional phrases used as object complement with habeo, see TLL s.v. 2427.42ff.; 2430. 30ff.;
2445.41ff.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
CHAPTER 10
Satellites
Satellites have been defined in a negative way as optional constituents that do not
belong to the nucleus of a clause, which in turn has been defined as the unit consisting
of the verb and its (obligatory) arguments (see § 2.10). Although from the structural
point of view satellites are optional in their clauses, this does not imply that they
contain unimportant information. On the contrary, they are often essential elements
in describing the world their clauses refer to.
From the formal point of view, a wide range of constituents may function as
satellites: nouns and noun phrases in a specific case, prepositional phrases, adverbs,
and a variety of subordinate clauses (both finite and non-finite). The semantic
functions they fulfil are equally diverse. They range from objective qualifications,
such as the place where an event happened, to subjective comments on the way an
action was performed. Some of these semantic functions can also be found within the
nucleus. Direction expressions, for example, are obligatory with a verb like me confero
‘I betake myself ’ (when used with such verbs they are called ‘direction arguments’,
see § 4.43), but optional with a verb like ambulo ‘to walk’.
Satellites also vary in their hierarchical position. Some of them are modifiers of the
state of affairs referred to by the verb and its arguments or of the verb itself (manner
expressions, for example). These are called ADJUNCTS in this Syntax (see § 2.10). Other
constituents are some form of comment on the content, the truth, or the appropri-
ateness of the clause as a whole (for example, recte ‘rightly’). On the one hand, they
reflect the way the language producer regards or wants to present the message he is
producing; on the other, they also play a role in engaging the addressee of that
message. These constituents are called DISJUNCTS (see § 2.10).
Adjuncts and disjuncts have different distributional properties. Adjuncts may occur:
(i) in each of the four sentence types discussed in Chapter 6;
(ii) as answers to questions introduced by a specialized question word (such as:
When did John arrive? # Yesterday.);
(iii) in all types of subordinate clauses;
(iv) in coordination patterns (such as: He stayed yesterday evening and this morning.);1
(v) under the scope of negation (such as: He came this morning, not yesterday
evening.).
1
See Pinkster (1972).
Satellites
Disjuncts are more restricted in their occurrence. In most of the contexts mentioned
for adjuncts they are excluded. This is obvious for words like probably in English or
fortasse ‘perhaps’ in Latin. Such words are not used in imperative clauses. They occur
in answers to a sentence question (such as: Will John pass his exam? # Probably.).
They are outside the scope of negation (such as: John will probably not pass his
exam.). In Latin, disjuncts are often the first constituent in their clauses.
There is another class of constituents which specify the relation of a sentence to its
wider context, either to a preceding or following segment of the text, or to knowledge
of the world in general, to which the language producer may refer in an indirect way.
An example is tamen ‘nevertheless’. In our grammars such constituents normally go
under the label ‘conjunction’ (together with what are called coordinators and sub-
ordinators in this Syntax; see § 3.25). From a sentential perspective they are clearly
optional and non-nuclear constituents (hence satellite-like). However, since their
main function is on the discourse level, they are more properly dealt with in the
chapter on discourse coherence and discourse structure (Chapter 24).
Lexical items may be used in more than one of the three ways just described, and
the use of specific lexical items may change in the course of time.
10.1 Adjuncts
Adjuncts have been defined in § 2.10 and in the introduction to this chapter as
optional constituents that modify in some way the state of affairs referred to by the
verb and its arguments. Formally, adjuncts show great variety. They include nominal
constituents, most frequently in the ablative case form (nouns, noun phrases, etc., also
ablative absolute clauses), prepositional phrases, adverbs, both underived words
such as nunc ‘here’ and productively derived words such as suaviter ‘agreeably’, and
also a range of finite and non-finite subordinate clauses. Adjunct constituents are
therefore typically an open class.
In § 2.9 restrictions are mentioned on the occurrence of certain adjuncts with
certain types of states of affairs. Thus (i) manner and (ii) instrument adjuncts
typically occur with controlled states of affairs. (iii) Beneficiary and (iv) purpose
adjuncts are restricted in their occurrence in the same way. Adjuncts denoting (v) the
time within which an event takes place occur only with terminative states of affairs,
whereas the same states of affairs exclude (vi) extent of time (duration) adjuncts.
Among adjuncts these six can be said to be more closely related to the verb and its
argument(s) than others, such as position in time adjuncts. The degree of semantic
cohesion with the nucleus might influence the position such constituents normally
have in the sentence. However, there is still much research to do on this issue. This
distinction in at least two types of adjuncts (sensitive or non-sensitive to the type of
state of affairs) is left out of account in the remainder of this section. The following
classification is made on semantic grounds.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Adjuncts
By using these criteria one can distinguish five types of adjuncts, each with a few
subtypes. Decisive proof for each of the distinctions cannot always be obtained, due to
the limitations of the extant corpus. Furthermore, for each individual expression that
one finds in the corpus, it is difficult to decide what class it belongs to. In light of these
difficulties, the distinctions made and the assignment of individual instances to the
semantic functions that are distinguished is not watertight. Prepositional expressions
are so manifold that some of them are difficult to assign. Further details are given in
the individual sections.2
First, a distinction is made between the types of adjunct listed below. Subtypes are
then presented in the sections that follow.
1. Space adjuncts
2. Time adjuncts
3. Process adjuncts
4. Contingency adjuncts
5. Respect adjuncts
2
For a slightly different classification of satellites, see Luraghi (2010), especially pp. 20–1.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
3 4
See Svennung (1935: 382–90) and Önnerfors (1955). See Petersmann ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
As (a) and (b) show, a clause may contain more than one position in space adjunct if
one can be seen as a specification of the other.5 The order of the position in space
adjuncts is not restricted. Further examples are (d)–(f).
(d) Nam Palaestrio / domi nunc apud me est, Sceledrus nunc autem est foris.
(‘For Palaestrio is at home at my place now, while Sceledrus is outside.’ Pl. Mil. 592–3)
(e) Habitasti apud Heium Messanae . . .
(‘You have stayed at Messana in Heius’s house . . . ’ Cic. Ver. 4.18)
(f) In Hispania ulteriore in Lusitania [ulteriore] sus cum esset occisus . . .
(‘When a sow was killed in Lusitania, in Farther Spain . . . ’ Var. R. 2.4.11)
Supplement:
Qui Syracusis perhibere natus esse in Sicilia . . . (Pl. Men. 408); Ubi tu nata es? Ut
mihi / mater dixit, in culina, in angulo ad laevam manum. (Pl. Per. 630–1); Cancer
ater, is olet et saniem spurcam mittit, albus purulentus est, sed fistulosus et subtus
suppurat sub carne. (Cato Agr. 157.3); . . . si polypus in naso intro erit, brassicam . . . in
malam conicito . . . (Cato Agr. 157.15); Boves perferi etiam nunc sunt multi in
Dardanica et M<a>edica et Thracia, asini feri in Phrygia et <Ly>caonia, equi feri
in Hispania[e] citeriore regionibus aliquot. (Var. R. 2.1.5); Deinde ut in curia
Syracusis . . . statua ex aere facta est . . . (Cic. Ver. 2.50); Res ad eum defertur esse
civem Romanum qui se Syracusis in lautumiis fuisse quereretur. (Cic. Ver. 5.160);
Namque et in Lysandri, qui Lacedaemoniorum clarissimus fuerat, statua, quae
Delphis stabat, in capite corona subito exstitit ex asperis herbis et agrestibus . . . (Cic.
Div. 1.75); Pernoctant venatores in nive in montibus. (Cic. Tusc. 2.40); Sed in ea
Pompei epistula erat in extremo ipsius manu ‘Tu censeo Luceriam venias. Nusquam
eris tutius’. (Cic. Att. 8.1.1); Agro autem Falisco via Campana in campo Corneto est
lucus in quo fons oritur, ibique avium et lacertarum reliquarumque serpentium ossa
iacentia apparent. (Vitr. 8.3.17); Constatque in Berenice, urbe Trogodytarum, et inde
stadiis MMMMDCCCXX in eadem gente Ptolemaide oppido . . . hoc idem ante
solstitium quadragenis quinis diebus totidemque postea fieri . . . (Plin. Nat. 2.183);
Direction expressions and source expressions and combinations of them can be used
in clauses where no motion is involved. In (g) and (h) the two adjuncts together
function as a position in space adjunct, answering the question ‘ubi?’ For a similar use
of a direction expression, see (i); for a source expression, (j). The source expression
may also be phrased in terms of the distance between two locations, as in (k).6
(g) Ac meus quidem (sc. reditus) is fuit ut a Brundisio usque Romam agmen
perpetuum totius Italiae viderit.
(‘My (return) was such that an uninterrupted crowd from every part of Italy watched
it all the way from Brundisium to Rome.’ Cic. Pis. 51)
5
The best survey of instances of juxtaposition of more than one position in space adjunct can be found
in Müller (1895). See also Otto (1912: 15–17, 27–35). For Vitruvius, see Wistrand (1933: 75–6).
6
See Torrego and de la Villa (2009).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(h) Latium antiquum a Tiberi Cerceios servatum est milia passuum L longitudine.
(‘Old Latium has preserved the original limits, extending from the Tiber to Cerceii, a
distance of 50 miles.’ Plin. Nat. 3.56)
(i) Nam et illi rumores de comitiis Transpadanorum Cumarum tenus caluerunt.
(‘Those rumours about elections in Transpadane Gaul were rife only as far as
Cumae.’ Cael. Fam. 8.1.2)
(j) Erat a septentrionibus collis . . .
(‘On the north side there was a hill . . . ’ Caes. Gal. 7.83.2)
(k) . . . ad castra Caesaris omnibus copiis contenderunt et a milibus passuum
minus duobus castra posuerunt.
(‘ . . . with all their forces they sped on to the camp of Caesar and pitched their own
less than two miles from it.’ Caes. Gal. 2.7.3)
Supplement:
. . . posteaquam a L. Scipione devictus Tauro tenus regnare iussus est . . . (Cic. Deiot.
36); Habitant ab Isthmo ad Scyllaeon Epidaurii . . . (Mela 2.49);
Sed speculabor, nequis aut hinc aut ab laeva aut a dextera / nostro consilio venator
assit . . . (Pl. Mil. 607–8); Eorum una pars . . . attingit etiam ab Sequanis et Helvetiis
flumen Rhenum, vergit ad septentriones. (Caes. Gal. 1.1.5); . . . Claudius dextro in
cornu, Livius ab sinistro pugnam instruit. (Liv. 27.48.4—NB: parallel expressions);
A septentrione Indus adluit, a meridie Acesines Hydaspi confunditur. (Curt. 9.4.8);
At si in medio flumine insula nata sit, haec eorum omnium communis est, qui ab
utraque parte fluminis prope ripam praedia possident. (Gaius Inst. 2.72);7
. . . conlocatis insidiis bipertito in silvis opportuno atque occulto loco a milibus
passuum circiter duobus Romanorum adventum ex<s>pectabant . . . (Caes. Gal.
5.32.1); Cassius noster . . . a milibus passuum XX castra habet posita Plfiø et
existimat se sine proelio posse vincere. (Cass. Fam. 12.13.4); Hannibal tumulum a
quattuor milibus inde . . . cepit. (Liv. 30.29.10); Postquam inceptum non succedebat,
castra propius hostem movit rex et a quinque milibus passuum communiit. (Liv.
42.58.1);
Appendix: Expressions indicating extent of space (for which see § 10.18) can be used
in combination with position in space adjuncts that consist of prepositional phrases
indicating relative position. Examples are (l)–(p). In (l), for example, citra Veliam
milia passuum III answers the question ubi?
(l) Erat (Brutus) enim cum suis navibus apud Haletem fluvium citra Veliam
milia passuum III.
(‘For he was with his ships at the mouth of the river Hales three miles north
of Velia.’ Cic. Att. 16.7.5)
(m) Ibi oculos arundinis pedes ternos alium ab alio serito.
(‘Plant the eyes of the reed there three feet apart.’ Cato Agr. 6.3)
7
More instances in OLD s.v. ab § 23; TLL s.v. ab 21.5ff.; 24.14ff.; Müller (1908: 100–1).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
10.4 The use of the bare locative and ablative in position in space adjuncts
The use of bare ablatives with common nouns is in Early and Classical Latin limited to a
couple of nouns, including mare and terra—usually coordinated—and locus, pars, and
regio, when they are modified; the latter two are not used in this way in Early Latin nor in
Cicero. Examples are (a)–(f). A few prepositional alternatives are given in the Supplement.
Topographical names in the locative and ablative cases are discussed below in § 10.17.
A few common nouns referring to a location are used in the same way as these topo-
graphical names, notably domus ‘home’ and rus ‘countryside’ (on which see also § 10.17).
(a) Sed pater illarum Poenus . . . / mari te<rraque> usquequaque quaeritat.
(‘But their Carthaginian father . . . has been looking for them everywhere by sea and
by land.’ Pl. Poen. 104–5)
(b) Natura sic ab iis investigata est ut nulla pars caelo, mari, terra, ut poetice
loquar, praetermissa sit.
(‘Natural philosophy has been investigated by them so thoroughly that no region in
sky or sea or land (to speak poetically) has been passed over.’ Cic. Fin. 5.9)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(c) . . . nec potest fieri tempore uno / homo idem duobus locis ut simul sit?
(‘ . . . (a thing which) is impossible, namely that one and the same man can be in two
places simultaneously at the same time?’ Pl. Am. 567–8)
(d) Peiore res loco non potis est esse quam in quo nunc sita’st.
(‘Matters couldn’t be in a worse place than the one in which they are now.’ Ter. Ad. 344)
(e) . . . conpluresque milites in viis urbis omnibus partibus interficiebantur.
(‘ . . . and many soldiers were killed in all parts in the streets of the town.’ Caes. Civ.
3.106.5)
(f) . . . nuntiosque ad eum celeriter mittit quibus regionibus exercitum
exposuisset . . .
(‘ . . . and he hastily sends him messages stating in what districts he had disembarked
his army . . . ’ Caes. Civ. 3.29.3)
Supplement:
. . . ut earum rationem rerum explicaret quae in mari caelove fierent. (Cic. Div. 1.13);
Faciunt hoc homines . . . ut multis in locis notas ac vestigia suorum flagitiorum relinqui
velint. (Cic. Ver. 2.115); Facile ex altera parte CCCC fuerunt. (Cic. Att. 1.14.5);
The bare ablative is regular when a common noun is modified by totus, as in tota urbe
‘throughout the whole city’, ‘all over the city’ in (g). The prepositional expression in
tota urbe has a different meaning ‘in every part of the city’ (the ‘throughout’
prepositional equivalent would be per totam urbem). The bare ablative is also often
found with omnis ‘all’, much less often cunctus ‘all’, universus ‘entire’, and exception-
ally plerusque ‘most’ (the use of a prepositional expression with in is more common),
as in (h), and also with a few spatial adjectives (imus ‘lowest’, medius ‘middle’, as in
(i), summus ‘highest’), where a prepositional expression with in is also more common.
In Early and Classical Latin the use of the bare ablative is rare outside these lexically
restricted classes of common nouns and modifiers. It becomes more common from
Livy onwards and poets use it freely.8 The nouns are usually modified. Classical
instances are sometimes emended.9 Examples are (j)–(n).
(j) Iamque mari magno classis cita / texitur.
(‘And now upon the mighty sea a fast fleet is built.’ Enn. scen. 65–6V=43–4J)
(k) Quod ut meliore tempore possimus facit Adriano mari Dolabella, Fretensi
Curio.
(‘I can do that at a better time of the year, thanks to Dolabella on the Adriatic and
Curio on the Straits.’ Cic. Att. 10.7.1)
(l) . . . aperto ac plano litore naves constituit.
(‘ . . . he grounded his ships on the even and open shore.’ Caes. Gal. 4.23.6)
(m) Pecunia a patre exacta crudeliter, ut divenditis omnibus bonis aliquamdiu
trans Tiberim veluti relegatus devio quodam tugurio viveret.
(‘The money was exacted from his father without pity, so that he was obliged to sell
all that he had and live for some time on the other side of the Tiber, like one
banished, in a certain lonely hovel.’ Liv. 3.13.10)
(n) Cernis custodia qualis / vestibulo sedeat?
(‘Do you see what sentry sits in the doorway?’ Verg. A. 6.574–5)
Supplement:
Est enim earum rerum omnium <in> (add. Wopkens) nostra urbe summa in publico
copia. (Cic. Tusc. 5.102); Quod facere non debes nec dubitare quin aut aliqua re
publica sis is futurus qui esse debes aut perdita non adflictiore condicione quam
ceteri. (Cic. Fam. 6.1.6—or is it an ablative absolute?); Nostri ex humili convalle bene
longe sunt egressi et planitie in aequiore loco constiterunt. (B. Hisp. 25.2); . . .
plerumque milites stativis castris habebat, nisi quom odor aut pabuli egestas
locum mutare subegerat. (Sal. Jug. 44.4); . . . praediscere . . . possumus . . . quando
(conveniat) armatas deducere classis / aut tempestivam silvis evertere pinum.
(Verg. G. 1.253-7); . . . quod orta ex lacu nebula campo quam montibus densior
sederat . . . (Liv. 22.4.6); Silvisque agrisque viisque / corpora foeda iacent . . . (Ov.
Met. 7.547–8); Aeolon ille refert Tusco regnare profundo . . . (Ov. Met. 14.223);
ET·PATER·HOC·TITULO·DEBUIT·ANTE·LEGI (CIL IX.3071.12 (Paeligni)); Sed haec domus
non orbis terrae uno angulo aedificatur sed per omnem terram. (August. Ep. 142.2);
8
For a survey of this use of the ablative (‘of extension’) in Virgil, see Malosti (1967).
9
See K.-St.: I.353–4.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
There is a large number of prepositions that indicate the relative position of two entities,
with a broad spectrum of meanings. A few prepositions govern the ablative: coram ‘in the
presence of ’, cum ‘with’, in in its meaning ‘in’, ‘at’, ‘on’ (when governing the accusative its
meaning is different), palam ‘openly in the presence of ’, prae ‘before’, pro ‘before’, sub
‘under’ (also with the accusative in another meaning), super ‘over’, ‘on top of ’ (it governs
the accusative in its meaning ‘above’). Most prepositions used in position in space
adjuncts, however, govern the accusative (see § 12.25). Examples are (a)–(d).
(a) Satis iam audivi tuas aerumnas ad portum mihi quas memorasti.
(‘I’ve already heard enough about your hard times, which you told me about at the
harbour.’ Pl. Capt. 929)
(b) . . . cum suo sibi gnato unam ad amicam de die / potare . . .
(‘ . . . he’s drinking at the same mistress’s place with his own son in broad daylight . . . ’
Pl. As. 825–6)
(c) Audire vocem visa sum ante aedis modo / mei Lampadisci servi.
(‘Just now I seemed to hear the voice of my good slave Lampadio in front of the
house.’ Pl. Cist. 543–4)
(d) Quid apud hasce aedis negoti est tibi?
(‘What business have you at this house?’ Pl. Am. 350)
Other prepositions that are used to mark position in space adjuncts are given in
Table 10.1.
Table 10.1 Prepositions (for those beginning with an a see the examples above) that
mark position in space adjuncts
circa and circum ‘round’, ‘around’ intra ‘within’ prope ‘near’
circiter ‘around’ (almost intro ‘within’ propter ‘near’
non-existent in this meaning) (very rare in this meaning)
cis and citra ‘on this side of ’ intus ‘within’ (very rare)10 secundum ‘after’, ‘behind’
contra ‘opposite’ iuxta ‘close to’ subter ‘below’ (rarely
governing an ablative)
extra ‘outside’ ob ‘before’ supra ‘above’
infra ‘below’ pone ‘behind’ trans ‘across’
inter ‘between’, ‘among’ post ‘behind’ ultra ‘beyond’
10
See Sz.: 278 and TLL s.v. 107.55ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The examples given above all indicate physical location. Examples of ‘abstract’
location are (e)–(i). The regular preposition is in ‘limitativum’.11 Examples (e) and (g)
are often dealt with as equivalent to the ablatival argument expressing the ‘nature of
supremacy’ (OLD s.v. praesto), but this is incorrect, as is shown by (g), where the two
expressions co-occur. Ex. (h), though, shows that the two expressions come very
close. For a similar use with a human being, ‘the case of an heir’, see (i).
(e) Plebes in hoc regi antistat.
(‘The commoners stand better than their king in this.’ Enn. scen. 338J (228V reads
antistat loco))
(f) . . . quod etiam in salute nostra nonnumquam facit cum aegrotamus
medicus.
(‘ . . . which even in matters of our health the physician sometimes does when we are
ill.’ Var. L. 5.8)
(g) . . . Socraten opinor in hac ironia dissimulantiaque longe lepore et humani-
tate omnibus praestitisse.
(‘ . . . I am of the opinion that Socrates far surpassed all others through his refined wit
in this strain of irony or assumed simplicity.’ Cic. de Orat. 2.270)
(h) . . . ut quo uno homines maxime bestiis praestent in hoc hominibus ipsis
antecellat?
(‘ . . . that by so doing he may surpass men themselves in that particular respect
wherein chiefly men are superior to animals?’ Cic. de Orat. 1.33)
(i) Quod in herede dicimus, idem erit et in ceteris successoribus.
(‘What applies to an heir applies to other successors as well.’ Ulp. dig. 14.4.9.pr.)
11
The term is used in the TLL s.v. in 783.59ff. Examples in OLD s.v. in §§ 41 and 42.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
12
‘Eo’ tamen ‘intus’ et ‘intro sum’ soloecismi sunt. (Quint. Inst. 1.5.50).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Trebatius quidem scribit se ab illo VIIII Kal. Febr. rogatum esse ut scriberet
ad me ut essem ad urbem.
(‘Trebatius writes that on 22 January Caesar asked him to write urging me that
I should be on my way to Rome.’ Cic. Att. 7.17.3)13
(e) . . . quaeque progredientibus ad septentrionem sunt nationes . . .
(‘ . . . and which nations there are as we move northwards . . . ’ Vitr. 6.1.7)
(f) HINCE·SUNT / NOUCERIAM·MEILIA·LI.
(‘From here to Nouceria it is 51 miles.’ CIL I2.638.3–4 (Lucania, 132 BC))
(g) . . . praeterea (dentes) in os vergentes, ne excidant cibi nullum habentibus
retinendi adminiculum.
(‘ . . . and also turned further into the mouth, so as to prevent morsels of food from
falling out, as they have no apparatus for retaining it.’ Plin. Nat. 11.162)
Supplement:
With the verb sum: . . . cum ad me bene mane Dionysius fuit (venit cj. Müller). (Cic.
Att. 10.16.1); Nec sane lavare potui. Fui enim hodie in funus. (Petr. 42.2 (Seleucus
speaking)); IN·CURIAM·FUERUNT·PONTIUS·CELSUS·DICTAT(OR) ·SUETONIUS·CLAUDIANUS·AED
<ILIS> . . . (CIL XI.3614.13 (Caere, AD 113)); Adesse in senatum (senatu D) iussit a. d.
XIII Kalendas Octobris. (Cic. Phil. 5.19); Ubi cum pervenissem fui ad episcopum .
. . (Pereg. 23.1);
Milestones: M(ILLIARIUS) T.·QUINCTIUS·T.F. / FLAMININUS / COS / PISAS (CIL I2.657 (near
Florence, 123 BC));
The verbs sum and habeo are used in a number of idiomatic prepositional expres-
sions with the preposition in governing an accusative abstract noun. Examples are
(h)–(j). In some cases it is not difficult to give parallels of the same prepositional
expression with a verb of motion, as in (h): in mentem fuit occurs alongside in
mentem venit and in the same way we find in potestatem habuit / venit, in possessio-
nem est / venit, in amicitiam sunt / veniunt, in gratiam habet / redit. These represent
the abstract counterpart of the examples of implied motion given above. Other
similar expressions are rather isolated, and parallels are not easy to find, but the
explanation must be the same. In some cases cited in this context in does not seem to
have its directional meaning, but its ‘final’ meaning (e.g. esse in vadimonium). Editors
often emend such cases.14
(h) Numero mihi in mentem fuit / dis advenientem gratias pro meritis agere
atque alloqui.
(‘Just in time it occurred to me to address the gods and to give them due
thanks on my arrival.’ Pl. Am. 180–1)
13
This is a much discussed passage. Shackleton Bailey translates ‘near Rome’, while rejecting the
manuscript tradition in Cic. Att. 10.16.1 (see the examples in the Supplement). For other instances, see
Siegert (1952).
14
Material in K.-St.: I.594, where also a few dubious instances are given of concrete nouns; for ‘in
finale’, see TLL s.v. in 763.35ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
More than one goal adjunct in a clause can be admitted if one can be understood as a
specification of the other. The relative order of the adjuncts is not restricted. Examples
are (k) and (l). The same phenomenon can be found in the case of goal arguments.
Some of the instances quoted below may be regarded as arguments.15
(k) Hic postquam in aedis me ad se deduxit domum . . .
(‘After this one took me into his house to him . . . ’ Pl. Mil. 121)
(l) . . . frequentes eos ad me domum adduxit.
(‘ . . . he brought them in large numbers to my house.’ Cic. Clu. 49)
Supplement:
2
EEIS·UTEI·AD·PR(AITOREM)·URBANUM / ROMAM·VENIRENT (CIL I .581.4–5 (SCBac.; Tiriolo,
186 BC)); . . . cur illi servi non ad Caelium domum venerint. (Cic. Cael. 61); Interea
Oppianicus cum . . . huc ad urbem profectus esset . . . cecidisse de equo dicitur . . . (Cic.
Clu. 126); . . . cohortesque ad se in castra traducere . . . (Caes. Civ. 1.21.1); . . . ferrum
intro clam in cubiculum tulit atque ita se traiecit. (B. Afr. 88); . . . nullo repugnante
captas naves Messanam in portum deduxerunt. (Liv. 21.49.3); Ancyram in stativa
oratores Tectosagum ad consulem venerunt . . . (Liv. 38.25.1);
The following categories of constituents may function as direction and goal adjuncts:
bare accusative case forms
prepositional phrases
adverbs
locative and ablative case forms (in post-Classical texts occasionally)
dative case forms (mainly poetic)
subordinate clauses (see § 16.6).
The most frequent way to mark direction and goal adjuncts is by a prepositional
phrase.
10.8 The use of the bare accusative case in direction and goal adjuncts
15
The best survey of instances of juxtaposition of more than one direction or goal adjunct can be
found in Müller (1895). See also Otto (1912: 27–35). For more direction arguments with the verb venio ‘to
come’, see Cabrillana (1997b: 125–31).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
notably domus ‘home’ and rus ‘countryside’ (on which see also § 10.17). This is
extended to other nouns referring to a location, especially in poetry, among other
things to avoid a prepositional expression. In the first prose instance, (a), the Greek
word astu may be said to have a topographical connotation. Lucretius’ ea loca in (b) is
a step further.
(a) Postquam astu (in astu cj. Nipperdey) venit . . .
(‘As soon as he arrived in the city . . . ’ Nep. Alc. 6.4)
(b) . . . ea quod loca cum venere volantes . . .
(‘ . . . because when they have come over those spots in their flight . . . ’ Lucr. 6.742)
(c) Ille autem ‘tua me, genitor, tua tristis imago / saepius occurrens haec limina
tendere adegit.’
(‘But he answered: “Your shade, father, your sad shade, meeting me repeatedly, drove
me to seek these portals.” Verg. A. 6.695–6)
Supplement:
. . . verba refers aures non pervenientia nostras. (Ov. Met. 3.462); Non tamen ut vastos
ausim temptare leones / aut celer agrestis comminus ire sues. (Prop. 2.19.21–2);
. . . tunc aethera tendit (tundit cj. Guyet) extremique fragor convexa irrumpit Olympi
. . . (Luc. 7.475); Ut vero infames scopulos silvamque nefandam / perventum . . . (Stat.
Theb. 3.121-2); . . . nec vos ulterius dubitabitis, si Thessaliam proximam civitatem
perveneritis . . . (Apul. Met. 1.5.2); Sed cum primam plateam vadimus, vento repentino
lumen, quo nitebamur, extinguitur . . . (Apul. Met. 2.32.1); . . . eadem loca profectus
reverti isdem celeri gradu permisit adminicula petituris exercitus validi. (Amm.
27.8.2); Insulam autem quandam decurrentes . . . (Vulg. Act. Apost. 27.16);16
Grammars also mention in this context idiomatic expressions consisting of one of
the verbs eo ‘to go’, do ‘to give’, and a few others and a number of abstract nouns, as
in (d)–(g). In none of these, however, is ‘direction’ involved. With more justification
are mentioned (h) and (i ).
(d) infitias ire ‘to deny’
(e) suppetias advenire, ire, proficisci, etc. ‘(to come) to one’s assistance’
(f) pessum abire, ire, sidere; dare ‘(to sink, to send) to the bottom’
(g) venum ire; dare ‘(to go, to put up) for sale’
(also dat. veno)
(h) malam crucem ire (also in malam crucem) ‘(to go) to be hanged’
(i) malam rem abire, ire (also in malam rem) ‘(to go) to hell’
In (very) Late Latin, limitations on the use of the bare accusative were no longer
observed, as appears from the use of the accusative with abstract nouns indicating a
position in society. An example is (j) from the seventh century.
16
The fullest collection of material is still Landgraf (1898).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Examples of prepositional phrases functioning as direction and goal adjuncts are (a)
and (b), respectively. For an abstract goal expression, see (c). Most of these preposi-
tions govern the accusative when marking a direction or goal adjunct (but see § 12.25
for tenus).
(a) Campester locus is melior qui totus aequabiliter in unam partem verget . . .
(‘A lowland farm is better that everywhere slopes regularly in one direction . . . ’
Var. R. 1.6.6)
(b) . . . nec meum pedem huc intuli etiam in aedis, ut cum exercitu / hinc
profectus sum ad Teloboas hostis eosque ut vicimus.
(‘I haven’t set foot here into this house ever since I and the army went away from here
to our enemy, the Teloboians, and defeated them.’ Pl. Am. 733–4)
(c) Qui in mentem venit tibi istuc facinus facere tam malum?
(‘What made it enter your mind to do such a thing, such an awful thing?’ Pl. Bac. 682)
Table 10.2 Prepositions that mark direction and goal adjuncts (for in see above)
ad ‘to’ extra ‘outside’ prope ‘near’
adversum/s ‘facing’ infra ‘below’ sub ‘to and under’
ante ‘to a position in front of ’ inter ‘among’ tenus ‘as far as’
circiter ‘near’ intra ‘inside’ trans ‘across’
cis and citra ‘to this side of ’ ob ‘against’ ultra ‘on the far side of ’
contra ‘against’ pone ‘to the rear of ’ versus/m ‘towards’
Examples of adverbs used as direction adjuncts are (a) and (b); for a goal adjunct, see (c).
(a) Sosia, age me huc aspice.
(‘Sosia, come on, look here at me.’ Pl. Am. 750)
(b) Quo ted hoc noctis dicam proficisci foras / cum istoc ornatu cumque hac
pompa, Phaedrome?
(‘Where should I say you’re going out to at this time of night, with that dress and
with this accompaniment, Phaedromus?’ Pl. Cur. 1–2)
17
Fredegar follows Greg. T. Hist. 5.30, who has, interestingly: Sophia voluit eum erigere in imperio [=
imperium]. For more instances in Fredegar and discussion, see Odelstierna (1926: 30–1).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 10.3 Adverbs functioning as direction and goal adjuncts (see also above)
eatenus (and ea intus ‘to within’ (usually quo ‘whither’
tenus) ‘so far’ denotes position in space)
eo ‘thither’ istoc and istuc ‘to or towards quorsum/s and quoversum/s ‘in
the place where you are’ what direction?’
extra ‘outwards’ nusquam ‘to no place’ compounds with –orsum/s
foris ‘to the outside’ obviam ‘in the way of ’ utroque(versum/s) ‘in both
directions’
illuc ‘thither’ peregre ‘abroad’
intra ‘inwards’ quatenus ‘how far’
10.11 The use of the dative case in direction and goal adjuncts
The dative is regularly used for the third argument of verbs of transfer to mark the
recipient of the entity transferred, so for instance with the verb do ‘to give’ and mitto
‘to send’ (see § 4.52). ‘Transfer’ implies ‘movement of an entity from one place to
another’. Accordingly, a purely directional expression (for example with the preposition
ad) was possible as well, which eventually became the normal expression in the Romance
languages (see § 4.52 and § 12.31). Conversely, the use of the dative was extended to verbs
of motion, without implying ‘transfer’. This extension is mainly poetic (Virgil alone has
some fifty examples) or poeticizing (but also present in the de Bello Hispaniensi) and was
probably influenced by Greek examples: it gave poets another means of avoiding
prepositions. Instances are found as late as Gregory of Tours.
On the basis of the meanings of the verbs involved, (a) and (b) are perfectly
understandable. However, the use of the dative dis immortalibus in (c) is different,
because no ‘transfer’ is involved. An interpretation as beneficiary or involved party
adjunct has been suggested (but see the manifestly directional expression ad vos in the
main clause). Virgil’s (d) is the example quoted most frequently to illustrate the dative
of direction.18
(a) Alterum apparet in funeribus indictivis, quo dicitur ‘ollus leto datus est’ . . .
(‘The other is heard in the case of funerals of which announcement is made, wherein
is said “that man has been given to death . . . ” Var. L. 7.42)
18
Virgil’s ex. (d) is partially influenced by Greek. See Coleman (1975: 131–2) and Calboli (2009:
100–4). See also Görler (1985: 266) on Virgil. The artificial character of this use of the dative is clearly
described by Löfstedt (1942: I.180–3). For the use of the dative with the verb venio ‘to come’, see
Cabrillana (1997b: 131–6), with references. For the history of the expression, see Sz.: 100–1.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(b) Quando ego te exemplis excruciaro pessumis / atque ob sutelas tuas te morti
misero . . .
(‘When I’ve tortured you in the harshest ways and put you to death for your
tricks . . . ’ Pl. Capt. 691–2)
(c) Tendit ad vos virgo Vestalis manus supplices easdem quas pro vobis dis
immortalibus tendere consuevit.
(‘To you a Vestal Maid extends in supplication the same hands which she has been
accustomed to extend to the immortal gods on your behalf.’ Cic. Font. 48)
(d) It clamor caelo, primusque accurrit Acestes . . .
(‘A shout mounts to heaven, and first Acestes runs forward . . . ’ Verg. A. 5.451)
Supplement:
Verum quorum liberi leto dati / sunt in bello, non lubenter haec enodari audiunt.
(Enn. scen. 334–5V=283–4J); Hunc . . . qui luci ediderat genitor Saturnius idem /
abdidit . . . (Cic. Div. 2.64—translating Hom. Il. 2.309 çø); . . . matres familiae
quae paulo ante Romanis de muro manus tendebant suos obtestari . . . coeperunt.
(Caes. Gal. 7.48.3); Quo de facto cum utrorumque copiae tumulum excellentem
petissent prohibiti a nostris sunt deiecti planitie. (B. Hisp. 24.3); . . . pelago Danaum
insidias suspectaque dona / praecipitare iubent . . . (Verg. A. 2.36–7); . . . aut haec in
nostros fabricata est machina muros / inspectura domos venturaque desuper urbi . . .
(Verg. A. 2.46–7); Inde Mago procul infensam contenderat hastam. (Verg. A. 10.521—
NB: human being); . . . et ipsum / pronum sterne solo portisque effunde sub altis. (Verg.
A. 11.485); . . . e curru saltum dedit ocius arvis . . . (Verg. A. 12.681); . . . ut formosa novo
quae parat ire viro. (Prop. 1.15.8—NB: human being); Pergere ad Treviros et externae
fidei. (Tac. Ann. 1.41.1—NB: coordination); Prohinc cubiculo te refer et lectulo lassitu-
dinem refove et ex arbitrio lavacrum pete. (Apul. Met. 5.2.3); . . . suis provinciis singulos
imperat revocari. (Vict. Vit. 3.60);
Locative forms and ablative forms are rarely found with verbs denoting movement
from one place to the other and have to be considered substandard. An example
is (a).19 Romae instead of Romam occurs until Late Latin. Sometimes, as in (b), it is
not clear whether this is a direction adjunct or a place adjunct (cf. the expressions
found with pono, § 4.81).
(a) Forte dominus Capuae exierat ad scruta scita expedienda.
(‘My master happened to have gone to Capua to look after some silly business or
other.’ Petr. 62.1 (Niceros speaking))
19
Sz.: 150–1; Adams (1977: 38, 2003a: 13, 2013: 332–7) and Mackay (1999) discuss a few instances.
A few late instances in TLL s.v. proficiscor 1712.18ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
More than one source adjunct in a clause is admitted if one can be understood as a
specification of the other. The relative order of the adjuncts is not restricted. Examples
are (j) and (k).20
(j) Haec omnia quae dixi signa, iudices, ab Heio e sacrario Verres abstulit.
(‘But Verres, gentlemen, carried off all the statues I have mentioned from the chapel
of Heius.’ Cic. Ver. 4.7)
(k) . . . (Latini) quibus ex Albano monte ex sacris carnem petere fuit ius . . .
(‘(Latins) who had the right to get a share of the meat at the sacrifices on the Alban
mount . . . ’ Var. L. 6.25)
Supplement:
Inveni duos solos libellos a L. Canuleio missos sociis ex portu Syracusis . . . (Cic. Ver.
2.182); VIII Id. Iun. Cularone ex finibus Allobrogum. (Planc. Fam. 10.23.7);
Topographical names in the ablative are discussed in § 10.17. A few common nouns
referring to a location are used in the same way as these topographical names, notably
domus ‘home’ and rus ‘countryside’ (on which see also § 10.17). This is extended both
to nouns indicating a location and, especially in poetry and in prose from Livy
onwards, to other nouns as well, among other things to avoid a prepositional
expression. Editors sometimes emend bare ablative expressions, as in (e).21 The
bare ablative expression gives way to prepositional expressions relatively early.22
(a) Nam qui in amorem praecipitavit, / peius perit quasi (=quam si) saxo saliat.
(‘The man who falls head over heels in love perishes in a worse way than if he were to
jump from the Rock.’ Pl. Trin. 265a–6)
20
For more instances, see TLL s.v. ex 1128.67ff. For a survey of instances of juxtaposition of more than
one source adjunct, see Müller (1895). See also Otto (1912: 27–35).
21
For more instances, see K.-St.: I. 361–3.
22
For the Vulgata, see Löfstedt (1985a), with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The use of the bare ablative is more often found with compound verbs, as in (f ), but is
even so very rare in Cicero, as in (g) (dubious) and in Caesar, as in (h). Examples from
other authors are given in the Supplement. In many instances one may regard the
source expression as an obligatory argument.
(f ) Ubiquomque usus siet, pectore expromat suo.
(‘Whenever necessary, he should draw a plan from his own mind.’ Pl. Bac. 653)
(g) . . . incommoda valetudo, <e> (edd.) qua iam emerseram . . .
(‘ . . . an indisposition from which I have now recovered . . . ’ Cic. Att. 5.8.1)
(h) . . . oppidani saxa quam maxima possunt vectibus promovent praecipitata-
que muro (de muro cj. Paul) in musculum devolvunt.
(‘ . . . the townsmen bring forward with cranes the largest possible stones, and roll
them headlong from the wall on to the gallery.’ Caes. Civ. 2.11.1)
Supplement:
. . . invenerit / aliquam causam, quam ob rem eiciat oppido. (Ter. An. 381–2); ‘cuius
ore sermo melle dulcior profluebat’. (Rhet. Her. 4.44); . . . quom interim Metellus
ignarus hostium monte degrediens cum exercitu conspicatur. (Sal. Jug. 49.4); Pro-
vehimur portu, terraeque urbesque recedunt. (Verg. A. 3.72); Eruimus terra solidum
pro frugibus aurum. (Ov. Am. 3.8.53); Hae litterae per forum ad tribunal praetoris
latae senatum curia exciverunt. (Liv. 27.50.9); . . . terra recipientes radices (caudices ?
cj. Fensterbusch) arborum in se sucum . . . (Vitr. 2.9.2); . . . eaeque fistulae castello
(<e> castello edd.) conlocentur ad castellum quod erit in moenibus. (Vitr.
8.6.4); . . . equi confossi hastis et dolore efferati iugum quatere et regem curru ex-
cutere coeperant . . . (Curt. 3.11.11); Milonem athletam, cum constitisset, nemo ves-
tigio educebat, malum tenenti nemo digitum corrigebat. (Plin. Nat. 7.83); Ea quasi
incensus nihil metuentem ferro transverberat et adcurrentem ancillam vulnere
absterret cubiculoque prorumpit. (Tac. Ann. 13.44.3);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Very rare and almost limited to Early Latin are ablatives of verbal nouns in combin-
ation with a verb of movement, as in (i), often regarded as supines in -u.23
(i) Redit eccum tandem opsonatu meus adiutor.
(‘Look, finally my helper’s returning from shopping.’ Pl. Cas. 719)
Supplement:
Primus cubitu surgat, postremus cubitum eat. (Cato Agr. 5); Et tunc venatu redi-
turum in limine primo / opperiens . . . (Stat. Ach. 1.119–20);
A few instances of source adjuncts where editors have inserted a preposition can
be found in the preceding section. The prepositional expressions with ab, de, and ex in
(a)–(c) parallel the examples in the beginning of that section. An example of implied
movement is (d).24 Some of the prepositions that are mainly used in position in space
adjuncts are also sometimes used to indicate source, as extra ‘from outside’ is in (e).
(a) In Sauracti <et> Fiscello caprae ferae sunt quae saliunt e saxo pedes plus
sexagenos.
(‘There are wild goats in Sauractus and Fiscellus which jump more than sixty feet
down from the rock.’ Cato hist. 52)
(b) . . . Liber agens celso Nysae de vertice tigris.
(‘ . . . Bacchus, driving his tigers down from Nysa’s lofty peak.’ Verg. A. 6.805)
(c) Et molientibus ab terra naves Philippus supervenit . . .
(‘And as they were struggling to cast off, Philip came upon them . . . ’ Liv. 28.7.7)
(d) Esne tu an non es ab illo militi Macedonio?
(‘Are you or are you not from that Macedonian soldier?’ Pl. Ps. 616)
(e) (loca) . . . quibus extra murum ex alto loco plano pede accessus fuerit ad
moenia oppugnanda.
(‘(places) . . . where there is an approach outside the wall from high ground by a level
footway for troops besieging the ramparts.’ Vitr. 1.5.6)
Latin has a few specialized pronominal source adverbs, viz. hinc ‘from here’, as in (a),
and dehinc ‘from here on’ (rare); inde ‘from there’, deinde (rare), and exinde ‘from
there’ (rare in Late Latin), indidem ‘from the same place’; also the interrogative and
relative adverb unde(cumque) and a few compounds with -unde. In addition, there are
23
For arguments not to consider them as such, see Kroon (1989a).
24
For further examples, see TLL s.v. a, ab 16.47ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
a few other source adverbs, like extrinsecus ‘from without’, as in (b). Finally, some of
the adverbs that are mainly used in position in space adjuncts are also found denoting
source, especially with motion verbs, as exemplified by intus in (c) (in Early Latin),
but also in other combinations, as extra ‘from outside’ is used with accipere in (d).
(a) Nunc, pater, ne peiierem / cura atque abduce me hinc ab hac quantum potest . . .
(‘Now, father, take care that I’m not perjuring myself and drag me away as quickly as
possible from here from this woman . . . ’ Pl. Bac. 1030–1)
(b) Haec autem in dicendo non extrinsecus alicunde quaerenda, sed ex ipsius
visceribus causae sumenda sunt.
(‘But the opening passage in a speech must not be drawn from some outside source
but from the very heart of the case.’ Cic. de Orat. 2.318)
(c) Evocate intus Cylindrum mihi coquom actutum foras.
(‘Call my cook Cylindrus out here to me at once.’ Pl. Men. 218)
(d) Qua propter cernere nemo / saepe supra potis est, at voces accipere extra.
(‘Therefore no one can see over a wall, though he can hear voices outside.’ Lucr.
4.610–11)
25
Notably Donatus 387.9K. See Pinkster (1972: 41).
26
Venio ad ‘Piraeea’, in quo magis reprehendendus sum quod homo Romanus ‘Piraeea’ scripserim,
non ‘Piraeum’ (sic enim omnes nostri locuti sunt), quam quod addiderim <‘in’>; non enim hoc ut oppido
praeposui sed ut loco; et tamen Dionysius noster et qui est nobiscum, Nicias Cous, non rebatur oppidum
esse Piraeea. (Cic. Att. 7.3.10).
27
Quod quo facilius efficeret aut necubi lectorem vel auditorem obturbaret ac moraretur, neque
praepositiones urbibus addere neque coniunctiones saepius iterare dubitavit, quae detractae afferunt
aliquid obscuritatis, etsi gratiam augent. (Suet. Aug. 86.1).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Generally speaking, there is much variation between individual authors. Poets and
certain prose authors such as Livy exploit the availability of the choice between bare
case and prepositional expressions. Regular instances with the names of towns are
(a)–(d). Ex. (a) has an ablative marking the source, (b) the accusative marking
direction, (c) a locative marking the place where, and (d) an ablative, also marking
the place where.
(a) Te pro filio / facturum dixit rem esse divinam domi, / quia Thebis salvos redierit.
(‘He said you were going to offer sacrifice at home for your son because he returned
safely from Thebes.’ Pl. Epid. 414–16)
(b) Sicut soror / eius huc gemina venit Ephesum et mater accersuntque eam.
(‘In fact, her twin sister and her mother have come here to Ephesus and want to fetch
her.’ Pl. Mil. 974–5)
(c) NOVIOS·PLAUTIOS·MED ROMAI·FECID
(‘Novius Plautius made me at Rome.’ CIL I2.561b (Praeneste, 250–235 BC))
(d) Ergo me potius in Hispania fuisse tum quam Formi<i>s cum tu profectus es
ad Pompeium!
(‘To think that I was in Spain instead of at Formiae when you left to join Pompey!’
Cael. Fam 8.17.1)
When the place name is modified by a topographical adjective the bare case expres-
sion is regularly used, and this is also often the case when the place name is modified
by totus ‘throughout the whole’ or by a pronoun. Poets especially use descriptive
adjectives with place names, with or without a preposition (see the Supplement).
Examples are (e)–(i).
(e) Cum Teanum Sidicinum venissem . . .
(‘When I got to Teanum Sidicinum . . . ’ Cic. Att. 8.11D.2)
(f ) Ego omnino, ut tibi proxime placuerat, Capuam veni eo ipso die quo tu
Teano Sidicino es profectus.
(‘For myself, pursuant to your wish expressed shortly beforehand, I went to Capua
the very day you left Teanum Sidicinum.’ Cic. Att. 8.11B.2)
(g) . . . alter autem eius filius Papia natus Teani Apuli . . .
(‘ . . . but his other son, by Papia, was being brought up under his mother’s care at
Teanum in Apulia.’ Cic. Clu. 27)
(h) Addidisti clausulam tota Tarracina tum omnibus in parietibus inscriptas
fuisse litteras tria L L L duo M M.
(‘You wound up by relating that three letters L.L.L. and two M.M. were inscribed on
every wall throughout the whole of Tarracina.’ Cic. de Orat. 2.240)
(i) Malo enim vel cum timore domi esse quam sine timore Athenis tuis.
(‘I would rather be frightened at home than secure in your Athens.’ Cic. Att. 16.6.2)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Topographical modifiers: Fateor me omnium / hominum esse Athenis Atticis
minimi preti. (Pl. Epid. 501–2); Victor exercitus opulentusque praeda Carthaginem
Novam in hiberna est deductus. (Liv. 21.5.4);
Pronouns: . . . (signa) quae ipsa Samo sublata sunt? (Cic. Ver. 1.51);
Descriptive modifiers: Et qui se sperat Romae regnare quadratae (Enn. Ann.
157V=150S—see ad loc.); . . . ditissimus agri / qui fuit Ausonidum et tacitis regnavit
Amyclis. (Verg. A. 10.563–4); Atque aliquis ‘iam nunc doctas eat’ inquit ‘Athenas.’
(Ov. Ep. 2.83); . . . et quos Hippocoon antiquis misit Amyclis . . . (Ov. Met. 8.314);
Other: Praecipue Carthagini prope ut captae tumultus fuit. (Liv. 29.28.4);
The use of prepositions with place names to indicate the place to which or from where
is rare in Classical Latin and is said to be restricted to the following situations: (i) when
the ab- and ad-phrases do not indicate the towns as such but their neighbourhood or
their position, as in (j)–(l); (ii) often when both the source and the goal of the action of
motion are expressed, the source is marked by a prepositional phrase, as in (n) and (o).
The first ‘rule’ seems to be confirmed by (j): Caesar was manifestly not in Gergovia. The
situations described in (k) and (l) are similar. But in an identical situation in (m) Caesar
has Dyrrachium without a preposition. As for the second ‘rule’, (p) serves as a counter-
example of (n)–(o). In light of this conflicting evidence it seems that these attempts to
explain the few exceptions in Caesar and Cicero are over-subtle. Prepositional phrases
(especially with ab, possibly to disambiguate source and position in space adjuncts) are
more common in other authors, certainly in Late Latin. Livy, for instance, has fifty-four
ab Roma vs zero Roma (abl.).28 A few are given in the Supplement.
(j) . . . consilia inibat quemadmodum a Gergovia discederet . . .
(‘ . . . he began to plan how he might withdraw from the vicinity of Gergovia . . . ’ Caes.
Gal. 7.43.5)
(k) Denuntiatum est ne Brutum obsideret, a Mutina discederet.
(‘He was solemnly warned not to besiege Brutus, to withdraw from Mutina.’ Cic.
Phil. 12.11)
(l) . . . nuntiant se . . . eum . . . cognovisse paratissimo animo ut cum suis copiis
iret ad Mutinam . . .
(‘ . . . they announce that they recognized that he was most ready to go to Mutina with
his forces . . . ’ Cic. Phil. 13.13)
(m) . . . (Caesar) mane Dyrrachium venit . . .
(‘ . . . (Caesar) arrived in the morning at Dyrrachium . . . ’ Caes. Civ. 3.41.5)
(n) Itaque ab Arimino M. Antonium cum cohortibus V Arretium mittit.
(‘And so he sends M. Antonius with five cohorts from Arminium to Arretium.’ Caes.
Civ. 1.11.4)
28
Figures taken from Adams (2013: 330).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
In poetry of all periods, but also in prose authors, names of countries or regions that
normally require a preposition—especially Aegyptus and Epirus—are often used without
a preposition. Compare (r) and (s) with (q). This usage is quite common in Late Latin.
The use of names of nations and rivers in this way is first attested in Virgil, as in (t).
(t) At nos hinc alii sitientis ibimus Afros, / pars Scythiam et rapidum cretae
veniemus Oaxen / et penitus toto divisos orbe Britannos.
(‘But we must go hence—some to the thirsty Africans, some to reach Scythia and the
chalk-rolling Oaxes, and the Britons, wholly sundered from all the world.’ Verg. Ecl.
1.64–6)
Supplement:
Ut <A>egypti in flumine quadrupes, sic in Latio nominati lu<t>ra et fiber. (Var. L.
5.79); Deinde Graeciae sicut apud nos delubra magnifica humanis consecrata
simulacris . . . (Cic. Rep. 3.14); . . . Thessaliae habitaverit et ibi opera eius latuerint.
(Plin. Nat. 34.68);
. . . qui Argum dicitur interemisse ob eamque causam Aegyptum profugisse . . .
(Cic. N.D. 3.56); . . . coniectans eum Aegyptum iter habere . . . (Caes. Civ. 3.104.1);
Hic clam praesidia Lusitaniam proficiscitur . . . (B. Hisp. 35.3); NB: Itaque ipse in
Lusitaniam proficiscitur . . . (B. Alex. 51.3); . . . cum Peloponnesum exercitum duxis-
set adversus Lacedaemonios . . . (Nep. Ep. 7.3); Arma virumque cano, Troiae qui
primus ab oris / Italiam fato profugus Laviniaque venit / litora . . . (Verg. A. 1.1–3);
Si iussa secutus / me vehis Hesperiam . . . (Luc. 5.533–4); Ipse praeceps Hiberos ad
patrium regnum pervadit. (Tac. Ann. 12.51.3); Unde contractis copiis omnibus
Mesopotamiam propere signa commovit . . . (Amm. 23.2.7); . . . Palaestinam reversus
Alexandriam perrexit, Aethiopiam adivit . . . (Hier. Ep. 53.1);
. . . nulla praeda, imperata tamen pecunia exercitum Britannia (v.l. ex Britannia)
reportabant. (Cic. Att. 4.18.5); . . . Aegypto venio . . . (Quint. Inst. 1.5.38); Et Britannia
Galliaque et Hispania auxilia Vitellius acciverat . . . (Tac. Hist. 3.15.1); . . . quod po-
pulum Aegypto profectum inhumane et inhospitaliter copiis defraudassent . . . (Tert.
Marc. 4.24.8);
Plautus uses prepositions alongside bare case forms for Greek towns, as in (u)–(w), which
can be compared with the bare case form Ephesum in the beginning of this section.
Supplement:
Plane hicine est qui / mihi in Epidauro virgini primus / pudicitiam perpulit. (Pl. Epid.
541–1b); Quin illam [quidem iam] in Sicyonem ex urbe abduxit modo. (Pl. Ps. 1098);
Quotumo die / ex Sicyone huc pervenisti? (Pl. Ps. 1173–4);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Bare case forms of the nouns domus ‘home’ (domi/ Late Latin domui, domum, domo,
and plural forms), humus ‘ground’ (humi, humum (exceptional), humo), and rus
‘countryside’, ‘estate’ (ruri/rure, rus, rure) are found as position, direction/goal, and
source expressions. They may be accompanied by a modifier. In the case of domus
this is understandably most often an indication of the possessor (a possessive
pronoun, genitive possessor, or an adjective referring to a possessor). Instead of
bare case expressions, prepositional phrases are also used, especially when the noun
is modified. A few illustrations of the use of domus are (x)–(ah).30
(x) At placuero huic Erotio, / quae me non excludet ab se, sed apud se occludet
domi.
(‘Erotium here will like me anyway. She won’t shut me out; oh no, she’ll shut me in
with her.’ Pl. Men. 670–1)
(y) Item hinc ultro fit, ut meret, potissumus / nostrae domi ut sit.
(‘I reciprocate of my own accord, as he deserves, so that he’s the most important
person in our house.’ Pl. Men. 359–9a)
(z) . . . qui istius domi erat educatus . . .
(‘ . . . who had been brought up in that man’s house . . . ’ Cic. Quinct. 21)
(aa) Attat, / quid illuc clamoris, opsecro, in nostra est domo?
(‘Goodness! What’s that shouting in our house, please?’ Pl. Cas. 619–20)
(ab) Nunc mea mater irata est mihi, / quia non redierim domum ad se . . .
(‘Now my mother’s angry with me, on the grounds that I didn’t return home to
her . . . ’ Pl. Cist. 101–2)
(ac) Sed quom inde suam quisque ibant divorsi domum . . .
(‘But when they all went to their homes on their different ways . . . ’ Pl. Rud. 1252)
(ad) Occiso Sex. Roscio primus Ameriam nuntiat Mallius Glaucia quidam . . . et
nuntiat domum non filii, sed T. Capitonis inimici.
(‘After Sextus Roscius had been killed, the news was first brought to Ameria by one
Mallius Glaucia and brought not to the son’s house, but to the house of his father’s
enemy, Titus Capito.’ Cic. S. Rosc. 19)
(ae) Quid igitur ego dubito? Aut quor non intro eo in nostram domum?
(‘Then why am I hesitating and why don’t I go into our house?’ Pl. Am. 409)
29
See Adams (2013: 337–44).
30
Detailed material about domus can be found in TLL s.v. 1956.22ff.; about humus s.v. 3124.21ff. The
observations in this section do not pertain to the adverb humi.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(af) Siquid uti voles, / domo aps te afferto, ne operam perdas poscere!
(‘If you want to use anything, just fetch it from home from your place; no need to
waste your energy asking.’ Pl. Aul. 340–1)
(ag) Iratus iste vehementer Sthenio atque incensus hospitium ei renuntiat, domo
eius emigrat atque adeo exit.
(‘Inflamed with violent anger against Sthenius, he renounced his hospitality and
moved out of his house—or rather stepped out of it.’ Cic. Ver. 2.89)
(ah) Tum ex Anniana Milonis domo Q. Flaccus eduxit viros acris.
(‘Then out came Q. Flaccus with some stout warriors from Milo’s house that he
inherited from Annius.’ Cic. Att. 4.3.3)
Supplement:
Domi suae nobilissimus vir . . . M. Drusus occisus est. (Cic. Mil. 16); . . . nonne mavis sine
periculo tuae domi esse quam cum periculo alienae? (Cic. Fam. 4.7.4); Domos
abeamus nostras, sultis, nunciam . . . (Pl. Poen. 814); . . . fragrantem Assyrio venit odore
domum . . . (Catul. 68.144); Mirumque, tradit barbaros suffitum huius herbae excipientes
ore lienes consumere et non egredi domibus nisi ab hoc suffitu . . . (Plin. Nat. 21.116);
. . . multa prius pelago multaque passus humo. (Ov. Tr. 5.3.12); . . . ut barbarorum
Claudius agmina / ferrata vasto diruit impetu / primosque et extremos metendo /
stravit humum sine clade victor . . . (Hor. Carm. 4.14.29–32); Qui primum eos humo
excitatos celsos et erectos constituit, ut deorum cognitionem caelum intuentes capere
possent. (Cic. N.D. 2.140); . . . summa tollebar humo. (Ov. Met. 2.587);
Nunc quoquo venias, plus plaustrorum in aedibus / videas quam ruri, quando ad
villam veneris. (Pl. Aul. 505–6); Cupiditates porro quae possunt esse in eo qui, ut ipse
accusator obiecit, ruri semper habitarit et in agro colendo vixerit? (Cic. S. Rosc. 39);
Rure meo possum quidvis perferre patique. (Hor. Ep. 1.15.17); Rus mane dudum
hinc ire me iussit pater . . . (Pl. Truc. 645); In Tiburte rus concesseramus hominis amici
divitis aestate anni flagrantissima . . . (Gel. 19.5.1); . . . equum conscendit et rus urba-
num, quod propter litus maris habebat, ignaris servis iussisque ad portam revertentem
opperiri, contendit. (Just. 31.2.3); Ipsos qui tibi subvectabant rure huc virgas ulmeas.
(Pl. As. 341); Rure suburbano poteris tibi dicere missa . . . (Ov. Ars 2.265);
Appendix: There are many prepositional expressions with place names for which no
bare case expression exists. A simple illustration is the prepositional phrase extra
Romam ‘outside Rome’ in (ai).
(ai) Sapiens existimari nemo potest in ea prudentia quae neque extra Romam
usquam neque Romae rebus prolatis quicquam valet.
(‘You cannot have a reputation for wisdom if it is based on knowledge which
has no value anywhere outside Rome and not even at Rome during the
vacation.’ Cic. Mur. 28)
Satellites
motion). They are regularly found with motion verbs like eo ‘to go’ and progredior ‘to
advance’, as in (b) and (c), but also with other actions that can be measured in some
way, as in (d) and (e). An example of an extent of space adjunct with a verb of motion
in a figurative sense is (f ). By way of illustration only instances of bare accusative
expressions are given: examples of other expressions follow below.31 (For extent of
space arguments, see § 4.84.)
(a) Quam longe est hinc in saltum vestrum Gallicanum? Naevi, te rogo. # ‘DCC
milia passuum.’ # Optime.
(‘How far is it from here to your pastures in Gaul? I ask you, Naevius. “Seven
hundred miles.” Quite right.’ Cic. Quinct. 79)
(b) Si hercle tu ex istoc loco / digitum transvorsum aut unguem latum
excesseris . . .
(‘Well, if you leave your place by just a finger’s or a nail’s breadth . . . ’ Pl. Aul. 56–7)
(c) . . . biduique iter progressus ad flumen Bagradam pervenit.
(‘ . . . and having completed a march of two days he arrived at the River Bagrada.’
Caes. Civ. 2.24.1)
(d) . . . scrobes tres in altitudinem pedes fodiebantur . . . (NB: vv.ll.)
(‘ . . . pits were dug three feet deep . . . ’ Caes. Gal. 7.73.5)
(e) . . . milia passuum decem novem murum in altitudinem pedum sedecim
fossamque perducit.
(‘ . . . he constructed a wall sixteen feet high and a trench over a distance of nineteen
miles.’ Caes. Gal. 1.8.1)
(f ) Videte . . . , iudices, quam longe videar ab consuetudine mea et cautione ac
diligentia discedere.
(‘Mark, members of the jury, how far I am departing from my customary habits of
caution and exactitude.’ Cic. Font. 12)
The following categories of constituents may function as extent of space adjuncts:
bare accusative case forms
bare ablative case forms
prepositional phrases
the adverb longe.
The most frequent way to mark extent of space adjuncts is by the bare accusative (but
see § 10.19). For the use of extent of space expressions in combination with position
in space adjuncts, see § 10.3, Appendix; for their use with adjectives and adverbs
indicating a dimension, see § 4.104 and § 11.92.
31
For a discussion of the internal properties of these expressions and their syntactic and semantic
status, see Torrego (2008).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of the use of the bare accusative have been given in the introductory
section. A few more can be found below in the Supplement. The ablative is uncom-
mon, with the exception of the nouns intervallum and spatium ‘distance’, which are
normally in the ablative; however, there is some individual variation. Examples of the
bare ablative are (a) and (b).32
(a) Clatros interesse oportet pede (pedem cj. Thielscher).
(‘The rack bars should be a foot apart.’ Cato Agr. 4)
(b) . . . (solum) eiusque abunde est gradum effodere tribus pedibus . . .
(‘ . . . and it is quite enough to dig up a spit of three feet . . . ’ Col. 11.3.10)
Supplement:
Bare accusatives: In Originum libro Cato scribit haec: ‘in Sauracti <et> Fiscello caprae
ferae sunt, quae saliunt e saxo pedes plus sexagenos’ (Var. R. 2.3.30); . . . MURUM /
2
AEDIFICANDUM·COIRAVERUNT·PED(ES) CCLXX . . . (CIL I .675.7–8 (Capua, 108 BC)); . . .
neminem esse qui possit biduo aut summum triduo septingenta milia passuum
ambulare . . . (Cic. Quinct. 78); . . . unius diei progressus iter . . . (Liv. 44.2.4); . . .
volutaeque ipsae sic caedantur altitudinis suae duodecimam partem. (Vitr.
3.5.7); . . . ut (Aetna) quinquagena, centena milia passuum harenas flammarum
globo eructet. (Plin. Nat. 2.234); . . . amnis Alpheus—navigatur VI milia passuum— . . .
(Plin. Nat. 4.14); Proditur Alexandrum nullo die minus stadia DC navigasse Indo . . .
(Plin. Nat. 6.60);
NB: elliptical: Hanc epistulam dictavi sedens in raeda cum in castra proficiscerer, a
quibus aberam bidui. (Cic. Att. 5.17.1);
Bare ablatives: Aut cur minimo declinent intervallo, maiore non? Aut cur declinent
uno minimo, non declinent duobus aut tribus? (Cic. Fat. 46); Tertia minor quidem
praedictis, sed multo spectatior, Aethiopicis lapidibus adsurgit CCCLXIII pedibus
inter angulos. (Plin. Nat. 36.80); Iugeri mensuram ducentos et quadraginta longi-
tudinis pedes esse dimidioque in latitudinem patere, non fere quisquam est qui
ignoret . . . (Quint. Inst. 1.10.42); . . . inter Mosam Rhenumque trium et viginti mi-
lium spatio fossam perduxit . . . (Tac. Ann. 11.20.2);
32
The fullest collection of instances is Müller (1908: 94–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
There are a few instances of the use of accusative neuter (singular) forms of descrip-
tive adjectives as extent of space adjuncts (the so-called accusativus adverbialis). This
usage is poetic and is also occasionally found in (post-Classical) prose authors.33
Examples are (e) and (f ). It is not always easy to distinguish such accusatives from
degree expressions.
(e) Creverat immensum.
(‘He grew to an enormous size.’ Ov. Fast. 5.537)
(f ) Extra inmane patens, tellus Mavortia, campi.
(‘Outside is a vast stretch of plain, the land of Mars.’ Stat. Theb. 4.434)
Supplement:
Mirum hercule, si cum . . . immensumque attolli provideret quem coniunctione tali
super alios extulisset, C. Proculeium et quosdam in sermonibus habuit . . . (Tac. Ann.
4.40.6); Transito fluvio modicum te progressa<m> textrices orabunt anus telam
struentes . . . (Apul. Met. 6.19.1);
Prepositional phrases with per ‘along a distance of ’ are found as an alternative of the
bare case expressions with actions of motion or implying motion from Livy onwards,
with an earlier example (a) in the de Bello Africo.
(a) . . . insecuti per III milia passuum . . . se ad suos recipiunt.
(‘ . . . chasing the enemy for three miles they retired to their homes.’ B. Afr. 78.8)
(b) Vallum fossamque, ingentis utramque rem operis, per tantum spatii
duxerunt.
(‘The rampart and the trench, each involving prodigious toil, they have carried all
that distance.’ Liv. 5.5.5)
(c) Nec non et Tityon, Terrae omniparentis alumnum, / cernere erat, per tota
novem cui iugera corpus / porrigitur . . .
(‘Likewise, one might see Tityos, nursling of Earth the mother of all. His body is
stretched over nine full acres . . . ’ Verg. A. 6.595–7)
33
Material in Müller (1908: 109–11).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
. . . ante vesperum licebit producere et docere per mille passus conposite ac sine
pavore ambulare. (Col. 6.2.5); Ipse per MCCXL milia passuum, parcissimis aucto-
ribus, navigatur . . . (Plin. Nat. 6.72); (an aquaduct) . . . supra terram per passus mille
ducentos quadraginta . . . (Fron. Aq. 10.7); PORTI/CUM·PER· PEDES·XXX (CIL III.976.6–7
(Alba Julia));34
The distance covered by a certain action must sometimes be inferred from prepos-
itional phrases with ad and in which indicate the limit or final point and which thus
resemble direction expressions. Examples are (d) and (e).35
The use of longe as an extent of space adjunct is illustrated at the beginning of § 10.18.
A few examples are added below.36
Illorum navis longe in altum abscesserat. (Pl. Rud. 66); Illa autem, Balbe, quae tu a
caelo astrisque ducebas quam longe serpant non vides. (Cic. N.D. 3.51); Invitati
praeda longius procedunt. (Caes. Gal. 6.35.7);
34
For more instances, see TLL s.v. per 1136.48ff.
35
For more instances, see Müller (1908: 94)
36
Procul ‘far’ is not used as an extent of space adjunct (see TLL s.v. 1557.41f.).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
The ablative nouns occurring as path adjuncts (the so-called ablativus viae or
ablativus prosecutivus) may be modified by any suitable modifier (determiner,
numeral, or topographical or descriptive adjective). The following five examples
illustrate these possibilities.
(a) Caute ut incedas via.
(‘Tread carefully on your road.’ Pl. Cur. 32)
(b) . . . exercitum eadem via qua adduxerat reduxit . . .
(‘ . . . he led the army back by the same way he had come . . . ’ Liv. 33.37.2)
(c) . . . luna dicitur esse, quae in caelo tribus viis movetur, in altitudinem et
latitudinem et longitudinem.
(‘ . . . she is said to be the moon, which moves in the sky by three ways, upwards,
sideways, and onwards.’ Var. L. 7.16)
(d) Itaque a. d. III Id. Nov. cum Sacra via descenderem, insecutus est me cum
suis.
(‘Accordingly, on 11 November as I was going down the Via Sacra, he came after me
with his men.’ Cic. Att. 4.3.3)
(e) Nunc iter conficiebamus aestuosa et pulverulenta via.
(‘At present we are en route on a hot and dusty road.’ Cic. Att. 5.14.1)
Supplement:
Common nouns (in alphabetical order by noun): . . . mox aere lapsa quieto / radit
iter liquidum . . . (Verg. A. 5.216–17); Abi istac travorsis angiportis ad forum. (Pl.
Per. 444); Statueram enim recta Appia Romam. (Cic. Att. 16.10.1); . . . aerius nascetur
Equus caeloque volabit . . . (Man. 5.633); It nigrum campis agmen. (Enn. Ann. 474V
=502S); Eadem autem celeritate adverso colle ad nostra castra . . . contenderunt.
(Caes. Gal. 2.19.8); . . . equitem Pedo praefectus finibus Frisiorum ducit. (Tac. Ann.
1.60.2); M’. vero Manilium nos etiam vidimus transverso ambulantem foro . . . (Cic.
de Orat. 3.133); Ipse de quarta vigilia eodem itinere quo hostes ierant, ad eos
contendit . . . (Caes. Gal. 1.21.3); Ille circumducto hoste qua voluit alio itinere ad
Capuam redit. (Liv. 25.19.8); Ipse iugis isdem exercitum reducit . . . (Liv. 22.15.4); Per
quem locum recta linea exire debebit a puncto circuli. (Hyg. agrim. p. 152); . . . neque
mari effugere ullo modo possent . . . (Cic. Ver. 5.91); . . . recta porta invadam
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
extemplo in oppidum antiquom et vetus. (Pl. Bac. 711); Si bene dicetis, vostra ripa
vos sequar. / Si male dicetis, vostro gradiar limite. (Pl. Poen. 631–2); . . . neque vero
quisquam potest hostis advolare terra, quin . . . (Cic. Rep. 2.6); Egressus est non viis,
sed tramitibus paludatus . . . (Cic. Phil. 13.19); Porta Collina urbem intravere sub
signis mediaque urbe agmine in Aventinum pergunt. (Liv. 3.51.10); (Hannibal) . . .
amnem vado traiecit . . . (Liv. 21.5.9);
Topographical names: . . . si tu Mesopotamia, nos Armenia circumgredimur
exercitum . . . (Sal. Hist. 4.69.21K); Omnes igitur undique clausi commeatus erant,
nisi quos Pado naves subveherent. (Liv. 21.67.5); Ex Tuscis frumentum Tiberi venit.
(Liv. 2.34.4);
NB: variation: . . . ut copiae per mare, dein Rhodano et Arare subvectae per eam fossam,
mox fluvio Mosella in Rhenum, exim Oceanum decurrerent . . . (Tac. Ann. 13.53.2);
Recto itinere in (f ) may be taken as a manner adjunct, because of its being combined
with per Umbriam in the same clause. In example (g) it may rather be taken as a path
adjunct. In a way, the difference is smaller than this description suggests. Path
adjuncts are usually taken as space adjuncts because the nouns refer to spatial entities,
but the relationship between the spatial entity and the action denoted by the verb and
its arguments rather resembles that of manner and instrument/means adjuncts in that
the subject is in control of the choice between several alternative routes.37
The only preposition that can be used in contexts in which the bare ablative is used is
per ‘along’. While the two expressions can be considered interchangeable and almost
synonymous when the nouns or noun phrases designate one-dimensional linear
entities like roads and rivers (compare per viam Sacram in (b) below, with (d) in
the preceding section), this need not be the case with other spatial entities. Esquilina
37
So e.g. K.-St.: I.350. Sz.: 130 organizes path adjuncts in the ablative as ‘Instrumental der Raum-
erstreckung’. More or less along the same lines Woodcock (1959: 31).
38
For discussion and references, see Mayer (1999: 161–3). See also § 4.11, (iii) on the ‘locative object’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
porta and per portam Capenam in (a) are probably conceptually different and not
synonymous: the use of per implies crossing a potential obstacle. This may explain
why with verbs of motion itinere is more common than per iter and why per fines is
more common than finibus. With nouns like porta and pons the bare ablative is more
common than the per-expression.39 In poetry the choice between the two alternatives
is also determined by metrical considerations.
(a) Lupus Esquilina porta ingressus frequentissima parte urbis cum in forum
decurrisset, Tusco vico atque inde Cermalo per portam Capenam prope
intactus evaserat.
(‘A wolf entered through the Esquiline Gate, passed through the most crowded part
of the city into the Forum, through the Vicus Tuscus and the Cermalus, and escaped
almost unharmed by the Porta Capena.’ Liv. 33.26.9)
(b) Dum descendunt per viam Sacram, interrogat Mercurius . . .
(‘As they passed downwards along the Sacred Way, Mercury asked . . . ’ Sen. Apoc.
12.1)
Supplement (in alphabetical order by noun): Volat ille per aera magnum / remigio
alarum . . . (Verg. A. 1.300–1); Abii illac per angiportum ad hortum nostrum clan-
culum. (Pl. Mos. 1044–5); Haec quoque per totum volitabit linea caelum . . . (Man.
1.654); Lento deinde agmine per fines Allobrogum ac Vocontiorum ductus
exercitus . . . (Tac. Hist. 1.66.3); . . . quos ipse per iter aliud ad diripienda tentoria
miserat Alamannica . . . (Amm. 27.2.9); . . . cum per Tiberim subiret navibus . . . (Vell.
2.45.5); . . . milites procul a flumine per viam minime ab regiis obsessam duxit . . .
(Liv. 24.40.9—NB: an ablative would be difficult in this case);
There are a few pronominal adverbs that occur as path adjunct. The idiom recta
‘directly’, ‘straight’ goes back to the ablative noun phrase recta via.
(a) Ea ibo opsonatum, eadem referam opsonium. / Per hortum utroque com-
meatus continet.
(‘That’s where I’ll go to buy provisions, and I’ll bring back the provisions the same
way; there’s a continuous path through the garden in both directions.’ Pl. St. 451–2)
(b) Iam ad regem recta me ducam resque ut facta est eloquar.
(‘This instant I’ll go to the king directly and tell him how this happened.’ Pl. Am. 1042)
Supplement:
Sequere hac me modo. (Pl. Am. 674); Abii illac per angiportum ad hortum nostrum
clanculum. (Pl. Mos. 1044–5); Thesprio, exi istac per hortum. (Pl. Epid. 660);
2
PORTICUMXQUA / INXARCEMXEITUR (CIL I .1529.5–6 (Aletrium, before 90 BC));
39
For the use of the ablativus prosecutivus, see Beckmann (1963: 16).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Just as with space adjuncts there is a broad spectrum of time adjuncts. The semantic
distinctions run more or less parallel to those discussed above in the sections on space
adjuncts. Certain expressions will appear to function in more than one of the
semantic subclasses distinguished below. An example is the adverb numquam
‘never’ which, depending on its context, may be interpreted as equivalent to ‘at no
point of time’ (position), ‘during no period of time’ (extent), and ‘zero times’
(frequency).40
More than one position in time adjunct can be used within a single clause if one can
be understood as a specification of the other. Examples are (d)–(g). The specification
most often follows, but the relative order of the adjuncts is not restricted. An example
of the reverse order, due to the contrast between ante lucem and vesperi, is (f ). The
two constituents are often immediately juxtaposed, but not necessarily so (see (g)).41
(d) Postero die mane a Pansa sum accersitus Bononiam.
(‘Early on the next day I had a message from Pansa summoning me to Bononia.’
D. Brut. Fam. 11.13.2)
(e) Hora nona illo ipso die tu es adoptatus.
(‘At the ninth hour on that same day you were adopted.’ Cic. Dom. 41)
40
For the various types of time adjuncts and the formal criteria that can be used to distinguish them,
see Torrego (1989b). For numquam, see Pinkster (2010).
41
The best survey of instances of juxtaposition of more than one position in time adjunct can be found
in Müller (1895). For the relative order of such constituents, see Spevak (2010b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(f ) Cum ante lucem VIII Kal. litteras ad te de Dionysio dedissem, vesperi ad nos
eodem die venit ipse Dionysius.
(‘After I had dispatched a letter to you about Dionysius before daybreak on the 22nd,
Dionysius himself arrived at my house that same evening.’ Cic. Att. 8.5.1)
(g) Pridie Idus Februarias haec scripsi ante lucem. Eo die apud Pomponium in
eius nuptiis eram cenaturus.
(‘I am writing this before dawn on 12 February. This evening I shall be dining with
Pomponius at his wedding.’ Cic. Q. fr. 2.3.7)
It is not always easy to decide whether one position in time adjunct specifies the
other, or if there is another explanation. In (h), ludis scaenicis is not a specification
but an apposition. In (i), summa senectute is not a specification of nostra memoria;
rather, the two position in time adjuncts seem to differ in scope. For position in time
adjuncts that function as settings, see Chapter 22.
(h) His ego tamen diebus, ludis scaenicis, . . . dirupi me paene in iudicio Galli
Canini . . .
(‘As for me, during these last few days, during the plays, I pretty well
ruptured my lungs defending Gallus Caninius . . . ’ Cic. Fam. 7.1.4)
(i) . . . quorum uterque nostra memoria summa senectute alter Oresten, alter
Pisonem adoptavit . . .
(‘ . . . each of whom adopted in our times, in extreme old age, Orestes and
Piso respectively . . . ’ Cic. Dom. 35)
Backward and forward span of time expressions and combinations of the two are also
used as position in time adjuncts, in the same way that direction and source
expressions are used as position in space adjuncts (see § 10.3). Examples are (j)–(l).
(j) Ego iam a principio in parsimonia . . . omnem adolescentiam meam abstinui . . .
(‘Already right from the beginning I restrained my entire youth in frugality . . . ’ Cato
orat. 128)
(k) Ibi tu ad hoc diei tempus dormitasti in otio.
(‘There you dozed in idleness until this time of the day.’ Pl. As. 253)
(l) A mani ad noctem usque in foro dego diem.
(‘I spend my entire day in the forum, from morning till night.’ Pl. Mos. 534)
Backward span of time expressions can be combined with expressions indicating the
extent of time between the time the event takes or took place and the time referred to in
the clause. Examples with abhinc in combination with a bare accusative are (m)–(n).42
42
The material can be found in Lundström (1961), where the putative, but non-existing bare ablative
expressions are discussed as well (the most difficult passages are Pl. Mos. 494; Cic. Q. Rosc. 37).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The most frequent way to mark position in time adjuncts is by the bare ablative.
10.28 The use of the bare ablative case in position in time adjuncts
The use of the bare ablative (the so-called ablativus temporis) is found not only
with nouns and noun phrases with a temporal meaning (year, month, day, etc.), as in
(a)–(c), but also with nouns and noun phrases indicating or implying an event or a
situation, as in (d)–(f ). In the latter case, the noun is usually modified, for example by a
numeral or an adjective with a temporal meaning or temporal implication. When there is
no modifier the preposition in is regularly used in Classical Latin, especially in Cicero and
Caesar. However, exceptions are found even in these authors—sometimes emended; and
the generalization ‘without modifier ! in, with modifier ! bare case’ becomes increas-
ingly void in the course of the development of Latin (see also the following section).
Satellites
43
For parallels, see TLL s.v. gladiator 2008.12ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
NB: a very exceptional use of principe in: Sexto demum consulatu Caesar Augustus,
potentiae securus, quae triumviratu iusserat abolevit deditque iura quis pace et
principe uteremur. (Tac. Ann. 3.28.2—NB: see Koestermann and see Woodman
and Martin ad loc. for other interpretations).
There are a number of prepositions that can be used to locate events in time in ways
for which no bare case alternatives exist. Examples are (a)–(d). The only one that can
be used as an alternative of the bare ablative expression indicating ‘a point of time at
which’ is in (+ abl.), which is treated separately below.
(a) (qui) . . . neque Athenas advenit umquam ante hesternum diem.
(‘ . . . and (who) never came to Athens before yesterday.’ Pl. Ps. 731)
(b) Post factum flector, qui antepartum perdidi.
(‘I’m changing course after the event, now that I’ve lost the property acquired in the
past.’ Pl. Truc. 343)
(c) . . . nec scribis quam ad diem te exspectemus.
(‘ . . . and you don’t say on which day I am to expect you.’ Cic. Att. 3.7.1)
(d) Postero die circa eandem horam in eundem locum rex copias admovit . . .
(‘On the following day around the same hour the king brought up his troops to the
same place.’ Liv. 42.57.10)
Other prepositions used in this way—many of them also used with a spatial
meaning—are circa ‘around’, citra ‘before’, cum ‘simultaneously with’, intra ‘within
the period of time terminated by’, ‘before’ (also with the same meaning in + abl.), per
+ acc. ‘in the course of ’ (see § 10.33 for another use of per),44 sub + acc. ‘just before’,
‘directly after’. Prepositions used in backward span of time expressions are ab ‘from’,
abusque ‘ever since the time of ’ (rare), de ‘starting at’, and ex ‘from’, ‘after’. In forward
span of time expressions ad ‘until’, adusque ‘right up to’ (rare), and usque ‘right until’.
With nouns that have no obvious temporal meaning and are not modified by one
of the types of modifiers mentioned in the preceding section, the use of in is regular,
certainly in Classical authors like Cicero and Caesar. Thus in is common with words
indicating political situation, age, and position in society: in bello ‘in war’, in pace ‘in
peace’, in proelio ‘in battle’; in adulescentia ‘in (one’s) youth’, in iuventute ‘during
(one’s) early manhood’, in pueritia ‘during (one’s) boyhood’, in senectute ‘in old age’;
in consulatu ‘during the consulship (of)’, ‘during (one’s) consulship’, in praetura
‘during the praetorship (of)’. Examples are (e)–(g). Exceptional bare ablative expres-
sions are given in the preceding section.
44
In combination with nouns denoting a season the bare ablative and the per expression are more or
less interchangeable, according to J. N. Adams (p.c.). See also TLL s.v. aestas 1087.9ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(e) Nostri enim sensus ut in pace semper, sic tum etiam in bello congruebant.
(‘For our sentiments, as ever in peace, so also in war, were in unison.’ Cic. Marc. 16)
(f ) Feci ego istaec itidem in adulescentia.
(‘I too did this in my youth.’ Pl. Bac. 410)
(g) (tribuni militum) . . . quorum in magistratu concordia domi pacem etiam
foris praebuit.
(‘During their administration domestic harmony insured peace abroad as well.’ Liv.
4.7.1)
Conversely, with nouns and noun phrases that have a temporal meaning, in (+ abl.) is
rarely used in Classical Latin prose, but is increasingly found in poetry and in later
authors in general. Examples of prepositional expressions are (h)–(j). In Late Latin in
die ‘in the daytime’ and in nocte ‘at night’, etc. are regular.
(h) Me nolo in tempore hoc videat senex.
(‘I don’t want the old master to see me just at the moment.’ Ter. An. 819)
(i) Disertissime Romuli nepotum, / quot sunt quotque fuere, Marce Tulli, /
quotque post aliis erunt in annis / gratias tibi maximas Catullus / agit . . .
(‘Most skilled in speech of the descendants of Romulus, all who are, and all who have
been, and all who shall be hereafter in other years, Marcus Tullius, to thee his
warmest thanks Catullus gives . . . ’ Catul. 49.1–5)
(j) Nunc ut in singulis mensibus sol signa pervadens auget et minuit dierum et
horarum spatia dicam.
(‘In the next place I will describe how the sun, going through the signs, augments and
diminishes in each month the length of the day and of the hour.’ Vitr. 9.2.4)
Supplement:
Scilicet arte madent simulacra et docta vagantur, / nocturno facere ut possint in
tempore ludos. (Lucr. 4.792–3); Haec fuit illa dies in qua Veientibus armis / ter
centum Fabii ter cecidere duo. (Ov. Fast. 2.195–6); Nutrix in tremulis ne quid
desideret annis / Parthenie. (Prop. 4.7.73–4); . . . satiatque semodius cibi in diebus
singulis vicenos et centenos turtures. (Col. 8.9.3); Eo in tempore Nero Anti agens
non ante in urbem regressus est . . . (Tac. Ann. 15.39.1); Ego quidem in hoc certe
tempore non fallo . . . (Gel. 1.3.3);
When the preposition in is used with temporal nouns like aetas, dies, tempestas, and
tempus modified by a demonstrative pronoun or a descriptive adjective, the prepos-
itional phrase is often interpreted as indicating the circumstances rather than
the position in time of an event. Ex. (i) is cited as an example of this interpretation
by K.-St. and so are (k)–(m). But, as K.-St. show, the same interpretation is also
possible in comparable bare ablative expressions, as in (n)–(p) (only expressions with
the noun tempus are given to illustrate this point).45
45
For further examples, see K.-St.: I.357–8; 360.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(k) Quae res in civitate duae plurimum possunt, eae contra nos ambae faciunt in
hoc tempore: summa gratia et eloquentia.
(‘Two things which have most power in the state—I mean great influence
and eloquence—are both working against us at this time.’ Cic. Quinct. 1)
(l) At vero in summo et periculosissimo rei publicae tempore etiam ab
inimicis eadem praetura laudatur.
(‘On the contrary, in a critical time of peril for the Republic even his enemies
extol that praetorship.’ Cic. Flac. 6)
(m) Iniquum esse et optimo cuique perniciosissimum non vitam honeste actam
tali in tempore quam plurimum prodesse . . .
(‘It is unfair and injurious to every good man that one’s honourable life in
the past should not be of the greatest possible help to him at such a time . . . ’
Cic. Inv. 2.36)
(n) . . . nedum his temporibus, his moribus, his magistratibus sine vestra sa-
pientia ac sine iudiciorum remediis salvi esse possimus.
(‘ . . . how much less at a time like the present, with such morals and such
magistrates, could we find safety, if it were not for your wisdom and the
redress provide by your courts?’ Cic. Clu. 95)
(o) Iste vero non procrastinat, locare incipit non proscripta neque edicta die
alienissimo tempore, ludis ipsis Romanis, foro ornato.
(‘But that one wasted no time; he proceeded with the tenders without any
previous advertisement or announcement of the day for tendering, at a most
unsuitable time, right in the middle of the Roman games, with the forum all
decorated.’ Cic. Ver. 1.141)
(p) Probas enim meum consilium negasque mihi quicquam tali tempore potius
faciendum fuisse.
(‘You approve my course of action and say that none better was open to me
in such circumstances.’ Cic. Att. 11.6.1)
With nouns denoting a phase such as initium and principium ‘beginning’, but also
with neuter substantival words like extremum ‘end’, the use of the preposition in
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
often implies ‘somewhere in’, but sometimes the distinction cannot be made.46
Further, sometimes it is not obvious whether one is dealing with a temporal or
with a spatial expression. Examples are (s)–(u).
(s) Sed amabo te, cura et cogita[tioni]—nihil novi, sed illud idem quod initio
scripsi.
(‘But do, I beg you, give care and consideration to—nothing new, only what
I wrote in the beginning.’ Cic. Fam. 2.7.2)
(t) Duo sunt, ut in initio dixi, quaestionum genera . . .
(‘Questions, as I said in the beginning, are of two kinds . . . ’ Cic. Part. 61)
(u) Quo etiam mihi durior locus est dicendi datus ut . . . ego in extremo non
partem aliquam agerem causae, sed de tota re dicerem quod mihi videretur.
(‘The position of my speech makes my task even harder than his, for I am
speaking last and am not dealing with any one part of the case but am saying
what I feel is required by it as a whole.’ Cic. Mur. 48)
The preposition intra ‘inside (a period of time)’ is used to indicate the limit of a period
in which an event occurs, as in (v) and (w).
(v) Quippe intra XII ferme annos huius temporis consules fuere Metelli aut
censores aut triumpharunt amplius duodeciens . . .
(‘For in fact within about twelve years of that time the Metelli were consuls or censors
or triumphed twelve more times . . . ’ Vell. 2.11.3)
(w) Ecce intra hos dies scripsit ut illi ex India semen boletorum mitteretur.
(‘Within the last few days, I may say, he has written for a cargo of mushroom spawn
from India.’ Petr. 38.4 (freedman speaking))
The prepositions ante ‘before’ (from Livy onwards) and post ‘after’ (from Cicero
onwards) are found with expressions of time, as in (x) and (y), respectively.47 For the
use of the adverbs ante and post see § 10.30.
(x) . . . M. Volscius Fictor, qui ante aliquot annos tribunus plebis fuerat, testis
extiterat . . .
(‘ . . . Marcus Volscius Fictor, who had been a tribune of the plebs a few years before,
had certified that . . . ’ Liv. 3.13.1)
(y) Cum eum cotidie videres, post biennium denique appellas.
(‘Although you saw him every day, you waited two years before you finally asked him
to pay.’ Cic. Quinct. 41)
46
For the material with in, see TLL s.v. in 778.60ff.
47
The material can be found in Lundström (1961: 92–7; 113–17).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Ante (and exceptionally post) in combination with the accusative diem ‘day’ and an
ordinal numeral functions as a complex preposition (comparable with pridie ‘the day
before’ and postridie ‘the day after’) that governs the accusative, as in (z). The entire
expression may be governed by another preposition, as in (aa).48
(z) Roma egreditur ante diem IIII Kalend. Februarias Quinctius Scipione et
Norbano coss.
(‘He set out and left Rome on January 27, during the consulship of Scipio
and Norbanus.’ Cic. Quinct. 24)
(aa) . . . diem edicti obire neglexit: in ante diem IV Kalendas Decembris distulit.
(‘ . . . he failed to appear on the day named in his edict, and he postponed the
meeting until 28 November.’ Cic. Phil. 3.20)
Position in time adverbs belong to several categories, including (i) former locative
forms of nouns denoting a period of time like heri ‘yesterday’, luci ‘by day’, temperi ‘at
the right time’ (also tempori, tempore, and in tempore in this meaning), vesperi ‘in the
evening’, diu ‘by day’, noctu ‘at night’; (ii) words that are also used as position in time
prepositions, like ante in (b)—compare also antea ‘previously’ and postea ‘after-
wards’; (iii) pronominal forms also used with a spatial meaning, like deinde ‘there-
after’, ‘next’ in (d) and ibi ‘there and then’; (iv) fossilized ablative forms of nouns and
noun phrases, like hodie ‘today’ in (a) and modo ‘just now’ in (b) and ilico ‘at once’,
originally a spatial adverb, in (e), also primo ‘at first’, ‘on the first occasion’ and
postremo ‘lastly’, ‘most recently’; (v) an adverb that originally expressed manner like
statim ‘at once’ in (f ). But there are many others which are less easily categorizable,
like dudum ‘some time ago’, tum ‘then’, nuper ‘recently’, umquam ‘ever’ and num-
quam ‘never’ (which is also used in other ways), aliquando ‘at some time or other’.
(a) Hanc fabulam inquam hic Iuppiter hodie ipse aget . . .
(‘This play, then, Jupiter will act himself here today . . . ’ Pl. Am. 94)
(b) Quid tecum, Stasime? # De istoc quod dixti modo. / Si ante voluisses, esses.
Nunc sero cupis.
(‘What are you talking to yourself about, Stasimus? # About what you’ve just said: if
you’d wanted it before, you’d be rich; now you wish for it too late.’ Pl. Trin. 567–8)
(c) . . . prioribus continuis diebus numquam ante horam quartam hostis
apparuerat.
(‘ . . . regularly on previous days the enemy had never appeared before the fourth
hour.’ Liv. 42.58.3)
48
For a discussion of the possible origin of this idiomatic expression, see Ramos (2007b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Many position in time adverbs are also used in other functions than position in time
adjuncts. For the discourse functions of notably the adverbs nunc and deinde, see
Chapter 24.
The adverbs ante (rarely antea) ‘before’, ‘in the past’ and post (rarely postea) ‘later’,
‘afterwards’ can be combined with a constituent in the ablative indicating the extent
of time (so-called ablativus mensurae—see Chapter 20), as in (g) and (h). The time
expression usually precedes the adverb. This type of expression is earlier attested than
the prepositional one mentioned above.49
(g) . . . vendiderat aedes eas quas ab eodem ipse paucis ante annis emerat.
(‘ . . . he sold the house which he himself had bought a few years ago from that same
man.’ Cic. Off. 3.67)
(h) . . . eaque is aegritudine / paucis diebus post Tarenti emortuo’st.
(‘ . . . and because of grief for him he died a few days later in Tarentum.’ Pl. Men.
35–6)
Adverbs used in backward span of time expressions are abhinc ‘from this time’ (see
§ 10.27) and illinc ‘from that time’ (very rare).
49
The material can be found in Lundström (1961: 100–2; 110–12).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Quod quaeris quam diu hic. Paucos dies, sed certum non habeo.
(‘You ask how long I shall be here. A few days, but I have nothing definitely fixed.’
Cic. Att. 12.40.5)
(b) Me nemo magis respiciet, ubi is [est] huc venerit, / quasi abhinc ducentos
annos fuerim mortuos.
(‘When he comes here, nobody will pay any more attention to me than if I’d been
dead for two hundred years.’ Pl. Truc. 340–1)
(c) Hanc domum / iam multos annos est quom possideo et colo / patri avoque
iam huius qui nunc hic habet.
(‘For many years already I’ve been occupying this house and protecting it for the
father and grandfather of the man who lives here now.’ Pl. Aul. 3–5)
(d) Annos iam triginta in foro versaris, sed tamen in Pergameno.
(‘For thirty years now you have practised at the Bar—I mean the Bar at Pergamum.’
Cic. Flac. 70)
50
The proportional use of these alternatives can be deduced from Müller (1908: 101–7).
51
Setaioli (2000: 84–5) found the following proportions in a sample of Seneca: abl. : per + acc. = 3 : 1;
abl. : acc. = 11 : 1.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Examples of the bare accusative are (a)–(d). Temporal nouns that are quantified are
modified by either a cardinal or an ordinal numeral, as can be seen in (d)–(e). The
nouns and noun phrases involved denote a period of time. The determiner hos ‘this’
in (c) indicates that the time of speaking is included in the period expressed by dies
multos; the adverb iam ‘already’ in (e) indicates that the period is counted from the
time the verb refers to (in this case the present).52 An example of an accusative neuter
form of a quantifying adjective with a genitive is (f ). (For the use of accusative forms
of substantival adjectives, see § 10.34.)
(a) . . . adeo ut aetatem ambo ambobus nobis sint obnoxii . . .
(‘ . . . so much so that both of them will be indebted to both of us for life . . . ’ Pl. As.
284)
(b) Plusculum annum / fui praeferratus apud molas tribunus vapularis.
(‘For a little longer than a year I was commander of beatings in the mill, shod with
iron.’ Pl. Per. 21–2)
(c) Sed mi intervallum iam hos dies multos fuit.
(‘But for me there has been a gap for these many days.’ Pl. Men. 104)
(d) Nam illaec me in alvo menses gestavit decem.
(‘For she carried me in her belly for ten months.’ Pl. St. 159)
(e) . . . sed ab illo tempore annum iam tertium et vicesimum regnat . . .
(‘ . . . but he has remained on the throne for twenty-two years from that date . . . ’ Cic.
Man. 7)
(f ) Nam Albinus . . . quantum temporis aestivorum in imperio fuit, plerumque
milites stativis castris habebat . . .
(‘For Albinus was generally keeping his soldiers in a permanent camp during the
time in the summer he was in command . . . ’ Sal. Jug. 44.4)
Supplement:
Septimum decimum annum ilico sedent. (Naev. Pun. 39); Ibi labore delassatum
noctem totam stertere. (Pl. As. 872); Quid, malum, me aetatem censes velle id
adsimularier? (Ter. Hau. 716); Fuisset tum illos mi aegre aliquot dies. (Ter. Ph.
159); Thebae, quae ante cataclysmon Ogygi conditae dicuntur, eae tamen circiter duo
milia annorum et centum sunt. (Var. R. 3.1.2); . . . eum ego Ephesi vidi fuitque
mecum familiariter triduum illud quod ego Ephesi commoratus sum . . . (Cic. Fam.
3.5.5); . . . discrepat aeternum tempus potuisse manere / innumerabilibus plagis
vexata per aevom. (Lucr. 1.582–3); . . . et ludi Romani biduum instaurati, item per
52
See Torrego (2010: 11–16).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
biduum plebeii ab aedilibus P. Aelio P. Villio . . . (Liv. 25.2.10); Quid dicam . . . nec
fora litibus strepere dies perpetuos . . . (Sen. Dial. 6.26.4); MIROR QUOD MIHI TOT
TEMPUS NIHIL RESCRIPSTI. (CEL appendix Vindol. Å.4–6 (Vindolanda, c. AD 100));
53
The bare ablative expressing extent of time is found from Varro onwards, as in (g),
with few instances in authors before Livy, after whom it becomes the regular bare case
expression. The nouns and noun phrases involved denote a period of time, as with the
bare accusative. Early examples are (g)–(j). Relatively common are examples like (k),
with ablative annis and accusative menses and dies next to each other.54
(g) . . . frumentum dicunt quindecim diebus esse in vaginis, quindecim florere,
quindecim exarescere, cum sit maturum.
(‘ . . . they say that the grain is in the sheath for fifteen days, blooms for fifteen days,
dries for fifteen days, and is then ripe.’ Var. R. 1.32.1)
(h) Itaque hic doctrina, consilio, eloquentia excellens, XL annis praefuit Athenis
et urbanis eodem tempore et bellicis rebus.
(‘And consequently he was eminent in learning, wisdom and eloquence, and for forty
years was supreme at Athens both in politics and at the same time in the conduct of
war.’ Cic. de Orat. 3.138)
(i) Hoc cum esset modo pugnatum continenter horis quinque . . .
(‘When they had fought in this way continuously for five hours . . . ’ Caes. Civ. 1.46.1)
(j) . . . ut liceat nobis tota perducere vita / aeternum hoc sanctae foedus
amicitiae.
(‘ . . . so that it may be our lot to extend through all our life this eternal compact of
hallowed friendship.’ Catul. 109.5–6)
(k) ·D(IS)· ·M(ANIBUS)· / CRESCE(N)S ET OLIMPIAS / PARENTES BENIGNO FI/LIO CARISSIMO
BE/NE MERENTI FECERUNT / QUI VIXIT ANNIS XVIII / MENSES TRES DIES / XV.
(‘To the shades of the departed. Crescens and Olimpias, the parents of the deceased,
have made this for their kind, dearly beloved, and well-deserving son, who lived
eighteen years, three months, and fifteen days.’ CIL VI.7901 (Rome, c.100 AD))
Supplement:
. . . quoi bini custodes semper totis horis occubant. (Pl. Mil. 212); Si Roma, unde eos
noverat Roscius qui Romam multis annis non venit neque umquam plus triduo fuit?
53
For tot = tantum, see Adams (1995a: 127).
54
For discussion and references, see Suárez (1994c).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(Cic. S. Rosc. 74); . . . multis saeclis verax fuisse id oraculum. (Cic. Div. 1.38); . . .
paucis diebus cibo se abstinuit atque ita interiit. (Hirt. Gal. 8.44.2); Metellus in isdem
castris quatriduo moratus saucios cum cura reficit . . . (Sal. Jug. 54.1); Namque ab eo
perpetua dissensit vita. (Nep. Cat. 1.3); His anxius curis ita se Africo bello . . . per
quinque annos, ita deinde novem annis in Hispania augendo Punico imperio gessit,
ut . . . (Liv. 21.2.1); Ex iis qui denis annis praefuerunt primus fuit Charops, ultimus
Eryxias, ex annuis primus Creon. (Vell. 1.8.3); Vivit autem longo tempore. (Plin.
Nat. 8.94); Gignit tota vita, quae est ei ad tricensimum annum. (Plin. Nat. 8.168);
Pistores Romae non fuere ad Persicum usque bellum annis ab urbe condita super
DLXXX. (Plin. Nat. 18.107); Nam si ad naturam rerum respexeris, etiam Nestoris
et Sattiae brevis est, quae inscribi monumento suo iussit annis se nonaginta
novem vixisse. (Sen. Ep. 77.20); Igitur amotus Cercinam, Africi maris insulam,
quattuordecim annis exilium toleravit. (Tac. Ann. 1.53.4); Vixit annis viginti
novem, imperavit triennio et decem mensibus diebusque octo. (Suet. Cal. 59.1);
NOMEN NON DICO NEC / QUOD [ = QUOT] VIXERIT ANNIS (CIL VIII.25006.2–3 (Carthage,
second or third century AD));
The main preposition that is used in extent of time expressions is per. Dictionaries
and grammarians distinguish two different meanings of the preposition per: ‘in the
course of ’ (mentioned in § 10.29) and ‘during the whole of ’.55 It is only in the second
meaning that per became equivalent to the bare accusative and ablative in extent of
time adjuncts (but see the note). The nouns and noun phrases involved denote a
period of time. A few examples will suffice.
(a) Sorbitione liquida hoc per dies septem dato.
(‘Administer it in a liquid form for seven days.’ Cato Agr. 157.12)
(b) . . . ut per biduum nemo hominem homo agnosceret . . .
(‘ . . . so that for two days no man could recognize his fellow . . . ’ Cic. N.D. 2.96)
(c) . . . Auster qui per biduum flaverat in Africum se vertit.
(‘ . . . the south wind, which had blown for two days, changed into a south-west wind.’
Caes. Civ. 3.26.5)
(d) . . . aut ideo durare aetatem posse per omnem . . .
(‘ . . . or else the reason why they can endure through all time . . . ’ Lucr. 3.811)
(e) Masinissa in spelunca occulta . . . duorum equitum latrocinio per dies aliquot
vixit.
(‘Masinissa in a hidden cave lived for some days on the booty of the two horsemen.’
Liv. 29.32.11)
55
So OLD s.v. per §6 and §7; TLL s.v. per 1142.58ff. and 1144.24ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Per has a wide range of meanings, many of which are not interchangeable with the
bare case expressions. In (f ) the per expression does not really specify the duration of
the action omnis potestas daretur but rather the period envisaged. Similarly in (g)
note the coordination with noctibus. In (h) the duration of the oratio is indeed meant,
but it is not very common to use a word with a non-temporal meaning as an extent of
time adjunct (unless we take per in the sense of ‘in the course of ’).
56
Material in Müller (1908: 109–10).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Unlike space adjuncts, time adjuncts expressing from what moment onwards and
until what moment a state of affairs lasts or lasted cannot be expressed by a bare
(ablative or accusative) case: prepositional expressions must be used instead, some-
times in combination with adverbs, in particular inde and usque.57 The resulting
combinations turned into idioms in which the original prepositions became super-
fluous and the adverbs function as prepositions.
In (a) these two types of time adjuncts are combined. In (b) a principio is combined
with inde. In the same way usque in (c) functions as a specification of adulescentiam. In
(d) and (e) inde and usque are combined with an ablative and accusative, respectively.58
(a) At ita me di deaeque omnes ament, / nisi mi supplicium virgarum de te datur
/ longum diutinumque, a mani ad vesperum . . .
(‘As truly as all the gods and goddesses may love me, unless I’m given your
punishment with rods as compensation, a long and enduring one, from dawn till
dusk . . . ’ Pl. Mil. 501–3)
(b) Inde a principio iam impudens epistula est.
(‘The letter is shameless right from the beginning.’ Pl. Bac. 1006)
(c) Nam is mecum a puero puer / bene pudiceque educatu’st usque ad
adulescentiam.
(‘From a young age till adulthood, the boy was brought up with me in a good and
decent fashion.’ Pl. Capt. 991–2)
57
For usque, see Wölfflin (1887b: 62–7)—recent editions to be consulted.
58
For disputed earlier instances of inde, see TLL s.v. 1110.24ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) . . . iam inde quadragesimi anni memoria replicabat post bellorum assiduos
casus . . .
(‘He was relating that, beginning forty years ago, after constant successes in war . . . ’
Amm. 18.5.7)
(e) . . . si id fieri non potest, sustinere tamen post cibum usque somni tempus . . .
(‘ . . . if this cannot be done, nevertheless to keep all back after food, until the time of
sleep . . . ’ Cels. 3.5.11)
Appendix: Prepositional phrases with ab may also indicate the relative order of
events and entities (‘after’), as in (f ).
The following categories of constituents may function as period of time within which
adjuncts:
bare ablative case forms
prepositional phrases.
The most frequent way to mark period of time within which adjuncts is by the bare
ablative.
10.37 The use of the bare ablative in period of time within which adjuncts
The use of bare ablative forms in period of time within which adjuncts is illustrated in
exx. (a)–(c).
(a) Numquam hoc uno die efficiatur opus, quin opus semper siet.
(‘The needs can never be met in one day without there being needs all the time.’ Pl.
Truc. 907)
(b) Nam ea quae Saturni stella dicitur . . . XXX fere annis cursum suum conficit . . .
(‘For the planet called Saturn completes its orbit in about thirty years . . . ’ Cic. N.D. 2.52)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(d) . . . id sum assecutus ut una hora qua coepi dicere reo audaci pecunioso
profuso perdito spem iudicii corrumpendi praeciderem . . .
(‘Within an hour after I began to speak I had cut short, in the mind of this
wealthy scoundrel and immoral spendthrift, his hopes of bribing the judges
who were to try him . . . ’ Cic. Ver. 1.20)
(e) . . . oppidum . . . paucis diebus quibus eo ventum erat expugnatum
cognoverant . . .
(‘ . . . the natives learned that a town . . . had been captured within a few days
of his arrival . . . ’ Caes. Gal. 3.23.2)
Bare ablative nouns and noun phrases can also be used to denote not so much the
entire period of time used to complete a state of affairs from its beginning to its end, as
above, but to indicate that a state of affairs will be or was terminated before or after
the end of a period of time in the future or in the past. Examples referring to the future
are (f ) ‘before the end of a period of two or three days’ and (g) ‘before a few days will
have passed’. An example referring to the past is (h): ‘the past two days’. Sometimes
the expression is interpreted (and translated) as equivalent to a prepositional phrase
with ante or post. So, for example, (i).59
(f ) Cura igitur ut valeas, et me hoc biduo aut triduo exspecta.
(‘Take care of yourself then, and expect me in two or three days’ time.’ Cic. Fam. 7.4)
(g) Dimidia pars exercitus si sibi permitteretur, (sc. se) paucis diebus Iugurtham
in catenis habiturum.
(‘If but half the army were put in his charge, he would have Jugurtha in fetters within
a few days.’ Sal. Jug. 64.5)
(h) Omnia erant facta hoc biduo laetiora.
(‘In the last couple of days everything has taken on a brighter look.’ Cic. Att. 7.26.1)
59
K.-St.: I.356–7 translate with ‘jetzt vor’ and ‘jetzt nach’. For instances of dies, see TLL s.v. 1037.21ff.
For instances of hora, see TLL s.v. 2956.49ff., but they are less precisely subcategorized.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The nouns involved have either an inherent quantifying element (like biduum ‘a
period of two days’) or have a quantifying adjective (like paucus ‘few’). They are often
modified by a determiner (like hic ‘this’) and are usually in the plural (annis ‘years’,
diebus ‘days’, horis ‘hours’).
Supplement:
. . . ut triduo hoc perpetuo prorsum e lecto nequeat surgere. (Ter. Ad. 520); . . . mors
Sex. Rosci quadriduo quo is occisus est Chrysogono nuntiatur. (Cic. S. Rosc. 105);
Eo biduo Caesar cum equitibus DCCCC quos sibi praesidio reliquerat in castra
pervenit. (Caes. Gal. 1.41.1); . . . illud argentum se paucis illis diebus misisse Lily-
baeum. (Cic. Ver. 4.39); Itaque paucis diebus cum auro et argento multo Romam
legatos mittit . . . (Sal. Jug. 13.6);
Apart from the preposition in + abl. illustrated in § 10.36 and in (a), the prepositions
that can be used to mark period of time within which adjuncts are inter and intra
‘inside (a period of time)’, as in (b) and (c).
(a) Credo equidem potesse te, scelus, / Massici montis uberrumos quattuor /
fructus ebibere in hora una. # ‘Hiberna’ addito.
(‘I do believe that you could drink up four very rich crops of Mount Massicius in one
hour, you crook. # Add “in winter”.’ Pl. Ps. 1302–4)
(b) Hoc constat, si quis ter in mense inter denos dies usus fuerit, in nocte non
laboraturum.
(‘This is certain, that if one has used it three times a month within ten days he will not
suffer.’ Larg. 159)
(c) . . . dimidiam partem nationum usque omnium / subegit solus intra viginti
dies.
(‘ . . . within twenty days he single-handedly subjected half of all nations on earth.’ Pl.
Cur. 447–8)
Satellites
Frequency expressions are either (i) pronominal adverbs like quotiens ‘how often?’, ‘as
often as’, quotiens(cum)que ‘as often as’, totiens ‘so often’, aliquotiens ‘several times’,
and (ii) numerical adverbs like semel ‘one time’, bis ‘twice’, ter ‘three times’, miliens ‘a
thousand times’, or (iii) adverbs with a more general meaning like crebro ‘frequently’,
‘often’, raro (rare) ‘rarely’ and its comparative and superlative forms, identidem
‘repeatedly’, saepe ‘often’ and its comparative and superlative forms, and in some
situations the zero quantifier numquam ‘never’ and semper ‘always’. The third group
seems to behave differently from the other two in that its members may co-occur with
members of the first two groups, while encompassing them within their scope, as in
(c) and (d). A further point of difference is that members of the first two groups are
regularly modified by a distributive expression indicating how often in a given period
the event occurs, as in (c)–(g). For the third group, by contrast, that is exceptional.
A few instances are found in Pliny the Elder, as in (h).
(c) Ita me miseram ad hunc modum / deciens die uno saepe extrudit aedibus.
(‘He drives a poor woman like myself out of the house like this, and often ten times in
a single day.’ Pl. Aul. 69–70)
(d) . . . quod bis accidit semper horarum XII spatio . . .
(‘ . . . which always happens twice in a space of twelve hours.’ Caes. Gal. 3.12.1)
(e) Reri’n ter in anno tu has tonsitari?
(‘Do you think they’re shorn three times a year?’ Pl. Bac. 1127)
(f ) Id aliquotiens in die cotidie facito . . .
(‘Do this daily several times in a day . . . ’ Cato Agr. 88.1)
(g) Et luxatum si quod est, bis die aqua calida foveto, brassicam tritam opponito.
(‘And if it is a dislocation, bathe it in warm water twice a day and apply a cabbage
poultice.’ Cato Agr. 157.4)
(h) Quare saepius die capulandum . . .
(‘Consequently it should be ladled several times a day . . . ’ Plin. Nat. 15.22)
As the examples show, the period within which expression is either a prepositional
phrase with in (+ abl.) or a bare ablative. In Early Latin and in Cicero the prepos-
itional expression is regular (but compare (f ) and (g)), but in other authors, especially
later ones, the bare ablative is regularly used.
Supplement:
Prepositional phrases: Quaen’ eapse deciens in die mutat locum, / ea ego auspicavi
in re capitali mea? (Pl. St. 501–2); Quidam has bis in anno tondent . . . (Var. R.
2.11.8); . . . unde vix ter in anno audire nuntium possunt. (Cic. S. Rosc. 132); . . . hic
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
alienus ovis custos bis mulget in hora. (Verg. Ecl. 3.5); In Ponto sunt quaedam albae,
quae bis in mense mella faciunt . . . (Plin. Nat. 11.59); . . . qui unum semel in anno
liberabant nunc cotidie populos conservant . . . (Tert. Bapt. 5.6); Lavacrum . . . eo quo
solet temporis intervallo tribuatur, hoc est semel in mense. (August. Ep. 211.13); Bis
quippe pariebant in anno. (August. Hept. I Quaest. Gen. 95, l. 1207); . . . cum haberet
necessitatem bis in die ingredi in sancta sanctorum . . . (August. Retract. 2.55, l. 19);
. . . certus undenos deciens per annos / orbis ut cantus referatque ludos / ter die claro
totiensque grata / nocte frequentis. (Hor. Saec. 21–4);
NB: Luxuriae ita indulsit ut saepius in die lavaret . . . (Suet. Gram. 23.5);
Bare ablative: Bis die cibum dant . . . (Var. R. 3.9.20); Bisque die numerant ambo
pecus, alter et haedos. (Verg. Ecl. 3.34); . . . bisque die lapsis Pagasaeae vertice silvae /
fontibus ora lavet . . . (Ov. Met. 12.412–13); . . . et fretum ipsum Euripi non septiens
die sicut fama fert temporibus statis reciprocat . . . (Liv. 28.6.10); Cibum bis die
assumere (Cels. 1.8); Euripon vocant, rapidum mare et alterno cursu septiens die
ac septiens nocte fluctibus in vicem versis adeo inmodice fluens, ut . . . (Mela 2.108);
Canum generi bis anno partus. (Plin. Nat. 8.151); Compitales Lares ornari bis anno
instituit vernis floribus et aestivis. (Suet. Aug. 31.4); . . . bis die, mane et vespere, ad
ecclesiam tuam sine ulla intermissione venientis . . . (August. Conf. 5.9); Pecudum
autem illius regionis fecunditas sicut Italarum tanta fertur ut bis anno pariant.
(August. Hept. I Quaest. Gen. 95, l. 1220);
NB: Ratio saepius anno serendi idem arvum. (Plin. Nat. 1, p. 56.49–50); Aegrotas
uno decies aut saepius anno . . . (Mart. 12.56.1);
Instances like die uno in (c) above and triduo and toto in anno in (i) and (j) below are
not easy to distinguish formally. Whereas in (c) regular repetition is implied, this is
not the case in (i) and (j): in these cases the bare ablatives and the prepositional
phrases are position in time adjuncts.60
(i) Se<r>vius Ocella nemini persuasisset se moechum esse nisi triduo bis
deprehensus esset.
(‘Servius Ocella would never have got anyone to believe that he went in for
adultery, if he had not been caught twice in three days.’ Cael. Fam. 8.7.2)
(j) . . . nec toto decies cacas in anno.
(‘ . . . and you don’t shit ten times in a whole year.’ Catul. 23.20)
A subtype of frequency adjuncts are those indicating ordinal rank, as in (k)–(m). With
the exception of iterum ‘for a second time’, the adverbs involved are derived from the
ordinal numbers, as in (k) and (l)—accusative neuter forms—and in (m), an ablative
form (less common). Less precise indicators of repetition are rursus and denuo.61
(k) . . . quasi qui nunc primum recipias te domum huc ex hostibus?
(‘ . . . as if you were coming home here from the enemy now for the first time?’
Pl. Am. 684)
60
For the juxtaposition of frequency and position in time adjuncts, see Pinkster (2010).
61
Rursus is more often used with the meaning of ‘restoring’ a previous situation and denuo is also used
with that meaning, albeit less often. See OLD and Bertocchi and Maraldi (2011a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
With terminative states of affairs the adverb may and sometimes must be taken as
indicating the reinstalment of a previous state (and not the repetition of the whole
event). This is the case in (n): Pelias had of course been young before. Ex. (o) is
ambiguous.62
(n) . . . Medea Peliam . . . dicitur / fecisse rursus ex sene adulescentulum . . .
(‘ . . . Medea . . . is said to have turned Pelias from an old man into a young one again . . . ’
Pl. Ps. 869–71)
(o) . . . usque mantant neque id faciunt, donicum / parietes ruont. Aedificantur
aedes totae denuo.
(‘ . . . they keep waiting and don’t do it until the walls come down. The entire house is
built over again.’ Pl. Mos. 116–17)
62
For this ‘restitutive’ interpretation, see Bertocchi and Maraldi (2011a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) . . . num aut in vilitate nummum arator quisquam dedit aut in caritate de
aestimatione frumenti questus est?
(‘ . . . and did a farmer ever pay him one penny when it was cheap or complain of the
rate of his commutation when it was dear?’ Cic. Ver. 3.216)
(f ) Veri simile non est ut, quem in secundis rebus, quem in otio secum semper
habuisset, hunc in adversis et in eo tumultu, quem ipse comparabat, ab se
dimitteret.
(‘It is not reasonable to suspect that in the midst of adversity and a revolution of
which he himself was the instigator he would part with the man whom he had always
had with him in times of prosperity and peace.’ Cic. Sul. 57)
Supplement:
Bare ablatives: Aere vespertino rursus pascunt ad solis occasum. (Var. R. 2.2.11);
Quid enim, obsecro te, terrane tibi hoc nebuloso et caliginoso caelo aut sata aut
concreta videtur tanta vis memoriae? (Cic. Tusc. 1.60); Cum aestu magno ducebat
agmen, laborabat. (Cic. Tusc. 2.35); Nobis enim hac infirmitate exercitus, inopia
sociorum, praesertim fidelium, certissimum subsidium est hiems. (Cic. Att. 5.18.1);
Inde Austro lenissimo caelo sereno nocte illa et die postero in Italiam ad Hydruntem
ludibundi pervenimus eodemque vento postridie (id erat a. d. VII Kal. Dec.) hora IIII
Brundisium venimus . . . (Cic. Fam. 16.9.2); At contra quis est omnium his moribus,
quin . . . (Sal. Jug. 4.7); Parte alia ventis et dis Agrippa secundis / arduos agmen agens.
(Verg. A. 8.682-3); . . . non audebam tantis occupationibus de architectura scripta . . .
edere . . . (Vitr. 1 pr. 1); Paulo enim post lucis exortum densitate praevia fulgorum
acrius vibratorum tremefacta concutitur omnis terreni stabilitas ponderis . . . (Amm.
26.10.16);
NB: without a modifier: Etiam tum tamen nubilo incertae fuere horae usque ad
proximum lustrum. (Plin. Nat. 7.215);
Prepositional phrases: Elusi militem, inquam, in alea. (Pl. Cur. 609); . . . ut quod in alea
perdiderat beneficio legis dissolveret. (Cic. Phil. 2.56); Quare in tranquillo tempestatem
adversam optare dementis est . . . (Cic. Off. 1.83); Producit servos quos in pabulatione
paucis ante diebus exceperat et fame vinculisque excruciaverat. (Caes. Gal. 7.20.9); Et cui
protinus in recenti morbo bilis atra vel infra vel supra se ostendit; cuive alterutr<o>
modo se prompsit, cum iam longo morbo corpus eius esset extenuatum et affectum.
(Cels. 2.6.8); Ipsa natura enuntiabit quales nos ante pabulum et potum in virgine adhuc
saliva exhibere consuerit rebus dumtaxat sensu agendis . . . (Tert. Ieiun. 6.1);
The ablatival adjuncts in this section are called ‘sociative’ by some linguists, ‘attend-
ant circumstances’ by others, while others again regard them as ‘normal’ ablative
absolutes.63 The difference between these instances and ‘real’ ablative absolutes as
described in this Syntax is that here one is dealing with nouns and noun phrases,
usually modified by an adjective or a demonstrative pronoun, whereas the ablative
63
Many of the examples given in the text occur in the list of ‘sociative’ ablatives in Flinck-Linkomies
(1929: 158ff.). Some are included in the list of instances of ‘attendant circumstances’ in Roby (1882: 100)
and regarded as ablative absolutes by K.-St.: I.410. For the concessive interpretation of such expressions,
see Iordache (1999).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
absolutes are clauses (see § 14.14). Obviously there are cases where one cannot easily
decide. The instances of in-phrases are called ‘in occasionis’ in the TLL.64 The
instances with an event noun or a noun implying an event (so in adulterio in (d))
resemble gerund and gerundival clauses with or without in (see § 16.102 and
§ 16.108, respectively). Note in (d) the parallelism with the temporal dum-clause.
64
S.v. 779.65ff., where also other semantic interpretations (e.g. concessive ones) are exemplified.
65
The wording in the section heading is taken from Quirk et al. (1985: 482).
66
For these two interpretations, see Torrego (1994b), also for discussion of the so-called iterative use
of adhuc.
67
For iam in general (also its non-temporal functions), see Kroon and Risselada (1998, 2002). For iam
in negative sentences, see Huitink (2005). For iam, etiam, nondum, and non amplius in negative contexts,
see Orlandini (2001: 192–210).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
What PROCESS ADJUNCTS have in common is that they are preferably combined with
controllable states of affairs (but see exceptions below). They answer the question quo
modo? ‘how?’, as in (a)–(b).
(a) Quo modo habeas, id refert, iurene anne iniuria.
(‘Whether you’ve got it lawfully or not, that’s what matters.’ Pl. Rud. 1069)
(b) Quid ergo addit amplius? # Suam familiam fecisse dicit. # Quo modo? # Vi,
hominibus armatis.
(‘Therefore what more is he adding? # He says that his slaves have done it. # How? #
With force, with armed men.’ Cic. Tul. 24–5)
The most common process adjuncts are manner (§ 10.43) and means and instrument
adjuncts (§ 10.53). Price adjuncts may be regarded as a subcategory of instrument
adjuncts, although it is difficult to imagine that they could be the answer to the
question quo modo? They are treated as a separate category of process adjuncts
(§ 10.56). Among process adjuncts are also counted norm adjuncts (§ 10.49), quantity
and degree adjuncts (§ 10.61), beneficiary adjuncts (§ 10.69), and associative adjuncts
(§ 10.72).
A clause may contain more than one process adjunct, as in (c). One might assume a
hierarchical difference, iniuste modifying loqui and per iocum modifying the whole
state of affairs (note also the variation ioco and per iocum).
(c) Heia, hau vostrum est iracundos esse quod dixi ioco. / # Nec tuom quidem
est amicis per iocum iniuste loqui.
(‘Goodness! You shouldn’t be angry about what I said in jest. # And you shouldn’t
insult your friends in jest.’ Pl. Poen. 572–3)
Quo modo? is a very general question that may also correspond to English ‘how
come?’. This explains some instances of its co-occurrence with real manner adjuncts,
like stulte in (d). But it may also co-occur with manner expressions in its literal sense,
as in (e). In addition to quomodo, other expressions containing modo can be
juxtaposed with process adjuncts in a similar way, as in (f ) and (g). In these cases
it must be assumed that there is a difference in the hierarchical position of the two
adjuncts.68
68
For these instances, see Ripoll (2007: 96–8), whose precise description differs from the one
presented in the text. For the hierarchical difference between the two uses of quomodo, see also Bodelot
(2012: 426).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(e) Sic qui se ad totam vitam instruxit non desiderat particulatim admoneri,
doctus in totum non enim quomodo cum uxore aut cum filio viveret, sed
quomodo bene viveret.
(‘So he who has equipped himself for the whole of life does not need to be
advised concerning each separate item, because he is now trained to meet his
problem as a whole; for he knows not merely how he should live with his
wife or his son, but how he should live aright.’ Sen. Ep. 94.3)
(f ) Nunc ad sociorum tabulas accepti et expensi, quas removere honeste nullo
modo potuerunt, et ad amicum tuum Carpinatium revertemur.
(‘We will now go back to the company’s accounts of receipts and expend-
iture, which they had no respectable means of suppressing, and to your
friend Carpinatius.’ Cic. Ver. 2.186)
(g) Ita prorsus, inquam; etsi ea quidem quae adhuc dixisti quamvis ad aetatem
recte isto modo dicerentur.
(‘ “Very true,” said I; “albeit the style of your discourse so far has been suited
to hearers of any age.” ’ Cic. Fin. 5.27)
Within the category of MANNER ADJUNCTS a distinction can be made between three
subcategories: subject-oriented, object-oriented, and verb-oriented. There are two
reasons to make this threefold distinction. The first is that there are several instances
of coordination of a subject- or object-oriented manner adverb with an adjective, as in
(c) (coordination with the subject complement invitus) and (d) (coordination with
the object vera) (for further examples, see Chapter 19 on coordination). The other
reason is that sometimes two or more manner satellites co-occur within the same
clause without being coordinated, as in (e). Studiose ‘deliberately’ is subject-oriented,
whereas maledice ‘in an abusive manner’ and contumeliose ‘in an insulting manner’
qualify the words used (they are object-oriented). Finally, per ridiculum ‘to raise a
laugh’ and severe ‘with severity’ may be taken as qualifying the action of speaking
(verb-oriented).
(c) At ille tarde quidem et invitus, sed postquam precibus etiam iracundiam
miscui . . . inquit . . .
(‘That boy, slowly to be sure and unwillingly and only after I had mixed anger with
entreaties, said . . . ’ Petr. 9.3)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Recte et vera loquere. Sed nec vere nec <tu> recte adhuc / fecisti umquam.
(‘Tell me the honest truth; but so far you’ve never behaved truthfully or honestly.’
Pl. Capt. 960–1)
(e) In primisque provideat ne sermo vitium aliquod indicet inesse in moribus
quod maxime tum solet evenire cum studiose de absentibus detrahendi
causa aut per ridiculum aut severe maledice contumelioseque dicitur.
(‘He should take care above all that his speech does not reveal that there is
some fault in his behaviour; in general that happens particularly when someone
speaks quite deliberately about people who are absent in an abusive or insulting
manner in order to disparage them, whether he does so to raise a laugh or
with severity.’ Cic. Off. 1.134—tr. M. T. Griffin and E. M. Atkins, Cambridge
1991)
Satellites
(iii) Verb-oriented manner adjuncts describe the way the action denoted by the
verb is executed. Examples are (k)–(n).
(k) Sed unde vos / ire cum uvida veste dicam, opsecro, / tam maestiter vestitas?
(‘But where should I say you’ve come from with wet clothes, please, dressed so
gloomily?’ Pl. Rud. 264–5a)
(l) Atque edepol tu me monuisti probe.
(‘Yes, you’ve given me good advice.’ Pl. Men. 385)
(m) . . . si (modiolus) leviter imprimitur, parum proficit. Si graviter, non
movetur.
(‘ . . . for if pressed lightly it makes little advance, if heavily it does not rotate.’ Cels. 8.3.2)
(n) Quinque consulatus eodem tenore gesti vitaque omnis consulariter acta
verendum paene ipsum magis quam honorem faciebant.
(‘Five consulships administered on the selfsame principles, and a life which had
throughout been of consular dignity, made the man himself almost more revered
than his office.’ Liv. 4.10.9)
Like other process adjuncts, manner adjuncts are restricted in their occurrence to
controlled states of affairs when human beings are involved. In other contexts the
notion of ‘control’ does not apply, so for example in (o) and (p).
(o) Quod per amoenam urbem leni fluit agmine flumen.
(‘Because the river flows with gentle train through a pleasant town.’ Enn. Ann.
173V=163S)
(p) . . . collem clementer et molliter adsurgentem . . . maxime probaverunt.
(‘ . . . they especially favoured hilly land rising gently and gradually.’ Col. 2.2.1)
The only seeming exceptions to the rule that states of affairs have to be + control when
human beings are involved are (q)–(s) with the verb sum. However, the first two
examples refer to the behaviour of the subjects. The third describes the sojourn of the
subjects (see § 4.95 on the copula sum and the notes on these specific instances in § 9.36).
Some lexical items used as manner adjuncts may also occur as attitudinal disjuncts
expressing a subjective evaluation (see § 10.104). An instance of this is comiter
(‘friendly’): whereas in (t) comiter characterizes the way the ambassadors were
received, in (u) it does not characterize the action of sending but evaluates the
whole event as a sign of friendliness. In English this would appear from word
order: they received them kindly vs they kindly sent them gifts.
(t) . . . legatos comiter acceptos . . .
(‘ . . . the ambassadors were courteously received . . . ’ Liv. 29.11.7)
(u) . . . legatis . . . comiter munera missa.
(‘ . . . presents were sent to the ambassadors as a sign of courtesy.’ Liv. 9.43.26)
One may sometimes be tempted to interpret evaluative adverbs like stulte ‘foolishly’ as
attitudinal disjuncts. However it seems better to interpret exx. like (v) and (w) as subject-
or verb-oriented adjuncts. Ex. (x), by contrast, does illustrate the use of stulte as an
attitudinal disjunct: here the whole event, including the negator non, is evaluated as
foolish. Similar options exist for recte ‘rightly’,69 repente ‘suddenly’, and subito ‘suddenly’.
(v) Num stulte anteposuit (sc. Damaratus) exilii libertatem domesticae
servituti?
(‘Surely he was not foolish in preferring freedom in exile to slavery at home?’
Cic. Tusc. 5.109—i.e. was Damaratus foolish in preferring . . . ?—tr. Douglas)
(w) . . . illa, quae in corpore excellerent, stulte antiquos dixisse per se esse
expetenda. Sumenda potius quam expetenda.
(‘ . . . and so also with bodily excellences, it was foolish of the ancients to call them
“desirable for their own sakes”; they were not “desirable” but “worth taking”.’ Cic.
Fin. 4.20—i.e. the statement of the ancients was foolish)
(x) Stulte non nosti detrimenti tui tempora.
(‘How foolish! You do not understand when your loss took place!’ Sen. Ben.
7.30.1)
It is not always easy to understand why a manner adjunct is chosen instead of a
secondary predicate (cf. Chapter 21). An example is (y).
(y) Parve per eos (eas [sc. stellas] cj. Soubiran) flectitur Delphinus.
(‘The Dolphin moves dimly among them.’ Vitr. 9.4.5)
69
In a slightly different way, Ripoll (2007: 135–40).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
The bare ablative is common with noun phrases containing a modifier, but prepos-
itional phrases, especially with cum are frequent as well. With unmodified nouns the
prepositional expression is more common, but there are many expressions (more or
less developed into idioms or even adverbs) in the bare ablative as well.
An interesting instance of coordination of these various expression types can be
seen in (z). In (aa) the use of the prepositional phrase cum virtute to paraphrase the
preceding adverb honeste illustrates the functional similarity of the two.
(z) Cum religione, caste, sanctisque moribus is locus debet tueri.
(‘That place must be kept with religious scruple, purely and with strict manners.’
Vitr. 1.7.2)
(aa) . . . idque esse beate vivere, honeste, id est cum virtute, vivere.
(‘ . . . and that to live honourably, that is virtuously, is to live happily.’ Cic. Fin. 3.29)
Supplement:
. . . quo (sc. agro) cum labore magno et misere viveret. (Pl. Aul. 14); Sapienter factum
et consilio bono. (Pl. Aul. 477); . . . uti curiosius summaque providentia solidationis
ratio habeatur. (Vitr. 7.1.1);
Co-occurrence of two manner expressions in the same clause, one being more specific
than the other, is occasionally found in Late Latin.70
The ablative as a marker of manner adjuncts (the so-called ablativus modi) is found
from Early Latin onwards. Examples of modified noun phrases in the ablative are (a)–
(d). Unmodified nouns are (e)–(h). Examples of manner adjuncts in clauses without a
human agent are (d) and (i)–(k). Ex. (k)—where rivis indicates in what form the
stream manifests itself—is typical of Virgil and other poets, but not restricted to
them.71 The nouns and noun phrases used in manner adjuncts include both nouns
denoting or implying an action, a state of mind, or an emotion of the subject of the
main clause (they are controller-oriented)—where often a prepositional phrase with
cum or per is possible as well—and nouns that denote a property of the verbal process.
An example of the latter is celeritate mirabili in (d), where a possible paraphrase
would be celeritas conficiendorum orbium mirabilis est ‘the speed with which they
complete their orbits is marvellous’. Finally, there are nouns that denote more or less
precisely manner themselves: modo or modis, ratione, and more or moribus +
modifier (a determiner, adjective, or noun phrase in the genitive), as in (c).
70
See Rosén (2005: 227–8).
71
See Görler (1985: 268), with references. Also Hillen on Seneca’s use of the so-called ablativus
explicativus (1989: ch. 2).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Noun phrases: . . . de ea re signa atque argumenta paucis verbis eloquar. (Pl. Am.
1087); At sci’n quo pacto me ad te intro abducas? (Pl. Bac. 1178); Eadem opera a
praetore sumam syngraphum. (Pl. Capt. 450); Vos omnis opere magno esse oratos
volo / benigne ut . . . (Pl. Cas. 21–2); Solet iocari saepe mecum illoc modo. (Pl. Men.
317); Omnibus hic ludificatur me modis. (Pl. Mer. 920); Nunc, ne quis erret vostrum,
paucis in viam / deducam . . . (Pl. Trin. 4–5); Graeco ritu fiebantur Saturnalia. (Cato
orat. 77); . . . polliceor hoc vobis, Quirites, bona fide. (Cic. Agr. 2.100); Quodsi iam
sint id quod summo furore cupiunt, adepti . . . (Cic. Catil. 2.19); Ac postea decerni-
tur, ac non varie sed prope cunctis sententiis . . . (Cic. Ver. 4.145); Nam fuit quoddam
tempus, cum in agris homines passim bestiarum modo vagabantur . . . (Cic. Inv. 1.2);
Cato, qui Siciliam tenere nullo negotio potuit . . . (Cic. Att. 10.16.3); Quanto studio
dignitatem tuam et in senatu et ad populum defenderim ex tuis te malo quam ex me
cognoscere. (Cic. Fam 12.7.1); Simili ratione Pompeius confecto eius diei itinere in
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
suis veteribus castris ad Asparagium consedit. (Caes. Civ. 3.76.2); (Marius) . . . pariter
atque in conspectu hostium quadrato agmine incedere. (Sal. Jug. 100.1); Sed non in
omnibus locis est aggeris ratio facienda, nisi quibus extra murum ex alto loco plano
pede accessus fuerit ad moenia oppugnanda. (Vitr. 1.5.6);
Unmodified nouns: . . . si eam senex anum praegnatem fortuito fecerit, / quid dubitas
quin sit paratum nomen puero Postumus? (Pl. Aul. 163–4); . . . ut omnes fere
Stoici prudentissumi in disserendo sint et id arte faciant sintque architecti
paene verborum . . . (Cic. Brut. 118); . . . quae beneficia aeque magna non sunt
habenda atque ea quae iudicio considerate constanterque delata sunt. (Cic. Off.
1.49); . . . prima confecta vigilia IIII milia passuum secundo flumine silentio progredi
ibique se ex<s>pectare iubet. (Caes. Gal. 7.60.1); Ac primo . . . strepitu velut hostes
adventare . . . (Sal. Jug. 53.7); (amnis) . . . fertur in arva furens cumulo. (Verg. A.
2.105); Merito abdicasti an immerito? (Sen. Con. 2.1.4); (sc. Delos) cingitur <V>
milia passuum, adsurgit Cynthio monte. (Plin. Nat. 4.66); Nox apud barbaros cantu
aut clamore, nostris per iram et minas acta. (Tac. Hist. 5.15.2); Immo et iudicium
frustra constituit deus, iniustitia utique puniturus . . . (Tert. Hermog. 11.3);
A subtype of this use of the ablative in manner adjuncts is constituted by the so-called
COGNATE ABLATIVE (sometimes called ‘figura etymologica’). Instances are found from
72
Early Latin onwards. Examples are (l)–(n). There are also instances in later periods.
This use of the ablative can be compared with the use of the accusative for the cognate
object (see § 4.10). Noun phrases are more common than unmodified nouns, but see
(n) and (o).
(l) Scis solere illanc aetatem tali ludo ludere.
(‘You know that people of that age normally play in such a manner of playing.’
(Pl. Mos. 1158)
(m) . . . vita ingenua (vitam ingenuam cj. Orelli) in beata civitate . . . viximus.73
(‘ . . . we lived an independent life in a happy community.’ Cic. Fam. 5.21.3)
(n) Nam hoc quidem edepol hau multo post luce lucebit.
(‘Yes, not much later it’ll be light with light.’ Pl. Cur. 182)
Supplement:
Noun phrases: Aequo mendicus atque ille opulentissumus / censetur censu ad
Accheruntem mortuos. (Pl. Trin. 493–4); Uno partu duos peperit simul. (Pl. Am.
1138); Laudant me maximis laudibus . . . (Cato frg. 15); . . . nosque malo solacio sed
nonnullo tamen consolamur quod . . . (Cic. Att. 4.6.1); Cura ut valeas meque ames
amore illo tuo singulari. (Cic. Fam. 15.20.3); . . . odissem te odio Vatiniano . . . (Catul.
14.3); Odisse plebem plus quam paterno odio. (Liv. 2.58.7); En ipsum imperatorem
clara voce vocare ad proelium et ire armatum ante prima signa. (Liv. 7.15.2); Aut
multum sudare maximeque frigido sudore. (Cels. 2.4.5); Bella res est mori sua
morte. (Sen. Ep. 69.6);
72
For the ablatival figura etymologica, see Rosén (1996b).
73
(Late) examples with vita can be found in Müller (1908: 20).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
74
For instances, see Hoppe (1903: 32 (=1985: 72)) and García de la Fuente (1990: 110).
75 76
Instances can be found in Müller (1908: 81–4). So, for example, Sz.: 40.
77
See Löfstedt (1967: 101–9), with a list of instances at pp. 103–6. For the Romance development, see
Hummel (2013).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Supplement:
Neuter singular: Nam eo magnum clamat. (Pl. Mil. 823); Qui volunt exclamare
maius . . . (Cic. Tusc. 2.56); Illa quam videtis / turpe incedere mimice ac moleste /
ridentem catuli ore Gallicani. (Catul. 42.7–9); . . . tibicen ubi canit Phryx curvo grave
calamo . . . (Catul. 63.22); Ecfugit horrendum stridens adducta sagitta . . . (Verg. A.
9.632); Talis equos alacer media inter proelia Turnus / fumantis sudore quatit,
miserabile caesis / hostibus insultans. (Verg. A. 12.337–9); Dulce ridentem Lalagen
amabo, / dulce loquentem. (Hor. Carm. 1.22.23–4); . . . et pavidum blandita ‘fer has,
fidissime, nostro’ / dixit . . . (Ov. Met. 9.569–70); Quis te tam lene fluentem / mo-
turum totas violenti gurgitis iras, / Nile, putet? (Luc. 10.315–17); BELLISSIMU <FU>-
TUERUNT (CIL IV.4884 (Pompeii)); Enimvero Tiberius torvus aut falsum renidens
vultu. (Tac. Ann. 4.60.2); . . . ego, quae dulce condiens et ollam et lectulum suave
quatere novi. (Apul. Met. 2.7.7); . . . martyrum animae placidum quiescunt . . . (Tert.
Scorp. 12.9); . . . implorans caelo iustitiam torvum renidens fundato pectore mansit
immobilis . . . (Amm. 14.9.6); . . . lugubre clamante praecone . . . (Amm. 28.2.13) . . . ibi
ore virgineo ad pudicitiae perpetranda naufragia Scyllaceum renidens libido blandi-
tur. (Hier. Ep. 14.6);
Neuter plural: Dux interea Gallorum . . . grandia ingrediens et manu telum recipro-
cans incedebat . . . (Quad. Hist. 12); . . . asper, acerba sonans, quo tota exterrita silvis /
diffugiunt armenta . . . (Verg. G. 3.149–50); Stabat acerba fremens ingentem nixus
in hastam / Aeneas . . . (Verg. A. 12.398–9); Nox ubi me thalamis ululantem et
acerba gementem / condidit . . . (Ov. Ep. 8.107–8); . . . ne vana labores . . . (Ov. Fast.
4.583); . . . quique ut catuli . . . incerta reptant . . . (Tert. Paen. 6.1);
The most common prepositional phrases functioning as manner adjuncts are those
with cum ‘with’, per ‘by’, and sine ‘without’. They are found from Early Latin onwards.
Those with cum mainly contain nouns denoting or implying either an action, a state
of mind, or an emotion of the subject of the clause (they are controller-oriented).
Typical examples are (a) and (b).78 They resemble secondary predicates (see
Chapter 21). They are also sometimes difficult to distinguish from accompanying
circumstance adjuncts (see §§ 10.74–10.76). They are found both juxtaposed (c) and
coordinated (a) with verb-oriented manner adjuncts. Cum-expressions are sometimes
used in explicit contrast with sine-expressions, as in (d).
(a) Agri reliquit ei non magnum modum / quo cum labore magno et misere
viveret.
(‘He did leave him a piece of land, not a big one, though, so that he could live on it
with great toil and miserably.’ Pl. Aul. 13–14)
78
For more instances, see TLL s.v. cum 1363.66ff, §§ III (‘de modo actionis’) A2 and 3, respectively (to
be used with restraint).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Manner adjuncts with per are mostly controller-oriented. Typical examples are
(e)–(g).79 With certain nouns the prepositional expression is (far) more common
than the bare ablative, from Early Latin onwards (for example, per iocum is more
common than ioco ‘in jest’, ‘jokingly’). Prepositional expressions with the accusative
neuter form of an adjective as in (h) are found from Virgil’s time onwards.
(e) Ego periurare me mavellem miliens / quam mi illum verba per deridiculum
dare.
(‘I’d rather perjure myself with solemn words a thousand times over than let him
deceive me in jest.’ Pl. Ps. 1057–8)
(f ) . . . utrum malit cervices <T.> Roscio dare an insutus in culleum [supplicium
parricidarum] per summum dedecus vitam amittere.
(‘ . . . whether he prefers to offer his throat to Titus Roscius or to be sewn up in a sack
and lose his life by a most infamous death.’ Cic. S. Rosc. 30)
(g) . . . deus explicat orbes / perque sinus crebros et magna volumina labens /
templa parentis init . . .
(‘ . . . the god unfolded his coils and, gliding on with many a sinuous curve and mighty
fold, entered his father’s temple . . . ’ Ov. Met. 15.720–2)
(h) Ceu septem surgens sedatis amnibus altus / per tacitum Ganges (sc. it) . . .
(‘Like Ganges, rising high in silence with his seven peaceful streams . . . ’ Verg. A.
9.30–1)
Cum and per are sometimes found with the same nouns in very similar contexts.
Compare (i)–(m) with the noun silentium ‘silence’. Note that Cicero (see (n)) has no
prepositional expressions and Plautus and Terence no bare ablatives. In light of this,
both individual variation and diachronic change seem to play a role in the choice
between the two types of expression
79
For instances, see TLL s.v. per 1153.29ff. , §§ IC (‘indicatur potius modus’) a and b, respectively. For
per + adj., ibidem 1154.46ff. For per in manner and instrument expressions, see Pinkster (1990).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Most sine expressions functioning as process adjuncts are interpreted as means and
instrument or accompanying circumstance adjuncts, but occasionally a manner
interpretation is possible, as in (o)–(r).
(o) Nimis haec res sine cura geritur.
(‘This business is being done far too carelessly.’ Pl. St. 310)
(p) Hortatus sum ut ea quae sciret sine timore indicaret.
(‘I urged him to reveal without fear what he knew.’ Cic. Catil. 3.8)
(q) . . . nec abiecte nec sine aliqua dignitate casum illum temporum et dolorem tuli.
(‘ . . . not abjectly nor without some dignity did I bear those chances of the times and
my own sorrow.’ Cic. Phil. 3.28)
(r) Equites Romanos severe curioseque nec sine moderatione recognovit . . .
(‘He revised the lists of Roman knights strictly and scrupulously, yet with due
moderation . . . ’ Suet. Cal. 16.2)
Apart from cum, per, and sine a few other prepositions are used in manner adjuncts,
mostly in a limited number of more or less idiomatic expressions. In the first place
there are prepositional expressions with modus ‘manner’, notably ad modum and in
modum + modifier.80 A few prepositions are used with a wide range of meanings and
are also found in manner adjuncts. They are (in alphabetical order): ad ‘after a pattern,
example, fashion’; de ‘from’, mainly after the Classical period, with nouns and ablative
neuter forms of adjectives; ex ‘from’, very rarely with noun phrases, but not uncom-
mon with ablative neuter forms of adjectives from Ovid and Livy onwards;81 in + abl.
80
For more instances, see TLL s.v. modus 1278.77ff.; OLD s.v. § 11.
81
For more instances, see TLL s.v. ex 1121.60ff. (noun phrases); 1124.3ff. (adjectives).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
‘in a certain form’ with noun phrases and ablative neuter forms of adjectives.82 A few
remarkable instances are given in the Supplement. The use of these prepositions
in manner adjuncts increases over the course of time, as with instrument adjuncts
(see § 10.55).
Supplement (in alphabetical order by preposition):
Plurumi ad illunc modum / periere pueri liberi Carthagine. (Pl. Poen. 988–9);
Propterea ornatus in novom incessi modum. (Pl. Am. 119);
. . . te . . . tamquam in schola prope ad Graecorum consuetudinem disputasse. (Cic.
de Orat. 2.13); Neque enim terras tibi, sed formam aliquam ad eximiam pulchritu-
dinem pictam videberis cernere. (Plin. Ep. 5.6.13); Sextus locus est, per quem
consulto et de industria factum demonstratur . . . (Cic. Inv. 1.102); Ea te expetere ex
opibus summis mei honoris gratia . . . (Pl. Mil. 620); Si ex facili spirat. (Cels.
2.3.1); . . . quae propter effectus claritatemque et magnitudinem in tanta subtilitate
nosci potuerunt . . . (Plin. Nat. 2.30); Ad quingentesimum <quinquagesimum> octa-
vum annum in promiscuo spectatum esse. (Liv. 34.54.5);
A wide variety of adverbs are used as manner adjuncts: (i) those productively derived
from adjectives with the suffixes -e and -(i)ter, as in (a) and (b); also comparative and
superlative forms, as in (c) and (d)—note that the comparative forms are formally not
distinct from neuter singular accusative forms of adjectives; (ii) non- or less product-
ive ones like clam ‘secretly’, palam ‘openly’, and adverbs in -tim and -tus; (iii) those
formed from pronominal stems like sic ‘so’, ita ‘so’, ut ‘how’, and qui ‘in what way’;83
(iv) forms of nominal and prepositional origin, as in (g) and (h). These types of
adverbs (with the exception of modo expressions) are in use in every period of Latin.84
(a) An vero non iusta causa est quor curratur celeriter . . .
(‘Well, isn’t there a just cause for running quickly . . . ’ Pl. Poen. 533)
(b) Nunc spectatores valete et nobis clare plaudite.
(‘Now spectators, farewell and applaud for us loudly.’ Pl. Men. 1162)
(c) Cum iste nihilo remissius atque etiam multo vehementius instaret
cotidie . . .
(‘When he abated nothing from his urgency, and indeed grew more violent . . . ’ Cic.
Ver. 4.76)
(d) Peregrinae sunt <et> lynces, quae clarissime quadripedum omnium cernunt.
(‘The lynx too is a foreign animal, and has keener sight than any other quadruped.’
Plin. Nat. 28.122)
82
For more instances, see TLL s.v. in 785.30ff. (noun phrases); 785.64ff. (adjectives).
83
For the various uses of interrogative ut and qui, see Ripoll (2014).
84
For the development of quo modo > quomodo, see Bodelot (2010b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
85
Torrego (2002b) deals with statim as a time within which adjunct.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In the course of time, such expressions with mente were treated as set phrases. The
order of the parts became fixed, mente always following the adjective, and they could
not be separated by other constituents. In such phrases mente lost its original meaning
and developed into an adverbial affix, replacing the old adverbial suffixes -e and -(i)ter.
That this process was well underway in the fourth century AD is proved by the fact that:
(i) mente expressions are used as a degree modifier of an adjective (see
§ 11.99);
(ii) mente expressions are found in clauses with an inanimate subject, as in (b).
(b) Audio Pontificis et populi voluntatem pari mente congruere.
(‘I hear that the will of the Pontifex and that of the people are in perfect harmony.’
Hier. Ep. 49.4)
86
For details of this development, see Karlsson (1981) and Norde (2009: 41–6). The literature about
this development is abundant: Löfstedt (1967: 94–6), Krenn (1993), Detges (1998), Hummel (2000, 2013),
and Bauer (2003). See also the historic section in TLL s.v. mens 737.69ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
The nouns occurring as norms adjuncts denote laws, customs, etc. Examples are (a)–(c).
(a) Multi more isto atque exemplo vivont . . .
(‘There are many people who live by that custom and example . . . ’ Pl. Bac. 540)
(b) Is qui omnium eruditorum testimonio totiusque iudicio Graeciae . . .
omnium fuit facile princeps . . .
(‘The person who on the evidence of all men of learning and the verdict of the whole
of Greece easily came out on top . . . ’ Cic. de Orat. 3.60)
(c) . . . quod legibus Haeduorum iis qui summum magistratum obtinerent
excedere ex finibus non liceret . . .
(‘ . . . inasmuch as it was not permitted by the laws of the Aedui for those who
exercised the highest office to leave the country . . . ’ Caes. Gal. 7.33.2)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Credo ego istoc exemplo tibi esse pereundum extra portam (Pl. Mil. 359); . . . dictus
filius tuos vostra voluntate . . . (Ter. Hau. 125); Neque eorum moribus turpius
quicquam aut inertius habetur quam ephippiis uti. (Caes. Gal. 4.2.4.);
Several prepositions are used in norm adjuncts, usually in combination with specific
nouns and noun phrases. The following are a few examples.
(a) . . . et hoc exigit ut ad legem suam quisque vivat . . .
(‘ . . . and it exacts of every man that he should live according to his own standards . . . ’
Sen. Ep. 20.2)
(b) Namque umeris de more habilem suspenderat arcum / venatrix . . .
(‘For from her shoulders she had slung the ready bow, a huntress in proper fashion.’
Verg. A. 1.318–19)
(c) Huncine erat aequom ex more illius an illum ex huius vivere?
(‘Should we expect this one to live by that one’s rules or that one by this one’s?’ Ter.
Hau. 203)
(d) Nam et frugi rectus est natura frux, at secundum consuetudinem dicimus ut
haec avis, haec ovis, sic haec frugis.
(‘For the nominative of frugi (fruit) is by nature frux, but by usage we say frugis, like
avis (bird), and ovis (sheep).’ Var. L. 9.76)
(e) . . . templum Iunonis Sospitae L. Iulius . . . de senatus sententia refecit ex
Caeciliae, Baliarici filiae, somnio.
(‘ . . . Lucius Julius, following the judgement of the senate, in accordance with a dream of
Caecilia, daughter of Balearicus, rebuilt the temple of Juno, the Saviour.’ Cic. Div. 1.4)
Supplement:
M. Flaccus . . . ex senatus sententia est interfectus . . . (Cic. Dom. 102); EX AUCTORITATE
/ IMP(PERATORIS) CAESARIS / VESPASIANI AUG(USTI) LOCA PUBLICA . . . T. SUEDIUS CLEMENS /
TRIBUNUS . . . / REI PUBLICAE / POMPEIANORUM RESTITUIT. (AE 2001, nr 796 (Pompeii));
The only adverb that can sometimes be interpreted as a norm adjunct is naturaliter
‘in accordance with the nature of things’, as in (a). There is one occurrence of rite
‘with all due formalities’ coordinated with more maiorum, (b), but both may be taken
as manner adjuncts.
(a) . . . nisi quod naturaliter audita visis laudamus libentius . . .
(‘ . . . unless we are more naturally inclined to praise what we have heard than what
has occurred before our eyes . . . ’ Vell. 2.92.5)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
In (e), a manner adjunct is juxtaposed to two means adjuncts: the actions of artificio
reperire and cura et cogitatione pertractare must be performed diligenter. It is rare for
two instrument adjuncts to occur in the same clause, but if the entities play a
sufficiently different role in the state of affairs, it appears to be allowed, as in (f )
and (g). In (f ) the action of rationem digitis computare is performed with the right
hand. Likewise in (g) per spongiam can be regarded as the external instrument applied
in performing the action of fovere vapore aquae calidae.
(e) Quare cum iudicatio et ea quae ad iudicationem oportet argumenta inveniri
diligenter erunt artificio reperta, cura et cogitatione pertractata, tum denique
ordinandae sunt ceterae partes orationis.
(‘Therefore when the point for decision and the arguments which must be devised for
the purpose of reaching a decision have been diligently discovered by the rules of art,
and studied with care and thought, then, and not till then, the other parts of that
oration are to be arranged in proper order.’ Cic. Inv. 1.19)
87
So Quirk et al. (1985: 482–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Human beings are rarely used as instrument adjuncts, but are acceptable if the social
position of the agent of the clause and the other human beings involved allow it.
Examples are (h) and (i). It is much more normal to find human beings expressed as
intermediaries in prepositional phrases with per ‘through the agency of ’, as in (j) and
(k). Some linguists posit a separate ‘intermediary’ function.
(h) Nam si negassent vim hominibus armatis esse factam, facile honestissimis
testibus in re perspicua tenerentur.
(‘For if they had denied that violence had been committed by armed men, they would
easily have been convicted in a plain case by most unimpeachable witnesses.’ Cic. Caec. 4)
(i) Interea ea legione quam secum habebat militibusque qui ex provincia
convenerant . . . fossam . . . perducit.
(‘Meanwhile he constructed a trench with the legion which he had with him and the
troops which had concentrated from the province.’ Caes. Gal. 1.8.1)
(j) Seu qui ipse ambissit seu per internuntium—
(‘Or if anyone should canvass himself, or through an intermediary—’ Pl. Am. 71)
(k) P. Clodium . . . per manus servulae servatum et eductum.
(‘P. Clodius was saved and led out by the hands of a servant girl.’ Cic. Att. 1.12.3)
The following categories of constituents may function as instrument adjuncts:
nouns and noun phrases in the ablative case
prepositional phrases
subordinate clauses (notably gerundial and gerundival clauses—see §§ 16.101,
16.107).
The use of noun phrases in the ablative as means and instrument adjuncts (the so-called
ablativus instrumenti) can be found in all periods of Latin, although especially with abstract
nouns a number of prepositional alternatives are available and increase in use. The nouns
that are used in means adjuncts are usually abstract nouns denoting or implying (human)
behaviour. The adjuncts are usually subject-oriented. Examples are (a)–(e).
(a) . . . vi summa ut quidque poterat rapiebat domum.
(‘ . . . (a relentless pimp) kept dragging to his home with greatest force whatever he
could.’ Pl. Mer. 45)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
88
For more instances in Vitruvius, see Wistrand (1933: 77–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
The preposition that is used to mark means adjuncts from Early Latin onwards is per
‘by means of ’. The nouns involved are abstract nouns, usually denoting or implying
human behaviour. Examples are (a) and (c)–(e). These expressions are used as
equivalents of bare case expressions, although no bare case alternative always exists.
Per dolum and dolo are both found with terminative states of affairs. See (a) and (b),
and note also the parallelism of fraude and the per-expressions in (a). Ex. (d)
illustrates that a strict classification of per-expressions is impossible; whereas per
vim can be taken as ‘in a violent manner’ on a par with temere and turbulente, it
can also be taken as ‘using violence’, as is the more likely interpretation in (c). This
interpretation seems excluded for per furorem in (d), which is rather a reason adjunct
(see § 10.84). The use of per in means expressions increased from Sallust and Livy
onwards, but the bare ablative remained the regular expression until long after the
period covered by this Syntax.89
(a) . . . artibus tribus tris demeritas dem laetitias de tribus / fraude partas, per
malitiam, per dolum et fallacias.
(‘I may give three delights by three arts won, from three men fraudulently gained,
through malice, through trickery and subterfuges.’ Pl. Ps. 705–5a)
(b) Sati’n med illi hodie scelesti capti ceperunt dolo?
(‘Haven’t those rascals who were caught caught me out well with their trick today?’
Pl. Capt. 653)
(c) Meamne hic Mnesilochus, Nicobuli filius, / per vim ut retineat mulierem?
(‘Should Mnesilochus, the son of Nicobulus, hold back my girl here by force?’ Pl. Bac.
842–3)
(d) . . . omnia contra leges moremque maiorum temere, turbulente, per vim, per
furorem esse gesta.
(‘ . . . the whole process had been carried out against the constitution and against
precedent, rashly, seditiously, forcefully, and angrily.’ Cic. Dom. 68)
(e) Alii credunt Hippocrati per calorem cibos concoqui.
(‘Others believe with Hippocrates that the food is cooked up by heat.’ Cels. pr. 20)
In a few exceptional instances the preposition cum is found with abstract nouns in an
expression that resembles the uses of per illustrated above. The most plausible ones
are (f ) and (g).90 Such prepositional phrases with cum are difficult to distinguish from
accompanying circumstance adjuncts.
89
For the history of the use of prepositions in instrument expressions, see Beckmann (1963) and
Adams (2013: 294–320); for the use of prepositional instruments in the Vulgata, see de la Villa (1998).
90
See Beckmann (1963: 38).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The use of per with concrete nouns denoting physical objects that can be used to
perform certain actions is restricted to certain lexical classes, such as letters and other
objects used in conveying messages (h), fluids used in the production of certain
materials (i), and objects used in performing an activity that is itself instrumental in
another action (j). In all these cases the ‘intermediary’ meaning of per is still relevant.
The use of per in real instrument adjuncts, as in (k), is rare and mostly late. The use of
cum in true instrument adjuncts is even more rare and definitely late: the first attested
instances date from the end of the fourth century AD, as in (l). The use of per and cum
seems to be more frequent with pronouns referring to instruments, probably because
they are more ambiguous.
(h) . . . qui de Dionis iniuriis gravissime per litteras cum Verre questus es?
(‘ . . . you who in a letter complained most bitterly to Verres about Dio’s wrongs?’ Cic.
Ver. 2.24)
(i) Cum opus est, teritur unus pastillus et immittitur per aquam decoctam
rosam et lentem habentem, si febricitabunt. Sin minus, ex vino eadem
incocta habente.
(‘When there is a need, one grinds one pastille and administers it with water
containing rose and lentil that has been boiled, if they are feverish. If not (administer
it) with wine having the same ingredients but unboiled.’ Larg. 114)
(j) Ecce suburbanae templum nemorale Dianae / partaque per gladios regna
nocente manu.
(‘Lo, hard by the city is Diana’s woodland shrine and the realm won by the sword
and guilty hand.’ Ov. Ars 1.259–60—cf. . . . partaque bellando gloria dulcis erat.
Ov. Ep. 3.122)
(k) . . . uno loco adducta per hamulum vena praeciditur.
(‘ . . . at one place the vein is drawn forward by the hook and cut away.’ Cels. 7.31)
(l) . . . caede caudam cum tabula aliqua ponderosa.
(‘Cut the tail with some heavy plank . . . ’ Veget. Mul. 1.27.2)
In (m) and (n) the expressions with per and cum are not instruments.
(m) Si huius terrae scribas, i littera sit extrema, si huic terrae, per e scribendum est.
(‘If you should write of this land, “i” would be the last letter, if you should
write to this land, it must be written by “e”. ’ Nigid. gram. 11)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Due to the spread of prepositions outside their original field, a number of them,
including those mentioned in the section on manner adjuncts (§ 10.46), came to be
used in means and instrument adjuncts as well. The widening of the range of
meanings of these prepositions is manifest already in Augustan poetry and elsewhere,
although instances of purely instrumental use as an alternative to e.g. gladio occido
aliquem ‘I kill someone with a sword’ are found only much later.
Instances of the purely instrumental use of ab, based on its use in separative
expressions, are found in Late Latin. The emergence of the rare instrumental use of
ad (normally meaning ‘reacting to’) can be seen in (o).91 A purely instrumental
example is (p). An early purely instrumental use of de, based on its use in separative
expressions as in the case of ab, is found in Apuleius (q), but its productive use is later.
Ex. (r) shows that it had become common enough by Augustine’s time to be used in a
gloss. It was much more normal than the other separative prepositions ab and ex—
which both disappeared, ex faster than ab—in this context.92 Ex. (s) seems to be the
first instance of instrumental ex.93 Ex is normal for ingredients that are used for the
production of entities, as in (t). The use of in in instrument adjuncts, as in (u), is
typical of biblical Latin and hence of Christian authors; in the Septuagint and in the
New Testament Greek K was used to translate the Hebrew locative preposition be,
which is also used with an instrumental meaning. A few other instructive examples
are given below in the Supplement.
(o) . . . reseret modo Delia postes / et vocet ad digiti me taciturna sonum.
(‘ . . . if Delia but unbars the door and summons me wordlessly with a finger’s tap.’
Tib. 1.2.33–4)
(p) Iuniores . . . ad omne genus exercebantur armorum, veteres autem . . .
exercebantur armis.
(‘The younger soldiers . . . used to practise with every type of weapon. However, the
veterans . . . used to practise in arms.’ Veget. Mil. 2.23.2—NB: parallelism with abl. armis)
(q) Sed ubi . . . advertit . . . inversa vite de vastiore nodulo cerebrum suum
diffindere . . .
(‘But when he perceived that he had reversed his staff and was splitting his skull with
the thicker end . . . ’ Apul. Met. 9.40.1)
(r) ‘In qua (sc. virga) percussisti’ dixit pro eo, quod dicimus ‘de qua percussisti’ . . .
(‘ “In which branch you struck” he said for this, which we say as “from which branch
you struck” . . . ’ August. Locut. Hept. 2.91)
91
For the instrumental use of ad in Vegetius and other texts, see Adams (2013: 308–14).
92 93
See the frequency table in Beckmann (1963: 70). See Skutsch ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(s) . . . si forte <feras> ex nare sagaci / sensit, voce sua nictit ululatque ibi acute.
(‘ . . . if by chance she has caught scent of wild quarry with her keen-scented nose,
then she lifts her voice in a whimper and straightway loudly gives tongue.’ Enn. Ann.
341–2V=333–4S)
(t) Item supra cameram materies ex calce et harena mixta subinde inducatur . . .
(‘Further, on the upper surface of the arch, mortar, mixed with lime and sand, is to be
spread . . . ’ Vitr. 7.3.2)
(u) Et virgam, in qua percussisti flumen, accipe in manu tua.
(‘And take the branch, with which you struck the river, into your hand.’ August.
Locut. Hept. 2.91 = Vet. Lat. Exod. 17.5)
Supplement (in alphabetical order by preposition):
Eodem momento quo friguntur et levantur ab aceto calido perfunduntur. (Apic.
11); . . . ab his te gladiis transfodere voluemus. (Greg. T. Hist. 8.29 fin.); Ficatum
praecidis ad cannam, infundis in liquamine. (Apic. 7.3.2); . . . episcopus sedens de
manibus suis summitates de ligno sancto premet . . . (Pereg. 37.2); O si occidas eos de
gladio bis acuto et non sint hostes eius. (August. Conf. 12.14.17); Battes de tabella
aliqua belle ponderosa diu . . . (Mulom. Chir. 11.30); . . . cum solarium vel descriptum
vel ex aqua contemplere . . . (Cic. N.D. 2.87); . . . id membrum perfricandum est ex
oleo et sale et nitro. (Cels. 8.10.7.o); LUCILLA EX CORPORE LUCRUM FACIEBAT (CIL
IV.1948 (Pompeii)); . . . corpora exanimata iam mortemve simulantia e cauterio
probat. (Tert. Nat. 1.10.47); Et nos regem et sacerdotem et pro nobis mortuum
Christum agnoscentes atque laudantes, tanquam in auro et thure et myrrha honora-
vimus. (August. Serm. 202.4);
Some manner adverbs imply the application of instruments, as cruciabiliter in (v).
It is also used in B. Afr. 46.2 with interfecti. Another example is irrumabiliter in a
Pompeiian inscription (CIL IV.1931).94
(v) Nisi effecero, cruciabiliter / carnufex me accipito.
(‘Unless I accomplish it, the hangman can give me a reception with torture.’
Pl. Ps. 950–50a)
94
Adams (2005b: 88–9) has a few more.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Instances of nouns and noun phrases in the ablative (the so-called ablativus pretii) are
given above in the introductory section. Most nouns involved have a commercial
meaning, as in (a), but see (b). Idiomatic is gratis (and gratiis) ‘for no reward but
thanks’, ‘gratis’.
(a) Tuo periclo sexaginta haec datur argenti minis.
(‘You can have her at your own risk for sixty silver minas.’ Pl. Per. 665)
(b) Locavit autem omnem (sc. agrum) frumento.
(‘He leased them all in return for grain.’ Liv. 27.3.1)
Supplement:
Em istuc verbum vile est viginti minis. (Pl. Mos. 297); Habe (sc. puellam) centum
minis. (Pl. Pers. 662); Pol mihi / eo pretio empti fuerant olim. (Pl. Mos. 820–1); Qui
cena poscit? Ecqui poscit prandio? (Pl. St. 222); Velaturam facere etiam nunc
dicuntur qui id mercede faciunt. (Var. L. 5.44); Addicitur opus HS DLX milibus . . .
(Cic. Ver. 1.144); Nunc demum intellego P. Clodi insulam esse venalem, cuius hic in
aediculis habitat decem, ut opinor, milibus. (Cic. Cael. 17); Nam Eriphyla auro viri
vitam vendidit. (Cic. Inv. 1.94); Habitent gratis in alieno. (Cic. Off. 2.83); . . . avum
tuum L. Paulum qui morte luit collegae in Cannensi ignominia temeritatem . . . (Cic.
Sen. 75); . . . qui vita bene credat emi quo tendis honorem. (Verg. A. 9.206); . . . pacti
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
95
On this inscription, see Solin (2013).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
The following four substantivally used neuter singular adjectives when functioning as
price adjuncts are only used in the genitive (the so-called genetivus pretii): quanti ‘for
how much’, tanti ‘for so much’, pluris ‘for more’, and minoris ‘for less’. Note in (a) the
co-occurrence of two price adjuncts.
(a) Quanti emi potest minimo?
(‘What’s the lowest price she can be bought for?’ Pl. Epid. 295)
(b) Multo minoris vendidit non modo quam tu . . .
(‘Not only did he sell these tithes for much less than you . . . ’ Cic. Ver. 3.43)
Supplement:
Qui datur, tanti indica. (Pl. Per. 661); Emit? Perii hercle. Quanti? # Viginti minis.
(Ter. Eu. 984); Nam P. Fabius nuper emit agrum . . . sane magno, dimidio fere pluris
incultum exustis villis omnibus quam quanti integrum atque ornatissimum carissi-
mis pretiis ipse Claudius <emerat>. (Cic. Tul. 14); . . . tamen hoc tibi praeclarum
putabis, te pluris quod non licebat, quam ceteros quod oportebat vendidisse? (Cic.
Ver. 3.40); Tu autem . . . tanti ista quattuor aut quinque sumpsisti. (Cic. Fam. 7.23.2);
Tanti enim pepigerat. (Liv. 38.24.7);
Exceptional: Tu mihi vitio dabis . . . quod paucioris habeo . . . (Apul. Apol. 21);
96
See Woytek ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Huiusce, inquam, pomarii summa sacra via, ubi poma veneunt contra
aurum, imago.
(‘The top of the Via Sacra I say, where fruit is being sold for gold, is a very
picture of this orchard.’ Var. R. 1.2.10)
Appendix: Price adjuncts in the accusative are a late phenomenon, though attention
has been drawn to a few earlier instances, the first of which is (e). However, this
instance may be a quantity and not a price expression.97
(e) Vendidit enim vinum quantum ipse voluit.
(‘For he sold his wine for any price he then wanted.’ Petr. 43.4)
Supplement:
ERGO <M>ERCA MINORE PRETIUM, ROGO . . . (CEL 114.17 (Karanis, AD 115)); FRONTINIUM
IULIUM AUDIO MAGNO LICE/RE PRO CORIATIONE QUEM HIC COMPARAVIT (DENARIOS) QUINOS.
(CEL appendix Vindol. Ł.38–41 (Vindolanda, AD 85–120));98 Apollonius ait: ‘Et
quantum me proscripsit (Hist. Apoll. 8RA 17 (NB: RB has quanti));
There are three types of QUANTITY ADJUNCT, which are dealt with separately.
97
So Parker (1994). Evidence to the contrary from the letters of Claudius Terentianus and the Tab.
Vindol. can be found in Adams (1977: 40–2, 1995a: 116, 2003a: 20). Late examples are discussed by
Löfstedt (1936: 170–4) and Norberg (1943: 103ff.).
98
According to Adams (2013: 321–3), these instances can better be explained as ablatives.
99
For the difficulty of distinguishing both categories, see Pinkster (1972: 50) and Ripoll (2007: 246–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
such as quantum ‘how much’, tantum ‘so much’, plus ‘more’, minus ‘less’, multum
‘very much’, ‘a lot’, paulum ‘(a) little’, pusillum ‘a tiny amount’, and exiguum ‘a small
amount’; the adverbs nihil ‘nothing’ and parum ‘too little’ are used as well. Descriptive
adjectives are rarely used in this way (but infinitum in ex. (d) may be an example),
although they are used with some frequency as modifiers of adjectives and adverbs
(see § 11.94). For a similar use of the pronoun aliquid, see (e).
(a) Menaechme, amare ait te multum Erotium . . .
(‘Menaechmus, Erotium says she very much likes you . . . ’ Pl. Men. 524)
(b) Nam bonum est pauxillum amare sane, insane non bonum est.
(‘Yes, it is good to be a little bit in love sanely, but being insanely in love is not good.’
Pl. Cur. 176)
(c) Nugae sunt. Si stimulos pugnis caedis, manibus plus dolet.
(‘That’s nonsense! If you beat cattle prods with your fists, your hands hurt more.’ Pl.
Truc. 768)
(d) . . . infinitum interest, utrum . . .
(‘ . . . it makes an infinite difference whether . . . ’ Sen. Ben. 2.6.1)
(e) . . . si in me aliquid offendistis . . .
(‘ . . . if you have taken offence at anything that I have done . . . ’ Cic. Mil. 99)
Supplement (in alphabetical order by quantity expression):
. . . videsne . . . ut etiam contendant et elaborent, si efficere possint ut aut non
appareat corporis vitium aut quam minimum appareat? (Cic. Fin. 5.46); L. etiam
Cotta . . . dicendi non ita multum laude processerat . . . (Cic. Brut. 137); Atque ille iis
semper utebatur, nos, nisi dum a populo auspicia accepta habemus, quam multum
iis utimur? (Cic. Div. 2.76); Nunc et Scaevola, quoniam in Tusculanum ire constituit,
paululum requiescet . . . (Cic. de Orat. 1.265); Agedum tu, Artamo, / forem hanc
pauxillulum aperi: placide, ne crepa. (Pl. Bac. 832–3); . . . pauxillum differt a cavilli-
bus. (Pl. Truc. 686); . . . longi duo validi asseres ex inferiore parte in terra defigeban-
tur, distantes inter se paulo plus quam quanta be<luae> latitudo est. (Liv. 44.5.3);
Expecta me pusillum, et de domo fiet numeratio. (Sen. Ep. 26.8);
Depending on the context, some of these quantity expressions may be interpreted as
frequency or duration expressions, as in (f)—note coordination with saepius. The
difference between the two expressions becomes apparent from (g ).100
100
See Ripoll (2007: 223ff.).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
10.64 The use of adverbs derived from descriptive adjectives as quantity adjuncts
Adverbs derived from descriptive adjectives normally function as manner adjuncts
(see § 10.47). However, with certain verbs, adverbs of this sort, when they indicate
quantity or size, may also function as quantity adjuncts. Examples are modice—note,
however, coordination with sensim—and abundantissime in (a) and largiter in (b), to
be compared with their use as manner adjuncts in (c)–(d).
(a) (sc. Nilus) Incipit crescere luna nova quaecumque post solstitium est, sensim
modiceque cancrum sole transeunte, abundantissime autem leonem . . .
(‘The Nile begins to rise at the next new moon after midsummer, the rise being
gradual and moderate while the sun is passing through the Crab and at its greatest
height when it is in the Lion . . . ’ Plin. Nat. 5.57)
(b) . . . tum color et facies hominum distare videntur / largiter et morbi gene-
ratim saecla tenere.
(‘ . . . so the colour and aspect of men are seen to be widely different and diseases to
possess the nations after their kind.’ Lucr. 6.1112–13)
(c) Qui distincte, qui explicate, qui abundanter, qui inluminate et rebus et verbis
dicunt . . .
(‘It is those whose speeches are clear, explicit and full, perspicuous in matter and in
language . . . ’ Cic. de Orat. 3.53)
(d) Stulte datum / reor, peccatum largiter.
(‘I think that it was stupid of you to pay and that you made a big mistake.’ Pl. Epid.
484–5)
Supplement:
Neque solum domi, sed etiam apud finitimas civitates largiter posse . . . (Caes. Gal.
1.18.4); . . . minus apud eum quam ego cum possent, immoderate sunt abusi. (Mat.
Fam. 11.28.2);
101
See Skutsch ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
by the fact they are also used with adjectives and adverbs (see § 11.98). Instances are
found until Late Latin, especially in Ammianus.102 Some of these expressions exist
alongside expressions with the copula expressed or understood. The subordinate
clause then normally has the subjunctive, as emerserint in (c) (see §§ 7.133–7.137
for the mood in indirect questions).
(a) O Phaedria, incredibile[st] quantum erum ante eo sapientia.
(‘Phaedria, you wouldn’t believe how much I surpass my master in wisdom.’
Ter. Ph. 247)
(b) Huic generi orationis aspergentur etiam sales qui in dicendo nimium quan-
tum valent.
(‘A speech of this kind should also be sprinkled with the salt of pleasantry, which
plays a rare great part in speaking.’ Cic. Orat. 87)
(c) Hac autem re incredibile est quantum civitates emerserint.
(‘You would hardly believe how much that has helped to drag the communities out of
the mire.’ Cic. Att. 6.2.4)
Supplement:
Immane quantum animi exarsere. (Sal. Hist. 2.44); Id mirum quantum profuit ad
concordiam civitatis iungendosque patribus plebis animos. (Liv. 2.1.11); . . .
inmensum quantum hinc Oceano, illinc Hiberico mari comprimentibus. (Plin.
Nat. 4.110); . . . mirabile quantum / gaudebat (Sil. 6.620–1); Et Flavianus exercitus
immane quantum <aucto> animo exitium Valentis ut finem belli accepit. (Tac. Hist.
3.62.1); . . . incredibile quantum procedit ad cumulum dilectionis. (Ambr. Off. 2.29);
. . . quorum varietate immane quantum exuberant. (Amm. 23.6.50); Huius nece
incredibile quantum laetitia gaudioque senatus ac plebes exsultaverint. (Aur. Vict.
Caes. 40.24);
DEGREE ADJUNCTS specify the degree or intensity with which a gradable event takes
place in relation to a certain subjective norm of the speaker. Examples are (a)–(c).
Many of the expressions involved are also used as modifiers of gradable adjectives,
adverbs, and prepositional phrases. See the examples (d)–(f ).103 Sometimes they
modify a particular constituent of another type, as in (g), where ex toto modifies
nihil. Many manner expressions, like mire ‘amazingly’ in (h), can by implication be
interpreted as indicating the intensity with which an event takes place, and thus serve
as a rich source of hyperbolic expressions. Indeed, many more or less specialized
degree expressions have developed from manner expressions (plane ‘absolutely’, sane
‘very’, and valde (valide) ‘intensely’, greatly’ are good examples). Degree expressions
102
For the development of such idioms, see Fruyt (1990b). For examples, see Bodelot (1990: 79–85).
Late Latin examples are in Bodelot (1999).
103
The examples of valde are taken from the OLD.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
104
The best discussion of the use of adverbs in exaggerating and hyperbolic expressions is Hofmann
(1951: 70–9, 192–3). Not all of these can be described as degree words. Most of the adverbs mentioned in
this section are taken from this study.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
10.67 The use of noun phrases and prepositional phrases as degree adjuncts
A few noun phrases and prepositional phrases have developed into degree expres-
sions, notably magnopere ‘greatly’ (< magno opere and opere magno), tantopere
(< tanto opere and opere tanto) ‘to such a great degree’, in maiorem modum ‘to a
greater (than usual) degree’, ‘especially’, and magna ex parte ‘to a large extent’ (see (b)
in § 10.66), ex toto ‘totally’.
(a) Vos omnis opere magno esse oratos volo, / benigne ut operam detis ad
nostrum gregem.
(‘I want to appeal earnestly to all of you to pay benevolent attention to our troupe.’ Pl.
Cas. 21–2)
(b) Unum reprehendo, quod otium <e> Gallia nuntiari non magno opere
gaudet.
(‘I have only one criticism, he is not overhappy at the news of peace in Gaul.’ Cic. Att.
1.20.5)
(c) Quapropter a te peto in maiorem modum ut me absentem diligas atque
defendas.
(‘Accordingly I would earnestly solicit your regard and protection during my
absence.’ Cic. Fam. 15.7)
(d) Atque in aliis quidem ex toto desinit (sc. scabies), in aliis vero certo tempore
anni revertitur.
(‘And whilst in some persons the scabies vanishes completely, in others it returns at a
definite time of the year.’ Cels. 5.28.16A)
When descriptive adverbs are used as adjuncts with gradable verbs they may be
understood as degree rather than as manner expressions. They belong especially to
certain semantic classes, e.g. adverbs meaning (i) ‘strong’, ‘energetic’, heavy’, etc.
(valde, developed from valide ‘vigorously’, is a good example of a degree expression
that developed from a manner expression), as in (i) and (j); (ii) ‘good’ and ‘bad’, as in
(k); and (iii) ‘amazing’, as in (l).
(i) . . . te . . . moderatissimum fuisse vehementissime gaudeo . . .
(‘ . . . it gives me the warmest pleasure that you have taken so temperate an attitude
. . . ’ Cic. Att. 1.20.1)
(j) . . . graviter se vulnerat occursuque satellitum inhibitus perfertur in regiam.
(‘ . . . gave himself a severe wounding . . . , and being stopped by the guards who ran
up, was brought into the royal apartment.’ Curt. 6.7.29)
(k) Egregie probo fore ut dum agamus ‘› º ‰æÆE ’ obrepat.
(‘I particularly like your remark that before we know it “the seasonable sailing” may
steal upon us.’ Cic. Att. 9.7.5)
105
For paene, see Bertocchi (1996b).
106
For further examples, see OLD s.v. quam § 3a and Bodelot (1990: 79–85).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
107
Quoted from Hüttemann (2010: 186).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
in this meaning as monovalent. This analysis is less obvious in the case of verbs that
only govern a dative argument, such as nubo ‘to get married to’ and parco ‘to spare’,
although some grammars take these as beneficiary as well (see also § 12.7).
A second class is formed by verbs of keeping off and defending such as arceo
‘to keep away’, (de)pello ‘to drive off ’, and prohibeo ‘to keep away’. Examples are
(e) and (f ). In this Syntax these dative expressions are regarded as beneficiary
adjuncts.
(e) Heus, ecquis hic est, maxumam qui his iniuriam / foribus defendat?
(‘Hey! Is anyone here to keep off the greatest assault from these doors?’ Pl.
Mos. 899–900)
(f ) Opsecro, parentes ne meos mi prohibeas.
(‘Please, don’t keep my parents from me.’ Pl. Cur. 605)
Appendix: From Early Latin onwards, reflexive pronouns are used as beneficiary
adjuncts with a less overt beneficiary meaning, for example in combination with
verbs denoting possession, taking into possession, demanding, or protecting one’s
possession like habeo ‘to have’, sumo ‘to take’, postulo ‘to demand’, and vindico ‘to lay
claim to’. This use of the reflexives is sometimes called ‘pleonastic’. It is less frequent
in Classical Latin, but through analogy it spreads to other verbs in Late Latin,
resulting in reflexive verbs (where the reflexive is no longer optional) in the Romance
languages. Early examples are (g)–(j).108
Dative reflexives are also found in Late Latin texts, mainly of a lower register, with
verbs of movement and verbs of sitting, as in (k). Originally these reflexives
were beneficiary adjuncts as well. This use of the reflexive also resulted in reflexive
verbs in the Romance languages. For verbs of this class with accusative se, see
§ 5.29.
108
For the history of these reflexive beneficiaries, see Dahlén (1964, 1977) and Adams (2013: 346–62);
for a summary of earlier studies, see Sz.: 294.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Although, as previously stated, the dative of interest most frequently refers to animate
entities, it is also used for inanimate entities, including abstract concepts, as is shown
by examples (n)–(q).111 Many scholars use the term dative of purpose (the so-called
dativus finalis) for this use of the dative of abstract nouns. The main semantic
difference that is assumed between the dative of interest and the dative of purpose
is that the latter signals an envisaged future result.112 A remarkable instance of a
beneficiary adjunct is (s), where the question is answered by a purpose ut-clause. (For
a survey of the various constructions for which the terms dative of purpose and
predicative dative are used, see Chapter 21.)
(n) Vita, quae fato debetur, saluti patriae potissimum solvatur.
(‘Above all, let me spend my life, which I owe to fate, for the salvation of my country.’
Rhet. Her. 4.55)
(o) Quid enim revocante et receptui canente senatu properet dimicare?
(‘For why, when the Senate’s trumpet sounds for recall, should it hasten to fight?’ Cic.
Phil. 12.9)
(p) Libertati enim ea praesidia petitis, non licentiae ad inpugnandos alios.
(‘For you seek in them guarantees of liberty, not of a licence to make attacks on
others.’ Liv. 3.53.6)
109
So Bulhart (1957: xvii).
110
Dahlén (1964: 164–5) takes sibi with dubitat, but this is incorrect. See Traina (1981).
111
The best diachronic survey is Landgraf (1893). Roby (1882: xxvff.) has a wealth of examples (which
he calls ‘predicative datives’).
112
So Önnerfors (1956: 15–17) and Sz.: 98.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Supplement:
Gustui dat dulce. Amarum ad satietatem usque oggerit. (Pl. Cist. 70); Holeris,
asparagis, lignis, aqua, itinere, actu domini usioni recipitur. (Cato Agr. 149.2);
Vinum faecatum sic facito. Fiscinas olearias campanicas duas illae rei habeto.
(Cato Agr. 153); Sed qui venationem voluptati suae claudunt contenti sunt . . .
munire vivarium semperque de manu cibos et aquam praebere. (Col. 9.1.1); Ante
brumam autem VII diebus totidemque post eam sternitur mare alcyonum feturae,
unde nomen ii dies traxere. (Plin. Nat. 2.125); Radix unguentis expetitur. (Plin. Nat.
12.110); Radix decoquitur in posca dolori dentium. (Plin. Nat. 27.29); Datur et equi
lingua inveterata ex vino praesentaneo medicamento . . . (Plin. Nat. 28.200—NB: with an
attribute); . . . alibi divitiis foditur . . . alibi deliciis . . . alibi temeritati . . . (Plin. Nat. 33.1 . . .
placuitque sic eam tabulam posteris tradi omnium quidem, sed artificum praecipuo
miraculo. (Plin. Nat. 35.83); Non vitae, sed scholae discimus. (Sen. Ep. 106.12);
In (t), Szantyr (: 98) assumes coordination of a beneficiary and a purpose adjunct, but
usurae meae can be taken as a beneficiary adjunct (as the fact that they are coord-
inated suggests).
Supplement:
. . . pugnaeque accinge pedestri. (Verg. A. 11.707); En, ego, quod vivis, poenae
crucianda reservor. (Ov. Ep. 14.119); . . . dum centurionem cognomento Sirpicum
illa morti deposcit, quintadecimani tuentur . . . (Tac. Ann. 1.23.5); Cum huic negotio
accingerentur apostoli . . . (Tert. Marc. 3.22.2); Quibus cognitis Galli, cum et ipsi se
proelio pararent, in auspicia pugnae hostias caedunt . . . (Iust. 26.2.2);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The use of the preposition pro ‘on behalf of ’ in beneficiary adjuncts is common from
Early Latin onwards. Examples are (a)–(c).113 Note in (b) juxtaposition of the pro
phrase and the recipient argument Dianae. The antonym of pro is contra ‘against’, as
in (d). For the use of ad, see (e).114
(a) Sed tu, qui pro tam corrupto dicis causam filio, . . .
(‘But you, who plea for a son so corrupted . . . ’ Pl. Bac. 420)
(b) POUBLILIA·TURPILIA·CN.·UXOR / HOCE·SEIGNUM·PRO·CN.·FILIOD / DIANAI·DONUM·
DEDIT.
(‘Publilia Turpilia, wife of Gnaeus, gave this statue to Diana on behalf of her son
Gnaeus.’ CIL I2.42 (Nemi, c.150 BC))
(c) Nullum bellum suscipi a civitate optima nisi aut pro fide aut pro salute.
(‘A war is never undertaken by the ideal State, except in defence of its honour or its
safety.’ Cic. Rep. 3.34)
(d) Phthisicis cyathi mensura quidam dederunt et contra veterem tussim.
(‘Some authorities have prescribed it in doses of one cyathus for people suffering
from consumption and for chronic cough.’ Plin. Nat. 24.39)
(e) Ex posca coquitur ad guttura.
(‘They are boiled in vinegar and water for sore throats.’ Plin. Nat. 22.145)
113
For the various interpretations of pro, see de la Villa (1995).
114
For the use of ad with patients and medicines, see Langslow (2000: 367–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
However, associative adjuncts are also found in other constellations: for example, to
denote the accompaniment of the second argument (animate or inanimate) of a verb,
both the object in active clauses, as in (e)–(g), and the subject in passive clauses, as in
(h). Instances abound from Early Latin onwards. It is sometimes difficult to decide
whether the cum-phrase is a modifier (see § 11.67) or an adjunct. Finally, although
they are rare, there are associative adjuncts that denote the accompaniment of an
indirect object, as in (i) and (j).115
(e) Argentum accipias, cum illo mittas virginem.
(‘Take your money and send the girl off with that man.’ Pl. Cur. 457)
(f ) Symbolum hunc ferat lenoni cum quinque argenti minis . . .
(‘He must bring this token with the five minas to the pimp . . . ’ Pl. Ps. 753)
(g) Eo circiter hominum numero sedecim milia expedita cum omni equitatu
Ariovistus misit . . .
(‘Towards it Ariovistus sent some sixteen thousand light-armed troops with all the
horse . . . ’ Caes. Gal. 1.49.3)
(h) Eae sunt surruptae cum nutrice parvolae.
(‘They were kidnapped with their nurse when they were little.’ Pl. Poen. 1346)
(i) Pompeiae cum Fausti liberis Caesar incolumitatem suaque omnia concessit.
(‘To Pompeia along with the children of Faustus, Caesar gave safety and all their
property.’ B. Afr. 95.3)
(j) ‘Illi cum illo habitationem lego’: perinde est ac si ita ‘illi et illi’ legasset.
(‘ “I bequeath the place to him with him” is the same as if he had bequeathed in the
following way: “to him and him”. ’ Alf. dig. 33.2.40)
115
For examples, see TLL s.v. cum 1347.9f.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
. . . domi suae defossam multis cum opibus / aulam invenit . . . (Pl. Aul. arg. I.2–3);
Domi domitus sum usque cum caris meis. (Pl. Men. 105); Id cum melle aut cum
mulso apponito. (Cato Agr. 80); . . . alter ulteriorem Galliam decernit cum Syria, alter
citeriorem. (Cic. Prov. 36); Id bellum cum causis et eventibus (etenim longius
provectum est) mox memorabimus. (Tac. Hist. 3.46.1);
For associative arguments with verbs like pugno ‘to fight’ and iungo ‘to join’ see
§ 4.38. For accompanying circumstance adjuncts, see §§ 10.74–10.76. For weapons
carried, clothes worn, and mental properties when performing a certain action, see
Chapter 21 on prepositional secondary predicates.
Appendix: Sometimes the combination of an associative expression with cum and a
subject constituent behaves as a plural constituent, as becomes apparent from the
verb ending. An example is (k). (For further examples, see Chapter 19 on comitative
coordination. See also (j), above.)
The types of adjunct in the following sections differ from process adjuncts in that they
do not specify details of the state of affairs of the clause itself, but rather indicate the
relation of the state of affairs of that clause to another state of affairs in terms of cause
and effect or otherwise.
116
Bennett’s term (1914: II.301). He deals with them under the heading ‘manner’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
The nouns used in accompanying circumstance adjuncts are abstract and verbal
nouns, many of which might also be used in manner adjuncts, e.g. laetitia in (c), if
it were to refer to the joy of a subject constituent rather than that of outsiders.
(a) Tuo ego istaec igitur dicam illi periculo.
(‘So I’ll tell him about this at your own risk.’ Pl. Bac. 599)
(b) Plane illuc te ire nisi tuo magno commodo nolo.
(‘Really I don’t want you to go there unless it’s to your great benefit.’ Cic. Att. 13.27.2)
(c) His senatus consultis perfectis dimisso senatu decemviri prodeunt in con-
tionem abdicantque se magistratu ingenti hominum laetitia.
(‘Having so decreed, the senate adjourned and the decemvirs went before the people
and laid down their office, to the great delight of all.’ Liv. 3.54.6)
Supplement:
. . . primum abs te hoc bona venia peto . . . (Ter. Ph. 378); Pace tua dixerim. (Cic. Leg.
3.29); Quinti fratris domus primo fracta coniectu lapidum ex area nostra, deinde
inflammata iussu Clodi inspectante urbe coniectis ignibus, magna querela et gemitu
non dicam bonorum, qui nescio an <n>ulli sint, sed plane hominum omnium. (Cic.
Att. 4.3.2); Quis enim audeat violare proposita cruce aut saxo, praesertim tantis
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The prepositions used to mark accompanying circumstance adjuncts are cum ‘with’
and sine ‘without’. They are found in this usage from Early Latin onwards.118
(a) Flagitio cum maiore post reddes tamen.
(‘You’ll return it later nevertheless, under greater and louder demands.’ Pl. Epid. 516)
(b) . . . ut illo itinere veniret Lampsacum cum magna calamitate et prope per-
nicie civitatis.
(‘ . . . he happened in the course of his journey to arrive at Lampsacum, with terrible
and almost ruinous consequences for the community.’ Cic. Ver. 1.63)
(c) . . . testamenta, quibus Trimalchio cum elogio exheredabatur.
(‘ . . . wills, in which Trimalchio was cut out with flattering terms.’ Petr. 53.9)
(d) Et ‘est’ autem et ‘erat’ et ‘fuit’ plerumque absunt cum elegantia sine detri-
mento sententiae.
(‘And in fact “is” and “was” and “has been” are often omitted with elegant effect and
without any loss of meaning.’ Gel. 5.8.7)
Supplement:
Exorditur magna cum expectatione veterem histrionis exponere societatem. (Cic.
Q. Rosc. 27); . . . ne per eam provinciam minore cum dignitate iter faceret . . . (B. Alex.
64.2); . . . L. Furio, qui contempta sua senectute et auctoritate foedissimo cum eventu
pugnasset . . . (Liv. 8.33.15); Aquae magnae bis eo anno fuerunt, Tiberisque agros
inundavit cum magna strage tectorum pecorumque <et> hominum pernicie. (Liv.
24.9.6); . . . testiculus . . . omnibus indecore, quibusdam etiam cum dolore dependet.
(Cels. 7.22.5); . . . ille autem contendat eum servum ab alio defendi cum satisdatione . . .
(Iulian. dig. 9.4.39.2); . . . vera . . . aliorsum subornant cum maxima iniuria veritatis . . .
117
See Woodman and Martin ad loc., also for Plin. Nat. 37.20.
118
For further examples, see TLL s.v. cum 1366.20ff. (‘cum condicione (extrinsecus accedente)’).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(Tert. An. 2.4); . . . quod pedibus necesse erat subiri cum labore . . . (Pereg. 11.4—NB:
perhaps a manner adjunct);
Nam qui dormiunt lubenter, sine lucro et cum malo quiescunt. (Pl. Rud.
923); . . . ut sine occulto consilio, sine nocte, sine vi, sine damno alterius, sine
armis, sine caede, sine maleficio fieri non potuerit, id sine dolo malo factum
iudicabitur? (Cic. Tul. 32); Ad haec <a> consulibus rescriptum sine detrimento rei
publicae abscedi non posse ab hoste. (Liv. 22.33.10); . . . licet filicem sine iniuria
vicini, etiam cum officio decidere et permiscere cum purgamentis cohortis . . . (Col.
2.14.6); . . . quoniam dii, quamvis potentium populorum arbitri, possessionem
Parthis non sine ignominia Romana tradidissent. (Tac. Ann. 15.24.1);
CAUSE ADJUNCTS denote inanimate entities that bring about a certain action or event.
Examples are (a)–(c). An example of a passive sentence with both a cause adjunct
(sermone huius) and a cause argument (ira) is (d). With bare ablative expressions
cause adjuncts are not always easy to distinguish from reason adjuncts (see § 10.87).
Some scholars make a distinction between inner causes (the subject’s mental attitude)
and outer causes.
(a) Homini ilico lacrumae cadunt / quasi puero gaudio.
(‘Tears of joy instantly fall from his eyes, like a small boy.’ Ter. Ad. 536–7)
(b) Nam ut lacrimae saepicule de gaudio prodeunt . . .
(‘Just as tears often flow from joy . . . ’ Apul. Met. 1.12.1)
(c) Nam quod scribis te audire me etiam mentis errore ex dolore adfici, mihi
vero mens integra est.
(‘You say you hear that I am afflicted with mental strain because of grief. No, my
mind is sound enough.’ Cic. Att. 3.13.2)
(d) Nimis sermone huius ira incendor.
(‘I’m getting really angry about his speech.’ Pl. Ps. 201)
10.79 The use of nouns and noun phrases in the ablative as cause adjuncts
Examples of the use of the ablative to mark cause adjuncts (the so-called ablativus
causae) are (a)–(c). An interesting example showing the closeness of cause and
purpose adjuncts is (c), where two ablative nouns are coordinated with the final
quo-clause. The use of the ablative to mark cause adjuncts is common from Early
Latin onwards. An alternative way to express the relationship between the subject of
the clause and its mental condition is shown in (d) and (e). Here passive participles
are used as a secondary predicate in combination with a cause argument. This
alternative is preferred in Classical prose. (For more cause arguments, see § 4.11.)
Satellites
est laborum et dolorum fuga. (Cic. Fin. 1.33); Sed dubium non est, quin haec
disciplina et ars augurum evanuerit iam et vetustate et neglegentia. (Cic. Leg.
2.33); Sed cum . . . Servius . . . ut ante dixi regnare coepisset, non iussu sed voluntate
atque concessu civium . . . (Cic. Rep. 2.38); Qua in fuga magna auctoritate Osaces,
dux Parthorum, vulnus accepit eoque interiit paucis post diebus. (Cic. Att. 5.20.3);
Hoc se colle interruptis pontibus Galli fiducia loci continebant . . . (Caes. Gal.
7.19.2); . . . aquae lignorum frumenti inopia conloquium petunt . . . (Caes. Civ.
1.84.1); . . . sed omnino cuncta plebes novarum rerum studio Catilinae incepta pro-
babat. (Sal. Cat. 37.1); . . . ni aviditate praedae in vacua Romana castra Numidae
devertissent. (Liv. 21.48.5); Hoc nuntio Marcellus duobus militum milibus Nolae in
praesidio relictis cum cetero exercitu Casilinum venit . . . (Liv. 24.19.5); . . .
Demetrium nullo alio crimine quam Romanae amicitiae initae occidit. (Liv.
41.23.11); Ex eo licet scire cibo atque aqua proprietates locorum naturaliter pesti-
lentes aut salubres esse (Vitr. 1.4.10); Pudore numquam nisi in abdito coeunt
(sc. elephanti), mas quinquennis, femina decennis. (Plin. Nat. 8.12); Alii taedio
coetus, quidam tristitia ingenii et metu comparationis, quia Augustus comiter
interfuisset. (Tac. Ann. 1.76.4); . . . nec nisi Tiberio imperitante deprecari senatum
ac principem ausus est M. Silani fratris potentia, qui per insignem nobilitatem et
eloquentiam praecellebat. (Tac. Ann. 3.24.3); Multi odio praesentium et cupidine
mutationis suis quoque periculis laetabantur . . . (Tac. Ann. 3.44.2);
119
See also TLL s.v. pro 1434.7ff.
120
For diachronic observations, see Luraghi (2005).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Si amavit umquam aut si parem sapientiam [hic] habet ac formam, per
amorem siquid fecero, clementi <hic> animo ignoscet.
(‘If he’s ever been in love or if he has as much wisdom as beauty, he’ll benevolently
forgive me if I do anything out of love.’ Pl. Mil. 1251–2)
(e) Credo, ut fit, misera, prae amore exclusti hunc foras.
(‘I suppose you shut him out through love, poor woman. It does happen.’ Ter. Eu. 98)
Supplement (in alphabetical order by preposition):
. . . QUOD·ARBOR / A VETUSTATE·DECIDIT . . . (CIL VI.2059.18–19 (Rome, AD 81)); . . . haud
scio an, qui tum animi ab decemvirorum infelicibus auspiciis Romanis hostibusque
erant, magno detrimento certamen staturum fuerit (Liv. 3.60.2); Exercitum alterum in
Aequos, non ad bellum —victos namque se fatebantur—sed ab odio ad pervastandos
fines . . . duxere . . . (Liv. 6.4.8); (pacem) . . . cuius impetratae ab insita animis levitate
ante deductum Cremera Romanum praesidium paenituit. (Liv. 2.49.12); Neque is rex
ab inopia id fecit. (Vitr. 2.8.10); Horum nomina non me a nimio favore sed a certo
posuisse iudicio scietis . . . (Sen. Con. 10 pr. 16); . . . etiam corpus ex calore emittit
sudores. (Vitr. 8.2.4); Ego me iniuriam fecisse filiae fateor tuae, / Cereris vigiliis, per
vinum atque impulsu adulescentiae. (Pl. Aul. 794–5); Ceterum, id quod non timebant,
per dolum ac proditionem prope libertas amissa est. (Liv. 2.3.2); Nam etiam ipsa die
accessimus et ad ceteros monachos valde sanctos, qui tamen pro aetate aut inbecillitate
occurrere in monte Dei ad oblationem faciendam non poterant. (Pereg. 5.10);
Agent adjuncts (human beings actively involved in a non-passive process) are very
rare. An example is (a).121 (For agent arguments with the passive, see § 5.8.)
(a) Moriuntur non alter ab altero, sed uterque a patre.
(‘They died, not at each other’s hand, but both at their father’s.’ Sen. Con. 5.3.1)
121
For further examples with the verb morior, see TLL s.v. 1493.32ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
There are a few instances of nouns and noun phrases in the genitive that can only be
understood as purpose adjuncts (the so-called genetivus causae). Ex. (a) is accepted by
some editors as the first attestation. (For gerundival clauses in the genitive, see also
§ 16.106.)122
(a) Utrum igitur avaritiae an egestatis accessit ad maleficium? Avaritiae?
(‘Was it then because of avarice or of need that he entered upon the crime? Avarice?’
Rhet. Her. 4.50—NB: Achard now reads causa avaritiae with part of the mss.)
Supplement:
. . . nunc iam, quoius (<causa> add. Iucundus) adhibetur ea cura, de fructu dicam.
(Var. R. 3.16.32) . . . cum annotinis privatisque, quas sui quisque commodi (causa v.l.)
fecerat . . . (Caes. Gal. 5.8.6); LIMEN / CONSERVO PIE/TATIS FECIT (CIL X.9.3–5 (Reggio C.));
Namque hoc genus orationis non capitis defendendi nec suadendae legis nec exercitus
adhortandi nec inflammandae contionis scribitur, sed facetiarum et voluptatis.
(Fro. Laudes fumi et pulveris 3 (vdH p. 215)—NB: coordinated with a genitive
gerundival clause);
122
For discussion and references, see Löfstedt (1942: I.169), Pasoli (1966: 63–8), Calboli (1975:
157–63, 2009: 91–2).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The prepositions that are most often used in purpose adjuncts are ad ‘to’, ‘for’ (OLD
§ G); in ‘with a view to’, ‘for the sake of (OLD § 21); propter ‘for the purpose of ’ (OLD
§ 7); causa ‘for the purpose of ’, ‘for the sake of ’ (although it is sometimes difficult to
distinguish these from cause adjuncts) (OLD § 17); and gratia ‘for the sake of ’. Ob
‘with a view to’ (OLD § 4) is sometimes understood as expressing purpose, especially
in later periods of Latin. An example from Cicero is (g).123 Most of the nouns
involved are abstract nouns.
(a) Inde Numidis Hispanisque ad praesidium simul castrorum simul Capuae
relictis cum cetero exercitu . . . descendit.
(‘Then leaving Numidians and Spaniards to defend the camp and Capua at the same
time, he came down with the rest of the army . . . ’ Liv. 24.12.4)
(b) . . . ex omnibus partibus . . . ii qui ad auxilium convenerant clamore et ululatu
suorum animos confirmabant.
(‘ . . . on every side . . . those who had come up to their assistance sought to encourage
their countrymen with shouts and yells.’ Caes. Gal. 7.80.4)
(c) Illam minorem in concubinatum sibi / volt emere miles quidam qui illam
deperit.
(‘A certain soldier, who is passionately in love with that younger one, wants to buy
her to become his concubine.’ Pl. Poen. 102–3)
(d) Inde M. Marcellum propraetorem cum iis copiis quas habebat Nolam in
praesidium misit.
(‘From there he sent M. Marcellus, the propraetor, with the forces which he had to
Nola, to serve as a garrison.’ Liv. 23.39.8)
(e) Quamquam sunt qui propter utilitatem modo petendam putant amicitiam.
(‘As a matter of fact there are some who think that friendship is to be sought solely
for advantage.’ Cic. Inv. 2.167)
(f ) Ad tria milia militum ibi Hannibalis, quae praesidii causa relicta erant,
oppressa.
(‘About three thousand of Hannibal’s soldiers, who had been left for the sake of a
garrison, were overpowered there.’ Liv. 27.1.2)
(g) . . . ob eiusque mulieris memoriam primo anno et vir et pater eius consul est
factus.
(‘ . . . and in gratitude to that woman’s memory both her husband and her father were
made consuls for the first year of the republic.’ Cic. Fin. 2.66)
123
See TLL s.v. ob 21.33ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
124
See TLL s.v. ad ‘de relatione finali’ (529.15ff.) and ‘de relatione generaliter’ (545.82ff.).
125 126
See TLL s.v. in (‘finale’) 763.35ff. See TLL s.v. causa 682.13ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Idcirco ‘therefore’ (OLD § 1, a), ideo ‘therefore’ (OLD § 2), and propterea ‘therefore’
(OLD § a) are used as purpose adjuncts both independently and in combination with
the final subordinators ut, quo, and ne. The interrogative and relative adverbs are cur
(quor and qur) ‘why’ and quamobrem ‘why’. These adverbs are also found as cause
adjuncts.
(a) Idcirco nemo superiorum attigit ut hic tolleret? Ideo C. Claudius Pulcher
rettulit ut C. Verres posset auferre?
(‘Did none of his predecessors touch those things for that reason, namely so that this
man might remove them? Did C. Claudius Pulcher give them back in order that
G. Verres might take them away again?’ Cic. Ver. 4.7)
(b) Suspicio est mi nunc vos suspicarier / me idcirco haec tanta facinora
promittere / quo vos oblectem . . .
(‘It’s my suspicion now that you suspect that I’m promising such great deeds for the
following reason, so that I may entertain you . . . ’ Pl. Ps. 562–4)
(c) Nam propterea <te> volo / scribere ut pater cognoscat litteras, quando legat.
(‘I want you to write for the simple reason that your father may recognize your
handwriting when he’s reading it.’ Pl. Bac. 729–30)
(d) Nam quor me miseram verberas? # Ut misera sis . . .
(‘Why on earth are you hitting me, miserable thing that I am? # So that you’re
miserable . . . ’ Pl. Aul. 42)
(e) Nunc quoius iussu venio et quam ob rem venerim / dicam . . .
(‘Now I’ll tell you on whose command and for what reason I’ve come . . . ’ Pl. Am.
17–18)
Satellites
(b) Cur ego non timuerim quaeris? Quia te illinc abisse constabat.
(‘Do you ask the reason for my own fearlessness? Because it was known that you had
left the scene.’ Cic. Dom. 8)
(c) Ipsum ‘meridiem’ cur non ‘medidiem’? Credo, quod erat insuavius.
(‘Why is meridiem itself not medidiem? I suppose because it would be less pleasant.’
Cic. Orat. 157)
(d) Dein, postquam (sc. potio) fervore aspernabatur, frigida in aqua adfunditur
venenum . . .
(‘Then, when he declined the drink because of its warmth, poison was poured into
the cold water.’ Tac. Ann. 13.16.2)
The use of the bare ablative to mark reason adjuncts is rare. Ex. (a) is sometimes taken
as a cause adjunct. Cause and reason cannot always be easily distinguished.127 An
unambiguous example is (b).
(a) Hi vel aetate vel curae similitudine patres appellabantur.
(‘These were called fathers by reason either of their age or of the similarity of their
duties.’ Sal. Cat. 6.6)
(b) . . . contra Tragurium Bova et capris laudata Brattia . . .
(‘ . . . opposite Tragurium Bova and Brattia celebrated for their goats . . . ’ Plin. Nat. 3.152)
Supplement:
Sed quoniam regale civitatis genus, probatum quondam, postea non tam regni quam
regis vitiis repudiatum est . . . (Cic. Leg. 3.15); Me tua matris egens damnataque
paelice proles / exanimat . . . (V. Fl. 2.153–4);128 Nam vis innumera, Lugii aliaeque
gentes, adventabant, fama ditis regni, quod Vannius triginta per annos praedatio-
nibus et vectigalibus auxerat. (Tac. Ann. 12.29.3); Deserta inde et damnata solitudine
domus totaque illi monstro relicta. (Plin. Ep. 7.27.6);
Appendix: Remarkable is the use of the genitive in (c), which is almost equivalent to
a reason adjunct, but is actually an extreme case of the use of the genitive that is
discussed in § 9.30.129
127
An interesting collection of examples is given by Nipperdey and Andresen ad Tac. Ann. 3.24.3.
128
This interpretation follows TLL s.v. damno 19.20. For other interpretations, see Poortvliet ad loc.
129
See Haverling (2006b: 354).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Velim credas hanc abstinentiam (sc. eligendi) non fuisse fastidii, sed quia . . .
(‘I should like you to believe that this failure to choose was not due to
contempt, but because . . . ’ Symm. Ep. 4.58.2)
The prepositions used in reason adjuncts are ob ‘because of ’ (OLD § 3), as in (a) and
(c), and propter ‘because of ’, ‘on account of ’ (OLD § 4), as in (b) and (c).130 Also prae
‘under the pressure of ’ (OLD § 5). Then there are the postpositions causa ‘on account
of ’ (OLD § 19), as in (b); gratia ‘because of ’ (OLD § 7e), as in (d)—rare; and ergo ‘on
account of ’ (OLD, s.v. ergo1), as in (e)—mainly found in Early Latin and in poetry as
an archaism. A remarkable use of ab, which shows the affinity of reason and cause, is
shown in (f ). Note the coordination in (c).
(a) . . . ea quae nosmet ipsi ob amorem in rem p(ublicam) incredibilem et
singularem pertulimus ac sensimus . . .
(‘ . . . and those things which I have borne and suffered because of an incredible and
unparalleled love for the Republic . . . ’ Cic. de Orat. 3.13)
(b) Alium alio pacto propter amorem ni sciam / fecisse multa nequiter, verear
magis / me amoris causa hoc ornatu incedere.
(‘If I didn’t know that other men had done many bad things in one way or another on
account of their love, I’d be more hesitant to parade myself in this get-up for the sake
of my love.’ Pl. Mil. 1284–6)
(c) Hieronymum ac postea Hippocraten atque Epicyden tyrannos cum
ob alia, tum propter defectionem ab Romanis ad Hannibalem invisos
fuisse sibi.
(‘As for Hieronymus, and later, Hippocrates and Epicydes, the tyrants, they said they
had themselves hated them for other reasons and especially because of their desertion
from the Romans to Hannibal.’ Liv. 26.30.2)
(d) Hancine aetatem exercere mei <me> amoris gratia?
(‘Should I tire you out at your age for the sake of my love?’ Pl. Mil. 626)
(e) . . . et Rhodiorum populus corona aurea virtutis ergo donatus . . .
(‘ . . . and the people of Rhodes, moreover, were presented with a golden crown
because of their valour.’ Liv. 31.15.6)
(f ) Id autem opus non modo vulgo, sed etiam in paucis a magnificentia
nominatur.
(‘And this building is famous because of its magnificence not only with the crowds
but with the experts.’ Vitr. 7 pr.15)
130
For ob and propter and their distribution in Cicero, see Luraghi (2005).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(g) Latro sic divisit: an pater ob ullam adoptionem accusari possit; an ob hanc
debeat.
(‘Latro’s division was: Can a father be accused on account of any adoption?
Should he be for this one?’ Sen. Con. 2.4.7)
(h) . . . ob hoc accusatur rei publicae laesae.
(‘ . . . and is therefore accused of harming the state.’ Sen. Suas. 2.21)
Causa and gratia originally were (causal) ablatives. In Late Latin a few more ablative
nouns in combination with genitive nouns became more or less idioms, e.g. beneficio
and merito + gen.131 Examples are (i) and (j).
(i) Competit enim actio non ideo, quia nunc abest (sc. servus), sed quia
umquam beneficio furis afu<i>t.
(‘The action lies not because he is now absent but because he ever was away
to the thief ’s advantage.’ Ulp. dig. 47.2.46)
(j) Et est ille de lyricis (Pindarum dico) qui Aesculapium canit avaritiae
merito . . . fulmine iudicatum.
(‘And then there is one of the lyric poets (I mean Pindar) who sings how
Aesculapius because of his avarice . . . was punished by a thunderbolt.’ Tert.
Apol. 14.5.)
The adverbs idcirco ‘therefore’ (OLD § 1, with § b), ideo ‘therefore’ (OLD § 2, with 1a),
and propterea ‘therefore’ (OLD § a) are used anaphorically to refer to a reason given
in the preceding context. They are also regularly used in combination with the causal
subordinators quia and quod—much less often, and only in texts later than the
Classical prose authors, with quoniam (see § 16.42)—and exceptionally cum.132 The
subordinate clauses more often follow their main clause, but they also regularly
precede. These adverbs are also used as purpose adjuncts. They include the inter-
rogative and relative adverbs quare (still two words qua re in Plautus) ‘for what
reason’, ‘why’ (OLD § 2 and § 4), cur (quor and qur) ‘for what reason’, ‘why’,
131
For the history of causa, see Miniconi (1951: 121–34), for gratia, Moussy (1966: 303–25). See also
Sz.: 133. For beneficio and merito, see Moussy (2012a).
132
For the relative frequency of these adverbs and the subordinators used with them in Caesar and
other authors, see Odelman (1972: 139–43). For the frequency of idcirco and ideo in the course of the
history of Latin, see TLL s.v. idcirco 172.30ff. See also Baños (2014: 88–95).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
quapropter ‘for what reason’, ‘why’, and the idiomatic phrase quam ob rem,
that developed into the interrogative and relative adverb quamobrem ‘for what
reason’, ‘why’. Exceptional is the use of unde ‘from what cause’ (OLD § 11). Examples
are (a)–(g).
(a) Equidem hercle nullum perdidi, ideo quia numquam ullum habui.
(‘I for one haven’t lost any such man because I’ve never had one.’ Pl. As. 622)
(b) Nam idcirco arcessor, nuptias quod mi adparari sensit.
(‘She’s sending for me because she’s heard about the preparations for my wedding.’
Ter. An. 690)
(c) Etiamne unguentis unguendam censes? # Minime feceris. / # Quapropter? #
Quia ecastor mulier recte olet, ubi nil olet.
(‘Do you think I should also apply ointments? # Don’t do that. # Why? # Because a
woman smells right when she doesn’t smell of anything.’ Pl. Mos. 272–3)
(d) Quare haec quoque praetermittas licet.
(‘You will do well, therefore, to also let this plea alone.’ Cic. Ver. 5.133)
(e) Quor dare ausu’s? # Quia mi lubitum est.
(‘Why did you dare to give it to him? # Because I felt like it.’ Pl. Epid. 710)
(f ) Quia ad tris viros iam ego deferam nomen tuom. # Quam ob rem? / # Quia
cultrum habes.
(‘Because I’ll report your name to the Board of Three now. # Why? # Because you
have a knife.’ Pl. Aul. 416–17)
(g) Hoc est unde abdicatus es . . .
(‘This is why you were disinherited.’ Sen. Con. exc. 8.5. init.)
Appendix: Quare is remarkably often used as a connecting relative expression. For its
development into French car, see L. Löfstedt (1966: 106–10).
The neuter accusative forms of the interrogative and relative pronouns (quid ‘for what
reason’, ‘why’, and—rarely—quod, respectively) are used as reason adverbs from
Early Latin onwards to ask for or to refer to the motive or reason for a certain action
or situation. Examples are (h) and (i).133
133
A survey of this use of quid can be found in Lodge: II.503Bff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Respect adjuncts indicate that the state of affairs the clause refers to does not apply in
full, but is limited to a well-defined area.134 They resemble disjuncts of qualified truth
(see § 10.101), but are more objective than those. Examples of respect adjuncts are
(a)–(c). In (b) the subject’s (Antonius) mistake concerns only the timing. These
adjuncts are mostly used with one-place verbs. It is not always easy to distinguish
them from cause, source, and means and instrument adjuncts. They are common in
all periods of Latin. The same expressions are also used at the noun phrase (see
§ 11.64) and adjective phrase (see § 11.92) levels.
(a) Quam (sc. the girl) si intellegam deficere vita, iam ipse / vitam meam tibi
largiar . . .
(‘If I were to see you running out of life, I’d immediately donate my life to you . . . ’ Pl.
As. 609–10)
(b) . . . in eo non tu quidem tota re sed, quod maximum est, temporibus errasti.
(‘In this you were not entirely wrong, but you were wrong about the timing, which is
all-important.’ Cic. Phil. 2.23)
(c) Bis perit amator, ab re atque <ab> animo simul.
(‘The lover perishes twice, with regard to his fortune and to his mind.’ Pl. Truc. 47)
134
Bennett (1914: II.365) uses the term ‘specification’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Quod toti Asiae iure occisus videbatur istius ille verbo lictor, re vera
minister improbissimae cupiditatis, pertimuit iste ne Philodamus Neronis
iudicio liberaretur.
(‘Since the whole of Asia regarded the slaying of that man—nominally
Verres’ lictor, really the instrument of his foul passions—as a just act, Verres
was much afraid that Philodamus would be acquitted by Nero’s court.’ Cic.
Ver. 1.72)
135
See Courtney (2004).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
136
For discussion of the Latin examples, see Hahn (1953) and Jacquinod (1992). For expressions
involving body parts in modern languages, see Lauric (2001).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
This expression has a parallel in the use of the so-called dativus sympatheticus discussed
in § 10.96, but it is structurally different and unproductive. These instances are often
treated as ‘double accusative’ constructions, but unlike the real ‘double accusative’
constructions with trivalent verbs discussed in § 4.71, the verbs involved are bivalent.
137
Instances in OLD s.v. ab § 25; s.v. ad § 40ff.
138
See TLL s.v. de ‘de relatione’ 67.25ff.; OLD s.v. de §§ 12, 13.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(e) Ut minime mirum futurum sit si, cum aetate processerit, aut in hoc oratio-
num genere cui nunc studet tantum quantum pueris reliquis praestet omni-
bus qui unquam orationes attigerunt aut . . .
(‘It would not at all surprise me if, as he grows older, he will excel all who have ever
engaged in oratory in this style which he now affects, as much as a man surpasses a
boy; or . . . ’ Cic. Orat. 41)
Supplement:
Ai’n tu te valere? # Pol ego hau perbene a pecunia. (Pl. Aul. 186); A morbo valui, ab
animo aeger fui. (Pl. Epid. 129); . . . si hoc actor tantum poterit a facundia . . . (Ter.
Hau. 13); Nihil enim isti adulescenti neque a natura neque a doctrina deesse sentio.
(Cic. de Orat. 3.229); Sumus enim flagitiose imparati cum a militibus, tum a
pecunia . . . (Cic. Att. 7.15.3); . . . verum etiam dicendi consuetudo longe et multum
isti vestrae exercitationi ad honorem antecellet. (Cic. Mur. 29); Dicito patri quo pacto
mihi cum hoc convenerit / de huius filio. (Pl. Capt. 395–6); Habet hoc virtus, quam
tu ne de facie quidem nosti, ut . . . (Cic. Pis. 81); . . . si quaeratur, cur doctissimi
homines de maximis rebus dissentiant. (Cic. de Orat. 3.114);
Appendix: As part of the general confusion concerning the use of the genitive it is
sometimes used instead of a de-expression in Late Latin.139
139
See Haverling (2006b: 357).
140
Examples taken from Ripoll (2007: 173–7) who uses the term ‘adverbes de domaine’. See also
Ricca’s ‘domain adverbs’ (2010: 153).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
141
Havers (1911: 170ff.) introduced the term. His collection of examples is still very complete, though
some categories that he distinguishes are dealt with in a different way in this Syntax. See also the examples
and discussion in Bennett (1914: II.134ff.). For Lucretius, see Reinhardt (2010: 217–19). Herslund (1988)
calls it ‘datif partitif ’. For the aspect of inalienability, see Lehmann (2005: 38–9).
142
However, as Happ (1976: 281) observes, with the dative constituent omitted, the remaining
utterance is sometimes odd.
143
See Bolkestein (2001a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
(Apul. Met. 3.24.16); Si equo fistula in corona fuerit natus, . . . (Pelag. 239); Ut equo
ungulae crescant. (Pelag. 231); Testiculus ei dexter tumebit et indurescet. (Veget.
Mul. 2.122.2);
With sum: Pol istuc tibi et tuo est ero in manu. (Pl. Poen. 912); Est atque non est mi
in manu, Megaronides. / Quin dicant, non est; merito ut ne dicant, id est. (Pl. Trin.
104–5); Cui caput infirmum est is . . . (Cels. 1.4.1); Mihi anima in naso esse. (Petr.
62.5 (Niceros speaking));
Inanimate dative constituents: At nitidae surgunt fruges ramique virescunt / arbo-
ribus, crescunt ipsae fetuque gravantur. (Lucr. 1.251–2); (herbae) . . . quibus flos ante-
quam caules exeant . . . (Plin. Nat. 1.69); . . . ne . . . crescat gurgiti (gurgitis rec.) altitudo.
(Plin. Nat. 8.11); Nulli arborum folia ibi decidunt . . . (Plin Nat. 12.40, cf. 16.78);
In a joke: Quid mihi futurum est, quoi duae ancillae dolent / quibus te donavi? (Pl.
Truc. 633–4);
With bivalent verbs meaning ‘touching’ or ‘wounding’ a part of someone’s body and
with verbs denoting the effect on someone’s physical or mental condition, the person
is regularly expressed in the dative case. Just as with the monovalent verbs above,
there is an implied relationship of inalienability between the dative constituent and
the second argument of the verb. As in the examples in the preceding paragraph, the
cor in (e) is the heart of the ‘I’. The dative constituent is again almost always animate
and functions as an experiencer. Examples are (d)–(g). (For related passive construc-
tions, see § 5.7.)
(d) Manu fustive si os fregit libero . . .
(‘If he has broken a freedman’s bone with hand or club . . . ’ Lex XII.8.3)
(e) Non res sed actor mihi cor odio sauciat.
(‘It’s not the success, but the actor that’s wounding my heart with tedium.’ Pl. Bac. 213)
(f ) (Dionysius) . . . qui cultros metuens tonsorios candente carbone sibi adurebat
capillum.
(‘ . . . through fear of the barber’s razor he used to have his hair singed off with a
glowing coal.’ Cic. Off. 2.25)
(g) Tum Tito Balventio . . . utrumque femur tragula traicitur.
(‘At this point, Titus Balventius . . . had both thighs pierced by a dart.’ Caes. Gal.
5.35.6)
Supplement:
Quid est? Quo modo? Iam quidem hercle ego tibi istam / scelestam, scelus, linguam
apscidam. (Pl. Am. 556–7); Neque illaec ulli pede pedem homini premat. (Pl. As. 775);
Oculos hercle ego istos, improba, effodiam tibi . . . (Pl. Aul. 53); Quin exta inspicere in
sole e<i> (sc. agno) vivo licet. (Pl. Aul. 565); Perii hercle, lembus ille mihi laedit latus.
(Pl. Bac. 281); . . . extemplo puer paedagogo tabula disrumpit caput. (Pl. Bac. 441);
Constringe tu illi, Artamo, actutum manus. (Pl. Bac. 799); Quis mihi subveniet tergo
aut capiti aut cruribus? (Pl. Cas. 337); Eripe oculum istic ab umero qui tenet ere, te
opsecro. (Pl. Men. 1011); Teneo ego huic oculum. (Pl. Men. 1014); Quia aps te abit,
animo male / factum est huic repente miserae. (Pl. Mil. 1331–2); . . . ni ei caput
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
exoculassitis . . . (Pl. Rud. 731); . . . flammea tempestas Siculum dominata per agros /
finitimis ad se convertit gentibus ora. (Lucr. 6.642–3); Cum penis mihi forte laesus
esset . . . (Priap. 37.3); Consuli primo tam novae rei ac subitae admiratio incluserat
vocem. (Liv. 2.2.8); Fugientibus . . . terga caesa. (Liv. 2.25.4); Ea desperatio Tuscis
rabiem magis quam audaciam accendit. (Liv. 2.47.6); Adice nunc quod <memoria>
illis nullius interventu excutitur. (Sen. Con. 9 pr. 3); Illi praecisas manus, illi erutos
oculos, illi fractos pedes. (Sen. Con. 10.4.3); Sane caput illi coperis et oculos illi ligas
(Pelag. 92.1); . . . cui caput erigere si volueris . . . (Mulom. Chir. 316—NB: . . . cuius caput
si erigere volueris . . . (Veget. Mul. 2.88.1)); . . . et laboranti cotidie pedes perunges
(Veget. Mul. 2.58.2);
The dative with a verb like abscido ‘to cut off ’ must sometimes be regarded as a third
argument, as in (h), just like the third argument with demo ‘to take away’ in (i).
A dativus sympatheticus is also possible with other satellites. Examples are (j) and (k).
In these examples, again, there is an implicational relationship (inalienability)
between the body part and the dative constituent. The dative constituent is an
animate entity, and the predicates involved are mostly the same as those that are
found in the paragraphs above.
(j) Nescioquoia vox ad auris mi advolavit.
(‘Someone’s voice has flown to my ears.’ Pl. Mer. 864)
(k) . . . mihi ad pedes misera iacuit . . .
(‘ . . . the poor creature lay down before my feet . . . ’ Cic. Ver. 5.129)
Supplement:
Vidin’ ego te modo manum in sinum huic meretrici / ingerere? (Ter. Hau. 563–4);
. . . omnibus istis latronibus . . . iam pridem de manibus arma cecidissent. (Cic. Phil.
14.21); Ea re impetrata sese omnes flentes Caesari ad pedes proiecerunt. (Caes. Gal.
1.31.2); At Venus Ascanio placidam per membra quietem / inrigat . . . (Verg. A.
1.691–2); Simile quiddam et viris in ventre gignitur, quod vocant scirron, sicut
Oppio Capitoni, praetorio viro. (Plin. Nat. 7.63); . . . rex Mithridates Aquilio duci
capto aurum in os infudit. (Plin. Nat. 33.48); . . . nascanturque quasi denuo suae
matri de ventre (Commod. Instr. 2.10.7);
It has often been stated that the dative expression is semantically equivalent to an
expression with a genitive attribute (a so-called possessive genitive—see § 11.47) or a
possessive pronoun, for example malae meae in (a), but differs from it in being ‘more
lively’ and in being ‘vulgar’ or ‘substandard’. However, whereas in many (probably
not all) cases of a dative expression the (possessive) genitive might be used as well, the
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
reverse substitution is definitely not possible, given the semantic restrictions men-
tioned above (the dative constituent is affected mentally or physically and stands in a
possessor : possessee relation to the noun involved). This shows that even where
substitution of the dative by the genitive is possible, this need not imply semantic (or
pragmatic) equivalence. The difference between the two expression types can also be
inferred from the fact that they sometimes co-occur.
The sociolinguistic difference is inter alios supported by Löfstedt (I. 1942: 225–35),
followed by Sz.: 94–6, but rightly criticized by Adams (1995b: 450ff.) and Pruvost-
Versteeg (forthc.). In Late Latin veterinarian texts dative expressions are found in
more or less the same context as genitive expressions and seem (in these contexts) to
be completely exchangeable from the semantic point of view, and, judging from their
use in various texts, there seems to be no sociolinguistic difference between them. See
Adams (1995b: 449–60) with examples and discussion. Bailey (in his commentary,
vol. I.92–3) gives instances from Lucretius, where he would have expected a genitive
instead of a dative. He points out that in sixteen out of twenty-six instances a genitive
alternative is impossible on metrical grounds, which therefore cannot be the (only)
explanation for Lucretius’ use of the dative. (Most of) the instances can be interpreted
as sympathetic datives, albeit of a statistically deviant type.
In Early Latin the dative expression type is much more common than the genitive.144
The dative possessor usually precedes the possessee noun phrase (whereas with the
genitive and the possessive pronouns the order is more often the reverse) and need
not be, though it often is, juxtaposed. The same can be observed for the Late Latin
author Pelagonius.145
It is often difficult to decide whether a dative constituent should be regarded as a
dativus sympatheticus (that is, a satellite) or as an argument of a complex expression.
There is also a diachronic aspect involved. Original dative satellites may develop into
arguments the more an expression becomes an idiom, as in (l) and (m).
144
Bennett (1914: II.145): ‘ . . . in Plautus we find only 11 genitives of nouns as against 70 Sympathetic
Datives of nouns’. He also notes that 75% of these datives are pronouns, of which two-thirds are mihi
and tibi (p. 144), but this may be due to the subject matter.
145
For Early Latin, see Bennett (1914: II.145). In the material given by Adams (1995b: 457–60)
genitives follow the nouns twenty-one times as against one preceding. Of the datives eighteen precede
and nine follow.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Disjuncts
with relational nouns as sympathetic datives is probably the fact that substitution by
a possessive pronoun or a genitive is possible (e.g. mea could be substituted for mihi
in (n)).146
(n) Ampsigura mater mihi fuit, Iahon pater.
(‘My mother was Ampsigura, and my father Iahon.’ Pl. Poen. 1065)
The label sympathetic dative is furthermore used in a wider sense for dative con-
stituents associated with nouns not denoting parts of the body as discussed in the
beginning of this section. Examples are (o) and (p).
(o) (ancillas) quae mihi comedint cibum?
(‘ . . . girls . . . to eat up my food?’ Pl. Truc. 534)
(p) . . . regique Thebano Creoni regnum stabilivit suom.
(‘And fixed his throne firm for Creon, the Theban king.’ Pl. Am. 194)
Although the dative constituents in these examples may be replaced by a possessive
pronoun and a genitive expression, respectively (meum cibum, regis Thebani Creonis
regnum—notice suom in the latter example), this Syntax treats them as beneficiary
satellites (see § 10.69), because there is no physical or mental effect on the referent of
the dative constituent involved.147
10.97 Disjuncts
Disjuncts are satellites that convey some form of comment by the speaker or writer
either on the content or on the form and communicative function of the sentence (or
clause) as a whole: that is, on the constituents that belong to the nucleus of the clause,
any adjuncts, and any form of negation. Disjuncts that comment on the content are
called ATTITUDINAL DISJUNCTS, those that comment on the form and communicative
function are called ILLOCUTIONARY DISJUNCTS in this Syntax (for the term see § 10.105).
Attitudinal disjuncts can be subdivided into two semantic types: expressions that
indicate to what extent the speaker or writer takes responsibility for the truthfulness
of the content of the sentence and expressions that indicate the subjective judgement
of the speaker or writer with respect to the content of the sentence.
146
Examples can be found in Bennett’s section (d) of his discussion of the sympathetic dative
(1914: II.142ff.).
147
Examples can be found in Bennett’s section (c) of his discussion of the sympathetic dative
(1914: II.140ff.).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
A very common type of disjunct is that by which the speaker or writer takes personal
responsibility for the content of the sentence (or clause) and signals how certain he is
about its truth—in other words, these disjuncts express the commitment of the
speaker or writer with respect to the content of the sentence.148 Adverbs and
prepositional phrases that function in this way are fors, forsan, forsitan, and fortasse
‘perhaps’,149 certe, profecto, vero ‘certainly’, and various expressions based on the stem
dub- in combination with some negation element meaning ‘beyond doubt’, such as
haud/nec/non dubie, procul/sine dubio, and in later periods, especially in Tertullian
and in legal texts from the third century AD onwards indubitate ‘undoubtedly’.
Nimirum ‘surely’ also expresses commitment. Examples are (a)–(e). Scilicet ‘of
course’, ‘obviously’ and videlicet ‘clearly’ do not indicate a degree of truth; rather
they refer to evidence for the content (they are ‘evidential’ adverbs). However, in an
indirect way they also express commitment.150
(a) Certe enim hic nescioquis loquitur.
(‘Certainly someone’s speaking here.’ Pl. Am. 331)
(b) Quamquam ego sum sordidatus, / frugi tamen sum, nec potest peculium
enumerari. # / Fortasse.
(‘Even though I look shabby I’m decent, and my money can’t be counted. # Perhaps.’
Pl. As. 497–9)
(c) Sine dubio, opinor.
(‘No doubt at all, if you ask me.’ Ter. Eu. 1044)
(d) Et sane asperiores in controversia partes fortasse recte declamatores relinquant.
(‘And to be sure declaimers may leave aside the tougher points in this controversy,
perhaps rightly.’ Quint. Decl. 270.2)
148
The formulation is taken from Dik (1997: I.242). The term ‘commitment’ is taken from Schrickx
(2011: 231). The adverbs involved are also called ‘epistemic modal adverbs’ (so Ricca 2010: 149–52). For
attitudinal disjuncts in general, see also Núñez (2002) and Pinkster (2004, 2005a).
149
For a discussion of this group of adverbs, see Calboli (2012: 148–55). For occasional fors et = forsit
in Virgil, Horace, and Propertius, see Kraggerud (2008).
150
For scilicet and videlicet, see Núñez (2001). For nimirum, scilicet, and videlicet, see Schrickx (2011).
For vero, see Kroon (1995: 300–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Disjuncts
(e) Nimirum occidor, nisi ego intro huc propere propero currere.
(‘Surely I’m being murdered unless I hastily make haste to run in here.’ Pl. Aul. 393)
Many of these words are also used in other functions, as manner and degree adjuncts
and also as modifiers of adjectives and adverbs, etc.
These disjuncts share a number of properties:
(i) they may be used as one-word sentences in answer to a sentence question, as
in (b) above, although this use is not attested for all of them;151
(ii) they do not occur in imperative sentences;152
(iii) they cannot be under the scope of negation themselves;153
(iv) they can occur in the same sentence with adverbs and adverbial phrases that
indicate an ethical or legal judgement such as recte (‘rightly’) and falso
(‘wrongly’), as in (d) above (on these, see § 10.104).
Supplement (in alphabetical order by adverb):
Quid enim diceres? Damnatum? Certe non. (Cic. Dom. 51); Ista vero . . . iucunda
certe mihi fuerunt . . . (Cic. Brut. 52); Respondebo tibi equidem, sed non ante quam
mihi tu ipse responderis, Attice, certen<e> non longe a tuis aedibus inambulans post
excessum suum Romulus Proculo Iulio dixerit se deum esse et Quirinum vocari
templumque sibi dedicari in eo loco iusserit . . . , et verumne sit ut Athenis non longe
item a tua illa antiqua domo Orithyiam Aquilo sustulerit. (Cic. Leg. 1.3); Moriendum
enim certe est et incertum an hoc ipso die. (Cic. Sen. 74); Miser ergo Archelaus?
# Certe, si iniustus. (Cic. Tusc. 5.35); Illud quidem certe nostrum consilium iure
laudandum est . . . (Cic. Fam. 1.9.13); Nunc quidem domi certo est. (Pl. Mil.
398); . . . quis umquam dixit aut veteris Academiae aut Peripateticorum . . . sapientem
nihil opinari? Certo nemo. (Cic. Luc. 113); Sed tum exspectabantur Kalendae
Ianuariae, fortasse non recte. (Cic. Phil. 5.31); Tribunus plebis fuit non fortasse
tam vehemens quam isti quos tu iure laudas, sed certe talis, quales . . . Cic. Planc. 28—
NB: non does not modify fortasse); Illam enim oderas, et iure fortasse. (Cic. Tusc.
3.26); Vide’n tu, Thais, quam hic rem agit? / Nimirum consilium illud rectum’st de
occludendis aedibus. (Ter. Eu. 783–4); Nemp’ tu illius servos es? # Planissume. (Pl.
Ps. 1169); Profecto de auro nil scio nisi ‘nescio’. (Pl. Bac. 324); Tu auferere hinc a me,
si perges mihi / male loqui, profecto, quoi ego nisi malum nil debeo. (Pl. Cur.
569–70); Ehodum dic mihi. Estne ut fertur forma? # Sane. (Ter. Eu. 360–1); Sed
visne . . . sermoni reliquo demus operam sedentes? # Sane quidem. (Cic. Leg. 2.1);
Tam ego homo sum quam tu. # Scilicet. (Pl. As. 490); A te scilicet nihil. Nemo enim
meorum. (Cic. Att. 14.7.1); Nunc, postquam nihili sum, id vero meopte ingenio
repperi. (Pl. Mos. 156); An quid est etiam amplius? # / Vero, amplius. (Ter. Ad.
468–9); Thesaurochrysonicochrysides. # / Videlicet propter divitias inditum id
nomen quasi est. (Pl. Capt. 285–6); Darius in fuga cum aquam turbidam et
151
See Schrickx (2011: 212). For diachronic developments, notably of sane and certe, see Müller (1997:
195–7).
152
For an exceptional instance in Verg. A. 8.532, see Pinkster (2004: 192, n.2).
153
For seeming exceptions, see Pinkster (2004: 193, 2005a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
A second type of truth value disjuncts are expressions that indicate to what extent or
under which conditions the content of the sentence is presented as true. Subordinate
clauses are the most common form of these disjuncts, as in (c) (see § 16.83) and (d) (see
§ 16.35), but other forms are used as well, like the ablative noun phrases in (a) and (b).
In (a) quidem underlines the fact that the opinion of the speaker is strictly personal. For
conditional, concessive, and reason clauses functioning as disjuncts of qualified truth,
see §§ 16.63, 16.71, and 16.42, respectively. A subtype are expressions that indicate
according to whom the content of the sentence is true, the so-called dativus iudicantis.
Prepositional phrases with prae ‘in comparison with’, ‘faced with’ belong here as well.
(a) Stultitia magna est, mea quidem sententia, / hominem amatorem ullum ad
forum procedere . . .
(‘It’s great stupidity, in my opinion at least, for any lover . . . to go to the forum.’ Pl.
Cas. 563–4)
(b) Nam meo iudicio pietas fundamentum est omnium virtutum.
(‘For in my opinion filial affection is the basis of all virtues.’ Cic. Planc. 29)
(c) Me quidem praesente numquam factum est, quod sciam.
(‘This never happened in my presence, as far as I know.’ Pl. Am. 749)
(d) Pulchriora Polycliti et iam plane perfecta, ut mihi quidem videri solent.
(‘Still more beautiful are the statues of Polyclitus, and indeed in my estimation quite
perfect.’ Cic. Brut. 70)
(e) Sed, quo modo video, . . . Romam tibi remigrandum est.
(‘But, as I see the situation, . . . you have got to move back to Rome.’ Cic. Fam. 9.18.4)
An alternative way of modifying the truth value of a sentence is to use parenthetical
first person singular verb forms like credo ‘I believe’ and opinor ‘I suppose’, as
in (f ),154 or subordinate clauses containing these verb forms and introduced by the
subordinator ut (‘as’).
154
For the semantic and phonological reduction of first person singular parenthetical verb forms
(shortened final -o), see Stephens (1986: 252–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Disjuncts
10.102 Dative nouns and noun phrases functioning as disjunct of qualified truth
The dative can be used to mark entities from whose perspective the content of a clause
is viewed (the so-called dativus iudicantis). Semantically such entities can be regarded
as experiencers on the level of the sentence.155 Examples are (a)–(c). There are a
number of instances of substantivized participles that function in this way, as in (c).
This usage became more popular from Livy onwards, possibly inspired by its use in
Greek authors. It is not always easy to distinguish these datives from beneficiaries.
(a) Oh melle dulci dulcior [mihi] tu es. # Certe enim tu vita es mi.
(‘Oh, you’re sweeter than sweet honey. # Certainly you are sweeter than life to me.’
Pl. As. 614)
(b) Quintia formosa est multis, mihi candida, longa, / recta est.
(‘To many people Quintia is beautiful, to me she is fair, tall, and straight.’ Catul.
86.1–2)
(c) Huius sacrae viae pars haec sola volgo nota, quae est a foro eunti primore
clivo.
(‘Of this Sacred Way, this is the only part commonly known, namely the part which
is at the beginning of the Ascent as you go from the Forum.’ Var. L. 5.8)
Supplement:
. . . tardo amico nihil est quicquam nequius, / praesertim homini amanti . . . (Pl. Poen.
504–5); Neque edepol tu is es neque hodie is umquam eris, auro huic quidem. (Pl.
Trin. 971); Iam lauta es? # Iam pol mihi quidem atque oculis meis. (Pl. Truc. 378);
An ille mihi liber, cui mulier imperat . . . (Cic. Parad. 36); . . . quod numquam tibi
senectutem gravem esse senserim quae plerisque senibus sic odiosa est . . . (Cic.
Sen. 4); Da veniam fasso, tu mihi Caesar eras. (Ov. Pont. 1.7.22); O miserabilem
illum, sibi certe. (Sen. Clem. 1.13); Quando autem istis divitibus non sordidati
sumus? (Sen. Con. 10.1.2); Alexander orbi magnus est. Alexandro orbis angustus
est. (Sen. Suas. 1.3); ‘Sed nobis’, inquitis, ‘dei sunt’. (Tert. Apol. 10.3);
. . . Caesar Gomphos pervenit, quod est oppidum primum Thessaliae venientibus ab
Epiro. (Caes. Civ. 3.80.1); Est urbe egressis tumulus templumque vetustum / desertae
Cereris . . . (Verg. A. 2.713–14); Ceterum vere aestimanti Aetolicum magis ad Ther-
mopylas bellum quam regium fuit. (Liv. 37.58.8); Dein petenti Melana sinum portus
Coelos et Panhormus et supra dicta Cardia. (Plin. Nat. 4.49); Eodem tractu Ophiussa,
Butoa, Ramnus circumvectisque Criu Metopon tres Acusagorus appellatae. (Plin.
Nat. 4.61); Ad sinistram enim intrantibus non longe ab ostiarii cella canis ingens,
catena vinctus, in pariete erat pictus . . . (Petr. 29.1);
155
As suggested by van Hoecke (1996: 17) and Bolkestein (2001a) for Latin, and Schmid (1988) for
German. For the history of the expression and examples, see Landgraf (1893: 50–5).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
156
For the disjunct function of these expressions, see Torrego (2002a: 267ff.), who also deals with
related uses of the prepositions ante and praeter.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Disjuncts
(b) Recte requiris, inquam, Attice; sed audies ex me fortasse quod non omnes
probent.
(‘ “Your question is a good one, Atticus,” I replied, “but you will get an answer from
me which perhaps everyone would not accept.” ’ Cic. Brut. 183)
(c) Itaque adeo iure coepta appellari est Canes.
(‘For that reason she rightly came to be called “the bitch”.’ Pl. Men. 718)
(d) Leviore flagitio legatum interficietis quam ab imperatore desciscitis.
(‘It will be a slighter crime to kill your general than it is to revolt from your emperor.’
Tac. Ann. 1.18.3)
(e) Nec enim ei cornix canere potuit recte eum facere quod populi Romani
libertatem defendere pararet.
(‘For the crow could not tell Deiotarus that he was doing right in preparing to defend
the liberty of the Roman people.’ Cic. Div. 2.78)
Supplement (in alphabetical order by evaluative adjunct):
Araneorum his non absurde iungatur natura, digna vel praecipua admiratione. (Plin.
Nat. 11.79); Illi igitur antiqui non tam acute optabiliorem illam vitam putant . . . (Cic.
Fin. 4.63); Bene opportuneque obviam es, Palaestrio. (Pl. Mil. 898); Melius peribimus
quam sine alteris vestrum viduae aut orbae vivemus. (Liv. 1.13.3); Miseram pacem vel
bello bene mutari. (Tac. Ann. 3.44.3); Igitur tertio quoque die cibus aegro commodis-
sime datur . . . (Cels. 3.5.3); In silvam non ligna feras insanius ac si / magnas Graecorum
malis inplere catervas. (Hor. S. 1.10.34–5); Iustius Pompeio Magno tribuatur DCCCXLVI
naves piratis ademisse. (Plin. Nat. 7.93); Ergo si iuvenis inbecillus est aut si mulier quae
gravida non est parum valet, male sanguis emittitur. (Cels. 2.10.2); . . . et opportun<i>us
post eorum accessiones cibus datur. (Cels. 3.4.12); Davum optume / video. (Ter. An.
335–6); Optume ipsa exit foras. (Ter. Hau. 722); Cum ex toto vero convaluerit, periculose
vitae genus subito mutabit . . . (Cels. 4.32.2); . . . tuque / rectius Iliacum carmen deducis in
actus / quam si proferres ignota indictaque primus. (Hor. Ars 128–30); . . . in hoc insaniae
genere solo recte vinum datur. (Cels. 3.18.24); Sed stulte istud credimus. (Arn. Nat. 2.53);
Turpius eicitur quam non admittitur hospes. (Ov. Tr. 5.6.13); Et lienibus inlinitur ex
aceto, infantibus vero e melle utilissime. (Plin. Nat. 20.129);
Fuficius enim mirum de his rebus primus instituit edere volumen . . . (Vitr. 7.
pr.14); NB: Talis equos alacer media inter proelia Turnus / fumantis sudore quatit,
miserabile caesis / hostibus insultans. (Verg. A. 12.337–9—see Tarrant ad loc.);
At tibi repente paucis post diebus, cum minime exspectarem, venit ad me Cani-
nius mane. (Cic. Fam. 9.2.1); Apud alterum uxor, liberi. Alter cotidie litteras mittere,
accipere, aperte favere Antonio. Hi subito hortari ad pacem, quod iam diu non
fecissent, non sine causa videbantur. (Cic. Phil. 12.2);157
Idque iucundius quam tutius fuit, quia inter impotentes et validos falso quiescas.
(Tac. Germ. 36.1); Amant ted omnes mulieres, neque iniuria, / qui sis tam pulcher.
(Pl. Mil. 58–9); An est quisquam qui hoc ignoret, cum de homine occiso quaeratur,
aut negari solere omnino esse factum aut recte et iure factum esse defendi? (Cic. Mil. 8);
157
For the use of repente and subito as disjuncts, see Torrego (2005).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
Deinde L. Flaccus et C. Pomptinus praetores, quod eorum opera forti fidelique usus
essem, merito ac iure laudantur. (Cic. Catil. 3.14);
Maiore animo non agnovit quam ignovisset. (Sen. Dial. 2.14.4);158 Non mea culpa
saepe ad vos oratum mitto . . . (Sal. Jug. 24.2);
These evaluative adverbs can be used independently, but not as an answer to a yes/no
question. Examples are (f ) and (g).159
158
For further examples in Seneca, see Axelson (1939: 162).
159
For further examples of recte, see OLD s.v. § 10. For this use of disjuncts in Petronius, see Rochette
(2007).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Disjuncts
aperte ‘frankly’ or vere ‘in accordance with the truth’. Examples are given in § 16.50.
Adverbs alone are uncommon. Breviter ‘briefly’ is used a couple of times by Pliny the
Elder, as in (a).160 The adverb denique ‘in short’ is used more or less in the same way,
as in (b).161 In (c), verius indicates that the wording trucidant is more appropriate
than rite sacrificant.
(a) (sc. Italia) . . . quae caelum ipsum clarius faceret . . . et tot populorum dis-
cordes ferasque linguas sermonis commercio contraheret ad conloquia et
humanitatem homini daret breviterque una cunctarum gentium in toto orbe
patria fieret.
(‘ . . . to make heaven itself more glorious, . . . to draw together in converse by com-
munity of languages the jarring and uncouth tongues of so many nations, to give
mankind civilization, and in a word to become throughout the world the single
fatherland of all the races.’ Plin. Nat. 3.39)
(b) Tibi aras, tibi occas, tibi seris, tibi item metis, / tibi denique iste pariet
laetitiam labos.
(‘You’re ploughing for yourself, harrowing for yourself, sowing for yourself, and
harvesting for yourself; finally, that labour will give joy to you yourself.’ Pl. Mer. 71–2)
(c) Pecora in fanis trucidant verius passim quam rite sacrificant.
(‘The beasts they butchered, scattered among the shrines, rather than that they
sacrificed according to ritual.’ Liv. 41.18.3)
160
For a few more instances, see TLL s.v. brevis 2187.18ff.
161
See TLL s.v. 531.27ff.
162
See Bolkestein (2001a), Joffre (1998, 2007b). For German, see Schmid (1988, 2006: 953–4) and
Wegener (1999); for the Romance languages, see Smith (2001).
163
See Volk’s (2002: 73ff.) section on ‘Teaching Speech’.
164
For a discussion of the properties of these expressions, see Rosén (forthc.).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Satellites
CHAPTER 11
The term ‘noun phrase’1 is sensu stricto used for a structure in which a noun
functions as the head and at least one other constituent is a modifier (or ‘attribute’)
of that noun (for the terminology, see § 2.4). In a wider sense, scholars also use the
term for head–modifier combinations with something other than a noun function-
ing as the head, and in a still wider sense also for nouns and pronouns without a
modifier. These various uses of the term noun phrase are shown in (a)–(d). In (a),
the determiner illi modifies the head minori adulescenti, itself a noun phrase in
which the adjective minori modifies the head adulescenti. In (b), the possessive
adjective tuom modifies the infinitive amare, which is the head. In (c), the noun
adulescens functions as a noun phrase (meaning either ‘the young man’ or ‘a young
man’), as do the demonstrative pronouns illum and hunc in (d). Modifiers may
precede or follow their head. Following modifiers are shown in (e). When there is
more than one modifier with the same head, each may be placed on different sides
of the head.
(a) Erant minori illi adulescenti fidicina et tibicina . . .
(‘That younger man had a lyre-girl and a flute-girl . . . ’ Pl. St. 542)
(b) Ita tuom conferto amare semper . . .
(‘Always handle your love in such a way . . . ’ Pl. Cur. 28)
(c) Adulescens huc iam adveniet.
(‘The young man will soon arrive here.’ Pl. Rud. 80)
(d) Boni miserantur illum, hunc inrident mali.
(‘Good men have pity on the former, bad men ridicule the latter.’ Pl. Vid. 103)
(e) Contra haec Pompeius naves magnas onerarias quas in portu Brundisino
deprehenderat adornabat.
(‘To meet this, Pompeius fitted out some large merchant-ships which he had seized
in the port of Brundisium.’ Caes. Civ. 1.26.1)
1
The abbreviations NP and NPs will be used now and then for ‘noun or noun phrase’ and ‘nouns or
noun phrases’, respectively.
The internal flexibility of the noun phrase and the absence of an article in Latin
have been taken as evidence that the concept of ‘noun phrase’ is less applicable to
Latin. Some scholars have suggested that modifiers behave more like appositions
and/or that in an earlier stage of the language modifiers were appositions.2 The
standard tests for recognizing a combination of words as a noun phrase, viz.
substitution, interrogation, and relativization, show that the noun phrase is none-
theless a valid unit of analysis. In (a) above, ei can be substituted for the entire
string minori illi adulescenti. It can also be given as answer to a question Cui erant
fidicina et tibicina? ‘Who had a lyre-player and a piper?’ In (e), the entire string
naves magnas onerarias, and not naves alone, is modified by the relative clause
quas . . . deprehenderat.3
The first major section of this chapter (§ 11.1) deals with head–modifier combin-
ations: the types of constituent that may function as head (§§ 11.2–11.23) or
attribute (§§ 11.24–11.99) in the noun phrase and their mutual relationship.
Often there is more than one modifier with the same head, each of which has
a different hierarchical position with respect to the head. This is discussed in
section § 11.75. The second major section (§§ 11.100–11.119) is devoted to the
referential properties of NPs. A separate section is reserved for independent pro-
nouns (§§ 11.120–11.160). Relative clauses (adnominal and autonomous) have
a chapter of their own (Chapter 18). The present chapter mainly deals with the
internal properties of the noun phrase; the way in which noun phrases function—
syntactically and pragmatically—on higher levels (in the clause or sentence) is
discussed elsewhere.4
Table 11.1 gives an overview of the constituents that can function as head of a noun
phrase. Details and examples are given in separate sections. Naturally, not all types of
heads can be modified by all types of modifiers. The combinations that are actually
allowed are discussed later in this chapter. For autonomous relative clauses, see
§ 18.15.
2
For discussion and references, see Touratier (2010) and Spevak (2014a: 218). For the ‘autonomy’ of
the adjective, see also de Carvalho (1991b).
3
For a typological treatment of the Latin noun phrase, see Lehmann (1991).
4
This chapter was written in frequent interaction with Olga Spevak at the same time as she was writing
her book on the Latin noun phrase (Spevak 2014a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 11.1 The main types of constituents that may function as head of a noun
phrase
Type of head Examples For more
examples,
see §
Noun
(i) common noun
(a) no modifier privata navis oneraria maxima 11.3
required (‘a very great merchant-vessel for private use’ cf.
Cic. Ver. 5.136)
(b) modifier Quis de C. Cethego atque eius in Hispaniam 11.70
required profectione . . . cogitat . . . ?
(‘Who thinks of Gaius Cethegus and his departure for
Spain . . . ?’ Cic. Sul. 70)
(ii) proper name Omnium primum iste qui sit Sosia, hoc dici volo. 11.4
(‘First of all I want to be told who that Sosia is.’
Pl. Am. 609)
Personal pronouns Tam infandum facinus, mea tu, ne audivi quidem. 11.6
(‘I’ve never even heard of such an unspeakable act, my
dear.’ Ter. Eu. 664)
Other pronouns . . . rogitas, / quo ego eam, quam rem agam, quid 11.7
negoti geram . . .
(‘ . . . you ask me where I’m going, what I’m doing,
what business I’m carrying out . . . ’ Pl. Men. 114–15)
Infinitive Ita tuom conferto amare semper, si sapis . . . 11.8
(‘If you’re smart, always handle your love in such a
way . . . ’ Pl. Cur. 28)
Various so-called
substantively used
expressions
(i) possessive Quamquam a te quidem cumulate satis fit et mihi et 11.10
adjectives meis omnibus.
(‘ . . . although your response is more than gratifying
both to myself and to all connected with me.’ Cic.
Fam. 13.32.1)
(ii) quantifiers Tantum gemiti et mali maestitiaeque / hic dies mi 11.11
optulit . . .
(‘This day’s brought me so much groaning and trouble
and sadness . . . ’ Pl. Aul. 722–3)
Atque haud scio an plerique nostrorum oratorum
ingenio plus valuerint quam doctrina.
(Continued )
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
name or when the proper name is used in a generalizing sense ‘a person’ or ‘persons
like X’.5 In the case of specific persons, determiners that serve to identify the person
carrying a personal name like iste and ille in (a) and (d) are common, as are possessive
adjectives (b). The indefinite determiners aliquis and quidam are not uncommon
either.6 Less common are evaluative adjectives, as in (c) and (d), and nouns (including
proper names) and noun phrases (typically in the genitive, as in (e)). Ex. (f ) has an
attributive noun phrase in the ablative (a so-called ablativus qualitatis) (but see the
note). For plural proper names, see the Supplement.7
(a) Omnium primum iste qui sit Sosia, hoc dici volo.
(‘First of all I want to be told who that Sosia is.’ Pl. Am. 609)
(b) Quam ob rem magnopere te hortor, mi Cicero, ut non solum orationes meas,
sed hos etiam de philosophia libros, qui iam illis fere se aequarunt, studiose
legas.
(‘And therefore, my dear Cicero, I cordially recommend you to read carefully not
only my orations but also these books of mine on philosophy, which are now about
as extensive.’ Cic. Off. 1.3)
(c) . . . quoniam et hunc tu oratorem cum eius studiosissimo Pammene, cum
esses Athenis, totum diligentissime cognovisti . . .
(‘ . . . since you yourself studied this orator thoroughly and with the utmost diligence
with his devoted admirer Pammenes during your stay at Athens . . . ’ Cic. Orat. 105)
(d) Haec cum legas, tum bellus ille et urbanus / Suffenus unus caprimulgus aut
fossor / rursus videtur.
(‘When you come to read these, the fashionable well-bred Suffenus I spoke of seems
to be nothing but any goatherd or ditcher, to look at him again.’ Catul. 22.9–11)
(e) Equidem Sosia Amphitruonis sum.
(‘I am Amphitruo’s Sosia.’ Pl. Am. 411)
(f ) Quis L. Philippum summo ingenio, opera, gratia, nobilitate a M. Herennio
superari posse arbitratus est?
(‘Who thought that a man of Lucius Philippus’ exceptional talents, public
service, high esteem, and nobility could be defeated by Marcus Herennius?’
Cic. Mur. 36)
In the case of descriptive adjectives, instead of an attribute an apposition of the form
homo/vir/mulier, etc. + adjective is more common, as in (g). Sometimes, when the
adjective is used alone, it is difficult to decide whether the adjective is an attribute or
5
See Orlandini (1995: ch. 6), Vallat (2002), and Biville (2008).
6
For the use of aliquis and quidam with proper names and the differences between them, see Bertocchi
and Maraldi (2006).
7
See Vallat (2005) and Biville (2008).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(g) Post mortem eius Sassia moliri statim, nefaria mulier, coepit insidias filio.
(‘After his death, the unspeakable Sassia immediately started to plot against
her son.’ Cic. Clu. 176)
(h) . . . cur omnium perfidiosissimus C. Marius Q. Catulum, praestantissuma
dignitate virum, mori potuit iubere?
(‘ . . . why had that monster of treachery Gaius Marius the power to order the
death of that noblest of mankind, Quintus Catulus?’ Cic. N.D. 3.80)
(i) Ex Macedonia CC erant quibus Rhascypolis praeerat, excellenti virtute.
(‘From Macedonia there were two hundred under the command of Rhascy-
polis, a man of marked valour.’ Caes. Civ. 3.4.3)
Supplement:
Singular: specific: Quis is homo est? # Menaechmus quidam. (Pl. Men.
650); . . . nemo huic ipsi nostro C. Mario, cum ei multi inviderent, obiecit umquam.
(Cic. Sul. 23); An tu existimas, cum esset Hippocrates ille Cous, fuisse tum alios
medicos . . . (Cic. de Orat. 3.132); . . . quod Cato ille sapiens queritur . . . (Cic. Div.
1.28); . . . cur omnium crudelissimus tam diu Cinna regnavit? (Cic. N.D. 3.81); Ita
sum ab omnibus destitutus ut tantum requietis habeam quantum cum uxore et filiola
et mellito Cicerone consumitur. (Cic. Att. 1.18.1); CICERO F. TIRONI SUO DUL-
CISSIMO S. (Cic. Fam. 16.21.init.); Ille cum Cotta saucio communicat, si videatur,
pugna ut excedant . . . (Caes. Gal. 5.36.3); . . . quom te talem virum di monuere, uti
aliquando pacem quam bellum malles neu te optumum cum pessumo omnium
Iugurtha miscendo conmaculares . . . (Sal. Jug. 102);
generic (a person like): . . . ostendam hanc Palatinam Medeam migrationemque
hanc adulescenti causam sive malorum omnium sive potius sermonum fuisse.
(Cic. Cael. 18); Neminem nominabo: putate tum Phormioni alicui, tum Gnathoni,
tum etiam Ballioni. (Cic. Phil. 2.15); Quem ita obsides, nove Hannibal . . .
(Cic. Phil. 13.25); Voluisti enim in suo genere unum quemque nostrum quasi
quendam esse Roscium . . . (Cic. de Orat. 1.258); . . . Mithridates, Ponticus rex . . .
semper animo maximus, consiliis dux, miles manu, odio in Romanos Hannibal . . .
(Vell. 2.18.1); Funerata est illa pars corporis qua quondam Achilles eram.
(Petr. 129.1);
Plural: specific: At enim multi Lesbonici sunt hic. (Pl. Truc. 919); Neque solum
Caelius sed etiam adulescentes humanissimi et doctissimi . . . Titus Gaiusque Coponii
qui ex omnibus maxime Dionis mortem doluerunt . . . (Cic. Cael. 24); Principio Ioves
tres numerant i qui theologi nominantur, ex quibus primum et secundum natos in
Arcadia, alterum patre Aethere, ex quo etiam Proserpinam natam ferunt . . . (Cic. N.
D. 3.53); Tres sunt praeterea eiusdem gentis: Xerxes et duo Artaxerxae, Macrochir
cognomine et Mnemon. (Nep. Reg. 1.3); . . . duosque Senecas unicumque Lucanum
facunda loquitur Corduba . . . (Mart. 1.61.7–8); Et quot fuerint Pythagorae nobiles,
quot Hippocratae. (Gell. 14.6.3);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
generic (people-like): Hinc illi Lycurgi, hinc Pittaci, hinc Solones atque ab
hac similitudine Coruncanii nostri, Fabricii, Catones, Scipiones fuerunt, non tam
fortasse docti, sed impetu mentis simili et voluntate. (Cic. de Orat. 3.56);
M’. Manilius patrum nostrorum memoria, ne semper Curios et Luscinos loquamur,
pauper tandem fuit? (Cic. Parad. 50); . . . meras proscriptiones, meros Sullas.
(Cic. Att. 9.11.3);
Proper names denoting a locality may also function as head of a noun phrase.
Common attributes are adjectives that specify a region like Gallia Narbonensis or
Hispania Citerior. But other modifiers can be used as well, as in (j)–(l).
(j) Homini post homines natos turpissimo, sceleratissimo, contaminatissimo
quis illam opimam fertilemque Syriam . . . dedit?
(‘Who ever committed the rich and fertile province of Syria . . . to the man who was of
all men ever born the most vile, the most wicked, the most polluted?’ Cic. Dom. 23)
(k) Qualis et, arentes cum findit Sirius agros, / fertilis aestiva Nilus abundet aqua.
(‘Or how, when Sirius splits the parching fields, through all the heats life-giving Nile
is full in flood?’ Tib. 1.7.21–2)
(l) Quot caelum stellas, tot habet tua Roma puellas.
(‘Your Rome has as many maidens as heaven has stars.’ Ov. Ars 1.59)
Supplement:
Plural: Quattuor sunt provinciae, patres conscripti . . . Galliae duae . . . Syria et
Macedonia . . . (Cic. Prov. 3); Interea . . . duas Hispanias confirmari . . . nolebat.
(Caes. Civ. 1.29.3); Deinde, ubi consumpto restabit munere pauper, / dic alias iterum
naviget Illyrias! (Prop. 2.16.9–10);
Appendix: Remarkable is the use of a proper name as secondary predicate modified
by the degree adverb tam in (m).8
The proper personal pronouns (ego, tu, nos, vos) cannot function as head of an NP.9
There are, however, a few remarkable instances in which the personal pronouns ego
and tu (no instances found for nos and vos) seem to function as head of a noun
phrase. In (a), ego can be seen as a metalinguistic reference to the preceding context.
8 9
See Biville (2008: 35). See Fugier (1974).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Note that the verb is third person.10 In (b), te appears to be modified by a prepos-
itional phrase. In (c), it is not clear whether the ablative of description is an attribute
with te or rather some type of parenthesis.
(a) Nec lact’ lactis magis est simile quam ille ego simil’est mei.
(‘Milk doesn’t resemble milk more than that me resembles this me.’ Pl. Am. 601)
(b) Quem Lesbia malit / quam te cum tota gente, Catulle, tua.
(‘Since Lesbia likes him better than you, Catullus, with all your kin.’ Catul. 79.1–2)
(c) Illud, Torquate, quale tandem videtur, te isto nomine, ingenio, gloria, quae
facis . . . non audere in conventu dicere?
(‘Or what, pray, are we to think of the situation if you, Torquatus, bearing the name
you do, and gifted and distinguished as you are, dare not profess before a public
audience the real object of all your actions . . . ?’ Cic. Fin. 2.74)
Appendix: Combinations like ego ille in (d) and (e) are different: ille is the determiner
of an apposition or of an autonomous relative clause. In dactylic poetry the favourite
order is ille ego, as in (f ).11
(d) Vide sis, ego ille doctus leno paene in foveam decidi, / ni hic adesses.
(‘Just look, I, that clever pimp, almost fell into a pit, if you hadn’t been
present.’ Pl. Per. 594)
(e) Itaque ego ille qui semper pacis auctor fui . . . , ego igitur pacis, ut ita dicam,
alumnus . . .
(‘Therefore, I am a man who has always been an advocate of peace . . .
Therefore I, a foster child of peace so to speak . . . ’ Cic. Phil. 7.7)
(f ) Ille ego liber, / ille ferox tacui!
(‘I, the one whom you know to be so open and savage-tempered, said
nothing.’ Ov. Met. 1.757–8)
Many pronouns can be used as head with a genitive noun or noun phrase as its
attribute, (i) when the genitival modifier denotes a plurality or collectivity of individ-
ual entities, as in (a) and (b), and the whole expression serves to single out one or
more entities (a genitive of the whole—see § 11.51); (ii) when the genitival modifier
denotes a concrete or abstract mass, a collectivity, or a plurality, as in (c) and (d), and
the whole expression serves to specify the quantity (a genitive of quantity—see
§ 11.59); (iii) further, ex. (e) illustrates the use of a genitive of description (see
§ 11.48). Alternative expressions can be used as well, as in (f ) with the preposition ex.
10
For two more instances from Plautus, see TLL s.v. ego 275.78ff.
11
For more instances, see TLL s.v. ego 275.78ff.; s.v. ille 360.12ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
modifiers, the first attestation of which is (a).12 Under the influence of Greek, in
which the use of the infinitive with an article is common, the use of determiners, of
the intensifier ipsum, and of totum increased in Cicero, especially in his philosophical
works and notably in Christian authors like Tertullian.13 In many texts this use of the
infinitive is entirely absent. The use of adjectives and genitives of pronouns is a rare
further exploitation of its possibilities (see the Supplement). Most instances are
present active (or deponent) infinitives. Perfect active infinitives are almost not
attested. The use of negators and adverbs with the infinitive is rare, but see (b) and
(c). (For the use of the infinitive as subject, object, or complement, see §§ 9.7, 9.15,
9.38, 9.47.)
(a) Ita tuom conferto amare semper . . .
(‘Always handle your love in such a way . . . ’ Pl. Cur. 28)
(b) . . . hoc non dolere solum voluptatis nomine appellaret . . .
(‘ . . . he might have confined the name of pleasure to this state of freedom from
pain . . . ’ Cic. Fin. 2.18)
(c) . . . ut totum hoc beate vivere in una virtute poneret.
(‘ . . . that he based the entire possibility of living happily upon virtue alone.’ Cic. Tusc.
5.33)
Supplement:
In qua permaneo, Catule, sententia meque, cum huc veni, hoc ipsum nihil agere et
plane cessare delectat. (Cic. de Orat. 2.24); Tibi enim adsentior vestrum esse hoc
totum diserte dicere . . . (Cic. de Orat. 2.39); Sed ego hoc ipsum velle miserius esse
duco quam in crucem tolli. (Cic. Att. 7.11.2); . . . interpositus annus alios induxit ut
victoriam sperarent, alios ut ipsum vinci contemnerent. (Cic. Fam. 15.15.2—NB:
passive); Ipsum cremare apud Romanos non fuit veteris instituti. (Plin. Nat. 7.187);
Multis ipsum / metuisse nocet . . . (Sen. Oed. 992–3—NB: perfect infinitive); Meum
enim intellegere nulla pecunia vendo. (Petr. 52.3 (Trimalchio speaking)); . . . quod
Academici quidem ipsum illud nihil posse comprehendi (NB: passive) quasi com-
prehendunt. (Gell. 11.5.8); Ergo totum hoc et spirare et vivere eius est cuius et vivere,
id est animae. (Tert. An. 10.7); Sapere enim nostrum licet obumbretur, non tamen
extinguitur . . . (Tert. An. 45.6); Nec hoc obsequi fuit aut ordinis aut honoris . . . (Min.
F. 4.6); Quo primitus pertinet ipsum velle . . . (August. Civ. 5.10.1); Velle meum
tenebat inimicus . . . (August. Conf. 8.5.10);
With adjectives: Postquam sapere urbi / cum pipere et palmis venit nostrum hoc
maris expers . . . (Pers. 6.38–9);14 Quid denique illud iners quidem, iucundum tamen
nihil agere nihil esse. (Plin. Ep. 8.9.1); . . . immaculatum cum Abraham vivere . . .
12
The best collection of examples of modifiers of infinitives is still Wölfflin (1886). See also Lundström
(1964: 177). Quintilian Inst. 9.3.9 calls the use of the infinitive ‘instead of a noun’ a Graecism (Mayer 1999:
165). See also Calboli (2009: 130–1).
13
For Tertullian’s use of the infinitive, see Hoppe (1903: 42 (=1985: 89–90)).
14
For this difficult passage, see Beikircher (1969: 69–77).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(Hier. adv. Pelag. 3.12); Ibi desiderans esse ubi est summum esse. (August. Psalm.
38.22);
With genitives: Fabius autem Maximus, cuius (cui cj. Madvig) non dimicare
vincere fuit . . . (V. Max. 7.3.7); . . . iratus mammae lallare recusas? (Pers. 3.18); . . .
illius iussisse fecisse est. (Hier. Abd. 17.18);
Appendix: Infinitives can also be part of a prepositional phrase, although in a very
limited way. They can be used in prepositional phrases with inter ‘between’ in some
form of grammatical observations, as in (d). In (e), ab ‘from’ is used in an etymological
observation. A poetic extension is (f ), with praeter ‘except’. In Late Latin a few
additional prepositions are found, probably under the influence of Greek.15
15
For infinitives in prepositional phrases, see Wölfflin (1886: 71–5, 78–80).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Possessive adjectives can be used as head of a noun phrase and can be modified by
various determiners, as in (a)–(d), and by quantifiers, as in (e) and (f ). Identifiers do
not seem to be attested.
(a) Abi intro ad vos domum / continuo, vide sitne istaec vostra intus.
(‘Go into your home at once and check if that girl of yours is inside.’ Pl. Mil. 535–6)
(b) Suom si quid sumet, id tu sinito sumere.
(‘If he takes anything of his own stuff, you must let him take it.’ Pl. Ps. 861)
(c) Sed ille noster est plane, ut supra dixi, Stoicus perpauca balbutiens.
(‘But our friend Carneades, as I said above, is a downright Stoic, though stammering
on a very few points.’ Cic. Luc. 137)
(d) . . . iste vester plane, ut mihi quidem videtur, inermis ac nudus est.
(‘ . . . your founder, as it seems to me, is completely unarmed and destitute.’
Cic. Fin. 1.22)
(e) . . . honos eius (sc. Milonis), in quo omnia mea posita esse decrevi.
(‘On this (sc. Milo’s success), I have decided, everything for me now depends.’ Cic.
Fam. 2.6.3)
(f ) . . . subfossisque equis compluribusque nostris deiectis reliquos in fugam
coniecerunt . . .
(‘ . . . by stabbing our horses and bringing down many of our troopers to the ground,
they put the rest to rout . . . ’ Caes. Gal. 4.12.2)
Supplement: Nostri in primo congressu circiter LXX ceciderunt . . . (Caes. Civ. 1.46.4);
Appendix: Substantival possessive adjectives without modifiers are very common.
Examples are (g)–(i).
Adjectives of amount such as multi ‘many’, pauci ‘few’, and plerique ‘most’ and
cardinal numerals such as unus ‘one’ and duo ‘two’ can be used as head in combin-
ation with NPs in the genitive in a part/whole relationship (the partitive genitive).
Examples are (a) and (b) (for further examples, see § 11.52). Such substantival
quantifiers may also be modified by a determiner, as in ex. (c).
(a) . . . ceteris . . . impositum vectigal est certum . . . , ut Hispanis et plerisque
Poenorum . . .
(‘ . . . in the other provinces . . . a fixed tax has been imposed . . . , as for example that
imposed on the Spaniards and most of the Carthaginians . . . ’ Cic. Ver. 3.12)
(b) Coram (cj. Gratwick, quorum P) eorum unus surrupuit currenti cursori
solum.
(‘In my presence, one of them stole the sole of a runner’s shoe from him while he was
actually running.’ Pl. Trin. 1023)
(c) Q. Fabius hortante filio ut locum idoneum paucorum iactura caperet, ‘visne’,
inquit, ‘tu ex illis paucis esse’?
(‘Quintus Fabius, upon being urged by his son to seize an advantageous position at
the expense of losing a few men, asked: “Do you want to be one of those few?” ’ Fron.
Str. 4.6.1)
For the substantival use of omnis, cunctus, and other words, see § 11.157.
Substantival identifiers can be used as head of a noun phrase and can be modified by
indefinite and interrogative determiners and (with the exception of alter) by quanti-
fiers. Examples with alius are (a)–(c). Genitival and prepositional partitive expres-
sions are possible as well.16
(a) Num quippiam aliud me vis?
(‘Do you want anything else from me?’ Pl. Per. 735)
(b) Quod numquam opinatus fui neque alius quisquam civium / sibi
eventurum . . .
(‘What neither I nor anyone else of our citizens ever believed would happen to us . . . ’
Pl. Am. 186–7)
(c) Quid aliud faciam?
(‘What else should I do?’ Pl. Mer. 568)
16
For attributes with alius, see TLL s.v. 1629.22ff.; 1632.14ff. (see also the Appendix in the text).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Nam quod egomet solus feci, nec quisquam alius affuit . . . (Pl. Am. 425); . . . si praeter
eam praedam quam dixi quicquam aliud causae inveneritis . . . (Cic. S. Rosc. 8);
Medici signa quaedam habent ex venis et spiritu aegroti multisque ex aliis futura
praesentiunt. (Cic. Div. 2.145); Hoc ego commodius quam tu, praeclare senator, /
milibus atque aliis vivo. (Hor. S. 1.6.110–11); . . . quicquid aliud (sc. Scipio) fecerit
quod cordi foret Masinissae id patres comprobare ac laudare. (Liv. 30.17.12);
Qui me Thebis alter vivit miserior? (Pl. Am. 1046); O mortalem malum, / alter hoc
Athenis nemo doctior dici potest. (Pl. Mos. 1071–2); Sed hoc verum est non solum si
ipsius viri et uxoris communis usus, sed etiam si liberorum eius aut alterius alicuius
communis usus fuerit. (Pompon. dig. 34.2.10);
. . . quod pol, soror, ceteris omnibus factum est. (Pl. Poen. 1183); Pompeius autem
cum ob ceteras causas plus potest unus quam ceteri omnes tum quod putatur ad
bellum Parthicum esse venturus. (Cic. Att. 6.1.3);
His ego litteris lectis in eadem opinione fui qua reliqui omnes, te cum omnibus
copiis ad Corfinium esse venturum. (Cic. Att. 8.11d.3); Reliquos omnes earum
turmarum aut interfecerunt aut captos ad Domitium deduxerunt. (Caes. Civ. 3.38.4);
Appendix: Alius is also found in combination with descriptive adjectives, but in that
case it is more likely that the adjectives are the head and alius the modifier, as in (d).
Some scholars17 take illi in (e) as determiner of substantival alii, but that would be an
exception: it is better to take illi as an adverb ‘there’ and to take alii with publicani.
(d) Quae cum omnia atque etiam multo alia maiora, quae consulto praetereo,
accidissent . . .
(‘Although these things and others of much greater importance, which
I purposely pass over, did happen . . . ’ Cic. Sest. 52)
(e) Hunc nos habemus publicum, illi alii sunt publicani.
(‘We occupy this public land, but over there there are other tax collectors.’
Pl. Truc. 151)
The identifier idem ‘same’ can be used as head of a noun phrase and can be modified
by determiners. Unlike the identifiers dealt with above, indefinite determiners are
excluded with idem. Examples with hic, ille, and iste are (f )–(i). A connecting relative
determiner is shown in (j).18
(f ) Sed hoc idem apud nos rectius / poteris agere.
(‘But you’ll be able to deal with this matter better at our place.’ Pl. Bac. 47–8)
(g) Atque hic idem alio loco aethera deum dicit.
(‘In another passage however this same (sc. Zeno) declares that the aether is god.’ Cic.
N.D. 1.36)
(h) Ille autem postquam filium sensit suom / eandem illam amare . . .
(‘But after the father realized that his son was in love with that same girl . . . ’ Pl. Cas. 60–1)
17 18
For example, Lodge: s.v. 96B, l. 11. For a survey, see TLL s.v. 200.20ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(i) Si idem istud nos faciamus, si [idem] imitemur, ita tamen vix vivimus / cum
invidia summa.
(‘If we do the same and imitate them, we still barely survive, and that with their ill
will.’ Pl. Cist. 27–8)
(j) Quod tu si idem faceres, magis in rem et vostram et nostram id esset.
(‘If you acted in the same way, it would be to your advantage and to ours.’ Ter. Hec. 249)
An example of idem seemingly modified by omnis is (k). Combinations with other
quantifying adjectives and also with ceterus ‘remaining’, as in (l), can be found in TLL
s.v. idem 186.39ff. In these cases it is not clear whether idem is really modified by the
other word.
(k) Non omnia eadem aeque omnibus, ere, suavia esse scito.
(‘Master, you should know that not everything is equally sweet for every-
one.’ Pl. As. 641)
(l) Cetera eadem in utroque casu facienda sunt . . .
(‘The rest that has to be done is the same in both conditions . . . ’ Cels. 3.27.1C)
The substantival use of adjectives is common from Early Latin onwards. Such
adjectives may be modified by determiners, quantifiers, identifiers, possessive adjec-
tives, and—much more rarely (except in poetry)—by other adjectives, and also by
combinations of these modifiers. Attributive NPs in the genitive, especially in a part/
whole relationship (the partitive genitive), are common with adjectives that more or
less require them on the basis of their meaning, such as medius ‘middle’, superlatives
like extremus ‘end’, summus ‘highest’, and ultimus ‘farthest’, and adjectives with a
superlative-like meaning, such as eximius ‘outstanding’.
For neuter singular adjectives, modification by a determiner, a possessive adjective,
or an adjective (and combinations of these) is common throughout Latinity. In (a),
substantival sollemne is modified by the determiner illud and the possessive adjective
nostrum. In (b), substantival tertium is determined by illud. In (c), substantival
malum is modified by magnum. In (d), an instance of a partitive genitive, substantival
dimidium requires modification by something that indicates the whole from which
the half is taken. Extension of the use of the partitive genitive to other adjectives like
incertum in (e) was popular in poetry and is found in prose from Sallust onwards.
Lucretius and Tacitus have many instances.19 One step further is the use of the
genitive in instances like (f ), where the combination of a head substantival adjective
X with an attribute noun Y is equivalent to attribute adjective X with head noun Y:
19
For Lucretius, see Bailey in his edition and commentary (1949: I.91–2); for Tacitus, see Draeger
(1882: 29–30).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(in) arido humi = (in) arida humo (PSEUDO-PARTITIVE).20 Genitive attributes other than
in a partitive relation are found as well, as in (g).
(a) Tantum igitur nostrum illud sollemne servemus ut ne quem istuc euntem
sine litteris dimittamus.
(‘So let me just keep up my old-established habit of not letting anyone go your way
without a letter.’ Cic. Att. 7.6.1)
(b) Est autem, ut ad illud tertium veniam, est plane oratoris movere risum.
(‘And again, to come to our third topic, it clearly becomes an orator to raise laughter.’
Cic. de Orat. 2.236)
(c) Verum tu malum magnum habebis . . .
(‘But you will have a big thrashing . . . ’ Pl. Am. 721)
(d) Pol me haud paenitet, si licet boni dimidium mihi dividere cum Iove.
(‘Well, I’m not upset if I can share half of the good with Jupiter.’ Pl. Am. 1125)
(e) Ceteris fuga tuta fuit incerto noctis et metu insidiarum.
(‘For the remainder safety was found through flight since they were protected by the
Romans’ uncertainty in the darkness and their fear of ambush.’ Sal. Hist. 2.87B—tr.
McGushin)
(f ) . . . arborum quae humi arido atque harenoso gignuntur.
(‘ . . . trees, which grow in a dry and sandy soil.’ Sal. Jug. 48.3)
(g) Teneamus enim illud necesse est, cum consequens aliquod falsum sit, illud
cuius id consequens sit non posse esse verum.
(‘For we are bound to hold that if a conclusion is false, the premise on which it
depends cannot be true.’ Cic. Fin. 4.68)
In combinations like quid . . . laetabile in (h), laetabile is usually regarded as the
attribute, and the fact that the interrogative pronoun is in the substantival form
(quid) and not in the adjectival form quod seems to support this, but from the
semantic point of view it is odd. Similarly, nihil is considered the head in (i).
20
For a discussion from a typological perspective of this type of ‘upgrading’ of the attribute, see
Malchukov (2000: 30–5).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
eos malo publico divites feci . . . (Cato orat. 173); Summum bonum esse erae putabam
hunc Pamphilum . . . (Ter. An. 717); Neque attendit eos ipsos, unde hoc simile
ducat . . . (Cic. de Orat. 2.316); Ille atomos quas appellat . . . censet in infinito inani . . .
ita ferri ut . . . (Cic. Fin. 1.17); Nam qui se ipse norit primum aliquid se habere sentiet
divinum . . . (Cic. Leg. 1.59); Atque, ut in eodem simili verser, ut ager quamvis fertilis
sine cultura fructuosus esse non potest, sic sine doctrina animus. (Cic. Tusc. 2.13); . . .
primordia rerum / quale sit in magno iactari semper inani. (Lucr. 2.121–2); . . . omnis /
impendunt curas denso distendere pingui . . . (Verg. G. 3.123–4); . . . alvus tamen sine ullo
acri ducenda est . . . (Cels. 8.9.1d);
NB: very odd combination, if correct: . . . humani divinique iuris sciens egregium
publicum et bonas domi artes dehonestavisset. (Tac. Ann. 3.70.3);21
With a genitive (often called ‘partitive’): . . . rectum animi servas? (Hor. S.
2.3.201); . . . quia serum diei fuerit . . . (Liv. 26.3.1); Tunc milites ad caedem missi
invenere in prominenti litoris nihil laetum opperientem. (Tac. Ann. 1.53.5); . . . donec
minor filius lubricum iuventae exiret. (Tac. Ann. 6.49.2); (sc. lacus) Certo anni
bitumen egerit . . . (Tac. Hist. 5.6.3—if correct; see Heubner and Fauth ad loc.); . . . cui
in hoc lubrico aetatis non praeceptor modo sed custos etiam rectorque quaerendus est.
(Plin. Ep. 3.3.4);
Pseudo-partitives: . . . et simulacra modis pallentia miris / visa sub obscurum
noctis . . . (Verg. G. 1.477–8); Rem ad ultimum seditionis erupturam . . . (Liv.
2.45.10); . . . haud dubius facilem in aequo campi tantum superanti multitudine
victoriam fore. (Liv. 5.38.4); Bellum scilicet! Aut diverso terrarum distineri . . . (Tac.
Ann. 3.59.3);
21
For discussion and conjectures, see Woodman and Martin ad loc.
22
For possibly Thucydidean influence, see Calboli (2009: 78), with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(n) Aulide quo pacto Triviai virginis aram / Iphianassai turparunt sanguine
foede / ductores Danaum delecti, prima virorum.
(‘ . . . as when at Aulis the altar of our Lady of the Crossroads was foully defiled by the
blood of Iphianassa, shed by chosen leaders for the Danai, chieftains of the host.’
Lucr. 1.84–6—NB = primi viri)
(o) Saepe ego multa tuae levitatis dura timebam . . .
(‘I dreaded oft much hardship from your fickleness . . . ’ Prop. 1.15.1)
Supplement:
. . . sed ad eius modi facinus in quo omnia nefaria contineri mihi atque inesse
videantur. (Cic. Ver. 4.60); Sic consequentibus vestris sublatis prima tolluntur.
(Cic. Fin. 4.55); . . . ut alia magna et egregia tua omittam . . . (Sal. Jug. 10.2); Nono
die in iugum Alpium perventum est per invia pleraque et errores . . . (Liv. 21.35.4);
Nec minus ex imbri soles et aperta serena / prospicere et certis poteris cognoscere
signis. (Verg. G. 1.393–4);
With a genitive (often called ‘partitive’): Interea sol albus recessit in infera noctis.
(Enn. Ann. 89V=84S); Pergamique in occultis ac reconditis templi . . . tympana
sonuerunt. (Caes. Civ. 3.105.5); . . . nocte proditione oppidum captum liberam per
aversa urbis fugam dederat . . . (Liv. 5.29.4); Inter cetera tristia eius anni consul alter
Ap. Claudius in ipso belli apparatu moritur. (Liv. 7.25.10); . . . quod vagus interim per
amoena Asiae atque Achaiae . . . scelerum probationes subverteret. (Tac. Ann. 3.7.1);
Immo, ut melius est, tu tuos exercitus rege, mihi bellum et proeliorum incerta
trade. (Tac. Hist. 2.77.2); . . . cum per secreta Thraciae exercitum duceret . . . (Suet.
Aug. 94.5);
Pseudo-partitives: . . . strataque iam volgi pedibus detrita viarum / saxea conspici-
mus. (Lucr. 1.315–16); . . . amittunt vera viai . . . (Lucr. 1.659); . . . miratur portas stre-
pitumque et strata viarum. (Verg. A. 1.422); . . . vi militis Romani ad excidenda
castellorum ardua . . . (Tac. Ann. 11.9.1);
Appendix: Substantival neuter singular and plural adjectives are also common
without a modifier. For the singular, adjectives of the second declension are frequent,
such as bonum ‘good’, malum ‘evil’, honestum ‘honourable’, iustum ‘just’, but
also utile ‘useful’, and insigne ‘remarkable’. Examples are (p)–(r). Very common
are prepositional phrases containing a substantival neuter singular adjective,
as in (s)–(u).23
(p) Neque est melius morte in malis / rebus miseris (cj. Pylades; meiseriis mss.).
(‘In a bad and wretched situation nothing is better than death.’ Pl. Rud.
675–6)
(q) . . . in ea varietate fere melius a deteriore facultate magis quam genere
distinguitur . . .
(‘ . . . in this variety the superior is distinguished from the inferior almost
more by capacity than by style . . . ’ Cic. de Orat. 3.34)
23
For a list of instances in Livy, arranged by order of prepositions, see Riemann (1885: 98–101).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(ac) Omnes non improbi humiles, quae magna in populo multitudo est, praesi-
dium sibi paratum vident.
(‘All the lowly who are not dishonest—and there is a large proportion of that sort
among the people—look upon such an advocate as a tower of defence raised up for
them.’ Cic. Off. 2.70)
(ad) . . . vestitus comitatusque vix ad privati modice locupletis habitum . . .
(‘ . . . clothed and attended in a style inferior to that of a private person of moderate
wealth . . . ’ Liv. 38.14.9)
With adjectives that require an argument, such as dignus ‘worthy’, one expects
instances of substantival use of the adjective with the argument expressed, parallel
to instances of those adjectives in attribute function (see § 11.92), but they are hard to
find. Examples might be (ag) and (ah). With adjectives that are gradable one expects
instances like (ai), but they are hard to find as well. A few examples of substantival
adjectives with their arguments, but without modifiers, are given in small type.
(ag) Nihil enim dicam reconditum, nihil exspectatione vestra dignum, nihil aut
inauditum vobis aut cuiquam novum.
(‘For I shall tell no mystery, nothing worthy of your waiting, nothing that you have
not heard already, or that is new to anyone.’ Cic. de Orat. 1.137)
(ah) Haec digna memoratu de Proculo didicisse me memini.
(‘This is all that I remember having learned about Proculus that is worthy of
mention.’ Hist. Aug. quatt. tyr. 13.6)
(ai) Sed quisnam est iste tam demens?
(‘But who is this lunatic that you have in mind?’ Cic. Marc. 21)
Supplement:
Oriundi ab Sabinis . . . sui corporis creari regem volebant. (Liv. 1.17.2); Ab Herman-
dica profugi exulibus Olcadum . . . cum se iunxissent . . . (Liv. 21.5.7); . . . maiusque id
peccatum ducis apud peritos artium militarium haberetur quam . . . (Liv. 23.18.13);24
24
For more instances from Livy, see Riemann (1885: 83–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Several types of substantival use can be distinguished, for example natalis ‘birthday’,
which is related to dies natalis ‘the day of one’s birth’ (sometimes called ‘elliptic sub-
stantivation’) and familiaris ‘a member of one’s household’ or falsum ‘an untruth’, where
the word denotes a person or thing to which the property of the adjective applies.25
Participles can be used as head of a noun phrase. This is common for present participles,
as in (a) and (b), but future participles and perfect participles (deponent and passive) can
be used as well, as is shown in (c)–(f ). In some cases it is difficult to decide whether a
participle is still a participle or whether it has developed into a noun. In Early Latin, there
are not many instances, except for forms of the present participle amans ‘someone in
love’ and plural forms of words like dictum ‘utterance’ and factum ‘deed’ (on which see
the Appendix in § 11.16). In Cicero, most substantival participles are masculine plural.26
Substantival participles can be used in any function in their clause, as the examples
below show. For a substantival participle as the subject of an ablative absolute clause,
see (g). As for the functions in which substantival participles are used by Cicero, a
distinction must be made between present and perfect participles. The former are
rarely used as subject or object, presumably because they might easily be interpreted
as secondary predicates. Perfect participles, on the other hand, are frequently found as
subject or object, with the genitive being especially rare.
Substantival participles may be modified in various ways. Note omnes in (e), quo in
(g), and, less common, the adjectives lascivi and expediti in (b) and (f ), respectively.
(a) Certo enim ego vocem hic loquentis modo mi audire visus sum.
(‘I’m sure I heard the voice of someone speaking here just now.’ Pl. Aul. 811)
(b) (saepimentum) . . . praetereuntis lascivi non metuet facem ardentem.
(‘A hedge has nothing to fear from the flaming torch of a mischievous passer-by.’
Var. R. 1.14.1)
(c) Quin solus omnium post memoriam humani <generis> supplicia in post
futuros conposuit . . .
(‘But more than that, he alone of all within the memory of man has devised
punishment for those yet unborn . . . ’ Sal. Hist. 1.55.6—tr. McGushin)
(d) Etenim quid tam est commune quam spiritus vivis, terra mortuis, mare
fluctuantibus, litus eiectis?
(‘For what is so common as breath to the living, earth to the dead, the sea to those
tossed by the waves, the shore to those cast up by the sea?’ Cic. S. Rosc. 72)
(e) . . . infra omnis mortuos amandatus esset.
(‘ . . . he had been removed to a lower depth than all the dead.’ Cic. Red. Pop. 10)
25
For examples, see K.-St.: I.222–33. For discussion, see Arias (1996).
26
For substantival participles in Cicero, see Laughton (1964: 70–83).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Substantival participles can be used with arguments and/or satellites to form participial
phrases, in the same way as when these verbs are used in finite clauses. The number of
such constituents is normally limited. The more complex phrases are especially found
in authors who use a compact form of presenting their information, like historians and
technical writers like Vitruvius (as in (k), but see also (h) from Cicero). In fact, more
than half of substantival masculine plural perfect participles in Vitruvius are used with
arguments and/or satellites. These are less common with present participles, probably
because relatively many of the present participles are genitival attributes, which would
become too heavy with arguments and satellites added.
(h) . . . huic se etiam summis honoribus usi contra suam dignitatem venditabant.
(‘ . . . even persons who had filled the highest offices sold themselves to him to the
detriment of their own dignity.’ Cic. Phil. 8.28)
(i) Cuius operis maxima parte effecta eodem fere tempore missi ad Pompeium
revertuntur.
(‘The main part of this work having been carried out, about the same time the
messengers who had been sent to Pompeius return.’ Caes. Civ. 1.18.6)
(j) Missique a praetoribus Damaratam Hieronis et Harmoniam Gelonis
filiam . . . interfecerunt.
(‘And men sent by order of the praetors put to death Damarata the daughter of Hiero
and Harmonia the daughter of Gelo . . . ’ Liv. 24.25.11)
(k) Eorum autem cogitata {utiliter} {hominibus}{ad vitam explicandam} e plu-
ribus singula paucorum uti exempla ponam . . .
(‘I will propose, as examples taken from a great number, several conceptions of a few
thinkers which have helped the furnishing of human life . . . ’ Vitr. 9.pr.3)
As was observed above, many former substantival participles have been lexicalized
and have become nouns, for example amans ‘lover’ and sapiens ‘wise man’, dictum
‘word’, nupta ‘married woman’, torrens ‘torrent’ (= fluvius torrens ‘rushing stream’).
But, of course, amans, etc. may also be used synchronically as a participle. There are
situations in which it is hard to decide.27
27
They are summarized by Adams (1973a: 116–20). A historical overview is also found in Adams
(1973a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
28
For the increase of the use of the present participle as a true alternative for verbal nouns in Silver and
Late Latin, see Adams (1973a), who also has a full list of the present participles involved. For Livy, see
Riemann (1885: 85–8); for Pliny the Elder, see Önnerfors (1956: 124–6). Hintzen (1993: 95–114) has
quantitative data on present participles and participial phrases in general in Cicero’s orations.
29
See Langslow (2000: 345–52, 404–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
est animus inferentis vim quam vim arcentis. (Liv. 21.44.3); Nec speciem adulantis
expaveris. (Tac. Hist. 2.76.2); . . . ut nemini parere animus bene informatus a natura
velit nisi praecipienti aut docenti aut utilitatis causa iuste et legitime imperanti. (Cic.
Off. 1.13); Namque Thaumaci a Pylis sinuque Maliaco per Lamiam eunti loco alto siti
sunt . . . (Liv. 32.4.3); Alia enim dare debemus feneranti, alia colenti agrum, alia
negotianti, alia regum amicitias sequenti . . . (Sen. Ep. 94.14); . . . ut non modo incli-
nantem excipere aut stantem inclinare, sed etiam adversantem ac repugnantem, ut
imperator fortis ac bonus, capere possit. (Cic. de Orat. 2.187); . . . sic iacere convenit
ut maxime cubantem iuvat. (Cels. 8.25.4); Lectio quoque non omnis nec semper
praeeunte vel interpretante eget. (Quint. Inst. 1.2.12);
Plural: Qui (sc. sunt) sancti, qui religionum colentes, nisi qui meritam dis immorta-
libus gratiam iustis honoribus et memori mente persolvunt? (Cic. Planc. 80—NB:
apparently the only instance of a substantival present participle in Cic. orations); Soli
enim ratione utentes iure ac lege vivunt. (Cic. N.D. 2.154); Ergo et Pythius et omnes
aliud agentes, aliud simulantes perfidi, improbi, malitiosi. (Cic. Off. 3.60); Semperne
in oratore probando aut improbando volgi iudicium cum intellegentium iudicio
congruit? (Cic. Brut. 183); Eaeque (sc. grues) in tergo praevolantium colla et capita
reponunt. (Cic. N.D. 2.125); . . . cum sentimus prudentibus fideliterque suadentibus
idem videri. (Cic. Fam. 2.13.1); Haec herba adalligata morbum regium habentibus ita
ut spectari ab his possit sanare id malum traditur. (Plin. Nat. 27.66); . . . ut quo modo
vis morborum pretia medentibus, sic fori tabes pecuniam advocatis ferat. (Tac. Ann.
11.6.2); Ex Hispania fugientis se excepisse et supplicio adfecisse dicit. (Cic. Ver.
5.151); (sc. medici) . . . qui leviter aegrotantes leniter curant . . . (Cic. Off. 1.83); Aureae
virgae et albae vestes regentes equos adornabant. (Curt. 3.3.11); Nec enim in con-
stituentibus rem publicam nec in bella gerentibus nec in impeditis ac regum domi-
natione devinctis nasci cupiditas dicendi solet. (Cic. Brut. 45);
Appendix: Substantival present participles usually differ in meaning from verbal
nouns derived from the same verb (for example with the suffix -tor). Whereas verbal
nouns often ‘imply a permanent or habitual quality or function’, substantival parti-
ciples are ‘used of a quality shown, or a function exercized, in particular circum-
stances.’30 This is illustrated by (g) and (h). However, as the main text indicates,
participles tended to be used as fully equivalent to nouns, and, conversely, verbal
nouns were sometimes used for very specific situations, as in (i).31
(g) Ceteri non exornatores rerum, sed tantum modo narratores fuerunt.
(‘The rest did not embellish their facts, but were chroniclers and nothing
more.’ Cic. de Orat. 2.54)
(h) Probabilis autem erit si personis, si temporibus, si locis ea quae narrabuntur
consentient . . . si probitas narrantis significabitur . . .
(‘And the statement will be convincing if the facts are narrated in accord-
ance with the persons, the times and the places . . . , if it indicates honesty in
the speaker . . . ’ Cic. Part. 32)
30
Quotation from Laughton (1964: 74). Adams (1973a: 119–20) holds that stylistic considerations are
more important than semantic ones. In reaction, Langslow (2000: 351–2) stresses the semantic differences.
31
Verbal nouns in -tor are restricted to verbs that denote an action (Torrego 1996b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(i) Mettius ille est ductor itineris huius, Mettius idem huius machinator belli,
Mettius foederis Romani Albanique ruptor.
(‘It is Mettius yonder who led this march; Mettius, too, who contrived this
war; Mettius who broke the treaty between Roman and Alban.’ Liv. 1.28.6)
It is not always clear whether in a given situation a present participle is substantival
or a secondary predicate with the entity to which it is related understood from
the context. An example is (j), where lacrumantem is better taken as a secondary
predicate.
32
See Önnerfors (1956: 23–7) and Adams (1995b: 320).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
constituti. (Cic. Luc. 23); Ipse autem homo ortus est ad mundum contemplandum et
imitandum—nullo modo perfectus sed est quaedam particula perfecti. (Cic. N.D.
2.37); Malagma ad luxum, contusum, canis morsum . . . (Larg. 365);
. . . reliqua quae ex empto aut vendito aut conducto aut locato contra fidem
fiunt . . . (Cic. N.D. 3.74); Nam Sullam consulem de reditu eius legem ferentem ex
composito tribunus plebis C. Herennius prohibuerat. (Sal. Hist. 2.21); . . . qui volun-
tatem bonam in solutum accipit . . . (Sen. Ben. 7.16.4);
In the plural, while masculine and feminine substantival expressions, as in (e) and (f ),
are not uncommon, neuter forms are quite frequent, especially those that are distinct
from the masculine and feminine forms. Examples are (g) and (h)—the latter a rare
dative form.
(e) Altera promulgata lex est ut et de vi et maiestatis damnati ad populum
provocent, si velint.
(‘Another law has been promulgated permitting persons convicted of violence and
treason to appeal to the people, if they wish.’ Cic. Phil. 1.21)
(f ) Nactus es ex perditis atque ab omni non modo fortuna, verum etiam spe
derelictis conflatam improborum manum.
(‘You have acquired as allies a band of evil men, swept together from men depraved
and abandoned not only by all fortune but even by hope.’ Cic. Catil. 1.25)
(g) Eidemque multa ex aliis causis aliquando a se acta, multa ab aliis audita
meminerunt.
(‘And they can also remember from other cases many arguments which they have
previously advanced and many which they have heard from other people.’ Cic. de
Orat. 2.355)
(h) Proxime accedere illum qui alterius bene inventis obtemperet.
(‘Next comes the man who accepts the good ideas of another.’ Cic. Clu. 84)
Supplement:
Plural: Masculine/feminine (ordered by case): (sc. partitio) . . . quam omnes ab
Aristotele et Theophrasto profecti maxime secuti sunt. (Cic. Inv. 1.61); . . . cum
Pythagoras atque eum secuti acceptam sine dubio antiquitus opinionem vulgaverint . . .
(Quint. Inst. 1.10.12); . . . tabulam in qua nomina civitate donatorum incisa essent
revelli iussisset . . . (Cic. Fam. 13.36.1); . . . at eius filii nec vivi probabantur bonis et
mortui numerum optinent iure caesorum. (Cic. Off. 2.43); . . . cum alios iudicio iniquo
circumventos convenirem! (Cic. Tusc. 1.98); Milites Romani perculsi tumultu insolito
arma capere alii, alii se abdere, pars territos confirmare, trepidare omnibus locis. (Sal.
Jug. 38.5); Item comitiali morbo correptos et caligine impeditos ex magna parte levat.
(Larg. 6); . . . hoc legale genus Hermagoras atque eum secuti vocant . . . (Quint. Inst.
3.5.4—NB: coordination); . . . et a conrogatis laudantur etiam quae non placent.
(Quint. Inst. 10.1.18);
Neuter: . . . tanta memoria praeteritorum futurorumque prudentia . . . (Cic. Sen. 78);
Qua re qui consumpta replere, erepta reciperare vellent . . . (Cic. Mur. 50); Vitabit
etiam quaesita nec ex tempore ficta sed domo allata quae plerumque sunt frigida.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(Cic. Orat. 89); Ambustis quoque prosunt recentibus . . . (Plin. Nat. 20.66); Sole
adusta sebo asinino aptissime curantur . . . (Plin. Nat. 28.222);
Appendix: Words like dictum, factum, and responsum are very common from the
earliest texts onwards (instances of dictum fill two columns in Lodge’s Plautus
lexicon), especially in the plural. They may be modified by all sorts of modifiers,
including adjectives, and are best regarded as deverbal nouns. Formally identical
substantival participles may be combined with satellites of various types. This results
in complex noun phrases such as those in (i)–(k).
(i) . . . multaque multorum facete dicta, ut ea quae a sene Catone collecta sunt,
quae vocantur IçŁªÆÆ.
(‘ . . . and many witty sayings of many men—like those collected by old Cato
under the title of Apophthegms.’ Cic. Off. 1.104)
(j) Multa eius et in senatu et in foro vel provisa prudenter vel acta constanter vel
responsa acute ferebantur.
(‘ . . . the many well-known instances wherein both in Senate and forum he
displayed shrewdness of foresight, resolution of conduct, or sagacity in
reply.’ Cic. Amic. 6)
(k) Lacrumans obtestatur per amicitiam perque sua antea fideliter acta ne super
tali scelere suspectum sese haberet.
(‘Bursting into tears, he begged the king by his friendship and his own
faithful service of old not to suspect him of such a crime.’ Sal. Jug. 71.5)
33
For Cicero’s usage, see Laughton (1964: 123–4). For Livy and other historians, see Riemann (1885:
90). For Seneca, see Westman (1961: 163–86).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Itaque, ut decet certae spei plenos et cum imparibus manus conserturos, pilis
ante pedes positis gladiis tantum dextras armemus.
(‘And so, as befits men who with sure confidence are about to fight with those who
are no match for them, let us lay our javelins at our feet and arm our right hands with
swords only.’ Liv. 6.12.8)
Supplement:
Masculine/feminine (ordered by case): Utinam qui divitias optaturi essent cum
divitibus deliberarent. Utinam honores petituri cum ambitiosis et summum
adeptis dignitatis statum! (Sen. Ep. 1.6.2); Hoc adicio ut dicturus intueatur cui,
apud quem . . . dicendum sit. (Quint. Inst. 4.1.52); Mire enim auditurum dicturi
cura delectat . . . (Quint. Inst. 11.3.157); (sc. Caesar) . . . ambiguus an urbem intraret,
seu, quia contra destinaverat, speciem venturi simulans. (Tac. Ann. 6.1.1); . . . ut per
nivem ituris bibere id perunguique eo praecipiant. (Plin. Nat. 26.117); . . . an aliquid
futuris reliquisset vetus avaritia. (Sen. Nat. 5.15.1); Tam sollicite recitaturis provi-
dendum est . . . (Plin. Ep. 6.15.4); . . . cuius id sit decus apud iudicaturos magno
certamine adfectant. (Mela 2.19); Non enim doceo, sed admoneo docturos. (Quint.
Inst. 1.4.17); . . . ne pari taedio lecturos adficerent verentur. (Tac. Ann. 6.7.5);
Neuter: Si omne futurum ex aeternitate verum est . . . (Cic. Fat. 32); Tuas (sc. litteras)
igitur exspecto, nec actorum solum, sed etiam futurorum. (Cic. Att. 15.4a.1); Acci-
pimus peritura perituri. (Sen. Dial. 1.5.7); Neque enim quicquam habet in se huius
materiae tractatio pulchrius, cum multa habeat futura usui, quam quod hominem
magnificentia sui detinet . . . (Sen. Nat. 6.4.2);
34
For more instances and references, see Aalto (1949: 104–5).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Ellipsis of a head noun that is modified by another noun phrase in the genitive is not
uncommon if the head noun can easily be inferred from the meaning of the noun in
the genitive or from other elements in the context. A clear case is the names of gods in
the genitive, which developed into an idiomatic expression (for which see § 11.20).
Other examples are the ellipsis of something like ‘book’, when referring to a book by a
certain author or about a certain subject, as in (a), or of the noun for ‘gardens’,
‘house’, or ‘baths’ of a specific person, as in (b) and (c), with the ellipsis resulting in
the substantival use of possessive genitives (for this use of the genitive, see § 11.47).
Less common is the substantival use of partitive genitives, as in (d) and (e) (for this
use of the genitive, see § 11.50–11.55).
(a) Eum video in Libonis praetorem P. Popilio P. Rupilio <consulibus>.
(‘I see in Libo that he was Praetor in the Consulship of P. Popilius and P. Rupilius.’
Cic. Att. 13.32.3, cf. in Libonis annali in Att. 13.30.2)
(b) Imago eius posita est in Quintiliorum in una tabula quinquiplex . . .
(‘His portrait was placed in the house of the Quintilii, representing him in five ways
on a single panel . . . ’ Hist. Aug. Tac. 16.2)
(c) TU·QUI LEGIS·VADE·IN·APOL<L>INIS / LAVARI.
(‘You who read this, go wash in the baths of Apollo.’ CIL XIII.1983.13 (Lyon, III AD))
(d) Certe satis constat . . . passim . . . venire victimarum, cuius adhuc rarissimus
emptor inveniebatur.
(‘There is no doubt that . . . sacrificial victims are on sale everywhere, though up till
recently scarcely anyone could be found to buy them.’ Plin. Ep. 10.96.10—NB: this is
a unique instance of a partitive genitive as subject; the text is usually emended.)35
(e) . . . item Macedones, Thessalos ac reliquarum gentium et civitatum
adiecerat . . .
(‘ . . . he had added . . . also Macedonians, Thessalians, and men of other nations and
states . . . ’ Caes. Civ. 3.4.6)
35
For this text and discussion in general, see Rosén (1986).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Noun phrases in the genitive denoting a god or goddess to whom a certain temple
(aedes, fanum, or templum) is dedicated are found without a head noun from Terence
onwards. The most common use is in prepositional phrases, as in exx. (a) and (b), but
there are also instances of genitives without a preposition, as in (c).
(a) Ubi ad Dianae veneris, / ito ad dextram.
(‘When you get to the temple of Diana, go right.’ Ter. Ad. 582–3)
(b) Nam ad me P. Valerius . . . scripsit . . . quem ad modum a Vestae ad Tabulam
Valeriam ducta esses.
(‘P. Valerius . . . has written to me an account . . . of how you were taken from the
Temple of Vesta to the Tabula Valeria.’ Cic. Fam. 14.2.2)
(c) Delata confestim materia omnis infra Veliam, et, ubi nunc Vicae Potae
(aedes add. edd.) est, domus in infimo clivo aedificata.
(‘Immediately the materials were all brought down below the Velia, and the house was
erected where the temple of Vica Pota is now, at the bottom of the slope.’ Liv. 2.7.12)
Supplement (in alphabetical order by preposition):
Intumus circus ad Murciae vocatur (Var. L. 5.154); . . . ut primo pugnatum ad Spei sit
aequo Marte, iterum ad portam Collinam. (Liv. 2.51.2);36 SELLAE / CURULIS·LOCUS . . . /
AD MURCIAE . . . DATUS / EST (CIL XI.1826.11–14 (Arezzo, early I AD)); In foro L. Antoni
statuam videmus, sicut illam Q. Tremuli, qui Hernicos devicit, ante Castoris. (Cic.
Phil. 6.13); Ad ea autem quae scripsisti . . . valde tibi adsentior, si multum possit
Octavianus, multo firmius acta tyranni comprobatum iri quam in Telluris . . . (Cic.
Att. 16.14.1); . . . IN PALATIO·IN DIVORUM . . . (CIL VI.2087.4 (Rome, II AD?)); Data venia
seducit filiam ac nutricem prope Cloacinae ad tabernas . . . (Liv. 3.48.5); Is est hodie
locus saeptus religiose propter Iovis pueri . . . (Cic. Div. 2.85);
There are a few fully convincing instances of prepositional phrases used substantively.
In (a), infra secundos has the same function as Homero and secundis: both secundis
and infra secundos are modified by horum.37 In (b), the prepositional phrase cum
armis aureis is determined by illum.38 Substantival prepositional phrases without
such contextual support are very rare, but see (c). In (d), ab epistulis ‘secretary’ and a
rationibus ‘treasurer’ are restrictive appositions with the proper names.39
36
For more examples, see TLL s.v. ad 522.46ff.; for the use of ad, see also Väänänen (1977).
37
See the discussion in Lebreton (1901a: 90–1). Horum . . . secundis is regarded as a Graecism by Kroll
ad loc. (E ø ıæØ).
38
For instances with ille, see TLL s.v. 356.32ff.
39
For prepositional phrases denoting public and private servants, see Väänänen (1977).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) An in poetis non Homero soli locus est . . . aut Archilocho . . . sed horum vel
secundis vel etiam infra secundos.
(‘Or among the poets there is room, not only for Homer or Archilochus . . . , but also
for those of a second or even lower rank.’ Cic. Orat. 4)
(b) Nempe illum dicis cum armis aureis / quoius tu legiones difflavisti spiritu . . .
(‘You mean the one with golden armour of course, whose legions you scattered with
a breath . . . ’ Pl. Mil. 16–17)
(c) Haud amice facis, qui cum onere offers moram.
(‘You’re not behaving in a friendly way since you cause me delay when I’m carrying a
burden.’ Pl. Poen. 852)
(d) (sc. Claudius suspexit) Sed ante omnis Narcissum ab epistulis et Pallantem a
rationibus . . .
(‘But most of all he was devoted to his secretary Narcissus and his treasurer Pallas . . . ’
Suet. Claud. 28.1)
Supplement:
. . . nostri, inquam, illi a Platone et Aristotele, moderati homines et temperati, aiunt
apud sapientem valere aliquando gratiam. (Cic. Mur. 63); ‘Non enim possum quin
exclamem’ ut ait ille in Trinummo. (Cic. de Orat. 2.39); Illa de ratione nummaria
non sunt eius modi (saepe enim audio ex ipso) ut non cupiat tibi praestare et in eo
laboret. (Cic. Att. 10.11.2); . . . quae sibi quisque facilia factu putat, aequo animo
accipit, supra ea veluti ficta pro falsis ducit. (Sal. Cat. 3.2); Adiunctis de suis
comitibus locum tutum reliquit. (Nep. Ag. 6.3); Desine, dulcium / mater saeva
Cupidinum, / circa lustra decem flectere mollibus / iam durum imperiis. (Hor.
Carm. 4.1.4–7); . . . amicum, / praesertim cautum dignos assumere, prava / ambitione
procul. (Hor. Sat. 1.6.50–2); . . . quia corpore in uno / frigida pugnabant calidis,
umentia siccis, / mollia cum duris, sine pondere habentia pondus. (Ov. Met.
1.18–20);40 RUFUM D(UUM)V(IRUM) I(URE) D(ICUNDO) A·BALNEO / ROG(ANT). (CIL
IV.7802.2–3 (Pompeii)); . . . occisis non modo e plebe, <s>ed militibus et centurione . . .
(Tac. Ann. 1.77.1); RESTITUTIANUS·CORNE/LIANUS·DE· XVI·A<B> AER(ARIO) . . . (CIL
XIV.4285.4–5 (Portus, end II AD)); Infit illa cum gladio: ‘ . . . ’ (Apul. Met. 1.12.4);
Qualis ille e limo, tales et qui de limo. (Itala I Cor. 48 = Cypr. ad Quirin. 3.11);
Dederunt nobis presbyteri loci ipsius eulogias, id est de pomis quae in ipso monte
nascuntur. (Pereg. 3.6);41
40
For this example, see Bömer ad loc.
41
For substantival use of partitive de phrases, see Väänänen (1987: 38).
42
TLL. s.v. ille 356.27ff. also quotes . . . ut illi paulo ante dicebant . . . (Sen. Nat. 6.14.2), but there is no
need to do so (illi is a determiner of the relative clause).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Attributes
Rare, but completely understandable is the use of various other expressions as head,
like the relative clause in (a), with the neuter determiner hoc.43
(a) Quibus autem hoc ‘qua de re agitur’ continetur, ea continentia vocentur . . .
(‘The arguments by which this “question at stake” is supported may be called
continentia (supports).’ Cic. Top. 95)
Supplement:
A duobus similibus similiter declinantur, ut fit in his nemus olus, nemora olera. (Var. L.
10.50);44 Quintia formosa’st multis. . . . Totum illud ‘formosa’ nego. (Catul. 86.1–3);
11.24 Attributes
In § 2.4 the term ATTRIBUTE is defined as the constituent of a noun phrase that is
optional with respect to the head of that phrase. Thus in (a), the genitival noun phrase
hominis nequam et improbi is optional with respect to parasitus: without the genitival
phrase the information value of the remaining words is slight, but structurally the
genitival noun phrase is optional. Similarly, within that genitival noun phrase the
coordinated adjectives nequam et improbi are optional with respect to hominis.
In certain configurations constituents that would in other contexts be regarded as
optional cannot be easily omitted without making the remaining utterance ungram-
matical and/or nonsensical. This is shown in (b) and (c). In (b), the noun phrase
infenso animo atque inimico functions as a secondary predicate (it is a so-called
ablativus qualitatis). Unmodified nouns in the ablative cannot normally function as
such.45 In (c), an instance of the possessive dative construction, compti can likewise
not easily be omitted. In none of these examples can the constituents under discussion
be said to be required by the meaning of the head noun.
43
For all sorts of expressions used substantivally in Cicero, see Lebreton (1901a: 84–91).
44
For Varro’s preference of hic in such expressions, see Rosén (1994: 132–3).
45
For exceptional unmodified animo in Cic. Pis. 23, see Pinkster (1982: 252).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Attributes
As stated in § 11.3, the combination of a noun with its adnominal arguments (and
satellites) can be modified by all sorts of optional modifiers, which is an indirect proof
of their different syntactic status. The two types of modifier can be shown to behave
differently in other respects as well. (i) The adnominal arguments have less freedom
in their position than optional modifiers: they tend to stand next to their head noun
(see Chapter 23). (ii) The relation between an optional attribute and its head noun can
be paraphrased in a subject + subject complement, or an object + object complement
expression; this is not possible in the case of adnominal arguments: see Table 11.246
Table 11.3 gives a survey of the main types of optional modifiers. The individual
sections are especially intended to illustrate combinations of multiple attributes with
the same head noun.
Table 11.3 Types of constituents that may function as modifier in a noun phrase
Type of modifier Examples For more
examples,
see §
Determiner
(i) demonstrative . . . ut sciretis nomen huius fabulae. 11.26
or anaphoric (‘ . . . that you should know the name of this play.’ Pl. As. 7)
(ii) indefinite Malum quod tibi di dabunt, sic scelestu’s. 11.27
(‘The gods will give you some trouble, you’re such a
criminal.’ Pl. Ps. 1130)
(iii) interrogative Quinam homo hic ante aedis nostras eiulans conqueritur 11.28
maerens?
(‘What man on earth is lamenting here in front of our
house, with wailing and sadness?’ Pl. Aul. 727)
(iv) relative Sentio, / nam qua nocte ad me venisti, eadem abis. 18.15
(‘I can feel that: the same night you’ve come to me you’re
going away.’ Pl. Am. 531–2)
(Continued )
46
For this test, see Bolkestein (1989: 13–14).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Determiners
The following sections start with attributes that show agreement with their head.
Nouns and noun phrases marked by the genitive or another case, prepositional
phrases, and adverbs will follow.47
(a) Est quidam homo qui illam ait se scire ubi sit. / # At pol ille a quadam
muliere, si eam monstret, gratiam ineat. / # At sibi ille quidam volt dari
mercedem. # At pol illa quaedam, / quae illam cistellam perdidit, quoidam
negat esse quod det. / # At enim ille quidam o<peram bonam magis>
expetit quam argentum. / # At pol illi quoidam mulieri nulla opera
gratuita est.
(‘There’s a certain man who says he knows where the casket is. # But if
he were to show it, he’d receive thanks from a certain woman. # But that
certain man wants to be given a reward. # But that certain woman who
lost that casket says she doesn’t have anything to give to the certain man.
# But that certain man is looking for a favour rather than money. # But
with that certain woman no favour goes without a reward.’ Pl. Cist.
735–40)
There are a few instances of a sequence hic ille in which at first sight hic seems to
determine ille, as in (b). In fact, however, this is a clever play on hic (est) ille
combinations, where ille has its exophoric function ‘the well-known’.
47
For a very complete description of the various types of attributes and the way they can be
coordinated in Tacitus, see Sánchez Martínez (2000: 189–344).
48
On determiners, see Orlandini (1995: passim); Touratier (2010: 131–2).
49
For a few more instances, see Lodge s.v. quidam § II B. For ille in combination with free-choice
indefinite determiners, see TLL s.v. ille 360.22ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Hunc illum fatis externa ab sede profectum / portendi generum paribusque
in regna vocari / auspiciis . . .
(‘ “This,” he thought, “must be he who, coming from a stranger’s home,
is predestined by the fates as my son-in-law, and called to sovereignty with
equal power . . . ” ’ Verg. A. 7.255–7)50
Appendix: A noun phrase with the determiner aliqui(s) can be modified by the
numeral unus and by alter, as in (c) and (d).
(c) Fortasse unum aliquod verbum inter eas iram hanc concivisse.
(‘It’s likely it was a single word which stirred up this quarrel between them.’
Ter. Hec. 313)
(d) Postremo praeceptoris est docere quemadmodum quaeri quidque conveniat
et unum aliquod aut alterum, non omnia, quae eius generis erunt, exempli
causa subicere . . .
(‘Finally, it is the instructor’s duty to teach the proper method of search in
each case, and . . . to add in illustration some one or two examples of its kind,
but not all.’ Rhet. Her. 3.39)
Latin has three demonstrative pronouns/determiners (hic, iste, and ille) and one
anaphoric pronoun/determiner (is). The demonstrative pronouns/determiners
have several uses, which vary for each individual item. Their best-known use is to
situate the entity they modify in space or time with respect to the speech participants,
the DEICTIC (or ‘demonstrative’ sensu stricto) use. In this case the entities involved are
directly perceivable in the communicative situation. However, they may also be
accessible for the addressee in some other way, either on the basis of the preceding
discourse (the anaphoric use) or on the basis of his knowledge of the world at large
(the EXOPHORIC or ‘encyclopedic’ use), especially ille. In their deictic use hic and iste
indicate that the entity is related in some way to the domain of the speaker and
the addressee, hic ‘this here’ being related to the speaker, iste ‘that (near you)’ to
the addressee. Ille ‘that (over there)’ indicates that the entity does not belong to the
domain of the speaker or addressee. The determiner is ‘this’ can be used as an
ANAPHORIC determiner to indicate that the entity is known from the preceding
discourse. Further details concerning the various functions these words fulfil when
functioning as determiners are discussed in § 11.103. The use of these words with
autonomous relative clauses is discussed in § 18.15.
Examples of deictic use of the demonstrative determiners are (a)–(c). Ex. (d) shows
the anaphoric use of is: eam indicates that the entity rem refers to is present in the
preceding discourse: in fact, it refers to the action described in the preceding sentence.
Examples of the anaphoric use of demonstratives are given in § 11.105. Examples of
50
See Horsfall ad loc. and Koestermann ad Tac. Ann. 14.22.2.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Determiners
these determiners modifying a noun phrase that contains one or more other modifiers
are given in the Supplement.51
(a) Haec urbs est Thebae. In illisce habitat aedibus / Amphitruo . . .
(‘This city is Thebes. In that house there lives Amphitruo . . . ’ Pl. Am. 97–8)
(b) Perii hercle, lembus ille mihi laedit latus.
(‘I’m done for. That fast-sailer is ramming me amidships.’ Pl. Bac. 281)
(c) Pater istius adulescentis dedit has duas mi epistulas, / Lesbonici.
(‘The father of that young chap, Lesbonicus, has given me these two letters.’ Pl. Trin.
894–5)
(d) Et meus pater nunc intus hic cum illa cubat, / et haec ob eam rem nox est
facta longior . . .
(‘And my father’s now lying with her inside the house, and for that very reason this
night’s been made longer . . . ’ Pl. Am. 112–13)
Supplement:
Nam hoc primum iter designatus rei publicae causa suscepit. (Cic. Sest. 71); Post
Aristoteles inlustravit omnem hunc c i v i l e m in disputando locum . . . (Cic. Leg.
3.14); Ac Romulus cum septem et triginta regnavisset annos et haec e g r e g i a duo
firmamenta rei publicae peperisset . . . (Cic. Rep. 2.17);
Illa in Pontum profectio et conatus tuus, quid? (Cic. Pis. 88); Itaque primum illum
actum istius vitae turpissimum et flagitiosissimum praetermittam. (Cic. Ver. 1.32);
Nam opinor <hoc> illi alteri flumini nomen esse. (Cic. Leg. 2.1); . . . cum in illo i p s o
acerbissimo miserrimoque spectaculo sibi timerent . . . (Cic. Tusc. 3.66);
Doce te petisse ab eo istam nescio quam innumerabilem pecuniam . . . (Cic.
Quinct. 37); Neque eos tam istius hominis perditi subita laetitia quam hominis
amplissimi nova gratulatio commovebat. (Cic. Ver. 21); Equidem in omnibus istis
conclusionibus hoc putarem philosophia nobisque dignum . . . (Cic. Fin. 4.52.4);
Quibus ex locis ad eas tris res quae ad fidem faciendam solae valent ducatur
oratio . . . (Cic. de Orat. 2.121); . . . sic in omnibus iis rebus quae coniectura investi-
gari videntur anceps reperitur oratio. (Cic. Div. 2.55);
Appendix: It appears that demonstrative and anaphoric determiners (and the other
determiners) may determine noun phrases that contain all sorts of attributes, but that
they cannot be part of a noun phrase that is then further modified. The only
exceptions seems to be when that modifier is omnis ‘each’, ‘all’, cunctus ‘all’, or
totus ‘entire’, as shown in (e)–(g).
51
More examples can be easily found in Merguet (Phil.)’s lemmata of these determiners in the section
‘attributiv’ ‘mit Zusatz’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The indefinite determiners aliqui ‘some’, ‘a’, qui ‘a’, quispiam ‘one or other’, quidam ‘a
certain’, nescioquis (also nescio quis) ‘some’, quisquam ‘any (single)’, and ullus ‘any at
all’ assert in various ways the existence of an entity. They can be combined with a
variety of other modifiers, as is shown in the examples below and in the Supplement.
Their syntactic behaviour is diverse. For details about the individual determiners, see
§§ 11.109–11.114. For free-choice indefinite determiners, see § 11.115.
Supplement:
Aliusne est aliquis improbis civibus peculiaris populus, cui nos offensi invisique
fuerimus? (Cic. Sest. 125); Si qua enim sunt privata iudicia summae
existimationis . . . (Cic. Q. Rosc. 16); Atque ille etiam si alia quapiam vi expelle-
retur illis temporibus ex hac urbe, facile pateretur. (Cic. Sest. 63); Ardeo . . .
incredibili quodam amore patriae . . . (Cic. Prov. 23); Nescio quem, mulier,
alium hominem, non me quaeritas. (Pl. Men. 406); Neque enim ratio adferri
potest cur, si cuiquam novo civi potuerit adimi civitas . . . (Cic. Caec. 101); Hunc
hominem ferum atque agrestem fuisse, numquam cum homine quoquam con-
locutum esse, numquam in oppido constitisse. (Cic. S. Rosc. 74); Non istanc
aetatem oportet pigmentum ullum attingere, / neque cerussam neque melinum,
neque aliam ullam offuciam. (Pl. Mos. 262–3);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Determiners
Zero quantifiers, also called negative indefinite pronouns, are also used as determiners:
nemo, nihil, nullus ‘no’, and neuter ‘neither’. They can be combined with several other
types of modifiers, as in (f )–(l). Note the combination nulla duo (sc. verba) in (k).
Nullus unus, nemo unus, and nihil unum (with substantival unus) are possible as well.52
The combination of nemo and homo is found from Early Latin onwards.
(f ) Hominem ego iracundiorem quam te novi neminem.
(‘I don’t know anyone more prone to anger than you.’ Pl. Mer. 141)
(g) . . . ita fuit inlustre notumque omnibus ut nemo tam rusticanus homo . . .
Romam ex ullo municipio vadimoni causa venerit quin . . .
(‘ . . . a fact so notorious, so well known to everyone, that . . . not even the simplest
rustic from the remotest provincial town came to Rome, in connexion with any case
heard in the law-courts, without . . . ’ Cic. Ver. 5.34)
(h) Nam mihi nil relicti quicquam aliud iam esse intellego.
(‘I understand that I have nothing else left.’ Pl. Mer. 666)
(i) Nullam ego me vidisse credo magis anum excruciabilem / quam illaec est.
(‘I don’t think I’ve seen an old woman deserving torture more than that one does.’ Pl.
Cist. 653–4)
(j) Sed <opsecro> te, nullusne est tibi amator alius quisquam?
(‘But please, don’t you have any other lover?’ Pl. Cist. 369)
(k) Non quo multa parum communis littera currat / aut nulla inter se duo sint ex
omnibus isdem . . .
(‘I do not say that there are very few common letters running through all, or that no
two words, if compared, are made up of elements all the same . . . ’ Lucr. 2.692–3)
(l) Neutra in re vobis difficultas a me erit.
(‘In either case you won’t get any problems from me.’ Ter. Hec. 667)
The negative indefinite pronouns nemo and nihil deny the existence of an entity that is
in some way involved in the state of affairs the clause refers to. Nemo may serve as an
answer to a question with quis ‘who’ that asks for the identity of an entity. This is shown
in (m). Something similar can be said about nullus. At the same time, they function as
zero quantifiers, as the opposites of omnis ‘all’ and aliquot ‘several’, as in (n).
(m) Interrogavi fratrem: Apud quem praetorem causam dixisti? ‘Apud nullum’
inquit. Quis accusator fuit? ‘Nemo.’ Quis testis? Immo qui testes? . . . ‘Nemo’
inquit. Quis de te pronuntiavit? ‘Nemo.’
(‘I asked my brother before which praetor he had pleaded his case. “None,”
he said. “Who was the accuser?” “No one.” “Who was the witness?—or
rather, witnesses . . . ?” “No one.” “Who passed a sentence on you?” “No
one.” ’ Sen. Con. 7.1.23)
52
See Sz.: 205.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
53
For the notion of possession, see Baldi and Nuti (2010: 239–46). For the range of uses of the
possessive adjectives of the first and second person, see Joffre (1998), with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
the house where the slave lives and does his job. Possessive adjectives may function
either as an optional attribute, as just shown, or as an adnominal argument. In the
phrase profectio mea ‘my departure’, as in (b), the relationship between the speaker
and the departure corresponds to that of the agent in a finite clause, in the same way
as a genitive noun functions in relation to a noun like profectio (the so-called
subjective genitive—see § 11.71). By contrast, in a phrase like conspectus meus ‘the
sight of me’ meus is used like an objective genitive, as in (c).54
(a) Uno eodemque tempore domus mea diripiebatur, ardebat . . .
(‘At one and the same moment my house was being plundered, was ablaze . . . ’ Cic.
Red. Sen. 18)
(b) . . . dolorem profectionis meae reditus dignitate consoler.
(‘I should mollify the sorrow of my departure by the magnanimity of my return.’ Cic.
Red. Sen. 23)
(c) Eo, etsi scio pol is fore meum conspectum invisum hodie.
(‘I’ll go, though I know they’re going to hate the sight of me today, by heaven.’ Ter.
Hec. 788)
This parallelism with the genitive in the amplitude of possible relations also explains
instances of coordination, as in (d), and combinations of a possessive adjective with a
secondary predicate in the genitive, such as tuum . . . adulescentis ‘of you in your
youth’ in (e), unius mea ‘of mine alone’ in (f ), and tuas, Fanni, solius ‘of you, Fannius,
alone’ in (g).55 Another indication of their relatedness is their use as subject comple-
ments (see § 9.28 and 9.30).
(d) Mihi uni necesse erit et meam et aliorum vicem pertimescere?
(‘Am I alone to be compelled to fear vicariously for others as well as individually for
myself?’ Cic. Dom. 8)
(e) . . . sed moleste fero me consulem tuum studium adulescentis perspexisse . . .
(‘ . . . but it irks me that when I was Consul I saw proof of your good will when you
were a young man . . . ’ Cic. Fam. 15.13.1)
(f ) . . . id in unius mea salute conservanda decernendum putavit.
(‘ . . . it thought fit to decree for my salvation alone.’ Cic. Sest. 128)
(g) Paulo ante M. Perpennae, P. Saturi tabulas poscebamus, nunc tuas, C. Fanni
Chaerea, solius flagitamus . . .
(‘A little time ago we asked for the account books of Marcus Perpenna and Publius
Saturius; we now urgently call for yours alone, Gaius Fannius Chaerea . . . ’ Cic.
Q. Rosc. 3)
54
For more instances, see TLL s.v. meus 721.20ff.
55
For more instances, see TLL s.v. meus 921.73ff. On the status of possessive adjectives, see Risselada
(1984: 227, n. 23). For the genitive case in general, see § 12.21.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
If it is beyond reasonable doubt who the possessor of a certain entity is, the possessive
adjective is normally not expressed. This holds for ‘inalienable’ entities like relatives
(pater ‘father’) and body parts (caput ‘head’). If in such cases a possessive adjective is
used, it is often for reasons of emphasis or contrast.56
Supplement:
(sc. Plancius) . . . cui nomen meum absentis honori fuisset, ei meas praesentis preces
non putas profuisse? (Cic. Planc. 26); Postquam arma dei ad Volcania ventum est . . .
(Verg. A. 12.739); Et flesti et nostros (= meos) vidisti flentis ocellos. (Ov. Ep. 5.45);
Unlike possessive adjectives in English and Dutch, those in Latin can modify a noun
phrase that is itself modified by the determiners discussed in § 11.25.57 In (h), istas
determines vostras leges urbanas, where vostras modifies the noun phrase leges
urbanas in which the adjective urbanas is the attribute. In (i), the relative determiner
cuius determines mei sensus.58
(h) Nescio, neque ego is tas vostras leges urbanas scio . . .
(‘I don’t know, and I don’t know those city laws of yours . . . ’ Pl. Rud. 1024)
(i) C uiu s mei sensus certissimus testis est hic idem qui custos capitis fuit, Cn.
Plancius . . .
(‘Here again infallible testimony is borne to these convictions of mine by Gnaeus
Plancius, who acted as the guardian of my person . . . ’ Cic. Red. Sen. 35)
Quantifiers too can modify noun phrases containing a possessive adjective, as in (j),
where duobus modifies the noun phrase oculis meis, and in (k) and (l). More examples
are given in the Supplement. The indefinite determiner ulla in (m) can hardly be
interpreted differently. In this same example the identifier alia also seems to modify
culpa mea (the structure seems to be: ulla <alia [(mea) culpa]>). Similarly in (n) and
(o). In (p), nostris seems to be preposed for reasons of emphasis, the structure being:
multis [(nostris) peccatis].
(j) Profecto vidi. # Tutin’? # Egomet duobus hi s oculis meis.
(‘I’ve really seen her. # You yourself? # I myself with these two eyes of mine.’ Pl. Mil. 290)
(k) Id enim est caput civilis prudentiae, in qua omnis h a e c nostra versatur
oratio . . .
(‘For this is the foundation of political wisdom, with which our whole discourse is
concerned . . . ’ Cic. Rep. 2.45)
(l) Multa mea in se, non nulla etiam sua in me proferebat officia.
(‘He cited the many services that I had done him and some too that he had done me.’
Cic. Sul. 18)
56
See Lehmann (2005).
57
For the indefinite status of the possessive adjective, see Baldi and Nuti (2010: 336–41). See also
Touratier (1991: 237–40).
58
For data on the relative order of determiners and possessive adjectives, see Iovino (2012: 94, 135).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(m) Magis adeo id facilitate quam alia ulla culpa mea contigit . . .
(‘Indeed it happened more by good nature than by any other fault of mine . . . ’ Cic. de
Orat. 2.15)
(n) At te eadem tua fortuna servavit.
(‘But that same “good luck” of yours preserved you.’ Cic. Deiot. 21)
(o) Ego ibo ad fratrem ad alios captivos meos . . .
(‘I’ll go to my brother’s to other prisoners of mine.’ Pl. Capt. 126)
(p) Scio nos nostris multis peccatis in hanc aerumnam incidisse.
(‘I know that I have fallen in this unhappy plight by my own mistakes, which have
been many.’ Cic. Att. 3.14.1)
Supplement:
. . . nullum meum minimum dictum, non modo factum, pro Caesare intercessit . . .
(Cic. Fam. 1.9.21); Reperti sunt complures nostri milites (om. b) qui in phalangem
insilirent . . . (Caes. Gal. 1.52.5); Ipse arcano cum paucis familiaribus suis
conloquitur . . . (Caes. Civ. 1.19.2); Nostrae naves duae tardius cursu confecto in
noctem coniectae . . . contra Lissum in ancoris constiterunt. (Caes. Civ. 3.28.1);
The use of the genitives of the personal pronouns nostrum and vostrum instead of the
corresponding possessive adjectives is rare. Examples are (q) and (r).
(q) Recordamini qui dies nudius tertius decimus fuerit, quantus consensus ves-
trum, quanta virtus, quanta constantia.
(‘Cast your minds back twelve days: what a day that was, what unity, what courage,
what constancy you displayed!’ Cic. Phil. 5.2)
(r) Is enim splendor est vestrum ut eadem postulentur a vobis quae ab amplis-
simis civibus.
(‘Your distinction is such that as much is expected from you as from the highest in
the land.’ Cic. Att. 7.13.3)
The use of the genitive of a third person reflexive pronoun (seemingly) instead of a
reflexive possessive adjective is first attested in (s), Cicero’s translation of Plato. In (t)
the content of what follows may explain its use. The same may be the case in (u).
However, in later authors, such as Seneca, Pliny, Tacitus, who has several instances,
and Apuleius, such contextual support often seems to be lacking.59 Nostri is first
attested in (w). (NB: for these forms as so-called objective genitives, see § 11.71,
Appendix.)
59
For further examples, see K.-St.: I.598–9. For Pliny the Elder, see p. I.544 of Mayhoff ’s edition. For
Seneca, see Axelson (1939: 69–70). For Apuleius and other authors, see Callebat (1968: 262–4). For
references to earlier publications, see Sz.: 61.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(s) Neque enim laetabitur umquam nec maerebit nimis, quod semper in se ipso
omnem spem reponet sui.
(‘For never will he give way to transports of either joy or grief, because he will always
base every hope he forms upon himself alone.’ Cic. Tusc. 5.36, cf. Plato Menex. 248A
Øa e Æ fiH ØŁ ÆØ)
(t) Lysander Lacedaemonius magnam reliquit sui famam, magis felicitate quam
virtute partam.
(‘Lysander the Lacedaemonian left a great reputation, gained rather by good fortune
than by merit.’ Nep. Lys. 1.1)
(u) Non aliud Scaevolae Mucio cognomen dedit et capto contra Porsennam
regem libertatem reliquit quam vilitas sui.
(‘What gave Mucius Scaevola his name, and left him free, though a prisoner, in the
face of King Porsenna, was his contempt for his own life.’ Sen. Con. 8.4.1)
(v) At Dolabella . . . primo sui incessu solvit obsidium . . .
(‘Dolabella, on the other hand, . . . raised the siege at his first advance . . . ’ Tac. Ann.
4.24.2)
(w) Non ullo gravius temptatur Cynthia damno / quam sibi cum rapto cessat
amore deus / praecipue nostri.
(‘No loss more keenly provokes Cynthia than when love has been stolen from her
and her god lies idle, especially when that love is mine.’ Prop. 1.4.25–7—decus cj.
Kraffert; nostro B)60
Supplement:
Nec haec vilitas sui est aut dementia pro uno capite tot milia excipere ferrum . . . (Sen.
Clem. 1.3.4); Locum enim sibi ad formam et habitum sui exculpit. (Sen. Nat. 4b.3.4);
Lanae et per se coactae vestem faciunt et, si addatur acetum, etiam ferro
resistunt, immo vero etiam ignibus novissimo sui purgamento. (Plin. Nat. 8.192);
(sc. Caesar) . . . adit castrorum vias, adsistit tabernaculis fruiturque fama sui . . . (Tac.
Ann. 2.13.1); Hunc sui finem multos ante annos crediderat Agrippina contempse-
ratque. (Tac. Ann. 14.9.3); Sed ipsam magnitudinem sui absconderat . . . (Tert. Marc.
4.25.2);
. . . teque ad tui similes musculos recondis, antequam nostri vim praesentariam
experiaris? (Apul. Met. 2.25.4);
Variously emended: (sc. naves) magno sui cum periculo deicerentur. (Caes. Gal.
4.28.2);
NB: Quod illi semper sui causa fecerant . . . quam diu intellegebant sese sibi et populo
Romano, non Verri et Apronio serere, impendere, laborare. (Cic. Ver. 3.121—instead
of sua causa);
60
On this much disputed passage, see Heyworth (2007: 22).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The third person possessive adjective is primarily used with a reflexive meaning,
indicating that the noun phrase it belongs to is coreferential with another constituent
of the clause, usually the subject, as in (a). The conditions of this reflexive use are
discussed in detail in § 11.130, together with the reflexive personal pronoun. How-
ever, it is also used in a variety of ways without any relationship with the subject.61 It
may have the meaning ‘characteristic’, ‘typical’, ‘convenient’, etc., depending on the
noun it modifies.62 Examples are (b) and (c). It is also used with an emphatic
possessive meaning ‘his/her/its/their own’ (usually taken as its original meaning), as
is shown in (d), where suam is in contrast with alienam.63 See also suus in this sense in
(e) and (f ). Suus is also the regular possessive adjective in combination with quisque
(whatever its function), as in (g). In all these cases a non-reflexive possessive modifier
(eius, for example) is excluded. In certain expressions it became idiomatic and lost its
third person element, as in (h), where the subject is first person plural. There are other
isolated examples in Late Latin.64
61
For these uses de Bruyne (1961) coined the term ‘pseudo-reflexief ’. He has a list of instances in
Cicero.
62 63
For examples, see OLD s.v. suus §§ 9–13. For this use of suus, see Fruyt (2008).
64 65
See Sz.: 176. On this example, see de Melo (2010: 83).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(g) Nunc adeo hoc factu est optumum, ut nomine quemque appellem suo . . .
(‘Now the best thing for me to do is to address each of you by name . . . ’ Pl. Ps. 185)
(h) Novare possumus aut ipsi, si sui iuris sumus, aut per alios qui voluntate
nostra stipulantur.
(‘We can novate in person, if we are independent, or through those who stipulate
through our will.’ Paul. dig. 46.2.20)
Suus is used for a singular and a plural possessor, whereas for the first and second
person distinct forms are available. This distinction is also available for the non-
reflexive third person substitute (eius : eorum/earum). In the Romance languages the
successors of suus are restricted to a singular possessor (and lack the reflexive
component).66 For the plural, non-reflexive genitives came to be used instead,
notably illorum (cf. Ital. loro, Fr. leur).
From Early Latin onwards (but not in Classical Latin) the reflexive possessive
adjective is used in combination with a dative personal pronoun. This is particularly
attested for the reflexive personal pronoun, with only one attestation of meus mihi.
Examples are (i)–(k). In none of these can the dative be interpreted as an argument or
a satellite at the clause level. The combination seems to be more emphatic than the
bare possessive. For substantively used suo, see (l).67
(i) In tabernam ducor devorsoriam, / ubi male accipiar mea mihi pecunia.
(‘I’m being led into an inn where I’ll get a bad reception from my own
money.’ Pl. Truc. 697–8)
(j) Ita nunc ignorans suo sibi servit patri.
(‘So now he’s his own father’s slave and doesn’t know him.’ Pl. Capt. 50)
(k) Uvas quae meridiem aut occidentem spectabunt, ne praeurantur, suo sibi
pampino tegito.
(‘Cover the bunches of grapes which face south or west with their own
tendrils, so that they may not be burnt by the sun.’ Col. Arb. 11.1)
(l) Et si tofus erit aut saxum, in suo sibi canalis excidatur . . .
(‘If the formation of the earth is of tufa or stone, the channel may be cut in
its own bed.’ Vitr. 8.6.3)
Supplement:
. . . suo sibi lautum sanguine tepido. (Acc. trag. 405); Suo sibi gladio hunc iugulo.
(Ter. Ad. 957); De suo sibi fricabis . . . (‘you will smear it by itself ’ Apic. 4.2.27);
Cf.: . . . sed solus sibi immotus Atreus constat . . . (Sen. Thy. 703);
Appendix: In its meaning ‘own’ suus resembles proprius, which, however, is not
restricted to the third person, as is shown in (m) (see also in (b) above propriis . . . ac
66
Some scholars think that there are forerunners of this development in Plautus, incorrectly. See de
Melo (2010: 89–99).
67
For more instances and references, see K.-St.: I.606. For the ‘pleonastic’ use of the dative personal
pronoun, see Dahlén (1964: 178–86) and, for Plautus, de Melo (2010: 80–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Identifiers
suis). An early instance of proprius in a purely possessive sense is (n). More clear-cut
instances of proprius = suus can be found in Late Latin, where it is also found in place
of sua in the idiom sua sponte, as in (o).68
(m) Sed persuade tibi de voluntate propria mea nihil esse remissum . . .
(‘But you may be sure that there has been no falling off in my personal
feelings . . . ’ Brut. ad Brut. 25.6)
(n) (sc. Tityos) . . . non tamen aeternum poterit perferre dolorem / nec praebere
cibum proprio de corpore semper.
(‘ . . . yet he will not be able to bear pain forever, nor to provide food from his
own body always.’ Lucr. 3.990–1)
(o) . . . sponte se propria dederunt.
(‘ . . . they surrendered of their own accord.’ Amm. 17.2.3)
68
For more examples of the purely possessive use of proprius, see TLL s.v. 2099.15ff. For a diachronic
overview, see Sz.: 179 and Moussy (2012c). For sponte + genitive (from Lucan onwards), see Kuntz (1962: 41).
69 70
For more examples, see TLL s.v. alius 1651.60ff. So, for example, Sz.: 188.
71
For more examples, see TLL s.v. idem 200.20ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
72
For a discusion of the literature on Latin quantifiers and their position, see Iovino (2012: 197–201,
221–6).
73
For the use of ad with numerals, see Torrego (1998) and Garcea (2012).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Quantifiers
Cardinal numerals can only be used with count nouns. Adjectives of amount (here
illustrated with multus) can be used with count nouns as well, usually in the plural, as
in (g). A (poetic) collective use of the sg. is (h). However, they can also be used in the
sg. with mass nouns, as in (i), and abstract nouns, as in (j). They are excluded with
collective nouns in the singular; adjectives of size are used instead, as in (k).
Both adjectives of amount and cardinal and distributive numerals may modify noun
phrases containing one or more adjectives, as in (l) and (m). These numerals may also
function at a lower level, the noun phrase to which they belong being modified, as in
(n) and (o).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(l) Supervenerunt his tres Rhodiae quadriremes et erant Atticae tres apertae
naves . . .
(‘In addition to these, three Rhodian quadriremes arrived, and there were three
Athenian vessels without protected decks . . . ’ Liv. 31.22.8)
(m) . . . dixit multos civis Romanos in lautumiis . . . morte esse multatos.
(‘ . . . he testified that many Roman citizens had been put to death . . . in the stone
quarries.’ Cic. Ver. 1.14)
(n) Primis quattuor legionibus populus tribunos creavit. In ceteras consules
miserunt.
(‘For the first four legions the people elected the tribunes. To the rest of the legions
tribunes were sent by the consuls.’ Liv. 27.36.14)
(o) Reliquae tres partes pronai ad antas parietum procurrant . . .
(‘The remaining three parts, constituting the pronaos, are to run forward to the antae
of the walls . . . ’ Vitr. 4.4.1)
(q) . . . scio scire te / quam multas tecum miserias mulcaverim. / Nunc hunc
diem unum ex illis multis miseriis / volo me Eleutheria capere advenientem
domum.
(‘ . . . I know that you know how many hardships I’ve forced down my throat in
dealing with you. Now for this one day in recompense for those many afflictions
I want to celebrate the Festival of Liberty on my arrival home.’ Pl. St. 419–22)
(r) . . . quod haud scio an gravissimum multis his annis huic ordini nuntiatum sit . . .
(‘ . . . which I am inclined to believe is the most solemn that has been announced to
this order for many years past . . . ’ Cic. Har. 10)
(s) Verum tamen ista multa iudicia quae sunt?
(‘But as for those many previous trials, what do they amount to?’ Cic. Clu. 88)
(t) . . . ut illam laudationem quam se vi ac metu coactos paucis illis diebus
decresse dicebant tollerent.
(‘ . . . to rescind the eulogy which, as they told me, they had been forced and
intimidated into decreeing only a few days before.’ Cic. Ver. 4.140)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Quantifiers
(v) Mihi de memet ipso tam multa dicendi necessitas quaedam imposita est ab illo.
(‘It was he who imposed upon me a kind of compulsion to say so much
about myself.’ Cic. Sul. 35)
Indefinite determiners cannot function as determiners of noun phrases containing an
adjective of amount, probably because indefinite determiners are by implication
understood as denoting (a low) number. See (w), where multus is contrasted with
aliqui.74 Aliqui can be combined with cardinal numerals, as in (x), in the sense of
‘round about’ (OLD s.v. § 9).
(w) Cur tam multos deos nihil agere et cessare patitur, cur non rebus humanis
aliquos otiosos deos praeficit . . .
(‘Why does it suffer so many gods to be idle and keep holiday? Why does it
not appoint some of the leisured gods . . . to superintend human affairs?’ Cic.
N.D. 3.93)
(x) Quid nunc vis? # Ut opperiare hos sex dies aliquos modo . . .
(‘What do you want now? # That you should wait for six days or so . . . ’ Pl. Ps. 321)
The universal quantifiers omnis and cunctus ‘all’ behave differently from the two types
distinguished above. When used with count nouns in the plural, omnis can be com-
bined with numerals (except for duo), as in (a) and (b) (for cunctus this seems not to be
attested). Both can be used as secondary predicates (see Chapter 21), including when
related to a noun phrase that itself contains a quantifier: (c) shows omnes as a secondary
predicate related to tres status (although it is difficult to prove). In (d), it is combined
with substantival plerique.75 For an example of cunctus with a count noun, see (e).
(a) . . . quodque omnes tres legiones adducturum Domitium putasset . . .
(‘ . . . and also because he had supposed that Domitius would bring up all three
legions . . . ’ B. Alex. 35.2)
(b) Omnes quinque et triginta tribus P. Scipioni id imperium decrevisse.
(‘ . . . that all thirty-five tribes had awarded that command to Scipio.’ Liv. 30.40.10)
(c) . . . quamquam tres status omnes cadere in hoc opus possint . . .
(‘ . . . although the three basic issues may all occur in this type . . . ’ Quint. Inst. 3.7.28)
74
There is one textually uncertain combination of quibusdam multis in Cic. Har. 62.
75
For more instances, see TLL s.v. omnis 622.7ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Both omnis and cunctus can also be used with collective nouns, such as senatus,
meaning ‘as a whole’, ‘entire’. This also holds for universus ‘the whole of ’, ‘entire’ and
(less common) solidus ‘whole’, ‘complete’. In this usage they resemble totus ‘whole’, as is
shown by (f ).76 This is also their meaning when they are used with abstract nouns.77
(f ) . . . et senatum totum et universum equestrem ordinem et cunctam plebem
Iovis Optimi Maximi pulvinaria petens comitem habuit.
(‘ . . . for seeking the couch of Jupiter Best and Greatest he was accompanied by the
senate, the entire equestrian order, and all the common folk.’ V. Max. 3.7.1e)
The use of omnis and cunctus with singular count nouns (meaning ‘each’, ‘every’), as
in (g)–(i), is uncommon and mainly poetical, but it is found from Early Latin
onwards. In this meaning they are close to the distributive determiner quisque ‘each
single’ (for which see § 11.36).78
(g) Non omnis aetas, Lyde, ludo convenit.
(‘Not every age is fit for the school, Lydus.’ Pl. Bac. 129)
(h) . . . decretum collegii vetus habemus omnem avem tripudium facere posse.
(‘ . . . we have an old ruling of our college which says that every bird can make a
tripudium.’ Cic. Div. 2.73)
(i) Omnia peccata esse paria. Omne delictum scelus esse nefarium . . .
(‘All misdeeds are equal; every misdemeanour is a heinous crime . . . ’ Cic. Mur. 61)
Cunctus is typical of literary authors. In Tacitus79 it is very rare in the singular and
much more often used substantivally, especially in the neuter, than as attribute.
In these respects it differs from his use of omnis. Other authors may have different
stylistic preferences. It is also used with some frequency (much less than omnis) in
parts of the Vulgate Old Testament, in imitation of the Hebrew original.
76
For semantic and statistical data concerning these words, see Hofmann (1948) and Callebat (1968:
288). See also TLL s.v. omnis 620.39ff.
77
For omnis, see OLD s.v. § 5 and TLL s.v. 612.82ff.; for cunctus, see OLD s.v. § 1.c and TLL s.v.
1398.7ff.; for solidus, see OLD s.v. §§ 8 and 9; for universus, see OLD s.v. §§ 1 and 3. For the instances in
Caesar and Cicero in which omnis and totus are both possible (though not necessarily synonymous), see
Lebreton (1901b: 22–4); For instances of omnis where it is more or less equivalent to totus in Apuleius, see
Callebat (1968: 287). For data on the position of omnis, see Iovino (2012: 207).
78
For the differences between omnis and quisque, see Bertocchi and Orlandini (1994); Bertocchi,
Maraldi, and Orlandini (2010: 123–41).
79
See Adams (1973b: 129–31).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Quantifiers
Ambo ‘both’ is not a proper numeral. It expresses the totality of duo ‘two’. It can be
used as an attribute, as in (a)—in combination with the demonstrative determiner
hasce—and in (b), in combination with the relative determiner quarum. It is much
more common as a secondary predicate (see Chapter 21). For the difference from duo
‘two’, see (c) and (d), where ambo functions as secondary predicate. (Italian has
ambedue.) In (e), ambo might seem to determine illi oratores (resembling the use of
omnis), but in fact it is preposed because of its contrast with preceding neutrum
illorum. For the difference between ambo and uterque, see § 11.37. Convincing
examples of substantival use are difficult to find.
(a) Aperite hasce ambas fores . . .
(‘Open this door, both leaves . . . ’ Pl. Capt. 831)
(b) Quarum ambarum rerum cum medicinam pollicetur, luxuriae licentiam
pollicetur.
(‘And, as he offers an antidote for both things, he virtually offers free indulgence for
sensuality.’ Cic. Fin. 2.70)
(c) Duas ergo hic intus eccas Bacchides. / # Quid? Duae? # Atque ambas sorores.
(‘Well, look, in here there are two Bacchises. # What? Two? # And both of them
sisters.’ Pl. Bac. 568–9)
(d) Iam hisce ambo, et servos et era, frustra sunt duo . . .
(‘Now both of them, slave and mistress, are fooled . . . ’ Pl. Am. 974)
(e) Ut duos summos viros eis qui neutrum illorum viderint, eorum quibus ambo
illi oratores cogniti sint . . . memoria teste commendemus.
(‘ . . . in order that I may recommend an illustrious pair to those who have never seen
either of them, on the testimony of the recollections of men to whom both these
famous orators were personally known . . . ’ Cic. de Orat. 2.9)
Quisque ‘any single’ is used as a so-called distributive quantifier from Early Latin
onwards, as in (a), including in more or less set phrases, as in (b).80 It can be
combined with superlative adjectives, as in (c), and with ordinal numerals, as in (d).
Similarly with substantival superlative adjectives and ordinal numerals, as in (e) and
(f ). It is also used as a secondary predicate, as in ubi quisque habitant in (a) (see also
Chapter 21). Its distributive meaning can be reinforced in the combination with unus,
as in (e). In Late Latin its function becomes very unclear.81
80
See OLD s.v. quisque § 7. For the differences between quisque and omnis, see Bertocchi and Maraldi
(2008); Bertocchi, Maraldi, and Orlandini (2010: 123–41).
81
See Bertocchi, Maraldi, and Orlandini (2010: 138–41).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Ubi quamque in urbem est ingressus, ilico / omnes meretrices, ubi quisque
habitant, invenit.
(‘When he gets to any city, he immediately finds all the prostitutes—wherever any of
them lives.’ Pl. Poen. 106–7)
(b) . . . quae studia amicorum, quae ex quoque ordine multitudo!
(‘ . . . what devotion from his friends, what a throng from every class in society!’
Cic. Sul. 73)
(c) Illud num dubitas, quin specimen naturae capi deceat ex optima quaque
natura?
(‘Can you doubt that properly our ideal of human nature should be formed from the
finest natures we meet with?’ Cic. Tusc. 1.32)
(d) . . . in quinto quoque sulco moriuntur boves.
(‘ . . . the oxen die in every fifth furrow.’ Pl. Trin. 524)
(e) Unum quodque istorum verbum nummis Philippis aureis / non potest
auferre hinc a me, si quis emptor venerit.
(‘If any buyer comes, he can’t take a single one of your words away from me for gold
Philippics.’ Pl. As. 153–4)
(f ) Vestram virtutem . . . is maxime comprobabit qui in iudicibus legendis opti-
mum et sapientissimum et fortissimum quemque delegit.
(‘Your courage . . . will meet with the high approval from him who, in his choice of
the jury, has fixed upon the best, the wisest, and the most brave.’ Cic. Mil. 105)
(g) . . . in foro vix decumus quisque est qui ipsus sese noverit.
(‘In the forum there’s hardly one in ten who knows himself.’ Pl. Ps. 973)
Quisquis in some rare instances is used in the sense of quisque, as in (h) and (i).82
Singuli ‘one each’, ‘every single’ is used as a distributive determiner from Early Latin
onwards, as in (j), including as a secondary predicate, as in (k). More often it is used
without its distributive meaning in the sense of ‘individual’, ‘single’, as in (l), and in
this use it is also found in the singular, but rarely.83 For the use of singuli as secondary
predicate, see also Chapter 21.
82
See OLD s.v. quisquis § 9. NB: In Rhet. Her. 3.31 editors now read quinto quoque.
83
See OLD s.v. singuli § 4b. For singuli, see Bertocchi and Orlandini (1994); Bertocchi, Maraldi, and
Orlandini (2010: 136–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Quantifiers
(j) Quid maceria illa ait in horto quae est, quae in noctes singulas / latere fit
minor . . . ?
(‘What does that wall say which is in our garden and which night by night becomes
one brick lower . . . ?’ Pl. Truc. 303–4)
(k) Et una / binae singulis quae datae nobis ancillae, / eae nos lavando eluendo
operam dederunt . . .
(‘And with us we had two slave-girls each that we were given—they took care of
washing and bathing us . . . ’ Pl. Poen. 221–3)
(l) Rex Creo vigiles nocturnos singulos semper locat.
(‘King Creon always posts individual sentries for the night.’ Pl. Am. 351)
The distributive numerals are used with pluralia tantum in the sense of cardinal
numerals. Examples are (m) and (n).
(m) Binas a te accepi litteras Corcyrae datas . . .
(‘I have received two letters from you, dispatched at Corcyra.’ Cic. Fam. 4.14.1)
(n) Ipse legionem unam in trina castra distribuit . . .
(‘He distributed one legion between three camps.’ Hirt. Gal. 8.36.2)
Latin has a rich system of binary quantifiers which may be used as pronouns and as
determiners. They are dealt with here together, see Table 11.4. There are interrogative,
positive, and negative binary quantifiers, as well as indefinite binary quantifiers. And
there is the binary relative pronoun/determiner. Several of the words involved are also
used as secondary predicates. Note also the ‘binary identifier’ alter ‘a second one’, ‘the
other of two’. Examples of uterque and neuter are given below. Exx. (a) and (d) are
pronouns, (b) and (e) determiners, (c) and (f ) secondary predicates.
84
For more instances, see K.-St.: I.427, 649.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
85
For discussion, see Bertocchi, Maraldi, and Orlandini (2010: 144–8).
86
For totus and possible factors contributing to its semantic development, see Bertocchi, Maraldi, and
Orlandini (2010: 115–23).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Equidem si totum exercitum meum mortem mihi optasse crederem hic
statim ante oculos vestros morerer . . .
(‘For my part, if I believed that the whole of my army had desired my death I should
die here at once before your eyes . . . ’ Liv. 28.27.10)
(d) Novi edepol nomen, nam mihi istoc nomine, / dum scribo, explevi totas
ceras quattuor.
(‘Yes, I know the name: while I was writing down that name, I filled four entire wax
tablets with it.’ Pl. Cur. 409–10)
(e) Hinc fuga coepta totam avertit aciem Romanam.
(‘ . . . they began a flight that spread to the whole Roman army.’ Liv. 10.36.6)
(f ) . . . cuius quidem clarissimas stellas totis noctibus cernimus . . .
(‘ . . . and the latter’s extremely bright stars, visible to us whole nights at a time . . . ’
Cic. N.D. 2.105)
Among the earliest instances of totus meaning more or less ‘all’ scholars cite (g), but
the use of totus in temporal expressions (‘during (the) whole’—OLD § 3b) is not
uncommon. Better examples are (h) and (i).87 In its meaning ‘whole’ totus was
succeeded by integer in Late Latin, as in ex. (j).
87
On the use of totus = omnis in Apuleius, especially in his Metamorphoses, see Callebat (1968:
287–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Contra haec Pompeius naves magnas onerarias quas in portu Brundisino
deprehenderat adornabat.
(‘To meet this Pompeius fitted out some large merchant-ships which he had seized in
the port of Brundisium.’ Caes. Civ. 1.26.1)
(b) Quemadmodum si quis statuas marmoreas muliebres stolatas, quae caryati-
des dicuntur, pro columnis in opere statuerit . . .
(‘For example, if anyone in his work sets up, instead of columns, marble statues of
long-robed women which are called caryatids . . . ’ Vitr. 1.1.5)
(c) Quamquam multi boni adulescentes illi homini nequam atque improbo
studuerunt.
(‘Though many estimable young men were devoted to that wicked and vicious man.’
Cic. Cael. 10)
(d) Quam ob rem in dissentiente populo noli putare nullos fuisse quorum
animos tuus ille fortis animus offenderet.
(‘In view of this, and in view of the wide divergence of popular feeling, you must not
think that none were scandalized by that uncompromising attitude of yours.’ Cic.
Planc. 53)
(e) Accedunt eodem multa privata magna eius in me merita, mea quaedam
officia in illum.
(‘Added to these are the many great private benefactions he’s done me and certain
kindnesses I’ve done him.’ Cic. Phil. 13.7)
88
For asyndetic coordination, see Ch. 19.
89
Von Nägelsbach and von Müller (1905: 337) characterize magna in (e) as some form of after-
thought: ‘und zwar bedeutend’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Itaque est cum Saturno, uti si quid liberum virile secus ei natum esset, ne quid
educaret. (Enn. var. 69–71—NB: the second quid is odd); Ego hanc clementem vitam
urbanam atque otium / secutu’ sum (Ter. Ad. 42–3); . . . qui malevoli / veteris poetae
maledictis respondeat. (Ter. An. 6–7); Labore alieno magno partam gloriam / verbis
saepe in se transmovet . . . (Ter. Eu. 399–400); Intro iit in aedis ater alienus
canis. (Ter. Ph. 706); . . . usque ad hoc tempus Syracusani festos dies anniversarios
agunt . . . (Cic. Ver. 4.107); Iste hominibus miseris innocentibus inici catenas imperat.
(Cic. Ver. 5.106); . . . tibi apud eosdem privata navis oneraria maxima publice est
aedificata . . . (Cic. Ver. 5.136); . . . cum consuleretur, utrum bono viro pauperi an
minus probato diviti filiam collocaret . . . (Cic. Off. 2.71); . . . conlocari iussit hominem
in aureo lecto strato pulcherrimo textili stragulo . . . (Cic. Tusc. 5.61); Quem locum
duplici altissimo muro munierant. (Caes. Gal. 2.29.3); Incidit Caesarianum civile
bellum . . . (Nep. Att. 7.1); Et tum quidem ab nocturno iuvenali ludo in castra
redeunt. (Liv. 1.57.11); . . . eademque hora duo exercitus, duae potentissimae et maxi-
mae (maxime, all but one ms.) finitimae gentes superatae sunt. (Liv. 2.53.3); . . .
columnaque inde aurea solida facta et sacrata est. (Liv. 24.3.6); Huius atrocitas
poenae duarum nobilissimarum in Italia Graecarum civitatium animos inritavit . . .
(Liv. 25.8.1); Et P. Licinio Varo praetori urbano negotium datum ut naves longas
triginta veteres reficeret. (Liv. 27.22.12);
90
For clues to the interpretation of participles as attributive in Cicero, see Laughton (1964: 52–69).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Non modo Siculorum nihil in hac re valuisse leges, sed ne ab senatu quidem
populoque Romano datas.
(‘In this matter it was not only the local Sicilian laws that were completely set aside,
but even the laws given by the Senate and Roman People.’ Cic. Ver. 2.121)
(d) . . . corporis aliquam commoditatem non natura datam, sed studio et in-
dustria partam.
(‘ . . . some bodily dexterity not given by nature but won by careful training and
practice.’ Cic. Inv. 1.36)
(e) Atqui haec, inquam, de incestu laudata oratio puerilis est locis multis . . .
(‘But to come back: this much praised oration on incest is in many places puerile . . . ’
Cic. Brut. 124)
91
For more Ciceronian instances, see Laughton (1964: 125–39).
92
For a few more instances, see Bennett (1910: I.430).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
forms of one-place verbs and indicate a property of the noun they modify, in (a) like a
quantifier (e.g. multas ‘many’), in (b) like a descriptive adjective (e.g. calidam ‘hot’).
Instances of present participles of two- and three-place verbs with arguments and
satellites added as in finite clauses become common from Cicero’s early rhetorical
treatise de Inventione and the Rhetorica ad Herennium onwards, where the influence
of Greek sources is evident and where we are dealing with texts that were composed
for readers. They are also common in other didactic texts, for instance in medical
texts. The participial phrase in (c) can be compared to the relative clause in (d).93
Supplement:
Nam si tibi necesse putas etiam adversariis amicorum tuorum de iure consulenti-
bus respondere . . . (Cic. Mur. 9); Semper in hac civitate nimis magnis accusatorum
opibus et populus universus et sapientes ac multum in posterum prospicientes
iudices restiterunt. (Cic. Mur. 59); Est enim lex nihil aliud nisi recta et a numine
deorum tracta ratio, imperans honesta, prohibens contraria. (Cic. Phil. 11.28); Sic
contendo, ex his parvis libellis apud unum magistrum societatis repertis vos iam
coniectura adsequi posse . . . (Cic. Ver. 2.184); Atque haec forsitan homini non
omnia quae sunt in natura rerum celeriter animo comprehendenti permulta
videantur . . . (Cic. de Orat. 2.136); Sed omnis nec claudicans nec quasi fluctuans
et aequabiliter constanterque ingrediens numerosa habetur oratio. (Cic. Orat.
198); Quae animi affectio suum cuique tribuens atque hanc, quam dico, societatem
coniunctionis humanae munifice et aeque tuens iustitia dicitur. (Cic. Fin. 5.65);
Non modo enim id virtutis non est, sed est potius immanitatis omnem humani-
tatem repellentis. (Cic. Off. 1.62); Ne [et] tu pusilli animi es et sordide se
consolantis . . . (Sen. Dial. 12.9.3);
93
For the development of attributive participles in Cicero, see Laughton (1964: 52ff.). Flinck-
Linkomies (1929: 43) holds training in Greek rhetoric responsible for the increased use of participles as
verbal forms.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Attributive perfect (passive and deponent) participles are also not frequent in Early
Latin.94 Examples are (a) and (b). More elaborate participial phrases become com-
mon from Cicero’s early writings onwards, both in didactic texts as in (c), and in his
orations (d). An interesting example of variation between relative clauses and a
participial phrase is (e).
(a) Ergo edepol <merito> meritam mercedem dabo.
(‘Then I’ll give him his deserved reward deservedly.’ Pl. Capt. 1020)
(b) . . . quia Argivi in ea delecti viri / vecti petebant pellem inauratam arietis . . .
(‘ . . . for that chosen Argive heroes were carried in it when they were seeking the
golden fleece of the ram . . . ’ Enn. scen. 250–1V=212–13J)
(c) Constitutio est prima conflictio causarum ex depulsione intentionis
profecta . . .
(‘The “issue” is the first conflict of pleas which arises from the defence or answer to
our accusation . . . ’ Cic. Inv. 1.10)
(d) Habetis consulem ex plurimis periculis et insidiis atque ex media morte non
ad vitam suam, sed ad salutem vestram reservatum.
(‘You have a consul kept from a multitude of dangers and plots, and from the very
jaws of death, not just to cling to life but to save you.’ Cic. Catil. 4.18)
(e) Habet honorem quem petimus, habet spem quam propositam nobis habemus,
habet existimationem multo sudore labore vigiliisque collectam . . .
(‘It holds the office to which I seek election; the ambition that I cherish in my heart; the
reputation which I have risen early and toiled in the heat to gain.’ Cic. Div. Caec. 72)
Supplement:
. . . veteres multo ante suscepti et provisi labores cum viro fortissimo et clarissimo
coniunxissent . . . (Cic. Dom. 29); . . . recordor quantum hae conquestiones in senatu
habitae punctorum nobis, Servi, detraxerint. (Cic. Mur. 72); . . . oblitus auspiciorum a
te ipso augure populi Romani nuntiatorum . . . (Cic. Phil. 1.31); Ego adulescentis bonos
et fortis, sed usos ea condicione fortunae ut . . . comitiorum ratione privavi. (Cic. Pis. 4);
. . . multas pecunias isti erogatas in operum locationes falsas atque inanis esse per-
scriptas. (Cic. Ver. 5.48); Nam et ex horreis direptum effusumque frumentum vias
omnis angiportusque constraverat . . . (Cic. Div. 1.69); Theophrastus mediocriterne
delectat, cum tractat locos ab Aristotele ante tractatos? (Cic. Fin. 1.6); Brutus noster
misit ad me orationem suam habitam in contione Capitolina . . . (Cic. Att. 15.1a.2);
Spargit rapida ungula rores / sanguineos mixtaque cruor calcatur harena. (Verg. A.
12.339–40); . . . LOCA·PUBLICA A PRIVATIS / POSSESSA ·T.·SUEDIUS·CLEMENS / TRIBUNUS . . . /
REI / PUBLICAE POMPEIANORUM / RESTITUIT. (CIL X.1018.4–9 (Pompeii));
94
For a few more instances, see Bennett (1910: I.436).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The first future participle used as an attribute is futurus, which is used quite often by
Cicero. Examples are (a) and (b). It became more popular later, especially in Seneca.95
(a) Etenim eo loco, Fanni et Scaevola, locati sumus ut nos longe prospicere
oporteat futuros casus rei publicae.
(‘For, my dear Fannius and Scaevola, we Romans are now placed in such a situation
that it is our duty to keep a sharp lookout for the troubles that may befall our State.’
Cic. Amic. 40)
(b) . . . sic futurae post mortem famae tamen esse propter rem, etiam detracto
usu, consulendum.
(‘ . . . so a man ought to give thought to his reputation even after death, for its own
sake, even apart from any advantage.’ Cic. Fin. 3.57)
Supplement:
Equi boni futuri signa (sc. sunt), si cum gregalibus in pabulo contendit in currendo
aliave qua re . . . (Var. R. 2.7.6); Est enim metus, ut aegritudo praesentis, sic ille futuri
mali. (Cic. Tusc. 4.64); Te, Cicero, rogo atque hortor ne defatigere neu diffidas,
semper in praesentibus malis prohibendis futura quoque, nisi ante sit occursum,
explores ne se insinuent. (Brut. ad Brut. 1.16.10); Quid hoc mihi prodest . . . si vindex
illius mali auctor exstitit alterius fundamentum et radices habituri altiores, si
patiamur . . . (Brut. ad Brut. 1.17.2); . . . ad suos tranavit, rem ausus plus famae habi-
turam ad posteros quam fidei. (Liv. 2.10.11); Sic verus ille animus et in alienum non
venturus arbitrium probatur. (Sen. Ep. 13.1); Ne aquam quidem habet nubes sed
materiam futurae aquae. (Sen. Nat. 1.5.4); Sed ubi lectio fortior erexit animum . . .
prosilire libet in forum, commodare alteri vocem, alteri operam, etiam si nihil
profuturam, tamen conaturam prodesse . . . (Sen. Dial. 9.1.12); Sub idem tempus
Calpurniam L. Pisonis filiam successuri sibi in consulatu duxit uxorem . . .
(Suet. Jul. 21.1);
The attributive use of gerundives is almost entirely restricted to those that are derived
from verbs denoting a personal evaluation (especially the ‘verba affectuum’). So one
finds contemnendus ‘to be despised’, laudandus ‘to be praised’, mirandus ‘remark-
able’, but not aedificandus ‘to be built’.96 As for their meaning they resemble adjec-
tives in -bilis. They are not very frequent but were in use throughout the history of
Latin in all types of texts. Examples are (a)–(c). The gerundives may be modified,
especially by degree adverbs; occasionally other constituents are found that can also
95
For Seneca, see Westman (1961: 135–62).
96
For a list of attributively used gerundives, see Aalto (1949: 102–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
be found with finite forms of the same verb. They can be coordinated with ‘normal’
adjectives, as in (b). Superlative forms occur in Late Latin.97
(a) . . . adduxit simul / forma expetenda liberalem virginem . . .
(‘ . . . has brought a free-born virgin with him, of desirable appearance . . . ’ Pl. Per.
520–1)
(b) Omnes sunt in illo rege virtutes, quod te, Caesar, ignorare non arbitror, sed
praecipue singularis et admiranda frugalitas.
(‘As I think you know well enough, Caesar, all virtues are present in that king, but
especially a remarkable and admirable sobriety.’ Cic. Deiot. 26)
(c) At non, Euandre, pudendis / vulneribus pulsum aspicies . . .
(‘But your eyes, Evander, will not look on your son routed with shameful wounds . . . ’
Verg. A. 11.55–6)
Supplement:
Pol haud miranda facta dicis. (Pl. Rud. 345); Pro di inmortales, facinus miserandum
et malum. (Ter. Ph. 1008—NB: coordination); Non minus res admiranda cum mi
esset dicta in Arcadia . . . (Var. R. 2.4.12); . . . nuntiatur terrae motus horribilis cum
quibusdam {multis metuendisque rebus. (Cic. Har. 62); O negotia non ferenda, quae
feruntur tamen. (Cic. Att. 15.12.2); Quae etiam si parum iusta tibi visa est, hanc
tamen habet rationem, non ut nimis concupiscendus honos sed tamen, si deferatur a
senatu, minime aspernandus esse videatur. (Cic. Fam. 15.6.2); Tum vox horrenda
per auras / excidit et Troum Rutulorumque agmina complet. (Verg. A. 9.112–13);
. . . votaque pro domina vix numeranda facit. (Tib. 3.10.12); Ita pessima atque
execranda res propaginem tamen docuit ac viviradicem. (Plin. Nat. 17.96—NB:
coordination); . . . ut alias pudendas clades exercituum taceam. (Liv. 25.6.10);
Cum deinde immanes res vix multis voluminibus explicandas C. Caesar in Gallia
ageret . . . (Vell. 2.46.1); . . . veteris qui tollunt grandia templi / pocula adorandae
rubiginis . . . (Juv. 13.147–8); . . . usurum me viribus meis clara liberaque voce clamavi,
ni abstineret a Gitone iniuriam mulier damnata et in toto navigio sola verberanda.
(Petr. 108.5); . . . PUELLA·DULCIS·AMICORUM / CONCUPIENDA·IOCIS . . . (CIL VI.20466.3–4
(Rome)); SIMULACRUM MINER<V=B>AE / ABOLENDO INCENDIO / TUMULTUS CIVILIS IGNI /
TECTO CADENTE CONFRACTUM / ANICIUS ACILIUS AGINATIUS / FAUSTUS . . . / RESTITUIT. (CIL
VI.526=1664 (Rome, end V));
97
For more instances, see Aalto (1949: 101–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
marking. The most common case for marking NPs as attribute is the genitive (see
§ 12.2 for statistics).
Latin grammars usually distinguish a fairly large number of uses of the genitive as
marker of attributes; in practice, however, these are characterizations of the semantic
relationship between the head and the attribute and thus have nothing to do with the
genitive as such.98 Without endorsing the notion that Latin has different genitives to
mark an attribute, a few of the most common labels are nevertheless used in this
Syntax. These are (i) the possessive genitive, which may be coordinated with a
possessive adjective; (ii) the genitive of description, which may be coordinated with
a descriptive adjective and function as answer to the question qualis ‘what kind’; (iii)
the partitive genitive, which may sometimes be replaced by a prepositional phrase
with ex and de with more or less the same meaning; (iv) the genitive of quantity.
Separate attention is also given to the genitive of definition.
For most combinations of head nouns and attributive NPs the precise semantic
function of the genitive attribute is wholly dependent on the meanings of the head
and attribute constituents involved and the way they are related in the extralinguistic
world. The scope of possible relations is infinite and no restrictions seem to apply: we
find animate, inanimate, concrete, abstract, alienable, and inalienable entities as
attributive genitive NPs. A few examples must suffice to illustrate the breadth of the
semantic relations that are possible. In (a) and (b), the relation is one of place. In (c), it
is one of the route taken through Asia and Syria, in (d) a night spent with, in (e) the
cargo transported.
(a) Cum servitiorum animos in Sicilia suspensos propter bellum Italiae fugiti-
vorum videret . . .
(‘Observing the unrest among the slaves of Sicily that was caused by the servile war in
Italy . . . ’ Cic. Ver. 5.14—NB: note the variation in Sicilia : Italiae.)
(b) Harum omnium rerum facultates sine ullo periculo pons Ilerdae
praebebat . . .
(‘The bridge at Ilerda . . . gave opportunites for all these measures without any risk.’
Caes. Civ. 1.49.2)
(c) Simul infamia duarum legionum permotus quas ab itinere Asiae Syriaeque
ad suam potentiam dominatumque converterat . . .
(‘Stirred, too, by the discredit attaching to his diversion of two legions from their
route by Asia and Syria and his appropriation of them for his own power and
supremacy . . . ’ Caes. Civ. 1.4.5)
98
The best description of the many uses of the genitive that are distinguished in grammars and studies
is de Groot (1956).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In its range of possible semantic relations the possessive genitive differs greatly from
certain uses of the dative, which some scholars regard as alternative expressions for
more or less the same semantic relationship of ‘possession’.99 See § 4.26 for the so-
called dativus possessivus and § 10.96 for the so-called dativus sympatheticus. Only
when the ‘possessor’ is an animate being and the ‘possessed entity’ an inalienable part
(especially a body part) of that entity are the genitive and dative expressions more or
99
For a critical discussion, see Bolkestein (1983) and (2001a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
less equivalent, as can be seen by comparing the genitive Caesaris in (f ) with the
dative Caesari in (g).
(f ) . . . cum . . . C. Marcellus se ad Caesaris pedes abiecisset . . .
(‘ . . . when . . . C. Marcellus had fallen at Caesar’s feet.’ Cic. Fam. 4.4.3)
(g) Ea re impetrata sese omnes flentes Caesari ad pedes proiecerunt.
(‘The petition was granted, and they all threw themselves in tears at Caesar’s feet.’
Caes. Gal. 1.31.2)
100
For the development in general, see Edwards (1902) and Vandvik (1942). For a detailed demon-
stration of the development with the noun ingenium ‘character’, see Edwards (1902). For Terence, see
Baloira (1981–3); for Symmachus, see Haverling (1988: 159ff.).
101
For such value expressions, see Laing (1920: 9–10).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
S. Rosc. 17); Ego autem casu, cum dedissem ad te litteras VI Kal. satis multis verbis,
tribus fere horis post accepi tuas et magni quidem ponderis. (Cic. Att. 14.14.1);
Omnino, si quaeris, ludi apparatissimi, sed non tui stomachi. (Cic. Fam. 7.1.2);
(Romanos) . . . qui tantae altitudinis machinationes tanta celeritate promovere et ex
propinquitate pugnare possent . . . (Caes. Gal. 2.31.2); . . . quin . . . patrem multo potius
M. Valerium spectatae virtutis et consularem legissent (Liv. 2.18.7); Huic bellum
adversus Aequos permissum est, pravae mentis homini . . . (Liv. 4.49.8); In veteres
quassasque naves paucis et parvi pretii rebus impositis . . . (Liv. 25.3.11—NB: coord-
ination); Illa adhuc in miserae sortis infantia timebantur. (Sen. Con. 10.4.21); . . .
homine acidae linguae et qui nimis liber erat. (Sen. Con. 10.5.22); Isdem ludis et
rhinoceros unius in nare cornus, qualis saepe, visus. (Plin. Nat. 8.71); . . . UNICAE
BONITATIS / ET GRAVITATIS VIRO . . . (CIL VI.41228. 4–5 (Rome, c. AD 250)); . . .
MARCIANAE / CONIUGI·OPTIMAE·ET·DULCIS/SIMAE·ET·DESIDERANTISSIMAE / RARISSIMI·EXEMPLI·
FEMINAE·FECIT /P. RAGONIUS·DAPHNUS . . . (CIL VI.9992.2–6 (Rome)); Hic adulescens et
semibarbarus et vix adhuc Latinae linguae, prope Thraecica imperatorem publice
petit, ut . . . (Hist. Aug. Maxim. 2.5);
NB: substantival use: O miserae sortis . . . (Luc. 2.45);102
The use of a noun in the genitive without an attribute is mainly a Late Latin
phenomenon.103 Examples are (e)—note the coordination—and (f ). In the Bible
translations and in ecclesiastical authors the relative frequency must be regarded as
due to Semitic influence.104
11.49 The partitive genitive and the genitive of quantity functioning as attribute
NPs in the genitive denoting a plural entity (for example, homines ‘men’) or a collective
entity (for example, exercitus ‘army’) may be optionally added to various categories of
head constituents. The head is part of the entity to which the genitive expression refers,
for example quis hostium ‘which among the enemies’ and maximum omnium flumi-
num ‘the largest of all rivers’. This comes under the overall category of the GENITIVE OF
THE WHOLE, often called PARTITIVE GENITIVE (genetivus partitivus).
Often dealt with under the same heading are NPs in the genitive that do denote a
plurality of entities or a collective entity as well, but occur with head nouns that
denote measure or quantity, or with various neuter singular forms of pronouns and
adjectives of amount, for example magnus numerus hostium ‘a large number of
enemies’ and multum temporis ‘much time’. Moreover, in this case the noun phrase
102
For this interpretation instead of as an exclamative genitive, see Mayer (1979: 341).
103
For more examples and discussion, see Haverling (1988: 163ff.).
104
See Suárez (2008) and Rubio (2009: 208–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
in the genitive may also be a mass noun (for example, libra farris ‘a pound of barley’).
As in some other grammars, this use of the genitive is called GENITIVE OF QUANTITY in
this Syntax. The noun phrases in these expressions are generic, whereas genitives of
the whole are non-generic (definite or indefinite). As an alternative for the partitive
genitive, prepositional phrases with de and ex are more or less common; for the
genitive of quantity they are less common (details in later sections). The same head
noun can occur in both expressions. This is illustrated with the nouns numerus
‘number’ and pars in (a)–(d). In (a), frumenti is generic; magnum numerum denotes
its quantity. Note the parallelism with the other quantitative expressions. In (b), the
context shows that magnum numerum is a subgroup of the whole of the enemies. In
(c), pars is understood as a unit for measuring the quantity of generic defrutum vetus,
which is needed for the production of a particular wine. In (d), suarum shows that the
troops are not generic; only part of them are involved in the action of traducere.
(a) Dico te maximum pondus auri argenti eboris purpurae, plurimam vestem Meli-
tensem, plurimam stragulam, multam Deliacam supellectilem, plurima vasa Co-
rinthia, magnum numerum frumenti, vim mellis maximam Syracusis exportasse.
(‘I assert that you exported from Syracuse a great weight of gold, silver, ivory, and
purple fabrics, a great deal of Maltese cloth and tapestries, a quantity of Delian wares,
a large number of Corinthian vessels, a large quantity of corn and an immense
amount of honey.’ Cic. Ver. 2.176)
(b) Quos inpeditos (om. β) integris viribus milites nostri consecuti magnum
numerum eorum occiderunt. Reliquos equites consectati paucos qui ex fuga
evaserant reliquerunt.
(‘Our troops pursued them, hampered as they were, with strength unimpaired and
slew a great number of them. The rest our cavalry hunted down, and left but a few,
who had got clear away from the rout.’ Caes. Gal. 3.19.4)
(c) Si helviolum vinum facere voles, dimidium helvioli, dimidium Apicii indito,
defruti veteris partem tricesimam addito.
(‘If you wish to make a straw-coloured wine, take equal parts of yellow and Apician
wine and add a thirtieth part of old boiled wine.’ Cato Agr. 24.1)
(d) Ibi vadis repertis partem suarum copiarum traducere conati sunt . . .
(‘There they found fords, and endeavoured to throw part of their forces across . . . ’
Caes. Gal. 2.9.4)
(iii) substantively used identifiers, e.g. alii ‘others’, reliqui ‘the others’, ‘the rest’;
(iv) substantively used adjectives and participles, e.g. fortissimi ‘the bravest’,
honestissimus ‘the most noble’;
(v) (rarely) nouns and noun phrases (including proper names), e.g. pars ‘part’.
adsunt forte miratur . . . (Cic. Div. Caec. 1); Nam quid interest inter Dolabellam et
quemvis Antoniorum trium? (Cic. ad Brut. 2.5.5); Hoc loco Sulpicius ‘insperanti’
inquit ‘mihi et Cottae, sed valde optanti utrique nostrum cecidit ut in istum
sermonem, Crasse, delaberemini.’ (Cic. de Orat. 1.66);
Neminem vestrum ignorare arbitror, iudices, hunc per hosce dies sermonem vulgi . . .
fuisse . . . (Cic. Ver. 1.1); Aliter sine populi iussu nullius earum rerum consuli ius est. (Sal.
Cat. 29.3); . . . quia neuter consulum potuerat bello abesse. (Liv. 9.44.2); Si illi loco cedant,
neminem eius exercitus non modo Italiam, sed ne Tagi quidem ulteriorem ripam
umquam visurum. (Liv. 39.31.5);
NB: Ibo intro, ut quid huius verum sit sciam. (Pl. Aul. 802); Minus mirandum est,
illaec aetas si quid illorum facit, / quam si non faciat. (Pl. Bac. 409–10);
Prepositional alternatives: Quasi nunc id agatur quis ex tanta multitudine occiderit . . .
(Cic. S. Rosc. 92); . . . QUAEQUE EX EIS MINUS ANNUM GNATAE ERUNT . . . (CIL I2.585.14 (Lex
Agr., c.111 BC)); Dicit quidam ex illis canibus quos iste Liguri dixerat esse circa se
multos . . . (Cic. Ver. 1.133); Quo nemo adaeque iuventute ex omni Attica / antehac est
habitus parcus . . . (Pl. Mos. 30–1); Illi ego ex omnibus optume volo. (Pl. Mos. 337);105
105
For more examples, see TLL s.v. ex 1116.72ff.
106
For the use of partitives with quantifiers, see Bertocchi and Maraldi (2010).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Sed uterque nostrum cedere cogebatur magnitudini animi orationisque gravitati . . .
(Cic. Phil. 9.9); Sunt vestrum aliquam multi qui L. Pisonem cognorint . . . (Cic. Ver.
4.56); Quorum non nulli etiam illum diem memoria tenebant . . . (Cic. Ver. 4.77);
Atque haud scio an plerique nostrorum oratorum ingenio plus valuerint quam
doctrina. (Cic. Orat. 143); . . . cum superioris etiam ordinis nonnulli ratione permoti
manendum eo loco et rem proelio committendam existimarent. (Caes. Civ. 3.74.2);
Quod ubi accidit, complures hostium magno nostrorum impetu perculsi
vulnerantur . . . (Hirt. Gal. 8.48.7); . . . quam (sc. caesariem) multis illa dearum / levia
protendens bracchia pollicita’st . . . (Catul. 66.9–10); Sed ubi plerumque noctis
processit . . . (Sal. Jug. 21.2); Sed quoniam humanarum rerum fortuna pleraque regit . . .
(Sal. Jug. 102.9); Pauci hostium evasere ex insidiis, praeda omnis recepta est.
(Liv. 3.3.8); . . . pluribus hostium in tentoriis suis quam in portis valloque caesis . . .
(Liv. 7.37.13); . . . patrum quoque plerosque captivi cognatione attingebant (Liv.
22.61.1); Romanorum aliquot vulnerati sunt. (Liv. 37.44.2); Paucis avium in iecore
(sc. est fel). (Plin. Nat. 11.194); Tradidere sese abeunti Turullius Cerialis cum com-
pluribus classicis et Iulius Briganticus cum paucis equitum . . . (Tac. Hist. 2.22.3);
NB: Facile ergo intellegimus, quanta apud te sit filiolae nostrae conciliatrix similitudo
utriusque nostri (Fro. Aur. 5.68vdH—NB: probably contamination);
. . . L. Sestium . . . cum ipse potestatem summam haberet, quod decemvirum unus
sine provocatione esset, vades tamen poposcit . . . (Cic. Rep. 2.61); . . . e regione unius
eorum pontium quos Vercingetorix rescindendos curaverat postero die cum dua-
bus legionibus in occulto restitit. (Caes. Gal. 7.35.2); . . . sum paulo infirmior, unus /
multorum. (Hor. S. 1.9.71–2); . . . pastorum unus progressus sub ipsas munitiones
inclamat alios . . . (Liv. 10.4.8); Ducenti ab Romanis, octingenti hostium cecidere. (Liv.
22.16.3—NB: parallelism between a prepositional phrase and a genitive); . . . arci
Babyloniae Agathone praesidere iusso cum septingentis Macedonum trecentisque
mercede conductis . . . (Curt. 5.1.43);
Appendix: In Caesar and Cicero, when several head constituents share one entity
that denotes the whole, as in example (e) above from Livy, the genitive is preferred
over prepositional expressions by far. In such a situation the genitive constituent
usually precedes its head and contains topical information. An additional example,
where the whole is expressed by a relative, is (f ).107 In general, the partitive expres-
sion (in which the noun is the modifier) has a different pragmatic structure from that
of the alternative, where the quantifier is the modifier (see also the next section).
(f ) Gallia est omnis divisa in partes tres, quarum unam incolunt Belgae, aliam
Aquitani, tertiam qui ipsorum lingua Celtae, nostra Galli appellantur.
(‘Gaul as a whole is divided into three parts, one of which is inhabited by the
Belgae, another by the Aquitani, and a third by a people called in their own
tongue Celtae, in the Latin Galli.’ Caes. Gal. 1.1.1)
The above quantifiers may, of course, also be used as attributes and are indeed found
in minimal pairs with the same nouns, such as partitive cum paucis equitum ‘with a
107
For more examples, see K.-St.: I.426.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
few of the horsemen’ (Liv. 27.30.8) and attributive cum paucis equitibus ‘with a few
horsemen’ (Liv. 24.30.10). However, the two expressions have different meanings. In
poetry, from Lucretius onwards, and in poeticizing prose, from Sallust till Ammia-
nus, cuncta ‘all’ (pl. n.) is used with a partitive genitive, as in (g). The use of other
forms of cunctus, and also of omnis, results in pseudo-partitive (see § 11.13) expres-
sions like (h) and (i).108 (Samnitium omnes ‘all of the Samnites’ in Liv.10.31.5 is
textually uncertain.)
(g) . . . Flora quibus mater praespargens ante viai / cuncta coloribus egregiis et
odoribus opplet.
(‘ . . . mother Flora . . . strewing the whole path in front and filling it with
brilliant colours and scents.’ Lucr. 5.739–40)
(h) Hic hominum cunctos ingenti corpore praestans / Iapetionides Atlas fuit.
(‘Here, far surpassing all men in huge bulk of body, was Atlas, of the stock of
Iapetus.’ Ov. Met. 4.631–2)
(i) Omnia non pariter rerum sunt omnibus apta . . .
(‘All things in the world are not suited to all men alike . . . ’ Prop. 3.9.7)
The pair plerique Romanorum ‘most of the Romans’ and plerique Romani ‘most
Romans’ (the latter is the statistically predominant expression; de and ex are used as
well) receives some attention in standard grammars.109 Plerique is also used as a
secondary predicate (see Chapter 21).
108
For the material, see TLL s.v. cunctus 1402.44ff.; s.v. omnis 619.3ff.
109
For example, K.-St.: I.427.
110
For partitive constructions with identifiers, see TLL s.v. alius 1629.48ff.; s.v. alter 1748.80ff.; s.v.
ceterus 967.78ff., 968.82ff. (contains also non-partitive expressions).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) At ille cohortem primam totam perdidit centurionemque primi pili, nobilem
sui generis, Asinium Dentonem et reliquos cohortis eiusdem . . .
(‘The result was that he lost his entire first cohort, including a chief centurion,
Asinius Dento, a distinguished man in his own class, and the other centurions of
the same cohort . . . ’ Cic. Att. 5.20.4)
Supplement:
Quod numquam opinatus fui neque alius quisquam civium / sibi eventurum, id
contigit . . . (Pl. Am. 186–7); Studiosiores ad opus fieri liberalius tractando aut cibariis
aut vestitu largiore . . . huiusce modi rerum aliis . . . (Var. R. 1.17.7); Uno partu duos
peperit simul. / Eorum alter, nostro qui est susceptus semine, / suis factis te
immortali adficiet gloria. (Pl. Am. 1138–40); . . . cum ferocissimi per acies aut pro-
scriptione cecidissent, ceteri nobilium, quanto quis servitio promptior, opibus et
honoribus extollerentur . . . (Tac. Ann. 1.2.1); . . . Pontus et nebulosus est et dulcior
aequorum ceteris . . . (Amm. 22.8.46); Reliquum scripturae consumetur in exemplis.
(Rhet. Her. 4.10); Ita fit ut illud breve vitae reliquum nec avide adpetendum senibus
nec sine causa deserundum sit. (Cic. Sen. 72); Reliquum anni quietum ab urbanis
motibus et ab externis mansit. (Liv. 3.72.7);
Prepositional alternatives: Necesse est enim sit alterum de duobus, ut . . . (Cic. Tusc.
1.97); Alius autem de discipulis eius ait illi . . . (Vulg. Mat. 8.21); Nam ex refectis
navibus alias in alium usum retineri ad urbem placuit. (Liv. 42.48.5); Quibus ex
generibus alteri se populares, alteri optimates et haberi et esse voluerunt. (Cic. Sest.
96); His cognitis rebus altera ex duabus legionibus, quae vernacula appellabatur, ex
castris Varronis adstante et inspectante ipso signa sustulit . . . (Caes. Civ. 2.20.4); Nam
hoc Verrem dicere aiebant, te non fato, ut ceteros ex vestra familia, sed opera sua
consulem factum. (Cic. Ver. 29);
(d) Nunc, ut Apelles Veneris caput et summa pectoris politissima arte perfecit,
reliquam partem corporis incohatam reliquit, sic . . .
(‘Like Apelles, who applied the utmost refinement of his art to perfecting the head
and bust of his Venus, but left the rest of the body a mere sketch . . . ’ Cic. Fam. 1.9.15)
(e) Hic homo est omnium hominum praecipuos . . .
(‘This man is exceeding all men . . . ’ Pl. Trin. 1115)
(f ) Maior Neronum mox grave proelium / conmisit . . .
(‘Soon too the elder Nero joined deadly battle . . . ’ Hor. Carm. 4.14.14–15)
Supplement:
HONC OINO·PLOIRUME·COSENTIONT·R<OMANE> / DUONORO·OPTUMO·FUISE·VIRO / LUCIOM·
SCIPIONE. (CIL I2.9.1–3 (Scip. Elog., Rome, c.200 BC)); Occisissimus sum omnium qui
vivont. (Pl. Cas. 694); Abi hinc sis ergo, pessumarum pessuma. (Pl. Cas. 793); Quid
ais tu, omnium hominum taeterrime? (Pl. Mos. 593); Quid porro in Graeco sermone
tam tritum atque celebratum est quam, si quis despicatui ducitur, ut ‘Mysorum
ultimus’ esse dicatur? (Cic. Flac. 65); Roscius, pater huiusce, municeps Amerinus
fuit . . . eius vicinitatis facile primus . . . (Cic. S. Rosc. 15); . . . omnium nationum ex-
terarum princeps Sicilia se ad amicitiam fidemque populi Romani adplicavit. (Cic.
Ver. 2.2); Cuius disputationis fuit extremum fere de inmortalitate animorum . . . (Cic.
Amic. 14); Indus vero, qui est omnium fluminum maximus . . . (Cic. N.D. 2.130);
Infima est quinque errantium terraeque proxuma stella Veneris . . . (Cic. N.D. 2.53);
Sed, mi Plance, incumbe ut belli extrema perficias. (Cic. Fam. 10.19.2); Ab eius
summo sicut palmae ramique late diffunduntur. (Caes. Gal. 6.26.2); Phaselus ille
quem videtis, hospites, / ait fuisse navium celerrimus . . . (Catul. 4.1–2); Nam aestatis
extremum erat. (Sal. Jug. 90.1); Quotiens in extrema periculorum ventum . . . (Liv.
7.29.2); Ganges, omnium ab Oriente fluvius eximius . . . (Curt. 8.9.5); . . . celeberrimo
fori effigiem divi Augusti amplecti . . . (Tac. Ann. 4.67.4); Hi ceterorum Britannorum
fugacissimi ideoque tam diu superstites. (Tac. Agr. 34.1—NB: contamination);
Comparative: Transmarini Epirotici non solum meliores totius Graeciae, sed etiam
quam Italiae. (Var. R. 2.5.10—NB: various emendations suggested111); Sicut Zac-
chaeus vidit illud bonum . . . maior publicanorum . . . (August. Psal. 75.9);
Prepositional alternatives (in alphabetical order by preposition): . . . noctu de servis
suis quem habuit fidelissimum ad regem misit . . . (Nep. Them. 4.3); Acerrimum
autem ex omnibus nostris sensibus esse sensum videndi. (Cic. de Orat. 2.357); In
testamento Ptolomaei patris heredes erant scripti ex duobus filiis maior et ex duabus
<filiabus> ea quae aetate antecedebat. (Caes. Civ. 3.108.4); Etenim rectum putabat
pro eorum honestate se pugnare propter quos ipse honestissimus inter suos numera-
batur. (Cic. S. Rosc. 16);
The partitive genitive may also be used when the superlative does not really function
as head of a noun phrase, but fulfils another function. In (g), for example, minimi is
an attribute within the subject complement minimi preti. The part/whole relationship
is between me and omnium hominum. In (h), honestissimum is the attribute of
111
See Giusta (2006: 127).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The use of positive forms of adjectives and participles with a partitive genitive is
illustrated in (k)–(m). It is very rare in Classical Latin (but see (l)) and is typical of
poets and poeticizing prose. In the case of (k), Homeric inspiration is obvious.
(k) . . . constitit inde loci propter sos (= eos) dia dearum . . .
(‘Thereupon she, hallowed among the holy goddesses, took her stand close to them.’
Enn. Ann. 22V=19S)
(l) Pergamique in occultis ac reconditis templi . . . tympana sonuerunt.
(‘At Pergamum in the secret and concealed parts of the temple . . . there was a sound
of drums.’ Caes. Civ. 3.105.5)
(m) Praemisso Caecina ut occulta saltuum scrutaretur pontesque et aggeres
umido paludum et fallacibus campis inponeret . . .
(‘Sending Caecina forward to explore the secret forest passes and to throw bridges
and causeways over the flooded marshes and treacherous levels . . . ’ Tac. Ann. 1.61.1)
Supplement:
. . . inter lucubrantes ancillas in medio aedium sedentem inveniunt. (Liv. 1.57.9);
. . . nocte proditione oppidum captum liberam per aversa urbis fugam dederat . . .
(Liv. 5.29.4); . . . alterum qui, si qui alibi motus exstitisset, ad subita belli mitti posset
. . . (Liv. 6.32.5); . . . cum expeditis peditum equitumque prius pugnam consererent
112
Bennett (1914: II.34) would probably take honestissimum hominem as an expression of ‘preëmi-
nence’, comparable with adjectives like praecipuus.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
cum hoste quam . . . (Liv. 28.14.16); Plani piscium quibus cauda non est . . . (Plin. Nat.
9.158); . . . ut iam minus temerarium videatur e profundo maris petere margaritas
atque purpuras. (Plin. Nat. 33.70); . . . alarios equites ac levis cohortium mittit in eos
qui . . . (Tac. Ann. 3.39.1);
Participles: . . . cum patrem Fabium appellasset circumfusosque militum eius totum
agmen patronos consalutasset . . . (Liv. 22.30.2); . . . perculsos deinde hostium <ut>
fudit . . . (Curt. 4.16.29); Quod Britannorum olim victis evenit. (Tac. Agr. 11.4); Ne
Othonianis quidem incruenta victoria fuit, quorum improvide secutos conversi
equites circumvenerunt. (Tac. Hist. 2.15.2);
Partitive expressions with a noun or proper name as head are regular in specific
pragmatic constellations.114 In (f ), the relative pronoun quorum by definition con-
tains topical information. In (g), Livy describes the array of the army: equitum is the
topic of the sentence. Likewise quadripedum is topical in (h), in a discussion of which
types of animals can see.
113
For more examples of mei, tui, sui, nostri, vostri, sui, see K.-St.: I.598.
114
For a discussion of these constellations from the perspective of word order, see Devine and
Stephens (2006: 368–76).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(i) Relatum inde ad patres a Quintiliano, tribuno plebei, de libro Sibullae, quem
Caninius Gallus quindecimvirum recipi inter ceteros eiusdem vatis . . .
postulaverat.
(‘A proposal was now put to the Fathers by the plebeian tribune Quintilianus
with regard to the Sibylline book; Caninius Gallus, of the Fifteen, demanding
its admission among the other verses of the same prophetess . . . ’ Tac. Ann.
6.12.1)
115
See Shackleton Bailey ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(j) Qui etiam dubitem an hic Anti considam . . . ubi quidem ego mallem duum-
virum quam Romae fuisse.
(‘I am even considering whether I might not settle down here at Antium . . .
where I would rather have been a duovir than at Rome.’ Cic. Att. 2.6.1)
Supplement:
. . . et quidem reliqui in ventre cellae uni locum, / ubi reliquiarum reliquias recon-
derem. (Pl. Cur. 387–8); Nam ducum taceo ducem, pro quo nihil debetur. (Sen. Med.
233–4); Mammas atque tatas habet Afra, sed ipsa tatarum / dici et mammarum
maxima mamma potest. (Mart. 1.100); Moesi quam feri, quam truces fuerint, quam
ipsorum etiam barbari barbarorum horribile dictu est. (Flor. Epit. 4.12.13R =
2.26M); Etiam ipse dominus oravit, cui sit honor et virtus in saecula saeculorum.
(Tert. Or. 29); Rex regum Sapor, particeps siderum . . . Constantio . . . salutem plu-
rimam dico. (Amm. 17.5.2);
116
For references, see Sz.: 55–6.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
11.56 Nouns and noun phrases in the genitive that function as attribute
with quantifying expressions (the genitive of quantity)
As is stated in § 11.49, the label ‘partitive genitive’ is often applied to NPs in the
genitive that denote a plurality of entities or a collective entity and which are used
with head nouns that denote measure or quantity or with various neuter singular
forms of adjectives and pronouns. However, there are several differences between
real partitive expressions and the expressions discussed in this section. In this
Syntax this use of the genitive is called ‘genitive of quantity’. The head constituents
involved are:
(i) nouns and noun phrases denoting quantity;
(ii) neuter singular forms of adjectives of amount;
(iii) neuter singular pronouns;
(iv) adverbs of quantity.
117
So K.-St.: I.429 on Cic. Fin. 4.76. See also Bennett (1914: II.14) on aula cineris in Pl. Am. fr. 4.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Appendix: In examples like (g)–(l) the head noun (phrase) is more specific and the
genitive noun more general. This phenomenon is often labelled ‘genitive of the
rubric’ (genetivus tituli).118
118
For the Appendix, see Woodcock (1959: 53). K.-St.: I.429, 453; Sz.: 71–2; Burkard and Schauer
(2000: 385), and others use the term as well but see it as a subcategory of the partitive genitive or the
genitive of quality. See also the discussion in Löfstedt (1942: I2.126–9). For lucri, see TLL s.v. lucrum
1722.63ff., where a different analysis is given.
119
For this text, see Speidel (1992); for the interpretation of the genitive, see Speidel (1996) ad loc.
120
See Wistrand (1933: 74–5).
121
For a discussion of these expressions, see Ripoll (2010b); he regards them as adverbs.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
genitive; with abstract nouns, as in (e), the whole expression functions rather like an
expression of degree.
(a) Paululum / praedae intus feci . . .
(‘I took a little bit of booty inside . . . ’ Pl. Poen. 802–3)
(b) . . . eo addito brassicae coliculos duos, betae coliculos II cum radice sua,
feliculae pullum, herbae mercurialis non multum, mitulorum libras II,
piscem capitonem et scorpionem I, cochleas sex et lentis pugillum.
(‘ . . . add two cabbage leaves, two beet plants with the roots, a shoot of fern, a bit of
the mercury-plant, two pounds of mussels, a capito fish and one scorpion, six snails,
and a handful of lentils.’ Cato Agr. 158.1)
(c) Ne ille haud scit hoc paullum lucri quantum ei damnum adportet.
(‘Truly he doesn’t know how much loss this little bit of gain will bring him.’ Ter. Hau. 747)
(d) Nam nunc lenonum et scortorum plus est fere / quam olim muscarum est
cum caletur maxume.
(‘Now there are practically more pimps and prostitutes than there are flies at the
height of summer.’ Pl. Truc. 64–5)
(e) Tantum gemiti et mali maestitiaeque / hic dies mi optulit . . .
(‘This day’s brought me so much groaning and trouble and sadness . . . ’ Pl. Aul. 722–3)
Supplement (in alphabetical order by governing expression):
Nouns in the genitive: Nec castra forent retenta, ni exiguum superfuisset diei. (Liv.
9.42.6); . . . quam minimum salis habeant. (Cato Agr. 135.3); . . . peregrinus est, /
minus potens quam tu, minus notus, minus amicorum hic habens. (Ter. Eu.
759–60); Ego autem iuris civilis studio multum operae dabam Q. Scaevolae, P. f. . . .
(Cic. Brut. 306); . . . cum anceps proelium in multum diei processisset, tum . . . (Liv.
9.44.11); In convivio rex erat et in multum vini processerat. (Liv. 37.7.12); . . .
multumque laboris exhaustum est, dum pervenitur ad solidum. (Sen. Ep. 52.5); Ipse
equos phalerasque et sagula donum principibus ferre et parvom auri, quod inter paucos
divideret . . . (Liv. 44.26.6); Ephelidem tollit resina cui tertia pars salis fossilis et paulum
mellis adjectum sit. (Cels. 6.5); Aquatiles autem piscium naturae . . . humoris habent
oppido quam paulum . . . (Vitr. 1.4.7); Nam non est veri simile, hominem pauperem /
pauxillum parvi facere quin nummum petat. (Pl. Aul. 111–12); Erat ei de ratiuncula /
iam pridem apud me relicuom pauxillulum / nummorum. (Ter. Ph. 36–8); . . . ut ne
legens quidem umquam senserit quantum flagiti commisisset? (Cic. Brut. 219);
Substantival neuter adjectives in the genitive: Ecastor minus / quam urbani, et multo
minus mali quaerunt sibi. (Pl. Mer. 716–17); Plus aegri ex abitu viri, quam ex adventu
voluptatis cepi. (Pl. Am. 641); Iubeo promi utrosque, ne quid plus mali nasceretur . . .
(Cic. Ver. 4.32); Sed calidi plus est illis quibus acria corda / iracundaque mens facile
effervescit in ira . . . (Lucr. 3.294–5); Non vides quantum mali ex <ea> re excites? (Ter.
Hau. 1013); Ecquid ergo intellegis, quantum mali de humana condicione deieceris?
(Cic. Tusc. 1.15); Tantum ego nunc porto a portu tibi boni. (Pl. Capt. 869);
Prepositional alternatives: Pulticula etiam, cum qua paulum ex favo vetere coctum
(v.l. cocti) sit . . . (Cels. 4.26.8);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The notion ‘quantity’ is less applicable in the following cases where the noun in the
genitive is an abstract noun, as in (e)–(h). Since the entities such genitives refer to are
not divisible, the entire expression describes rather the type or the degree of the entity
in the genitive. For the use of these genitival expressions as subject and object
complement see § 9.32 and § 9.42, respectively.
(e) . . . rogitas / quo ego eam, quam rem agam, quid negoti geram . . .
(‘ . . . you ask me where I’m going, what I’m doing, what business I’m carrying out . . . ’
Pl. Men. 115–16)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(f ) Si res tibi forte notior est quam mihi aut si Plancum rogandum putas, scribas
ad me velim ut quid rei sit et quid rogandum sciam.
(‘If you happen to know more about it than I do or if you think I ought to ask
Plancus, would you please write so that I know what it’s about and what I ought to
say?’ Cic. Att. 12.52.1)
(g) Nam cum omnis iuventus, omnes etiam gravioris aetatis in quibus aliquid
consilii aut dignitatis fuit eo convenerant . . .
(‘For, on the one hand, all the fighting men, nay, all the older men who had any
sagacity or distinction, had there assembled . . . ’ Caes. Gal. 3.16.2)
(h) . . . id modo si mercedis / datur mi, ut meus victor vir belli clueat.
(‘ . . . if only my reward is that my husband is renowned as winner in war.’ Pl. Am.
646–7)
Supplement:
Quid nunc consili captandum censes? (Pl. As. 358); Nam hoc negoti clandestino ut
agerem mandatum est mihi. (Pl. Mil. 956); . . . observes filium / quid agat, quid cum
illo consili captet. (Ter. An. 169–70); Quid illos, bono genere gnatos, magna virtute
praeditos, opinamini animi habuisse . . . (Cato orat. 58); . . . tuum est, Caesar, qui pro
multis saepe dixisti, quid mihi nunc animi sit ad te ipsum referre . . . (Cic. Deiot. 7);
Tibi autem idem consili do quod mihimet ipsi . . . (Cic. Fam. 9.2.2); Temptandum
existimavit, quidnam Pompeius propositi aut voluntatis ad dimicandum haberet.
(Caes. Civ. 3.84.1); . . . prius in hostium castris constiterunt quam plane ab his videri
aut quid rei gereretur cognosci posset. (Caes. Gal. 3.26.3); . . . quibus id consilii fuisse
cognoverint ut, si flumen transissent, una ex parte ipsi, altera Arverni se circumsis-
terent. (Caes. Gal. 7.5.5); . . . inflatusque amplius XX legionum numero in id furoris
processerat ut . . . (Vell. 2.80.2);
NB: . . . aliud alii commodi aliquo adiuncto incommodo muneratur. (Cic. Inv. 2.3);
Instead of abstract nouns, genitive forms of substantival neuter adjectives can be used
as modifiers of neuter pronouns and neuter forms of adjectives of amount, for
example aliquid novi ‘something new’. This is in principle limited to adjectives of
the second declension, as in (i)–(k). (Note in (k) that tantum mali is determined by
hoc.) Third declension adjectives are exceptionally so used in parellel expressions
such as those in (l), where eminentis is in the genitive.
(l) Illud video pugnare te species ut quaedam sit deorum quae nihil concreti
habeat, nihil solidi, nihil expressi, nihil eminentis . . .
(‘I am aware that what you maintain is that the gods possess a certain outward
appearance, which has no firmness or solidity, no definite shape or outline . . . ’ Cic.
N.D. 1.75)
Supplement:
Quin incommodi plus malique ilico adsit, boni si optigit quid. (Pl. Am. 636); Aut
quid is iniqui fit? (Pl. Rud. 647); Hic siquid nobis forte advorsi evenerit . . . (Ter. Hau.
355); Quid in ea boni sit salubritatisque. (Cato Agr. 157.1); . . . petivi saepe per
litteras . . . a Curione tribuno, non ut decernatur aliquid novi, quod solet esse diffici-
lius, sed ut ne quid novi decernatur . . . (Cic. Fam. 2.7.4);
. . . tantum adest boni improviso, verum commixtum malo. (Pl. As. 310); Non tu
scis quantum isti morbo nunc tuo facias mali? (Pl. Men. 911); Sed si ex ea re plus
mali’st quam commodi . . . (Ter. An. 547);
Appendix: The normal expression for third declension adjectives in -is is exemplified
by senile in (m), with a substantival neuter adjective determined or modified by a
pronoun or an adjective of amount (see § 11.13). This option also exists for second
declension adjectives, as in (n) (note the variation!) and is the regular one in the case
of parallel structures, as in (o), where tantum . . . bonum is in parallel with tam
gaudium grande. It is the regular alternative for second declension adjectives if
they are modified in some way, as ineptum is by tam in (p), but see the Supplement.
(m) Ut enim adulescentem in quo est senile aliquid, sic senem in quo est aliquid
adulescentis, probo.
(‘For just as I approve of the young man in whom there is a touch of age,
so I approve of the old man in whom there is some of the flavour of youth.’
Cic. Sen. 38)
(n) Quidquid incerti mi in animo prius aut ambiguom fuit . . .
(‘Whatever was uncertain or unclear in my heart before . . . ’ Pl. Ps. 759)
(o) Tantum a portu adporto bonum, tam gaudium grande adfero . . .
(‘I’m bringing such a great good from the harbour and such great joy . . . ’ Pl.
St. 295)
(p) Miror quo modo / tam ineptum quicquam tibi venire in mentem, mi vir,
potuerit.
(‘I’m amazed that such a silly idea could have entered your head, my dear
husband.’ Ter. Hau. 1004–5)
Supplement:
. . . si quid med erga hodie falsum dixeris . . . (Pl. As. 20); Quare omnis vos oratos volo
/ ne plus iniquom possit quam aequom oratio. (Ter. Hau. 26–7); Nihil enim dicam
reconditum, nihil exspectatione vestra dignum, nihil aut inauditum vobis aut
cuiquam novum. (Cic. de Orat. 1.137); Eadem ratio habet in se quiddam amplum
atque magnificum . . . (Cic. Fin. 2.46); Nam qui se ipse norit primum aliquid se
habere sentiet divinum . . . (Cic. Leg. 1.59);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Cum essent duo Terentii aut plures, discernendi causa, ut aliquid singulare haberent,
notabant . . . (Var. L. 9.60); Nihil humile de Tarquinio, nihil sordidum accepimus.
(Cic. Phil. 3.10); Aut si rei dignitas adimet iocandi facultatem, aliquid triste, novum,
horribile statim non incommodum est inicere. (Cic. Inv. 1.25);
. . . si quicquam in vobis, non dico civilis, sed humani esset . . . (Liv. 5.3.9); ‘Fulvus’
autem videtur de rufo atque viridi mixtus in aliis plus viridis, in aliis plus rufi habere.
(Gel. 2.26.11);
There is much discussion about why the genitive of third declension adjectives in -is
is avoided in the partitive construction, except when used in combination with
second declension adjectives. Probably, the -is form might be misunderstood as
nominative and is therefore avoided. It is also avoided in other constructions, except
in parallel pairs, as in (q) and (r).122
(q) . . . ea pars senatus, cui potior utilis quam honesti cura erat . . .
(‘ . . . that part of the senate to whom the pursuit of advantage was more
important than that of honour . . . ’ Liv. 42.47.9)
(r) . . . sine ulla mora veniat illa turpis honestique distinctio.
(‘ . . . the great distinction between base and honourable action presents itself
without any delay.’ Sen. Ben. 7.2.1)
Another question is whether there is a difference in meaning between aliquid novi
and aliquid novum.
Appendix: There are a number of idioms formed by the combination of a singular
neuter pronoun with the genitive of a noun; these idioms can perform various
functions in the clause. Examples are (s) and (t). Note also the adverb ubi in (u).
(s) Haec eri immodestia / coegit, me qui hoc noctis a portu ingratiis excitavit.
(‘My master’s lack of restraint has forced me into it, who routed me out of
the harbour against my will at this time of night.’ Pl. Am. 163–4)
(t) Neque puduit eum id aetatis sycophantias / struere . . .
(‘And he was not ashamed to play tricks at his age . . . ’ Pl. As. 71–2)
(u) Non edepol nunc ubi terrarum sim scio . . .
(‘I don’t know where on earth I am’ Pl. Am. 336)
122
The examples are taken from Svennung (1935: 207). For discussion, see Baldi (1977, 1983).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Fluit aes rivis aurique metallum . . . (Verg. A. 8.445); Lucus ibi frequenti silva et
proceris abietis arboribus saeptus . . . (Liv. 24.3.4); Scrobe<m> quoquoversus pedum
IIII ab arbore olivae tam longe fodito, ut . . . (Col. 5.11.13);
123
For Cicero’s use of the ablative (some 445 instances) and genitive (twenty-seven), see Stegmann
(1885: 243–7, 1887: 265–8).
124
Some hundred attestations are critically discussed by Edwards (1902: 469ff.). On ex. (f ), much
discussed among Latinists, see Pasoli (1966: 14–33).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) Quo verius in causa nostra vir magni ingenii summaque prudentia, L. Cotta
dicebat . . .
(‘For this reason Lucius Cotta, a man of great talent and the highest wisdom, was all
the more surely correct when he said about my own case . . . ’ Cic. Leg. 3.45)
(f ) Lucreti poemata ut scribis ita sunt, multis luminibus ingeni, multae tamen
artis.
(‘Lucretius’ poetry is as you say—sparkling with natural genius, but also marked by
plenty of technical skill.’ Cic. Q. fr. 2.10[9].3)
Supplement:
Carthaginienses fratres patrueles duo / fuere, summo genere et summis ditiis. (Pl.
Poen. 59–60); Eu edepol specie lepida mulierem. (Pl. Rud. 415); Rufamne illam
virginem, / caesiam, sparso ore, adunco naso? (Ter. Hau. 1061–2); . . . Lucium vero
Lamiam, equitem Romanum, praestanti dignitate hominem . . . consul imperiosus
exire ex urbe iussit. (Cic. Red. Sen. 12); . . . vir incredibili fide, magnitudine animi,
constantia, L. Ninnius, ad senatum de re publica rettulit . . . (Cic. Sest. 26); Sulpicius,
et summa auctoritate et optime sentiens, non adest. (Cic. Fam. 12.2.3—NB:
coordination); . . . qui ingenti magnitudine corporum Germanos, incredibili virtute
atque exercitatione in armis esse praedicabant . . . (Caes. Gal. 1.39.1); Cum ergo tam
magna potentia reges non contempserint latericiorum parietum structuras . . . (Vitr.
2.8.16); CORPORE FUSCO, FA<C>IE <D>DED<U>CTA, NASO RECTO, LENTIGO125 MALO /
D<E>XTRO ( = mala dextra) (CPL 210.2–3 (Karanis, second half II AD));
Coordination of ablative and genitive: . . . posteroque die Thuyn, hominem maximi
corporis terribilique facie, quod et niger et capillo longo barbaque erat promissa,
optima veste texit . . . (Nep. Dat. 3.1); Glaucion in Syria et Parthia nascitur, humilis
herba, densis foliis fere papaveris, minoribus tamen sordidioribusque, odoris taetri,
gustus amari cum adstrictione. (Plin. Nat. 27.83); Fine anni excessere . . . et
Q. Haterius, familia senatoria, eloquentiae quoad vixit celebratae. (Tac. Ann. 4.61.1);
There are no sure cases of nouns in the ablative without an attribute functioning as
attribute of another noun.126
125
The nominative follows exactly a formula used in Greek documents. See Adams (2003b: 81–2).
126
See Sz.: 119.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
NPs in the dative can be used as attributes in a way that resembles beneficiary
adjuncts at the clause level (see § 10.70). Such dative constituents can be animate,
as in (a), and abstract, as in (b). In (a), the food is intended for the sheep. In (b),
receptui can hardly be understood as a clause constituent with audire, whereas it must
be so understood with dedit in (c). Both in (a) and in (b) a genitive constituent is
possible as well, which does not imply that they have exactly the same meaning. In the
case of signum, for example, as in (d), the genitive does not mean: ‘the sign (ordering)
to leave’, but ‘the sound that typically signals leaving’ (as opposed to other sounds
used in other situations). In Late Latin the dative is replaced by prepositional
alternatives with ad, following earlier instances like (e).128 (For the dative in com-
bination with a so-called predicative dative, see § 9.34.)
127
For more instances in Vitruvius, see Wistrand (1933: 69–70).
128
For optional attributes in the dative in Early Latin, see Bennett (1914: II.183–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Accusative NPs are rarely used as optional attribute, with the exception of extent of
time expressions, such as dies quindecim in (a) and triduum in (b), and extent of space
expressions, as in (c).
(a) Ob easque res ex litteris Caesaris dies quindecim supplicatio decreta est,
quod ante id tempus accidit nulli.
(‘And for those achievements, upon receipt of Caesar’s despatches, a fifteen days’
thanksgiving was decreed, an honour that had previously fallen to no man.’ Caes.
Gal. 2.35.4)
(b) . . . supplicatio triduum pro collegio decemvirorum imperata fuit . . .
(‘ . . . a three-day period of prayer was proclaimed in the name of the college of
decemvirs . . . ’ Liv. 38.36.4)
129
For references, see Sz.: 95.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
130
For instances of prepositional attributes in general (both arguments and optional attributes), see
K.-St.: I.213–18. Wharton (1996) has statistical observations on Cato, Varro, Cicero, Livy, Columella, and
Tacitus. Additional data in Wharton (2009). For prepositional phrases modifying verbal nouns in Early
Latin, see Rosén (Hannah) (1981: 92–6, 1999: 23–5). For Vitruvius, who has many prepositional
attributes, see Wistrand (1933: 70) and Adams (forthc.). The TLL articles are usually very helpful. For
a list of prepositional phrases that function as arguments with emotion and action nouns in Cicero, see
Froehlich (1889: 7–11).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The semantic relations between the head NPs and their optional attributive prepos-
itional phrases include local relations (place, time, direction, source, provenance, etc.)
and more abstract relations, such as interpersonal relations, purpose and result,
content, and partitive relations (for the latter, see § 11.68). It is clear from this list
of relations that a small number of prepositions are used very frequently in this
context: ab, ad, cum, de, ex, in, post,131 and sine. In what follows, a number of
examples of NPs (including proper names) functioning as head are given first;
pronominal and other types of heads are then dealt with separately later on.
In (f ), the prepositional phrase ex Rhodo indicates the place where the ship comes
from. In (g), the nouns impietatem and iniuriam are modified by prepositional
phrases with in expressing an interpersonal relation (some scholars might prefer to
consider iniuria bivalent). In (h), the use of the optional prepositional phrase ad
gloriam with the monovalent noun ardor (mentis) runs parallel with the preceding
bivalent noun cupiditate (vincendi). In (i), the locative prepositional phrase is an
optional attribute with fundus. Ex. (j) has a partitive expression and (k) illustrates the
‘content’ relationship. Ex. (l) illustrates the semantic relation of ‘provenance’ or
‘origin’. Ex. (m) illlustrates the increasing use of the preposition de instead of a
regular (‘possessive’) genitive in Classical Latin.
(f ) . . . atque ego conspicor / navem ex Rhodo qua est heri advectus filius.
(‘ . . . and I spot the ship from Rhodes on which my son arrived yesterday.’ Pl. Mer.
256–7)
(g) Ita ad impietatem in deos in homines adiunxit iniuriam.
(‘Thus to impiety towards the gods he added injustice towards men.’ Cic. N.D. 3.84)
(h) De impetu animi loquor, de cupiditate vincendi, de ardore mentis ad
gloriam.
(‘I am speaking about his impetuosity, his eagerness to win, his ardent desire for
glory.’ Cic. Cael. 76)
(i) . . . si propter pastiones tuus fundus in Rosia probandus sit . . .
(‘ . . . if your place in the Rosea is to be commended for its pasturage . . . ’ Var. R.
3.2.10)
(j) . . . homo ex numero disertorum postulabat ut . . .
(‘ . . . one of our accomplished advocates applied for . . . ’ Cic. de Orat. 1.168)
(k) Quo in genere severitatem maiorum senatus vetus auctoritas de
Bacchanalibus . . . declarat.
(‘The strictness of our ancestors in this matter is shown by the ancient decree of the
Senate with respect to the Bacchanalia . . . ’ Cic. Leg. 2.37)
131
See TLL s.v. post 184.56ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(q) Duae factiones senatum distinebant, una, cuius principes erant defectionis a
populo Romano auctores, altera fidelium civium.
(‘Two factions kept the senate divided: one of these had for leaders the men who had
inspired the revolt from Rome, the other was composed of loyal citizens.’ Liv. 9.16.6)
It is not always easy to decide whether a prepositional phrase modifies a noun
(phrase) or has a function of its own in its clause. Two examples with ab are (r)
and (s). Similarly ambiguous cases with ‘partitive’ de are (t) and (u). The contextual
features mentioned above (including the order of the constituents involved) are often
absent. (For more partitive expressions, see next section.)
Supplement:
Prepositional attributes (in alphabetical order by preposition): Praeterea autem,
dum illum quaero, a villa mercennarium / vidi. (Ter. Ad. 541–2); Legationes autem a
regibus, cum putant eos sua causa reticere, sumptus atque pecunias maximas
praebent . . . (Gracch. orat. 44M); . . . litterae quoque ab Hannibale ad Philippum
inventae . . . (Liv. 22.34.7); De epistula ad Caesarem quid egeris exspecto. (Cic. Att.
12.52.2); Post pugnam ad Regillum non alia illis annis pugna clarior fuit. (Liv.
2.31.3); Interim accidit illa calamitas apud Leuctra Lacedaemoniis. (Nep. Ag.
6.1); . . . cum omni argento et auro Capuam et urbes circa Capuam occuparint?
(Cic. Agr. 1.22); Sic enim habebit . . . circa cellam ambulatio auctoritatem. (Vitr.
3.3.6); . . . qui mi intro misti in aedis quingentos coquos, / cum senis manibus, genere
Geryonaceo. (Pl. Aul. 553–4); Sequitur deinceps cum officio selectio . . . (Cic. Fin.
3.20); Ita sunt multi, quibus videmus optabilis mortes fuisse cum gloria. (Cic. Tusc.
1.116); Unam mehercule tecum apricationem in illo lucrativo tuo sole malim quam
omnia istius modi regna . . . (Cic. Att. 7.11.1); Ibi turres cum ternis tabulatis erigebat.
(Caes. Civ. 1.26.1); Iam et incesta donabis propter Loth et fornicationes cum incesto
propter Iudam . . . (Tert. Pud. 6.10); Nam eam compressit de summo adulescens loco.
(Pl. Rud. 28); Nam mihi quidem cottidie augescit magis / de filio aegritudo . . . (Ter.
Hau. 423–4); Relicuum est igitur crimen de veneno. (Cic. Cael. 56); . . . et iam cetera
de genere hoc quaecumque secuntur . . . (Lucr. 3.481); . . . cum ille Agrippae iuvenis
132
More instances of ab can be found in TLL s.v. 16.59ff.; of de in TLL s.v. 56.6ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(v) Tu qui urnam habes aquam ingere . . . / Te cum securi caudicali praeficio
provinciae.
(‘You there, with the pail, fetch water. You with the axe I make head of the
wood-cutting department.’ Pl. Ps. 157–8)
(w) Hos ad magistros si qua te fortuna, Cato, cum ista natura detulisset . . .
(‘If some happy chance, Cato, had carried a man of your character off to
these masters . . . ’ Cic. Mur. 64)
(x) Nil moror eum tibi esse amicum cum eius modi virtutibus.
(‘I don’t care for you having him as your friend with virtues of this kind.’
Pl. Trin. 337)
133
‘Talia saepe in sacr. script.’ TLL s.v. 184.69.
134
For Ovid’s use of prepositional attributes and adverbs, see Bömer ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Prepositional phrases that denote the whole from which the head constituent is taken
or to which it belongs are common from Early Latin onwards. Examples are (a)–(i).
(a) Si moreretur, ne expers partis esset de nostris bonis.
(‘ . . . so that, if she died, she would not have been without a part of our possessions.’
Ter. Hau. 652)
(b) Nam ex illa pecunia magnam partem ad se vertit, mulieri reddidit quantu-
lum visum est.
(‘He appropriated a considerable part of that sum of money, only returning a modest
fraction to the woman herself.’ Cic. Div. Caec. 57)
(c) Quos ubi qui proximi Rhenum incolunt perterritos senserunt, insecuti
magnum ex his numerum occiderunt.
(‘And when the tribes which dwell next to the Rhine perceived their panic, they
pursued and slew a great number of them.’ Caes. Gal. 1.54.2)
(d) Quid si de vostro quippiam orem?
(‘What if I were to ask for something from your personal possessions?’ Pl. Truc. 6)
(e) Nam cum quidam ex eius adsentatoribus, Damocles, commemoraret in
sermone copias eius . . .
(‘For when one of his flatterers, named Damocles, dilated in conversation upon his
troops . . . ’ Cic. Tusc. 5.61)
(f ) Vellem aliquis ex vobis robustioribus hunc male dicendi locum suscepisset.
(‘I could wish that one of the more hardened among you accusers had taken upon
himself the part of slanderer.’ Cic. Cael. 7)
(g) Nemo de nobis unus excellat.
(‘Let none of us distinguish himself alone.’ Cic. Tusc. 5.105)
(h) Pleraque enim de iis quae ab isto obiciebantur cum confiterer, tamen ab illo
maiestatem minutam negabam.
(‘For, while admitting most of our friend’s indictment, I still maintained that the
defendant was not guilty of treason.’ Cic. de Orat. 2.107)
(i) Te unum ex omnibus amat.
(‘You alone she loves out of all men.’ Pl. Truc. 186)
Supplement (in alphabetical order by preposition):
Scuto ab novissimis uni militi detracto . . . (Caes. Gal. 2.25.2—disputed, see K.-St.:
I.495);
Tametsi quidam de Italicis, quos propter amplitudinem praestare dicunt, [ad]
victimas faciunt . . . (Var. R. 2.5.10); Dicat qui volt hodie de illo populo concitato . . .
(Cic. Clu. 90); . . . est nescio quis de eis quos amicos nobis arbitramur qui nostra
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
consilia ad adversarios deferat. (Cic. Clu. 143); Itaque cum quidam de collegis nostris
agrariam curationem ligurrirent, disturbavi rem . . . (Cic. Fam. 11.21.5); Qua re sic tibi
eum commendo ut unum de meis domesticis et maxime necessariis. (Cic. Fam.
13.71.1); Pleraque enim de eis quae ab isto obiciebantur cum confiterer . . . (Cic. de
Orat. 2.107); Et pauci de nostris cadunt. (Caes. Gal. 1.15.2);
Qui autem ex numero civium dementia aliqua depravati hostes patriae semel esse
coeperunt, eos . . . nec vi coercere nec beneficio placare possis. (Cic. Catil. 4.22); An
non intellegis . . . quot ex his qui vivunt eodem crimine in summum periculum
capitis arcessas? (Cic. Rab. Perd. 26); . . . te non fato, ut ceteros ex vestra familia,
sed opera sua consulem factum . . . (Cic. Ver. 29); Sed his commemorandis etiam illud
adsequor, ut intellegatis primum ex omni numero quam non multi ausi sint dicere,
deinde ex iis ipsis quam pauci fuerint laude digni. (Cic. Brut. 270); . . . plerique ex his
ne in angustias inciderent X pedum munitione se in fossas praecipitabant . . . (Caes.
Civ. 3.69.3); Sed compluribus interfectis cupidius insecuti nonnullos ex suis ami-
serunt. (Caes. Gal. 5.15.2); Tres ex his condemnati sunt. (Liv. 33.42.10); . . . quamvis
pleraque ex his sanguinem supprimunt. (Cels. 5.26.21B); Epaminondas . . . plerosque
ex militibus suis muliebri ornatu inmiscuit. (Fron. Str. 3.2.7);
. . . magnumque ex iis numerum occidit atque omnes armis exuit. (Caes. Gal.
5.51.5); . . . solent totam bulen atque etiam e plebe non exiguum numerum vocare . . .
(Plin. Ep. 10.116.1);
NB: . . . mox M. Servilium e consularibus aliosque testis inlexit ad proferenda quae
velut reicere voluerat. (Tac. Ann. 3.22.2);
(a) . . . et si non solum inpulsu scutorum neque conflictu corporum neque ictu
comminus neque coniectione telorum sed . . .
(‘ . . . and if . . . it is not only by the weight of the enemy’s shields nor the shock of
impact, not by blows struck in close combat nor missiles hurled from a distance . . . ’
Cic. Caec. 43)
(b) . . . fugam magis retro quam proelium aut hostem spectante milite . . .
(‘ . . . the soldiers were thinking more of retreating back (to the camp) than of the
battle or the enemy . . . ’ Liv. 8.19.7)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Intuemini enim horum deinceps annorum vel secundas res vel adversas.
(‘For consider these pasts few years in order, with their successes and reverses.’ Liv. 5.51.5)
(d) Atenim illa tunc caecitas longe a finibus saeculi habebatur.
(‘And yet that blindness then was felt long before the ends of the world.’ Tert. Uxor.
1.5.4)
(e) Deinde consules Carbo tertium et C. Marius, septiens consulis filius, annos
natus XXVI . . .
(‘Carbo now became consul for the third time, in conjunction with Gaius Marius,
now aged twenty-six, the son of a father who had been seven times consul.’ Vell.
2.26.1)
(f ) Si nimia tristitia, prodest lenis sed multa bis die frictio . . .
(‘If there is excessive depression light and prolonged rubbing twice a day is
beneficial . . . ’ Cels. 3.18.22)
Less convincing is (g), generally accepted as an instance of attributive use of nunc, but
the only evidence offered is the fact that it can be translated into Greek H F
I Łæ ø . A borderline case is (h) where the position of ante supports taking it as the
attribute of malorum as does the fact that at that time Greek influence (and in Bible
translations Semitic influence) is not unlikely.135
(g) Non tu nunc hominum mores vides?
(‘Can’t you see the ways of people nowadays?’ Pl. Per. 385)
(h) O socii (neque enim ignari sumus ante malorum).
(‘O comrades—for ere this we have not been ignorant of previous misfortunes . . . ’
Verg. A. 1.198)
Supplement:
Saturnum et Opem ceterosque tunc homines humanam carnem solitos esitare. (Enn.
var. 113); Eho tu, Glycerium hinc civem esse ais? (Ter. An. 908); Sub rostro duas ut
mammulas pensiles habeant. (Var. R. 2.3.2); Praefuisse classi populi Romani Sicu-
lum, perpetuo sociis atque amicis Syracusanum. (Cic. Ver. 5.131); Quae si in extremo
breviora sunt, infringitur ille quasi verborum ambitus. (Cic. de Orat. 3.186); Nec
tamen ea species corpus est sed quasi corpus, nec habet sanguinem sed quasi
sanguinem. (Cic. N.D. 1.49); Ex Asia rediens cum ab Aegina Megaram versus
navigarem, coepi regiones circumcirca prospicere. (Sulp. Ruf. Fam. 4.5.4); Ita pridie
duorum centurionum interitio hac adversariorum poena est litata. (B. Hisp. 24.6);
Atque ita terrigenis rigido de fratribus unum / comminus ense ferit; iaculo cadit
eminus ipse. (Ov. Met. 3.118–19); Populus superamur ab uno / vixque viro (Ov. Met.
12.499–500); Omnia enim inde suspecta erant. (Liv. 6.22.1); Haud dubia res visa
quin per invia circa nec trita antea quamvis longo ambitu circumduceret agmen.
135
More unconvincing instances can be found in K.St.: I.220. The best discussion of the topic is Ripoll
(2010a). For the use of adverbs as modifiers in Bible translations, see Rubio (2009: 214–15).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(Liv. 21.36.4); . . . per nudam infra glaciem fluentemque tabem liquescentis nivis
ingrediebantur. (Liv. 21.36.6); Postero ac deinceps aliquot diebus ad portas maior
prope mulierum quam virorum multitudo stetit . . . (Liv. 22.7.11); . . . precibus age-
bant ne proderent patriam tyranni ante satellitibus et tum corruptoribus exercitus.
(Liv. 24.32.5); Crocodilus quoque inde ob argumentum hoc Caesareae in Iseo dicatus
ab eo spectatur hodie. (Plin. Nat. 5.51); A Vergiliarum exortu significant Caesari
postridie arcturi occasus matutinus, III id. Mai. fidiculae exortus . . . (Plin. Nat.
18.255); . . . non obscuris, ut antea, matris artibus, sed palam hortatu. (Tac. Ann.
1.3.3); . . . hostilibus circum litoribus aut ita vasto profundo, ut credatur novissimum
ac sine terris mare. (Tac. Ann. 2.24.1); . . . iubetque praevenire conatus consulis,
occupare velut arcem eius . . . (Tac. Ann. 15.69.1); Nam illa cottidie tua Lorium
ventio . . . (Aur. Fro. Aur. 2.19.1 vdH2); . . . quae omnium retro principum devota in
Deum definivit auctoritas . . . (Gesta coll. Carth. 1.4, l. 13–14);
Much more common is the use of adverbs with NPs that function as apposition,
attribute, or secondary predicate. A few examples are given below. In (i), the degree
adverb admodum modifies the noun senex that functions as secondary predicate. In
(j), the truth value adverb plane modifies the noun vir that functions as an appos-
ition.136 In (k), the degree adverb magis modifies the noun asinos that functions as an
attribute with homines.
(i) . . . cum quidem ille admodum senex suasor legis Cinciae de donis et mune-
ribus fuit.
(‘ . . . and he, at that very time, though far advanced in age, made speeches in favour of
the Cincian law on fees and gifts.’ Cic. Sen. 10)
(j) At vero C. Marius, rusticanus vir, sed plane vir, cum secaretur . . .
(‘But as a matter of fact C. Marius, a countryman by extraction yet undoubtedly a
man, when under the surgeon’s knife . . . ’ Cic. Tusc. 2.53)
(k) Neque ego homines magis asinos numquam vidi.
(‘I've never seen human beings who were such asses.’ Pl. Ps. 136)
136
TLL s.v. planus 2342.50ff. gives more instances (hesitantly called ‘degree’ expressions s.v.
2341.24ff.).
137
For the various constellations of adnominal arguments, see Bertocchi and Maraldi (1990: 73–5). See
also Spevak (2014b) on noun valency.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 11.5 Types of constituents that may function as adnominal arguments with
verbal nouns
Type of modifier Examples For more
examples,
see §
Nouns and noun Quis de C. Cethego atque eius in Hispaniam 11.71
phrases and various profectione ac de volnere Q. Metelli Pii cogitat . . . ?
pronouns (‘Who thinks of Gaius Cethegus and his departure
for Spain, and of his assault upon Quintus
Metellus Pius . . . ?’ Cic. Sul. 70)
Opulento homini hoc servitus dura est . . .
(‘This is why being a wealthy man’s slave is hard . . . ’
Pl. Am. 166)
Prepositional Quid tibi ad hasce accessio aedis est prope aut pultatio? 11.72
phrases (‘What’s the meaning of your coming near this house
or knocking on it?’ Pl. Truc. 258)
Possessive Sed tua profectio spem meam debilitat. 11.73
adjectives (‘But your departure weakens my hope of success.’
Cic. Att. 5.4.1)
Adverbs Quid tibi huc ventio est? 11.74
(‘Why are you coming here?’ Pl. Truc. 622)
Various types of Cum a me quoque id responsum tulisses me nullo modo Ch. 17
embedded clauses posse isdem parietibus tuto esse tecum . . .
(‘And when you had received answer from me that
I could not possibly be safe in the same house with
you . . . ’ Cic. Catil. 1.19)
(d) Quid tibi hanc curatio est rem, verbero, aut muttitio?
(‘Why are you interfering in this matter, you whipping stock, or why are you
muttering?’ Pl. Am. 519)
(e) Serva erum, cave tu idem faxis alii quod servi solent, / qui ad eri fraudatio-
nem callidum ingenium gerunt.
(‘Save your master, don’t do the same as other slaves do, who have cunning ways only
in order to cheat their master.’ Pl. As. 257–8)
(f ) Offendes dissignationem Tyrannionis miri ficam librorum meorum . . .
(‘You will find that Tyrannio has made a wonderful job of arranging my books.’ Cic.
Att. 4.4a.1)
(g) Scipionem eadem spes provinciae atque exercituum impellit . . . simul iudi-
ciorum metus . . .
(‘Scipio is stimulated by the same hope of a province and of armies . . . ; also by the
dread of the law courts . . . ’ Caes. Civ. 1.4.3)
(h) (sc. senex) . . . difficilis, querulus, laudator temporis acti / se puero, castigator
censorque minorum.
(‘ . . . peevish, surly, given to praising the days he spent as a boy, and to reproving and
condemning the young.’ Hor. Ars 173–4)
(i) Differor / cupidine eius.
(‘I'm being torn apart by my desire for her.’ Pl. Poen. 156–7)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Distinguishing between the subjective and objective use of the genitive does not
entail that Latin had two (or more) separate genitives. The role of the genitives in ex.
(j) is the same in all three parallel noun phrases: the genitive marks one noun as the
attribute of the other. It is not easy to distinguish verbal event nouns from others.
Some scholars regard more nouns as such than others do, including e.g. officium
‘obligation’, ‘duty’, and opera, as in (k).138
(j) Non ego illam mi dotem duco esse quae dos dicitur / sed pudicitiam . . . /
deum metum (objective genitive), parentum amorem (objective genitive) et
cognatum concordiam (subjective genitive) . . .
(‘I don’t consider that to be my dowry which is called a dowry, but
chastity, . . . fear of the gods, love for my parents, friendship with relatives . . . ’
Pl. Am. 839–41)
(k) Venit imber, perlavit parietes, perpluont, / tigna putefacit, perdit operam fabri.
(‘Rain comes and soaks the walls; they let in water; the rain rots the timbers
and destroys the builder’s work.’ Pl. Mos. 111–12—NB: It is more likely that
opera is a result noun and the genitive possessive.)
It is commonly held that if a noun governs both a subjective and an objective
genitive, the former precedes the head noun and the second follows. It is unlikely
that there is a syntactic explanation for such an ordering. The one that precedes is
usually known from or inferrable from the preceding context. See Chapter 24.139
(iii) An example of a trivalent noun (with the agent unexpressed) is (l). The second
argument is in the genitive, the argument corresponding to the third argument in a
clause with trado is in the dative. Note the presence of the manner/instrument
adjunct nexu. In (m), translatio has two arguments in the genitive, corresponding
to the agent and the patient in a clause with transfero, as well as the prepositional
phrases a iustis dominis and ad alienos, the status of which is not immediately clear
(argument or satellite?). In (n), which contains a support verb construction, the
patient with deditionem is in the genitive, the recipient in the dative. Note the
secondary predicate vivorum with eorum. A rare instance of marking with the genitive
what would be a third argument at the clause level is maiestatis in (o).
(l) Abalienatio est eius rei . . . traditio alteri nexu . . .
(‘Abalienatio is . . . transfer of a thing to another person with legal obligation . . . ’ Cic.
Top. 28)
(m) Quare L. Sullae, C. Caesaris pecuniarum translatio a iustis dominis ad
alienos non debet liberalis videri.
(‘The conveyance of property by Lucius Sulla and Gaius Caesar from its rightful
owners to the hands of strangers should, for that reason, not be regarded as
generosity.’ Cic. Off. 1.43)
138
A very complete list can be found in Bennett (1914: II.50–6).
139
See also Spevak (2010a: 271).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(n) (sc. sacerdotes) . . . qui quos ad mortem devovissent, eorum deditionem vi-
vorum hosti fecissent . . .
(‘ . . . priests, who had delivered alive to the enemy those whom they had consecrated
to death . . . ’ Liv. 31.18.6)
(o) Respondebo igitur superiori prius, in qua scribis ad me de absolutione
maiestatis.
(‘I shall therefore reply first to the former, in which you write of your acquittal on the
charge of lèse-majesté.’ Cic. Fam. 3.11.1)
Instances of nouns with more than one argument are relatively rare, because in terms
of information distribution they tend to be too heavy and difficult to process.140 At
the same time, they have the advantage of presenting information in a concentrated
way and are therefore attractive to authors of didactic and technical texts.141
The normal case form of noun phrases that fill an argument position with a noun is
the genitive, the case form for adnominal relations in general (see § 12.14).142 This is
the case both when the corresponding verb governs an accusative object and when it
governs a non-accusative object. Examples of the former type can be found through-
out this section. An example of a genitive that corresponds to an ablative in the clausal
construction is (p).
(p) . . . virtus in usu sui tota posita est. Usus autem eius est maximus civitatis
gubernatio . . .
(‘ . . . the existence of virtue depends entirely upon its use; and its noblest use is the
government of the State . . . ’ Cic. Rep. 1.2)
There are two exceptions to this general rule, viz. (i) instances where the adnominal
argument is in the same case (or has the same preposition) as it would be in a clausal
construction and (ii) instances where prepositions are used that cannot be used in the
same way at the clause level.
Examples of the first type of exceptions are (q)–(s). In (q), nos is in the accusative,
as is the object with the verb tango. In (r), alteri is in the dative, as is the object with
the verb obloquor. One might say that in the combination of the verb sum and the
noun the latter determines the case form (some sort of ‘support verb’ construc-
tion).143 Instances like (q) are almost limited to Plautus.144 Ex. (s) shows a source
argument in the ablative with discessu.
(q) Sed quid tibi nos tactio est, mendice homo?
(‘But why did you touch us, beggar?’ Pl. Aul. 423)
140
Quantitative data about adnominal arguments can be found in Bolkestein (1989: 17).
141
For Vitruvius, see Wistrand (1933: 67ff.).
142
On the genitive with verbal nouns, see Torrego (1991).
143
See Rosén (1981: 152ff.).
144
For the material (to be checked in recent editions), see Landgraf (1898: 399–402).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Good examples of the second type of exceptions are the prepositions erga and in +
acc. ‘towards’, ‘for’ to mark patients at the noun phrase level, as in (t) and (u),
respectively. These correspond to the use of the accusative at the clause level; with
the verb amo, the object cannot be marked by erga or in. In (t), erga distinguishes the
patient constituent from the agent constituent (which is in the genitive), and thereby
avoids any potential confusion: both constituents are animate and might have fulfilled
either role. This is less likely to have been a factor in the case of (u). Further details can
be found in § 11.72.145
(t) . . . urbem hodierno die deorum immortalium summo erga vos amore . . .
restitutam videtis.
(‘ . . . the city has, as you see, been restored . . . on this very day, by the great love that
the immortal gods hold for you . . . ’ Cic. Catil. 3.1)
(u) . . . postea quam L. Flacci amor in patriam perspectus esset . . .
(‘ . . . after Lucius Flaccus’ devotion to his fatherland had been proved . . . ’ Cic. Flac. 2)
Supplement:
Arguments in the genitive (in alphabetical order by governing noun):
Corresponding to a first argument: Sperabit sumptum sibi senex levatum esse
harunc abitu. (Ter. Hau. 746); Nimis paene inepta atque odiosa eius amatio est.
(Pl. Rud. 1204); Placide egredere et sonitum prohibe forium et crepitum cardinum . . .
(Pl. Cur. 158); Probe horum facta inprudens depinxit senex. (Ter. Ph. 268); . . . res in
Graecia tranquillas . . . profectio Attali fecerat . . . (Liv. 28.8.14); . . . si aberis ab eri quaes-
tione. (Pl. Trin. 1012);
Corresponding to a second argument: (event nouns): Et principium orationis ab
accusatione stultitiae orsus suae . . . (Liv. 44.31.13); . . . neque in hac subigitationes
sunt neque ulla amatio / nec pueri suppositio nec argenti circumductio . . . (Pl. Capt.
1030-1); Iam igitur amota ei erit omnis consultatio / nuptiarum . . . (Pl. Epid. 282–3);
Catonem veteres inimicitiae Caesaris incitant et dolor repulsae. (Caes. Civ. 1.4.1);
Quod ad rem publicam pertinet, omnino multis diebus exspectatione Galliarum
actum nihil est. (Cael. Fam. 8.8.4); Erant autem magnae partis hominum ad favorem
eius inclinati animi . . . (Liv. 42.5.2); . . . qui . . . ex praetoris Cn. Fulvi exercitu ob
similis iram fugae missi eo ab senatu fuerant. (Liv. 27.7.13); Numquam illum
aspectum dicebat quin mihi Telamo iratus furere luctu fili videretur. (Cic. de Orat.
2.193); . . . quam multos divini supplicii metus a scelere revocarit. (Cic. Leg. 2.16);
145
For prepositional phrases that are equivalent to objective genitives, see Torrego (1991: 286–91).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Se tamen ante illud tempus eam rem numquam in medium propter periculi metum
protulisse. (Cic. Fam. 15.2.6); O salutis meae spes. (Pl. Rud. 680);
(actor nouns): Verum si augeam / aut etiam adiutor si<e>m eius iracundiae, /
insaniam profecto cum illo. (Ter. Ad. 145–7); Bonorum extortor, legum contortor!
(Ter. Ph. 374); . . . ut meus victor vir belli clueat. (Pl. Am. 647);
Corresponding to a third argument: Quapropter neminem dubitare existimo quin
illa damnatio ambitus nulla ex parte cum causa Cluenti vestroque iudicio coniuncta
esse possit. (Cic. Quinct. 98); Possunt in pactionibus faciendis non legem spectare
censoriam sed potius commoditatem conficiendi negoti et liberationem molestiae.
(Cic. Q. fr. 1.1.35);
Arguments in the dative: (event nouns): Quid mihi scelesto tibi erat auscultatio . . .
(Pl. Rud. 502); Quodsi iustitia est obtemperatio scriptis legibus institutisque
populorum . . . (Cic. Leg. 1.42); Quid Pompei statuae plausus infiniti? Quid duobus
tribunis plebis qui vobis adversantur? (Cic. Phil. 1.36); Quid tibi aucupatio’st
argumentum aut de amore verbificatio’st patri? (Caecil. com. 62);
(with trivalent nouns): Stylobatisque adiectio quae fit per scamillos inpares, ex
descriptione quae supra scripta est in libro tertio sumatur. (Vitr. 5.9.4); . . . ne qua
exprobratio cuiquam veteris fortunae discordiam inter ordines sereret. (Liv.
23.35.7); . . . sibi ipsi responsio . . . (Cic. de Orat. 3.207); Is eo tempore erat Ravennae
exspectabatque suis lenissimis postulatis responsa . . . (Caes. Civ. 1.5.5);
(actor nouns): Eam vero inspirat et docet verus Deus, dator vitae aeternae veris
cultoribus suis. (August. Civ. 6.4.1);
Arguments in the accusative: (event nouns): . . . quae misera in exspectatione est
Epignomi adventum viri. (Pl. St. 283); . . . QUIS·VOLET·PRO·IOUDICATOD·NI (N. L cj.
Mommsen) / MANUM INIECT<I>O·ESTOD. (CIL I2.401.5–6 (Luceria, Rep.)); Quid tibi
hunc receptio ad te est meum virum? (Pl. As. 920); Quid tibi hanc digito tactio est?
(Pl. Poen. 1308);
(actor nouns): . . . a me consilium petis quid sim tibi auctor, in Siciliane subsidas
an ut ad reliquias Asiaticae negotiationis proficiscare. (Cic. Fam. 6.8.2); Denique
caverat pristinum doctor de emendatione sua et manet chirographum apud psychi-
cos, apud quos tunc gesta res est. (Tert. Prax. 1.6);146
Arguments with verbal nouns denoting movement: Quid tibi hanc aditio est? /
Quid tibi hanc notio est, inquam, amicam meam? (Pl. Truc. 622–3); Et utinam
subicere vestris oculis, Achaei, possem concursationem regis magni ab Demetriade
nunc Lamiam in concilium Aetolorum, nunc Chalcidem. (Liv. 35.49.9); Tantam
spem attulerat exploratae victoriae tua praeclara Mutina eruptio . . . (Cic. Fam.
11.14.1); . . . ad otium convertor et, quemadmodum pecoribus fatigatis quoque, velo-
cior domum gradus est. (Sen. Dial. 9.1.11); Nocturnus introitus Zmyrnam quasi in
hostium urbem . . . (Cic. Phil. 11.5); Inde iter Alexandriam contra senatus auctorita-
tem, contra rem publicam et religiones. (Cic. Phil. 2.48); Qui vero inde reditus
Romam, quae perturbatio totius urbis! (Cic. Phil. 2.108); Provehimur pelago vicina
Ceraunia iuxta, / unde iter Italiam cursusque brevissimus undis. (Verg. A.
3.506–7); . . . sed Rhodiorum classi propinquum reditum ac domum itionem dari.
146
For arguments with nouns in -tor, see Fruyt (1990a: 64). For Late Latin instances of accusative
arguments with these nouns, see Sz.: 34, with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(Cic. Div. 1.68); . . . domum reditionis spe sublata . . . (Caes. Gal. 1.5.3); Nam illa
cottidie tua Lorium ventio . . . (Aur. Fro. Aur. 2.19.1 vdH2);
Arguments in the ablative: Atque etiam scriptum a multis est, cum terrae motus
factus esset, ut sue plena procuratio fieret vocem ab aede Iunonis ex arce extitisse.
(Cic. Div. 1.101);
More adnominal arguments with one noun: Tanta tamen universae Galliae con-
sensio fuit libertatis vindicandae et pristinae belli laudis recuperandae, ut . . . (Caes.
Gal. 7.76.2); . . . tanta fuit omnium exspectatio visendi Alcibiadis, ut . . . (Nep. Alc. 6.1);
Intellegere sese, tametsi pro veteribus Helvetiorum iniuriis populi Romani ab his
poenas bello repetisset, tamen . . . (Caes. Gal. 1.30.2); Est in eadem urbe et ferreus
Hercules, quem fecit Alcon laborum dei patientia inductus. (Plin. Nat. 34.141);
Appendix: If the adnominal (second) argument is a pronoun, the genitives mei, tui,
and sui are used, as in (v). For the plural first and second person there are two
options: nostri, vestri and—rarely—nostrum, vestrum. Examples are (w) and (x),
respectively.147
(v) Mirum me desiderium tenet urbis, incredibile meorum atque in primis tui,
satietas autem provinciae . . .
(‘I have a marvellous longing for Rome, and miss my family and friends, you
especially, more than you would believe. I am sick and tired of the province . . . ’
Cic. Fam. 2.11.1)
(w) Grata mihi vehementer est memoria nostri tua quam significasti litteris.
(‘I am truly gratified by your remembrance of me, which you signify in your
letter.’ Cic. Fam. 12.17.1)
(x) Qua re noli me ad contentionem vestrum vocare, Laterensis.
(‘Do not, therefore, Laterensis, challenge me to institute a comparison
between you.’ Cic. Planc. 16)
Ex. (a) has a directional prepositional phrase that functions as an argument of the
verbal noun accessio. Note the presence of the adverb prope as well. Ex. (b) shows two
prepositional phrases modifying the trivalent noun translatio. The agent and patient
are expressed as well (in the genitive). (For exceptional agent expressions with ab,
see § 5.8.)
(a) Quid tibi ad hasce accessio aedis est prope aut pultatio?
(‘What’s the meaning of your coming near this house or knocking on it?’
Pl. Truc. 258)
147
The latter forms are normal in combination with omnium, for example: omnium nostrum vitam
(Cic. Catil. 1.14).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Possessive adjectives are used as adnominal arguments from Early Latin onwards. In
(a) and (b) meo and meae respectively correspond to the first argument of a one-place
verb at the clause level. In (c) and (d), by contrast, the possessive adjectives corres-
pond to a second argument.148 Meo in ex. (e) is usually taken as corresponding to a
148
For Cicero, see Lebreton (1901a: 98–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
second argument (see the translation), but since Jupiter is speaking these words
meo may correspond to the first argument as well: ‘through my intimidation’.149
For the second argument the use of the genitive of the personal pronoun is more
common, as in (f ).
(a) Maxumam praedam et triumphum is adfero adventu meo.
(‘With my coming I’m bringing them the greatest booty and triumph.’ Pl. As. 269)
(b) Exponam vobis breviter consilium et profectionis et reversionis meae.
(‘I shall briefly explain to you my rationale both for setting out on my journey and for
turning back.’ Cic. Phil. 1.1)
(c) Eo, etsi scio pol is fore meum conspectum invisum hodie.
(‘I’ll go, though I know they’re going to hate the sight of me, by heaven.’ Ter. Hec. 788)
(d) Nunc ego illam me velim / convenire, postquam inanis sum, contemptricem
meam.
(‘Now I’d like her to meet me, now that I have nothing, that woman who despises
me.’ Pl. Bac. 530–1)
(e) ‘Exsurgite’ inquit ‘qui terrore meo occidistis prae metu.’
(‘ “Rise,” he said, “you who have fallen down in terror of me, out of fear.” ’ Pl. Am.
1066)
(f ) Quintus misit filium non solum sui deprecatorem sed etiam accusatorem mei.
(‘Quintus has sent his son not only to make his own peace but to accuse me as well.’
Cic. Att. 11.8.2)
Supplement:
Nam neque neglegentia tua neque odio id fecit tuo. (Ter. Ph. 1016); Ipsae (sc. leges)
enim te a cognitione sua iudicio publico reppulerunt. (Cic. Balb. 33); . . . cum ea quae
faciebat tua se fiducia facere dicebat . . . (Cic. Ver. 5.176); . . . L. Papirius Paetus, vir
bonus amatorque noster, mihi libros eos quos Ser. Claudius reliquit donavit. (Cic.
Att. 1.20.7); Et cum ferae bestiae cibum ad fraudem suam positum plerumque
aspernentur . . . (Liv. 41.23.8); Caesar flexo in maestitiam ore suam invidiam tali
morte quaesitam apud senatum . . . (Tac. Ann. 3.16.2);
When two arguments are expressed with the same head, one of them being a
possessive adjective, the other a genitive, the possessive adjective corresponds to the
first argument, as in (g) and (h).150
(g) Nicias te, ut debet, amat vehementerque tua sui memoria delectatur.
(‘Nicias loves you as he ought, and is delighted by your remembrance of him.’ Cic.
Att. 13.1.3)
149
For more instances of meus, see TLL s.v. 921.10ff.
150
For disambiguating factors for the interpretation of adnominal arguments, see Bertocchi and
Maraldi (1990: 76–80).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(h) Quid ergo illa tua tum obtestatio tibicinis, quid foculus . . . valuerunt?
(‘What then was the significance of your fervent appeal to the flute-player, of your
brazier . . . ?’ Cic. Dom. 125—see Nisbet ad loc.)
Supplement:
Noun phrases with more than one (sometimes hierarchically differing) attribute,
at least one of which is prepositional: Cedo tris mi homines aurichalco contra cum istis
151
For the hierarchical structure of the noun phrase, see Lavency (1991).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
moribus. (Pl. Mil. 658); . . . melium, id est cingulum circum collum ex corio firmo cum
clavulis capitatis quae . . . (Var. R. 2.9.15); Studium est autem animi assidua et
vehementer ad aliquam rem adplicata magna cum voluptate occupatio. (Cic. Inv.
1.36); Extrema est . . . assumptio iudicationis et de ea per amplificationem ex deliber-
ationis praeceptis dictio. (Cic. Inv. 2.82); Illud unum derectum iter ad laudem cum
labore qui probaverunt prope soli iam in scholis sunt relicti. (Cic. Cael. 41); . . .
derideatur de rege Pyrrho triumphus M.' Curi, de Philippo T. Flaminini . . . (Cic.
Mur. 31); Alter vero, non ille Serranus ab aratro, . . . nomen suum de tabula sustulit.
(Cic. Sest. 72); Iucundus enim Balbo nostro sermo tuus contra Epicurum fuit. (Cic.
N.D. 3.2); . . . Oileus ille apud Sophoclem . . . (Cic. Tusc. 3.71); Est igitur Zenonis haec
definitio ut perturbatio sit, quod Ł ille dicit, aversa a recta ratione contra
naturam animi commotio. (Cic. Tusc. 4.11); Et quoniam meam tuorum erga me
meritorum memoriam nulla umquam delebit oblivio . . . (Cic. Fam. 2.1.2); Omnium
temporum iniurias inimicorum in se commemorat . . . (Caes. Civ. 1.7.1); . . . Uticenses
pro quibusdam Caesaris in se beneficiis illi amicissimi . . . (Caes. Civ. 2.36.1); . . .
ancipitem terrorem intra extraque munitiones . . . (Caes. Civ. 3.72.2); . . . horribilis
de saxo iactus deorsum . . . (Lucr. 3.1016); Huic tam pacatae profectioni ab urbe
regis Etrusci . . . (Liv. 2.14.1); Et huic proelium cum Tuscis ad Ianiculum erat crimini.
(Liv. 2.52.7); Inter has iras plebis in patres cum tribuni plebi nunc illud tempus esse
dicerent . . . (Liv. 5.12.8); Cato censorius in illa nobili trium sapientiae procerum ab
Athenis legatione audito Carneade quam primum legatos eos censuit dimittendos .
. . (Plin. Nat. 7.112); Lyncas vulgo frequentes et sphingas, fusco pilo, mammis in
pectore geminis Aethiopia generat . . . (Plin. Nat. 8.72); . . . quam<quam> Tiberio
nulla vetus in Arruntium ira. (Tac. Ann. 1.13.1); Denis in diem assibus animam et
corpus aestimari. (Tac. Ann. 1.17.4); . . . initio veris et repentino in Chattos excursu
praecepit. (Tac. Ann. 1.55.1); Sed et adversus convicia malosque rumores et famosa
de se ac suis carmina firmus ac patiens . . . (Suet. Tib. 28.1); . . . eandem habent
causam eandemque utique mussitationem Pharisaeorum erga commercium domini
et ethnicorum. (Tert. Pud. 9.4); . . . nemo mundus a peccato coram te, nec infans
cuius est unius diei vita super terram? (August. Conf. 1.7.11); . . . longinquioris apud
Carthaginem peregrinationis sumptus praeparabantur . . . (August. Conf. 2.3.5);
152
Most examples are taken from the excellent TLL article on medius.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
PARTITIVE use. As (a) shows, with medius used in its non-partitive meaning, the entity
in the middle of which the head noun is situated can be expressed either in the
genitive, as here, or with a prepositional phrase with inter, as in (d). A noun phrase
containing non-partitive medius can be modified by a descriptive adjective like
magnus, as in (e). A partitive adjective can modify a noun phrase containing another
modifier, as in (f ), where summo modifies cornu dextro. Note in (g) partitive media
modifying summa arce, where summa is partitive as well.153
(a) Nam terra nona inmobilis manens una sede semper haeret complexa
medium mundi locum.
(‘For the earthly sphere, the ninth, remains ever motionless and stationary in its
position in the centre of the universe.’ Cic. Rep. 6.18)
(b) Ad caput amnis (sc. ii) quod de caelo exoritur sub solio Iovis. / # Sub solio
Iovis? # Ita dico. # E caelo? # Atque <e> medio quidem.
(‘To the source of the river which rises from heaven under Jupiter’s throne. # Under
Jupiter’s throne? # That’s what I’m saying. # From heaven? # And right from its
centre at that.’ Pl. Trin. 940–1)
(c) Item figura verbi qualis sit refert, quod in figura vocis alias commutatio fit in
primo verbo sŭit modo sūit, alias in medio, ut curso cursito, alias in extremo,
ut doceo docui . . .
(‘Likewise the character of the form of a word is important, because in the form of the
spoken word a change is sometimes made in the first part of the word, as in sŭit
“sews” and sūit “sewed”; sometimes in the middle, as in curso “I run to and fro” and
cursito, of the same meaning; sometimes at the end, as in doceo “I teach” and docui “I
have taught.” ’ Var. L. 10.25)
(d) Cum venit in medium spatium inter brumam et solstitium, quod dies aequus
fit ac nox, aequinoctium dictum.
(‘When the sun has arrived midway between the bruma and the solstitium, it is called
the aequinoctium “equinox”, because the day becomes aequus “equal” to the nox
“night.” ’ Var. L. 6.8)
(e) . . . fossas duas derectas non ita magno medio intervallo relicto IIII pedum
altitudinis in eum locum deduxit . . .
(‘ . . . he carried two straight trenches, four feet deep and spaced not so very far apart,
as far as the point . . . ’ B. Alex. 38.3)
(f ) In cornu autem summo dextro Veneris et Mercurii fanum (sc. est) . . .
(‘On the right wing at the top is a temple of Venus and Mercury . . . ’ Vitr. 2.8.11)
(g) In summa arce media Martis fanum (sc. est) . . .
(‘In the middle of the top of the citadel is a temple of Mars . . . ’ Vitr. 2.8.11)
153
For a few more examples and discussion, see Romero (1996), Hoffmann (1997b: 184–6), and
Lehmann (1998).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Non-partitive medius can function as subject or object complement, as in (h) and (i),
and as secondary predicate, as in (j) where medium is related to the relative pronoun
quem (for this criterion, see Chapter 21). Partitive medius cannot function as subject
or object complement, but it is not uncommon as secondary predicate, as in (k),
where medium is also related to the relative pronoun quem.
(h) . . . rotundum ut caelum terraque ut media sit eaque sua vi nutuque teneatur,
sol ut eam circumferatur . . .
(‘ . . . that the sky is a round vault, with the earth at its centre, held stationary by its
own force and stress; and the sun travels round it . . . ’ Cic. de Orat. 3.178)
(i) Hi certo anni tempore in finibus Carnutum, quae regio totius Galliae media
habetur, considunt . . .
(‘These druids, at a certain time of the year, meet within the borders of the Carnutes,
whose territory is reckoned as the centre of all Gaul . . . ’ Caes. Gal. 6.13.10)
(j) Solio tum Iuppiter aureo / surgit, caelicolae medium quem ad limina ducunt.
(‘Then from his golden throne Jupiter rises, and the celestial company gather round
and escort him to the threshold.’ Verg. A. 10.116–17)
(k) Lucus erat quem medium ex opaco specu fons perenni rigabat aqua.
(‘There was a grove watered by a perennial spring which flowed through the midst of
it, out of a dark cave.’ Liv. 1.21.3)
The distributional differences mentioned demonstrate that these words have two
different meanings. Of course, this does not mean that no cases of ambiguity exist.
Word order does not offer a solution: both in their non-partitive and in their
partitive use, these adjectives predominantly precede the noun.
In the Supplement examples can be found of those adjectives that are regularly used
with a partitive meaning (including a few instances of secondary predicates).154
Supplement (in alphabetical order by adjective):
. . . ut in extremo tertio (sc. libro) scribis . . . (‘ . . . at the end of the third book . . . ’ Cic.
Div. 1.9); . . . litterae mihi dantur a te, quibus in extremis scriptum erat . . . (Cic. Fam.
7.5.2—NB: secondary pred.); Saepe illam perhibent ardenti corde furentem / cla-
risonas imo fudisse e pectore voces . . . (Catul. 64.124–5); . . . uti schema subscriptum
est in ima pagina . . . (Vitr. 9.pr.5); In foro infimo boni homines atque dites ambulant,
/ in medio propter canalem, ibi ostentatores meri. (Pl. Cur. 475-6); Quod inferiore
epistula scribis me Id. Sept. Pompeio legatum iri . . . (Cic. Q. fr. 3.1.18); Traxit ex
intimo ventre suspiritum. (Pl. Truc. 600); Ibi ego nescio quo modo / iratus videor
mediam arripere simiam. (‘to grab the monkey around the middle’ Pl. Rud. 608–9—
NB: secondary pred.); . . . platea ampla latitudine facta, in qua media Mausoleum . . .
est factum . . . (Vitr. 2.8.11—NB: secondary pred.); Vix agmen novissimum extra
munitiones processerat . . . (Caes. Gal. 6.8.1); More maiorum date plausum postrema
in comoedia. (Pl. Cist. 787); Atque etiam modo vorsabatur mi in labris primoribus.
154
The best study of partitive adjectives is Vaughan (1942). For frequency data, see his ch. V.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(Pl. Trin. 910); Prata primo vere stercerato luna silenti. (Cato Agr. 50.1); (sc. spatula)
In qua summa capitulum est rotundum . . . (Cels. 8.15.2—NB: secondary pred.); Sed
hic quis est senex quem video in ultima platea? (Ter. Ph. 215);
Appendix: The partitive use of adjectives is often described as a reversal of the role of
the adjective and the noun within the noun phrase, thus as some form of ‘upgrading’
of the adjective (see also § 11.13).155 A paraphrase of e.g. summus mons in its
partitive meaning ‘top of the mountain’ by summum montis is indeed possible.
However, ‘upgrading’ cannot explain the use of these adjectives as secondary predi-
cates. Further differences between the non-partitive and partitive expressions are that
in the former the adjective can both more easily be left out without making the
remaining expression ungrammatical and it can be replaced by a relative clause.156
There are also other expressions that are sometimes called ‘partitive’: for example, the
use of the adjective of amount multus illustrated by (l).
Two further words that are usually mentioned in this context are the identifiers
ceterus and reliquus, but this may be a side effect of their meaning. Two examples are
(m) and (n).
155
So already K.-St.: I.233 (‘Attributive Adjektive an der Stelle von Substantiven’).
156
For disambiguating tests, see Amacker (1986).
157
For the history of the terminology and discussion of many examples, see Hey (1906). See also
Maurach (1983: 96–8). For references to the abundant litterature, see Sz.: 159–60. For a discussion in a
broader context, see Burkhardt (1971).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In (a), Phrygio ‘Trojan’ is obviously a property of Alexandri and not of the head noun
pectore. In (b), angusti fits better with the head noun claustra. In poetry, Greek
influence and metrical considerations certainly played a role; the whole phenomenon
has to be regarded as stylistic. Examples from Classical prose cited in the literature are
(c) and (d), but they are not very convincing. More convincing is (e). Sometimes an
adjective can semantically belong to either noun. There is no need to regard (f ) as an
example of transference of the adjective.158
(a) . . . ignis Alexandri Phrygio sub pectore gliscens . . . (< Alexandri Phrygii sub
pectore)
(‘ . . . fire . . . glowing beneath the breast of Phrygian Alexander . . . ’ Lucr. 1.474)
(b) . . . angusti rarescent claustra Pelori . . . (< angusta claustra Pelori)
(‘ . . . the barriers of narrow Pelorus will open out . . . ’ Verg. A. 3.411)
(c) . . . cautum est in Scipionis legibus ne plures essent in senatu ex colonorum
numero quam ex vetere Agrigentinorum.
(‘ . . . the laws of Scipio provided that the number of settlers in the Senate should not
exceed that of the original inhabitants of Agrigentum.’ Cic. Ver. 2.123)
(d) Qui omnes fere iis nominibus civitatum appellantur quibus orti ex civitati-
bus eo pervenerunt.
(‘Nearly all of these latter are called after the names of the states from which they
sprang when they went to Britain.’ Caes. Gal. 5.12.2)
(e) . . . ad maiora (maiorum cj. Novak) rerum initia ducentibus fatis . . .
(‘ . . . with fate guiding him to greater beginnings of things . . . ’ Liv. 1.1.4)
(f ) . . . nec adiri usquam ad iusti cursum poterat amnis . . .
(‘ . . . afforded nowhere any access to the regular channel of the river . . . ’ Liv. 1.4.4)
Supplement:
Poetry: Videte, cuncti, tuque, caelestum sator, / iace, obsecro, in me vim coruscam
fulminis! (Cic. Tusc. 2.21—NB: tr. of Soph. Trach.); Quid Nemeaeus enim nobis nunc
magnus hiatus / ille leonis obesset . . . (Lucr. 5.24–5);159 Daphni, quid antiquos sig-
norum suspicis ortus? (Verg. Ecl. 9.46); . . . Tyrrhenusque tubae mugire per aethera
clangor. (Verg. A. 8.526); Quodsi dolentem nec Phrygius lapis / nec purpurarum
sidere (Sidone cj. Nisbet) clarior / delenit usus . . . (Hor. Carm. 3.1.41–3); Luctus erat
falsaeque patres in crimine caedis . . . (Ov. F. 2.497); Accipe supremo dictum mihi
forsitan ore, / quod tibi qui mittit non habet ipse, ‘vale’. (Ov. Tr. 3.3.87–8); . . . MULTIS
GENERUM POMIS . . . (CLE 492.24 (Nicopolis, first half III AD—NB: Cugusi and Sblen-
dorio Cugusi 2008: 68 now read multigenerum)); Tum iubar Arcturi languet, tum
fulva Leonis / ira perit . . . (Claud. 1.25–6);
158
For more examples of this kind, see K.-St.: I.221.
159
For Lucretius, see Bailey (1949: I.144).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
11.79 Apposition
160
For more instances in Vitruvius, see Wistrand (1933: 79–81).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Thus, in (a), erum and Amphitruonem are both governed by the preposition ad; either
of them may be omitted (ad erum and ad Amphitruonem are both grammatically
correct); they both refer to the same person. A similar argument can be made for
servom . . . Sosiam. Similarly, in (b), Syncerastum and lenonis servom are both subject
of recipere se; either of them may be omitted; they refer to the same person.
(a) Iam ille illuc ad erum quom Amphitruonem advenerit, / narrabit servom
hinc sese a foribus Sosiam / amovisse.
(‘In a moment, when that chap gets there and meets his master Amphitruo, he’ll say
the slave Sosia drove him away from the door.’ Pl. Am. 466–8)
(b) Attat, e fano recipere video se Syncerastum, / lenonis servom.
(‘Goodness! I can see Syncerastus, the pimp’s slave, returning from the shrine.’ Pl.
Poen. 821–2)
Although the appositive phrases in both examples consist of a proper name and a
common noun (phrase), the relation between these two elements is different. In (a),
erus indicates the social category to which Amphitruo belongs; at the same time,
Amphitruo specifies which master is being referred to. In (b), lenonis servom identifies
Syncerastus as ‘the slave of the pimp’, answering an underlying question ‘who is
Syncerastus?’. In another context it might indicate what kind of person Syncerastus is
(he is ‘a slave of the pimp’ or ‘a slave of a pimp’—the noun phrase would be
indefinite). In the first interpretation of (b) the two constituents do refer to the
same person, in the second lenonis servom rather describes Syncerastus and does
not properly refer to a person. Whereas in (a) the two constituents relate to each other
in terms of ‘more general’ / ‘more particular’, this is not the case in (b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
The first type of apposition is often called RESTRICTIVE APPOSITION (also called
‘attributive’ apposition); the second type, NON-RESTRICTIVE APPOSITION, is also called
‘explicative’ apposition. The two types of apposition differ in a number of respects, the
most conspicuous of which are indicated in Table 11.6 (for further details, see the
following sections).161 Since the two types have different properties, they can occur
together with the same head, as in (c): praetorem is restrictive, fortem virum is non-
restrictive, and the NP as a whole is further modified by a relative clause.
(c) . . . C. Sulpicium praetorem, fortem virum, misi qui ex aedibus Cethegi si
quid telorum esset efferret.
(‘ . . . I sent the praetor Gaius Sulpicius, a gallant man, to get from the house of
Cethegus any weapons that were there.’ Cic. Catil. 3.8)
In spoken modern languages the two types of nominal apposition differ in inton-
ational contour. In modern written texts, punctuation sometimes fulfils the same
function. The traditions in this respect differ from one language community to
another, which is reflected in the way Latin texts are edited: editors tend to follow
the conventions for punctuation in the particular modern language they are working
in and insert such punctuation according to their interpretation of the Latin text.
Comparison of translations of Latin texts, in the same or in different languages,
shows that these interpretations vary considerably.
One complicating factor is word order: the orders servos Sosia and Sosia servos are
both allowed (the former is more frequent) and discontinuity is allowed as well (at
least in comedy—see § 11.81), as in (d).
(d) Ego servi sumpsi Sosiae mi imaginem / qui cum Amphitruone abivit . . .
(‘I’ve taken on the slave Sosia’s image, who went away to the army with
Amphitruo . . . ’ Pl. Am. 124–5)
However, instead of taking servi Sosiae as an instance of restrictive apposition one
might prefer a different interpretation: in Mercury’s preceding words no mention has
been made of a slave nor of a slave called Sosia. Instead of as ‘the slave Sosia’ his words
might be interpreted as: ‘I’ve taken on a slave’s, (to be precise) Sosia’s, image . . . ’.
161
For these discriminating characteristics, see Fugier (1973: 99).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
162
This is how the term is used by Bennett (1914: II.5–7). There is an abundant literature on the
definition and classification of appositions. For Latin, Fugier (1973), Longrée (1990, 2007), Lavency
(1991), Touratier (1994: 441–6), Heberlein (1996), Rosén (2002a), van Laer (2007), and Spevak (2014a:
ch. 4, with further references), deserve special mention. For English (and the term ‘appositive’), see Quirk
et al. (1985: 1300–21).
163
Bauer (2008: 44) claims that there is a change from Roma urbs to urbs Roma, but this seems difficult
to prove, since the sequence Roma urbs is only once attested on PHI (Vell. 1.8.4) and once in Eutropius
(9.3—Bal, p.c.), the other sequence more than 150 times.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
semantic or pragmatic difference between minimal pairs like (d) and (e). For a plural
general lexeme, see (f ).
(a) Ait se ob asinos ferre argentum atriensi Saureae / . . .
(‘He said he was bringing money as payment for donkeys to the steward Saurea . . . ’
Pl. As. 347)
(b) Hoc D. Silanus consul in hoc ordine, hoc meus etiam collega dicebat.
(‘Such was the language used in this house by the consul Decimus Silanus, and even
by my own colleague.’ Cic. Pis. 56)
(c) Allienus autem praetor putabat aliquem, si ego non, ex collegis suis.
(‘Praetor Allienus thinks that if I don’t, one of his colleagues will.’ Cic. Att. 10.15.3)
(d) Nimis velim / certum qui id mihi faciat, Ballio leno ubi hic habitat.
(‘I’d very much like someone to tell me for certain whereabouts here the pimp Ballio
lives.’ Pl. Ps. 598–9)
(e) Si tu quidem es / leno Ballio.
(‘If indeed you are pimp Ballio.’ Pl. Ps. 1154–5)
(f ) Cum L. Octavius C. Aurelius consules aedis sacras locavissent . . .
(‘When the consuls Lucius Octavius and Gaius Aurelius had made contracts for
temple maintenance . . . ’ Cic. Ver. 1.130)
The appositive phrase can be modified by one or more determiners, as in (g) and (h),
or by a possessive adjective, as in (i). In (j), the appositive phrase contains a
(possessive) genitive. Discontinuity of the possessive adjective and the noun, as in
(k), is normal.164 It is not clear whether with a noun phrase in the (possessive)
genitive an appositive phrase can be interpreted as restrictive as well, as in (k):
Lampadisci is best regarded as a restrictive apposition.
(g) . . . nec praesente nobis alius quisquam est servos Sosia.
(‘ . . . and when the two of us are present, there’s no other slave Sosia.’ Pl. Am. 400)
(h) Sed lenone istoc Lyco, / illius domino, non lutum est lutulentius.
(‘But dirt isn’t dirtier than that pimp Lycus, her master.’ Pl. Poen. 157–8)
(i) Noster item aequalis D. Silanus, vitricus tuus, studi ille quidem habuit non
multum . . .
(‘Another contemporary, Decimus Silanus, your stepfather, Brutus, lacked in
application . . . ’ Cic. Brut. 240)
(j) Sed Amphitruonis illi[c] est servos Sosia.
(‘But there is Amphitruo’s slave Sosia.’ Pl. Am. 148)
164
For parallels, see Brix ad Pl. Capt. 875.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(k) Audire vocem visa sum ante aedis modo / mei Lampadisci servi.
(‘Just now I seemed to hear the voice of my good slave Lampadio in front of the
house.’ Pl. Cist. 543–4)
Supplement (in alphabetical order by noun):
. . . quaestor C. Mancini consulis cum esset . . . (Cic. Har. 43); Evocate intus Cylin-
drum mihi coquom actutum foras. (Pl. Men. 218); Quid ille dux Leonidas dicit? (Cic.
Tusc. 1.101); Ubi ego te noverim, / qui amicam habes eram meam hanc Erotium? (Pl.
Men. 299–300—NB: or non-restrictive?); Nam me hodie oravit Argyrippus filius / uti
sibi amanti facerem argenti copiam. (Pl. As. 74–5); Sed eccum fratrem Pamphilip-
pum, incedit cum socero suo. (Pl. St. 527); Et suppressit impetum Romanorum
volnus imperatoris Ap. Claudi . . . (Liv. 26.6.5); L. Plancus interim legatus petit a
Caesare uti sibi daret potestatem . . . (B. Afr. 4.1); . . . M. Petreio legato exercitum
permittit. (Sal. Cat. 59.4); Submonuit me Parmeno / ibi servo’ quod ego arripui.
(Ter. Eu. 570-1); Tu cum Alcumena uxore antiquam in gratiam / redi. (Pl. Am.
1141–2); . . . quae misera in exspectatione est Epignomi adventum viri. (Pl. St. 283);
165
Combinations with terra are treated as pleonastic by K.-St.: II.568.
166
For mons, see TLL s.v. 1433.41ff.; for oppidum, see TLL s.v. 758.35ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
altissimo, qui est inter Sequanos et Helvetios, tertia lacu Lemanno et flumine
Rhodano, qui provinciam nostram ab Helvetiis dividit.
(‘ . . . the Helvetii are confined: on one side there is the river Rhine, exceeding broad
and deep, which separates the Helvetian territory from the Germans; on another the
Jura range, exceeding high, lying between the Sequani and the Helvetii; on the third,
the Lake of Geneva and the river Rhone, which separates the Roman Province from
the Helvetii.’ Caes. Gal. 1.2.3)
167
See Spevak (2014a: 278).
168
For a few examples, see Rosén (2002a: 211). Prepositions are sometimes inserted in the text of Pliny
the Elder, e.g. 4.16 Cretam <in> insulam add. Jan.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
What these three types have in common is that a proper name or a common noun is
in some form classified by another (common) noun. Only one classifier of this type is
allowed. Some of these appositive phrases can be modified by a determiner and/or a
possessive adjective and/or a descriptive adjective.171 The phrase as a whole can be
modified by one or more non-restrictive appositives.
169
K.-St.: II. 568 deal with some of them as pleonasms. Sz.: 157–8 discusses them in a section on
adjectival use of nouns.
170
For arbor, see TLL s.v. 426.32ff.; for homo, see TLL s.v. 2885.60ff. and Bennett (1914: II.6); for piscis,
see TLL s.v. 2209.44ff.
171
Further research is necessary.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
Other types of appositive phrase are those containing the name of a god and a title,
like Mars pater, as in (v), or containing a noun that denotes an animal and a gender
specification, as in (w). Different again are combinations like nos duo in (x).
172
The text has long been disputed, but see now Dyck ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Whereas in the above examples the first constituent is short, often only one word, and
the second relatively long, the reverse order is also quite common.173 In such cases the
more specific constituent follows. Examples are (h)–(l). A second appositive (or
173
For instances from Plautus onwards, see Otto (1912: 22–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
more) may follow that constituent, like nomine in (k), added to avoid confusion with
the common noun lamia ‘witch’. Remarkable is (l).
(h) . . . ut ego oculis rationem capio quam mi ita dixit erus meus, miles . . .
(‘ . . . that’s how I draw the conclusion with my eyes from what my master the soldier
told me . . . ’ Pl. Ps. 596)
(i) Quid de L. Mummio, qui urbem pulcherrimam atque ornatissimam, Co-
rinthum, plenissimam rerum omnium, sustulit . . .
(‘What of Lucius Mummius, who took the beautiful city of Corinth, full of art
treasures of every kind . . . ’ Cic. Ver. 1.55)
(j) Lautissimum oppidum nunc municip[i]um honestissimorum, quondam colo-
norum, Suessam, fortissimorum militum sanguine implevit.
(‘He made Suessa, a very thriving town, once of colonists, now of most respectable
burghers, run with the blood of terribly brave soldiers.’ Cic. Phil. 13.18)
(k) Mulier est Segestana perdives et nobilis, Lamia nomine. Per triennium isti
plena domo telarum stragulam vestem confecit . . .
(‘At Segesta there is a lady of wealth and rank named Lamia, who for three years had
her house full of looms making woven fabrics for him . . . ’ Cic. Ver. 4.59)
(l) Certo haec mulier aut insana aut ebria est, Messenio, / quae hominem
ignotum compellet me tam familiariter.
(‘This woman is definitely either mad or drunk, Messenio: she addresses me, a total
stranger, so intimately.’ Pl. Men. 373–4)
Supplement:
Aedium dominum, lenonem Ballionem, quaerito. (Pl. Ps. 1140); Sed quid istum
remoramini, / mulieres, Trachalionem? (Pl. Rud. 1208–9); Itaque hoc, iudices, ex
homine clarissimo atque eloquentissimo, L. Crasso, saepe auditum est . . . (Cic. Ver.
3.3); . . . hominem suae civitatis in primis honestum ac nobilem, Andronem Centu-
ripinum, legarunt ad Apronium . . . (Cic. Ver. 3.114); Inter bina castra Pompei atque
Caesaris unum flumen tantum intererat, Apsus . . . (Caes. Civ. 3.19.1); . . . excepit
orationem eius princeps Aetolorum, Alexander, vir ut inter Aetolos facundus. (Liv.
32.33.9); Praeclarum excellentis viri factum C. Sentii Saturnini, circa ea tempora
consulis, ne fraudetur memoria. (Vell. 2.92.1);
174
Taken from Rosén (2002a: 210).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(m) . . . quosdam etiam senatores summo loco natos non in hortis aut suburbanis
suis, sed Neapoli, in celeberrimo oppido, in {maeciapella{ saepe videri . . .
(‘ . . . even some senators of eminent family were often seen wearing ?, not in their country
seats or their suburban villas, but in the populous town of Naples . . . ’ Cic. Rab. Post. 26)
(n) (sc. Archias) . . . primum Antiochiae—nam ibi natus est loco nobili—celebri
quondam urbe et copiosa atque eruditissimis hominibus liberalissimisque
studiis adfluenti, celeriter antecellere omnibus ingeni gloria contigit.
(‘ . . . first at Antioch, where he had been born of gentle parents, a place which in those
days was a renowned and populous city, the seat of brilliant scholarship and artistic
refinement, his intellectual pre-eminence rapidly gained for him a commanding
position among his contemporaries.’ Cic. Arch. 4)
(o) . . . quod cum concursum plorantium ac tempestatem querellarum ferre non
posses, in oppidum devium, Beroam, profugisti?
(‘ . . . that here, unable to endure the crowds who besieged you with their wailing and
bombarded you with their complaints, you took refuge at the town which lay off your
road, Beroa?’ Cic. Pis. 89)
The combinations (i), (ii), and (iv) are regular in Classical Latin, (ii) being more
common in Cicero than (i). In Late Latin combination (ii) is the preferred one.175
175
For more examples, see K.-St.: I.480.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
Supplement:
Type (i): Certe licuit Tusculi se in otio delectare, salubri et propinquo loco. (Cic. Rep.
1.1); . . . Capuam colonia deducetur, urbem amplissimam atque ornatissimam. (Cic.
Agr. 2.76); Nec sum ignarus esse quosdam qui Seleucia, urbe Syriae, accitum (sc.
deum esse perhibeant) . . . (Tac. Hist. 4.84.4);
Type (ii): Albae constiterunt, in urbe opportuna, munita, propinqua, fortissi-
morum virorum, fidelissimorum civium atque optimorum. (Cic. Phil. 4.6); (sc.
Demaratus) . . . fugisse cum magna pecunia dicitur ac se contulisse Tarquinios, in
urbem Etruriae florentissimam. (Cic. Rep. 2.34); . . . antiquitas, quam Tusculo, ex
clarissimo municipio, profectam in monumentis rerum gestarum incisam ac nota-
tam videmus . . . (Cic. Font. 41);
Type (iii): Nobili Graecorum et ampla civitate, Ephesi, lex vetusta dicitur a maioribus
dura condicione sed iure esse non iniquo constituta. (Vitr. 10.pr.1); Ita firmissimum
eius praesidium, Ateguam, proficiscitur. (B. Hisp. 6.1);
Type (iv): . . . teque, Iuno regina, cuius duo fana duabus in insulis posita sociorum,
Melitae et Sami, sanctissima et antiquissima . . . (Cic. Ver. 5.184); . . . mittitur a con-
sule Sextius quaestor in oppidum Iugurthae, Vagam. (Sal. Jug. 29.4);
Type (v): Obiit autem septuagesimo sexto anno morte communi in oppido Campa-
niae, Atella. (Eutr. Brev. 7.8.4);
Type (vi): (sc. non inpluit) . . . in Nea, oppido Troadis, circa simulacrum Minervae.
(Plin. Nat. 2.210); . . . implentur orcae in Asia, cadi autem <in> Ruspina, Africae
urbe . . . (Plin. Nat. 15.82);
The types of non-restrictive appositive phrases dealt with so far contain for their
descriptive component noun phrases (and an occasional adjective phrase). However,
gerundial and gerundival clauses and infinitival clauses can be used as well. Examples
are (p), (q)–(r), and (s), respectively. In (p), an instance of ‘partitive apposition’, the
preposition ad is repeated; in (r), in is not repeated. In (s), there is an accusative and
infinitive clause.176
(p) Hi non viderunt . . . hominem ad duas res, ut ait Aristoteles, ad intellegen-
dum et agendum, esse natum . . .
(‘They failed to see that . . . man, as Aristotle observes, is born for two purposes, for
thinking and acting . . . ’ Cic. Fin. 2.40)
(q) Nisi me credo huic esse natum rei, ferundis miseriis.
(‘Unless I’m to believe I was born for this, to endure misery.’ Ter. Ad. 545)
(r) In quo enim maxime consuevit iactare vestra se oratio . . . claris et fortibus
viris commemorandis eorumque factis . . . laudandis . . .
(‘There is one field in which the eloquence of your school has been wont especially to
vaunt itself . . . I mean the stories of illustrious and heroic men . . . and the applause of
their actions . . . ’ Cic. Fin. 1.36)
176
For more examples of infinitival clauses, see K.-St.: I.665.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(s) Quae istaec audacia est, te sic interdius / cum corolla ebrium incedere?
(‘What impudence is that, that you strut around like this in daytime, with a garland
and drunk?’ Pl. Ps. 1298–9)
Supplement:
Quo modo autem dicatur id est in duobus, in agendo et in eloquendo. (Cic. Orat. 55);
Et erat aeris alieni magna vis re damnosissima etiam divitibus, aedificando, contracta.
(Liv. 6.11.9); Haec ne dicerentur a me, uno modo vitare potuisti, non quaerendo
quam ob causam exercitus in Macedoniam traicerentur . . . (Liv. 42.40.10);
. . . ad alia . . . traducebantur opera, foros in circo faciendos cloacamque maxi-
mam, receptaculum omnium purgamentorum urbis, sub terra agendam. (Liv.
1.56.2); Itaque ad remedium iam diu neque desideratum nec adhibitum, dictatorem
dicendum, civitas confugit. (Liv. 22.8.5);
Nam qui amat quod amat si habet, id habet pro cibo, / videre, amplecti,
osculari, alloqui. (Pl. Mer. 744–5); Gravis etiam illa est et plena dignitatis [dicendi]
facultas, . . . posse consilio atque oratione et senatus et populi . . . mentes permovere.
(Cic. Mur. 24);
177
For the text, see Shackleton Bailey (1959: 201).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
The term DISTRIBUTIVE APPOSITION is used for a number of expressions that differ in
several respects from each other. It has nothing to do with apposition as described
in § 11.80. In one type a plural or collective entity is introduced to be split up into two
or more entities in (partial) parallel clauses, as in (a) with alteram . . . alteram ‘the
one . . . the other’. In this case the term ‘distributive’ has its proper meaning. Another
type is shown by (b). There is a plural entity, referred to explicitly with qui; quisque
‘each’ indicates that the individuals belonging to that plural entity are each involved in
(separate instances of) the state of affairs. This type is dealt with in this Syntax as a
secondary predicate (Chapter 21). A third type is illustrated by (c), implying some
form of reciprocity. Note that types (b) and (c) can co-occur, as in (d). (For agreement
properties of verbs with distributive appositives, see § 13.11.)
(a) Si enim fatum fuit classes populi Romani bello Punico primo, alteram
naufragio, alteram a Poenis depressam, interire . . .
(‘For, if it was the will of Fate that the Roman fleets in the First Punic War should
perish—the one by shipwreck and the other at the hands of the Carthaginians . . . ’
Cic. Div. 2.20)
(b) . . . qui suam quisque rem publicam constituissent legibus atque institutis
suis . . .
(‘ . . . who had each established their commonwealths with their laws and institutions . . . ’
Cic. Rep. 2.2—NB: following the reading of the codex, see Ziegler et al. app.)
(c) Alius alium percontamur: ‘Quoia est navis?’
(‘We ask one another: “Whose ship is it?” ’ Pl. St. 370)
(d) Haec se meritos dicere, exprobrantes suam quisque alius alibi militiam.
(‘These, they cried, were the rewards they had earned, and they bitterly rehearsed to
one another the campaigns they had each served in various places.’ Liv. 2.23.11)
Supplement:
Duplices igitur similitudines esse debent, unae rerum, alterae verborum. (Rhet. Her.
3.33); Tabulae testamenti unae per legatos eius Romam erant adlatae . . . , alterae
eodem exemplo relictae atque obsignatae Alexandriae proferebantur. (Caes. Civ.
3.108.6); Sine signis, sine ordinibus equites peditesque permixti cedere alius, alius
obtruncari, multi contra advorsos acerrume pugnantes ab tergo circumveniri. (Sal.
Jug. 97.5); . . . qui in suo magistratu cum Hannibale, alter ad Ticinum amnem, ambo
aliquanto post ad Trebiam pugnaverint. (Liv. 21.15.4); Consules Marcellus retro unde
venerat Nolam red<i>it, Fabius in Samnites . . . processit. (Liv. 24.20.3); Multitudo
pars procurrit in vias, pars in vestibulis stat, pars ex tectis fenestrisque prospectant et
quid rei sit rogitant. (Liv. 24.21.8—NB: note the change from sg. to pl.);
NB: with an object constituent: Fulvius legionem de qua supra dictum est quinque
cohortes in dextram viae partem direxit, quinque ad sinistram . . . (Fron. Str. 1.6.1);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In Latin first and second person verb forms can be combined with proper names or
NPs. In (a) it is likely that a change of construction occurs (pars spectatorum scit >
scitis vos), but this is not the case in (b)–(d), which look like descriptions in non-
restrictive appositive phrases.178
(a) . . . et pars spectatorum scitis pol haec vos me hau mentiri.
(‘ . . . and plenty of you spectators are well aware that I’m not lying.’ Pl. Truc. 105)
(b) Philosophi sumus exorti . . . qui amori auctoritatem tribueremus.
(‘We philosophers have come forward . . . to attribute authority to love.’ Cic.
Tusc. 4.71)
(c) HEIC·SITUS·SUM·LEMISO / QUEM·NUMQUAM·NISI·MORS / FEINIVIT·LABORE.
(‘Here I Lemiso lie, whom nothing but death ever stopped from my work.’ CIL
VI.6049 (Rome, end Rep.))
(d) Maxima pars vatum, pater et iuvenes patre digni, / decipimur specie recti.
(‘Most of us poets, o father and ye sons worthy of the father, deceive ourselves by the
semblance of truth.’ Hor. Ars 24–5)
Supplement:
Nec nostrum quisquam sensimus, quom peperit, nec providimus. (Pl. Am. 1071);
Themistocles veni ad te . . . (Nep. Them. 9.2); Non iam prima peto Mnestheus neque
vincere certo . . . (Verg. A. 5.194); Hoc tibi iuventus Romana indicimus bellum. (Liv.
2.12.11); . . . duo exercitus Aventinum insedistis. (Liv. 9.34.4); M(ARCUS) LOLLIUS
PHILIPPUS SCRIPSI ME ACCEPISSE MUTUA . . . (TPN 39.5.3 (Pompei, AD 45)); Magna pars
studiorum amoenitates quaerimus. (Plin. Nat. pr. 14); (on a small earthen equipoise)
ANT/IOC/US·FI/XI [= finxi] TE (CIL I .2371 (Mantua)); MAURICIUS V(IR) R(LARISSIMUS)
2
Appositives can be used with various forms of address, proper names, as in (a), nouns
and noun phrases, pronouns, as in (b), or with the subject that the verb ending refers
to, as in (c). More examples can be found in Chapter 22.
(a) Mi Libane, ocellus aureus, donum decusque amoris, / amabo, faciam quod
voles . . .
(‘My dear Libanus, my golden eye, love’s gift and glory, please, I’ll do what you
like . . . ’ Pl. As. 691–2)
178
For more examples, see Sz.: 412 and K.-St.: I.244.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
(b) Tu, urbanus vero scurra, deliciae popli, / rus mihi tu obiectas?
(‘You city loafer, you darling of the streets, you are throwing the farm in my face?’
Pl. Mos. 15–16)
(c) . . . augebis ruri numerum, genus ferratile.
(‘ . . . you’ll increase the population on the farm, as a shackled labourer.’ Pl. Mos. 19)179
11.88 The use of measure, number, and weight nouns in partitive apposition
179 180
The interpretation is disputed. See Adams (2006).
181
See Wistrand (1933: 74–5).
182
See Shackleton Bailey ad Galba [Cic.] Fam. 10.30.1.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) Lepidus ad Forum Voconi castra habet, qui locus a Foro Iuli quattuor et
viginti millia passus abest . . .
(‘Lepidus is encamped at Forum Voconii, twenty-four miles from Forum
Julii . . . ’ Planc. Fam. 10.17.1)
The term ‘apposition’ is also applied to various other types of collocation of two NPs
that seem to have the same function in their context. Ex. (a) looks like a theme
constituent followed by nemo. Ex. (b) looks like a variation on Corinthum oppidum, a
real restrictive apposition. In (c), chorus and formosae . . . heroinae probably have to
be understood as non-restrictive appositives.183
(a) Itaque proavos (NB: nom. sg.) ac superiores de Tremel[l]iis nemo appellatus
Scrofa . . .
(‘Hence neither my great-grandfather nor any of the Tremelii who preceded him was
called by this surname of Scrofa . . . ’ Var. R. 2.4.2)184
(b) ‘quae Corinthum arcem altam habebant matronae opulentae, optimates’
(‘rich and noble dames that dwelt in Corinth, lofty citadel’ Cic. Fam. 7.6.1—NB:
possibly a paraphrase of Ennius)185
(c) Illic formosae veniant chorus heroinae . . .
(‘There, even if in a troupe to greet me come all the lovely heroines . . . ’ Prop. 1.19.13)
183 184
For more instances, see K.-St.: I.250–1. See Flach ad loc.
185 186
See Shackleton Bailey ad loc., with references. K.-St.: I.248.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
However, not all scholars interpret (a) in this way. The OCT and Teubner editors, for
example, have a period after voluptatem and continue: Rem videlicet difficilem et
obscuram! They seem to interpret it as an independent (exclamatory) sentence (with
an exclamation mark). Exclamatory sentences in the accusative are common (see
§ 6.35), but there are certain restrictions. We would have to regard the adjectives as
evaluative (not excluded perhaps for difficilem), but the presence of the evidential
disjunct videlicet makes it difficult to take the utterance as an exclamatory sentence.
A few more examples of such appositive clauses which serve as a comment and are
integrated in the overall linguistic structure are (b)–(e). In (b), the appositive coun-
terpart of rem is more or less ‘making this classification’. Its case is probably
determined by faciunt. Taking it as an independent exclamatory sentence is
unattractive, since it does not contain an evaluative adjective. In (c), repeated from
§ 11.79, rem non difficilem is clearly a comment on the content of ut . . . existimem, the
third argument with admoneor, which explains the accusative case of rem non
difficilem. In (d), the appositive counterpart of rem incredibilem is the preceding
accusative and infinitive clause, which must be understood as the object of an implied
verb of communication. This explains the accusative case of rem. Tacitus, in his
endeavour to concentrate the wording to the minimum, has cases like (e). Here, the
appositive noun phrase refers to manus intentantes: by their action those who shook
their hands created or caused a causa . . . , whence the accusative.187 Note that the
sentence continues with an appositive secondary predication.
(b) Deinde quinque faciunt quasi membra eloquentiae, invenire quid dicas . . .
tum ad extremum agere ac pronuntiare; rem sane non reconditam.
(‘After that they set forth a sort of fivefold division of rhetoric, to choose what to
say . . . and in the end actually to deliver it—assuredly no mysterious progress.’ Cic. de
Orat. 2.79)
(c) Cuius hoc dicto admoneor ut aliquid etiam de humatione et sepultura
dicendum existimem, rem non difficilem . . .
(‘And this saying suggests the thought that I ought to say a word about interment and
burial—no difficult matter . . . ’ Cic. Tusc. 1.102)
187
The fullest account of Tacitus’ usage is still Nipperdey-Andresen ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Levant quidam regis facinus; in tesserarum prospero iactu vocem eius ambi-
guam, ut occidi iussisse videretur, a Fidenatibus exceptam causam mortis
legatis fuisse, rem incredibilem, interventu Fidenatium, novorum sociorum
consulentium de caede ruptura ius gentium, non aversum ab intentione lusus
animum nec deinde in errorem versum facinus.
(‘Some persons seek to palliate the king’s act, saying that an ambiguous expression of
his upon a lucky throw of dice, which made him seem to order them to kill the
envoys, was heard by the Fidenates and was responsible for the men’s death. But it is
quite incredible that the king on being interrupted by the Fidenates, his new allies,
come to consult him about a murder that would violate the law of nations, should not
have withdrawn his attention from the game, and that the attribution of the crime to
a mistake did not come later.’ Liv. 4.17.3–4)
(e) Postremo deserunt tribunal, ut quis praetorianorum militum amicorumve
Caesaris occurreret manus intentantes, causam discordiae et initium
armorum, maxime infensi Cn. Lentulo, quod . . .
(‘At last they left the tribunal, shaking their fists at any guardsman, or member of the
Caesar’s staff, who crossed their road, in order to supply a ground of quarrel and initiate
a resort to arms. They were bitterest against Gnaeus Lentulus . . . ’ Tac. Ann. 1.27.1)
Supplement:
Eumen<en>, quoius amicitiam gloriose ostentant, initio prodidere Antiocho, pacis
mercedem. (Sal. Hist. 4.69.8); . . . nullo in bellum adhortante, multis ad transitionem,
qui suas centurias turmasque tradere, donum victori et sibi in posterum gratiam,
certabant. (Tac. Hist. 3.61.1);
Whereas in the above examples one might hold that the appositive noun phrase refers
to an entity with which it shares its syntactic function, it is less easy to see why in (f )
one should take indigna homine dubitatio as an appositive, as K.-St. do. The noun
phrase is in the nominative, and there is no way to explain this on the basis of the
preceding context. Editors, apparently, take it as an independent sentence (without
the verb sum). It is an evaluating comment, like the preceding ones, but syntactically
it is entirely different (and not an appositive at all). Other examples of this type are
(g)—likewise in K.-St.—and (h) (with res).
(f ) Existit autem hoc loco quaedam quaestio subdifficilis num quando amici
novi, digni amicitia, veteribus sint anteponendi, ut equis vetulis teneros
anteponere solemus. Indigna homine dubitatio.
(‘But at this point there arises a certain question of some little difficulty: Are new
friends who are worthy of friendship, at any time to be preferred to old friends, as we
are wont to prefer young horses to old ones? The doubt is unworthy of a human
being.’ Cic. Amic. 67)
(g) Negat ullum esse tempus quo sapiens non beatus sit. Omnia philosopho
digna sed cum voluptate pugnantia.
(‘He says that there is no season when the wise man is not happy: all thoughts worthy
of a philosopher but at variance with pleasure.’ Cic. Tusc. 3.49)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Apposition
(h) Frumentariam legem C. Gracchus ferebat. Iucunda res plebei. Victus enim
suppeditabatur large sine labore.
(‘Gaius Gracchus brought forward a corn law: a measure agreeable to the masses. For
it provided food in abundance without work.’ Cic. Sest. 103)
The literature mentions as an instance of ‘clausal apposition’ in the dative (i) from
Tacitus, in some editions printed with a comma after evenisse. The dative can be
explained as a predicative dative functioning as subject complement with evenisse
(instead of fuisse), parallel to secunda in the preceding coordinate clause. Striking and
unexpected as the phrase may seem, by its content and its position, it has nothing
‘appositive’ about it; rather it is a full constituent of its clause.
(i) Multa Romanis secunda, quaedam Parthis evenisse documento adversus
superbiam.
(‘The Romans had had numerous successes, but some things had gone the Parthians’
way as a lesson against haughtiness.’ Tac. Ann. 15.27.2—tr. A. J. Woodman)188
Another type of expressions called ‘clausal apposition’ is illustrated by (j) and (k), a
type favoured by the historians but also found in Augustan and later poetry. Here,
the interpretation neither as an apposition nor as an independent sentence is likely
(in (j) one might paraphrase the constituent as a clause: ‘these are gifts of your rich
masters’). In this Syntax such expressions are treated as tail constituents (see Chapter 22).
(l) Sensit, ut ipsa suis aderat Venus aurea festis, / vota quid illa velint et, amici
numinis omen, / flamma ter accensa est apicemque per aera duxit.
(‘But golden Venus (for she herself was present at her feast) knew what that
prayer meant; and, as an omen of her favouring deity, thrice did the flame
burn brightly and leap high in the air.’ Ov. Met. 10.277–9)
188
For more examples in Tacitus, see Furneaux in his commentary (Introd. 1896: 47)—to be used with
caution.
189
See Klingner (1955: 197).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(m) Per idem tempus L. Piso pontifex, rarum in tanta claritudine, fato obiit . . .
(‘About the same time, the pontiff Lucius Piso—rare accident in one of his
great fame—died in the course of nature.’ Tac. Ann. 6.10.3)
A further type of expressions called ‘clausal apposition’ is illustrated by (n) and (o).
These expressions are secondary predicates that indicate in what way the entities to
which they are related function: the sedes in (n) served as a payment for the minions’
crimes. Ut mercedem would have been possible as well (see Chapter 21).
In conclusion, the term ‘clausal apposition’ is too easily used for any constituent that
is loosely connected to its surrounding context, especially if it is positioned at the end
of a sentence. For further examples of loosely attached utterances, especially parti-
ciples and ablative absolute clauses, see Chapter 22.
In this section most examples are adjective phrases functioning as attribute. Examples
of arguments marked by a specific case are (a)–(c) (genitive, dative, and ablative,
respectively). Arguments marked by a preposition are illustrated by (d) and (e).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Another type of obligatory constituents with adjectives is found with the comparative
and superlative forms, which require the expression of a basis of comparison, for
example with the comparative a constituent containing quam, as in (g), and with the
superlative a constituent in the genitive, as in (h) (for comparison, see Chapter 20).
adjunct (cf. § 10.3); in (j), a position in time adjunct (cf. § 10.27). Exx. (k)–(n) have
respect adjuncts (cf. § 10.90). Most common are ablatives of respect (ablativus limi-
tationis), as in (k), and prepositional phrases, as in (n). The genitive of respect is
common from Early Latin onwards in expressions like (l) with animi and adjectives
indicating mental or physical condition, but it is found in other combinations as well
and its use is expanded in poetry (among other things to avoid prepositions). The use
of the accusative of respect (often called accusativus Graecus) starts with the Augustan
poets and is clearly a stylistic development. In (o) there occurs a manner adjunct
signalling the subjective judgement of the speaker (cf. § 10.47).190
(i) Quis Platorem . . . hominem in illis locis clarum ac nobilem, legatum
Thessalonicam . . . venisse nescit?
(‘Is it not an acknowledged fact that one Plator . . . a noble and a celebrity in his own
land, came . . . to Thessalonica as legate?’ Cic. Har. 35)
(j) (sc. bona) . . . duobus milibus nummum sese dicit emisse adulescens vel
potentissimus hoc tempore nostrae civitatis, L. Cornelius Chrysogonus.
(‘ . . . a young man, at the present time perhaps the most powerful in the State, claims
to have bought the same for 2000 sesterces—I mean Lucius Cornelius Chrysogonus.’
Cic. S. Rosc. 6)
(k) Quis P. Octavio Balbo ingenio prudentior, iure peritior, fide, religione, officio
diligentior aut sanctior commemorari potest?
(‘Whom can you call to mind more gifted by nature, more versed in law, more careful
or more scrupulous in honour, faith, and duty than P. Octavius Balbus?’ Cic. Clu. 107)
(l) Ita me Amor lassum animi ludificat / . . .
(‘Love mocks me in my weariness of soul . . . ’ Pl. Cist. 214–15)
(m) . . . et te flava comas frugum mitissima mater / sensit equum . . .
(‘The golden-haired mother of corn, most gentle, knew thee as a horse.’ Ov. Met.
6.118–19)
(n) Vehementem accusatorem nacti sumus, iudices, et inimicum in omni genere
odiosum ac molestum.
(‘We have got an energetic prosecutor, gentlemen, and an enemy in every way
offensive and dangerous.’ Cic. Flac. 13)
(o) Retinere incipit tabulas Theomnastus quidam, homo ridicule insanus, quem
Syracusani Theoplactum vocant.
(‘A man called Theomnastus took the tablets and held on to them—an amusingly
crazy fool whom the Syracusans call Theoplactus.’ Cic. Ver. 4.148)
190
For this use of adverbs as manner adjuncts, see von Nägelsbach and von Müller (1905: 372–6), who
interpret some as degree adjuncts.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Space and time: Heraclius est Hieronis filius Syracusanus, homo in primis domi suae
nobilis et ante hunc praetorem vel pecuniosissimus Syracusanorum, nunc nulla alia
calamitate nisi istius avaritia atque iniuria pauperrimus. (Cic. Ver. 2.35); Grave
crimen est hoc et vehemens et post hominum memoriam iudiciaque de pecuniis
repetundis constituta gravissimum . . . (Cic. Ver. 3.130); . . . atque huius potentiae
causa matrem in Biturigibus homini illic nobilissimo ac potentissimo collocasse . . .
(Caes. Gal. 1.18.6);
Respect: Vidi simul cum populo Romano forum comitiumque adornatum ad spe-
ciem magnifico ornatu, ad sensum cogitationemque acerbo et lugubri. (Cic. Ver. 58);
Haec cum graviter tum ab exemplis copiose, tum varie, tum etiam ridicule et facete
explicans . . . (Cic. Brut. 198); Nec hoc tam re est quam dictu inopinatum atque
mirabile. (Cic. Parad. 35); . . . quique pedum cursu valet et qui viribus audax / aut
iaculo incedit melior levibusque sagittis . . . (Verg. A. 5.67–8); Una namque ire videbat
/ egregium forma iuvenem et fulgentibus armis . . . (Verg. A. 6.860–1); Vicensimo et
altero die cum sol est exortus, luna tenet circiter caeli medias regiones et id quod
spectat ad solem, id habet lucidum reliquis obscura. (Vitr. 9.2.4); Estque tum scrotum
et rotundum et tactu leve. (Cels. 7.18.5); Per ea tempora Mithridates, Ponticus rex,
vir neque silendus neque dicendus sine cura, bello acerrimus, virtute eximius,
aliquando fortuna, semper animo maximus, consiliis dux, miles manu, odio in
Romanos Hannibal, occupata . . . cum terribilis Italiae quoque videretur imminere,
sorte obvenit Sullae Asia provincia. (Vell. 2.18.1–3);
Sati’n tu es sanus mentis aut animi tui / . . . (Pl. Trin. 454); Quod quidem me
factum consili incertum facit. (Ter. Ph. 578); . . . ne quid aeger animi ferox iuvenis
gravius consuleret . . . (Liv. 30.15.9); Viderat, obscura vallum dum nocte pererrat /
aeger consilii curisque novissima volvens . . . (Stat. Theb. 11.140–1);
Nam ego vos novisse credo iam ut sit pater meus, / quam liber harum rerum
multarum siet . . . (Pl. Am. 104-5); . . . ne ille ecastor hinc trudetur largus lacrumarum
foras. (Pl. As. 533); Ingens ipse virium atque animi. (Sal. Hist. 3.91); Lucus in urbe
fuit media, laetissimus umbrae . . . (Verg. A. 1.441); Tum Drances idem infensus . . .
/ . . . / largus opum et lingua melior . . . surgit. (Verg. A. 11.336–42);191 Ut rapit in
praeceps dominum spumantia frustra / frena retentantem durior oris equus. (Ov.
Am. 2.9b.29–30); . . . utque deorum / spretor erat mentisque ferox . . . (Ov. Met.
8.612–13); Sed femina ingens animi munia ducis per eos dies induit . . . (Tac. Ann.
1.69.1); At in superiore Germania Caecina, decorus iuventa, corpore ingens, animi
immodicus, scito sermone, erecto incessu, studia militum inlexerat. (Tac. Hist.
1.53.1—NB: variation);
Olli serva datur operum haud ignara Minervae, / Cressa genus, Pholoe. (Verg. A.
5.284–5); Poenorum qualis in arvis / saucius ille gravi venantum vulnere pectus / tum
demum movet arma leo . . . (Verg. A. 12.4–6—cf. . . . iam vulgatum actis quoque:
‘saucius pectus’. (Quint. Inst. 9.3.17)); Nec mihi displiceat (sc. bos) maculis insignis
et albo, / aut iuga detrectans interdumque aspera cornu / et faciem tauro propior . . .
(Verg. G. 3.56–8); Candidas ales modo movit alas, / dulcior vocem moriente cygno.
(Sen. Phaed. 301–2); (sc. feminae) . . . nudae brachia ac lacertos. (Tac. Ger. 17.2); Alius
191
For Virgil, see Görler (1985: 266).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
manum aeger eodem deo auctore ut pede ac vestigio Caesaris calcaretur orabat. (Tac.
Hist. 4.81.1);
Mauris omnibus rex Bocchus imperitabat, praeter nomen cetera ignarus populi
Romani . . . (Sal. Jug. 19.7); . . . omnia Mercurio similis, vocemque coloremque / et
crinis flavos et membra decora iuventa. (Verg. A. 4.558–9); . . . iniuriosus alia et
iracundus . . . (Amm. 31.14.6);
Manner: Quid huic tam loquaciter litigioso responderet ille unde petebatur non
habebat. (Cic. Mur. 26); Quid est enim verius quam neminem esse oportere tam
stulte adrogantem ut . . . (Cic. Leg. 2.16); Sapientis autem civis fuit causam nec
perniciosam et ita popularem ut non posset obsisti perniciose populari civi non
relinquere. (Cic. Leg. 3.26); Una de multis face nuptiali / digna, periurum, fuit, in
parentem / splendide mendax . . . (Hor. Carm. 3.11.33–5); Adversus haec Tempani
oratio incompta fuisse dicitur, ceterum militariter gravis . . . (Liv. 4.41.1); Pavidus
tribunus, quippe qui . . . illum praevalidum iuvenem et, quod haud minus timendum
erat, stolide ferocem viribus suis cerneret . . . (Liv. 7.5.6); (sc. dux erat) In ludo
praeterea militari . . . comiter facilis. (Liv. 7.33.2—comis ac facilis cj. H. J. Müller); . . .
C. Curio . . . , vir nobilis, eloquens, audax, suae alienaeque et fortunae et pudicitiae
prodigus, homo ingeniosissime nequam et facundus malo publico . . . (Vell. 2.48.3);
Alias deinde (sc. eum) effeminatum et muliebriter timidum, alias proditorem ama-
toris appellans . . . (Curt. 6.7.11); Quid ergo inter nos et illos interest, ut Ariston ait,
nisi quod nos circa tabulas et statuas insanimus, carius inepti? (Sen. Ep. 115.8);
Quam grata est facies torva viriliter et pondus veteris triste supercili! (Sen. Phaed.
798–9); Nam (sc. Horatius) et insurgit aliquando et plenus est iucunditatis et gratiae
et varius figuris et verbis felicissime audax. (Quint. Inst. 10.1.96); Nam senem
Augustum devinxerat adeo uti nepotem unicum, Agrippam Postumum, in insulam
Planasiam proiecerit, rudem sane bonarum artium et robore corporis stolide fero-
cem, nullius tamen flagitii compertum. (Tac. Ann. 1.3.4); Caninius Rebi<l>us . . .
cruciatus aegrae senectae misso per venas sanguine effugit . . . ob libidines muliebriter
infamis. (Tac. Ann. 13.30.2); Ego perdidi viginti milia nummum meo nomine, sed
cum effuse in lusu liberalis fuissem, ut soleo plerumque. (Suet. Aug. 71.3);
Pinus, piceae, alni . . . si non integantur, cito senescunt, mirum in modum fortiores,
si umor extra quoque supersit. (Plin. Nat. 16.224);
Cause: Tum Cererem corruptam undis Cerealiaque arma / expediunt fessi rerum.
(Verg. A. 1.177–8); Iamque aegra timoris / Roma tuos numerat lacrimandos ma-
tribus annos. (Sil. 3.72–3);
Ibi Alexander cum animadvertisset portum naturaliter tutum . . . (Vitr. 2.pr.4); Fit
autem barbarismus principaliter modis duobus, pronuntiatione et scripto: <pronun-
tiatione,> si aut naturaliter longas syllabas breviter proferamus, ut Romam, aut si
naturaliter breves producamus, ut rosam. (Serv. G.L. IV.444.12);
In cases like (p) and (q) it is not necessary to assume that the adverb and the adjective
form an adjective phrase together: the adverbs may be taken as manner adjuncts at
the clause level.192
192
For these puns in Plautus, see Gehlhardt (1892: ch. 3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(p) . . . quia nullus umquam amator adeo est callide / facundus quae in rem sint
suam ut possit loqui.
(‘ . . . because no lover has ever been so clever at speaking that he could say
the things that benefit him.’ Pl. Mer. 35–6)
(q) Nunc huic ego graphice facetus fiam.
(‘Now I’ll become witty for him in a charming way.’ Pl. Per. 306)
Examples of adjective phrases containing a disjunct are (r)–(t). Exx. (r) and (s) have
an attitudinal disjunct; (t) an illocutionary disjunct.
(r) Quin etiam parens tuus . . . Catilinae fuit advocatus, improbo homini, at
supplici, fortasse audaci, at aliquando amico.
(‘Furthermore, your father . . . was counsel for Catiline, a rogue, but a suppliant,
reckless perhaps, but once a friend.’ Cic. Sul. 81)
(s) Quo quidem in genere familiaris noster M. Buculeius, homo neque meo
iudicio stultus et suo valde sapiens et ab iuris studio non abhorrens . . . erravit.
(‘In that kind of action too our friend Marcus Buculeius, no fool in my opinion, and
mightily wise in his own, and a man with no distaste for legal studies . . . went wrong.’
Cic. de Orat. 1.179)
(t) C. Fimbria temporibus isdem fere sed longius aetate provectus habitus est
sane, ut ita dicam, luculentus patronus.
(‘Gaius Fimbria, at about the same time with these, though he lived much longer, was
accounted an excellent, if I may use the word, pleader.’ Cic. Brut. 129)
Supplement:
Homo enim videlicet timidus aut etiam permodestus vocem consulis ferre non
potuit. (Cic. Catil. 2.12); Mihi nunc vos, inquit Crassus, tamquam alicui Graeculo
otioso et loquaci et fortasse docto atque erudito quaestiunculam de qua meo arbi-
tratu loquar ponitis? (Cic. de Orat. 1.102); Accedit quod orationis etiam genus habent
fortasse subtile et certe acutum, sed ut in oratore exile, inusitatum . . . (Cic. de Orat.
3.66); (sc. expecto annum) Plane aut tranquillum nobis aut certe munitissimum.
(Cic. Q.fr. 2.15.2); Sunt tibi, M. Tulli, liberi propinquique digni quidem te et merito
tibi carissimi. (Cas. Fam. 12.13.2); . . . honestos / fascibus et sellis nollem mihi
sumere, demens / iudicio vulgi, sanus fortasse tuo . . . (Hor. S. 1.6.96–8); . . . excepit
orationem eius princeps Aetolorum Alexander, vir ut inter Aetolos facundus.
(Liv. 32.33.9);
The quantity and degree expressions that can be used at the clause level (see § 10.61)
are also common in adjective phrases. Following the order of § 10.61, five types can
be distinguished: (i) accusative neuter forms of adjectives (not only adjectives of
quantity and size), as in (a); (ii) adverbs formed from descriptive and other
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
‘insufficiently’, vix ‘hardly’ (aliquam ‘a fair number or amount of ’ is only used with
forms of multus ‘many’, ‘much’). Examples are (a)–(c).
(a) Quam vero indignus videor?
(‘How undeserving do I seem?’ Pl. As. 697)
(b) Ex quo uno genere totus est Tutor, mimus vetus, oppido ridiculus.
(‘The Guardian, an ancient and exceedingly droll farce, was entirely made up of this
sort of thing.’ Cic. de Orat. 2.259)
(c) . . . me amisisti a te vix vivom domum.
(‘ . . . since you let me get back home from you, even if barely alive.’ Pl. Mos. 432)
Supplement (in alphabetical order by adverb):
Itaque me malum esse oportet, callidum, astutum admodum . . . (Pl. Am. 268); Erat
admodum amplum [et excelsum] signum cum stola. (Cic. Ver. 4.74); Vitium tibi
istuc maxumum est, blanda es parum. (Pl. Cas. 584); Itaque nunc illi homines
honestissimi propter istius insidias parum putantur cauti providique fuisse. (Cic.
S. Rosc. 117); . . . quin scelestus sacerdotem anum praecipes / reppulit propulit per-
quam indignis modis . . . (Pl. Rud. 671–2); Tum Crassus ‘Atqui vide’ inquit ‘in
artificio perquam tenui et levi quanto plus adhibeatur diligentiae quam . . . (Cic. de
Orat. 1.129); Qualem porro illam feminam fuisse putatis, iudices, quam pudicam, quae
cum Verre loqueretur, quam religiosam, quae sacrarii spoliandi rationem ostenderet?
(Cic. Ver. 4.102); . . . nec satis liber sibi videtur nec satis frugi nec sat honestus . . . (Pl.
Per. 839); . . . te pugna Cannensis accusatorem sat bonum fecit. (Cic. S. Rosc.
89); . . . quae parentes tam in angustum tuos locum compegeris. (Pl. Rud. 1147);
Tu Apronium, hominem vix liberum . . . verbo quidem graviore appellasses . . . (Cic.
Ver. 3.134);
NB: Vulnera nodo Herculis praeligare mirum quantum ocior medicina est . . . (Plin.
Nat. 28.64);
NB: cf. with adverbs: Oppido opportune te obtulisti mi obviam. (Ter. Ad. 322); Ut
adsimulabat Sauream med esse quam facete! (Pl. As. 581);
The scalar adverb quamvis ‘however’ is found with gradable adjectives and adverbs
from Plautus onwards, as in (a) and (b). Its development from the degree adverb
quam + vis ‘in which degree do you wish?’ is apparent from the existence of other
inflected forms of the verb volo ‘to wish’, as quam voltis in (c). The use of quamvis
with a gradable lexeme in a sentence with a concessive subjunctive, followed by
another sentence (that may contain tamen ‘nevertheless’) is the starting point of its
use as a concessive subordinator, as in (d) and (e)—see § 16.75.193 Quamlibet is used
in a similar way as a scalar adverb.
193
For quamvis as a degree adverb see Bertocchi (2002) and Spevak (2005b: 74–81), with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
194
More examples in OLD s.v. quam § 3.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
195
On this example, see Adams (2005b: 87).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Quid dixit? # Si tu illum solem sibi solem esse diceres, / se illum lunam
credere esse et noctem qui nunc est dies.
(‘What did he say? # That if you told him that the sun up there is the sun, he’d believe
that it’s the moon and that what’s now day is the night.’ Pl. Bac. 699–700)
(b) Neque enim pauci neque leves sunt qui se duo soles vidisse dicant . . .
(‘For those who claim to have seen two suns are neither few nor untrustworthy . . . ’
Cic. Rep. 1.15)
(c) Et bulla aurea est pater quam dedit mi natali die.
(‘And there's a golden locket which my father gave me on my birthday.’ Pl. Rud.
1171)
(d) Tuos pater bellissumus / amicam adduxit intro in aedis.
(‘Your terribly charming father has brought a mistress into the house.’ Pl. Mer.
812–13)
can be referred to in other ways than by NPs. Apart from proper names and personal
pronouns, which have been already mentioned, one may think of relative pronouns
and of verb endings, especially those of the first and second person. The determiners
discussed in §§ 11.103–11.117 can be used to explicitly indicate definiteness and
indefiniteness. Thus, in (e), the demonstrative pronoun haec points to a young lady
who just appeared on the scene. By using the demonstrative determiner illa, the
speaker verifies with the addressee whether she is ‘that kidnapped girl’ they talked
about before, and not just ‘a kidnapped girl’ (illa is used anaphorically). In (f ), the
anaphoric determiner eam refers to Euclionis filia mentioned by the addressee, and
not just to ‘a girl’. In (g), aliquem indicates that some nuntius may be involved, but
that the addressee (pater) is not supposed to know who he is or, as in this case, that his
identity is irrelevant (note also the subjunctive veniat). In (h), aliquos contains no
suggestion that the interlocutors might be aware of specific gods, only that ‘some’ or ‘a
few’ gods might be involved. Without aliquos, otiosos deos probably would be
interpreted as ‘the gods just mentioned’. In (i), quoidam indicates that the speaker
has a particular merchant in mind, but for some reason does not (want to) tell who he
is. Illa and eam serve as definiteness markers, aliquem, aliquos, and quoidam as
indefiniteness markers. Details are given in the following sections.
(e) Haecine illa est furtiva virgo?
(‘Is this that kidnapped girl?’ Pl. Per. 545)
(f ) . . . siquidem ita est ut praedicas, / te eam compressisse vinolentum virginem.
(‘ . . . if it is as you tell me and you did violence to this girl when you were drunk.’ Pl.
Aul. 688–9)
(g) Nam pater expectat aut me aut aliquem nuntium / qui hinc ad se veniat.
(‘For your father is expecting either me or some messenger to come from here to
him.’ Pl. Capt. 382–3)
(h) Cur tam multos deos nihil agere et cessare patitur, cur non rebus humanis
aliquos otiosos deos praeficit . . . ?
(‘Why does it suffer so many gods to be idle and keep holiday? Why does it not
appoint some of the leisured gods . . . to superintend human affairs?’ Cic. N.D. 3.93)
(i) . . . cepi tabellas, consignavi, clanculum / dedi mercatori quoidam . . .
(‘ . . . I took tablets, sealed them, and secretly gave them to a certain merchant . . . ’ Pl.
Mil. 130–1)
A considerable percentage of definite nouns and noun phrases, in some texts even
the majority, have no overt marking of definiteness. This means that the interpret-
ation of these NPs as definite depends on the meanings of the NPs themselves and/or
on contextual or situational information. Overt markers that bring about an
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In the text that surrounds this narrative (for details, see the note) the use of definite
determiners is less rare. There are fifteen instances of hic,196 four of iste, six of ille, and
one of is. In addition, there are forty-four possessive expressions. Hic is almost only
used as a demonstrative determiner with temporal (hac noctu/nocte ‘this night’) or
spatial reference, as in (e) and (f ). The only exception is hoc onus in (g), which
anaphorically refers to the difficult life of slaves just narrated by the speaker (Sosia)
and for which the strictly anaphorical id seems not necessarily to be excluded, albeit
with a difference in meaning (onus alone would not do); broadly speaking, hic refers
to entities that are part of the situation in which the speaker finds himself (see also
§ 11.104). In (f ), something like hoc is essential to avoid an indefinite interpretation of
pallium ‘a cloak’ and to make it clear that it is Sosia’s cloak. In both (e) and (f ) the
actor may actually point to or touch the objects in question.
(e) Haec sic dicam erae. / Nunc pergam eri imperium exsequi et me domum
capessere. / # Attat, illic huc ituru’st. Ibo ego illi obviam, / neque ego huc
hominem hodie ad aedis has sinam umquam accedere.
(‘That’s how I’ll tell my mistress the story. Now I’ll continue to carry out master’s
command and go home. # Oho, he’s going to come here. I’ll go and meet him. I’ll
never let him come here to this house today.’ Pl. Am. 261–4—NB: Studemund
emended the text: hunc hominem <huc> hodie, but the use of a determiner hunc
with hominem after illi is very odd; cf. 295 for the normal anaphoric use of homo and
§ 11.155)
(f ) Mi in mentem venit, / illic homo <hodie> hoc denuo volt pallium detexere.
(‘It looks as if he wants to weave my cloak again today.’ Pl. Am. 293–4)
(g) Habendum et ferundum hoc onu’st cum labore.
(‘You must take up and bear this wearisome burden.’ Pl. Am. 175)
The use of iste is what one would expect: it refers to an entity associated with the
addressee in the communicative situation (see also § 11.104). An interesting case is
(h). Sosia directly addresses the audience where the scortatores he is talking about are
located or whose identity the audience knows.197 The four instances of illic homo ‘that
fellow’ are all in asides, directed to the audience, as in (i). Ille here refers to an entity
that does not form part of the interaction between speaker and addressee (see also
§ 11.104). A less clear example is (j), where the use of strictly anaphoric eum does not
seem impossible; however, some form of implicit contrast may explain illum. And
numerum alone would seem odd. The only instance of is (k) is resumptive with a
preceding relative clause (see also § 11.138).
(h) Ubi sunt isti scortatores qui soli inviti cubant?
(‘Where are those lechers who are lying alone against their will?’ Pl. Am. 287)
196
Including four instances of the expression hoc noctis ‘at this time of night’.
197
Compare the use of iste in Pl. Mos. 356–61, on which see Fraenkel (1922: 141). For the involvement
of the audience, see also del Vecchio (2008), with references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 11.8 Definiteness and indefiniteness marking of nouns and noun phrases
(including proper names) in Caesar Gal. 1.2–12 and in Peregrinatio 1–4 (in %)198
Non-referring Referring
Indefinite Definite
ø ø proper names is, hic, ille ø
Caesar (N = 378) 10 15 28 27 19
Peregrinatio (N = 306) 8 16 14 29 33
ø = no overt marking
198
Source: Pinkster (1996: 145).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
NPs). The table shows that Caesar has relatively many proper names and, corres-
pondingly, relatively few unmarked definite common NPs, but the use of the various
devices in these two texts, with more than four centuries between them, is otherwise
very similar. In roughly half of the instances there is no marking at all. The reader has
to base his interpretation on the syntactic and semantic information the text contains
(and that must not have been a problem).
Demonstrative and anaphoric determiners, just like the corresponding pronouns,
indicate that the NP they modify is identifiable for the addressee, either (i) because it
is directly perceivable, or (ii) because it is present in the preceding context, or (iii)
because it is identifiable on the basis of the knowledge and experience the addressee
has. These various sources of identifiability are dealt with below. Anaphoric is is
actually limited to type (ii). For type (iii), ille is almost unique.
The frequency with which these words function as determiners or as pronouns varies
from text type to text type and also diachronically. The data for Plautus and for Cicero’s
orations can be found in Table 11.9. For Plautus and Cicero, in the absence of more
precise data, the figures indicate the number of columns in Lodge’s and Merguet’s
lexica, respectively, by approximation. In the case of Plautus the use of these words as
determiners of relative clauses could be found easily and therefore the relevant data are
included (in parentheses). Added are the data on Apuleius, in absolute numbers.199 For
Pliny the Elder total numbers are given, with a specification only for iste.
A few observations: (i) The distribution of anaphoric is on the one hand and demon-
strative hic/ille/iste on the other correlates with the type of text and seems also to
199
Figures taken from Callebat (1968: 264), who uses Wolterstorff (1917).
200
Source: Rosumek and Najock (1996), with two quotations from Cato excluded.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
change in the course of time (see also § 11.108). (ii) The proportion of the use of the
words as determiner or pronoun varies. (iii) In these authors iste is the least frequent.
Appendix: Combinations of ille and iste with ecce ‘behold’ are at the basis of Italian
quello and questo and related forms in other Romance languages. An early instance is (l).
These formations are back formations from combinations in which ille and iste behave
as accusative objects. This is attested already in Plautus, as in (m), but there need not be a
direct connection between the use in Plautus and in the Romance languages.201
(l) Libertus eccille qui clavis eius loci in hodiernum habet et a vobis stat
numquam se ait inspexisse . . .
(‘There is your freedman who still has charge of the keys of the place; he is one
of your witnesses, but he says that he has never examined . . . ’ Apul. Apol. 53.8)
(m) Ubinam est is homo gentium? #/ Eccillum video.
(‘Where on earth is that chap? # Look, I can see him.’ Pl. Mer. 434–5)
201
For the diachronic development, see Adams (2013: 465–81). For the Plautine material, see
Perdicoyianni-Paléologou (2006).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Nescio pol quae illunc hominem intemperiae tenent. (Pl. Aul. 71); Agedum pulta
illas fores. (Pl. Cist. 637); Omnes illae columnae quas dealbatas videtis machina
adposita nulla impensa deiectae eisdemque lapidibus repositae sunt. (Cic. Ver.
1.145); . . . si meminisse voltis . . . illos colles quos cernitis patriae vestrae esse . . . (Liv.
7.40.6—NB: direct speech);
Iam tibi istuc cerebrum dispercutiam, excetra tu . . . (Pl. Cas. 644); At ego mi
anulum dari istunc tuom volo. (Pl. Mil. 771); Nam cum hic Sex. Roscius esset
Ameriae, T. autem iste Roscius Romae, cum hic filius adsiduus in praediis esset
cumque se voluntate patris rei familiari vitaeque rusticae dedisset, ipse autem
frequens Romae esset . . . (Cic. S. Rosc. 18); Has ut hodie, ut in isto templo finiatis
simultates quaesumus vos universi . . . (Liv. 40.46.7—NB: direct speech);
In a broader sense hic is used to refer to an entity that is physically present near the
speaker or is related in some way to the actual situation of the speaker, iste to refer to
an entity that is physically near or is in some way related to the actual situation of the
addressee, whereas ille is used to refer to an entity the speaker regards or presents as
not belonging to the actual situation, that is as somehow distant. The choice between
these ways of referring to an entity is to some extent up to the speaker, as is illustrated
by (d), a passage from Plautus’ Pseudolus.202 The impostor Simia and the pimp Ballio
have entered Ballio’s house (and therefore have left the stage). The slave Pseudolus
(who, posing as Harpax, has asked for Simia’s help to trick the pimp out of a slave girl)
is speaking. Pseudolus’ use of hic for his associate and of ille for his enemy is
understandable. Less obvious is illum hominem in l. 1019 (contrast with ego?; a
person you’re afraid of is more distant?). Illi (Simia) in l. 1023 can be explained as
contrastive.
(d) Peiorem ego hominem magisque vorsute malum
numquam edepol quemquam vidi, quam hic est Simia.
Nimisque ego illum hominem metuo et formido male 1019
ne malus item erga me sit ut erga illum fuit,
ne in re secunda nunc mi obvertat cornua,
si occasionem capsit qui sit <mihi> malus.
Atque edepol equidem nolo, nam illi bene volo. 1023
Nunc in metu sum maximo, triplici modo:
primum omnium iam hunc comparem metuo meum
ne deserat med atque ad hostis transeat;
...
quom haec metuo, metuo ne ille huc Harpax advenat 1030
prius quam hinc hic Harpax abierit cum muliere.
(‘I’ve never seen a worse man and bad in a more wicked way than this Simia is. And
I’m terribly afraid and scared that that man [Simia] could be bad to me in the same
way he was bad to that chap [Ballio], and that he might turn his horns against me
202
The same sequence is used in TLL s.v. hic 2703.16ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
now in his success, if he gets an opportunity to be bad like this. And I for one don't
want it, as I’m well disposed to him [Simia]. Now I’m in greatest fear in a threefold
way: first of all I’m afraid now that this comrade of mine [Simia] could desert me and
go over to the enemy. . . . And while I fear this, I fear that the real Harpax might come
here before my Harpax [Simia] has left with the woman.’ Pl. Ps. 1017–31)
Supplement:
Tantum gemiti et mali maestitiaeque / hic dies mi optulit, famem et pauperiem. (Pl.
Aul. 721–2); Em istuc si potes / memoriter meminisse, inest spes nobis in hac astutia.
(Pl. Capt. 249–50); Quid enim est huic reliqui quod eum in hac vita teneat . . . (Cic.
Sul. 89); Satis tu quidem in Consolatione es lamentatus. Quam cum lego, nihil malo
quam has res relinquere . . . (Cic. Tusc. 1.76); Ite obviam Hannibali, exornate urbem
diemque adventus eius consecrate, ut hunc triumphum de cive vestro spectetis. (Liv.
23.40.8—NB: direct speech);
Satiu’st me queri illo modo servitutem. (Pl. Am. 176); Quid illum putas, / natura
illa atque ingenio? (Pl. Trin. 811–12); Si enim mihi hodie respondere ad haec quae
dico potueris . . . posse te et illi quoque iudicio non deesse . . . arbitrabor. (Cic. Div.
Caec. 47); Est autem etiam de Brundisio atque illa ora Italiae providendum. (Cic.
Phil. 11.26); Id munimentum illo die fortuna urbis Romanae habuit. (Liv. 2.10.2);
Temperare istac aetate istis decebat artibus. (Pl. Merc. 982); Sequere hac. # Ubi
istic leno est? (Pl. Rud. 1357); P. Servili, Q. Metelle, C. Curio, L. Afrani, cur hunc non
audistis tam doctum hominem, tam eruditum, prius quam in istum errorem indu-
ceremini? (Cic. Pis. 58); Tu ad me de istis rebus omnibus scribas velim quam
diligentissime. (Cic. Fam. 7.13.2); ‘Enimvero’ inquit consul ‘ferri iam ludificatio
ista non potest.’ (Liv. 38.14.11);
A remarkable instance of the use of hic and ille to refer to the same person is (e). One
explanation is that the parasite who is speaking is changing position on the stage. The
other is that hic is appropriate from the parasite’s position on the stage, ille from the
position of the audience he is addressing.203
(e) Periuriorem hoc hominem si quis viderit / aut gloriarum pleniorem quam
illic est, / me sibi habeto . . .
(‘If anyone has seen a man more perjured than this one or more boastful
than that one is, he can have me for himself.’ Pl. Mil. 21–3)
Appendix: From Vitruvius onwards a few instances of is are found in contexts in
which one would expect hic, referring to the actual situation and without its
anaphoric meaning. Two examples cited in this context are (f ) and (g).204
203
So Brix, Niemeyer, and Köhler ad loc.
204
For more instances, see Wistrand (1961, 1963); also Pieroni (2010a: 444).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(g) Quid enim futurum est, si nec dii nec homines eius (huius cj. Scheffer)
coloniae miserentur?
(‘Whatever is to happen if neither the gods nor man will take pity on this
town?’ Petr. 44.16 (Ganymedes speaking))
(c) Ai’n tu meum virum hic potare, opsecro, cum filio / et ad amicam detulisse
argenti viginti minas / meoque filio sciente id facere flagitium patrem?
(‘I beg you, are you saying that my husband’s drinking here together with his son?
And that he’s brought twenty silver minas to a mistress? And that he, the father, is
committing such an outrage with the full knowledge of my son?’ Pl. As. 851–3)
(d) . . . lotium conservato eius qui brassicam essitarit: id calfacito, eo hominem
demittito. Cito sanum facies hac cura: expertum hoc est. Item pueros pusillos
si laves eo lotio, numquam debiles fient. Et quibus oculi parum clari sunt, eo
lotio inunguito: plus videbunt. Si caput aut cervices dolent, eo lotio caldo
lavito: desinent dolere. Et si mulier eo lotio locos fovebit, numquam m(e)n
(ses) seri fient.
(‘ . . . save the urine of a person who eats cabbage habitually, heat it, and bathe the
patient in it, he will be healed quickly; this remedy has been tested. Also, if babies are
bathed in this urine they will never be weakly; those whose eyes are not very clear will
see better if they are bathed in this urine; and pain in the head or neck will be relieved
if the heated urine is applied. If a woman will warm the privates with this urine, her
periods will never be late.’ Cato Agr. 157.10–11)
(e) Iam ea res me horrore afficit, / erilis praevortit metus.
(‘The previous events were already filling me with terror, but the fear of my mistress
prevailed.’ Pl. Am. 1068–9)
(f ) Et primoribus qui iam diu publicis consiliis aberant propalam minabantur
nisi venirent in senatum circa domos eorum ituros se et in publicum omnes
vi extracturos esse. Is timor frequentem senatum magistratui praebuit.
(‘And they openly threatened the leading citizens, who for a long time had been
absent from public considerations, that if they did not come into the senate, they
would make the rounds of their homes and forcibly bring them all out into the
streets. The fear of that gave the magistrate a full session of the senate.’ Liv. 26.13.1)
The anaphoric use of demonstrative determiners resembles the use of the anaphoric
determiner to some extent. However, whereas the anaphoric determiner merely links
its noun phrase to the preceding discourse, the demonstratives also contribute the
meaning they have when used deictically.205 So hic often contributes something like
‘the entity I am talking about’, iste ‘the entity you are talking about’ or ‘the one
I mentioned to you’. Ille often contains an element of contrast. A few examples from
Plautus are given below in the Supplement. Ille and iste in particular are used when
the earlier mention of the entity is farther away. When hic and ille are used in the
same context, hic refers to the entity that was last mentioned. Further differences are
dealt with in Chapter 24.
205
For the difference between the anaphoric use of demonstrative pronouns and determiners and the
use of the anaphoric pronoun/determiner is, see de Jong (1996, 1998) and Kroon (forthc.), with
references.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
Quo in loco haec res sit vides, / Epidice. (Pl. Epid. 81-2); Pallam ad phrygionem fert
confecto prandio / vinoque expoto, parasito excluso foras. / Non hercle ego is sum
qui sum, ni hanc iniuriam / meque ultus pulchre fuero. (Pl. Men. 469–72); (preced-
ing: is adulescens [78] . . . duae filiae [84]) Earum hic adulescens alteram efflictim
perit . . . (Pl. Poen. 96);
Immo est <quod satis est>, et di faciant ut siet / plus plusque, <et> istuc sospitent
quod nunc habes. # / (Aside) Illud mihi verbum non placet, ‘quod nunc habes’. / Tam
hoc scit me habere quam egomet. (Pl. Aul. 545–8); Edepol facinus improbum. # At
iam ante alii fecerunt idem. / Erit illi illa res honori. (Pl. Epid. 32–3—NB: relatively
rare in Plautus); (preceding: senex quidam [426] . . . quidam adulescens [427]) Ita ille
est quoi emitur, senex. / Sanus non est ex amore illius. Quod posces feres. / # Certe
edepol adulescens ille quoi ego emo efflictim perit / eius amore. # Multo hercle ille
magis senex, si tu scias. / # Numquam edepol fuit nec fiet ill’ senex insanior / ex
amore quam ille adulescens cui ego do hanc operam, pater. (Pl. Mer. 442–7—NB: ille
may be demonstrative; the passage is quite complicated);
Comesse panem tris pedes latum potes, / fores pultare nescis. Ecquis in aedibu’st? /
Heus, ecquis hic est? Ecquis hoc aperit ostium? / Ecquis exit? # Quid istuc? Quae
istaec est pulsatio? / <Quid?> Quae te mala crux agitat, qui ad istunc modum / alieno
viris tuas extentes ostio? (Pl. Bac. 580–5); Deos volo bene vortere / istam rem vobis.
(Pl. Cur. 658–9); Abi in malam rem maxumam a me / cum istac condicione. (Pl. Epid.
78–9); Numquam edepol quisquam me exorabit quin tuae / uxori rem omnem iam,
uti sit gesta, eloquar. / Omnes in te istaec recident contumeliae. (Pl. Men. 518–20); Ut
mihi aedis aliquas conducat volo / ubi habitet istaec mulier. (Pl. Mer. 560–1 (preceding:
adulescentem [873])); Papae! Divitias tu quidem habuisti luculentas. # / Miserum istuc
verbum et pessumum est, ‘habuisse’ et nihil habere. (Pl. Rud. 1320–1); Pater istius
adulescentis dedit has duas mi epistulas, / Lesbonici. Is mihi est amicus. (Pl. Trin. 894–5);
206
For statistical information concerning preparative and resumptive demonstratives in combination
with clauses, see Bodelot (1996). For hic, see TLL s.v. 2729.81ff.; for ille, see TLL s.v. 347.82ff.; for is, see
TLL s.v. 474.6ff.; for iste, see TLL s.v. 498.72ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
part of the complex sentence. In (f ), however, illa serves to announce a new episode
that follows.
(a) Nunc domum ibo atque ex uxore hanc rem pergam exquirere, /{quis fuerit
quem propter corpus suom stupri compleverit}.
(‘Now I’ll go home and continue questioning my wife about this matter, who it was
she filled her body with shame for.’ Pl. Am. 1015–16)
(b) Atque etiam hoc praeceptum officii diligenter tenendum est, {ne quem
umquam innocentem iudicio capitis arcessas}.
(‘Again, the following rule of duty is to be carefully observed: never prefer a capital
charge against any person who may be innocent.’ Cic. Off. 2.51)
(c) . . . cum subest ille timor {ea neglecta ne dignitatem quidem posse retineri}.
(‘ . . . when there is a covert fear lest if expediency be neglected worth will also have to
be abandoned.’ Cic. de Orat. 2.334)
(d) Accusat M. Cato qui . . . ea condicione nobis erat in hac civitate natus {ut eius
opes, ut ingenium praesidio multis . . . esse deberet}.
(‘He is accused by Marcus Cato . . . whose birth had given him a position with us in
Rome from which he should direct his resources and native talent to the protection of
many . . . ’ Cic. Mur. 56)
(e) Quae istaec audacia est, {te sic interdius / cum corolla ebrium incedere}?
(‘What impudence is that, that you strut around like this in daytime, with a garland
and drunk?’ Pl. Ps. 1298–9)
(f ) Illa vero eius cupiditas incredibilis est.
(‘There is another thing for which he had an incredible passion.’ Cic. Ver. 4.58)
The term ‘cataphoric’ is also applied to the use of determiners with autonomous
relative clauses. See § 18.15.
207
See Dziatzko and Hauler’s note ad loc.: ‘dem (früheren) milden Vater’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Tuus autem ille amicus (Sci’n quem dicam? De quo tu ad me scripsisti . . . ) . . .
(‘As to that friend of yours (you know whom I mean? The person of whom you write
to me . . . ) . . . ’ Cic. Att. 1.13.4)
Supplement:
Nam istos reges ceteros / memorare nolo, hominum mendicabula: / ego sum ille rex
Philippus. (Pl. Aul. 702–4); . . . aequo mendicus atque ille opulentissumus / censetur
censu ad Accheruntem mortuos. (Pl. Trin. 493–4); Ego ille agresti’, saevo’, tristi’,
parcu’, truculentu’, tenax / duxi uxorem. (Ter. Ad. 866–7); Et rex / et misellus ille
pauper amat habetque ignem intus acrem. (Var. Men. 210 Cèbe [= 205]); Ego me,
illum acerrimum regum hostem, ipsum cupiditatis regni crimen subiturum time-
rem? (Liv. 2.7.9); Hoc est illud tempus quo Cicero insito amore Pompeianarum
partium Caesarem laudandum et tollendum censebat, cum aliud diceret, aliud
intellegi vellet. (Vell. 2.62.6); Non aliud Codrum illum ceteris imperatoribus exem-
plum dedit . . . (Sen. Con. 8.4.med.);208 Procumbo ad genua vestra, iudices, ille con-
tumax qui cum vapularem non rogavi. (Sen. Con. 9.4.1); Hoc est autem medium illud
genus nec dimittendi divitem nec accusandi. (Sen. Con. 10.1.10); Timeo nescioquem
illum qui patrem meum occidit. (Sen. Con. 10.1.12); Nullum ornamentum principis
fastigio dignius pulchriusque est quam illa corona ob cives servatos, non hostilia
arma detracta victis, non currus barbarum sanguine cruenti, non parta bello spolia.
(Sen. Cl. 1.26.5); Hinc felix illa Campania . . . (Plin. Nat. 3.60); Adeoque tempestivam
admirationem intulit ut ille proscriptor animus, modo et contumelia furens, non
aliud in censu magis ex fortuna sua duceret. (Plin. Nat. 7.56); Monstraverunt etiam
locum ubi factus est vitulus ille. (Pereg. 5.3);
Exophoric use of hic is rare. Ex. (c) is regarded as such.209 Huius tyranni refers to
Caesar. But perhaps huius can also be attributed to the contrastive parallel phrase
reliquorum . . . tyrannorum: ‘the tyrant of our days’. An early instance is (d).
(c) Nec vero huius tyranni solum, quem armis oppressa pertulit civitas paretque
cum maxime mortuo, interitus declarat quantum odium hominum valet ad
pestem, sed reliquorum similes exitus tyrannorum, quorum haud fere quis-
quam talem interitum effugit.
(‘The death of this tyrant, whose yoke the state endured under the constraint of
armed force and whom it still obeys more humbly than ever, though he is dead,
illustrates the deadly effects of popular hatred; and the same lesson is taught by the
similar fate of all other despots, of whom practically no one has ever escaped such a
death.’ Cic. Off. 2.23)
(d) Deleo omnis de(h)inc ex animo mulieres. / Taedet cotidianarum harum
formarum.
(‘From now on I banish all other women from my mind. I’ve had enough of these
everyday beauties.’ Ter. Eu. 296–7)
208
For ille in Seneca Rhetor, see Pinkster (2005c). See also Dupraz (2012), who uses the term
‘anamnestique’.
209
See TLL s.v. 2723.51ff.: ‘der bewußte’. See also Joffre (2012b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
‘Cenabo’ inquit ‘apud te’, huic lusco familiari meo, C. Sextio ‘uni enim locum esse
video.’ (Cic. de Orat. 2.246); Tum Catulus ‘erras’, inquit ‘Crasse, si aut me aut horum
quemquam putas a te haec opera cotidiana et pervagata expectare’. (Cic. de Orat.
3.188—NB: see also ex. (g)); Orpheum poetam docet Aristoteles numquam fuisse, et
hoc Orphicum carmen Pythagorei ferunt cuiusdam fuisse Cerconis. (Cic. N.D. 1.107);
In potu autem quibus serrati dentes lambunt, et mures hi vulgares, quamvis ex alio
genere sint. (Plin. Nat. 10.201); Panis hic ipse quo vivitur innumeras paene continet
medicinas. (Plin. Nat. 22.138); Solebat dicere Fabianus, non ex his cathedraris
philosophis sed ex veris et antiquis, contra adfectus impetu, non suptilitate
pugnandum . . . (Sen. Dial. 10.10.1);
Iste is used in a ‘derogatory’ sense in instances like (e)–(g).210 But in some instances
the element of contempt is not so obvious.
(e) Verum ita sunt <morati> isti nostri divites. / Si quid bene facias, levior pluma
est gratia. / Si quid peccatum est, plumbeas iras gerunt.
(‘But that’s the way our rich men are: if you do them a good turn, the thanks is lighter
than a feather; if you make a mistake, their anger is as heavy as lead.’ Pl. Poen.
811–13)
(f ) Hominem conmonstrarier / mihi istum volo aut ubi habitet demonstrarier. /
# Nemp’ Phormionem? # Istum patronum mulieris.
(‘I want you to point the fellow out to me or show me where he lives. # You mean
Phormio? # The fellow who championed the girl.’ Ter. Ph. 305–7)
(g) ‘At enim vereor’, inquit Crassus ‘ne haec aut difficiliora istis ad persequen-
dum esse videantur aut, quia non traduntur in volgari ista disciplina, nos ea
maiora ac difficiliora videri velle videamur.
(‘ “O, but in fact,” said Crassus, “I am nervous lest either these doctrines may seem to
our young friends to be too difficult to carry out in practice, or we may ourselves be
thought to want them to seem grander and more difficult because they are not taught
as part of that common curriculum.” ’ Cic. de Orat. 3.188)
Indefiniteness of nouns and noun phrases is normally not marked. Just as the
demonstrative and anaphoric determiners used with definite NPs add their
own meaning to the NP, the indefinite determiners each contribute their own
210
See TLL s.v. iste 500.52ff; OLD s.v. iste § 3.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
meaning.211 A good illustration of this is (a). All three NPs are indefinite, but whereas
both quispiam and aliquis limit the number of indefinite candidates, medicus without
a determiner does not: in fact, there are multi ‘many’ to whom the content of the
clause might apply.
(a) Neque enim si de rusticis rebus agricola quispiam aut etiam, id quod multi,
medicus de morbis aut si de pingendo pictor aliquis diserte scripserit aut
dixerit, idcirco illius artis putanda est eloquentia.
(‘For even though some farmer may have written or spoken with address upon
country matters or perhaps a medical man upon pathology, as many have done, or a
painter upon painting, it does not therefore follow that eloquence belongs to the
particular art.’ Cic. de Orat. 2.38)
Since the indefinite determiners have their own meaning, it is to be expected that their
distribution varies with the type of text (and possibly also over time). The relative
frequency of the five indefinite determiners/pronouns in five authors can be seen in
Table 11.10.212 Qui(s) is much more often used as a pronoun than as a determiner,
certainly in Plautus. For aliqui(s), use of the pronoun predominates in Plautus, but this is
hardly the case in Cicero. Quidam is much more often used as a determiner, quisquam
overwhelmingly as a pronoun. With quispiam pronominal use is much more frequent.
Nescio qui(s) (and nescioqui(s)) ‘one or other’ (lit.: I don’t know who) also functions
as an indefinite determiner/pronoun.213
Appendix: There are instances (especially in Early Latin and in Late Latin) with more
than one indefinite determiner/pronoun in a clause, as in (b).214 They are best
regarded as pleonastic. The use of aliquam after interrogative quam in (c) and
quisquam after the interrogative pronoun quis in (d) are different: aliquam reflects
the wording of the order Epidicus has received (see l. 315); quisquam adds its usual
emphatic meaning.215
211
For a discussion of the indefinite determiners/pronouns and references to other publications, see
Bertocchi, Maraldi, and Orlandini (2010). For the ‘additional’ meaning of the indefinite determiners, see
Meisterfeld (2000: 315–19).
212
Source: Codoñer (1968: 23), who, however, for Plautus gives other data at p. 11. Her sources are the
usual lexica.
213 214
See OLD s.v. nescio §§ 5 and 6. Instances with aliqui in TLL s.v. 1612.16ff.
215
See Löfstedt (1933: II.191–4).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Persectari hic volo, / Sceledre, nos nostri an alieni simus, ne dum quispiam /
nos vicinorum imprudentis aliquis immutaverit.
(‘Sceledrus, I want to find out if we are ourselves or other people, in
case one of our neighbours has changed us without our knowledge.’ Pl.
Mil. 430–2)
(c) Sed me una turbat res ratioque, Apoecidi / quam ostendam fidicinam
aliquam conducticiam.
(‘But one matter and concern troubles me, which is the “some lyre-girl” for
hire I can show Apoecides.’ Pl. Epid. 312–13)
(d) Quis me Athenis nunc magis quisquam est homo quoi di sint propitii?
(‘What man is there in Athens now who the gods are more well disposed to
than me?’ Pl. Aul. 810)
216
See OLD s.v qui1 § 24, s.v. quis2 § 4. 217
See Bertocchi and Maraldi (2005).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) Quid in deserto quo loco (sc. faciet is homo), nactus quem multo auro
spoliare possit imbecillum atque solum?
(‘What will he do if, in some desolate spot, he meets a helpless man, unattended,
whom he can rob of a fortune?’ Cic. Leg. 1.41)
Appendix: In the four instances of malum quod in Plautus and Terence, as in (f ),
quod is understood as a connective particle ‘as to which’, as a relative pronoun, or as
an indefinite determiner by the various commentators. Since it is an isolated idiom
there is no way to decide.
218
For examples of aliqui(s) in negative sentences and after sine, see TLL s.v. 1612.56ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(f ) In eum locum res deducta est ut, nisi qui deus vel casus aliquis subvenerit,
salvi esse nequeamus.
(‘We have reached the point when we cannot survive unless some God or accident
comes to our rescue.’ Cic. Fam. 16.12.1)
In certain contexts aliqui(s) may be interpreted as signalling the opposite of some-
thing else. In ex. (g) aliquam follows two other precise alternatives. The translation
‘some other’ is contextually defined. See also (h). Normally to express ‘some other’
the combination ‘aliquis alius’ (and quis alius) must be used. In other contexts aliqui(s)
may be interpreted as signalling that the existence of the entity is valued as important
or of only moderate importance, as in (i) and (j),219 respectively. With plural NPs or
collective NPs it is interpreted as ‘a few’, as in (k). Note the possibility of combining this
quantitative use with the demonstrative determiner hic in (l). Aliqui(s) can be com-
bined with a numeral, meaning ‘approximately’, as in (m). Combinations with unus
are relatively common, as in (n).220
(g) Hinc prohibitus non ad Caesarem ne iratus, non domum ne iners, non
aliquam in regionem ne condemnare causam illam quam secutus esset
videretur: in Macedoniam ad Cn. Pompei castra venit . . .
(‘Debarred thence, he goes not to Caesar, lest he should be thought resentful,
not homewards, lest he should be thought apathetic, not to some other
country, lest he should be thought to condemn the cause he had followed; he
goes to Pompeius’ camp in Macedonia . . . ’ Cic. Lig. 27)
(h) Plura enim multo homines iudicant odio aut amore aut cupiditate aut
iracundia aut dolore aut laetitia aut spe aut timore aut errore aut aliqua
permotione mentis quam veritate aut praescripto aut iuris norma aliqua aut
iudicii formula aut legibus.
(‘For men decide far more problems by hate, or love, or lust, or rage, or
sorrow, or joy, or hope, or fear, or illusion, or some other inward emotion,
than by reality, or authority, or any legal standard, or judicial precedent, or
statute.’ Cic. de Orat. 2.178)
(i) . . . quid sentias, quaerimus, existimesne artem aliquam esse dicendi?
(‘ . . . we ask you to tell us your opinion—whether you hold that there is any
such thing as an “art” of oratory?’ Cic. de Orat. 1.102)
(j) Ipse etiam Fufidius in aliquo patronorum numero fuit.
(‘This Fufidius too had some place in the ranks of advocates.’ Cic. Brut. 113)
(k) Saltem si mihi / mulierculae essent salvae, spes aliquae forent.
(‘If at least my women were safe, I’d have some hopes.’ Pl. Rud. 552–3)
(l) Immo amabo ut hos dies aliquos sinas / eum esse apud me.
(‘No, I kindly ask you to let him be with me for the next few days.’ Pl. Truc. 872–3)
219
For more examples, see OLD s.v. aliqui1 § 4; K.-St.: I.640–2.
220
For more examples, see TLL s.v. aliquis 1612.31ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(o) . . . cum esses interrogatus quid sentires de consulatu meo, gravis auctor,
Calatinus credo aliquis aut Africanus aut Maximus et non Caesoninus
Semiplacentinus Calventius, respondes . . .
(‘ . . . when asked for your views as to my consulship you, with a sage
sententious air—another Calatinus, one would have thought, an Africanus
or a Maximus, instead of a Caesoninus Semiplacentinus Calventius—you
made answer . . . ’ Cic. Pis. 14)
221
For more examples, see OLD s.v. aliqui1 § 1a, TLL s.v. aliquis 1612.9ff., Hoppe (1903: 105 (= 1985:
197)).
222
For the history, see Sz.: 196; for Apuleius, see van Mal-Maeder (1994: 223–4); for examples, see
OLD s.v. quispiam1.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Vix finieram et ilico me perducit ad domum quampiam, cuius ipsis foribus
obseptis per quandam brevem posticulam intro vocat me et conclave quod-
dam obseratis luminibus umbrosum <intrans> demonstrat matronam . . .
(‘I had scarcely finished talking when he led me quickly off to a house whose entrance
was bolted shut. He invited me in through a tiny back door, and when we had entered
a darkened room with tightly barred windows, he pointed out a woman.’ Apul. Met.
2.23.5)
(e) Erat tum nobiscum in eodem ambulacro homo quispiam sane doctus.
(‘There was with us at the time in the same promenade a man of considerable
learning.’ Gel. 3.1.7)
223
So OLD s.v. quisquam § 6, also for more examples.
224
For the difference, see Wagenvoort (1922), who also notes that quisquam is not found in nisi clauses.
225
For more examples, see OLD s.v. ullus.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
226
The term ‘specifying’ is used to avoid confusion with ‘specific’, which is used as the opposite of
generic (see § 11.101).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
the case: in (c) the matrona sarcastically tells her husband Menaechmus that the man
whose identity he wants to know is ‘a certain Menaechmus’ (himself). There are no
restrictions on the type of entities that quidam can modify. Exx. (a) and (b) have a
count noun. Note in (a) the use of suo. In (b), Mercury refers to Sosia, who thinks
Mercury has not seen him. Proper names usually have unique reference, which does
not mean that they are identifiable for everybody. Quidam signals (correctly or not)
that the addressee does not know the person referred to and/or that details concern-
ing the identity do not matter, as in (d), where the relative clause tells the reader/
hearer that she was quite well known.227 In (e), quaedam determines res, which refers
to a complex situation, a third-order entity.
(a) Hac nocte in somnis mea soror gemina est germana visa / venisse Athenis in
Ephesum cum suo amatore quodam.
(‘Last night I dreamed that my twin sister had come from Athens to Ephesus with
some lover of hers.’ Pl. Mil. 383–4)
(b) Olet homo quidam malo suo.
(‘Someone’s smelling here, and it won’t do him any good.’ Pl. Am. 321)
(c) Quis eam surrupuit? # Pol istuc ille scit qui illam apstulit. / # Quis is homo
est? # Menaechmus quidam. # Edepol factum nequiter. / Quis is Menaech-
mu’st? # Tu istic, inquam.
(‘Who stole it? # Well, the man who took it away knows that. # Who is it? # A certain
Menaechmus. # Honestly, an appalling act. Who is that Menaechmus? # You there,
I’m telling you.’ Pl. Men. 649–51)
(d) <Erat> Pipa quaedam, uxor Aeschrionis Syracusani, de qua muliere plurimi
versus qui in istius cupiditatem facti sunt tota Sicilia percelebrantur.
(‘Among these ladies was some Pipa, the wife of Aeschrio of Syracuse, a woman
about whom a number of lampoons in verse were composed, in allusion to Verres’
passion for her, which are in circulation throughout Sicily.’ Cic. Ver. 5.81)
(e) Nam est res quaedam quam occultabam tibi dicere. Nunc eam narrabo, /
und’ tu pergrande lucrum facias.
(‘There’s a certain matter I was keeping to myself. Now I’ll tell you about it, so that
you can make a great profit from it.’ Pl. Per. 493–4)
227
The compatibility of quidam with uniquely referring proper names was a reason for Serbat (1984)
to reject the element ‘known to the speaker’ as part of the meaning of quidam. See also Orlandini
(1995: 160).
228
For more examples, see OLD s.v. quidam § 3 and § 4a.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
229
For more examples, see OLD s.v. § 4b.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
from Livy onwards. Another contextual interpretation is shown in (p), where the
sequence quosdam . . . alios is more or less equivalent to alios . . . alios.
(m) Nihil me clamor iste commovet sed consolatur, cum indicat esse quosdam
civis imperitos sed non multos.
(‘The outcry that I hear does not perturb me, nay, it consoles me; for it shows there
are some uninstructed citizens but not many . . . ’ Cic. Rab. Perd. 18)
(n) (sc. Caesar) . . . deductis quibusdam cohortibus ex praesidiis eodem venit.
(‘Caesar . . . brought down some cohorts from the garrisons and came to the same
place.’ Caes. Civ. 3.65.3)
(o) (sc. Hercules) . . . agere porro armentum occepit. Inde cum actae boves quae-
dam ad desiderium, ut fit, relictarum mugissent, reddita inclusarum ex
spelunca boum vox Herculem convertit.
(‘Hercules . . . made ready to drive his herd. As the cattle were being driven off, some
of them lowed, as usually happens, missing those which had been left behind. They
were answered with a low by the cattle shut up in the cave, and this made Hercules
turn back.’ Liv. 1.7.6–7)
(p) . . . militesque et centuriones quosdam puniit, alios praemiis adfecit . . .
(‘ . . . and he also punished some soldiers and centurions, others he rewarded . . . ’ Tac.
Hist. 4.50.3)
230
For the use of quidam in Apuleius, see van Mal-Maeder (1994); for its use in presentational
sentences, see Rosén (1998).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In Late Latin quidam is sometimes used in contexts where the other indefinite
determiners would have been used in Classical Latin. This is a sign of the decline of
the indefinite system as a whole.231
Appendix: Certus in the meaning ‘a certain’, which was to take the place of quidam in
Romance, is used from Cicero onwards, as in (t).232
(t) Sed mihi videntur certi homines . . . vos universos qui me antetuleritis nobi-
litati vituperaturi.
(‘But there appear to me to be certain men who . . . will be ready to reproach
you all for having preferred me to my noble competitors.’ Cic. Agr. 2.6)
231
For later developments, see Bertocchi, Maraldi, and Orlandini (2010: 67–71).
232
For more examples, see OLD s.v. certus § 3; TLL s.v. certus 904.48ff.
233
For more examples, see OLD s.v. quicumque § 8; quilibet11; quisquis § 9; quivis1. For discussion, see
Bertocchi, Maraldi, and Orlandini (2010: 95–114).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The indefinite use of quicumque and quisquis extends in the course of Latin. Its origin
can be traced to their use as relative pronouns in relative clauses with something like
est or fieri potest.234 For their use as generalizing relative pronouns and determiners,
see §§ 18.31–18.33.
Many instances of quivis can still be understood as relative clauses, with the
relative governed by the verb volo. This is the case in (e) (translated as an indefinite
determiner). With other forms of volo ambiguity is excluded, as in (f ).
234
See K.-St.: II.199.
235
For the distinction between alteruter and utervis, see Bertocchi, Maraldi, and Orlandini (2010:
148–52).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Nam si parti utrilibet omnino alteram detrahas, natura etiam sine doctrina
multum valebit, doctrina nulla esse sine natura poterit.
(‘For if you isolate one of the pair from the other altogether, nature will be able to do
a lot without teaching, but without nature there can be no teaching.’ Quint. Inst.
2.19.2)
(d) . . . utriusvis rei me aut adiutorem velim esse aut certe non expertem.
(‘ . . . I should in both of these cases wish to help or at any rate not to be on the
outside.’ Cic. Att. 7.3.2)
236
For the idea, see Graur (1969). Criticism in van Mal-Maeder (1994), Rosén (1998: 728–31), and
Bertocchi, Maraldi, and Orlandini (2010: 67–71), where references to other studies can be found as well.
237
So Wehr (1984: 44).
238
For the interpretation of unus here as ‘articloïde’, see Väänänen (1987: 57). A comparative study is
Givón (1981).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
From Early Latin onwards there are a few other non-numerical uses of unus, (i) with
a superlative ‘one in particular’, ‘one above all’, as in (e) and (f ), (ii) ‘an ordinary’ as
in (g)—a non-referring noun phrase—and maybe (iii) ‘some’, as in (h).239 For its
meaning ‘only’, see § 11.33.240
(e) Sed est huic unus servos violentissumus / qui . . . clamore apsterret abigit.
(‘But he has one most savage slave who . . . frightens us and drives us away
with his shouting.’ Pl. Truc. 250–2)
(f ) Interea inter mulieres / quae ibi aderant forte unam aspicio adulescentulam /
forma—# Bona fortasse.
(‘Meanwhile among the women who were present I caught sight of one
young lass whose looks were—# Not bad, perhaps.’ Ter. An. 117–19—NB:
the speaker is interrupted)
(g) . . . de hoc uno minime est facile praecipere non mihi modo, qui sicut unus
pater familias his de rebus loquor, sed etiam ipsi illi Roscio . . .
(‘ . . . it is especially difficult to give instructions on this topic—not only for
me who am talking about these things like a head of a household would, but
even for Roscius himself . . . ’ Cic. de Orat. 1.132)
(h) Sunt autem etiam fontes uti vino mixti, quemadmodum est unus Paphlago-
niae, ex quo eam aquam sine vino potantes fiunt temulenti.
(‘There are also springs, as it were, mixed with wine, such as one in Paphla-
gonia, and persons drinking it without wine become drunk.’ Vitr. 8.3.20)
239
For these three uses, see OLD s.v. unus §§ 8b, 10, 11, respectively. For the development, see Pinkster
(1988), Meisterfeld (2000: 316–29), and de la Villa (2010: 226–33). For examples, see Wehr (1984: 39–46).
240
See also OLD s.v. unus § 7.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Reliqua (sc. animalia) ut libitum est gradiuntur, leo tantum et camelus
pedatim.
(‘Whereas the other animals walk as they like, only the lion and the camel pace with
one foot after the other.’ Plin. Nat. 11.253)
(c) Abies e cunctis amplissima est et femina etiam prolixior . . .
(‘The biggest of the entire group is the fir, the female being even taller than the
male . . . ’ Plin. Nat. 16.48)
(d) Atque, ut nunc sunt maledicentes homines, uxori meae / mihique obiectent
lenocinium facere.
(‘And, given what crooked gossipers people are nowadays, they would disapprove of
my wife and myself on the grounds that we were keeping a brothel.’ Pl. Mer. 410–1)
(e) Cornibus tauri, apri dentibus, cursu leones, aliae fuga se, aliae occultatione
tutantur, atramenti effusione saepiae, torpore torpedines, multae etiam
insectantes odoris intolerabili foeditate depellunt.
(‘Bulls protect themselves with their horns, boars with their tusks, lions by rushing,
some species by flight, some by hiding, the cuttlefish by emitting an inky fluid, the
stingray by causing cramp, and also a number of creatures drive away their pursuers
by their insufferably disgusting odour.’ Cic. N.D. 2.127)
(f ) Quaedam in montanis prolixiora nascuntur ac firmiora propter frigus, ut
abietes ac sappini . . .
(‘Certain trees, such as the fir and the pine, flourish best and are sturdiest in the
mountains on account of the cold climate . . . ’ Var. R. 1.6.4)
(g) Egon ab lenone quicquam / mancupio accipiam, quibus sui nihil est nisi una
lingua / qui abiurant si quid creditum est?
(‘I should take anything formally from a pimp? They have nothing of their own
except for the bare tongue with which they swear off if anything’s been entrusted to
them.’ Pl. Cur. 494–6)
Generic use of adjectives and participles is equally possible. Examples are (h)–(k).241
241
For further examples, see K.-St.: I.222–5.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Nouns and noun phrases are used in a non-referring (that is: non-specific and non-
generic) way in a variety of constructions to describe the property or quality of an
entity or a state of affairs. Illustrations are (a)–(g). Nauta in (a), a subject comple-
ment, describes the subject in terms of his social position or job. Lenonem in (b), an
object complement, is comparable. Urbis proditor in (c) is a non-restrictive appos-
ition. Bono animo in (d) is a subject complement (a so-called ablativus qualitatis).
Ornamento, praesidio, and detrimento in (e) are also subject complements (so-called
predicative datives). Summi laboris and summae industriae in (f ) are both attributes
(of hominum), instances of the so-called genetivus qualitatis. In (g), the manner
adjunct incredibili celeritate describes a property of the action involved. As the
examples show, there is no formal marking of non-reference in Latin.
(a) Nam ill’ qui lanam ob oculum habebat <laevom> nauta non erat.
(‘That chap who had a patch on his left eye was not a sailor.’ Pl. Mil. 1430)
(b) . . . numquam hercle quisquam me lenonem dixerit, / si te non ludos pessi-
mos dimisero.
(‘ . . . no one shall ever call me a pimp if I don’t make great sport of you before I let
you go.’ Pl. Rud. 790–1)
(c) Ei erat hospes par sui, Siculus senex / scelestus, Agrigentinus, urbis proditor.
(‘He had as a guest the counterpart of himself, an old man from Sicily, a wicked
fellow, from Agrigentum, the type that would betray his city.’ Pl. Rud. 49–50)
(d) Bono animo es.
(‘Be of good courage.’ Pl. Mil. 1143)
(e) Amicitiam populi Romani sibi ornamento et praesidio, non detrimento esse
oportere . . .
(‘The friendship of the Roman people ought to be a distinction and a security to him,
not a hindrance . . . ’ Caes. Gal. 1.44.5)
(f ) Cognoscetis Entellinorum, hominum summi laboris summaeque indus-
triae, dolorem et iniurias.
(‘You shall learn of the wrongs and sufferings of the strenuous and hard-working
people of Entella.’ Cic. Ver. 3.103)
(g) . . . incredibili celeritate ad flumen decucurrerunt . . .
(‘ . . . with incredible speed the enemy rushed down to the river . . . ’ Caes. Gal. 2.19.8)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
11.120 Pronouns
Pronouns derive their name from the fact that they can function like nouns, as
argument or satellite at the clause level, as well as as attribute at the noun phrase
level. As shown in §§ 11.6–11.7, most pronouns cannot function as head of a noun
phrase. Most of them can be singular or plural, just like nouns (and, unlike nouns,
they may be characterized for gender as well). For personal pronouns the situation is
more complicated (see § 11.121). With the exception of personal pronouns, most
lexemes involved function also as determiner of a noun or noun phrase (see §§ 11.25–
11.28). For general discussion of the term ‘pronoun’, see § 3.8.
The first and second person pronouns ego ‘I’, tu ‘you’, for the singular, and nos ‘we’,
vos ‘you’, for the plural, can even less than most other pronouns be regarded as
equivalent to nouns as far as their functioning at the clause level is concerned.242 For
their use as subject (in the nominative) see §§ 9.2–9.3. Also, nos is not really the plural
of ego, nor is vos the plural of tu: nos includes the speaker ego and vos includes the
addressee tu. The same holds for the first and second person verb endings.
Latin has no third person pronoun (apart from the reflexive pronoun). When in a
particular situation reference has to be made to a third person, the demonstrative
pronouns hic, ille, and iste serve as a substitute, each with its specific meaning (see
§§ 11.134–11.141). For the conditions of use of explicit third person subjects, see
§ 9.9. Typical examples of ille functioning as subject are (a) and (b). In (a), there is
contrast between ego and ille, supported by a gesture ‘over there’. In (b), Cicero
narrates the discussion between Atticus, Brutus, and himself; ille marks topic change.
(a) Quibus ego si me restitisse dicam, nimium mihi sumam et non sim ferendus:
ille, ille Iuppiter restitit. Ille Capitolium, ille haec templa, ille cunctam
urbem, ille vos omnis salvos esse voluit.
(‘If I were to say that I foiled them, I should be taking too much credit for myself—an
intolerable presumption. It was Jupiter, the mighty Jupiter, who foiled them; it was
Jupiter who secured the salvation of the Capitol, of these temples, of the whole city
and of you all.’ Cic. Catil. 3.22)
(b) Tum Brutus: ‘ . . . Attici litterae . . . ’ # ‘Istae vero’, inquam, ‘Brute . . . ’ # ‘Salutem?’
inquit ille. ‘Quodnam . . . ’ # ‘An mihi potuit’, inquam, ‘esse . . . ’ # Tum ille:
‘Nempe . . . ’ # ‘Istum ipsum’, inquam, ‘Brute . . . ’ # Tum Atticus ‘ . . . ’ # ‘Ille vero
et nova’, inquam, ‘ . . . ’ # Tum ille ‘ . . . ’ # ‘Mihi quoque’, inquit Brutus, ‘ . . . ’
(‘Brutus replied, “ . . . writing of Atticus . . . ” # “Yes, Brutus,” I replied, “it . . . ” #
“Salvation?” he said. “What . . . ” # I asked, “Could any . . . ” # “Ah yes,” he replied . . .
242
See Fugier (1974).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
# I said, “Yes, Brutus, that is . . . ” # Then Atticus said . . . # “Much indeed that was
novel,” I replied“ . . . ”# He replied, “ . . . ” # “I too,” said Brutus, “ . . . ” ’ Cic. Brut. 13–17)
In a continuous text the anaphoric pronoun is functions as a substitute for the third
person pronoun. The demonstrative pronouns mentioned are used as well (see
§ 11.136). Two examples of is are (c) and (d).
(c) Haec urbs est Thebae. In illisce habitat aedibus / Amphitruo, natus Argis ex
Argo patre / . . . Is prius quam hinc abiit ipsemet in exercitum, / gravidam
Alcumenam fecit uxorem suam.
(‘This city is Thebes. In that house there (points) lives Amphitruo, born in Argos of
an Argive father . . . Before he himself went away to the army, he made his wife
Alcumena pregnant.’ Pl. Am. 97–103)
(d) Salve. Sed quem quaeritas? / # Bacchidem. # Utram ergo? # Nil scio nisi
Bacchidem. / Paucis: me misit miles ad eam Cleomachus . . .
(‘Hello. But who are you looking for? # Bacchis. # Which one now? # I only know
Bacchis. In short: the soldier Cleomachus has sent me to her . . . ’ Pl. Bac. 587–9)
243
For nos instead of ego in Cicero, see Pieroni (2010b). For references, see Sz.: 19–20.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(Cic. Orat. 105); Haec properantes scripsimus. (Cic. Att. 4.4.1); Dum stabat regno
incolumis regumque vigebat / conciliis, et nos aliquod nomenque decusque / gessimus.
(Verg. A. 2.88–90—NB: the person speaking is Sinon);
11.123 The use of the first and second person plural instead of the
singular by and for persons of a high social standing
The first instances of the use of the so-called PLURAL OF MAJESTY (pluralis maiestatis) by
dignitaries to underline their authority are found in the third century AD.244 This use
of the first person plural (personal pronouns, possessive adjectives, and finite verb
endings) becomes quite common in the fifth century. An example is (a). Its counter-
part, the use of the plural as an expression of politeness (pluralis reverentiae) becomes
common from Symmachus onwards. An example is (b).245
(a) Nam pecunias navibus vectas non omnes iudicamus mercatores debere promere,
quippe in usu tantum publico pecunias constitutas permittimus convehi . . .
(‘For We judge that merchants must not export all kinds of money carried in their
ships; in fact, We permit only money established in public use to be so conveyed.’
Cod. Theod. 9.23.1.2)
(b) Quod apud germanum clementiae vestrae divum principem non sileremus . . .
(‘If the Roman world were still enjoying the presence of your Clemencies’ brother,
I should not keep silence . . . ’ Symm. Rel. 4.1—tr. Barrow)
Supplement: Hinc vos munere salutationis impertio doque nuntium propere nos diis
volentibus esse redituros. Fors fuat huiusce promissi. Vestra tamen indulgentia
affatum saepe tribuat, quasi diutius abfuturis. Vale. (Symm. Ep. 1.3.5—NB: addressee
is his father; note also nos); Sed et vobis, gloriosi principes, cum sitis absolute
mirabiles, aliquid tamen additur, cum vos omnia regna venerantur. (Cassiod. Var.
10.19.3—NB: Theodahadus addressing Iustinianus);
When the subject of a finite clause and another constituent of that clause have the
same reference, a reflexive pronoun or adjective must be used. In the case of the first
and second person (singular and plural) the same forms are used for non-
coreferential and coreferential constituents. This is shown for the first person singular
in (a) and (b), respectively. In (a), pater and me refer to different persons (hence the
244
The fullest collection of examples can be found in Sasse (1889). For further references, see Sz.: 20–1.
245
For the use of second person plural vos instead of tu as an expression of courtesy in the fifth century
AD and later, see Haverling (1995). For a historical survey, see Ehrismann (1900–1). For further references,
see Sz.: 20–1.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
different indexes x and y). In (b), ego and me refer to the same person. In both cases
the same form, me, is used. However, in the case of the third person, different forms
must be used for non-coreferentiality and coreferentiality, as is shown in (c) and (d),
respectively (NB: eum is just one of the possibilities). Somewhat similar is the use of
possessive adjectives, as in (e). The following sections are exclusively devoted to the
third person reflexive pronoun and the third person reflexive possessive adjective and
the conditions of their use. In § 5.23 a distinction is made between ‘true reflexive’ and
other types of reflexive. ‘True reflexive’ refers to the use of a reflexive pronoun in a
position where another noun phrase is possible as well, as can be seen in (e), where the
reflexive pronoun is coordinated with a noun phrase. Only ‘true reflexive’ instances
are taken into account.
(a) Paterx huc mey misit ad vos oratum meus . . .
(‘My father’s sent me here to plead with you . . . ’ Pl. Am. 20)
(b) . . . egox mex deorsum duco de arbore . . .
(‘ . . . I climbed down from the tree . . . ’ Pl. Aul. 708)
(c) Nam eumy paterx eiusy subegit.
(‘For his father forced him.’ Pl. Cist. 101)
(d) Nam senexx / rus abdidit sex, huc raro in urbem commeat.
(‘The old man has buried himself in the country and rarely comes here into town.’
Ter. Hec. 174–5)
(e) Mulier quaex sex suamxque aetatem spernit speculo ei usus est.
(‘A woman who is dissatisfied with herself and her age, she has need of a mirror.’ Pl.
Mos. 250)
In all situations in which the reflexive third person pronoun or the reflexive possessive
adjective has the same reference as the subject of that clause, this is called DIRECT use of
the reflexive pronoun and the reflexive possessive adjective. Further details are given
in § 11.126.
However, in subordinate clauses the third person reflexive pronoun and possessive
adjective must be or are usually used when there is coreference with the subject
(sometimes another constituent) in the governing clause. This is called the INDIRECT
use of the reflexive pronoun and the reflexive possessive adjective. It is illustrated by
(d) and (e). In (d), se is the subject of the accusative and infinitive clause se dictatores
aut reges futuros. It is coreferential with consules, the subject of the governing clause
consules sperant. In (e), se, the subject of the ablative absolute clause se vivo, is
coreferential with Postumus, the subject of the main clause. More examples follow
in § 11.127. (There is also coreferentiality at the phrase level, which is illustrated in
§ 11.129.)246
246
For a very complete description of the use of the reflexive pronoun in Cicero, see Lebreton (1901a:
111–49). For Livy, see Riemann (1885: 115–57).
247
So Pieroni (2010a: 437–43), with references.
248
For statistics and discussion, see Sznajder (1981).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
11.126 The direct use of the reflexive third person personal pronoun
The reflexive third person personal pronoun is in simple clauses (whether main or
subordinate clauses) required for any constituent that is coreferential with the
grammatical subject of the clause. In addition to the examples given in the introduc-
tory section, (a) shows se as the object. In (b), sibi is a beneficiary satellite. In (c) and
(d), se is part of a prepositional phrase that functions as satellite. In (d), sibi is required
by the subject complement carus; in (e), sui is an adnominal argument of memoria, a
cause argument in its clause.
(a) Ita vorsipellem se facit quando lubet.
(‘He changes appearances like that whenever he wants.’ Pl. Am. 123)
(b) Illic homo a me sibi malam rem arcessit iumento suo.
(‘That man is asking me to give a thrashing to his beast of burden for him.’ Pl. Am. 327)
(c) Pro se quisque id quod quisque <et> potest et valet / edit . . .
(‘Each man inflicted for himself what he could . . . ’ Pl. Am. 231–2)
(d) . . . per se sibi quisque carus est.
(‘ . . . each man is dear to himself on his own account.’ Cic. Amic. 80)
(e) Niciasx te, ut debet, amat vehementerque tua suix memoria delectatur.
(‘Nicias loves you as he ought, and is delighted by your remembrance of him.’ Cic.
Att. 13.1.3)
The reflexive can also be coreferential with a non-subject constituent in its clause.
Examples are (j) and (k).
(j) Nam is est servos ipse, nec praeter sex umquam eix servos fuit.
(‘He himself is a slave and there never was a slave in his possession besides himself.’
Pl. Capt. 580)
(k) . . . ut . . . huius tamen diei vocem testem rei publicaex relinquerem meae
perpetuae erga sex voluntatis.
(‘ . . . so that I might leave the words I speak today as witnesses to the Republic of my
abiding loyalty to it.’ Cic. Phil. 1.10)
Supplement:
Quae mens eumx aut quorum consilia a tanta gloria, sibix vero etiam necessaria ac
salutari, avocarit . . . (Planc. Fam. 10.24.6); Has adversus copias spes omnis consis-
tebat Datamix in sex locique natura. (Nep. Dat. 8.3);
11.127 The indirect use of the reflexive third person personal pronoun
Pronouns
249
For the use of indirect reflexives in indirect discourse, see Fruyt (1987, 2002a) and Poirier (1989).
250
See Sz.: 175 and Pieroni (2010a: 437–43) with references. For ille, see TLL s.v. 350.81ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
More examples of the reflexive pronoun in autonomous relative clauses in an
argument position: Eum fecisse aiunt sibi quod faciundum fuit. (Pl. Poen. 956);
Mater quod suasit sua / adulescens mulier fecit. (Ter. Hec. 660–1); Dexo hic, quem
videtis, non quae publice Tyndaride, non quae privatim sibi eripuisti, sed unicum
miser abs te filium optimum atque innocentissimum flagitat. (Cic. Ver. 5.128); . . .
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
homines prius ipsum illud quod proprium sui iudici est audire desiderant. (Cic. de
Orat. 2.213); . . . Epaminondas, Thebanorum imperator, ei qui sibi ex lege praetor
successerat exercitum non tradidit . . . (Cic. Inv. 1.55); . . . nil rectum nisi quod placuit
sibi ducunt . . . (Hor. Ep. 2.1.83); Caesar . . . omnibus qui contra se arma tulerant
ignovit . . . (Vell. 2.56.1);
Appendix: For an example of a reflexive pronoun in a comparative clause, see (o).
(v) Nam paterx expectat aut me aut aliquem nuntium / qui hinc ad sex veniat.
(‘Your father is expecting either me or some other messenger who is coming to him
from here.’ Pl. Capt. 382–3)
(w) Dicitx capram quam dederam servandam sibix / suai uxoris dotem ambedisse
oppido.
(‘He said that the goat that I had given him to watch over had completely eaten up his
wife’s dowry.’ Pl. Mer. 238–9)
(x) . . . FACITOX·QUEI·PRO·SE·PRAES STAT·PRAEDES·PRAEDIAQUE·AD·IIII·VIR(UM)·DET·QUOD·
X X
SATIS / SIT·, QUAE·PEQUNIA·PUBLIC<A SA>CRA·RELIGIOSA·EIUS·MUNICIPI·AD·SE ·IN SUO ·MA-
GISTRATU / PERVENERIT·, EAM·PEQUNI<A>M·MUNICIPIO·TARENTINO· SALVAM·RECTE·ESSE·
FUTUR<A>M . . .
(‘[sc. the magistrate] is to see that whoever is standing surety for him furnish to a
IIIIvir [member of the Board of Four] sureties and securities as may be sufficient, that
whatever public, sacred, or religious money of that municipium shall have come to
him in his magistracy, that money will be duly secure for the municipium of
Tarentum . . . ’ CIL I2.590.9–11 (Tarentum, early I BC—tr. Crawford (ed.)))251
Supplement:
Hunc sibi ex animo scrupulum qui se dies noctesque stimulat ac pungit ut evellatis
(sc. Roscius) postulat . . . (Cic. S. Rosc. 6); Centum boves militibus dono dedit qui
secum in expeditione fuerant. (Liv. 7.37.3);
251
For the date of this inscription, see Crawford (1996: I.302).
252
As noted by K.-St.: I.601. For more instances of multiple reflexives in subordinate clauses in Cicero,
see Lebreton (1901a: 124–5). For ipse, see Fruyt (2010: 70–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(d) Modo magis Pausaniasx perturbatus orare coepit ne (sc. Argilius) enuntiarety
nec sex meritum de illoy optime proderety.
(‘Pausanias, still more disturbed, began to beg him not to report and betray one who
had always deserved well of him.’ Nep. Paus. 4.6)
Supplement:
An quisquam esse testis potest quem accusatorx sine cura interroget nec metuat ne
sibix aliquid quod ipsex nolit respondeat? (Cic. Flac. 22); Diviciacusx multis cum
lacrimis Caesarem complexus obsecrare coepit ne quid gravius in fratremy statueret:
scire sex illa esse vera, nec quemquam ex eo plus quam sex doloris capere, propterea
quod, cum ipsex gratia plurimum domi atque in reliqua Gallia, illey minimum
propter adulescentiam posset, per sex crevissety . . . (Cic. Gal. 1.20.1–2);
The reflexive pronoun can also be used at the phrase level, especially with participles,
as in (a) and (b). Its use can be explained in the same way as at the clause level. Ex.
(c) has a two-place verbal noun. In (c), suas is coreferential with D. Bruti, the agent of
adventu. Ex. (d) has the two-place adjective dignus functioning as attribute. Ex. (e) has
the two-place adjective functioning as a secondary predicate.
(a) Proelium fuit quale inter fidentes sibixmet ambo exercitusx . . .
(‘It was a battle such as takes place when both armies are confident . . . ’ Liv. 3.62.6)
(b) Habetis ducemx memorem vestri, oblitum suix . . .
(‘You have a leader mindful of you, heedless of himself . . . ’ Cic. Catil. 4.19)
(c) Quamvis enim tu magna et mihi iucunda scripseris de D. Brutix adventu ad
suasx legiones . . .
(‘For while your news about D. Brutus joining his troops is important and welcome . . . ’
Cic. Att. 14.13.2)
(d) . . . habeo dotem unde dem, / ut eamx in sex dignam condicionem collocem.
(‘ . . . I certainly have the means to give a dowry to his daughter so that I can give her
in a marriage worthy of her.’ Pl. Trin. 158–9)
(e) Non illumx miserum, ignarum casus suix, redeuntem a cena videtis . . . ?
(‘Do you not see that unfortunate man returning from supper, without suspicion of
the fate that awaits him . . . ?’ Cic. S. Rosc. 98)
(i) Examples of coreferentiality with the grammatical subject of the clause are
(a)–(d). In (e) and (f ) suus is coreferential with a ‘logical subject’.
(a) Nam ecastor numquam satis dedit suae quisquam amicae amator . . .
(‘No lover has ever given enough security to his girlfriend . . . ’ Pl. Truc. 239)
(b) Malus et nequam est homo qui nihili eri imperium sui servos facit . . .
(‘A slave who thinks nothing of his master’s command is bad and worthless . . . ’ Pl.
Ps. 1103)
(c) Ut Alexandrum regem videmus, qui cum interemisset Clitum, familiarem
suum, vix a se manus abstinuit.
(‘As for instance we see king Alexander did, who could scarcely keep his hands off
himself after he had killed his friend Clitus.’ Cic. Tusc. 4.79)
(d) (sc. Fadius) . . . a me diligitur propter summam suam humanitatem et
observantiam.
(‘ . . . Fadius is highly regarded by me for his good nature and attentive courtesy.’ Cic.
Fam. 15.14.1)
(e) . . . eosque qui secus quam decuit vixerunt peccatorum suorum tum maxume
paenitet.
(‘ . . . while those who have lived otherwise than as they should, feel, at such a time,
the keenest sorrow for their sins.’ Cic. Div. 1.63)
(f ) . . . cui possit exploratum esse de sua sanitate?
(‘ . . . who can be satisfied of his own sanity?’ Cic. Luc. 54)
Pronouns
(k) Me a portu praemisit (sc. Amphitruo) domum, ut haec nuntiem uxori suae /
ut gesserit rem publicam ductu, imperio, auspicio suo.
(‘He’s sent me ahead home from the harbour so that I could report to his wife how he
managed affairs of state through his leadership, command, and authority.’ Pl. Am.
195–6)
(l) Acuto homine nobis opus est et natura usuque callido, qui sagaciter perves-
tiget quid sui cives eique homines . . . cogitent . . .
(‘We require a man of sharpness, ingenious by nature and experience alike, who with
keen scent will track down what his fellow-citizens and those men . . . think.’ Cic. de
Orat. 1.223)
(m) Velatis manibus orant ignoscamus peccatum suom . . .
(‘With covered hands they asked us to forgive them their transgression.’ Pl. Am. 257)
(n) (sc. Flaccus) . . . metuens ne suum quoque exercitum sicut Hannibalis nimia
urbis amoenitas molliret . . .
(‘(Flaccus) . . . feared that the great charms of the city might weaken his army also, as
they had Hannibal’s.’ Liv. 27.3.2)
(o) Si scit et persuasus est, quid irascitur ei qui aliquid scripsit contra suam
voluntatem . . .
(‘If he knows and is persuaded of it, why is he angry with somebody for writing
something to his displeasure . . . ’ Caecin. Fam. 6.7.2)
(p) . . . quod ad perniciem suam fuerat cogitatum, id ad salutem convertit.
(‘ . . . thus he made the plot which had been devised for his ruin the means of his
safety.’ Nep. Dat. 6.8)
In addition to § 11.30 a few more examples are given below of suus indicating
coreferentiality with a different constituent at the clause level than the subject of
that clause or of a governing clause. Suus is especially used in this way when it is
emphatic, for example in an unexpected situation as in (v). But it is also often used
without emphasis, as exx. (w)–(z) show. Ex. (aa) is an illustration of the indirect use of
suus. Ex. (ab) shows the regular use of suus in an apposition related to a prepositional
phrase with cum, used as a comitative coordinator (on which see Chapter 19). Ex. (ac)
has the familiar combination suum cuique.
(v) Hunc (sc. Hannibalem) sui cives e civitate eiecerunt.
(‘His fellow-citizens drove him out.’ Cic. Sest. 142)
(w) (sc. Amphitruo) . . . regique Thebano Creoni regnum stabilivit suom.
(‘He has secured the kingship for the Theban king, Creon.’ Pl. Am. 194)
(x) Quem ego hominem, siquidem vivo, vita evolvam sua.
(‘As truly as I live, I’ll send him spinning out of his life.’ Pl. Men. 903)
(y) Huic puero qui est ei vita sua multo carior metuit . . .
(‘He fears for this boy, who is much dearer to him than his own life, . . . ’ Cic. Sul. 88)
(z) . . . omne imperium nostri penes singulos esse voluerunt, quorum ipsum
nomen vim suae potestatis indicat.
(‘ . . . our people have preferred that all the power should be granted to one man. And
this man’s title shows the character of his power.’ Cic. Rep. 1.63)
(aa) (sc. Teloboi) . . . respondent bello se et suos tutari posse, proinde uti / propere
<irent>, de suis finibus exercitus deducerent.
(‘They replied that they could protect themselves and their families by war; so our
men had better remove their troops from their territory quickly.’ Pl. Am. 214–15)
(ab) Dicaearchum vero cum Aristoxeno, aequali et condiscipulo suo, doctos sane
homines, omittamus.
(‘But as for Dicaearchus, along with his contemporary and fellow-pupil Aristoxenus,
in spite of their undoubted learning, let us ignore them.’ Cic. Tusc. 1.41)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
253
For the role of the speaker as a determining factor for the choice between the reflexive and the
anaphoric pronoun, see Bolkestein (1990: 90) and Bertocchi (1994). For the opposite position, see Calboli
(1997: 294–314).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Cui (sc. Catilinae) cum adfuit (sc. Torquatusx) post delatam ad eumx primam
illam coniurationem, indicavit se audisse aliquid, non credidisse.
(‘When he was helping him, after that first conspiracy was reported to him, he
indicated that he had heard something but did not believe it.’ Cic. Sul. 81)
(b) Haec cum maxime summa hominum frequentia testificaretur (sc. Apollo-
niusx), ut quivis intellegere posset eumx, quod pecuniam non dedisset,
idcirco illa tam acerba iniuria affici . . .
(‘It was when he (sc. Apollonius) was proclaiming this fact in the hearing of a large
crowd, so that anyone could see that he was suffering this cruelty simply because he
would not pay up . . . ’ Cic. Ver. 5.17)
(c) (sc. Verresx) Milesios navem poposcit quae eumx praesidii causa Myndum
prosequeretur.
(‘Verres demanded from the people of Miletus a ship to accompany and protect him
as far as Myndus.’ Cic. Ver. 1.86)
(d) Scripsit enim filiusx sex idcirco profugere ad Brutum voluisse quod, cum sibix
negotium daret Antoniusy ut eumy dictatorem efficeretx, praesidium occu-
paretx, id recusassetx.
(‘His son writes giving as his reason for wanting to take refuge with Brutus that when
Antony gave him the job of making him Dictator and seizing a strong point he had
refused.’ Cic. Att. 15.21.1)
(e) Barbarus eumx quidam palam ob iram interfecti ab eox domini obtruncavit.
(‘A certain barbarian slew him openly on account of his anger at the slaughter of his
master by that one.’ Liv. 21.2.6)
Supplement:
Legatusx in pace profectus est. In provincia pacatissima ita se gessit ut eix pacem esse
expediret. (Cic. Lig. 4); Nam si esset, non commi[s]sisset (sc. Antoniusx) ut eix
senatus tamquam Hannibali initio belli Punici denuntiaret ne oppugnaret Saguntum.
(Cic. Phil. 6.6—NB: remarkable instance of is in an argument clause); (Haeduix
speaking) . . . ita sex omni tempore de populo Romano meritos esse, ut . . . liberi eorumx
(del. Meusel) in servitutem abduci, oppida expugnari non debuerint. (Caes. Gal.
1.11.3);254 . . . Ambiorixx statim cum equitatu in Atuatucos, qui erant eiusx regno fini-
timi, proficiscitur. (Caes. Gal. 5.38.1); Hac victoria Lysanderx elatus . . . sic sibix indulsit
ut eiusx opera in maximum odium Graeciae Lacedaemonii pervenerint. (Nep. Lys. 1.3);
In Cicero, the genitive forms of the anaphoric pronoun is are more common than the
reflexive possessive adjective suus.255 Here, too, is is sometimes used in order to avoid
ambiguity. Examples are (f )–(i). In (f ), eorum is coreferential with the subject of the
clause. In (i), sua instead of eius would create ambiguity. A much debated case is (j),
where various emendations of eius have been proposed.
254
For instances of the anaphoric pronoun in such contexts in Caesar, see Lebreton (1901b: 16–17).
255
For the instances in Cicero, see Lebreton (1901a: 142–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(f ) Iam membrorum . . . aliax videntur propter eorumx usum a natura esse donata . . .
(‘Again, of the members . . . some appear to have been bestowed on us by nature for
the sake of their use . . . ’ Cic. Fin. 3.18)
(g) Antoniumx . . . in mediis eiusx iniuriis et cupiditatibus mors oppressit.
(‘Now Antonius . . . in the midst of his career of greed and injustice died suddenly.’
Cic. Ver. 3.213)
(h) Quid loquar quanta ratio in bestiisx ad perpetuam conservationem earumx
generis appareat?
(‘Why should I speak of the amount of rational design displayed in animals to secure
the perpetual preservation of their kind?’ Cic. N.D. 2.128)
(i) Regemx ipsum suspectum habebant pro eiusx crudelitate perfidiaque . . .
(‘The king himself they looked on with suspicion, for his cruelty and treachery . . . ’
Liv. 32.19.7)
(j) Orator autem, vivis eiusx aequalibus, proxumus optumis numerabatur (sc.
Curiox) . . .
(‘Still as an orator he was regarded by his contemporaries as next in rank to the
best . . . ’ Cic. Brut. 220)
Supplement:
(sc. Siculix) Me saepe esse pollicitum, saepe ostendisse dicebant, siquod tempus accidis-
set, quo tempore aliquid a me requirerent, commodis eorumx me non defuturum. (Cic.
Div. Caec. 2); (sc. Pompeiusx) . . . ipse cunctae Italiae cupienti et eiusx fidem imploranti
signum dedit ut ad me restituendum Romam concurreret. (Cic. Mil. 39);
(sc. Gaditanix) Testantur . . . hunc praesentem Cn. Pompeium, quem procul ab illo-
rumx moenibus acre et magnum bellum gerentem commeatu pecuniaque iuverunt . . .
(Cic. Balb. 40); Nam mihi isx [ter] dixit te scripsisse ad se mihi ex illiusx litteris rem illam
curae fuisse, quod ei pergratum erat. (Cic. Att. 5.11.6);
Appendix: The reflexive pronoun usually follows the constituent it is coreferential
with. There are a number of instances in which the reflexive is not used in spite of
coreferentiality with another constituent. Some of these instances occur in participial
phrases functioning as secondary predicate and in prepositional phrases. In some of
these cases the coreferential constituent precedes, as in (k), but that explanation is
not sufficient to explain all cases, for example (l).256
(k) Quin etiam <M.> Favonio, fortissimo viro, quaerenti ex eox qua spe fureret
Milone vivo, responditx triduo illum . . . esse periturum.
(‘But when the gallant Marcus Favonius asked him what he hoped for in his
frenzy, so long as Milo lived, he replied that in three days Milo . . . would be
dead.’ Cic. Mil. 26)
256
So K.-St.: I.601, where more examples can be found.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(l) Omne igitur quod vivit . . . idx vivit propter inclusum in eox calorem.
(‘Every living thing therefore . . . owes its vitality to the heat contained within
it.’ Cic. N.D. 2.24)
From Celsus onwards, instances are found in which sibi is more or less equivalent to
per se, denoting spontaneous or independent processes, states, or actions, as in (d).
(e) Nam scorpio sibi ipse pulcherrimum medicamentum est.
(‘In fact scorpio itself is an excellent drug.’ Cels. 5.27.5)
Supplement:
Ex carnibus eorum non manducabitis et sibi mortua eorum non tangetis. (‘You will
not eat their meat nor touch their flesh if they died naturally.’ Vet. Lat. Lev. 11.8); Qui
et sibi quidem post unam horam sani fiunt. (‘There are also some that after an hour
recover by themselves.’ Mulom. Chir. 502);258
In Late Latin the use of dative sibi with a stressing function is also found in the almost
fixed combination se sibi.259 In a similar way sibi is found in combination with suus
and other words to stress the individual involvement of the subject (see § 11.30).
257
See Sznajder (1981: 20).
258
For sibi = per se, see Adams (2013: 353–6), with references.
259
Dahlén (1964: 187–201).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
The sections below follow the same order of discussion as the treatment of the
determiners in §§ 11.104–11.107.
260
For a discussion of the function of these pronouns, see Joffre (2011).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
As for the choice of a particular demonstrative pronoun, it is the speaker who decides
from which perspective he wants to present an entity or entities within a certain
situation. The result may be that the same entity is referred to by the same speaker
with different demonstrative pronouns. This is illustrated by (e).262 Diniarchus has
just said farewell to his girlfriend Phronesium (Vale) and continues in what seems to
be a monologue reflecting upon the situation. The passage is not a piece of continuous
narrative, so it is no surprise that the anaphoric pronoun is is absent. Haec in l. 436
expresses the close relationship between the speaker and his lover who just left the
scene. Illam and illi in l. 441 are contrastive ($ egone). Huic in l. 442 is understand-
able. Really difficult are isti and istam in ll. 443–4. The best explanation is that they are
part of an interaction between Diniarchus and the audience: that girl I am telling you
about. In l. 446 illi is in contrast with mihi in the next line.
261
The description presented here follows essentially Bach (1888), K.-St.: I.619–20 (from whom some
of the examples in the Supplement are taken); TLL s.v., and de Jong (1998). For other views, see Keller
(1946), Kurzová-Jedličková (1961), Fontán (1965), and Eckert (2003: 127–45). For a survey of the
literature, see Eckert (2003: 30–44).
262
For discussion, see de Jong (1998: 21–3, 31–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(e) Vale.
Pro di immortales, non amantis mulieris,
sed sociai unanimantis, fidentis fuit
officium facere quod modo haec fecit mihi, 436
suppositionem pueri quae mihi credidit,
germanae quod sorori non credit soror.
Ostendit sese iam mihi medullitus.
Se mi infidelem numquam, dum vivat, fore.
Egone illam ut non amem? Egone illi ut non bene velim? 441
Me potius non amabo quam huic desit amor. 442
Ego isti non munus mittam? Immo ex hoc loco 443
iubebo ad istam quinque perferri minas, 444
praeterea opsonari <una> dumtaxat mina.
Multo illi potius bene erit quae bene volt mihi, 446
quam mihi, qui mihimet omnia facio mala.
(‘Be well. Immortal gods! It wasn’t the duty of a woman in love, but of a trusting
soulmate, to do what she just did for me: she confided the smuggling in of the boy to
me, something which a true sister doesn’t confide to her sister. She revealed herself to
me from the depths of her heart: that she’ll never be unfaithful to me for as long as
she lives. Should I not love her? Should I not wish her well? I’d rather not love myself
than that she should be lacking in love. Should I not send her a present? No, this very
instant I’ll have five minas brought over to her, and in addition I’ll have food bought
for at least one mina. I’d rather she, who wishes me well, have a good time than I,
who do all bad things to myself.’ Pl. Truc. 433–47)
Supplement:
Sed quis hic est? (Pl. As. 378); Tu si hic (‘this person’ = ‘me’) sis aliter sentias. (Ter.
An. 310); Nam dudum ante lucem et istunc et te vidi. (Pl. Am. 699); Ne iste hercle ab
ista non pedem discedat, si licessit, / qui nunc festinat atque ab hac minatur sese
abire. / # Sermoni iam finem face tuo, huius sermonem accipiam. (Pl. As. 603–5—
NB: note the change of perspective); (aside) Certe edepol, quom illum contemplo et
formam cognosco meam, / quem ad modum ego sum (saepe in speculum inspexi),
nimis simil’est mei. (Pl. Am. 441–2); Quid nunc, mulier? audi’n illum? # Ego vero, ac
falsum dicere. # / Neque tu illi neque mihi viro ipsi credis? (Pl. Am. 755–6);
Heus, ecquis hīc est? (Pl. Am. 1020); Age, alter istinc, alter hinc assistite. (Pl. Rud.
808); Etiam tu hoc respondes quid istic tibi negoti’st? (Ter. An. 849); Illuc regredere
ab ostio. (Pl. Aul. 46);
shown for their use as determiners in § 11.105) that in certain contexts they are used
with their proper meanings, which are different from the semantic contribution of the
anaphoric pronoun.263 This is shown for iste, ille, and hic, in that order.
Neuter demonstrative pronouns can be used to refer to words or actions mentioned
in the preceding context (that is, when they refer to a third-order entity). Examples of
iste are (a) and (b). With istuc in (a), Amphitryon refers to what Alcumena says she
has heard from him: ‘the information provided by the addressee’. Note that in the
preceding text Amphitryon uses id and therefore does not associate the information
he is referring to with Alcumena (or a third party). Likewise, in (b) istaec refers to
what Alcumena has just said (it is plural because the information is more complex).
(a) (Amphitryon) An etiam id tu scis? # (Alcumena) Quipp’ qui ex te audivi,
ut urbem maxumam / expugnavisses regemque Pterelam tute occideris. /
# (Amphitryon) Egone istuc dixi?
(‘You know about it? # Of course! I heard from you how you conquered the great city
and slew King Pterela yourself. # Did I tell you about this?’ Pl. Am. 745–7)
(b) Vera istaec velim.
(‘I wish these things were true.’ Pl. Am. 835)
The situation is more complicated in (c). The speaker of the prologue tells the story
about an exposed girl (puellam in l. 41). After a first (re)introduction with illam in l.
42 (not eam, because of intervening mulierem), the girl is referred to by the anaphoric
pronoun several times, except in l. 53 (istanc). Istanc must mean ‘that girl you are
familiar with’,264 but in l. 56 it is eam again, in a very similar context. Maybe after the
excursus-like details about the father and the son in ll. 50–1 the anaphoric pronoun
was not sufficient to ensure proper identification of the girl. (NB: eandem illam in l. 61
does not pose a problem: the anaphoric pronoun is excluded in this combination.)265
(c) Is servos—sed abhinc annos factum est sedecim
quom conspicatu’st primulo crepusculo
puellam exponi. Adit extemplo ad mulierem 41
quae illam exponebat, orat ut eam det sibi. 42
Exorat, aufert. Detulit recta domum,
dat erae suae, orat ut eam curet, educet.
Era fecit, educavit magna industria,
quasi si esset ex se nata, non multo secus.
Postquam ea adolevit ad eam aetatem ut viris
placere posset, eam puellam hic senex
amat efflictim et item contra filius.
263
The distinction between the anaphoric use of the demonstratives and the anaphoric pronoun is the
main point of de Jong (1996).
264
MacCary and Willcock ad loc. paraphrase ‘that girl about whom you (the audience) have been
hearing’. For doubts about the appropriateness of istanc in this context, see TLL s.v. 504.38ff.
265
See TLL s.v. idem 203.63ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
In the above passage, istanc seems to be an instance of anaphoric use of the pronoun
in continuous narrative, but this is a special type of narrative: it is a story that is orally
told to an audience. In this kind of story the speaker may appeal to the knowledge of
the addressee, as is the case with istanc in l. 53. In narrative without an obvious
addressee iste has no place. In Caesar’s de bello Gallico and de bello Civili, there is only
one instance of iste, in a speech (Gal. 7.77.5).
The role of ille in a continuous text is shown in (d)–(f ). In (d), Mercury explains
the situation to the audience; illa refers to the immediately preceding Alcumenae. It is
chosen instead of anaphoric ea, because there is a change of topic (and subject). Illum
refers to pater meus (Jupiter), who is not Alcumena’s husband, contrary to what she
thinks (illum is also contrastive). Ille occurs relatively often in the nominative, due to
its frequently marking topic (and subject) change.266 In Plautus co-occurrence of two
266
See Pinkster (1987b); de Jong (1996: 502); Pennell Ross (1996).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
forms of ille in the same clause is relatively frequent.267 In (e), the lar familiaris tells
the story of the family he belongs to. After the introduction of the daughter, she is
referred to by ea. The next reference is illam: the lar has acted not in the interest of the
father, but of the daughter (implied contrast). At the end of the passage illa illum can
be explained as double contrast: he knows her : she doesn’t know him. In (f ), there are
two references to a piece of text. Istuc can be explained as above: Lysimachus refers to
‘what you—the cook—are talking about’. Illud, in Lysimachus’ words to his wife
Dorippa, refers to the same cook’s words: ‘what he is talking about’. Note also istaec
‘your actions and sentiments’.
(d) Quae illi ad legionem facta sunt memorat pater / meus Alcumenae. Illa illum
censet virum / suom esse, quae cum moecho est.
(‘My father’s telling Alcumena what happened there during the campaign. She
believes he’s her husband, while she is with an adulterer.’ Pl. Am. 133–5)
(e) Huic filia una est. Ea mihi cottidie / aut ture aut vino aut aliqui semper
supplicat, / dat mihi coronas. Eius honoris gratia / feci thesaurum ut hic
reperiret Euclio, / quo illam facilius nuptum, si vellet, daret. / Nam com-
pressit eam de summo adulescens loco. / Is scit adulescens quae sit quam
compresserit, / illa illum nescit, nec compressam autem pater.
(‘The man who lives here has one daughter. She worships me every single day with
incense or wine or something else and gives me garlands. It’s in order to honour her
that I let this man here, Euclio, find the treasure, so that he might give her more easily
in marriage should he wish to do so: a young man of very high rank has raped her.
This young man knows who the girl he raped is, but she doesn’t know him, and her
father doesn’t even know that she’s been raped.’ Pl. Aul. 23–30)
(f ) (Coquus) Nempe uxor ruri est tua, quam dudum dixeras / te odisse [aeque]
atque anguis. # (Lysimachus) Egone istuc dixi tibi? # / (Coquus) Mihi quidem
hercle. # (Lysimachus) Ita me amabit Iuppiter, / uxor, ut ego illud numquam
dixi. # (Dorippa) Etiam negas? / Palam istaec fiunt te me odisse.
(‘Surely your wife is in the country, who you said a while ago you hate as much as you
hate snakes. # Did I say that to you? # Yes, to me. # As truly as Jupiter will love me,
my wife, I never said that. # You still deny it? It’s coming out into the open that you
hate me.’ Pl. Mer. 760–4)
267
See Lodge: s.v. ille § 2.
268
For statistical information about the use of the plural and singular of these pronouns in Caesar, see
de Jong (1996: 507).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
Early Latin examples. Ex. (h) and a few other are usually emended.269 Examples from
narrative prose are (i) and (j). In both instances the context indicates that there are
several persons available for the next development in the narrative, but that ‘this
particular’ individual is the one selected for it. In both cases the anaphoric pronoun
would be less suitable.270
(g) (Sosia) Iam vos rediistis in concordiam? # / (Iuppiter) Derides qui scis haec
[iam] dudum me dixisse per iocum. #/ (Sosia) An id ioco dixisti? Equidem
serio ac vero ratus. #/ (Iuppiter) Habui expurigationem. Facta pax est. #
(Sosia) Optume est.
(‘Have you returned to harmony now? # You’re mocking me; you know that I said
this as a joke a while ago. # You said it as a joke? I thought you said it in earnest and
seriously. # I’ve apologized. Peace has been made. # That’s excellent.’ Pl. Am. 962–5)
(h) Fugitivos ille, ut dixeram ante, huius patri / domo quem profugiens do-
minum apstulerat vendidit. / Hic (is cj. Fleckeisen) postquam hunc emit,
dedit eum huic gnato suo / peculiarem, quia quasi una aetas erat.
(‘As I said before, that runaway slave sold his master he’d snatched when he was
running away from home to the father of this chap here. After this man bought him,
he gave him to his son here as his own because their age was roughly the same.’ Pl.
Capt. 17–20)
(i) Erat una cum ceteris Dumnorix Haeduus, de quo a nobis antea dictum est.
Hunc secum habere in primis constituerat . . .
(‘Among the others there was Dumnorix of the Aedui, of whom we have spoken
before. Caesar had determined to keep Dumnorix in particular with him . . . ’ Caes.
Gal. 5.6.1)
(j) Erat unus intus Nervius nomine Vertico, loco natus honesto, qui a prima
obsidione ad Ciceronem perfugerat suamque ei fidem praestiterat. Hic servo
spe libertatis magnisque persuadet praemiis ut litteras ad Caesarem deferat.
(‘There was a Nervian in the camp, named Vertico, born to an honourable estate,
who at the very beginning of the blockade had fled to Cicero for refuge, and had since
proved his loyalty to him. He persuaded a slave by the hope of freedom and by great
rewards to deliver a despatch to Caesar.’ Caes. Gal. 5.25.2–4)
Supplement:
Haec (‘the things I mentioned’) igitur Epicuri non probo, inquam. (Cic. Fin. 1.26);
Nam qui septem appellantur eos qui ista (‘the things you do’) subtilius quaerunt,
in numero sapientium non habent. (Cic. Amic. 7); Sic prorsus, inquit, existumo
atque istum (‘the person you mention’) de superioribus paene solum lego. (Cic.
Brut. 125); . . . quod isti (‘the man you are praising’) contigit uni . . . (Cic. de Orat.
2.228); Atqui ego ista (‘the things you are talking about’) sum omnia dimensus.
(Cic. Sen. 59);
269 270
See also TLL s.v. hic 2708.44ff. For discussion, see de Jong (1996: 504–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(d) Vester iste puer praetextatus in provincia quem ad modum fuisset non
dicerem si pueri esse illam culpam ac non patris existimarem. Tune, cum
te ac tuam vitam nosses, in Siciliam tecum grandem praetextatum filium
ducebas, ut, etiamsi natura puerum a paternis vitiis atque a generis simi-
litudine abduceret, consuetudo tamen eum et disciplina degenerare non
sineret? Fac enim fuisse in eo C. Laeli aut M. Catonis materiem atque
indolem.
(‘The behaviour of that young boy of yours in Sicily is such as I would not
mention, if I thought the boy himself was to blame for it and not his father.
How could you, Verres, knowing yourself and the life you lead, take with you to
Sicily a young man who was no longer a child, so that, even if his natural bent
tended to wean him from his father’s vices and make him unlike his family,
habit and training might nevertheless keep him true to type? Suppose there had
been in him the stuff and the disposition to make a Laelius or a Cato of him.’ Cic.
Ver. 3.159–60)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
271
For statistical observations on the use of these cataphorically used pronouns in various authors and
text types, see Bodelot (1996). In her sample there are 84 instances of is, 210 of hic, 126 of ille, 12 of iste.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Enim vero illud praeter alia mira miror maxume, / {si haec habet [pateram]
illam}.
(‘Honestly, I am astonished if she has that bowl more than I am astonished about the
other astonishing things.’ Pl. Am. 772–3)
(c) Scelestissume, audes mihi praedicare id, / {domi te esse nunc qui hic ades}?
(‘You hardened criminal, you dare tell me that you, who are here, are at home now?’
Pl. Am. 561–2)
(d) Em istuc ago, {quo modo argento intervortam et adventorem et Sauream}.
(‘Well, I’m at it, how I can swindle both the newcomer and Saurea out of the money.’
Pl. As. 358–9)
(e) Id, amabo te, huic caveas. # Quid isti caveam? # {Ut revehatur domum . . . }
(‘Please, do take care of this for her. # What should I take care of for her? # That she’s
taken back home . . . ’ Pl. Bac. 44)
In this section the diachronic developments of the four words involved (is, hic, ille,
iste) are discussed both for their use as a determiner and as a pronoun. The outcome
of the various developments is the (almost complete) disappearence in the Romance
languages of anaphoric is and of demonstrative hic. The function of the former was
continued by ille, the anaphoric use of which expanded to indicate close-distance
anaphora, the typical area of is. The deictic function of hic (relatedness to the speaker)
was taken over by iste which expanded its original meaning (relatedness to the
addressee). What the two had in common was their relation to the interaction
between speaker and addressee.273
272
For more examples, see TLL s.v. ille 344.42ff.
273
For a discussion of the common features of hic and iste (both related to the communicative
situation) in Apuleius, see Joffre (2012a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
From the beginning, the determinative use of is was much less frequent than its use
as pronoun (see Table 11.9), which is understandable: reintroducing a noun that is
already present or inferrable from the context in combination with an anaphoric
pronoun seems to be redundant. Furthermore, a considerable portion of determina-
tive instances is constituted by its functioning as determiner of an autonomous
relative clause to mark the function of that clause in its sentence (see § 18.16). This
de facto non-anaphoric use of is remains remarkably stable until Late Latin.274 In its
purely anaphoric determinative use, is is freely combined in Plautus’ time with all
sorts of nouns and noun phrases. Over time, the range diminishes. After the Classical
period the determinative use seems to become restricted to a number of set phrases
(ea res, eo die, eo genere, etc.), in which ille is used (much) less frequently, sometimes
even less frequently than hic (in expressions like id est, hoc gradually replaces id). In
its anaphoric pronominal use, already in the Classical period certain forms were more
frequent than others. The nominative of is is relatively infrequent (in comparison
with the oblique forms) if compared with the nominative of ille.275 This relative
infrequency is related to the fact that for the expression of subjects that are already
introduced in the preceding context, is competes with zero subjects and the connect-
ing relative. In certain expressions id remained in use: id est and ob id, for example.
Certain forms are replaced less often by forms of ille (for example eius). However, as
early as the Elder Seneca, the pronoun ille is found where Early and Classical Latin
required is. An example is (a), where it is not possible to assign more than an
anaphoric sense to illam.276
(a) Utcumque res tulerat, ita vivere, nihil vocis causa facere, non illam per gradus
paulatim ab imo ad summum perducere . . .
(‘ . . . to live as circumstances suggested, without doing anything for the sake of his
voice—such as gradually taking it up from low to high . . . ’ Sen. Con. 1.pr.16)
Educated writers continued to use is until the end of the period covered by this
Syntax. The progression of ille as a close-distance anaphoric determiner and pronoun
(alongside most of its other uses) is visible in, for example, the letters of Claudius
Terentianus, where alongside thirty-one instances of ille there are only four of is, and
in the passages spoken by the freedmen in the Cena Trimalchionis. In Augustine’s
Sermones, intended for an audience that was mainly uneducated, the proportion of
anaphoric ille : is (70/30) is the opposite of the proportion in De civitate Dei, intended
for a learned public (30/70).277
274
For comparative data about Cicero and Augustine, see André (2010) and André and Fruyt (2012).
For detailed statistics in general, see Meader (1900, 1901); Selig (1992) with Pinkster (1996).
275
See Pinkster (1996).
276
For is and ille in The Elder Seneca, see Pinkster (2005c). For their use in Petronius, Claudius Terentianus,
and other documentary texts, see Adams (1977: 44–7, 2003a: 13–17) and Halla-aho (2009: 160–3). For
the Peregrinatio, see Nocentini (1990). For ille in the Peregrinatio and in Augustine, see Pieroni (2014).
277
The best statistical material for the frequency of individual forms of is can be found in Meader
(1900, 1901).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 11.11 shows the frequency of the nom.sg. and acc.sg. forms of is, ille, hic, iste in
a number of texts (data taken from LASLA August 29, 2012). Although there is
considerable variation, depending on the type of text among other things, the low
frequency of the nom.sg. of is is obvious, as well as the high frequency of nom. sg. ille
(if less so in Seneca).
Table 11.11 Frequency of nom.sg. and acc.sg. forms of anaphoric and demonstrative
determiners/pronouns
is ille hic iste
nom.sg. acc.sg. nom.sg. acc.sg. nom.sg. acc.sg. nom.sg. acc.sg.
Pl. Am. 31 43 29 23 54 48 2 7
Cic. S. Rosc. 61 86 27 28 45 83 16 7
Caes. Gal. 53 218 27 19 40 75 1 0
Sen. Ep. 123 185 376 443 510 634 97 32
Whereas the short-distance anaphoric use of ille increased, its deictic function
decreased, which was compensated for by the development of eccille (see the Appen-
dix in § 11.103). Its exophoric use proved useful to signal the grammatical function of
uninflected Hebrew words, as in (b), with precedents as early as Cicero (c).278 Much
later successors of this determinative use of ille became the definite article in most
Romance languages.279
(b) . . . egerunt filii Israhel cum his qui erant ex captivitate illud phase (Gr. e
åÆ) quartadecima luna mensis primi . . .
(‘The people of Israel who came from exile kept the passover on the fourteenth day of
the first month . . . ’ Vet. Lat. III Esdr. 7.10)
(c) Illo enim addito ‘iuste’ fit recte factum, per se autem hoc ipsum reddere in
officio ponitur.
(‘The addition of the qualification “on principle” makes it a right action: the mere
restitution in itself is counted as an appropriate act.’ Cic. Fin. 3.59)
Ille is used in the Vetus Latina versions of the Bible where the Greek original,
following the Hebrew original, has a definite article.280 Jerome did not translate the
article. Compare (d) with (e).
(d) Adducite vitulum illum saginatum / pastum.
ŒÆd çæ τὸν å τὸν Ø ı .
(‘Bring the fattened calf.’ Vet. Lat. Luc. 15.23—NB: the mss. vary)
278
‘Quasi adminiculum pronominale’ (TLL s.v. ille 356.14ff.).
279
For the development and the chronology, see Banniard (1995); Pieroni (2010a: 462–76), with
references.
280
See Abel (1971); Rubio (2009: 213).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
The overall conclusion must be that ille did not function as a fully fledged article in
the period covered by this Syntax. The same must be concluded for ipse.281
The identifier idem is more frequently used as a pronoun than as a determiner, at least
in Cicero.282 It can be determined by demonstrative and (connecting) relative pro-
nouns, as in (a)–(c), as well as by omnis, as in (d).
(a) Multae aliae idem istuc cupiunt / quibus copia non est.
(‘Many others have that same desire, yet no possibility of achieving it.’ Pl. Mil. 1040–1)
(b) Ecquis inventus est postea praetor qui idem illud ediceret?
(‘Was any praetor found afterwards to draw up an edict in similar style?’ Cic. Ver. 1.111)
(c) Quod tu si idem faceres, magis in rem et vostram et nostram id esset.
(‘If you acted in the same way, it would be to your advantage and to ours.’ Ter. Hec. 249)
(d) Non omnia eadem aeque omnibus, ere, suavia esse scito.
(‘Master, you should know that not everything is equally sweet for everyone.’ Pl. As. 641)
Idem can also be used to add one or more states of affairs in which one and the same
entity is involved or one or more properties that pertain to an entity, in the sense of
‘also’, ‘likewise’, ‘at the same time’. Examples are (e)–(h).283 In (e), a second property,
era, is assigned to tu. Similarly, in (f ) the adjective delicatam adds a second property
to Asiam. In (g) and (h) a second activity of the subject, tu and me respectively, is
introduced.
(e) Tuo facit iussu, tuo imperio paret. Mater tu, eadem era es.
(‘She does it on your order, she’s obeying your command: you’re her mother, you’re
also her mistress.’ Pl. As. 147)
(f ) . . . Asiam istam refertam et eandem delicatam sic obiit . . .
(‘He traversed that Asia of yours, so crammed with wealth and luxurious as well . . . ’
Cic. Mur. 20)
(g) Tu argentum eluito, idem exstruito.
(‘You must clean the silverware and also set it out.’ Pl. Ps. 162)
281
For a detailed discussion of the evidence, see Adams (2013: 482–527). For an early ninth-century
text with abundant use of anaphoric ille, see Bischoff (1981).
282
In Merguet (Reden) s.v. idem 52/3 columns determiner, 91/2 pronoun.
283
For more examples, see TLL s.v. idem 191.27ff.; OLD s.v. §§ 7 and 8.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
This usage is also found with personal pronouns, as in (i), and with relative pronouns,
as in (j) and (k).284
(i) At ego faciam tu idem ut aliter praedices, / Amphitruo . . .
(‘But I’ll make you speak differently too, Amphitruo . . . ’ Pl. Am. 1085–6)
(j) Ego sum ille Amphitruo, cui est servos Sosia, / idem Mercurius qui fit . . .
(‘I am that Amphitruo who has a slave Sosia who also becomes Mercury . . . ’ Pl. Am.
862–3)
(k) Nullum genus est ioci quo non ex eodem severa et gravia sumantur.
(‘There is no source of laughing-matters from which austere and serious thoughts are
not also to be derived.’ Cic. de Orat. 2.250—NB: remarkable word order)
More or less certain instances of idem used in the sense of the anaphoric pronoun is
can be found in TLL s.v. idem 205.41ff. An example is (l).285
(l) Nos ergo cum venissemus in eodem campo, peraccessimus ad locum ipsum . . .
(‘When we reached this plain, we went on to the very spot . . . ’ Pereg. 10.7—
tr. Wilkinson)
The comparative constructions with which idem can be used are dealt with in
Chapter 20.
284
For more examples, see TLL s.v. idem 193.43ff.
285
For this use of idem in Late Latin, especially technical texts, see Adams (2013: 494–8).
286
For more instances of ipsum with an adverb, see TLL s.v. 347.59ff.; OLD s.v. ipse § 8c.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
In all these examples the contribution of ipse is to indicate that it is ‘precisely’ this
element, that this element is ‘the very X’ suited to the context, that it is ‘this one and
no other’. Often contrast with another constituent in the context is involved or the
element to which ipse is related is unexpected (counter-presuppositional). In (f ),
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
there is a chain of zero-anaphora: ipse is not the subject of conscendit but serves to
underline the contrast between what the leno does with the young man and what
the young man does himself. In (h), one does not expect a person to be beatus when
being tortured and certainly not to be beatior than a person drinking on a couch of
roses. In this Syntax the use of the intensifier ipse demonstrated in (a)–(i) is called its
PREGNANT use.
Another use of ipse is shown in (j)–(l). In (j), ipsum is related to the autonomous
relative clause that functions as subject in the accusative and infinitive clause. Ipsum
implies ‘without the assistance of others’, ‘alone’. In (k), ipse implies that the rhythm
follows ‘by itself ’. In (l), ipse is related to the agent of rogando, which is identical with
the subject of fatigasset. It implies that he did it ‘in person’. This use of ipse is called
DISCRETIVE in this Syntax. The discretive use seems only possible when related to
animate or other active entities in subject (or similar) function. This restriction does
not hold for the pregnant use of ipse. The two uses of ipse can co-occur in the same
sentence, as in (m).
(j) Nam qui cum ingeniis conflictatur ei<u>s modi / . . . / scias posse habere iam
ipsum s<uae> vitae modum.
(‘When a man is involved with characters like that . . . you can be sure that he is now
capable of taking control over his own life himself.’ Ter. An. 93–5)
(k) . . . cum ita structa verba sunt, ut numerus non quaesitus sed ipse secutus esse
videatur . . .
(‘ . . . when the words are so placed that the rhythm seems not to be planned but to
come naturally . . . ’ Cic. Orat. 219)
(l) . . . cum patrem primo allegando, deinde coram ipse rogando fatigasset . . .
(‘ . . . when he had worn her father out first by sending messengers, then by personal
requests delivered verbally . . . ’ Liv. 36.11.1)
(m) Atque hic quidem ipse et sese ipsum nobis et eos qui ante fuerunt in medio
posuit, ut ceteros et se ipsum per se cognosceremus.
(‘And this man personally put at our disposal both himself and his predecessors so
that thanks to him we became acquainted with the others and himself.’ Cic. Inv. 2.7)
In the scholarly literature these two aspects have received considerable attention.
Exx. (a)–(i) are sometimes called ‘inclusive’, (j)–(k) ‘exclusive’,287 or ‘climax’ vs
‘contrast’.288 Priscian already made a distinction between significantia ‘pregnant
force’ and discretio ‘discrimination’ as functions of ipse (12.5 = II.579.30K). McGlynn
s.v. makes a distinction between ‘ut emphasin indicet’ and ‘cum oppositione’. The
distinction is also present (without further elaboration) in Sz.: 189 ‘ipse . . . ist von
Haus aus “er und kein anderer”, dann “für sich allein”, “von selbst” u.ä.’ TLL s.v.
makes a distinction between ipse ‘demonstrativum’, on the one hand, and ipse
287
So also Bertocchi (1996a, 1999, 2000), Fruyt (2010), and already Meader (1910: 44).
288
So Pieroni (2010a: 448–59), with references. TLL s.v. uses the term ‘vis exclusiva’ for ex. (l).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
‘intensivum’ and ipse ‘determinativum’, on the other. The latter two terms seem to
cover what is here called ‘pregnant’ and ‘discretive’, but not entirely so.
The difference can also be formulated in terms of ‘scope’. In (a)–(g), ipse has the
elements it is related to in its scope. In (h)–(j), ipse functions at the clause level.
Ipse may be reinforced by the focusing suffix -met, as in (n) (for -met, see Chapter 22).
(n) Is prius quam hinc abiit ipsemet in exercitum, / gravidam Alcumenam
uxorem fecit suam.
(‘Before he himself went away to the army, he made his wife Alcumena pregnant.’ Pl.
Am. 102–3)
289
See TLL s.v. et 909.21ff.
290
For more instances, see TLL s.v. ipse 314.50ff., from which most examples are taken.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
13.475–6); . . . et ipsis etiam reis dare laudatores licet . . . (Quint. Inst. 3.7.2); Una
omnino interpositio difficilior est, quam ne ipse quidem facile legere soleo, de
quadrimo Catone. (Cic. Fam. 16.22.1);
Additive subjuncts: . . . quod eo facilius faciam, quod et ipse pecuarias habui grandes . . .
(Var. R. 2.pr.6); . . . tum et ipsi—perpauci autem supererant—cum Crispino consule
duobus iaculis icto et Marcello adulescente saucio et ipso effugerunt. (Liv. 27.27.7);
Alius Latio iam partus Achilles, / natus et ipse dea. (Verg. A. 6.89–90); . . . scio
Niceronem nihil nugarum narrare. Immo certus est et minime linguosus. Nam et
ipse vobis rem horribilem narrabo. (Petr. 63.2 (Trimalchio speaking)); Sed nescio
quomodo huius . . . exiliores orationes sunt . . . quam aut Laeli <aut> Scipionis aut
etiam ipsius Catonis. (Cic. Brut. 82); Forsitan ipse etiam Cepheus referetur in actus.
(Manil. 5.469); . . . Maharbal cum maiore robore virorum missus nec ipse eruptionem
cohortium sustinuit. (Liv. 23.18.5); Cogitavi Decium, qui nec ipse patri cessit. (Sen.
Con. 10.2.1); Quamquam huius quoque ipsius rei, quemadmodum scis, praecepta
sunt. (Cic. Brut. 112); Quorum (sc. medicaminum) ipsorum quoque pondera paria
miscentur. (Cels. 5.19.10);
‘Discretive’ ipse can be combined with expressions that indicate the independence of
action of the entity ipse is related to, such as sua sponte and per se. Examples are (e)–
(g). Moveatur in (g) is an autocausative passive.291 The personal involvement can be
underlined by words indicating presence, such as coram in (l) in § 11.144.292 The
personal involvement can also appear by its being opposed to something else, as in
(h). Ipse in this sense can be the scope of negation, as in (i).
(e) Atque illud optandum est oratori ut aliquam permotionem animorum sua
sponte ipsi adferant ad causam iudices . . .
(‘Another desirable thing for the advocate is that the members of the tribunal, of their
own accord, should carry with them to the Court some mental emotion . . . ’ Cic. de
Orat. 2.186)
(f ) (sc. Fabius) . . . praepositoque castris L. Scipione . . . Romam ipse ad consul-
tandum de bello rediit, sive ipse sponte sua . . . sive senatus consulto accitus.
(‘ . . . putting Lucius Scipio in charge of the camp . . . Fabius himself returned to Rome
to consult about the war, either voluntarily, . . . or, it may be, summoned by the
senate . . . ’ Liv. 10.25.12)
(g) Deinde cum mens moveatur ipsa per sese . . .
(‘Then, since the mind is capable of entirely self-originated motion . . . ’ Cic. Ac. 2.48)
(h) . . . seu qui ipse ambissit seu per internuntium . . .
(‘ . . . or if anyone should canvass himself, or through an intermediary . . . ’ Pl. Am. 71)
(i) Non ipse emam, / sed Lysimacho amico mandabo.
(‘I won’t buy her myself, but commission my friend Lysimachus to do so.’ Pl. Mer. 466–7)
291
For more instances, see TLL s.v. ipse 336.47ff.
292
For a few more instances, see TLL s.v. 327.73f.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
Supplement:
Neminem coegeris, ipsi convenerint sua sponte. (Cic. Caec. 59); Sed tamen huic malo
populus Romanus ipse nullo haruspicum admonitu sua sponte prospexit. (Cic. Har.
56); Huc si perveneris me tibi primum quidque concedente, meum vitium fuerit, sin
ipse tua sponte processeris, tuum. (Cic. Ac. 2.49); . . . praesertim cum is ardor qui est
mundi non agitatus ab alio neque externo pulsu sed per se ipse ac sua sponte
moveatur. (Cic. N.D. 2.31); Corripuit tremulis altaria flammis / sponte sua . . . cinis
ipse. (Verg. Ecl. 8.105-6); (sc. Tarquinius) Bellum, pacem, foedera, societates per se
ipse cum quibus voluit, iniussu populi ac senatus, fecit diremitque. (Liv.
1.49.7); . . . qui se ultro ad poenam ipsi obtulissent . . . (Liv. 43.4.3);
(o) Quo quidem auctore nos ipsi ea gessimus, ut omnibus potius quam ipsis
nobis consuluerimus.
(‘It was with his support that I carried out this accomplishment, which was for all
men’s interest rather than my own.’ Cic. Fin. 2.62)
(p) Dies noctesque aliquot obsessi vallum armis se ipsi tutati vallo sunt.
(‘They were besieged for several days and nights, and protected the rampart with
their swords, and themselves with the rampart.’ Liv. 22.60.23)
Cicero’s and Livy’s usage have received much attention. Exx. (n) and (o) are often
regarded as counter-intuitive options, just like (p), due to the fact that the distinction
between the two uses of ipse is not made.293
A further distinction between the pregnant and the discretive use seems to be that in
the former, ipse is as a rule adjacent to the element to which it is related, whereas in
the latter ipse seems to be more mobile.
The distinction discussed above is not always clear-cut. Within each type further
subtypes can be (and have been)294 distinguished, but it is difficult to find formal
correlates to these semantic distinctions. Translation into a modern language requires
quite a few different words, as is shown in the OLD.
293 294
See K.-St.: I.632, with references. See, for example, the TLL article.
295
So, for example, K.-St.: I. 628ff., who call it the pronoun of contrast (‘Pronomen des Gegensatzes’),
and Sz.: 189.
296
For examples, see TLL s.v. ipse 352.73ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(b) Siquidem ille ipse non volt et tu sen[si]sti in eo esse, Myrrina, / peccatum,
aderam . . .
(‘Even if he is unwilling and you felt that the wrong was on his side, Myrrina, I was
here . . . ’ Ter. Hec. 560–1)
(c) Miser homo est qui ipse sibi quod edit quaerit et id aegre invenit . . .
(‘Wretched is the man who has to look for his food himself and has a hard time
finding it . . . ’ Pl. Capt. 461)
(d) Atque etiam tu quoque ipse si esses percontatus / me ex aliis, scio pol
crederes nunc quod fers.
(‘And you yourself too, if you’d enquired about me from others, I know you’d entrust
me with what you’re carrying now.’ Pl. As. 502–3)
(e) Quis sibi ipse vim adhibet?
(‘Who can apply force to himself?’ Sen. Con. 9.3.11)
Supplement:
Subjects: Comparet? # / Id ipsum. (Ter. Hau. 778–9); Cum primum illi ipsi debue-
rint potius accusari de pecuniis repetundis quam ambitus. (Cic. Clu. 114); Quem
enim appellem . . . ? Senatumne? At is ipse auxilium petit a vobis . . . (Cic. Flac. 4);
Alium fac iam, quoniam sibi hic ipse desperat. (Cic. Mur. 45); . . . vide ne hoc ipsum
non sit necesse. (Cic. Phil. 13.15); Tantisper hoc ipsum magni aestumo quod polli-
cetur. (Cic. Tusc. 5.20); Cui tamen illud ipsum non accidisset, si in re p. quiescens
Epicuri legibus paruisset. (Cic. Tusc. 5.108);
Atque ut audias quid ego ipse sentiam, non solum mihi videris eorum studiis qui
[tuis] litteris delectantur, sed etiam patriae debere hoc munus . . . (Cic. Leg. 1.5); Nec
ego nunc ipse aliquid adferam melius. (Cic. N.D. 1.60); . . . cum tuam auctoritatem
tute ipse edicto provinciali repudiabas. (Cic. Ver. 1.112); De te . . . tu ipse quem ad
modum existimes vide etiam atque etiam, et tu te collige, et qui sis et quid facere
possis considera. (Cic. Div. Caec. 37);
Quis homo est me insipientior, qui ipse egomet ubi sim quaeritem? (Pl. Trin.
929); . . . neque solum ingratus, quod ipsum grave est, verum etiam impius appelletur
necesse est. (Cic. Red. Pop. 23); Homo, ut haec audivit, sic exarsit ad id quod non
modo ipse numquam viderat, sed ne audierat quidem ab eo qui ipse vidisset, ut
statim ad Philodamum migrare se diceret velle. (Cic. Ver. 1.64);
Nocte sub obscura quis te spoliavit amantem, / quis tetigit ferro durior ipse latus?
(Eleg. Maec. 1.28–9);
Non-subjects: Te cruci ipsum adfigent propediem alii. (Pl. Per. 295); Me ipso
praesente et Lycone factum est. (Pl. Cur. 714—NB: ablative absolute clause); Quo
ego illam? Ad nos scilicet. / # Ad patrem[ne]? # Ad eum ipsum. (Ter. Hau. 312–13);
Hunc quoque asserva ipsum, ne quo abitat. (Pl. Rud. 777); ‘Iratus’ inquit ‘ob hoc
ipsum fui, quod hoc scelere etiam tyrannicidium inquinaveras. (Sen. Con. 1.7.14);
Nam isti quid suscenseam ipsi? (Pl. As. 146); Istuc volo ergo ipsum experiri. (Ter. Ad.
172); Ego enim ipse cum eodem (sc. Platone) ipso non invitus erraverim. (Cic. Tusc.
1.40—NB: rare); Sed curiosi sunt hic complures mali / alienas res qui curant studio
maximo, / quibus ipsis nulla est res quam procurent sua. (Pl. St. 198–200); Nam et
publicani . . . suas rationes et copias in illam provinciam contulerunt, quorum ip-
sorum per se res et fortunae vobis curae esse debent. (Cic. Man. 17);
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
All these examples show that ipse does not belong to any of these pronominal lexical
categories. The question then remains what sort of relationship exists between ipse
and these pronouns when they occur in the same clause, often adjacent to each other
(and ipse usually following); neither of them is the head, and neither the attribute.
The other major category of constituents with which ipse co-occurs are nouns
(including proper names) and noun phrases.297 In (f ), ipse is related to the subject,
probably a case of discretive use. In (g), ipsis is related to a source argument, clearly a
case of pregnant use. In (h), with an inanimate noun, ipsum is placed between the
determiner illud and the phrase gravissimum . . . pietatis. Something similar is the case
within the relative phrase quod . . . templum in (i). Both are instances of the pregnant
use of ipse. Especially in (h) and (i) the question is whether ipse functions as an
attribute within the noun phrase. This is not the case; in all the examples (f )–(i),
leaving out the noun and its proper attributes results in a grammatical utterance.
(f ) Ipse in meo collo tuos pater cruminam collocavit.
(‘Your father put the wallet onto my neck in person.’ Pl. Epid. 360)
(g) . . . ex ipsis dominis meis pugnis exculcabo furfures.
(‘ . . . I’ll knock the husks out of their owners themselves with my fists.’ Pl. Capt. 810)
(h) . . . cum illud ipsum gravissimum et sanctissimum nomen pietatis levius mihi
meritis erga me tuis esse videatur?
(‘ . . . when even this solemn, sacred word “piety” seems to me inadequate for what
I owe you?’ Cic. Fam. 1.9.1)
(i) Erat eodem tempore senatus in aede Concordiae, quod ipsum templum
repraesentabat memoriam consulatus mei . . .
(‘At the same time the Senate had assembled in the Temple of Concord, the very
temple that recalled the memory of my consulship . . . ’ Cic. Sest. 26)
Supplement:
Subjects: Scio quam rem geram. Hoc / ipsus magister me docuit. (Pl. Aul. 412–12a);
Vix ipsa domina hoc, ut scias, exoptare ab dis audeat. (Pl. St. 296); Commodum / ipse
exit Lesbonicus cum servo foras. (Pl. Trin. 400–1); Si sapis, / neque praeter quam quas
ipse amor molestias / habet addas . . . (Ter. Eu. 76–8); . . . eum denique nos agrum
P. Rullo concessisse, qui ager ipse per sese et Sullanae dominationi et Gracchorum
largitioni restitisset. (Cic. Agr. 1.21); . . . cum T. Annius ipse magis de rei publicae
salute quam de sua perturbetur . . . (Cic. Mil. 1); Numerandus est ille annus denique
in re publica, cum . . . neque civis unus ex civitate sed ipsa civitas tuo et Gabini sceleri
furorique cessisset? (Cic. Pis. 26); . . . hos (sc. campos) ita vastatos nunc ac desertos
videbam ut ager ipse cultorem desiderare ac lugere dominum videretur. (Cic. Ver.
3.47); Nam ego nihil umquam vidi quod tam e manibus elaberetur quam mihi tum
est elapsa illa ipsa causa. (Cic. de Orat. 2.202); Eaque ipsa adfinitas haud spreta
gratiam Fabio ad volgum quaesierat. (Liv. 6.34.5);
297
For combinations of ipse with infinitives, see § 9.7.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
Non-subjects: Nam vigilante Venere si veniant eae, / ita sunt turpes, credo ecastor
Venerem ipsam e fano fugent. (Pl. Poen. 322–3); Ego, nisi ipsi Ballioni, nummum
credam nemini. (Pl. Ps. 644); Nempe ab ipso id accepisti Charmide? (Pl. Trin. 966);
Tua pietas nobis plane auxilio fuit, / quom huc advenisti hodie in ipso tempore. (Pl.
Poen. 1137–8); Ubi ad ipsum veni devorticulum, constiti. (Ter. Eu. 635); . . . cum
volgo essent et illo ipso et multis ante diebus admoniti et rogati? (Cic. Pis. 55); Qui
potuerunt ista ipsa lege quae de proscriptione est . . . bona Sex. Rosci venire . . .
potuerunt? (Cic. S. Rosc. 125); Hoc dico, hanc ipsam Cererem antiquissimam,
religiosissimam . . . a C. Verre ex suis templis ac sedibus esse sublatam. (Cic. Ver.
4.109);298 Nisi forte putamus neminem eodem tempore ipso et conceptum et natum
quo Africanum. (Cic. Div. 2.95); Triginta dies erant ipsi cum has dabam litteras per
quos nullas a vobis acceperam. (Cic. Att. 3.21.1);299 Qua re permagni hominis est et
cum ipsa natura moderati tum vero etiam doctrina atque optimarum artium studiis
eruditi sic se adhibere . . . (Cic. Q. fr. 1.1.22); Id ubi impetravere, post eos ipsos montes
unde averterant hostem exercitu omni coacto repente . . . castra Romanorum
oppugnare . . . adgressi sunt. (Liv. 40.25.5); Atque idem Atticus non prohibet
eodem ipso tempore inserere . . . (Col. 4.28.4);
The third type of context in which ipse can be used consists of instances in which
ipse appears alone. In clauses with a first or second person finite verb form, as in (j),
ipse can only be considered some sort of secondary predicate. In clauses with a
third person finite verb form, as in (k) and (l), there are no formal signs to prove
whether ipse is the subject itself (as a substitute third person pronoun or as an
anaphoric pronoun),300 or whether it is a secondary predicate, the subject being
inferred from the context (by zero-anaphora). However, it seems reasonable to
assume that in such cases ipse is not itself the subject. In (k), mater is the subject,
in (l) erus.
(j) Nunc quoius iussu venio et quam ob rem venerim / dicam simulque ipse
eloquar nomen meum.
(‘Now I’ll tell you on whose command and for what reason I’ve come, and at the
same time I’ll tell you my name myself.’ Pl. Am. 18–19)
(k) Mater irata est patri vehementer, quia scortum sibi / ob oculos adduxerit in
aedis, dum ruri ipsa Ø abest.
(‘My mother’s terribly angry with my father because she thinks he brought a
prostitute into the house before her very eyes, while she herself was away on our
country estate.’ Pl. Mer. 924–5)
298
For instances of combinations of ipse and noun phrases determined by various pronouns, see
Merguet (Reden) s.v. ipse 753–4.
299
For the translation of ipse when related to a noun phrase containing a numeral (‘whole’, and not
‘exactly’), see Hunter (1913: 94–7) and TLL s.v. ipse 332.13ff.
300
Burkard and Schauer (2000: 116) take it as a substitute third person pronoun, Joffre (2007c) as an
anaphoric pronoun. See also TLL s.v. ipse 319.59 ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(l) . . . ut apsente ero rem eri diligenter / tutetur, quam si ipse Ø assit aut rectius.
(‘ . . . so that when his master is away he guards his master’s business as diligently as if
he were present in person or even better.’ Pl. Men. 968–9)
Instances of ‘independent’ ipse can also be used in other functions in the clause, as in
(m)–(q), as well as at the noun phrase level, as in (r). It is usually not difficult to
identify the entity to which ipse is related. As the examples show, ipse usually closely
follows the clause in which the entity it refers to is situated. Poets allow a larger span,
as in (s). In all these examples ipse functions in a secondary predicate-like manner.
(m) . . . med amantem ex aedibus deiecit huius mater. / Argenti viginti minae
med ad mortem appulerunt / quas hodie adulescens Diabolus ipsi Ø daturus
dixit, / ut hanc ne quoquam mitteret nisi ad se hunc annum totum.
(‘ . . . her mother’s thrown me out of her house here, me, her daughter’s lover. Twenty
silver minas has driven me to my death; young Diabolus said he’d give it to her today
so that she wouldn’t send her anywhere except to him for a whole year.’ Pl. As.
632–5)
(n) Aedium dominum, lenonem Ballionem, quaerito. / # Quisquis es, adulescens,
operam fac compendi quaerere. / # Quid iam? # Quia tute ipsus ipsum Ø
praesens praesentem vides. / # Tune is es?
(‘I’m looking for the master of this house, pimp Ballio. # Whoever you are, young
man, spare yourself the trouble of looking. # Why? # Because you can see him face-
to-face. # Are you him?’ Pl. Ps. 1140–3)
(o) Pater istius adulescentis dedit has duas mi epistulas, / Lesbonici. Is mi est
amicus. . . . / # Ab ipson’ Ø istas (sc. epistulas) accepisti? # E manibus dedit
mi ipse Ø in manus.
(‘The father of that young chap, Lesbonicus, has given me these two letters. He’s a
friend of mine . . . # Did you receive the letters from him in person? # He himself put
them from his hands into mine.’ Pl. Trin. 894–902)
(p) Etenim omnes boni, quantum in ipsis fuit, Caesarem occiderunt.
(‘And so, all decent men killed Caesar so far as it was in them to do so.’
Cic. Phil. 2.29)
(q) Credet his equitibus Romanis populus Romanus qui . . . ipsis inspectantibus
ab isto civem Romanum . . . sublatum esse in crucem dixerunt.
(‘The Roman people will believe those Roman knights who . . . affirmed that a Roman
citizen . . . was crucified before their own eyes.’ Cic. Ver. 1.13)
(r) Quid erat ei nomen? # Thensaurochrysonicochrysides. / Videlicet propter
divitias inditum id nomen quasi est. / # Immo edepol propter avaritiam
ipsius Ø atque audaciam.
(‘What was his name? # Goldtreasure-Goldwinson. # I take it that that name, so to
speak, was given to him for his wealth. # No, for his greed and audacity.’ Pl. Capt.
285–7)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(s) Namque fluentisono prospectans litore Diae, / Thesea cedentem celeri cum
classe tuetur / indomitos in corde gerens Ariadna furores, / necdum etiam
sese quae visit visere credit / . . . / quem procul ex alga maestis Minois ocellis, /
saxea ut effigies bacchantis, Ø prospicit, eheu, / Ø prospicit et magnis
curarum Ø fluctuat undis / . . . / . . . / omnia quae toto delapsa e corpore
passim / ipsius ante pedes fluctus salis alludebant.
(‘For there, looking forth from the wave-sounding shore of Dia, Ariadne sees
Theseus, as he sails away with swift fleet, Ariadne bearing wild madness in her
heart. Not yet can she believe she beholds what yet she does behold . . . At whom afar
from the weedy beach with streaming eyes the daughter of Minos, like a marble
figure of a bacchanal, looks forth, alas! looks forth tempest-tossed with great tides of
passion . . . All these, as they slipped off around her whole body, before her very feet
the salt waves lapped.’ Catul. 64.52–67)
Supplement:
Nam tu quidem hercle certo non sanu’s satis, / Menaechme, qui nunc ipsus male
dicas tibi. (Pl. Men. 312–13); Ipsus, inquam, Charmides sum. (Pl. Trin. 988);
C(AIO)·POPLICIO·L(UCI)·F(ILIO)·BIBULO·AED(ILI)·PL(EBIS)·HONORIS / VIRTUTISQUE·CAUSSA·SENA-
TUS / CONSULTO·POPULIQUE·IUSSU·LOCUS / MONUMENTO·QUO·IPSE·POSTEREIQUE / EIUS·INFER-
2
RENTUR PUBLICE·DATUS·EST (CIL I .834 (Rome, early II BC)); Sed estne haec Thais quam
video? Ipsa est. (Ter. Eu. 848); M. Catoni, homini in omni virtute excellenti, de ipsius
accusatione, de senatus consulto, de re publica respondebo. (Cic. Mur. 54); Nam
consules modesti legumque metuentes impediebantur lege, non ea quae de me, sed ea
quae de ipsis lata erat . . . (Cic. Red. Sen. 4); . . . et hanc hereditatem fraternam et
omnia eius tibi commendo, in primisque ipsum, virum optimum mihique
familiarissimum . . . (Cic. Fam. 13.30.1); Ob eas causas ei munitioni quam fecerat
T. Labienum legatum praefecit. Ipse (sc. Caesar) in Italiam magnis itineribus
contendit . . . (Caes. Gal. 1.10.3);
NB: ipse as part of a non-finite subordinate clause (indicated by curly brackets):
Ergo in eadem voluptate eum, qui alteri misceat mulsum {ipse non sitiens}, et eum,
qui illud {sitiens} bibat? (Cic. Fin. 2.17); Quid quod semper in auctores redundat
timor nec quisquam metuitur {ipse securus}? (Sen. Dial. 4.11.3);
The position taken in this Syntax is that ipse is a lexical item of its own, with its own
properties. In many languages its equivalent ‘-self ’ has a special position as well. The
function it fulfils in its clause is that of some form of secondary predicate (see § 2.12
and Chapter 21). This is especially attractive for ipse in its discretive use.301
Appendix: In the section ‘pronomen demonstrativum’ of the article ipse the TLL
makes a distinction between purely deictic use (gestu ‘with a gesture’) and demon-
strative use with only words, without a gesture.302 Four examples of this deictic use
are (t)–(w). They belong to recognition scenes and gestures are certainly not
301
In Quirk et al. (1985: 356, 360–1) the partly comparable use of himself is called ‘appositional’. TLL
s.v. ipse sometimes uses the term ‘(magis) praedicative’. See also Pieroni (2010a: 452). For the status of
intensifiers (including ipse) from a typological perspective, see König (2010).
302
TLL s.v. ipse 299.46–300.39.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
excluded, but need not be part of the linguistic meaning of ipse. There are no signs
that they cannot be understood as normal instances of pregnant use.
303
For met, see Ch. 22.
304
For both developments, see Meader (1900, 1910), Väänänen (1981: 120–3, 1987: 50), Adams
(1995a: 92), Fruyt (2003), Joffre (2007c), Pieroni (2010a: 459–62), and, for examples, TLL s.v. 307.70ff.
For the Romance developments, see Renzi (2000) and Sornicola (2012).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
Supplement:
Enimvero si (sc. filia) emancipata est, non potest neque nuntium remittere neque
quae dotis causa data sunt condicere: ipsa enim filia nubendo efficiet dotem esse . . .
(Ulp. dig. 23.1.10—not very convincing); Primum quidem, cum populus de Aegypto
[libere] expeditus vim regis Aegypti per aquam transgressus evadit, ipsum regem cum
totis copiis aqua extinguit. (Tert. Bapt. 9.1); (Precede: summitas, ecclesia) Cum ergo
iubente Deo persubissemus in ipsa summitate et pervenissemus ad hostium ipsius
ecclesiae, ecce et occurrit presbyter . . . (Pereg. 3.4); Decoques in vino iunci radices et
marrubium, colas ipsut vinum et . . . (Pelag. 366);
As for the development ipse > idem, there are contexts in which both words can be
used (with a difference in meaning), and there are also contexts in which they co-
occur, obviously also with a difference in meaning (examples in § 11.146). Signs of the
later development have been seen in instances like (d)–(f ), and in translations of Gr. ›
ÆP ‘the same’ by ipse. Of these, examples (d) and (e) are not at all convincing. In
(f ), it is difficult to explain ipse in the ‘classical’ way. Whether it is equivalent to idem
or purely anaphoric is difficult to say. In the Peregrinatio most instances of ipse are
intensifying.305
(d) . . . maceriae, quas curatores aquarum causa cognita ne demolirentur do-
minis permiserunt quibus inscripta insculptave essent ipsorum qui permi-
sissent curatorum nomina, maneant, hac lege nihilum rogat[i]o.
(‘ . . . as to the fences, which the commissioners after due process have exempted their
owners from tearing down, and on which have been inscribed or carved the names of
the commissioners who gave their permission—as to all these, nothing in this
enactment prevents their remaining.’ Fron. Aq. 129.10 (= Lex Quinctia, 9 BC))
(e) (M. Cicero . . . somnium . . . narrabat) Deinde repente Augusto viso . . . ipsum
esse, cuius imago secundum quietem sibi obversata sit.
(‘Just then suddenly catching sight of Augustus, . . . he declared that he was the very
one whose form had appeared to him in his dream.’ Suet. Aug. 94.9)
(f ) Sed non ipsa parte exire habebamus qua intraveramus . . .
(‘But we did not have to come out by the way we had gone in . . . ’ Pereg. 4.5—tr.
Wilkinson)
Supplement:
Tu quidem felix et ipsa tanti mali ignorantia beata sedes incuriosa periculi tui . . .
(Apul. Met. 5.17.2);306 Nam et ipsi unitatem animae tuentur quae in totum corpus
diffusa et ubique ipsa velut flatus in calamo per cavernas, ita per sensualia variis
modis emicet . . . (Tert. An. 14.5); Cum inflaveris, virga molli armum caedito et sale et
oleo confricato ipso die et postero die tracta uteris . . . (Pelag. 44); Iesus Christus heri
et hodie ipse et in saecula. (Vulg. Hebr. 13.8—NB: › ÆP; Nestle-Aland read idem).
305
See Fruyt (2003). For the use of ipse in the sense of idem in Late Latin texts, see Adams (2013:
493–8).
306
So Callebat (1968: 283–4), hesitantly.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Appendix: Like the demonstrative and anaphoric pronouns, ipse is also used with
uninflected Hebrew names, as in (g).307
The order of the following sections is the same as those on indefinite determiners
(§ 11.109–11.116).
307
For more examples, see TLL s.v. ipse 310.12ff.
308
See Neue-W.: II.438ff.; Löfstedt (1933: II.92); Lundström (1986: 9–10).
309
So Bertocchi, Maraldi, and Orlandini (2010: 29), where also statistical information concerning the
types of clauses can be found (pp. 31ff.).
310
See Bertocchi and Maraldi (2005).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(e) Itidem ut saepe iam in multis locis / plus insciens quis fecit quam prudens
boni.
(‘Just as often before in many places a person’s done more good unknowingly than
knowingly.’ Pl. Capt. 44–5)
(f ) Sed quocumque quis tempore destinaverit inserere . . .
(‘But at whatever time one intends to graft . . . ’ Col. 4.29.5)
In interrogative clauses the indefinite determiners/pronouns ecquis and numquis
(also written num quis) are used, as in (g) and (h).
311
For instances in Classical Latin, see TLL s.v. aliquis 1613.27ff. For the Peregrinatio, see Väänänen
(1987: 54). For the increase of aliquis in Late Latin, see Bortolussi (2012).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(e) . . . si denique aliquid [quod] non contra ac liceret factum diceretur, sed
contra atque oporteret, tamen esset omnis eiusmodi reprehensio a vobis,
iudices, repudianda.
(‘If, in fact, it were asserted that something had been done not contrary to what was
legal, but to what was befitting, nevertheless, gentlemen, every such objection ought
to be overruled by you.’ Cic. Balb. 7)
(f ) Aut erit omnino incredibile, ut si aliquis quem constet esse avarum dicat
alicuius mediocris officii causa se maximam pecuniam neglexisse aut . . .
(‘Or the statement may be wholly incredible, as in the case of a man known by
everyone to be avaricious who says that he neglected great financial returns for the
sake of doing some humble duty, or . . . ’ Cic. Inv. 1.80)
312
For discussion, see Bertocchi, Maraldi, and Orlandini (2010: 48–55).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
Quisquam is also found in conditional clauses, as in (j) and (k). In other contexts it is
rare313 but can be found from Early Latin onwards, as in (l)–(o). In (o), where
quicquam governs the genitive of quantity lucis, it is interpreted as ‘the smallest
quantity’.314 In Late Latin quisquam is freely used in affirmative clauses. Quisquam
cannot be used in a clause with a potential subjunctive (*dixerit quisquam). It is only
attested in the singular.
(j) Eorum si quoiusquam scrofam in publico conspexero, / ex ipsis dominis
meis pugnis exculcabo furfures.
(‘If I see a sow of any of them in public, I’ll knock the husks out of their owners
themselves with my fists.’ Pl. Capt. 809–10)
(k) Di me perdant, si ego tui quicquam apstuli . . .
(‘May the gods destroy me if I carried away anything belonging to you . . . ’ Pl. Aul.
645)
(l) Quod quispiam ignem quaerat, exstingui volo, / ne causae quid sit quod te
quisquam quaeritet.
(‘In case anyone should request fire, I want it to be extinguished so that there’s no
reason why anyone should ask you.’ Pl. Aul. 91–2)
(m) Quam diu quisquam erit qui te defendere audeat, vives . . .
(‘So long as there remains a single man bold enough to defend you, you will live . . . ’
Cic. Catil. 1.6)
(n) . . . nunc angor quicquam tibi sine me esse iucundum.
(‘ . . . now it hurts me to find that you can enjoy anything without me.’ Cic. Fam.
7.15.1)
(o) Consul quoque nusquam remisso certamine dum quicquam superfuit lucis
hostem tenuit.
(‘Neither did the consul relax his efforts anywhere, but kept the enemy engaged as
long as there was any light.’ Liv. 4.39.5)
The sequence et nemo instead of nec quisquam is absent from Early and rare in
Classical Latin (et nihil is less rare), except in cases like (p) with repeated et . . . et. An
example of isolated et nemo is (q), emended in various ways. In Late Latin et nemo
becomes the norm.
313
For the use of quisquam in non-negative contexts, see Wagenvoort (1922).
314
Formulation taken from Bertocchi, Maraldi, and Orlandini (2010: 88–9), where further discussion
of quisquam and ullus can be found as well.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(p) Verum homines notos sumere odiosum est, cum et illud incertum sit
velintne hi sese nominari, et nemo vobis magis notus futurus sit quam est
hic Eutychus et . . .
(‘But it is a breach of good manners to take as examples men who are well
known, since it is uncertain whether they would like their names to be given;
besides, none of them is likely to be better known to you than Eutychus . . . ’
Cic. S. Rosc. 47)
(q) Quod autem satis est, eo quicquid accessit, nimium est; et nemo nimium
beatus est; ita nemo beato beatior.
(‘Any addition to what is enough makes it too much; and no one has too
much happiness; therefore no one can be happier than happy.’ Cic. Fin. 5.81)
For the combination of nemo + quisquam and nihil + quicquam, see § 8.33 and
(r) and (s).
(r) Lepidiorem uxorem nemo quisquam quam ego habeo hanc habet.
(‘Nobody has a more charming wife than I have.’ Pl. Cas. 1008)
(s) Nil ego tibi hodie consili quicquam dabo.
(‘I won’t give you any advice at all today.’ Pl. Bac. 1036)
element ‘known to the speaker’ is absent.315 Plural quidam also occurs in parallel
pairs, as in (h), and in alternation with alii, as in (i). Here the element ‘known to the
speaker’ still seems to be in force. In Late Latin quidam is used in contexts where the
other indefinites were common.
(e) Ego enim fateor in ista ipsa potestate inesse quiddam mali . . .
(‘And indeed I acknowledge that there is an element of evil in the very power of the
tribunate . . . ’ Cic. Leg. 3.23)
(f ) In quo (sc. libro) non sane multa mutavi sed tamen quaedam.
(‘A book in which I have made changes, not many to be sure, but still some.’ Cic. Att.
13.21a.1)
(g) Plurimi namque Bituricam, multi spioniam, quidam baliscam, nonnulli
arcelacam laudibus efferunt.
(‘For very many people are high in their praise of the Bituric, many of the Spionian,
some of the Basilic, and several of the Arcelacan.’ Col. 3.7.1)
(h) Quidam ludere eum simpliciter, quidam haud dubie insan<ire ai>ebant.
(‘Some said that he was playing childish tricks, some that he was unquestionably
insane.’ Liv. 41.20.4)
(i) Quae enim pertinacia quibusdam, eadem aliis constantia videri potest.
(‘For what appears to some as obstinacy may be strength of purpose in the eyes of
others.’ Cic. Marc. 31)
315
For more examples, see OLD s.v. quidam2 § 3.
316
For more instances, see OLD s.v. quicumque § A4; s.v. quilibet2; s.v. quisquis § A7; s.v. quivis2.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
equivalent to aliquis. The development took place ages after the period taken into
account in this Syntax.317
(a) Hercle miserum est ingratum esse homini id quod facias bene.
(‘It really is wretched if the good you do someone is not received with thanks.’ Pl.
Epid. 136)
(b) . . . si tamen labor dici potest, ubi homo desiderium suum compleri videt.
(‘ . . . if one can really speak of trouble when one sees one’s wishes fulfilled.’ Pereg. 13.1)
The determinative use of the words discussed in this section is discussed in § 11.33.
Often these pronouns are presented as a subcategory of the indefinite pronouns.
317
Rightly sceptical is TLL s.v. homo 2882.58ff. For the earlier literature, see Sz.: 198. Add Weerenbeck
(1943). More recent publications are Pepe (1975–6) who suggests ex. (a), Setaioli (2000), and Ramat and
Sansò (2010).
318
For the substantival use of omnis, see TLL s.v. 617.63ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
Supplement:
. . . castigatis oblique patribus quod cuncta curarum ad principem reicerent . . . (Tac. Ann.
3.35.1); Omnia non pariter rerum sunt omnibus apta . . . (Prop. 3.9.7); . . . id omnium
maxime tegendum, . . . obliviscendum, pro non dicto habendum esse. (Liv. 23.33.9);
319
Rosén (Haiim) (1998) considers quisque in instances like (b) a relative pronoun. For further
examples, see § 18.33.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The interrogative pronouns quis (masc. and fem.) ‘who’, quid (n.) ‘what’, and uter/
utra/utrum ‘which of the two’ are also discussed in § 6.19. They are all used in direct
and indirect questions. The distinction between masc./fem. quis and masc. qui / fem.
quae (normally the determiner forms) is not clear-cut (see § 11.28), as can be seen in
(b): qui and quae are used where quis is the regular form. Neuter quid and quod are
clearly distinguished.
(a) Sed quis haec est muliercula et ille gravastellus qui venit?
(‘But who is that girl and that grey-haired man who’s coming along?’ Pl. Epid. 620)
(b) Ergasile! # Ergasilum qui vocat? / # Respice. # Fortuna quod tibi nec facit nec
faciet, [hoc] me iubes. / Sed quis est? # Respice ad me, Hegio sum.
(‘Ergasilus! # Who’s calling Ergasilus? # Kindly look at me. # Look at you kindly?
You’re telling me to do what Good Fortune isn’t doing for you, or ever will do. But
who is it? # Kindly look at me. I’m Hegio.’ Pl. Capt. 833–5)
(c) . . . haec illaec est ab illa quam dudum <dixi>. / # Qua ab illarum?
(‘ . . . she’s that maid from the woman I told you about some time ago. # From which
of them?’ Pl. Mil. 1046-7)
(d) Quid nomen tibi est?
(‘What’s your name?’ Pl. Am. 364)
Supplement:
Cedo, / si vera memoras, quae fuit mater tua? (Pl. Cur. 641–2); Nunc speculabor quid
ibi agatur, quis eat intro, qui foras / veniat. (Pl. Truc. 708–9); Qui nominat me? Hem
quid istuc turbae’st, obsecro, / mi vir? (Ter. Ph. 990–1); . . . in somnis vidit ipsum
deum dicentem qui id fecisset. (Cic. Div. 1.54);
Uter ‘which of the two’ can fulfil all syntactic functions that quis can (argument,
adjunct, attribute), as is shown in (e), but may (just like uterque and quisque) also
function in the way secondary predicates do, as in (f ) and (g) (see also Chapter 20).
An example of plural use of uter (to refer to one of two groups) is (h).
(e) Sed uter vestrorum est celerior?
(‘But which of you is faster?’ Pl. Aul. 321)
(f ) Uter eratis, tun’ an ille, maior?
(‘Which of you was older, you or he?’ Pl. Men. 1119)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Pronouns
(i) Quam ob rem etiam atque etiam considera, C. Piso, quis quem fraudasse
dicatur. Roscius Fannium!
(‘Therefore I ask you, Gaius Piso, consider again and again who is said to
have cheated whom. Roscius cheated Fannius!’ Cic. Q. Rosc. 21)
(j) De praemiis autem quaeruntur duo: an ullo sit dignus qui petit, an tanto: ex
duobus, uter dignior, ex pluribus, quis dignissimus.
(‘As to rewards, two questions arise: whether the claimant deserves a reward
at all, and if so, such a large one. If two claim, we ask which is worthier; if
more, which is most worthy.’ Quint. Inst. 7.4.21)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
CHAPTER 12
The aim of this chapter is to sketch a picture of the Latin case and preposition system as
a whole. It does not contain the well-known catalogues of all the usages of the cases and
of the prepositions as one finds them in most grammars. In fact, many of the individual
usages of cases and prepositions are described elsewhere in this Syntax, notably in
Chapter 4 (verb frames), Chapter 9 (arguments), Chapter 10 (satellites), and
Chapter 11 (the noun phrase). The table of contents of these chapters will refer the
reader to the individual usages. Further references may be found in the Index. The
question underlying this chapter is: ‘How did the case and preposition system function?’
CASE is here used in the narrow sense of the information contained in the endings
of nouns and other nominal constituents apart from the information concerning
‘number’ (and—in the case of adjectives and comparable constituents—the informa-
tion concerning ‘gender’) (see § 3.2). PREPOSITION is used as a general term for
‘prepositions’ (which precede the constituent they govern, such as in ‘at’, ‘into’1),
‘postpositions’ (which follow the constituent they govern, such as gratiā ‘for the sake
of ’), and words that may either precede or—in certain combinations—follow, such as
cum ‘with’. A blanket term sometimes used for these words is ADPOSITIONS. In this
Syntax, unless specified otherwise, ‘preposition’ will be used to cover both preposi-
tions and postpositions. For the purposes of this chapter ‘case’ will mean a ‘bare’ case,
i.e. one not governed by a preposition. ‘Preposition’ will be used to refer to the
combination of the preposition proper and the case or cases that are associated
with it. Cases and prepositions mark the syntactic and/or semantic function of a
constituent in its clause or phrase.
Cases and prepositions belong to the class of relators, to which subordinators and
agreement markers also belong. Subordinators are discussed in detail in Chapter 14.
Agreement is discussed in Chapter 13. Cases, prepositions, and subordinators share a
number of properties. In the first place, all three mark constituents at the clause or
sentence level (arguments and satellites) and at the noun phrase and adjective phrase
levels. For each type of relator, examples of marking of arguments are (a)–(c); of
satellites (d)–(f ); and of modifiers of noun phrases (g)–(i) (for adjective phrases, see
§§ 11.91–11.99).2
1
For very rare exceptions, see TLL s.v. in 805.32ff.; s.v. gratia 2237.34ff.
2
Examples (a)–(f ) are taken from LSS: § 5.1.
At the noun phrase level a similar line of argumentation can be given: the semantic
relation between the two nouns that together form a noun phrase is largely deter-
mined by their lexical meanings: whereas the relation between patris ‘father-genitive’
and domus ‘house’ in the phrase domus patris is most likely one of ownership: ‘the
house of the father’, this is not the case in amor patris ‘love for the father’ (it may also
be ‘love from the father’) (see § 11.46 for details).
The semantic contribution of prepositions looks at first sight straightforward; one
might think of a preposition like sine ‘without’ or extra ‘outside’ (which is also used as
an adverb with a clear meaning). However, the diversity of the uses of prepositions,
certainly for the most frequently used ones, is considerable.3 This is illustrated by
(c)–(e). The de-phrases in (c) and (d) are satellites, the one in (e) the third argument
with accusabant. In (c), de vehiculo is understood in a spatial sense ‘down from the
carriage’, an interpretation of de which is likely because the use of vehiculum implies
that the subject was in a higher position. The phrase might be given in answer to a
question unde? ‘from where?’ In (d), de prandio is understood in a temporal sense
(the corresponding question would be quando? ‘when?’). This interpretation of de
‘immediately after’ (OLD s.v. de § 4) is only possible with nouns meaning a point of
time or with event nouns, as in this case: ‘taking lunch’. In (e), de veneficiis cannot be
3
See the discussion in LSS: § 5.3.3. For the syntactic functions of de phrases, see Berdasco (2009).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In this Syntax, the relation of a nominal constituent to the verb and the meaning of
the constituent (or constituents) itself are regarded as essential elements for the
description of the system of cases and prepositions as a whole. Cases and prepositions
are regarded as sources of supplementary information on top of the lexical informa-
tion provided by the noun or noun phrase and other components of the clause,
especially verbs and adjectives.4 (NB: cases and prepositions can also be used to mark
non-finite clauses as argument or satellite at the clause, the noun phrase, and the
adjective phrase levels, in the same way as they mark nominal constituents.)
4
In the words of Touratier (1995: 313) cases ‘n’appartiennent pas au niveau de la première articula-
tion’. For the idea that cases supply supplementary information, see also Blake (2001) for case systems in
general and Schøsler (2008, 2012), and many earlier publications, on Old French (in comparison with
Latin).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
6.7
11.3
32
12.2
16.2
21.6
Figure 12.1 The overall frequency of cases and prepositions in a number of Latin
texts (in percentage)
grammatical system (as with accuso in the previous section), and by the author’s
individual preferences, texts diverge considerably in the use they make of cases and
prepositions and in the use of individual cases and prepositions. The aggregate use,
based on an analysis of c.6000 nouns (including pronouns) and noun phrases and
prepositional phrases in a corpus of various texts, can be seen in Figure 12.1.5
The corpus used for this analysis (see below) has its limitations. So the conclusions
drawn below for the system as a whole cannot be taken as absolute statements, but
they are nevertheless not without importance. The corpus naturally does not produce
instances of all the individual usages of cases and prepositions that are attested in the
entire literature that we possess.
In order to analyse the corpus, a number of decisions had to be taken to be able to
work in a unified way. Very important for the overall picture is the decision to take
implicit nominative subjects into account. These are especially important in Plautus
and in the letters of Cicero (first and second person subjects). In this way the number
of overt nominatives (1,379) was augmented with 578 non-expressed subjects. For
the purpose of this chapter, nominative subjects of passive clauses are counted as first
arguments. Another important decision was to regard the nominative and accusative
case of subject and object complements as due to agreement and not to independent
case assignment. Forms of address, including vocatives, are left out of account
because they are extra-clausal constituents (see § 12.22). Locative forms are left out
5
Figure 12.1 is taken from Brightbill and Kosch (2012). For similar exercises, see LSS: § 5.2.1. For
statistical data about Early and Classical Latin, see also Serbat (1996a: 116–17, 254, 435).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
of account, because in this corpus they are very infrequent. The texts taken into
consideration are the following.
Prose: Cic. Mil. 1–15; Att. 1.1–5; Caes. Gal. 1.1–10; Nep. Att. 1–9; Plin. Nat.
2.1–28; Pereg. Part I, Part II;
Poetry: Pl. Capt. 1–210; Verg. A. 1.1–183; Hor. Carm. 4.1–7; Ov. Ars 1.1–198; Sen.
Ph. 1–235.
The overall distribution of the cases and prepositions can be seen in Figure 12.1. It
shows, among other things, that the dative is the least-used relator of nominal
constituents (although in some texts it is more frequent than in others) and that
prepositions play an important role, coming next in frequency after the accusative.
The same data can be viewed from the perspective of the constituents at the various
levels. This results in the following. At the clause level, first arguments are the most
numerous (more than one third). Almost one quarter are adjuncts, followed by
second arguments (more than one fifth). Third arguments are less than 5%. Subject
and object complements not marked by a nominative or an accusative through
agreement are very rare. This is illustrated by Figure 12.2.
0
1
24
35
22
Figure 12.2 Types of constituent at the clause level in a number of Latin texts (in
percentage)
The use of cases and prepositions in each of these positions at the clause level is as
follows:
(i) The first argument is marked by the nominative (91%) or the accusative
(9%)—in the accusative and infinitive clause.
(ii) The second argument is marked by an accusative (81%), a preposition (9%), a
dative (6%), an ablative (4%), or a genitive.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
100
80
60
40
20
0
arg. 1 arg. 2 arg. 3 adjunct disjunct complement
Figure 12.3 The use of cases and prepositions at the clause level in a number of
Latin texts
(i) At the noun phrase level the genitive predominates (89%), followed by
prepositions (8%), the ablative (less than 3%), and the accusative and
dative.
(ii) At the adjective phrase level the ablative (38%) and the dative (31%) are the
most common, followed by the genitive (23%) and prepositions (8%). There is
an occasional accusative.
It was observed above that there is considerable variation between texts. Three points
may illustrate this variation. (i) It is well known that in poetry prepositions are
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
relatively unpopular.6 This of course has influence on the frequency with which ‘bare’
cases are used. It is also the case that in poetry adjuncts are less frequent than in prose.
(ii) The accusative and infinitive is relatively uncommon in poetry, so there are not
many accusative first arguments. (iii) Noun phrases consisting of two nouns are
relatively uncommon in poetry. As a consequence, genitives are relatively infrequent.
The differences are shown in Figure 12.4.
nom.
35
30
25
prep. 20 gen.
15
10
5
0
abl. dat.
acc.
prose poetry
Figure 12.4 The average distribution of cases and prepositions in poetry and prose
(in percentage)
6
For Virgil, see Görler (1982: 72–3). For general observations, see Denooz (1988).
7
See LSS: § 5.2.2.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
If one leaves the prepositions out of account, the structure of the case system becomes
even clearer. This is visualized in a slightly different way in Figure 12.5.
100
80
60
40
20
0
arg.1 arg.2 arg.3 adjunct noun phrase
Figure 12.5 The use of cases at the clause and noun phrase levels in a number of
Latin texts (in percentage)
This picture shows an almost ideal case system. The clause level and the noun phrase
level are nicely separated with the genitive being almost entirely reserved for the noun
phrase. Adjuncts are separated from arguments by the use of the ablative. The first
argument and the second argument are always distinct, except in the case of the
accusative and infinitive (see below), and if there is both a second and a third
argument they are also distinct from each other, the second argument in that case
(with a few exceptions—see below) being an accusative.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
For a correct interpretation of the figures and the generalizations given above, one
must be aware that the distinction between arguments (especially third arguments)
and adjuncts is not quite as easy as the figures suggest (see § 4.3). Secondly, the term
‘case’ is used in a narrow sense (see the beginning of this chapter). Finally, the case
forms themselves are sometimes ambiguous: in the plural, dative and ablative forms
of nouns are never distinct; they are also not distinct in the second declension
singular in Classical and later Latin; in the course of time the distinction between
singular accusative and ablative forms of nouns disappears. The fact that in the plural
there are only four (and for the neuter only three) distinct cases means that the
relatively clear distinction between third argument and adjunct in Figure 12.5
disappears.
There are two types of exceptions to the generalization that with three-place verbs the
second argument is an accusative and the third argument distinctly marked. The first
type of exception is a few verbs that govern two non-accusatives such as interdico
alicui aliqua re (dat. + abl.), as in (a), which can be explained as being a merger of two
frames (see § 4.86).
(a) Verum si quis est qui etiam meretriciis amoribus interdictum iuventuti
putet . . .
(‘However, if there is anyone who thinks that youth should be forbidden affairs even
with courtesans . . . ’ Cic. Cael. 48)
The second type of exception consists of verbs that govern two accusatives, as in (b).
Statistically this type of exception concerns less than 1% of all third arguments.
Various subtypes of verbs that govern two accusatives must be distinguished, of which
the most important are rogo aliquem aliquid ‘to ask someone for something’ and
doceo aliquem aliquid ‘to teach someone something’, and ditransitive compound
verbs with a preverb such as traduco (transduco) excercitum flumen ‘to lead the army
across the river’, which can be regarded as the causative counterpart of transeo ‘to
cross’, which also governs an accusative argument. Interestingly, for both types of
exceptions other frames are attested that do conform to the generalization. For details,
see §§ 4.71–4.79.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In finite clauses the first argument (if explicit) is marked by a nominative. Examples
are (a)–(c).The nominal part of a passive or deponential finite verb form is also
marked by the nominative, like mortui in (b), and so is the subject complement, like
mater tua in (c) (see Chapter 13 on agreement).
(a) Carthaginienses fratres patrueles duo / fuere . . .
(‘There were two first cousins from Carthage . . . ’ Pl. Poen. 59–60)
(b) Mirum quin ab avo eius aut proavo acciperem, qui sunt mortui.
(‘Odd that I didn’t get it from his grandfather or great-grandfather, who have died.’
Pl. Trin. 967)
(c) Cedo, / si vera memoras, quae fuit mater tua?
(‘Tell me, if you speak the truth, who was your mother?’ Pl. Cur. 642–3)
Appendix: From the end of the fourth century AD onwards a few accusative forms are
found in non-literary, among them substandard, texts as the marker of the subject of
clauses with a one-place, a passive two-place, or a copular verb—that is, in contexts
in which the precise case form is irrelevant for the understanding of the clause. It is a
marginal phenomenon, with many textual uncertainties.8 Examples are (d)–(f ).
If expressed,10 the subject of an accusative and infinitive clause is, as the name of the
construction reflects, in the accusative. Examples are (g)–(i). In (g), the entire
sentence se . . . dies, corresponds to quid in the preceding sentence and functions as
the object of dixit. Se . . . credere itself governs an accusative and infinitive clause as
well. In (h), the accusative and infinitive clause functions as subject with dictum est.
In (i), scalpellum . . . posse is the subject with manifestum est; in (j), me . . . amari of
8
See Väänänen (1981: 115–16); Pinkster (1985c: 168). For the fullest discussion, with references, see
Adams (2013: 239–49). For instances of the use of the accusative for subject and subject complement
constituents in the Casae litterarum piece in the Corpus Agrimensorum (pp. 310–38L), see Josephson
(1950: 176–82).
9
The introductory passage or title is missing in most manuscripts.
10
For unexpressed subjects in accusative and infinitive clauses, see § 14.8.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
tempus est. For further details of the accusative and infinitive construction, see
§§ 15.91–15.109.
(g) Quid dixit? # Si tu illum solem sibi solem esse diceres, / se illum lunam
credere esse et noctem qui nunc est dies.
(‘What did he say? # That if you told him that the sun up there is the sun, he’d believe
that it’s the moon and that what’s now day is the night.’ Pl. Bac. 699–700)
(h) In hac habitasse platea dictum’st Chrysidem . . .
(‘I was told that Chrysis lived on this street . . . ’ Ter. An. 796)
(i) . . . scalpellum adhiberi sine pernicie non posse manifestum est.
(‘ . . . then it is clearly impossible that the knife be employed without injury to the
patient.’ Cels. 7.20.2)
(j) Sed iam tempus est me ipsum a me amari, quando ab illis nullo modo
possum.
(‘But now it’s time for me to be loved by myself since they won’t love me whatever
I do.’ Cic. Att. 4.5.3)
Although in instances like (g) the use of the accusative for the subject of the accusative
and infinitive clause prevents confusion with the nominative of the governing clause,
there is the risk of confusing the accusative subject with the accusative object in the
accusative and infinitive clause. The risk of ambiguity was real enough to be signalled
by Quintilian, as is shown by (k).
(k) Vitanda in primis ambiguitas, non haec solum, de cuius genere supra dictum
est, quae incertum intellectum facit, ut ‘Chremetem audivi percussisse
Demean’, sed illa quoque quae etiam si turbare non potest sensum in idem
tamen verborum vitium incidit, ut si quis dicat visum a se hominem librum
scribentem. Nam etiam si librum ab homine scribi patet, male tamen
composuerit, feceritque ambiguum quantum in ipso fuit.
(‘Above all, ambiguity is to be avoided, not only ambiguity of the kind discussed
above which makes understanding uncertain—“I heard that Chremes Demea
struck”—but also that which, although it cannot confuse the sense, falls into the
same verbal fault. So if you were to say, for example, “I saw a man a book writing”,
although it is obvious that the man is writing the book, it will be a bad piece of
composition, and you will have made it as ambiguous as you could.’ Quint. Inst.
8.2.16)11
The question of why the accusative is used has received considerable attention in the
literature.12 There are various historical and structural explanations. One common
historical explanation is the following: the accusative marks the object with three-
place verbs that may have a prolative infinitive for the third argument, as with
11
On this passage, see Roberts (1912). For ways to avoid ambiguity, see § 14.8.
12
An extensive discussion of the various proposals can be found in Lavency (2003b: 125–35).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
admoneo ‘to advise’ in (l). This pattern was copied for verbs like dico. Of course,
synchronically, se in (g) above is not the object of dixit: the whole accusative and
infinitive clause is object. In fact, with dico, an addressee (in the dative) may be used as
well, like illi in (m) (NB: the subject of the accusative and infinitive clause eam is
coreferential with illi). This analogical reasoning obviously does not apply to (h)–(j),
which then must be regarded as further developments.
(l) Nonne te . . . Quinta illa Claudia aemulam domesticae laudis in gloria mu-
liebri esse admonebat . . .
(‘Did not even the famous Quinta Claudia admonish you to rival our family’s glory in
the splendid achievements of its women?’ Cic. Cael. 34)
(m) Immo ipsus illi dixit conductam esse eam / quae hic administraret ad rem
divinam tibi.
(‘Yes, he himself told her that she was hired to assist you here in your sacrifice.’ Pl.
Epid. 417–18)
There are several other historical explanations. (i) sentio eum venire ‘I notice that he
is coming’ is a combination of sentio eum ‘I notice him’ and sentio venire ‘I notice
coming’. Sentio eum : venire became ‘reanalysed’ to sentio : eum venire.13 (ii) The
infinitive was originally a direction expression. Volo te venire ‘I want you to come’
derives from ‘I want you toward coming’. This development was stimulated by the
parallelism of doceo te litteras ‘I teach you letters’ and ‘doceo te sapere ‘I teach you
being wise’.14 (iii) Dico te bonum ‘I call you good’ was reanalysed as dico te bonum
esse ‘I say that you are good’.15 (iv) The accusative reflects the fact that (in most cases)
the accusative and infinitive clause functions as object.16 Hettrich (1992, 1997) shows
that in Greek and Latin formally more or less identical expressions have resulted
from different developments.17
One structural explanation uses the concept of RAISING. The subject of the subordinate
clause is said to be ‘raised’ or ‘promoted’ to a position in the main clause and to
receive the formal marking of an object in the accusative case in the absence of a
normal nominal object: dico tu venis > dico te venire. This reasoning resembles to
some degree what is actually the case with pseudo-objects, as in (n), the difference
being that in (n) the subordinate clause is finite (for pseudo-objects, see § 9.17).18
13
For this explanation, see Blatt (1952: 320–1).
14
For this explanation, see Sz. 344, 353. To some extent along the same lines is Heberlein (1998).
15
For this explanation, see Hahn (1950: 12); Calboli (1978: 209, 2002); Torrego (1987: 79).
16
This is a slightly different formulation than that of Kurzová-Jedličková (1989). See also Fugier (1989:
140–1, 1998: 349–51).
17
For the historical development, see also Coleman (1985).
18
For this explanation, see Maraldi (1983). For counter-arguments, see Bolkestein (1979).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Another structural explanation is that in the Latin case system the accusative is the
most likely non-nominative case to use: it is the most frequent case and the one that is
maximally used for marking arguments. Also, with few exceptions, there is no further
accusative constituent in the main clause.19 Exceptions are (o) and (p), with the three-
place verbs admoneo ‘to remind that’ and doceo ‘to inform’. Note in (o) the corefer-
entiality of te . . . te.
(o) Tantum te admonebo, si illi absenti salutem dederis, praesentibus te his
daturum.
(‘I’ll merely remind you that in granting life to the absent Ligarius you will grant it to
all these here present.’ Cic. Lig. 38—NB: following most edd. and the g mss.)
(p) Docuerunt enim me periti homines . . . ad Volusium traferri nomen a Valerio
non potuisse.
(‘Experts informed me . . . that the debt could not be transferred from Valerius to
Volusius.’ Cic. Fam. 5.20.3)
In finite clauses the second argument is always distinct from the first argument (for
accusative and infinitive clauses, see § 12.5; for copular clauses, see §§ 4.91–4.97).
There are two questions to be answered concerning the fact that several cases and
prepositions are used. The first is how to account for the individual ‘deviant’ non-
accusative marker (for example, why is the second argument of noceo ‘to harm’ in the
dative, whereas with laedo ‘to injure’ it is in the accusative). The second is how to
account for the fact that with some verbs the second argument may be either accusative
or non-accusative (for example, with metuo ‘to be afraid’ both the accusative and the
dative can be used). As for the latter question, it is shown in § 4.46 that the difference in
marking with metuo corresponds to a difference in meaning: accusative ‘to be afraid of ’,
dative ‘to be afraid for’. This brings us back to the first question.
There are also verbs where the variation dative/accusative seems to be free, for
example with the verb adulor ‘to flatter’, which is found governing dative plebi
‘people’ in Liv. 3.69.4, accusative plebem in Liv. 23.4.2. According to Quint. Inst.
9.3.1 there is a diachronic difference: the accusative is older (and Ciceronian). For
moderor, see § 4.46.
Apart from the alternation between accusative and non-accusative, there are also
cases of fluctuation between two non-accusatives, sometimes within the same mean-
ing class. This is illustrated for verbs of abundance and lacking where fluctuation
exists between genitive and ablative. The verb abundo is (with two poetic exceptions
19
For this explanation, see Pinkster (LSS § 5.2.4. fin. and § 7.6.2). See also Calboli (1996). For a
hypothesis in a formal framework, see Mellazzo (2005).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
in the genitive) used with an ablative for the second argument. Careo is similar (with
a few exceptions in the genitive or, more rarely, the accusative). With egeo the
ablative predominates, but there are instances with the genitive. With the adjective
egenus the genitive predominates, which is also much more common with the
adjective plenus ‘full’.20 With the trivalent verb compleo ‘to fill’ the earliest instances
have a genitive; with privo ‘to deprive of ’ the ablative is the norm (a single exception).
Diachronically the ablative increases. An obvious semantic explanation is lacking.
From a systematic point of view the genitive, as the noun phrase case par excellence
(see § 12.21), seems more problematic than the ablative at the clause level as an
alternative of the accusative.
Several attempts have been made to find a common denominator for the non-
accusative cases used with two-place verbs.21 What the second arguments of these
verbs have in common is that they are affected objects (and not effected objects) (see
also § 4.23 for other common properties). The second common feature is that
(personal) passivization is excluded. It has been proposed that (in the terminology
used in this Syntax; see § 2.9) the use of an accusative object is typical for + control
verbs, the use of a non-accusative for – control verbs. However, this is not a sufficient
explanation for all the verbs involved (in both directions).22 It has also been suggested
that what the non-accusative verbs have in common is that they lack the semantic
value of the accusative, which is described as ‘transitive’.23 But transitivity is not a
clear semantic concept either. An argument against assuming a semantic difference is
the fact that with verbal nouns derived from these verbs, no such distinction is made
in the cases of nouns that depend on them: such adnominal arguments are normally
in the genitive, just as with adnominal arguments depending on verbal nouns related
to accusative-governing verbs. Thus the adnominal argument eius with favorem in
(a), the verbal noun derived from dative-governing faveo ‘to be well-disposed
(towards)’, is in the genitive just like eius with amor, derived from accusative-
governing amo ‘to love’, in (b) (for details, see § 11.71).
(a) Erant autem magnae partis hominum ad favorem eius inclinati animi . . .
(‘Moreover, the sympathies of a large proportion of the people were inclined to
favour him . . . ’ Liv. 42.5.2)
(b) Flos veteris vini meis naribus obiectu’st. / Eius amor cupidam me huc prolicit
per tenebras.
(‘My nostrils have been offered the flowery scent of old wine. Love for it is driving my
eager self out here through the darkness.’ Pl. Cur. 96–7)
20
For statistical data about the distribution of the gen. and abl. with plenus, see TLL s.v. 2407.23ff.
21
For critical discussion, see Pinkster (1989). For the opposite position, see Moralejo (1996).
22
For such an approach, see Heilig (1978), with discussion in LSS: § 5.2.4 (d). For faveo and iuvo
and the problem of case variation in general, see also Álvarez Huerta (2010b).
23
For this characterization of the accusative, see Ebeling (1957: 135–6).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Table 12.1 The use of the same case in different but semantically related categories
Case Adjective Two-place verb Three-place verb
Dative vicinus ‘cognate’ haereo ‘to cling to’ iungo ‘to fasten to’
Ablative inanis ‘devoid of ’ vaco ‘to be devoid (of)’ privo ‘to deprive of ’
The preference for a specific case also manifests itself in the case of antonyms.
This is illustrated in Table 12.2, which also contains both two-place and three-place
antonyms.
That Roman speakers did indeed associate specific meaning classes with a certain case
(or preposition) appears from the fact that sometimes the second argument of a
‘hardcore’ accusative governing verb is used with a non-accusative case, as with iuvo
‘to help’ + dat. in Late Latin (especially medical texts), on the analogy of prosum ‘to be
good (for)’ and its antonym noceo ‘to harm’.24 The use of the dative for the second
argument with some compound verbs can also be explained by assuming that on the
basis of their meaning they fit in with meaning classes with which the dative was
common.25 The opposite tendency, to treat all second arguments in the same way by
using the accusative, was much stronger, as the Romance languages show (for details,
see § 12.31).26
In conclusion, a semantic explanation for the use of the dative, ablative, and
genitive cases to mark second arguments is difficult to arrive at. Their use seems to
24
See TLL s.v. 747.74ff. See also Adams (2007: 438) on the use of the dative with adiuto by the
freedman Niceros in Petr. 62.11.
25
For the use of the dative with compound verbs, see Pinkster (2011).
26
For the two principles mentioned in the text, see García Hernández in several studies, including
(2002).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
27
For the attention devoted by ancient grammarians to the relationship between specific cases and
semantic classes of verbs, see Rosellini (2010).
28
Examples taken from Woodcock (1959: 42).
29
TLL s.v. condono 157.1 makes a distinction between ‘cuius gratia’ and ‘cui’.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
30 31
See the explanation in Woodcock (1959: 1–2, 41). See K.-St.: I.307–8.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
reficiendas in (c). Many of the nouns that function as object of utor, including nouns
referring to human beings, could not function as instrument adjunct at all.32
(a) . . . saltasse eum commode scienterque tibiis cantasse.
(‘ . . . that he danced gracefully and played the flute with skill.’ Nep. praef. 1)
(b) . . . nec vero qui fidibus aut tibiis uti volunt ab haruspicibus accipiunt earum
tractationem, sed a musicis.
(‘Again, persons who want to learn to play on the harp or on the flute take lessons,
not from a soothsayer, but from a musician.’ Cic. Div. 2.9)
(c) . . . quae gravissime adflictae erant naves, earum materia atque aere ad
reliquas reficiendas utebatur . . .
(‘ . . . he utilized the timber and bronze of the ships which had been most severely
damaged to repair the rest . . . ’ Caes. Gal. 4.31.2)
Here, too, the reasoning given above is intended by some in a diachronic sense
(the instrument was ‘absorbed’ by the verb), by others also in a synchronic sense
(see also § 12.11).
32
For a discussion of the use of the ablative for second arguments and the inadequateness of
explaining it as an ‘instrument’, see Burkard and Schauer (2000: CIX–XXI).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Out of the 117 prepositional arguments in the corpus, thirteen are not related to verbs
of (figurative) location and movement. Exx. (c) and (d) show arguments with cum
and pro (more examples in § 4.38). Comparable with verbs of lacking like careo,
which govern a bare ablative, is tempero + ab + abl. in (e) (more examples in § 4.40).
(c) . . . Helvetii . . . fere cotidianis proeliis cum Germanis contendunt . . .
(‘ . . . the Helvetii . . . struggle in almost daily fights with the Germans . . . ’ Caes. Gal.
1.1.4)
(d) (sc. Milo) semper pro bonis contra improbos senserat.
(‘ . . . his sympathies had always been on the side of patriots against agitators.’ Cic.
Mil. 5)
(e) Neque homines inimico animo . . . temperaturos ab iniuria et maleficio
existimabat.
(‘Nor did he believe that men of unfriendly disposition . . . would refrain from outrage
and mischief.’ Caes. Gal. 1.7.5)
12.11 The use of cases and prepositions for the third argument
The case par excellence for third arguments is the dative. It is the normal form to mark
recipient arguments with verbs of transfer (and their opposites) and addressee
arguments with verbs of communication. Examples are (a)–(c). More examples are
given in § 4.52.
(a) Tacitus conscripsit tabellas, / obsignatas mi has dedit. / Tibi me iussit
dare . . .
(‘Quietly he wrote this letter, sealed it, and gave it to me. He told me to give it to
you . . . ’ Pl. Bac. 984–5)
(b) Aurumque ei ademit hospiti . . .
(‘And he took away the gold from his guest . . . ’ Pl. Mos. 481)
(c) Eadem istaec verba dudum illi dixi omnia.
(‘I said exactly all the same words to him a while ago.’ Pl. Bac. 1018)
The ablative is common with verbs of treatment or supplying (with) and their
opposites, as in (d) and (e). More examples are given in § 4.53. With some verbs
both the ‘transfer’ and the ‘supply’ frames are possible, among them instruo ‘to equip
with’, as in (d) and (f ). More examples are given in § 4.59.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The ablative is also used with verbs that require a position or a source argument.
Examples are (g) and (h). More examples are given in § 4.81 and § 4.83, respectively.
(g) . . . exercitum equitatumque castris continuit.
(‘ . . . he kept the army and the cavalry in camp.’ Caes. Gal. 2.11.2)
(h) Tuum consilium vehementer laudo quod non prius exercitum Apollonia
Dyrrachioque movisti . . .
(‘I strongly commend your prudence in not moving your army from Apollonia and
Dyrrachium . . . ’ Cic. ad Brut. 1.2)
As to the explanations given for the use of the ablative with these types of verbs, the
most common way is to identify them with one of the uses of the ablative as a marker of
adjuncts. In the case of instruo in (d), the ablative is usually considered instrumental
(for instrument adjuncts, see § 10.53). Many of the nouns found as third arguments
with instruo can indeed function as instrument adjunct, such as pecunia in (i). In
addition, instruo in the sense of ‘to equip’33 can be used without an indication of the
equipment involved, as in (j) and (k), which suggests that it is a two-place verb.34
(i) Corrupisse dicitur A. Cluentius iudicium pecunia . . .
(‘Aulus Cluentius is charged with having corrupted the court with money . . . ’
Cic. Clu. 9)
(j) Iam ubi liber ero, igitur demum instruam agrum atque aedis, mancupia . . .
(‘As soon as I’m free, I’ll organize a farm and a house and slaves . . . ’ Pl. Rud. 930)
(k) . . . instituit accusatores, instruit testis . . .
(‘ . . . she organizes the prosecution and marshals the witness . . . ’ Cic. Clu. 18)
Nevertheless, from the semantic point of view the ablative constituents are not
really instruments. Instances like (j) and (k) can better be explained as ‘absolute’
use: in (j), for example, the ager will be provided with ‘the usual things’ and the testis
in (k) received ‘the usual instruction’.
33
This is part II of the TLL s.v. 2017.46ff., which more or less corresponds to §§ 5–8 of the OLD.
34
For more examples, see OLD s.v. § 6 and § 8.c.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Diachronically one might want to say that constituents that were originally instru-
ment adjuncts got ‘absorbed’ by what was originally a two-place verb frame and
so increased the number of arguments to three.35 There seem to be no restrictions
on the meaning of the third argument of instruo. A few examples may illustrate
this: . . . tamen facile declaratur, utrum is qui dicat tantum modo in hoc declamatorio
sit opere iactatus an ad dicendum omnibus ingenuis artibus instructus accesserit.
(Cic. de Orat. 1.73); Itaque naves omnes quas paratas habuerant ad navigandum,
propugnatoribus instruxerunt . . . (B. Alex. 10.4); Ibi eum parata instructaque remi-
gio excepit navis. (Liv. 33.48.2); Iphicrates, dux Atheniensium, classem suam hostili
habitu instruxit . . . (Fron. Str. 4.7.23);
The ablative portorio in (e) is usually identified with the so-called separative use of the
ablative (the ablative proper), but semantically this is counter-intuitive: the portitorem
in the accusative is rather the person from whom the portorium is taken away. It is
therefore better to explain the use of the ablative with this and related verbs in the
same way as with their antonyms (see Table 12.2).
The use of the ablative for place and source arguments is usually identified with the
(locative and separative) use of the bare ablative to mark position and source adjuncts
(for these, see § 10.4 and § 10.14, respectively). However, in the case of verbs
indicating position or movement it is not always easy to decide whether an ablative
constituent in the clause is an argument or an adjunct.
For the rare use of the genitive for third arguments, see §§ 4.61–4.67.
Prepositions are much less frequent as markers of third arguments than the dative
and the ablative. Unlike the situation with the second argument, where Caesar and the
Peregrinatio were found to have a much higher percentage of prepositions than
the other texts (§ 12.10), there is no such variation between texts with the third argument.
Prepositions sometimes mark spatial arguments, as in (l), but most of them are figurative,
like those in (m)–(o). The direction arguments with some verbs can also be marked by the
dative, as with affero. For more examples of spatial third arguments, see §§ 4.80–4.84.
Ex. (p) has an associative third argument. For more examples, see § 4.69.
(l) . . . optata potiuntur Troes harena / et sale tabentis artus in litore ponunt.
(‘ . . . the Trojans gain the welcome beach and stretch their brine-drenched limbs
upon the shore.’ Verg. A. 1.172–3)
(m) (sc. Rhenus) . . . qui agrum Helvetium a Germanis dividit . . .
(‘ . . . which separates the Helvetian territory from the Germans . . . ’ Caes. Gal. 1.2.3)
(n) Etsi vereor, iudices, ne . . . minime . . . deceat . . . me ad eius causam parem
animi magnitudinem adferre non posse . . .
(‘Although I am afraid, gentlemen of the jury, that . . . it is in the last degree
unbecoming . . . that I should be unable to bring to his case a greatness of spirit
equal to his . . . ‘ Cic. Mil. 1)
35
For the idea of ‘absorption’, see Serbat (1981). See also Hettrich (1990) on comparable expressions in
the Rgveda.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
With some verbs the third argument may be marked by either a bare case form or a
preposition that governs the same case form. See § 12.28.
Prepositions are used slightly more frequently than bare cases to mark adjuncts (see
Figure 12.3). Among the bare cases the ablative is by far the most frequent (see
Figure 12.5). In the corpus the dative is only used for beneficiary adjuncts, as in (a)
and (b) (for more instances, see § 10.70). The accusative can be used for direction
adjuncts, as in (c) (especially for towns—see § 10.17) and for extent of space and time
adjuncts, as in (d) and (e) (for more instances, see § 10.19 and § 10.32, respect-
ively). The genitive is not used to mark nominal adjuncts. The ablative is used for
most of the types of adjuncts distinguished in Chapter 10, with the exception of
beneficiary adjuncts (dative) and direction adjuncts (accusative). For extent of space
and time adjuncts it competes with the accusative (see § 10.19 and § 10.32,
respectively). The adjuncts marked by the ablative are almost always inanimate:
the exception is soldiers being used by their commander to perform a certain
action, as in (f ) (for more instances, see § 10.54). The adjuncts marked by the
dative are for the most part animate (but for an inanimate beneficiary, see (b)).
Prepositions in general can be used for all sorts of adjuncts, animate or inanimate.
Certain individual prepositions can be used in a variety of adjuncts (see at the end
of this section).
(a) Sed si repperero, o Fides, / mulsi congialem plenam faciam tibi fideliam. / Id
adeo tibi faciam. Verum ego mi bibam, ubi id fecero.
(‘But if I do find it, dear Good Faith, I’ll offer you a six-pint pot filled to the brim with
honey-wine. Yes, I’ll offer it to you; but I’ll drink it for myself as soon as I’ve done so.’
Pl. Aul. 621–3)
(b) Vita, quae fato debetur, saluti patriae potissimum solvatur.
(‘Above all, let my life, which I owe to fate, be spent for the salvation of my country.’
Rhet. Her. 4.55)
(c) Aetolos pacem velle. De ea re oratores Romam profectos.
(‘That the Aetolians wanted peace. That concerning that matter envoys left for
Rome.’ Cato orat. 130)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Since the ablative can be used for a wide range of adjuncts, the question arises how
one can establish the semantic function of a specific NP, or—less technically—how
does one know what an NP in the ablative means in its context. For example, in the
case of the morphologically unambiguous noun phrases in (g), what clues are there
that these ablatives are being used to indicate ‘when’ Pompeius’ actions took place
(that is that they function as position in time adjuncts)?
(g) Ita tantum bellum . . . Cn. Pompeius extrema hieme apparavit, ineunte vere
suscepit, media aestate confecit.
(‘Thus, so great a war . . . was organized by Gnaeus Pompeius at the end of
winter, started at the beginning of spring, and finished by the middle of summer.’
Cic. Man. 35)
The negative clues are that (i) the verbs are all two-place verbs. Besides the explicit
subject, they have an explicit object that makes perfect sense with these verbs and is in
the case form that they require. (ii) The ablatives are not governed by a preposition or
by another type of constituent. Therefore, the ablative constituents must be optional
elements (satellites) that specify in some way the actions referred to. The positive clue
is that the nouns refer to a period of the year. Consequently, the noun phrases must
specify in some way the period ‘at’, ‘for’, or ‘within’ which the actions took place.
Since the actions involved are neither durative nor terminative, the most likely and
satisfying interpretation is ‘at’ or ‘during’, that is, as position in time adjuncts. The
preposition in + abl. can be used in this sense, but it is very rarely used with the nouns
involved. Ex (h) is attested in some of the manuscripts of Vitruvius and adopted by
editors.36 In the case of plural case forms like aestatibus, which are morphologically
ambiguous, it is easy to interpret them as ablatives functioning as position in time
adjuncts, and not as dative beneficiary adjuncts. In aestatibus is not attested.
36
For this temporal use of in, see TLL s.v. 778.46ff. (especially 778.57–8) and OLD § 35a,b.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(h) Cum enim in hieme anni sereno caelo in ea (sc. praesepia) traducuntur
matutino boves, ad solem pabulum capientes fiunt nitidiores.
(‘For when in the winter while the weather is fair the cattle are led up for their
morning feed, they become more sleek by eating their fodder in the sun.’ Vitr. 6.6.5)
The more transparent and precise the meaning of a noun or noun phrase in the
ablative is, the easier it is to establish its role in its clause and the less need there is for a
preposition to specify its role. Nouns referring to an entity that is often used as an
instrument, such as gladius ‘sword’ and culter ‘knife’, are marked by the bare ablative
when they function as adjunct in their clause. The preposition per + acc. is attested for
the first time with such nouns in (i). By contrast, from Early Latin onwards abstract
nouns like iocus ‘joke’ and dolus ‘trickery’ are more often marked by per, as in (j), than
by the bare ablative, as in (k), when functioning as adjunct.37
(i) Ubi iam idem quacumque varices sunt factum est, uno loco adducta per
hamulum vena praeciditur . . .
(‘When the same thing has been done wherever there are swellings, at one place the
vein is drawn forward by the hook and cut away.’ Cels. 7.31.3)
(j) Derides, qui scis haec [iam] dudum me dixisse per iocum.
(‘You’re mocking me; you know that I said this as a joke a while ago.’ Pl. Am. 963)
(k) An id ioco dixisti? Equidem serio ac vero ratus.
(‘You said it as a joke? I thought you said it in earnest and seriously.’ Pl. Am. 964)
The meaning of the noun or noun phrase in the ablative is not the only important clue
to understanding it; the verb and its arguments may also provide information that
makes the interpretation of an optional constituent in the clause easier. This is shown
by a closer examination of price adjuncts. Typical examples of this class of adjuncts
are (l)–(o).38
(l) Quia de illo emi virginem / triginta minis . . .
(‘Because I bought a girl from him for thirty minas . . . ’ Pl. Cur. 343–4)
(m) . . . conduxit in Palatio non magno domum.
(‘ . . . he took a house on the Palatine at a moderate rent.’ Cic. Cael. 18)
(n) . . . Alabandensem Apollonium, qui cum mercede doceret, tamen . . .
(‘ . . . Apollonius of Alabanda, who, though teaching for hire, nevertheless . . . ’ Cic. de
Orat. 1.126)
(o) . . . Flaccus . . . locavit autem omnem (sc. agrum) frumento . . .
(‘ . . . Flaccus . . . moreover leased them all in return for grain . . . ’ Liv. 27.3.1)
37
See Pinkster (1990) and Luraghi (2010: 54–7, 87–9). Muhr (1971) shows that Sallust rarely uses the
preposition per in instrument expressions with concrete nouns (e.g. per arma once versus armis—almost
twenty instances).
38
This section follows LSS: § 5.2.4 (f ).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The adjuncts in (l)–(n), certainly (l) and (n), by their meaning could hardly be
interpreted in any way other than as price adjunct. This is the more likely in (l) and
(m), because the verbs refer to commercial activities. This is not the case in (n), but
the meaning of mercede is sufficient for the correct interpretation. Ex. (o) shows the
opposite situation: frumento by its meaning does not necessarily refer to some sort of
commercial compensation, but locavit refers to a commercial activity, and there is
little room for another interpretation than as ‘price adjunct’. The preceding expos-
ition is summarized in Table 12.3.
Interaction of the meaning of the verb (and its arguments) with the meaning of the
noun or noun phrase in the ablative is necessary for a correct interpretation of the
semantic function of the noun or noun phrase especially in the case of so-called
process adjuncts (see § 10.42). In these cases normally the first argument is an agent
that has control over the state of affairs and often is also intentionally involved.
The semantic function of adjuncts marked by a preposition is on the whole easier
to determine, which comes as no surprise, since the degree of semantic specialization
of even the c. fifteen most common prepositions (out of about seventy-five preposi-
tions in total) is considerable. In order to illustrate the different role of prepositions
and bare cases in marking adjuncts, examples (p)–(v) with the noun amor ‘love’ may
be useful. In (p), amore is the cause of the person’s perishing from love for the young
lady, comparable to the use of the bare ablative in passive clauses (see § 2.12 and
§ 5.8). Although in (q)–(t) the relation between amor and the states of affairs is also in
some way causal, they are different from (p). In (q) the insanity of the person is the
result of amor. In (r)–(t), amor is the reason for the person’s behaviour. In (u), per
amorem is a means adjunct: ‘by showing, practising my feelings of love’.
A prepositional phrase with ab + abl. ‘because of ’, ‘as a result of ’ (OLD s.v. § 15.a)
seems to come closest to the bare ablative, as in (v).40
(p) Is amore misere hanc deperit mulierculam . . .
(‘He’s miserably pining away with love for this woman . . . ’ Pl. Cist. 131)
39
‘Verbs’ is a simplification. The arguments may play a role as well.
40
Note that without ab the noun phrases may be interpreted as means adjuncts.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
With the more frequent prepositions the contribution of the meaning of the noun or
noun phrase is essential for a correct understanding. For per, for example, Lodge41
distinguishes eight different types of adjunct, among which are the following:
(i) de loco, with nouns referring to a locality, as in per urbem ‘through the
town’;
(ii) de tempore, with nouns having a temporal meaning, as in per tempus
‘at a suitable time’;
(iii) indicat instrumentum, with a distinction between personam, as in per
nuntium ‘by messenger’ and per rem, as in per dolum ‘through a trick’;
(iv) vi modali, with abstract nouns, as in per pacem ‘in peace’;
(v) vi causali, with abstract nouns, as in per metum ‘out of fear’.
Within these categories considerable variation exists between the nouns involved,
and the assignment of particular instances to one of these types is debatable (as Lodge
indicates). This is a well-known methodological problem. In the end one has to
decide between simply listing all occurrences and trying to organize them into a
limited number of types. See § 10.1.42
However, many prepositions, among them relatively recent developments such as causa
and gratia ‘for the sake of ’ and merito ‘thanks to’ (rare—from Tertullian onwards) signal
very precisely what the contribution of the prepositional phrase to the clause is.
41 42
See Lodge s.v. per. For the methodological problem, see Torrego (1996a: 206).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
In the corpus used for this chapter disjuncts marked by a bare case or a preposition
are very rare. In § 10.102 and in § 10.107 examples are given of the use of the dative
for disjuncts, the so-called dativus iudicantis, as in (a), and the so-called dativus
ethicus, as in (b). An example of the ablative is (c) (for more examples, see § 10.101).
Prepositional phrases are (d), with prae (for more examples, see § 10.103), and (e),
with sine ‘without’ (for more examples, see § 10.100). Example (a) in particular
contains very few, if any, positive clues for its correct interpretation.
(a) Iam lauta es? # Iam pol mihi quidem atque oculis meis.
(‘Are you washed now? # Yes, at least to myself and my own eyes.’ Pl. Truc. 378)
(b) At tibi repente paucis post diebus, cum minime exspectarem, venit ad me
Caninius mane.
(‘Well, you’re not going to believe it, but a few days later Caninius arrived at
my house early in the morning out of the blue, when I was least expecting him.’
Cic. Fam. 9.2.1)
(c) Nam meo iudicio pietas fundamentum est omnium virtutum.
(‘For in my opinion filial affection is the basis of all virtues.’ Cic. Planc. 29)
(d) Pithecium haec est prae illa et spinturnicium.
(‘She’s a little monkey, a little ugly duckling compared with her mistress.’
Pl. Mil. 989)
(e) Etenim sine dubio, iudices, in hac causa ea res in discrimen adducitur.
(‘For undoubtedly, gentlemen, that issue is being brought to a decision in this case.’
Cic. Ver. 1.6)
As Figure 12.5 shows, the genitive is the case that is predominantly used at the noun
phrase level to mark nominal constituents as modifier of another NP. The genitive
makes no contribution of its own to the understanding of the relationship between the
modifier and its head noun: it simply indicates that one is the modifier of the other. The
semantic relation between the two constituents is entirely determined by the meaning
of each and by the communicative situation (see § 11.46 for several illustrations).
In broad terms the constituent in the genitive may ‘define, describe, or classify’43
43
The wording is Woodcock’s (1959: 50).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
the entity the head noun refers to. Nevertheless, the genitive is not the only case
and there are prepositions to mark modifiers as well. At the same time, the genitive
is used as a marker of third arguments, second arguments, subject and object
complements, and also at the adjective phrase level. How can this diversity be
understood?
As for modifiers at the noun phrase level, a distinction is made in Chapter 11
between adnominal arguments and optional modifiers. Adnominal arguments are
usually in the genitive, but sometimes the case or preposition used at the clause level is
maintained. Early Latin expressions like hanc rem curatio ‘interference in this matter’
(an accusative as with the object in a clause with the verb curo) illustrate the point
(see § 11.71). A genitive (huius rei curatio) would be interpreted in the same way,
as a so-called objective genitive. With optional modifiers, however, the situation
is different. They may have a satellite-like relation to the noun they modify.
Examples are:
(a) curatio manu ‘treatment with the hand’ (ablative, an instrumental relation—
see § 11.64)
(b) pabulum ovibus ‘fodder for sheep’ (dative, a beneficiary relation—see
§ 11.65)
(c) supplicatio triduum ‘a prayer during three days’ (accusative, an extent of time
relation—see § 11.66)
Whereas in (a) and (c) the head nouns are verbal nouns with which the modifiers
have the same relation as they would have at the clause level, a clausal equivalent
is less obvious in (b), yet it is easy to understand. In none of these cases could the
genitive, if possible at all, be used to express these relationships in the same
unambiguous way. The same holds for the far more frequent use of prepositional
phrases as optional modifiers of nouns, like navis ex Rhodo ‘a ship from Rhodes’
(see the note in § 11.67). All these instances of ‘deviant’ marking at the noun
phrase level can be explained as manifestations of the PRINCIPLE OF TRANSCATEGORIAL
PARALLELISM: the use of the same expression for the same semantic relation,
independently of the category to which the structure in which it is used belongs
(see below).
Appendix: The genitive is much less predominant at the adjective phrase level than it
is at the noun phrase level. (See also § 12.2.) This is not surprising. (i) Many of the
adjectives involved govern arguments in the same case or with the same preposition
as semantically related verbs: egenus ‘deprived of ’, ‘in need of ’ + gen./abl. ~ egeo ‘to
lack’, ‘to need’ + gen./abl., another manifestation of the principle of transcategorial
parallelism. (There are no adjectives that govern an accusative argument.) (ii)
Optional constituents at the adjective phrase level usually have the form that
semantically equivalent satellites at the clause level would have. Nevertheless, there
is a tendency to use the genitive for all sorts of semantic relations at the adjective
phrase level. For details, see § 11.92.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The preceding sections have shown that it is difficult to assign a unique meaning to
the individual cases. As discussed in § 12.1, cases add information about how the
noun or noun phrase they mark is to be understood in its clause in addition to the
lexical meaning already conveyed by the noun/noun phrase itself and the verb (+ its
arguments). If the noun or noun phrase is an argument, the contribution of the case
by which it is marked to the understanding of the clause as a whole is modest: the
semantic relations are usually determined by the verb. If, on the other hand, the noun
or noun phrase is a satellite, it is the meaning of the noun that is decisive for its
interpretation, although for process adjuncts the meaning of the verb and its argu-
ments is not unimportant. Nevertheless, the cases (not governed by a preposition)
add some sort of ‘global’ information, as indicated in § 12.3. An accusative form, for
instance, creates the expectation in the hearer that the word or phrase so marked
functions as second argument of a verb, an expectation that may turn out to be wrong
and may have to be revised when more information becomes available. An ablative
(most often marking an inanimate entity) will initially be taken as a signal that the
noun or noun phrase contains subsidiary information. The availability of such
information licenses a greater flexibility in the positioning of nouns and noun phrases
(and non-finite clauses) in their sentence.
The different roles of the cases when marking arguments and satellites have been
noted by several linguists who worked on the Latin case system or case systems in
general, notably by Kuryłowicz (1949). In his view, some cases are predominantly
syntactic (the accusative, for example, which is typically a marker of second argu-
ments), whereas others are predominantly semantic (the ablative, for example). In
principle, all cases have both syntactic and semantic values, but in different propor-
tions. Other linguists have tried to find one meaning for each case, which necessarily
leads to a high level of abstraction, so for example Hjelmslev (1935–7).44 Indo-
Europeanists try to formulate one semantic value for each of the cases that can be
reconstructed. The accusative is, for example, regarded as indicating ‘goal’. However,
unless interpreted in an abstract sense, this semantic value is not really useful in
explaining the most common use of the accusative as a marker of the second
argument, and some form of ‘desemantization’ must be assumed from the earliest
periods onwards.45
Exercises by Latin scholars to find one central meaning for each case have led to
similar abstract results.46 A few tendencies can be discerned in descriptions of the
44
For the history of case theories, see Calboli (19752), Serbat (1981, 1996a: 3–26), Viparelli (1993), and
Blake (2001). For a critical discussion of the structural and functional approach to case theory, see
Torrego (1996a).
45
For a discussion of the semantic and syntactic uses of the accusative in the double accusative
construction from the Indo-European perspective, see Hettrich (1994), also (1990).
46
See, for example, Serbat (1996a: 26).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
cases by Latinists. In the first place, grammars tend to attach (mostly semantic) labels
to the names of the cases that correspond to the various uses of those cases. For the
ablative, K.-St. have no less than fourteen different labels, such as ablativus loci,
ablativus separativus, ablativus instrumenti, and ablativus qualitatis.47 These labels
may be understood as indications of the semantic function fulfilled by the noun or
noun phrase in the ablative, but the implication is that the ablative is providing the
semantic information. As indicated above, in the case of the ablative it is rather the
meaning of the noun or noun phrase that is decisive for the interpretation of its role in
the clause.
Secondly, for those cases for which multiple uses are distinguished, grammars tend
to offer common denominators or to indicate the ‘chief function’ of a particular case.
Both approaches are not without problems. This can be illustrated by the way K.-St.
and Woodcock deal with the dative.48 K.-St. make a distinction between two types of
‘objects’, one type that ‘completes’ (ergänzt) the meaning of the verb and another that
optionally ‘determines’ (bestimmt) it. This distinction resembles that between argu-
ments and satellites. As examples of the first type they give among others: scribo
epistulam ‘I am writing a letter’ (accusative) and philosophia medetur animis ‘phil-
osophy heals minds’ (dative). Note that both verbs are two-place verbs. Then, further
on, the accusative is said to be of the first type (completion). It is the ‘direct object’.
The dative, by contrast, ‘indicates a more precise determination of the action the
clause is about, the entity the action bears upon or in consideration of which it takes
place, the indirect or more remote object . . . ’.49 The example used to illustrate this is,
surprisingly, omnes homines natura libertati student. In the chapter on the dative,
K.-St. add that the dative constituent is actively involved in the action of the clause
and counteracts the action of the subject. Among the examples are tibi impero ‘I order
you’, legibus pareo ‘I obey the laws’, and tibi servio ‘I serve you’. This description of
the meaning of the dative is quite enigmatic and difficult to apply in practice. The
examples would perfectly illustrate objects of the first type (compare animis above).
Woodcock writes: ‘The chief function of the dative is to denote the person or, more
rarely, the thing for whose benefit (or the reverse) the action of the verb is performed,
or for whose benefit a thing or a state exists, or whom the quality of a thing affects: e.g.
Librum mihi dat. “He gives a book to me.” Est mihi liber. “There is a book for me.” (or
“I have a book.”) Donum mihi suavissimum, “a gift most acceptable to me”.’ The three
descriptions and related examples are quite diverse, and to speak of one chief function
requires some imagination. Nevertheless, the first type (with do) is the most common
47
For a discussion of K.-St.’s approach, see LSS: § 5.2.6 (b).
48
K.-St.: I. 251–5, 307; Woodcock (1959: 38–40).
49
‘Der Dativ bezeichnet die nähere Bestimmung der im Satze ausgesprochenen Handlung, den
Gegenstand, dem die Handlung gilt oder mit Rücksicht auf den sie vollzogen wird, das mittelbare oder
entferntere Objekt . . . ’ (: I.253). A more detailed description is given at I.307: ‘Der durch den Dativ
bezeichnete Gegenstand wird als ein Tätiges aufgefasst, das der Tätigkeit des Subjektes gegenübertritt, sich
mit dem Subjekte in tätiger Wechselbeziehung befindet, bei der Handlung gleichsam persönlich wirkend
beteiligt ist.’ In Dutch, the indirect object is called ‘meewerkend’ or ‘belanghebbend’ ‘voorwerp’ (‘cooper-
ating’ or ‘interested’ ‘object’).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
use of the dative, so in that sense ‘chief function’ seems defensible. The main question
is how to relate the chief function to the other six uses Woodcock recognizes. Some of
the difficulties are discussed in § 12.6.
The ablative is the most difficult case for which to find one common denominator.
Grammars agree that the various uses of the ablative can be assigned to three types:
LOCATIVE, SEPARATIVE (or: purely ablative), and INSTRUMENTAL uses. However, there is no
general agreement on how to distribute the individual uses over these three types.
This distinction into three types of uses is usually supported by historical consider-
ations and comparison with other Indo-European case systems. In fact, in the first (-a
stem) and second (-o stem) declensions the locative, separative, and—briefly—the
instrumental were formally distinct in the archaic period: *domin-ei, *domin-ōd,
and domin-ō, respectively. Already in Plautus’ time there remained only one
‘ablative’ form.50
This leaves the question open whether Roman speakers were aware of this earlier
distinction and which clues they had to make such a distinction. After all, the system
had only one case to cover all these uses.
Leaving the vocative out of account, the Latin case system has five cases. In the
plural, however, masculine and feminine nouns have maximally four different forms
(dative and ablative are the same) and in certain declensions only three (with res
‘thing’, for example, accusative and nominative are the same in the plural). Neuter
nouns have only three plural forms. In the singular, most masculine and feminine
nouns have only four forms (in the case of dominus ‘master’, for example, dative and
ablative are the same; with domina ‘mistress’, genitive and dative are the same, which
means that the distinction between clause level and noun phrase level cannot be
formally expressed). Neuter nouns have maximally four forms (fulgur ‘flash of
lightning’), minimally two (cornu ‘horn’), while most have three (donum ‘gift’). In
addition, in some declensions the genitive singular and the nominative plural are not
distinct (with dominus for example). Since most cases (with the exception of the
nominative) can be used at all levels—clause and phrase—one may ask how easy it
was to arrive at the correct interpretation of a clause (and a sentence) and how
ambiguity was avoided.
The first part of the answer is a repetition of what was said above: the backbone of a
clause is formed by the meanings of the words involved. The presence of these words
is also often related to the wider context and the expectations of the participants in the
communicative situation. Another factor is the relative simplicity of the clauses in
terms of the number of constituents: on average a clause has one or two arguments
and one or two satellites. So the number of semantic functions of satellites marked by
the dative and ablative, for example, is rather large, but the chance of having many
such satellites in one clause is small. A third element is that prepositions play a larger
role, especially in marking satellites, than most grammars seem to be aware of. Their
50
For the development, see Porzio Gernia (1974). For the (pre)history of the Latin case system in
general, see Weiss (2009: 194–212).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(ii) The absence of a clear case marking would cause the distinction between word
group level and clause level to disappear, as in (d). Here the noun phrase tres menses,
functioning in the example as an attribute (in the genitive), could also function as a
duration adjunct (in the accusative).
(d) Trium mensum molita cibaria sibi quemque domo efferre iubent.
(‘They commanded every man to take for himself from home a three months’
provision of victuals.’ Caes. Gal. 1.5.3)
51
For a very complete list of factors that contribute to a correct understanding of the arguments in a
clause, see Schøsler (2012, forthc.).
52
For an illustration with the verbs venio ‘to come’ and eo ‘to go’, see Cabrillana (1997a).
53
See LSS: § 5.2.6 (a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The percentage mentioned above seems low, but in reality this is difficult to
determine. Language needs an overdose of information to compensate for all sorts
of disturbance of the perception and digestion of information. For the importance of
the endings of words for a correct understanding, the judgement of Vitruvius in his
handbook on architecture is interesting:
This section deals with the individual cases, including the vocative. It will mostly
consist of references to more detailed discussions elsewhere in this Syntax, supple-
mented with a few uses that do not belong to the clause or the phrase level. This is
especially the case for the nominative and the vocative, considered as casus recti in the
Roman grammatical tradition, as opposed to the other four, the casus obliqui.
Already Varro L. 5.4 uses it for the nominative . . . recto casu quom dicimus inpos . . .
(‘when we say “inpos” in the nominative’). It is the first form one learns, on the basis
of which the other forms can be formed (Var. L. 8.6). The term casus nominandi or
nominativus is also used by Varro (L. 10.23).54
12.17 Nominative
The regular uses of the nominative are as subject in finite clauses (and clauses with a
historic infinitive) (see § 9.2) and as subject complement in finite clauses (see § 9.20).
(See also § 4.96 for so-called nominal sentences.) The nominative is (like other cases)
also used as a marker of Theme and Tail constituents (for the terms see § 2.11). An
example of the former is (a) (for more examples, see Chapter 22); an example of the
latter is (b) (for more examples, see Chapter 22). These constituents have no function
in the clause, but they are closely related to it and together they form a sentence.
Another extraclausal use of the nominative is as a marker of address constituents (see
54
See Serbat (1996a: 29–30).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Chapter 22) for those words for which no specialized vocative form exists (see § 12.22
for details).
(a) Cancer ater, is olet et saniem spurcam mittit . . .
(‘The black ulcer, that has a foul odour and exudes putrid pus . . . ’ Cato Agr. 157.3)
(b) Ibi auditur P. Deci eventus, ingens hortamen ad omnia pro re publica
audenda.
(‘There they learned of Decius’s death, a great incentive to dare everything for the
republic.’ Liv. 10.29.5)
Another type of use consists of independent nominatives, for example, the names of
persons on a painting, such as Ulixes and Thrax55 or the name of a shop, as in (c). In
such cases some scholars speak of a ‘naming’, ‘designating’, or ‘referential’ function of
the nominative.56 A further ‘naming’ use is as title of a book or as chapter heading, as
in Brutus, Orator, and Tusculanae Disputationes (Cicero) and Miltiades (Nepos), but
the book carrying a title, when used in a clause, can be expressed in the normal way, as
the object oratorem meum in (d).
(c) HOSPITIUM / C(AI) HUGINI FIRMI
(‘The inn of Gaius Huginus Firmus’ CIL IV.3779 (Pompeii)—NB: assuming it is a
nominative)
(d) ‘Oratorem’ meum tanto opere a te probari vehementer gaudeo.
(‘I am greatly pleased that my Orator is esteemed so highly by you.’ Cic. Fam. 6.18.4)
Some scholars regard this ‘naming function’ as the basic function of the nominative,
but one might hold that the material context supplies information of the type ‘this
is . . . ’ (so the nominative is there because it marks a subject complement). Something
similar is the case with the names of deceased people on graves, as in (e): they can be
regarded as subjects (‘here lies’) in their material context.
(e) SALVIUS PUER / VIXIT ANNIS VI
(‘(Here lies) the boy Salvius. He lived six years.’ CIL X.1032 (Pompeii))
55
The examples are taken from Serbat (1996a: 33).
56
So Suárez (1994a) and Serbat (1996a: 32).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The nominative is sometimes used in a list that has been announced in the preceding
clause, in another case or in a prepositional phrase, as in (k).
(k) Quid de illis dicam quae certe cum ipso homine nascuntur, linguae solutio,
vocis sonus, latera, vires, conformatio quaedam et figura totius oris et
corporis?
(‘What shall I say of those other attributes which undoubtedly are innate in the man
himself: the ready tongue, the ringing tones, strong lungs, vigour, suitable build and
shape of the face and body as a whole?’ Cic. de Orat. 1.114)
A phenomenon that is sometimes not sufficiently distinguished from the use of the
nominative in theme constituents is illustrated by (l). Here the sentence starts with
two constituents in the nominative. After the parenthesis the sentence continues with
reliquas etiam virtutes in the accusative, as object of continet. This is clearly a failed
construction (an ‘anacoluthon’), for which sometimes the term NOMINATIVUS PENDENS
is used. A few of these failures are even attested in Cicero.
(l) Omnis enim abstinentia, omnis innocentia (quae apud Graecos usitatum
nomen nullum habet, sed habere potest IºØÆ; nam est innocentia
adfectio talis animi quae noceat nemini)—reliquas etiam virtutes frugalitas
continet.
(‘For all abstinence and inoffensiveness (and this with the Greeks has no customary
term, but it is possible to use IºØÆ, harmlessness; for inoffensiveness is a
disposition of the soul to injure no one)—well, “frugality” embraces all the other
virtues as well.’ Cic. Tusc. 3.16)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Writers of Latin with an imperfect knowledge of its grammar sometimes use the
nominative of names instead of an oblique case. This is not uncommon in inscrip-
tions. There are also several examples in the ostraca found at Bu Njem, as in (m).
Further instances are known from the Albertini tablets (end VI AD). Such examples
show that the nominative form was used as a ‘base’ form, and was possibly the first
form for learners of Latin as a second language. As such they were sometimes
inserted automatically in formulaic contexts such as the ostraca and tablets just
mentioned.57
Appendix: Grammars pay some attention to the fact that Latin differs from English
and other modern languages in its treatment of words that are mentioned rather than
used. Often these are typically expressed as appositives and merely marked typo-
graphically, whereas Latin marks them as genitive modifiers, as in vox voluptatis ‘the
term “pleasure” ’ and cognomen Sapientis ‘the surname “Sapiens” ’. Accusatives with
verbs of shouting and saying, on the other hand, are explained as quotations in the
accusative, as in (n) (translated as: ‘he shouted: “Cicero!” ’ by K.-St.) and (o) (‘they
shouted: “Victory” ’),59 but this is not necessary.
Literal quotations in Latin may be of any form, including nominative forms, as in (p)
and (q). In Late Latin this is also attested for verbs like appello ‘to call (something X)’,
as in (r).60
(p) Ut enim ipsa philosophia et multa verba Graecorum, sic sorites satis Latino
sermone tritus est.
(‘For, just as the word “philosophy” and many other words are of Greek origin
and are in general use as Latin words, so it is with sorites.’ Cic. Div. 2.11)
57
See Sz.: 27–30, with references, Väänänen (1981: 115–16), and Adams (1995a: 96–102, 2003b: 236,
2013: 205–15) for more instances and discussion.
58
For the meaning of the word sbitualis, see Marichal (1992: 101). VII is probably a mistake for VIIII.
59
K.-St.: I.254–5. Cf. OLD s.v. exclamo § 3; s.v. clamo § 3f.
60
For further examples and discussion, see Adams (2013: 224).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Nominative nouns and noun phrases in exclamatory sentences (see § 6.35) do not
exactly give names; rather, they are evaluations: nugae! ‘(that is) nonsense’.
Up to this point discussion has focused on nominatives that mark nominal constitu-
ents, but the nominative case may also mark a non-finite clause that functions as subject
of its sentence. An example is (s) (for more examples, see § 15.13). Similarly, it may
mark clauses with an adjective, like gnarus in (t) (for more examples, see § 15.14).
(s) Careo . . . omnibus amicis quorum benevolentiam nobis conciliarat per me
quondam te socio defensa res publica . . .
(‘I lack . . . all the friends whose goodwill I won when I defended the state with you at
my side.’ Cic. Fam. 4.13.2)
(t) Augebat metum gnarus Romanae seditionis et, si omitteretur ripa, invasurus
hostis.
(‘The fact that the enemy was cognizant of the disaffection in the Roman ranks and was
going to invade if the Rhine band was abandoned increased their fear.’ Tac. Ann. 1.36.2)
12.18 Accusative
The accusative is almost entirely limited to the clause level, mostly as a marker of
second arguments, but also of first arguments in accusative and infinitive clauses, and
of third arguments. It can be used as a marker of only a few satellite functions.
Examples of arguments are (a)–(d). Ex. (a) shows the most common use of the
accusative, as a marker of the object. For the accusative as a marker of the subject
of an accusative and infinitive clause, as in (b), see § 12.5. For the use of the accusative
as a marker of a third argument (the so-called double accusative), as in (c), see § 12.4
with further references. An example of the use of the accusative as a marker of
direction arguments is (d) (for more examples, see § 4.43). For the use of the
accusative in exclamations, see § 6.35.
(a) Equidem sana sum et deos quaeso ut salva pariam filium.
(‘I am sane and I ask the gods that I may safely give birth to a son.’ Pl. Am. 720)
(b) Nam meus formidat animus nostrum tam diu / ibi desidere neque redire
filium.
(‘It frightens me that our son’s been sitting there for so long and isn’t coming back.’
Pl. Bac. 237–8)
61
See Fedeli ad loc.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Examples of satellites are (e)–(h). Ex. (e) shows the accusative as the marker of a
direction adjunct. For detailed discussion, see § 10.8 (common nouns) and § 10.17
(topographical names). Exx. (f ) and (g) show local (see § 10.19) and temporal (see
§ 10.32) extension, respectively. In both cases the ablative can be used as well. Ex. (h)
shows the mainly poetic use of the accusative as a marker of a respect adjunct. For
details, see § 10.92. In the remainder of this section this use will be ignored.
(e) Pater ad mercatum hinc me meus misit Rhodum.
(‘My father sent me away from here to Rhodes on a trading trip.’ Pl. Mer. 11)
(f ) . . . edixitque ut ab urbe abesset milia passuum ducenta . . .
(‘ . . . he issued an edict that he keep two hundred miles away from the city . . . ’ Cic.
Sest. 29)
(g) Ibi cenavi atque ibi quievi in navi noctem perpetem . . .
(‘There I had dinner and there I rested the entire night on the ship.’ Pl. Am. 732)
(h) Quin etiam subito vi morbi saepe coactus / ante oculos aliquis nostros, ut
fulminis ictu, / concidit et spumas agit, ingemit et tremit artus, / desipit,
extentat nervos, torquetur, anhelat / inconstanter . . .
(‘Moreover, we have often seen someone constrained on the sudden by the violence
of disease, who, as if struck by a thunderbolt, falls to the ground, foams at the mouth,
groans while his limbs shudder, raves, grows rigid, twists, pants irregularly . . . ’ Lucr.
3.487–91)
It is not easy to discern a common element in the argument and satellite uses of the
accusative. Some scholars regard the use of the accusative in extent adjuncts ((f ) and
(g)) as related to its use for the so-called cognate object (for which see § 4.10),
referring to instances like (i)–(k). Passivization, as in (j) and (k), both poetical, is
generally speaking a good proof of objecthood.62 Note also the variation in (k).
(i) . . . tridui viam progressi rursus reverterunt . . .
(‘They proceeded for a three days’ journey, and then returned . . . ’ Caes. Gal. 4.4.4)
62
For a few more examples, see K.-St.: I.282, n.1. Serbat (1996a: 159–66) also suggests that the use of
the accusative for direction adjuncts is related to its use for the cognate object (‘objet interne’).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(j) Nobis, cum semel occidit brevis lux, / nox est perpetua una dormienda.
(‘For us, when the short light has once set, remains to be slept one unbroken night.’
Catul. 5.5–6)
(k) . . . vixi / annos bis centum. Nunc tertia vivitur aetas.
(‘ . . . I have lived for two centuries; now the third is being lived.’ Ov. Met. 12.187–8)
For some time, the use of the accusative as a marker of direction arguments and
adjuncts was regarded as the basis for its use as a marker of the object (as most cases
were supposed to have developed from ‘concrete’ uses).63 However, most scholars
now agree that there is no common element that unifies these two uses. The most
common explanation is that in prehistoric times two cases merged, one case that
indicated direction or goal (a so-called lativus) and one that marked the object. The
Latin accusative is thought to continue these two uses. This is not a satisfactory
explanation.
This explanation is open to the general objection against the assumption of a merger
(or: syncretism): linguistic signs do not merge unless the meanings have merged. See
§ 7.30 on the Latin perfect tense.64
The position taken in this Syntax is that there is no need to exclude the possibility of
the same form being used for multiple purposes, if the context is sufficiently clear.
Here, too, there is much information that prevents an accusative satellite from being
perceived as the object of its clause or as the subject of an accusative and infinitive
clause. The direction and extent uses of the accusative are characterized by strong
lexical restrictions, both on the meaning of the noun (it denotes place, usually a
place with a certain name, or a measure of space or time) and on the meaning of the
verb.
In Early Latin the accusative is also used at other levels of the clause. It is used at the
noun phrase level as a marker of adnominal arguments (see the example hanc rem
curatio ‘interference in this matter’ in § 12.14), which can be explained as an instance
of transcategorial parallelism. The same principle holds for the use of the accusative as
a marker of extent of place or time at the adjective level, as in (l) and (m), and at the
adverb level, as in (n).
(l) . . . ad hunc modum vasa componito: arbores crassas pedes II, altas pedes
VIIII cum cardinibus; foramina longa pedes III semis, exculpta digitos VI.
(‘ . . . lay off the vats as follows: anchor-posts 2 feet thick, 9 feet high, including tenons;
openings hollowed out 3 ½ feet long, 6 fingers wide.’ Cato Agr. 18.1)
63
For a discussion of nineteenth-century scholarship, see Bennett (1914: II.191–5).
64
For the implausibility of the assumption of a merger, see Serbat (1996a: 159–60); For the merger
approach, see Sz.: 30. For the idea that the object function of the accusative is derived from its direction
function, see Woodcock (1959: 1–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The accusative is rarely used for optional attributes at the noun phrase level (for a few
examples, see § 11.66).
The accusative can also be used as a marker of non-finite clauses functioning as
object. An example is (o) (for more examples, see § 15.133).
(o) Non illi paucitatem nostrorum militum, non iniquitatem loci . . . , non absci-
sum in duas partes exercitum . . . causae fuisse cogitabant.
(‘They did not reflect that the cause of their success had been the small number of our
troops, the unfavourable conditions of the site . . . , the severance of the army into two
parts . . . ’ Caes. Civ. 3.72.2)
In inscriptions from Pompeii isolated nouns in the accusative are used as exclam-
ations or as indications of merchandise (e.g. panem ‘bread for sale’); these uses are
related to the normal uses of the accusative.65 See also the list of dishes in the
accusative in (p).66
(p) Cena haec fuit: ante cenam echinos, ostreas crudas quantum vellent,
peloridas . . .
(‘This was the dinner: as a prelude sea-urchins, raw oysters (as many as they
wanted), cockles . . . ’ Met. Pius in Macr. 3.13.12)
In substandard texts in Late Latin the accusative seems to function as a ‘base’ form
that is also used in contexts in which other cases were required in standard Latin.67
12.19 Dative
As is indicated in § 12.3 (introduction), the dative is more evenly distributed over the
levels of the clause than the other cases. It is used as a marker of the second argument
with essentially four semantic classes of verbs (see § 12.7 with references) and as a
marker of third arguments with verbs of transfer and communication (see § 12.11). In
addition to these uses the dative can also be used to mark the second argument with
the so-called impersonal verb licet, as in (a) (for more examples, see § 4.14). The
dative can also be used as a marker of the possessor with the verb sum ‘to be’, as in (b)
65
For examples, see Väänänen (1966: 116–17).
66
I owe the reference to Jim Adams.
67
For discussion, see Adams (2003b: 62–3).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(for more examples, see § 4.26). The dative is also common with combinations of the
verb sum ‘to be’ and a so-called predicative dative, as in (c) (for more examples, see
§ 9.34). Note that the meaning of the combination resembles that of the verb auxilior
‘to help’, which also governs a dative. Furthermore, the dative can be used to mark the
agent with the passive and the gerundive, as in (d) (the so-called dativus auctoris—for
more examples, see § 5.8).
(a) Amovi a foribus maxumam molestiam, / patri ut liceret tuto illam
amplexarier.
(‘I removed a terrible nuisance from the door so that my father can embrace that
woman in safety.’ Pl. Am. 464–5)
(b) Subduxi ratiunculam / quantum aeris mihi sit quantumque alieni siet.
(‘I’ve done the reckoning of how much money I have and how much debt I have.’ Pl.
Cur. 371–2)
(c) Opsecro vos ego, mi auxilio, / oro optestor, sitis . . .
(‘I beg you, I entreat you, I beseech you: be a source of aid to me . . . ’ Pl. Aul. 715–16)
(d) Adeundus mihi illic est homo.
(‘That man must be approached by me.’ Pl. Rud. 1298)
As for its use with satellites, the dative can be used as a marker of both a variety of
adjuncts (see below) and of two types of disjuncts (see § 12.13). The most common
adjunct function marked by the dative is beneficiary (see § 12.12). The dative is used
in poetry to mark direction adjuncts, as in (e) (for more examples, see § 10.11). The
dative also functions as a marker of another type of adjunct, as in (f ) (the so-called
dativus sympatheticus—for more examples, see § 10.96).
(e) Consurgunt studiis Teucri et Trinacria pubes. / It clamor caelo . . .
(‘Eagerly the Teucrians and men of Sicily rise up; a shout mounts to heaven . . . ’ Verg.
A. 5.450–1)
(f ) Ei mihi, / quom nihil est qui illi homini dimminuam caput.
(‘Dear me! Not to have anything I can smash this person’s head with!’ Pl. Men.
303–4)
At the noun phrase level the dative functions as a marker of both adnominal arguments
and optional attributes. An example of the former is tibi in (g) (for more examples, see
§ 11.71). This is an instance of transcategorial parallelism. The relationship between
integumentum and its optional modifier aetati meae in (h) resembles that of a benefi-
ciary adjunct at the clause level (for more examples, see § 11.65).
(g) Quid mihi scelesto tibi erat auscultatio . . .
(‘What induced a wretch like me to listen to you?’ Pl. Rud. 502)
(h) Istaec ego mi semper habui aetati integumentum meae.
(‘I’ve always considered this to be a shield for my youth.’ Pl. Trin. 313)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
At the adjective level the dative is used as a marker of arguments of a large number of
bivalent adjectives which are semantically close to the classes of verbs mentioned in
§ 12.6, again a form of transcategorial parallelism. Most often these adjectives
function as subject complement with the copula sum. Examples of such adjectives
are adfinis ‘bordering’, amicus ‘friendly’, carus ‘dear’, fidelis ‘faithful’, gratus ‘pleasing’,
morigerus ‘obliging’, opportunus ‘advantageous’, propitius ‘well-disposed’, similis
‘similar’, utilis ‘useful’. Illustrations of adjectives that are modifiers of a noun at the
noun phrase level are (i) and (j) (for further examples of bivalent adjectives, see
§ 4.100). Some of these adjectives are also used with other cases, such as the genitive
(for example, similis ‘similar’).
The adverbs with which the dative is used as a marker of dependent nouns or noun
phrases are semantically related to the adjectives. The most common are adversum
‘opposite’ and obviam ‘in the path (of)’. An example with opportune is (k).
The dative can also be used as a marker of non-finite clauses, both those functioning
as argument and those functioning as satellite. Examples of arguments are (l)–(n). Ex.
(l) contains a participial clause (for more examples, see § 15.133); (m) a gerundial
clause (for more examples, see § 15.136); and (n) a gerundival clause (for more
examples, see § 15.140). For clauses in the dative with bivalent adjectives, see
§ 15.138 (gerundial clauses) and § 15.141 (gerundival clauses).
68
For a defence of the ‘motion towards’ meaning of the dative, see Bennett (1914: I.101–3). For
counter-arguments, see Woodcock (1959: 39–40). Serbat (1996a: 436–7) presents a synchronic variant of
the ‘motion towards’ idea.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
An example of a gerundial satellite clause in the dative is (o) (for more examples, see
§ 16.100); of a gerundival satellite clause (p) (for more examples, see § 16.106).
(o) . . . vitandoque imbres et aestus (sc. parietes) tegebant harundinibus et
fronde.
(‘ . . . (others) covered the walls with reeds and foliage for the purpose of escaping the
rain and heat.’ Vitr. 2.1.3)
(p) Serviendae servituti ego servos instruxi mihi . . .
(‘I have schooled my servants to do me servant’s service . . . ’ Pl. Mil. 745)
12.20 Ablative
The use of the ablative as a marker of second arguments is discussed in § 12.8; its use
for third arguments in § 12.11. The most frequent use of the ablative is as a marker of
satellites (see § 12.3 introduction). The spectrum of semantic relations of ablative
constituents in their clause appears to be unrestricted: ‘the freedom of this case is
immense’.69 How comprehensibility is more or less guaranteed is discussed in § 12.12.
In authors who aim at compactness of information, such as Pliny the Elder, anything
inanimate can be put in the ablative. It is sometimes difficult to classify ablative
constituents using the traditional labels. Examples are (a) and (b).70 Chapter 10
contains a detailed list of uses, which need not be repeated here.
(a) (sc. Delos) Cingitur V̅ passuum, adsurgit Cynthio monte.
(‘It measures five miles in circumference. Its only eminence is Mount Cynthius.’ Plin.
Nat. 4.66)
69
Multa sunt quae Romani per ablativum casum solent efferre: immensa enim est huius casus licentia.
(Diomedes I.310 GLK) See already: Sed ablativo ipsi . . . inest naturalis amphibolia. ‘Caelo decurrit aperto’
utrum ‘per apertum caelum’ an ‘cum apertum esset’. (Quint. Inst. 7.9.10).
70
Nutting (1930) coined the term ‘stenographic ablative’. For Pliny the Elder’s use of the ablative, see
Pinkster (2005b: 244–8).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(b) Ritu<s> naturae hominem capite gigni, mos est pedibus efferri.
(‘It is Nature’s method for a human being to have been born head first, and it is the
custom for him to be carried to burial feet first.’ Plin. Nat. 7.46)
For the use of the ablative as a marker of disjuncts, see § 12.13, with references.
At the noun phrase level, instances of adnominal arguments in the ablative are
hardly attested. An example is usura governing quibus in (c). Instances of optional
modifiers in the ablative with verbal nouns are less rare (see § 11.64). An example of a
modifier which functions like an instrument adjunct at the clause level is (d).
Descriptive modifiers of the type (e) are relatively common (a so-called ablativus
qualitatis) (for more examples, see § 11.63).
(c) (ignavi) . . . quibus, nisi ad hoc exemplum experior, non potest usura
usurpari.
(‘(good-for-nothings) . . . no possible use can be made of you, unless I treat you in this
fashion!’ Pl. Ps. 135)
(d) Quid tibi hanc digito tactio est?
(‘Why do you touch this girl even with a finger?’ Pl. Poen. 1308)
(e) Ibi amare occepi forma eximia mulierem.
(‘There I fell in love with a woman of outstanding beauty.’ Pl. Mer. 13)
Ex. (j) illustrates the use of the ablative as the marker of a participial clause function-
ing as object (for more examples, see § 15.133). Ex. (k) contains a gerundial clause in
the ablative functioning as argument (for more examples, see § 15.136). Gerundival
clauses in the ablative functioning as argument seem not to be attested. Non-finite
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Among the satellite uses of the ablative, non-finite clauses form an important
component. Examples of this use are (l)–(p). Exx. (l)–(n) are instances of the so-
called ablative absolute, (l) with a perfect participle (for more examples, see § 16.88),
(m) with an adjective, (for more examples, see § 16.117), (n) with a noun (for more
examples, see § 16.116). In (o), the ablative forms of the gerund function as cause
adjuncts (for more examples, see § 16.101). Ex. (p) has a gerundival clause in the
ablative (for more examples, see § 16.107).
(l) Victores victis hostibus legiones reveniunt domum / duello exstincto ma-
xumo atque internecatis hostibus.
(‘The enemy conquered, as conquerors our legions return home, now that a mighty
war’s been brought to an end and the enemy’s been exterminated.’ Pl. Am. 188–9)
(m) Cur ego te invito me esse salvom postulem?
(‘Why should I demand to be well against your will?’ Pl. Capt. 739)
(n) Atqui aut hoc emptore vendes pulchre aut alio non potis.
(‘Either you’ll sell her nicely with this man as purchaser or you won’t be able to with
any other purchaser.’ Pl. Per. 580)
(o) Lumbi sedendo, oculi spectando dolent, / manendo medicum dum se ex
opere recipiat.
(‘My loins hurt from sitting, my eyes from watching, from waiting for the doctor to
return from work.’ Pl. Men. 882–3)
(p) . . . aggerundaque aqua sunt viri duo defessi.
(‘ . . . and from bringing us water two men got exhausted.’ Pl. Poen. 224)
12.21 Genitive
The genitive is the case of the noun phrase level par excellence (see § 12.14). As for its
use with two- and three-place verbs, this is restricted to a limited number of semantic
classes and varies over time.71 The genitive is also occasionally found through analogy
71
For the use of the genitive with these verbs in a number of languages see Haspelmath and Michaelis
(2008).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
with verbs which normally govern another case (see § 12.6); these will be ignored in
the present discussion. The two-place verbs with which the genitive is used to mark
the second argument concern personal and impersonal verbs of emotion like misereor
‘to feel pity (for)’and me miseret ‘I feel pity (for)’ (§ 4.35 and see § 4.48) and verbs of
remembering and forgetting like obliviscor ‘to forget’ (see § 4.36). In alternative
expressions with the personal verbs the accusative is used, the normal case for the
second argument. There are no restrictions on the type of entities that are allowed as
second argument with these verbs. For the Roman speaker and hearer the use of the
genitive with these verbs must have constituted an idiom.
The three-place verbs with which the genitive is used to mark the third argument
belong to four semantic classes. The first are verbs of reminding, the causative variant of
the aforementioned verbs of remembering and forgetting, to which the same observa-
tions apply, that is, they were idioms (see § 4.66). A second class are verbs of valuing.
They can be used with a limited class of words as value arguments, the same class that
can be used with two-place sum ‘to be worth’ (see § 4.32), of which these verbs can be
considered causative variants (see § 4.65). The idiomatic character of these expressions
is even more clear. Thirdly, there are two classes of verbs that belong to the juridical
domain (the so-called genetivus forensis or criminis), the first class containing verbs
such as accuso ‘to accuse’, where the genitive phrase denotes the charge (see § 4.63—
also for alternative prepositional expressions, especially with de). The second class
contains verbs like damno ‘to condemn’, where the genitive phrase denotes the penalty
(see § 4.64—also for alternative ablative expressions). In both cases the nouns and noun
phrases functioning as third argument normally belong to a limited class. As with two-
place verbs governing a genitive second argument, for the Roman speaker and hearer
this use of the genitive at the clause level must have constituted an idiom.
The genitive is also used at the clause level as a marker of price adjuncts. This
concerns a very few expressions like pluris vendo aliquid ‘to sell something at a higher
price’ (see § 10.59). Both the genitival expressions and the verbs involved belong to a
small and recognizable class, which makes a ‘price’ interpretation almost inevitable.
The genitive is also used as a marker of non-finite adjunct clauses. Ex. (a) illustrates
the use of a participial clause in the genitive as a third argument (for more examples,
see § 15.133). Ex. (b) contains a similar clause at the noun phrase level (for
more examples, see Chapter 17). Ex. (c) contains a gerundial clause depending on a
bivalent adjective (for more examples, see § 15.138). A corresponding case with a
gerundival clause is (d) (for more examples, see § 15.141).
(a) . . . quo facilius indicio regis Scauri et relicuorum quos pecuniae captae
arcessebat delicta patefierent.
(‘ . . . in order that through the testimony of the king the offences of Scaurus and the rest
who were accused of taking bribes might the more readily be disclosed.’ Sal. Jug. 32.1)
(b) Sibi enim bene gestae, mihi conservatae rei publicae dat testimonium.
(‘He gives witness to his good service to the state, but to my preservation of it.’ Cic.
Att. 2.1.6)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
The question arises whether some unity can be discovered in the diverse uses of the
genitive at the clause level and, if so, how this can be related to the use of the genitive
at the noun phrase level.72
Starting with the clause level, most scholars agree that ‘there is no single sense-
value underlying these . . . uses of the genitive’.73 One term that has been applied to
some of the clause-level uses (and to some of the uses at the adjective level) is ‘genitive
of relation’ (genetivus relationis): the genitive marks the domain within which the
action of the verb takes place. This is regarded as the semantic value underlying the
use of the genitive with judicial verbs and verbs of emotion.74 The term itself is not
very precise, and it is difficult to see why it would not be applicable to the uses of other
cases at the clause level.75
Another term that is used to describe (some of) the uses of the genitive at the clause
level is ‘partitive’, usually meant as a diachronic (prehistorical) explanation. The
genitive with verbs of remembering and forgetting, for example, is sometimes
regarded as partitive. It is clear that ‘partitive’ suffers from the same sort of vagueness
as the term ‘relation’.76
As for the relationship between the uses at the clause and noun phrase level, some
scholars assume that the genitive with judicial verbs can be explained as due to ellipsis
of words such as crimine ‘charge’, iudicio ‘action’, comparing examples like (f )
and (g).77 Although idiomatization in such a well-defined semantic field that more-
over refers to a well-defined part of social life is not unlikely, the use of the genitive
72
The most extensive discussion of the problems involved in identifying the semantic value(s) of the
genitive is in Touratier (1995: 307–13), with references to earlier publications.
73
In Woodcock’s words (1959: 51). Similarly, Touratier (1995). For an Indo-European perspective, see
the discussion in Sz.: 50–1 and Serbat (1992).
74
So Sz.: 75 and 82. The German term is ‘Genitiv des Sachbetreffs’.
75
For critical discussion, see Serbat (1996a: 387ff.).
76
For the verbs of remembering and forgetting, see Sz.: 81. Alongside ‘inclusive’ and ‘ablatival’ uses,
Serbat regards a large part of the use of the genitive at the clause level as a trace of an old partitive, but
‘synchroniquement . . . le g(énitif) part(itif) ne possède aucun signifié syntaxique’ (1996a: 402).
77
So, among others, K.-St.: I.464–5 and Perotti (1987). Discussion in Serbat (1996a: 355ff.).
A historical survey of the various opinions and proposals can be found in Fraile (2013).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
without these words is already firmly established in Early Latin, so the ellipsis would
have to be situated in prehistoric times.
(f ) (sc. Alcibiades) . . . postulabat . . . potius de praesente quaestio haberetur,
quam absens invidiae crimine accusaretur.
(‘Alcibiades . . . begged them . . . to conduct the investigation while he was present,
rather than bring forward in his absence charges inspired by malice.’ Nep. Alc. 4.1)
(g) Atque etiam hoc praeceptum officii diligenter tenendum est, ne quem
umquam innocentem iudicio capitis arcessas.
(‘Again, the following rule of duty is to be carefully observed: never arraign someone
who may be innocent on a capital charge.’ Cic. Off. 2.51)
In sum, attempts to arrive at one semantic explanation for all the uses of the genitive
end in abstract notions which are difficult to falsify.78
12.22 Vocative
The vocative is an exceptional case in that it only exists in part of the nominal system
(details below). Where it is lacking, the nominative is used instead. This is shown in
(a)–(c). In (a), stulte is the singular masculine vocative form of stultus. In (b), stulta is
the singular feminine nominative form; in (c), stulti is the plural masculine nomina-
tive form. In all three examples the word or words are not a constituent of the clause,
although they may be incorporated into the clause, as in (b). Their extraclausal status
was probably marked by intonation, as in modern languages.79
(a) O stulte, stulte, nescis nunc venire te?
(‘O you poor poor fool, don’t you know that you’re being sold now?’ Pl. Bac. 814)
(b) Quid, stulta, ploras?
(‘What are you crying for, stupid woman?’ Pl. Cur. 520)
(c) O stulti, febrem creditis esse?
(‘Fools, do you believe this is fever?’ Mart. 2.40.8)
The function of these constituents is ADDRESS (for which see Chapter 22), directed at
one or more participants in the communicative situation, as in (a)–(c), or—as an
invocation—directed at a god or the gods, as in (d)–(e). The vocative is limited to this
function of address.
78
So, for example, the notion of dependence (‘dépendance’) on either a noun or a verb (‘sans être
directement gouverné’) in Fugier (2001). Also the notion ‘extraction’ in Serbat (1996a: 423–5).
79
For the extraclausal status of vocative expressions, see Vairel-Carron (1981b), Panhuis (1986), and
Serbat (1996a: 103–4). For the opposite view, see Fugier (1985). For a discussion of the literature, see
Cabrillana (1996).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
80
For later developments and references, see Sz.: 25–6. For Virgil’s use of the vocative, see Görler
(1985: 265).
81
See Dickey (2000b). The sequence ěě, except in words like creet (subjunctive of the verb creo ‘to
procreate’) is rare generally.
82
See Dickey (2000a).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Supplement:
I, puere, pulta / atque atriensem Sauream, si est intus, evocato huc. (Pl. As. 382–3);
Ere, salve. (Pl. As. 619); Bone serve, salve. (Pl. Bac. 775); Cupite atque exspectate /
pater, salve. (Pl. Poen. 1261–2); Orator est, Marce fili, vir bonus, dicendi peritus.
(Cato inc. 14); Salve, exoptate gnate mi. (Pl. Capt. 1006); Dixit Abraham ‘Deus
providebit sibi victimam holocausti, fili mi.’ (Vulg. Gen. 22.8); Caesar, ades voto,
maxime dive, meo! (Ov. Tr. 3.1.78); Dic mihi, dive Claudi . . . (Sen. Apoc. 10.4);
83
For a very thorough discussion, see Svennung (1958: 246–86).
84
The vocative (ocelle) auree would yield a form that is avoided (see note 81). For further examples
and discussion of the ‘nominative of apposition’, also in non-literary texts, see Adams (2013: 216–20).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Prepositions
From the end of the first century AD, the use of the nominative instead of a vocative
increases.
In African non-Latin or mixed inscriptions the vocative of Latin proper names is
used regardless of the function of the proper name in its context, thus functioning as
an uninflected ‘base’ form. An example in a fourth-century Latino-Punic inscription
from Sirte is (m).
12.23 Prepositions
As was said in the introduction to this chapter, the term ‘preposition’ is used as a
general term for ‘prepositions’ (by far the majority), ‘postpositions’, and words that
can be used in both positions (in certain circumstances). Details about their position
are given in Chapter 23. Prepositions form a unit with the constituent they govern:
the PREPOSITIONAL PHRASE. In (a), the two prepositional phrases in patriam and ad
patrem can both function as an answer to a question containing the interrogative
adverb quo ‘to which place?’, and they can both be substituted for by the anaphoric
adverb eo ‘to that place’. The fact that both phrases are direction adjuncts depends
crucially on the meaning of the prepositions. Neither of the two constituents of a
prepositional phrase can be omitted without making the remaining part ungrammat-
ical (for a seeming exception, see § 12.28). Prepositions normally govern nominal
constituents, as in (a), but they can also be used with non-finite clauses, as in (b);
occasionally (in poetry) with (substantival) finite clauses, as in (c);86 and with
infinitives, as in (d) (for more examples, see § 11.8).
(a) Ergo animum advortas volo / quae nuntiare hinc te volo in patriam ad
patrem.
(‘Well then, I’d like you to pay attention to what I want you to report home to my
father from here.’ Pl. Capt. 383–4)
(b) Mox hercle vero, post transactam fabulam, / argentum si quis dederit, ut ego
suspicor, / ultro ibit nuptum . . .
(‘But soon, after the completion of the play, I suspect if anyone gives her money she’ll
marry him willingly . . . ’ Pl. Cas. 84–6)
85
For the use of vocative forms instead of nominative forms in Latin and Punic, see Adams (2003b:
512–15, 2013: 211). For the use of the vocative instead of the nominative in a number of languages, see
Stifter (2013), especially pp. 55–9.
86
For more instances of per, see TLL s.v. 1158.33ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(c) Quod te per superos et conscia numina veri, / per si qua est quae restet
adhuc mortalibus usquam / intemerata fides, oro . . .
(‘But I beseech you, by the gods above, by the powers that know the truth, by faith, if
there is any which still remains unstained anywhere for mortals . . . ’ Verg. A.
2.141–3)
(d) . . . ut inter optime valere et gravissime aegrotare nihil prorsus dicerent
interesse . . .
(‘ . . . resulting in their claiming that there was absolutely no difference between being
in excellent health and being gravely ill . . . ’ Cic. Fin. 2.43)
When two prepositions govern the same noun or noun phrase, they can be coordin-
ated, as in (e), but the phenomenon is very rarely attested. The regular situtation is
illustrated in (f ), with repetition of the relative pronoun.87 In (g), extra is usually
taken as the adverb.
(e) . . . dum in XV diebus ante et post brumam, ut pleraque, ne facias.
(‘ . . . provided you do not do this, or in fact most things, during the fifteen days
preceding and following the solstice.’ Var. R. 1.35.2)
(f ) Materiam enim rerum, ex qua et in qua omnia sint, totam esse flexibilem et
commutabilem . . .
(‘ . . . that matter, from which and in which all things exist, is in its entirety flexible
and subject to change . . . ’ Cic. N.D. 3.92)
(g) Sed tamen et in corpore et extra esse quaedam bona.
(‘Yet still some bodily and external things are good.’ Cic. Fin. 2.68)
For coordination of prepositional phrases sharing the same preposition, see Chapter 19.
Prepositions are used in Latin from the earliest period onwards. It is generally
assumed that many of these prepositions go back to earlier adverbs. This explains
why a number of them correspond to preverbs in compound verbs: ab, ad, cum, de,
ex, in, ob, per, for example.88 Others have clearly different origins, such as causa ‘for
the sake of ’ and fini (fine) ‘as far as’ (grammaticalized ablative forms of causa ‘reason’
and finis ‘boundary’),89 adversus ‘against’ (grammaticalized perfect participle of
adverto ‘to turn towards’), adversum (with the same origin, but probably initially an
accusative neuter form, functioning as a manner adjunct—see § 10.45),90 excepto
87
For instances of coordination with et, see TLL s.v. 872.6f.; von Nägelsbach and von Müller (1905:
536).
88
For a discussion of Latin prepositional syntax from an Indo-European perspective, see Coleman (1991).
89
OLD s.v. finis § 3 takes fini/e as a ‘quasi-preposition’.
90
For the grammaticalization path of versus, see Fruyt (2011: 247–9).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Prepositions
‘with the exception of ’ (in Late Latin, developed from an ablative absolute construc-
tion), and una ‘together’ (unclear adverbial formation, from unus ‘one’, used as a
preposition in Late Latin).91 In the course of the development of Latin many old
prepositions disappeared and new ones were formed.92 A remarkable formation are
compound preposition/adverbs such as abante ‘in front of ’ and depost ‘behind’ (both
found from the end of the fourth century onwards, notably in Christian texts).
A number of prepositions are homonymous with adverbs.93 There is much vari-
ation among the individual lexemes in the balance between prepositional and adver-
bial use and there is also diachronic variation. Whereas most lexemes in their
adverbial use function as adjuncts at the clause level, like contra ‘in return’ in (a),
ad ‘approximately’ functions as a degree adverb with numerals at the noun phrase
level, as in (b) (similarly, circa).
(a) Edepol me uxori exoptatum credo adventurum domum, / quae me amat,
quam contra amo . . .
(‘I really believe my wife will be waiting eagerly for my arrival. She loves me and
I love her in return.’ Pl. Am. 654–5)
(b) Ad quadraginta fortasse eam posse emi minimo minis.
(‘Perhaps she can be bought for around forty minas at the lowest.’ Pl. Epid. 296)
In some cases there may be doubt whether one is dealing with an adverb governing a
noun phrase in a certain case form or a preposition. An example of an adverb
governing a noun phrase (in the accusative or genitive, later even in the ablative)94
is (c). In the end the criterion whether the word in question is an adverb or a
preposition is whether it is normally used without a case form or not.
(c) . . . cum scias esse horum neminem qui nesciat te pridie Kalendas Ianuarias
Lepido et Tullo consulibus stetisse in comitio cum telo . . .
(‘ . . . when you know that there is not one man present who does not know that on
the last day of December, in the consulship of Lepidus and Tullus, you were standing
in the comitium armed with a weapon . . . ’ Cic. Catil. 1.15)
The following lexemes function as preposition and as adverb: ad (see above),
adversus / adversum (also exadversum), ante, post, pone, secundum, contra, circa,
circum, prae, prope, propter, iuxta, intra, supra, infra, subter, subtus, super, citra,
ultra, trans. Mostly adverbial are simul, procul, coram, clam, palam, usque, foras,
foris.95
Appendix: The accusative Kalendas Ianuarias in (c) is difficult to explain. Even more
remarkable is the combination ante diem tertium Kalendas Ianuarias.96
91
See Sz.: 272.
92
On the development of prepositions in the Romance languages, see Brea (1985).
93
On the relationship between prepositions and adverbs, see Pinkster (1972: ch. 9).
94
For the case forms attested, see TLL s.v. pridie 1231.32ff.
95 96
This list is based on K.-St.: I.575–8. See Ramos (2007b).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
With a few exceptions, Latin prepositions govern the accusative case, the ablative case,
or both (with a difference in meaning). The genitive is only regular with a few
prepositions that developed from ablative forms of nouns, such as the postposition
causa ‘for the sake of ’ (see § 12.27) with which the noun in the genitive originally
functioned as a modifier. The dative is not used with prepositions, except as an
ungrammatical variant, sometimes under Greek influence. (For a proper understand-
ing of the use of the genitive and dative with prepositions in Late Latin texts, it is
useful to note that Greek has prepositions that govern these cases.) The nominative
and the vocative are not governed by prepositions. The data concerning the most
common individual prepositions are presented in Table 12.4. Diachronic variations
exist. In Late Latin many accusative governing prepositions are used with the ablative,
and the reverse situation is even more common (due also to the loss of distinctive
formal properties in the singular forms).97
97
For details and references, see Sz.: 214–85; see also Väänänen (1981: 112) on the development of the
accusative as the prepositional case ‘par excellence’.
98
See OLD s.v. secus2 § B.
99
For the history of the word, see Haverling (1988: 45–8); for putative instances in Cicero and
Quintilian, see Haverling (1989).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Prepositions
the presence of ’, prae ‘in front of ’, pro ‘in front of ’, ‘on behalf of ’, procul
‘far away from’, se (sed) ‘without’ (only in a few legal idioms), simul
‘together with’ (from Sal. onwards), sine ‘without’, tenus ‘as far as’ (also
with gen. (and acc.?100))
Accusative or clam ‘without the knowledge of ’, in ‘in’, ‘into’, sub (desub) ‘under’, ‘up to’,
ablative subter ‘at (to) a position below’, super ‘above’
Genitive beneficio ‘because of ’ (from Silver Latin onwards),101 causa ‘for the sake
of ’, ergo ‘on account of ’, fini (fine) ‘as far as’ (rarely abl.), gratia ‘for the
sake of ’, merito ‘because of ’ (from Tert. onwards)
There is much discussion about the relationship between prepositions and cases. The
common explanation is that originally, in prehistoric times, the adverbial predeces-
sors of the prepositions were added to constituents in the accusative or ablative case,
as some sort of specification of the meaning relation of that constituent to its clause.
In the course of time the relationship was reanalysed and the former adverb
understood as forming a unit with the noun. This can be expressed as in Figure 12.6.
In a way the original situation resembles the relation between the adverb intus and
the position in space adjunct in the ablative totis aedibus in (a) (for such cases of
specification, see § 10.3 and elsewhere in Chapter 10). Further phenomena referred
to in support of the ‘specification hypothesis’ are separation of the preposition from
its noun or noun phrase (so-called TMESIS), as in (b),102 and inversion of the order of
noun (phrase) and preposition, as in (c). Both phenomena, however, are relatively
rare and highly stylized. (For details concerning inversion of prepositions, see
Chapter 23.)
100 101
See Kleywegt ad V. Fl. 1.538. See TLL s.v. beneficium 1888.53ff.
102
For the problems with tmesis, see Löfstedt (1961: 276–93), also for adverbialization of prepositions.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Whatever the merits of this explanation concerning the origin of at least some
prepositions, from the earliest documents onwards prepositions cannot be con-
sidered as specifications of cases: the prepositions govern a case (or two cases), and
the reason why it is that specific case is difficult to explain on the basis of the uses of
that case without a preposition (see also § 12.26).103
This distinction corresponds to the fact that the bare ablative can be used to mark
position in space adjuncts (see § 10.4) and the bare accusative to mark direction and
goal adjuncts (see § 10.8). The preposition in might therefore be regarded as a
specification of the meaning of the cases. Note also the difference in meaning between
ad and in in (c). However, several prepositions that govern the accusative mark
position in space adjuncts, for example ante in (e). Ad can be used both for direction
adjuncts, as in (c), and for position in space adjuncts, as in (f ), without a difference in
case form.104
103
For discussion, see Pinkster (1972: ch. 9); LSS: § 5.3.2. For the relationship between preposition and
noun in a prepositional phrase (government or modification?), see Luraghi (2010: 89–93), with references.
104
See OLD s.v. § A and § C, respectively.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Prepositions
The conclusion is as stated in the note of § 12.25. When these prepositions govern
the ablative they are mostly used with states of affairs that denote ‘position’; when
they govern the accusative mostly movement (in a certain direction) is involved.
The difference in case form was therefore largely redundant, even when the accusative
and ablative forms were still clearly distinct (see §§ 12.30–12.34 for the diachronic
changes).
105
See Callebat (1968: 239–42).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
libero, as can be seen in (a) and (b). Ex. (a) illustrates the bare ablative expression, ex.
(b) the preposition a + ablative.
(a) Q. Fabius Maximus consul . . . febri quartana liberatus est in acie.
(‘The consul Quintus Fabius Maximus was rid of a quartan ague in action.’ Plin. Nat.
7.166)
(b) Te a quartana liberatum gaudeo itemque Piliam.
(‘I am glad you are rid of your fever and Pilia too.’ Cic. Att. 10.15.4)
Such pairs are another element in the discussion about the relationship between cases
and prepositions. It looks as if a in (b) is optional. However, it can hardly be
maintained that a makes the separative relation between libero and quartana more
specific: it is rather an idiom. Although the two expressions look synonymous, the
prepositional expression seems to be less restricted (used for human and non-human
entities), whereas the bare ablative seems to be excluded with human entities. It has
also been noted that with such verbs the prepositional expression is more common
with concrete entities, the bare case with abstract entities.106
Whereas in the previous paragraph two alternative expressions did involve the
same case in some way, this is not the case in (c) and (d), which are both instances of a
partitive expression at the noun phrase level. In (c) the partitive expression consists of
the prepositional phrase with de + ablative; in (d) the partitive expression is in the
genitive. Clearly, the specification hypothesis does not work here.
(c) . . . concedite pudori meo ut aliquam partem de istius impudentia reticere
possim.
(‘Please indulge my sense of modesty so far as to allow me the liberty of holding my
tongue about some small part of his shameless career.’ Cic. Ver. 1.32)
(d) . . . ne residere in te ullam partem iracundiae suspicemur.
(‘ . . . that we may have no misgivings that any particle of resentment lingers in your
heart.’ Cic. Deiot. 8)
106
See Théoret (1982) and LSS: § 5, n. 51 (= 52, Spanish edition).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Prepositions
The usual expression of the second argument with evado is as a source argument
(see § 4.44) with either a prepositional phrase, as in (a) and (b), where the same
element is as it were repeated, or with a bare ablative, the same case that is required by
the preposition ex, as in (c). In (d) the second argument is a patient in the accusative,
as with the verb desero ‘to depart from’ (the passive is not attested). In (e) it is in the
dative, by analogy with verbs of taking away (for more instances, mainly poetic, see
§ 4.25).
With the verb invado the second argument can be expressed as a direction argument
with the preposition in, as in (f ), or as a patient argument in the accusative
(passivization is possible), as in (g) (comparable with the verb intro ‘to enter’).
Here too the dative can be used, as in (h) (rare) (see also § 10.11 for dative direction
adjuncts).
(f ) . . . recta porta invadam extemplo in oppidum antiquom et vetus.
(‘ . . . I’ll instantly storm into the old and ancient town right through the gate.’
Pl. Bac. 711)
(g) (sc. Iugurtha) Igitur ex inproviso finis (Adherbalis) . . . cum magna manu
invadit . . .
(‘Therefore when Jugurtha suddenly invaded Adherbal’s territory with a large force .
. . ’ Sal. Jug. 20.3)
(h) Quae vastitudo haec aut unde invasit mihi?
(‘What ravage this? And whence has it assailed me?’ Acc. trag. 455)
These examples demonstrate that synchronically it is difficult to relate the form of the
argument that is governed by a compound verb to the case that is normally required
by the preposition that is at the basis of the preverb. It may be the same, but often it is
not. It may be a prepositional phrase with the same preposition, but it may be another
preposition just as well.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
107
For surveys of the developments see Väänänen (1981: 110–17); Luraghi (2010: 94–101); Fruyt
(2011); Haverling (2011).
108
See Weiss (2009: 232).
109
For Pompeii, see Väänänen (1981: 31); also Adams (2007: 260–5) for later developments.
110 111
See TLL s.v. careo 454.67ff. See Adams (1977: 36–7).
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Tandem non ego illam caream, si sit opu’, vel totum triduom?
(‘Couldn’t I in the end do without her for a whole three days if I had to?’ Ter. Eu. 223)
(b) . . . ET FA / CTUM EST ILLI VENIRE ALEXANDRIE CON TIRONES ET ME RELI / QUID CON
MATREM MEAM . . .
(‘ . . . and it happened that he came to Alexandria with recruits and left me with my
mother . . . ’ CEL 146.21–3 (Karanis, 1st quarter II AD))
(iii) A third factor is the gradual spread of the use of prepositions. Prepositions are
an essential element of the Latin language from the earliest period onwards, but the
percentage of nouns and noun phrases marked by a preposition rather than merely by
a bare case form increases markedly in the corpus, as, conversely, the use of bare cases
decreases. Thus, whereas in Classical texts the genitive overall marks 15% of all nouns
and noun phrases, it marks less than 6% in the Peregrinatio Egeriae. At the same time
prepositional phrases that were important in the replacement of the (bare) genitive
and dative (de and ad, respectively) increase enormously.112 Here, too, the increase in
use differs from preposition to preposition. Prepositions are often said to function as a
substitute for the phonologically merged case forms, but this position is too simplistic.
The phonologically least threatened genitive and dative cases were also (in various
degrees) replaced by prepositional expressions. Furthermore, in Early Latin, preposi-
tions were especially favoured with abstract nouns, without any phonological motiv-
ation, and their use widened gradually to other categories (see § 12.12).
There is, however, also a reverse trend, with a bare case assuming a function that
was in Classical Latin fulfilled by a prepositional phrase. This is visible not only in
less-educated authors, but also in literary authors such as Symmachus. For instance,
the genitive expanded into the domain of de.113 This development probably has to be
understood as some form of hypercorrect use of a prepositionless expression and as
an indirect proof of the reduction of bare case expressions.
(iv) The fourth factor is one that might be called ‘morphosystemic’. Diachronic
developments apparently are not unrestricted. The need for transparency in the
morphological system as a whole and the avoidance of semantic ambiguity seem to
function as constraints on potential changes. So, in the case of Latin, the distinction
between singular and plural, which is semantically important, was left intact by
various forms of adjustment, and otherwise likely phonological developments were
prevented from taking place. Likewise, distinctions available in one declension class
were maintained or created in other declension classes as well, or abandoned if they
were marginal to the overall system.114
112
See Iso (1994: 351) for statistics. See also the data in Brightbill and Kosch (2012).
113
See Haverling (1988: 144–202; 2006b: 352–9).
114
See Iso (1994) for a possible sketch of the chronological order of the various developments. It is
difficult not to describe the developments as organized and almost logical and unavoidable. In reality, of
course, such developments are spontaneous and unconscious.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(v) A fifth factor is the need to maintain transparency in the marking of the central
arguments in clauses/sentences with bivalent or trivalent verbs—that is, between the
subject and the second and third arguments. This need is not equally strong for all
lexical items that may be found in these functions. It has been shown that ‘the loss of
declension starts in proper nouns before it reaches common nouns, . . . inanimate
nouns before it reaches nouns indicating humans and animals, feminines before
masculines, adjectives before nouns, nouns before articles, nouns and articles before
pronouns’,115 plurals before singulars. Signals of this trend are observable in Late
Latin.
(vi) Finally, it should be observed that some sort of causal relationship is commonly
assumed between the loss of case marking and the fixation of the order of constitu-
ents, which then came to serve as a marker of syntactic relations. However, word
order is still flexible and predominantly determined by pragmatic factors in Late
Latin, as in the Classical period. See Chapter 23 for discussion.116
It is difficult to make a precise estimate of what the situation in the spoken language
of uneducated speakers was by the end of the period covered by this Syntax. Around
AD 400 it may have been the following. With nouns the nominative and the accusative
(or the merger of accusative and ablative) were still productive, and used for first and
second arguments. The genitive was still alive, although the use of the preposition de
for the partitive relation had increased. The dative case was limited to certain
expressions, but for many of its uses prepositional expressions with ad were an
alternative. The old bare uses of the ablative (to mark instrument adjuncts, for
instance) had for the most part been replaced by prepositional expressions. The
normal case with prepositions was the accusative (or the merger of accusative and
ablative). Note that there was no phonological reason for merging the accusative and
the ablative in the plural. The final outcome was, however, that for the plural the
accusative form was the ‘winner’.117 In written texts the case forms played a more
important role. Some of the developments leading to this situation are sketched
briefly below.
For the first argument not much changed. There is inscriptional evidence for the drop
of final -s of the nominative, like the forms ending in -u in the curse tablet (a), but on
the whole the nominative remained intact until after the period covered by this
Syntax. (NB: The third declension names Victore and Elegante are oblique (accusa-
tive/ablative) forms.)
115
Quoted from Schøsler (2001: 284).
116
For pragmatic factors as decisive for the order of constituents in Late Latin, see Pinkster (1991) and
Spevak (2005a).
117
This follows the line of argument of Iso (1994), where further details can be found.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
(a) Alumnu cadat, Ideu Centauru cadant, Lydeu cadat, / Virgineu Bracatu
cadant, Adamatu cadat, Lyceu / Epafu cadat, Victore cadat, Elegante
cadat, Lydeu . . .
(‘Let Alumnus fall, let Ideus and Centaurus fall, let Lydeus fall, let Virgineus and
Bracatus fall, let Adamatus fall, let Lyceus Epafus fall, let Victor fall, let Elegans fall, let
Lydeus . . . ’ Tab. devot. Audollent 283.9–11 = Kropp 11.2.1/20 (Sousse, II/III AD))118
For the second argument the merging of the accusative and ablative (and partly the
dative) in a sense removed a problem, since the use of the ablative with certain verbs
such as utor lacked a clear semantic justification. Even in Late Latin texts, in which the
distinctions in case marking made in Classical Latin are preserved, the ablative and
dative second arguments are often replaced by the accusative, as in (b) for the verb
noceo ‘to harm’.119 However, with some dative governing verbs a prepositional
alternative (notably with ad) was developed as well, suggesting that these verbs
were recognized as semantically different from the accusative governing verbs and
needed a separate marking. An example is (c).120 This use of ad is the basis for French
expressions like aider à, nuire à.121 (In Medieval Latin noceo ad is ‘not rare’.122)
(b) . . . ne iumentum noceant.
(‘ . . . that they do not harm the mule.’ Mulom. Chir. 997)
(c) Haec res et ad febricitantes prosunt.
(‘These things are also beneficial for people who have a fever.’ Mulom. Chir. 379)
Another instance of replacement of a dative argument by an ad-phrase is (d).
In the case of three-place verbs, a consequence of the merger of the accusative and the
ablative (and partly the dative) was that the second and third arguments of these verbs
were no longer formally distinct. Although there already existed a number of double
accusative verb frames (see § 12.4, and details in §§ 4.71–4.78) and such frames are
attested with some of these verbs, prepositional alternatives were preferred. With the
verb privo ‘to deprive of ’, ab- and de-phrases are found instead of the bare ablative
(see § 4.54), an example of which is (e).123 With verbs of transfer and communication
118
For discussion see Herman (1987: 103–4), with references. See also Cennamo (2001).
119
For rare passive instances of noceo, starting with Vitr. 2.7.3, see Neue-W.: III.6; also for the
influence of Greek ºø Ø on the use of the accusative in ecclesiastical texts.
120
For earlier instances of prosum + ad, especially in medical texts, see TLL s.v. prosum 2249.10ff. The
instances given there do not concern human beings, as it does in ex. (c).
121
For alternative expressions for the dative in Late Latin, see de la Villa (1999b). For the Romance
development, see Grevisse (1993: 394ff.) and Sornicola (1997).
122 123
So Stotz (1998: 277). See TLL s.v. privo 1406.49ff.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
ad-expressions compete with the dative from Early Latin onwards, and they become
interchangeable in the course of time. A remarkable case of variation is (f ). With
verbs of the opposite meaning (‘taking away’), an ab-expression instead of the dative
is once attested in Cicero, which is given in (g) below, but more instances are found
from Ulpian onwards.
(e) . . . eos a paterna pietate et a nostra communicatione privare.
(‘ . . . to deprive them of fatherly piety and of fellowship with us.’ Cypr. Ep. 56.2)
(f ) Ita hortis ad inrigandum vel ad salinas ad temperandum praebetur aquae
multitudo.
(‘Thus an abundant supply of water is furnished for irrigating gardens, or for diluting
salt in salt pits.’ Vitr. 10.4.2)124
(g) . . . cum iste a Syracusanis quae ille calamitosus dies reliquerat ademisset.
(‘ . . . when Verres had taken away from the Syracusans everything that the former
day of disaster had left them.’ Cic. Ver. 4.151)
Prepositional phrases increasingly become the norm as markers of satellites. The best
illustration are instrument adjuncts, which in Classical Latin are normally marked by
the ablative. Ex. (a) provides an illustration of the use of ex ‘out of ’ to mark an
instrument adjunct. For further examples, see § 10.55.
(a) Metae fiunt duae, una solida, una cava, ex torno ita perfectae ut . . .
(‘Two cones are made, one solid, one hollow, and so finished by the lathe that . . . ’
Vitr. 9.8.6)
For beneficiary adjuncts the preposition pro is used from Early Latin onwards
alongside the dative (see § 10.71).
The most successful prepositional alternative for the adnominal genitive was the
preposition de (‘from’), originally indicating source or origin, as in (a).
(a) Heus, memento ergo dimidium istinc mihi de praeda dare.
(‘Hey, then remember to give me half of the booty.’ Pl. Ps. 1164)
124
This is the reading of the Loeb edition. In the Teubner edition Krohn’s emendation salinis is
printed.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
Although there are instances of dimidium with a ‘partitive’ genitive (for example
dimidium mali in Pl. Ps. 452), there is no reason to assume that in (a) the prepos-
itional expression with de is equivalent to a genitive expression. It may well be taken
as an optional source adjunct with (dimidium) dare (‘half has to be taken away from
the booty and given’) and not as an attributive prepositional phrase with the noun
dimidium. In a similar way one can say dare ex praeda (Liv. 10.30.10).125 However, in
(b) de legione must be regarded as a prepositional attribute and its meaning ‘coming/
originating from’ comes very close to ‘belonging to’, the typical context for the
genitive.
(b) Tum miles ‘non sum’ inquit ‘tiro, Labiene, sed de legione X veteranus.’
(‘ “I’m not a recruit, Labienus,” replied one soldier, “but a veteran of the Tenth
legion.” ’ B. Afr. 16.2)
Even in examples from Late Latin such as (c) and (d) from the Peregrinatio Egeriae,
where a genitive might have been used, a source interpretation (‘from amidst’) is not
excluded altogether by some scholars. In (d) de may have been chosen to avoid a
sequence of two genitives.126 In the Apicius text (end of the fourth century) de is
used, as in (e), in a few instances in which a source interpretation is likely. In that
work the genitive is the normal expression for a partitive relation.127
Apart from this (optional) use of de for a partitive relation at the noun phrase level, the
genitive continued to be used for the most part as it was in the Classical period. There is
no sign of the use of de instead of the possessive genitive until (very) Late Latin.128
In Late Latin the accusative became the primary case governed by prepositions, with
the exception of the denominal prepositions of the type causa and gratia, with which
the genitive remained in use.129 The number of such denominal expressions increased
125
Full discussion in Väänänen (1981: 113–14).
126
See Väänänen (1987: 37–8).
127
See Milham (1961) and Adams (1995b: 430–4). Likewise expressions with ex are very rare.
128
See Adams (1976: 50–1) and Sas (1937: 21–30) for Merovingian Latin.
129
No deviations are recorded in the TLL.
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
considerably: for example, sub obtentu ‘on the pretext of ’ + genitive (obtentu is
already used by Tacitus).130 Due to ignorance of the classical norm and/or imitation
of Greek examples, the traditional prepositions are now and then used with the
genitive and even the dative.131
The number of compound prepositions (related to compound adverbs, which are
more common) increased enormously in Late Latin, and many of them survived in
the Romance languages (for example, abante > Fr. avant).132 An Early Latin example
is insuper in (a).
(a) Insuper arbores stipitesque trabem planam imponito . . .
(‘Above the anchor-posts and the guide-posts lay a horizontal beam . . . ’ Cato Agr. 18.5)
130
For more instances, see Blaise (1955: 94–5).
131
For the use of in with the genitive and the dative, see TLL s.v. 799.12ff.
132
For a very complete survey, see Hamp (1888) and Campos (1972).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
CHAPTER 13
Agreement
Of these two types of agreement the grammatical one is dominant, certainly in Cicero
and Caesar, and instances of notional agreement are often treated as exceptions.
1
For the terms, see Corbett (2006: 4).
Agreement
Editors tended (and tend) to choose grammatical agreement when there is variation
in the manuscripts or even against readings in the manuscripts.2
nominal agreement:
(ii) anaphoric(ally used) pronouns and relative pronouns agree with the con-
stituent they refer to in the preceding clause or sentence (NB: this does not
hold for case);
(iii) subject and object complements agree with the subject and object they are
associated with;
(iv) secondary predicates agree with the constituent they are associated with;
(v) pronominal subjects may agree with the subject or object complement of
their clause;
(vi) adjectives, determiners, quantifiers, and similar words agree with the head
noun in their noun phrase;
(vii) appositions agree with the constituent they are associated with.
The form of the verb in principle matches the properties of the subject constituent of
its clause in person/number, and (as far as nominal forms are concerned) in number/
gender/case.3 This general rule is relatively simple in the case of subject constituents
that consist of a single noun, pronoun, or other subject constituent (simple subjects),
but is more complicated in the case of subject constituents that consist themselves of
two or more coordinated constituents (compound subjects).
2
For a large collection of readings chosen against the readings of the manuscript, see Baehrens (1912:
452–68).
3
For a general introduction to person/number agreement, see Siewierska (2004: 120–72).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
This rule of agreement also holds when the subject has to be inferred from the
preceding context or from the situation (see § 13.26).
In the examples above, gender is either an inherent feature of the nouns involved,
or—in the case of the proper name Amphitruo in (c)—determined by convention:
male person > masculine gender. Similarly, in (f ) factus is masculine, because in the
specific situation the person tu refers to is a male person. In (g), the gender of the
demonstrative pronoun hic is determined by the referent (a male person) with which
it agrees in gender and number: accepturus in turn agrees with hic.
(f ) Tum igitur tu dives es factus?
(‘So you’ve become rich?’ Pl. As. 330)
(g) Certe advenientem hic me hospitio pugneo accepturus est.
(‘He’s definitely going to receive me with a fisty welcome on my arrival.’ Pl. Am. 297)
In accusative and infinitive clauses, agreement of the verb with its subject is impos-
sible with present infinitives and perfect active infinitives, as is shown in (h) present
active infinitive, (i) present passive infinitive, and (j) perfect active infinitive.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
However, just as in (c)–(g) above, the nominal parts of compound perfect, future, and
gerundival forms do show agreement with the subject in number/gender/case. See
(k)–(n) for instances of the perfect deponent, perfect passive, periphrastic future, and
gerundive, respectively.
(h) . . . donec se adiurat anus / iam mihi monstrare.
(‘ . . . until the old woman swore she’d show her to me right away.’ Pl. Cist. 582–3)
(i) Istac lege filiam tuam sponde’n mihi uxorem dari?
(‘Do you promise your daughter will be given in marriage to me under that
condition?’ Pl. Trin. 1162)
(j) Eam suam esse filiam / seque eam peperisse sancte adiurabat mihi.
(‘She gave me a solemn oath that she was her daughter, and that she had given birth
to her.’ Pl. Cist. 568–9)
(k) Commenta mater est esse ex alio viro / nescioquo puerum natum . . .
(‘The mother has made up a story that the girl has a child by some other man . . . ’ Ter.
Ad. 657–8)
(l) Postquam Syracusas de ea re rediit nuntius / ad avom puerorum, puerum
surruptum alterum / patremque pueri ess’ Tarenti emortuom, / immutat
nomen avos huic gemino alteri.
(‘After the news about this came back to Syracuse to the grandfather of the boys, that
the one boy had been kidnapped and that the boy’s father had died in Tarentum, the
grandfather changed the name of this other twin.’ Pl. Men. 37–40)
(m) Mysis, / per omnis tibi adiuro deos, numquam eam me deserturum . . .
(‘Mysis, I swear to you by all the gods that I will never abandon her . . . ’ Ter. An. 693–4)
(n) Qui autem civium rationem dicunt habendam externorum negant ii diri-
munt communem humani generis societatem.
(‘Others again who say that regard should be had for the rights of fellow-citizens, but
not of foreigners, would destroy the universal brotherhood of mankind.’ Cic. Off. 3.28)
There are a few instances of a plural imperative form in combination with singular
aliquis, as in (o).
of only one constituent, the verb of a clause with a compound subject agrees with that
subject. However, in this case it is not immediately evident what the properties of the
compound subject are in terms of person/number and, if there is a nominal part of the
verb, in terms of number/gender/case.
In this section most attention will be given to the factors that determine the number
of the verb in a clause with a compound subject. In the case of a compound subject
that consists of two or more coordinated personal pronouns, agreement in person
is relevant as well (see § 13.6, par. (ii)). Agreement in gender is discussed separately
in § 13.9. Two factors will turn out to be decisive for the form of the verb
with a compound subject: (a) animacy or inanimacy of the members of the com-
pound subject; (b) the relative order of the compound subject or its members
and the verb.
In some cases the use of the plural or the singular seems to have a semantic
explanation. Thus, in (a), Milphio and vilicus, both singular (and masculine) together
determine the plural form of egrediuntur. In this case the two members are indeed
together involved in the same state of affairs, as appears from the presence of una
‘together’. The plural is therefore understandable. In (b), dissentiunt is a verb that
normally requires (see § 4.38) two persons to disagree, so the plural is again
understandable. In (c), there is no indication that the father and mother died at
the same time and together, but the verb is plural nonetheless, as if its subject were
parentes ‘parents’. Note that mortui is plural masculine, although pater is masculine
and mater feminine (details in § 13.9). Such an explanation is not available in cases like
(d), where the persons involved definitely did not cooperate, and yet the verb is plural.
Agreement
Turning now to singular verb forms in agreement with a compound subject, diligitur
in (e) is applicable to both mater and soror, both singular (and feminine), but the
feelings of the two agents are most likely for each of the women individually.4 The
verb is singular, in agreement with the nearest member of the compound subject
and is understood with the other member (a form of conjunction reduction—see
Chapter 19). The same reasoning may hold for the singular in (f ). Sometimes the
combination of the two members may be understood as referring to one complex entity,
as senatus populusque Romanus in (g), more or less understood as ‘the government’.
In fact, a semantic explanation of the type given in the preceding note does not
account for all the data. In most cases animacy and relative order turn out to be the
decisive factors. In Classical Latin, the majority of verbs are plural when (i) the subject
is compound and consists of two or more singular members that refer to animate
beings, and (ii) the entire subject precedes the verb. In all other situations agreement
with the nearest member is much more common, especially with inanimate subjects.
In Cicero, out of more than 1,000 compound subjects consisting of two or more
abstract nouns (a subcategory of non-animate nouns), a few more than ten have their
corresponding verb in the plural.5 The most likely explanation is that these abstract
nouns are not conceived of as acting in cooperation. Occasionally a verb that follows a
compound subject agrees with the first member. Latin, therefore, presents a broad
spectrum of possibilities.6 There is, moreover, considerable variation between authors
and within individual authors, so it is better to speak of more or less strong tendencies
than of rules.
The above observations are based purely on numerical data. A semantic analysis has
never been made, as far as this author is aware. On the basis of these figures the plural
seems to be the normal option. However, in the system it may well have been the
exception: the preserved corpus is very anthropocentric and subjects are much more
often animate than inanimate.
4
For a discussion of the reality behind this sentence, see Shackleton Bailey ad loc.
5
See Lebreton (1901a: 2ff.).
6
Latin has therefore all three types of agreement with compound subjects distinguished by Corbett
(2006: 62–3, 168–70).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
13.6 The subject consists of two or more animate (or equivalent) entities
(i) This section begins with cases in which the entire compound subject precedes the
verb. When the compound subject is formed by two or more singular members referring
to animate beings (or entities that are treated as equivalent to animate beings), it regularly
entails plural agreement if the members are coordinated by one of the conjunctive
coordinators et, atque, -que, or by zero coordination (so-called asyndetic coordination).
Examples are (a)–(d). Much less common is agreement of the verb with the nearest
member of a compound subject, as in (e). In comedy from Terence onwards and in
poetry, agreement with the nearest member is more common than with the whole subject.
Agreement
(ii) The situation is more complicated if first and second person personal pronouns
are involved. If a compound subject consists of a first person pronoun (singular or
plural) plus a second person pronoun (singular or plural) or a third person entity
(singular or plural), the verb is normally first person plural, as shown in (h) and (i). If
the compound subject consists of a second person pronoun (singular or plural) plus a
third person entity (singular or plural), the verb is normally second person plural.
This is shown in (j) and (k). Less common is (l)—third person plural verb in spite of
ego—and (m), where the verb agrees with third person plural homines, in spite of tu.
In such cases the verb agrees with the nearest member.
(h) Iam diu ego huic bene et hic mihi volumus et amicitia est antiqua.
(‘I have been well disposed to him, and he to me for a long time, and our friendship is
old.’ Pl. Ps. 233)
(i) Pater ego fratres mei . . . pro vobis arma tulimus.
(‘My father, myself, my brothers . . . have borne arms in your cause.’ Liv. 37.53.24)
(j) Nam quod idem Segulius (sc. dicit) veteranos queri quod tu et Caesar in
decem viris non essetis, utinam ne ego quidem essem! Quid enim molestius?
(‘As for this same Segulius’ story that the veterans are grumbling because you and
Caesar are not on the Commission of Ten, I only wish I was not on it! A most
tiresome business!’ Cic. Fam. 11.21.2)
(k) Si tu et Tullia, lux nostra, valetis, ego et suavissimus Cicero valemus.
(‘If you and our beloved Tullia are well, darling Marcus and I are well too.’ Cic.
Fam. 14.5.1)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(l) Itaque nunc . . . aspiravit nemo eorum quorum ego concursu itemque ii
consules qui post me fuerunt rem publicam defendere solebant.
(‘So now . . . not one of those men lifts a finger—those men by whose rallying both
I and the consuls who succeeded me used to defend the state.’ Cic. Att. 2.1.8)
(m) Quanti autem facere debeam et tu et omnes homines sciunt.
(‘You and all the world know how highly I ought to regard him.’ Cic. Fam. 13.8.1)
(iii) If the verb precedes a compound subject, it usually agrees with the first (and
nearest) member of that subject, as in (n)–(q), where the verbs are singular. Note that
in (n) the subject complement consul is singular as well. In (o), the apposition is in the
plural. In ex. (p), the last member of the subject is plural. In (q), an accusative and
infinitive clause, the nominal part of the infinitive (convictam) also agrees with the
first member of the subject in gender. However, a preceding verb may also be in the
plural, as in (r) and (s). Compare also (t).
(iv) If the verb is positioned between the members of a compound subject, as in (u)
and (v), it agrees with the member that precedes it.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(u) Namque ego fui illi in re praesenti et meus quom pugnatum est pater.
(‘My father and I were there at the scene of action when the fighting took place.’
Pl. Am. 249)
(v) . . . siquidem Homerus fuit et Hesiodus ante Romam conditam.
(‘ . . . seeing that Homer and Hesiod lived before the foundation of Rome.’ Cic. Tusc. 1.3)
(v) Up to this point the discussion has dealt with ‘normal’ coordination. In the case
of emphatic coordination by means of et . . . et or neque . . . neque (which is infrequent)
the verb usually agrees with the nearest member. This is understandable since this
form of coordination underlines the individuality of the members. The position of the
verb with respect to the compound subject is irrelevant. Examples are (w)–(z). Note
also (aa) with neve.
There are, however, also instances of emphatic coordination in which the verb
agrees with the compound subject as a whole in the way discussed above. Examples
are (ab)–(ae).
(ac) Incredibile est quanti faciamus et ego et frater meus, qui mihi carissimus est,
M. Laenium.
(‘You would hardly believe how much I and my dearest brother think of M. Laenius.’
Cic. Fam. 13.63.1)
(ad) . . . quodque ei et filius et fratris filius a Caesare remissi essent . . .
(‘ . . . because both a son and a nephew had been sent back to him by Caesar . . . ’ Caes.
Gal. 5.27.2)
(ae) . . . maximeque indignari quod in decem viris neque Caesar neque ego habiti
essemus atque omnia ad vestrum arbitrium essent collata.
(‘ . . . they are particularly indignant that neither Caesar nor I have been put on the
Commission of Ten, and that everything has been placed in the hands of you
gentlemen.’ D. Brut. Fam. 11.20.1)
There are a few instances of plural verbs with compound subjects coordinated by a
disjunctive coordinator. Examples are (af) and (ag). Note also (ah) with an. Plural
verb forms are occasionally also found in other parallel structures, as in (ai).
(af) . . . nec quemquam hoc errore duci oportet ut siquid Socrates aut Aristippus
contra morem consuetudinemque civilem fecerint locutive sint idem sibi
arbitretur licere.
(‘ . . . and no one ought to make the mistake of supposing that, because
Socrates or Aristippus did or said something contrary to the manners and
established customs of their city, he has a right to do the same.’ Cic. Off. 1.148)
(ag) Iam aut Callipho aut Diodorus quo modo poterunt tibi istud concedere . . .
(‘Or how will Callipho or Diodorus be able to grant your premise . . . ’ Cic.
Fin. 4.50)
(ah) Tuque dubitas, Cimberne Annius an Veranius Flaccus imitandi sint tibi . . .
(‘Can you doubt whether Annius Cimber or Veranius Flaccus ought be
imitated by you . . . ?’ Suet. Aug. 86.3)
(ai) Quid enim ista repentina adfinitatis coniunctio, quid ager Campanus, quid
effusio pecuniae significant?
(‘What else signifies this sudden marriage connection, or the Campanian
Domain, or the pouring out of money?’ Cic. Att. 2.17.1)
Conjunction reduction may apply in parallel clauses when two or more singular
subjects are engaged in similar states of affairs. The verb may be plural in such cases,
as in (aj) and (ak).
(aj) Non cum Latinis decertans pater Decius, cum Etruscis filius, cum Pyrrho
nepos se hostium telis obiecissent.
(‘The elder Decius in desperate conflict with the Latins, his son in conflict
with the Etruscans, his grandson fighting Pyrrhus would not have flung
themselves upon the weapons of the enemy.’ Cic. Tusc. 1.89)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
It is far less common for the verb to be in the plural, agreeing with these compound
subjects as a whole, the more so for the (rare) combination animate/abstract. Two
examples are (h) and (i), the latter possibly the only instance in Cicero. Exx. (j)–(l)
illustrate combinations of inanimate entities. Note that in (k) and (l) the neuter plural
form of the nominal parts is in agreement with two coordinated nouns of different
gender (details below § 13.9).
(h) . . . qui scire posses aut ingenium noscere, / dum aetas metus magister
prohibebant?
(‘ . . . for how could you know him or judge his character before, while age, appre-
hension, and a tutor were restraining him?’ Ter. An. 53–4)
(i) Messalla languet, non quo aut animus desit aut amici, sed coitio consulum
<et> Pompeius obsunt.
(‘Messalla is lagging, not for lack of spirits or friends, but the Consuls’ pact and
Pompey are against him.’ Cic. Att. 4.15.7)
(j) <Et> quem ad modum temeritas et libido et ignavia semper animum
excruciant et semper sollicitant turbulentaeque sunt . . .
(‘And just as Rashness, Licence and Cowardice ever torment the mind, ever awaken
trouble and discord . . . ’ Cic. Fin. 1.50)
(k) Introducti autem Galli ius iurandum sibi et litteras a Lentulo, Cethego,
Statilio ad suam gentem data esse (v.l. datas) dixerunt . . .
(‘The Gauls were then brought in and said that an oath and a letter to their people had
been given to them by Publius Lentulus, Cethegus, and Statilius . . . ’ Cic. Catil. 3.9)
(l) . . . quod Fregenis murus et porta de caelo tacta erant . . .
(‘ . . . the wall and gate at Fregellae had been struck by lightning . . . ’ Liv. 32.29.1)
7
No further examples in K.-St.: I.53 and Lebreton (1901a: 23).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(ii) A remarkable instance of agreement of the verb with the first pronoun of a
compound subject is (b): ego (the fetialis) represents the populus Romanus, and
assimilates it to the first person. Ex. (c) is also quite remarkable.
(b) . . . ob eam rem ego populusque Romanus populis Priscorum Latinorum
hominibusque Priscis Latinis bellum indico facioque.
(‘ . . . I therefore and the Roman People declare and make war on the tribes of the
Ancient Latins and the men of the Ancient Latins.’ Liv. 1.32.13)
(c) O noctes cenaeque deum! Quibus ipse (sc. ego) meique / ante Larem
proprium vescor . . .
(‘O nights and feasts divine! When before my own Lar we dine, my friends and I . . . ’
Hor. S. 2.6.65–6)
Grammatical agreement in gender of the nominal part of the verb with a compound
subject is no different from agreement of subject and object complements and of
secondary predicates, which is dealt with separately in §§ 13.15–13.18. If agreement is
not with the subject as a whole, but with the nearest member of the compound
subject, gender agreement is straightforward. When the agreement is with the subject
as a whole and the members have the same gender there is also no problem. One
expects: masculine + masculine > plural masculine—see (a); feminine + feminine >
plural feminine—see (b); neuter + neuter > plural neuter—see (c).
(a) . . . in nostra acie Castor et Pollux ex equis pugnare visi sunt . . .
(‘ . . . Castor and Pollux were seen fighting on horseback in our ranks.’ Cic. N.D. 2.6)
(b) . . . sapientiam, temperantiam, fortitudinem copulatas esse docui cum
voluptate . . .
(‘Wisdom, Temperance, and Courage I have shown to be so closely linked with
pleasure . . . ’ Cic. Fin. 1.50)
(c) At initio primi belli Punici Firmum et Castrum colonis occupata . . .
(‘At the outbreak of the First Punic War Firmum and Castrum were occupied by
colonies . . . ’ Vell. 1.14.8)
However, when the members of the compound subject differ in gender and the
solution of agreement with the nearest member is not chosen, a resolution of the
conflict is required.8 The following resolutions are found.
8
For typological aspects, see Corbett (2006: § 8.3). For the phenomenon of resolution in Latin, see
Johnson (2013).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(i) When the members of a compound subject refer to individual animate entities
and differ in gender, masculine gender takes precedence over feminine gender. This is
illustrated by (d) and (e)—the situation doesn’t arise very often.
(d) Mimos dico et mimas, patres conscripti, in agro Campano conlocatos.
(‘I tell you, Members of the Senate, that mime actors and mime actresses have been
settled on Campanian land.’ Cic. Phil. 2.101)
(e) Hic cum augurato liberaretur Capitolium, Iuventas Terminusque maxime
gaudio patrum vestrorum moveri se non passi.
(‘Here, when the Capitol was being cleared with augural rights, Juventas and
Terminus, to the vast joy of your fathers, refused to be moved.’ Liv. 5.54.7)
(ii) When the members of a series of parallel nouns functioning as subject refer to
animate beings and differ in gender, but one of them is a collective noun, the verb may
agree with the nearest member. This is shown in (f ) and (g).
(f ) Pro me non . . . adulescentes filii, non propinquorum multitudo populum
Romanum est deprecata.
(‘No young sons of my own, no crowd of kinsfolk were there to entreat the Roman
people on my behalf.’ Cic. Red. Sen. 37)
(g) Nullum externum periculum est, non rex, non gens ulla, non natio perti-
mescenda est.
(‘There is no danger from without, no king, no people, no nation, is to be feared.’ Cic.
Agr. 1.26)
(iii) When the members of a compound subject refer to inanimate entities and the
members differ in gender, the nominal part of the verb is neuter plural. Examples are
given in § 13.7 ((k) and (l)). An additional example is (h). This also holds for other
nominal constituents that are related to the compound subject, like the secondary
predicate integra in (i) and the appositive adjective dissimillima in (j).
(h) Deinde, inter conscios ubi locus veneficii tempusque composita sint, eo
audaciae provectum ut . . .
(‘Later, when a place and time for the mortal dose had been agreed upon among the
conspirators, he carried audacity to the point of . . . ’ Tac. Ann. 4.10.2)
(i) . . . sic anima atque animus quamvis [est] integra recens <in> / corpus eunt . . .
(‘ . . . so spirit and mind, even though they enter whole into a new body . . . ’ Lucr. 3.705-6)9
(j) Labor voluptasque, dissimillima natura, societate quadam inter se naturali
sunt iuncta.
(‘Toil and pleasure, most unlike in nature, have been linked together in a sort of
natural bond.’ Liv. 5.4.4)
9
For Lucretius’ practice, see Bailey in his Commentary (1949: I.94).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
Remarkably, from Sallust onwards the neuter plural form is also chosen, especially in
historical prose, when the members are both or all feminine abstract nouns, possibly
inspired by Greek usage.10 An example is (k). Ex. (l) is different: fugienda is
substantival: ‘things to be avoided’.
(k) Deinde ubi ira et aegritudo permixta sunt, cum maxuma cura ultum ire
iniurias festinat.
(‘Then, when anger and sorrow were mingled, he devoted all his thoughts to
prompt vengeance for the outrage.’ Sal. Jug. 68.1)
(l) Stultitiam autem et timiditatem et iniustitiam et intemperantiam cum
dicimus esse fugienda (fugiendas cj. Schiche) propter eas res quae ex ipsis
eveniant . . .
(‘On the other hand when we say that folly, cowardice, injustice, and
intemperance are to be avoided because of the consequences they entail . . . ’
Cic. Fin. 3.39)
Supplement:
Nox atque praeda castrorum hostis quo minus victoria uterentur remorata sunt.
(Sal. Jug. 38.8); . . . postquam pax et concordia speciosis et inritis nominibus iactata
sunt . . . (Tac. Hist. 2.20.2);
(iv) Grammatical agreement in gender of the nominal part of the verb with the first
member of a preceding compound subject is very rare. Clear examples are (m) and
(n). In (m), masc. pl. loricandi agrees with masc. pl. parietes; in (n), masc. sg. amor
with masc. sg. adlaturus. Less clear is (o): membra and diffusa are both n. pl., but n. pl.
is the normal solution for a compound subject with members with different gender
(see the preceding section).
10
See Sz.: 434–5; K.-G.: I.77–8.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
ipsum, quod paranti expeditionem Othoni campus Martius et via Flaminia, iter belli,
esset obstructum, . . . in prodigium . . . vertebatur. (Tac. Hist. 1.86.3);
NB: poetic exploitation: . . . obvia nescio quot pueri mihi turba minuta / venerat (hos
vetuit me numerare timor). (Prop. 2.29.3–4);
Grammatical agreement with the head: . . . salutemque tibi plurimam adscribit et
Tulliola, deliciae nostrae. (Cic. Att. 1.5.8); Epicrates Rhodius . . . censuit, et Patara,
caput gentis, in societatem adiungenda. (Liv. 37.15.6); Mylae, proximum oppidum,
ita munitum ut inexsuperabilis munimenti spes incolas ferociores faceret, non portas
claudere regi satis habuerunt . . . (Liv. 42.54.1); Ipsa Bactra, regionis eius caput, sita
sunt sub monte Parapaniso. (Curt. 7.4.31); Eodem anno Frisii, Transrhenanus
populus, pacem exuere, nostra magis avaritia quam obsequii impatientes (NB: sec-
ondary predicate). (Tac. Ann. 4.72.1);
Related to agreement with an apposition is agreement with a constituent of a non-
restrictive relative clause, as in (d)–(f ), a rare phenomenon.
Something similar is possible with alter . . . alter and alius . . . alius functioning as
distributive appositives (for which see § 11.85). For the regular agreement with the
subject, see (b).
(a) Ut enim pictores et ii qui signa fabricantur et vero etiam poetae suum
quisque opus a vulgo considerari vult . . .
(‘For, as painters and makers of statues, and indeed even poets, each wants his own
work to be inspected by the public . . . ’ Cic. Off. 1.147)
(b) . . . ambo exercitus, Veiens Tarquiniensisque, suas quisque abirent domos.
(‘ . . . both armies, the Veientine and the Tarquiniensian, marched off every man to
his own home.’ Liv. 2.7.2)
Supplement:
Si quidem istis, cum summi essent oratores, duae res maxumae altera alteri defuit.
(Cic. Brut. 205); Hinc ceteri particulas arripere conati suam quisque videri voluit
afferre sententiam. (Cic. Fin. 5.72); . . . quando duo ordinarii consules eius anni, alter
morbo, alter ferro perisset . . . (Liv. 41.18.16);
However, agreement with the subject complement occurs as well. In the standard
grammars this phenomenon is often called ATTRACTION (the term is used for various
other phenomena as well). This option is chosen particularly often (though not
required) if the subject complement is positioned nearer to the verb and either
immediately precedes it, or—less frequently—follows. In (c), the subject complement
onus precedes the copular verb visum est with which it agrees, instead of the subject
paupertas. Ex. (d) shows a different order; the verb dico is a three-place verb requiring
an object complement in the active. Similar in this respect is (e). In (f ), the subject is
plural; the verb is singular in agreement with the subject complement.
(c) . . . paupertas mihi onus visum’st et miserum et grave.
(‘ . . . poverty appeared to me a burden so grievous and so insupportable.’ Ter. Ph. 94)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(d) In terris dictum (sc. est) templum locus augurii aut auspicii causa quibusdam
conceptis verbis finitus.
(‘On the earth, a place set aside and limited by certain formulaic words for the
purpose of augury or the taking of the auspices is called templum.’ Var. L. 7.8)
(e) Minime, nisi honos ignominia putanda est.
(‘By no means, unless an honour is to be thought a disgrace.’ Cic. Balb. 7)
(f ) Cuius (sc. consulatus) fuit initium ludi compitalicii tum primum facti post
L. Iulium et C. Marcium consules contra auctoritatem huius ordinis.
(‘Of which the Compitalician Games were the beginning, then for the first time
celebrated since the consulship of Lucius Julius and Gaius Marcius, against the
authority of this order.’ Cic. Pis. 8)
Supplement:
Pr<a>esidium dictum (sc. est) qui extra castra praesidebant in loco aliquo . . . (Var.
L. 5.90); . . . unus homo plures esse homines iudicarentur (iudicaretur edd.).
(Cic. Caec. 62); . . . ut <sit> iam universus hic mundus una civitas communis deorum
atque hominum existimanda. (Cic. Leg. 1.23); Contentum vero suis rebus esse
maximae sunt certissimaeque divitiae. (Cic. Par. 6.51); Summa omnium fuerunt
ad milia CCCLXVIII. (Caes. Gal. 1.29.3); . . . possedere ea loca quae proxuma
Carthagine[m] Numidia appellatur. (Sal. Jug. 18.11); . . . unumque erat omnia vul-
nus. (Ov. Met. 15.529); Gens universa Veneti appellati (sc. sunt). (Liv. 1.1.3); Taurini
Semigalli proxuma gens erat in Italiam degresso. (Liv. 21.38.5); Idem iocosis nomine
Gryllum deridiculi habitus pinxit, unde id genus picturae grylli vocantur. (Plin.
Nat. 35.114); Magnae divitiae sunt lege naturae composita paupertas. (Sen.
Ep. 4.10); Litterae thesaurum est . . . (Petr. 46.8 (Echion speaking)); . . . iudicii mei
documentum sit non meae tantum necessitudines, quas tibi postposui, sed et tuae.
(Tac. Hist. 1.15.2);
A very rare exception is (c), in which the second person plural head vos is followed by
a third person plural verb in the relative clause.11
(a) Ille qui illas perdit salvus est. Ego qui non perdo pereo.
(‘The one who loses them is saved; I, who don’t lose them, am lost.’ Pl. As. 637)
(b) . . . vostram iterum imploro fidem / qui prope hic adestis quique auditis
clamorem meum . . .
(‘I implore your protection again, you who are close by and can hear my shouting!’
Pl. Rud. 622–3)
(c) Vosque, Lares, tectum nostrum qui funditus curant.
(‘And you, House-Gods, who make our home, from floor to roof, their care.’ Enn.
Ann. 620V= 619S)
11
References to late instances in Sz.: 430.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
As in the case of agreement of a verb with a compound subject, especially with respect
to its nominal part, (§ 13.5, especially § 13.9), the situation is more complicated if the
agreement is with a compound noun phrase, certainly if the members of that phrase
are different in number and/or gender. Complications also arise if the agreement is
with a constituent that consists of a head and non-restrictive apposition. (The
relevant instances are given in § 11.81.) Factors that may determine the form of
agreement chosen are relative position with respect to the members of the compound
noun phrase and animacy of the members.
Agreement
ablative absolute clause. If one of the members is masculine that will normally have
precedence over the feminine gender of the other member(s). An example is (b)—a
so-called ab urbe condita clause with the participle in agreement with the masculine
member. However, the subject complement, etc. may also agree with the last member
only, as in (c): postulante is sg. in agreement with Domitio. See also (d), with the
secondary predicate in agreement with the second (feminine) member.
If the subject complement precedes the compound noun phrase, agreement with the
nearest member is more common. In (e), the subject complement consul only agrees with
Tullius. In (f ) the secondary predicate insperanti agrees only with mihi. In (g) amici is
masc. plural in agreement with pavones. However, in (h), plural formosi agrees in gender
with masculine singular verris (precedence of masculine, as in § 13.6 and § 13.9).
Finally, ex. (i) shows an ablative absolute clause, with the complement consule in
agreement with the first member Sulla which it follows.
(i) Atque huic anno proxumus Sulla consule et Pompeio fuit.
(‘The year following this was the consulship of Sulla and Pompeius.’ Cic. Brut. 306)
Exceptionally rare are instances where a preceding complement is plural. An
example is (j), where patria must be regarded as personified.12
(k) Non mihi uxor aut filius patre et re publica cariores sunt . . .
(‘Neither my wife nor my son is dearer to me than my father and my
country . . . ’ Tac. Ann. 1.42.1)
12
See Lebreton (1901a: 10).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(a) Recte igitur, si omnibus in rebus temeritas ignoratioque vitiosa est, ars ea
quae tollit haec virtus nominata est.
(‘If therefore rashness and ignorance are in all matters fraught with mischief, the art
which removes them is correctly entitled a virtue.’ Cic. Fin. 3.72)
(b) Nam clementia, iustitia, benignitas, fides, fortitudo in periculis communibus
iucunda est auditu in laudationibus.
(‘For mercy, justice, kindness, fidelity, courage in common dangers are acceptable
topics in a panegyric.’ Cic. de Orat. 2.343)
(c) Etsi quid te horum fugit, legibus, iudiciis, senatu sublato libidines, audacias,
sumptus, egestates tot egentissimorum hominum nec privatas posse res nec
rem publicam sustinere?
(‘But you must be well aware of all this—that after the subversion of laws, law courts,
and Senate neither private wealth nor common can satisfy the lusts, impudence,
extravagance, and necessities of such a multitude of starvelings.’ Cic. Att. 9.7.5)
Supplement:
. . . quam in omnibus rebus difficilis optimi perfectio atque absolutio ex eo quod
dicam existimari potest. (Cic. Brut. 137); Corporis igitur nostri partes totaque figura
et forma et statura quam apta ad naturam sit, apparet . . . (Cic. Fin. 5.35);
(ii) When there is agreement with the compound noun phrase as a whole and the
members have different gender, a possible conflict is resolved by using neuter plural
forms. This is illustrated by (d) and (e). Ex. (d) has one neuter and two feminine
members; the subject complement is neuter plural. In (e), the members of the
compound noun phrase are feminine and masculine. There is no conflict between
the feminine and neuter member in (f ) because the ablative plural of all genders is the
same. For a preceding subject complement, see (g).
(d) Igitur iis genus aetas eloquentia prope aequalia fuere . . .
(‘So their good birth, age, and eloquence were about equal . . . ’ Sal. Cat. 54.1)
(e) . . . exemplo fuit collegis se eumque intuentibus, quam gratia atque honos
opportuniora interdum non cupientibus essent.
(‘ . . . thus showing his colleagues, as they considered his case and their own, that
favour and high office sometimes come more easily when men do not covet them.’
Liv. 4.57.6)
(f ) Igitur quod nunc intra murum fere patres familiae correpserunt relictis falce
et aratro . . .
(‘As therefore in these days practically all the heads of families have sneaked within
the walls, abandoning the sickle and the plough . . . ’ Var. R. 2.pr.3)
(g) Natura inimica inter se esse liberam civitatem et regem.
(‘That by nature a free state and a king were things hostile to each other.’ Liv. 44.24.1)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Supplement:
. . . in itinere de proelio facto Brutoque et Mutina obsidione liberatis audivi. (Planc.
Fam. 10.11.2); . . . BASIM·GRADUM·ARAM·SUA·PECUNIA / FACIUNDA· COER(AVERUNT)·EIDEMQ
(UE)·PROBAVER(UNT) (CIL X.5159.5–6 (Casinum, 40 BC)); . . . propositurum, ut iudi-
cium censorum ac pudor sponte cedentium permixta (cj. Ritter, permixti ML)
ignominiam mollirent. (Tac. Ann. 11.25.3); Fraudem et dolum obscura eoque
inevitabilia. (Tac. Hist. 4.24.2);
NB: remarkable plural with disjunctive coordination: . . . cerebri membranas mitigat
farina seminis nitro aut sale aut cinere additis. (Plin. Nat. 20.250);
In § 13.9 examples are given of the use of the neuter plural form in agreement with a
compound noun phrase consisting of two or more feminine members. Related
examples of subject complements are (h) and (i). Note that the coordinated nouns
are abstract, which may in these two cases explain the use of the neuter forms as
substantival (less convincing for nominal forms of verbs). Remarkable too is the use
of the neuter plural form cultiora with three masculine (abstract) nouns in (j).
In the case of compound phrases that consist of two or more inanimate or mixed
animate/inanimate entities a nominal subject complement sometimes agrees with the
first member. In (k), inbecillos—subject complement—is masculine, as is motus.
The same occurs with object complements and with secondary predicates. In (l),
masculine lectus—a secondary predicate—agrees with lucus.
(l) Lucus quidem ille et haec Arpinatium quercus agnoscitur saepe a me lectus
in Mario.
(‘Surely that grove yonder and this oak tree of Arpinum are recognized by me as
those of which I have read so often in the “Marius”.’ Cic. Leg. 1.1)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
13
For possible Greek influence, see Calboli (2009: 156–7). For examples of the use of the dative in
Greek, see K.-G.: II.24. For ex. (e), see K.-St.: I.680.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
qui respectum habent . . . vobis necesse est fortibus viris esse . . . (Liv. 21.44.8); Quo
tempore, mihi fratrique meo . . . destinari praetoribus contigit . . . (Vell. 2.124.4);
L. quoque Sulla, cum imperator esset, chlamydato sibi et crepidato Neapoli ambulare
deforme non duxit. (V. Max. 3.6.3); Nescio ut Ismenias tibiis canere, sed non pudet
me tibicinem non esse; nescio ut Apelles coloribus pingere, sed non pudet me non
esse significem. (Apul. Soc. 21);
(Secondary predicates): Sed quia non est omnibus stantibus necesse dicere . . . (Cic.
Marc. 33); . . . quod nihilo magis ei liciturum esset plebeio rem publicam perdere
quam similibus eius me consule patriciis esset licitum. (Cic. Att. 2.1.5); . . . ut sibi
triumphanti liceret urbem invehi. (Liv. 36.39.5);
(Lack of agreement): Nam expedit bonas esse vobis. Nos, quibu’cum est res, non sinunt.
(Ter. Hau. 388); Cur his per te frui libertate sua, cur denique esse liberos non licet? (Cic.
Flac. 71); Is enim erat annus quo per leges ei consulem fieri liceret. (Caes. Civ. 3.1.2); . . . ne
sibi interesse certaminibus eorum armaque aut haec aut illa abnuentem alteram socie-
tatem sequi necesse sit. (Liv. 29.23.9—NB: secondary predicate); . . . cui post exilium
consulem creari proscriptoque facere proscriptionem contigit. (V. Max. 6.9.14);
With three-place verbs governing an accusative object and a prolative infinitive, such
as doceo ‘to teach’, the subject complement that goes with the infinitive agrees with
the accusative object, as in (f ). With three-place verbs governing a dative recipient/
addressee and a prolative infinitive the subject complement is in the dative, as in (g).
(f) Agite dis inmortalibus gratias quod eum docetis esse crudelem qui non potest
discere.
(‘Thank the immortal gods that you are teaching a person to be cruel who cannot
learn this.’ Sen. Ep. 7.5)
(g) Mediocribus esse poetis / non homines, non di, non concessere columnae.
(‘Neither men nor gods nor booksellers ever allowed poets to be middling.’ Hor. Ars
372–3)
Supplement:
. . . da mihi fallere, da iusto sanctoque videri . . . (Hor. Ep. 1.16.61); Primum ego me
illorum dederim quibus esse poetis / excerpam numero. (Hor. S. 1.4.39–40); Vobis
inmunibus huius / esse mali dabitur. (Ov. Met. 8.690–1);
Appendix: With the expressions of naming, such as nomen est alicui ‘someone is called’
and nomen do alicui ‘I give someone a name’ the name agrees either with the dative or
with nominative or accusative. Examples of both are (h)–(k). Authors vary in their
preferences.14 For a few examples of lack of agreement with the object, see § 12.17.
14
For examples and the individual variation, see K.-St.: I.420–1; Sz.: 90–1.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
Genitives are attested after the Classical period, as in (l) and (m).
(l) . . . fuit interim effugium legionibus in castra quibus Veterum nomen est.
(‘ . . . and so the legions meanwhile were able to escape to the camp that was
called Vetera.’ Tac. Hist. 4.18.3)
(m) . . . Q. Metellus praetor, cui ex virtute Macedonici nomen inditum erat.
(‘ . . . Quintus Metellus the praetor, who received the cognomen of Macedo-
nicus due to his valour in this war.’ Vell. 1.11.2)
15
For all sorts of errors, also on inscriptions, see Sz.: 426.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
cherishing its own native citizens, thought that it was honoured itself in his honour
and was magnified by his dignity.’ Cic. Planc. 22)
(b) Nostrae naves duae tardius cursu confecto in noctem coniectae . . . contra
Lissum in ancoris constiterunt.
(‘Our two ships, overtaken by night owing to the slow progress of their course, . . .
anchored opposite Lissus.’ Caes. Civ. 3.28.1)
(c) . . . C. Heium iuratum dicere audistis isti navem onerariam maximam Mes-
sanae esse publice coactis operis aedificatam.
(‘ . . . Gaius Heius, the most distinguished person in that city, has stated on oath in
your hearing that a large cargo ship was built for Verres at Messana by workmen
officially impressed.’ Cic. Ver. 2.13)
(d) Quemadmodum si quis statuas marmoreas muliebres stolatas, quae carya-
tides dicuntur, pro columnis in opere statuerit . . .
(‘For example, if anyone in his work sets up, instead of columns, marble statues of
long-robed women which are called caryatids . . . ’ Vitr. 1.1.5)
It is much more common for the modifier to agree with only one of the members of
the compound noun phrase, almost always with the nearest one. The various options
are: M(odifier) N(oun or noun phrase) & N(oun or noun phrase), N & N M, N M &
N, N & M N. In such sequences it is not always evident that the modifier has to be
understood with all members of a compound noun phrase. With possessive adjec-
tives, determiners, and quantifying expressions this is more obvious than with
16
For a collection of examples of the various types in this section, see Meister (1916: 106–10).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
descriptive adjectives (unless in certain syntactic configurations). The four types are
illustrated below.
(i) M N & N In (d) and (e), the modifier precedes the compound noun phrase and
agrees with the first member in number/gender/case. In (d), the possessive adjective
nostra, which agrees with vicinitate, has to be understood with amore as well. In (e),
the ablatival noun phrase summo . . . copia modifying vir (a so-called ablative of
description—see § 11.63) requires a (descriptive) modifier, which is expressed only
once with ingenio, but is to be understood with the others.
(d) . . . quod visus sum debere pro nostra vicinitate et amore scribere potissi-
mum ad te.
(‘ . . . thinking that in view of our nearness and our affection it is to you particularly
that I should dedicate it.’ Var. R. 3.1.9)
(e) . . . Aristoteles, vir summo ingenio, scientia, copia . . .
(‘ . . . Aristotle, a man of supreme genius, knowledge, and fertility of speech . . . ’ Cic.
Tusc. 1.7)
(ii) N & N M In (f ) and (g), the modifier follows the compound noun phrase and
agrees with the last member in number/gender/case. In (f ), the modifier is a posses-
sive adjective. In (g), the quantifying adjective omne goes with both nouns. In (h), the
combination of populus and plebs is a standard one, and the geographical adjective
Romanae belongs to both. Note that further in the same sentence, Latino only goes
with nomini. Ex. (i) is also cited as an illustration of this type,17 but there is no proof
that vacuae consilio etc. modifies the whole compound noun phrase: it is the context
that suggests such an interpretation. Ex. (j) is cited as a combination of types one and
two, but there is no proof that nimius also modifies querella.
(f ) Praestet idem ingenuitatem et ruborem suum verborum turpitudine et
rerum obscenitate vitanda.
(‘He must also testify to his own good birth and modesty, by avoiding all unseemly
language and offensive gestures.’ Cic. de Orat. 2.242)
(g) Quaenam balaena meum voravit vidulum, / aurum atque argentum ubi
omne compactum fuit?
(‘What whale swallowed my trunk, where all my gold and silver were packed.’ Pl.
Rud. 545–6)
(h) ‘Divi divaeque’ inquit (sc. Scipio) . . . ‘vos precor quaesoque uti quae in meo
imperio gesta sunt geruntur postque gerentur, ea mihi populo plebique
Romanae, sociis nominique Latino . . . bene verruncent . . .
(‘Ye gods and goddesses . . . I pray and beseech you that whatever under my authority
has been done, is being done, and shall henceforth be done, may prosper for me, for
the Roman people and the commons, for allies and Latins . . . ’ Liv. 29.27.1)
17
See K.-St.: I.54.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(i) Nam divitiae, nomen, opes vacuae consilio et vivendi atque aliis imperandi
modo dedecoris plenae sunt et insolentis superbiae . . .
(‘For riches, names, and power, when they lack wisdom and the knowledge of
how to live and to rule over others, are full of dishonour and insolent pride . . . ’
Cic. Rep. 1.51)
(j) Noli ista meditari atque illud cave potius ne tua ista querella dolorque nimius
ab illis sapientissimis viris reprendatur.
(‘Never think about those things. Take care rather lest that querulousness and
excessive grief of yours be reproved by those men of consummate wisdom.’ Cic.
Planc. 51)
(iii) N M & N In (k) and (l) the modifier follows the first member of the compound
noun phrase and has to be understood with the second member: in (k) a possessive
adjective, in (l) a determiner, and in (m) an evaluative adjective magni, which is a
modifier in a genitival noun phrase modifying vir (a required modifier in this type of
phrase; see § 11.48).
(k) Sed tamen cum omnis gradus aetatis recordor tuae cumque vitam tuam ac
studia considero . . .
(‘But all the same when I recall all the stages of your career, and when I contemplate
your life and pursuits . . . ’ Cic. de Orat. 3.82)
(l) Difficile enim dictu est quaenam causa sit cur . . . ab iis celerrime fastidio
quodam et satietate abalienemur.
(‘For it is hard to say why exactly it is that . . . we are most speedily estranged from
these things by a feeling of disgust and satiety.’ Cic. de Orat. 3.98)
(m) . . . Gaius Volusenus, tribunus militum, vir et consilii magni et virtutis . . .
(‘ . . . Gaius Volusenus, a military tribune, a man of great sagacity and courage . . . ’
Caes. Gal. 3.5.2)
(iv) N & M N The fourth type, with the modifier combined with the last member, is
one of the forms of the so-called figura apo koinou (Ie ίF).18 Apart from a few
debatable instances in Cicero and Tacitus,19 this type is almost restricted to poetry. In
contrast to the other three types, the modifiers are mainly descriptive adjectives.
Among the instances cited several are not very convincing. Early examples cited in the
literature are (n) and (o). In (n), aerumnae can be understood without a possessive
adjective. In (o), there is no need to understand the quantifier omni(bus) with opibus:
opibus alone is a normal expression. (NB: Coordination of opes with copiae in the
plural is well attested, but this does not prove that omni has to be taken with opibus.)20
18
For this figura, see Sz.: 834–6, with references to studies on individual authors.
19
See K.-St.: I.54–5; II.559.
20
For parallels, see TLL s.v. ops 813.4ff. For a different view, see Leo (1896: 7–11).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
Examples from poetry are (p)–(r). In (p), auribus is not immediately interpreted as
Cato’s ears, so it is likely that Catullus wanted ‘your’ to be understood with auribus as
well. Exx. (q) and (r) have descriptive adjectives. There is no proof that the modifiers
have to be interpreted with the first members, although there is nothing wrong with it
and this is probably what the authors wanted.
(p) O rem ridiculam, Cato, et iocosam / dignamque auribus et tuo cachinno.
(‘O, Cato, what an absurdly funny thing, worthy of your ears and laughter!’ Catul. 56.1–2)
(q) Fit quoque ut interdum venti vis missa sine igni / igniscat tamen in spatio
longoque meatu . . .
(‘It comes to pass, too, sometimes that the force of the wind, starting without fire, yet
catches fire on its course and its long wandering . . . ’ Lucr. 6.300–1—tr. Bailey)
(r) Iam satis terris nivis atque dirae / grandinis misit Pater . . .
(‘Enough already of dire snow and hail has the Father sent upon the earth . . . ’ Hor.
Carm. 1.2.1–2)
There is one clear instance (s) in which the modifier is found after the compound
noun phrase in agreement with the first member, which is said to be more important
in its context.21 The two nouns go closely together, but there may also be a structural
reason: putting relatively complex magna magis quam firma before corpora would be
strange, and magnos magis quam firmos in agreement with animos could create the
impression that the modifier only belongs to animos.
(s) Gens est (sc. Gallorum) cui natura corpora animosque magna magis quam
firma dederit.
(‘That people . . . to whom nature has given bodily size and courage, great
indeed, but vacillating.’ Liv. 5.44.4)
Related is the question concerning the agreement of a gentilicium with two or more
praenomina. Gentilicia developed from adjectives to nouns. This explains the exist-
ence of instances like (t) in the pre-Classical period, with sg. Pomplio (adjectives with
compound nouns were regularly singular). When the gentilicia came to be regarded
as substantival (names), they were put in the plural (which is normal for nouns in
apposition). So in a later inscription, (u), Minucieis. This is the rule in Classical texts,
as in (v) and (w)—the latter with explicit coordination—with a few problematic
exceptions.22 Cognomina behave as nouns (with notable exceptions).
21
So K.-St.: I.55. The other two examples in K.-St. can be interpreted differently.
22
For this entire note, see Meister (1916: 99–112).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Whereas in (a) and (b) the shared noun is singular, (c) has plural portas, which is
understandable, since there is more than one entity involved. In fact, Livy has the
noun in the plural when it precedes, as in (c), and in the singular when it follows, as in
(d). Later authors vary. In (e), the plural may have been necessary because sg. statuae
would have suggested that there was only one statue representing a horseman and a
foot-soldier.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(‘On the occasion of the dedication of two statues, one on horseback and one
standing, he gave olive oil to the people of both sexes.’ CIL V.7007.12–13 (Torino,
after AD 70))
There are a few problematic instances. In Cato Agr. 2.5 Rationes putare argentariam
frumentariam, with an apparently remarkable plural noun, the two adjectives may be
regarded as independent specifications. In Caes. Civ. 3.110.3, Huc accedebant collecti
ex praedonibus latronibusque Syriae Ciliciaeque provinciae finitimarumque regio-
num, need not be an exceptional instance of postposition of the singular noun
provinciae, if it is only taken with Ciliciae.
In Cic. Agr. 2.96, Agros vero Vaticanum et Pupiniam cum suis opimis atque
uberrimis campis conferendos scilicet non putabunt, feminine Pupiniam is strange
(for Pupinia = ager Pupinius, see OLD s.v.).
In clauses with the copula sum or other copular verbs in which the subject is a
demonstrative, anaphoric, relative, or interrogative pronoun and the subject comple-
ment a noun or noun phrase, Latin has two options for the form of the pronoun. The
most common option, at least in Classical Latin, is to make the pronoun agree with
the subject complement in number and gender (of course also in case). In the
standard grammars this phenomenon is labelled ASSIMILATION or ATTRACTION. Another
term is BACK AGREEMENT.23 An example with a relative pronoun is (a), with a demon-
strative (b). For speakers of a modern language like English the use of quod in (a) in
23
The term is used by Corbett (2006: 63).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Authors vary in the degree to which they follow the ‘agreement’ or the ‘independent’
strategy. Caesar follows the agreement strategy almost without exception (there is a textual
problem in Gal. 5.49.1). Cicero more often follows the agreement strategy, but prefers the
independent strategy for the resumptive pronoun id when it refers to the autonomous
relative clause, as in (e) (id not ea), and to refer to abstract ideas, as in ex (f ).
(e) Quid tum? Quod ita erit gestum, id lex erit?
(‘And then what? Will something transacted in this fashion be a law?’ Cic. Phil. 1.26)
(f ) Fac sane esse summum bonum non dolere—quamquam id non vocatur
voluptas, sed non necesse est nunc omnia— . . .
(‘Suppose, if you like, that the highest good is absence of pain; although that is not termed
pleasure—but there is no need to go into everything now— . . . ’ Cic. Tusc. 3.40)
Examples of the four types of pronouns mentioned at the beginning of this section in
which the agreement option is chosen are (g)–(j), a demonstrative, anaphoric,
relative, and interrogative pronoun, respectively.
(g) Attat hoc illud est, / hinc illae lacrumae, haec illa’st misericordia.
(‘Ay, ay, this is it; hence those tears, hence that sympathy.’ Ter. An. 125–6)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(h) . . . verum tamen / is est honos homini pudico meminisse officium suom.
(‘ . . . but still it is an honour for a decent man to remember his duty.’ Pl. Trin. 696–7)
(i) Hoc Epicurus in voluptate ponit, quod summum bonum esse vult . . .
(‘Epicurus locates this quality in pleasure, which he maintains is the highest good . . . ’
Cic. Fin. 1.29)
(j) Quae illaec res est?
(‘What’s that?’ Pl. Mos. 935)
Similarly, a pronoun that is the object in its clause agrees with its object complement
in number and gender. Examples are (k) and (l).
(k) Eas divitias, eam bonam famam magnamque nobilitatem putabant.
(‘This they considered riches, this fair fame and high nobility.’ Sal. Cat. 7.6)
(l) . . . non plausum illum sed iudicium puto.
(‘ . . . I do not regard it as applause but as a verdict.’ Cic. Phil. 1.37)
Supplement:
Demonstrative and anaphoric pronouns: Quid illud est? # Pro deum fidem quid est,
si haec (hoc cj. Bentley) non contumelia’st? (Ter. An. 237); Sed haec erat spinosa
quaedam et exilis oratio longeque ab nostris sensibus abhorrebat. (Cic. de Orat. 1.83);
Hanc NÆ appellat Epicurus, id est aequabilem tributionem. (Cic. N.D. 1.50—
NB: Greek word); . . . non erit ista amicitia sed mercatura quaedam utilitatum
suarum. (Cic. N.D. 1.122); Galli . . . ad Caesarem omnibus copiis contendunt. Haec
erant armata[e] (corr. Nipperdey) circiter milia LX. (Caes. Gal. 5.49.1–2); . . .
Labienus . . . cum IIII legionibus Luteciam proficiscitur. Id est oppidum Parisiorum . . .
(Caes. Gal. 7.57.1); Sed haec inter bonos amicitia, inter malos factio est. (Sal. Jug. 31.15);
Itaque . . . Calavios Ovium Noviumque—ea capita coniurationis fuerant— . . . mors . . .
iudicio subtraxit. (Liv. 9.26.7—NB: Prius animadversum in eos qui capita coniurationis
fuerant a consule scribit Claudius. (Liv. 8.19.13)); Leucade haec sunt decreta—id caput
Acarnaniae erat—eoque in concilium omnes populi conveniebant. (Liv. 33.17.1); Vivant
Galataeque Syrique, / Cappadoces Gallique extremique orbis Hiberi, / Armenii, Cilices.
Nam post civilia bella / hic populus Romanus erit. (Luc. 7.540–3); Numero (sc.
armentorum et pecorum) gaudent, eaeque solae et gratissimae opes sunt. (Tac. Ger. 5.1);
Relative pronouns: In nutricatu, quam porculationem appella[ba]nt, binis mensibus
porcos sinunt cum matribus. (Var. R. 2.4.13); . . . incedunt in locum unum, quod alii
vocant peristerona, alii peristerotrophion . . . (Var. R. 3.7.2—NB: Greek words); . . .
est animi lucem splendoremque fugientis iustam gloriam, qui est fructus verae
virtutis honestissimus, repudiare. (Cic. Pis. 57); Sed demus tibi istas duas sumptiones
(ea quae λήμματα appellant dialectici, sed nos Latine loqui malumus) . . . (Cic. Div.
2.108—NB: Greek word); Huc enim pertinet, animal hoc providum . . . plenum
rationis et consilii, quem vocamus hominem, praeclara quadam condicione gene-
ratum esse a supremo deo. (Cic. Leg. 1.22); . . . mundus hic totus, quod domicilium
quamque patriam di nobis communem secum dederunt . . . (Cic. Rep. 1.19); . . . morbi
conficiuntur, quae vocant illi νοσήματα . . . (Cic. Tusc. 4.23—NB: Greek word); . . .
eamque, quam lenitatem nos dicimus, vitioso lentitudinis nomine appellant. (Cic.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Greek (and other foreign) words and geographical names deserve special mention. Of
these no general knowledge could be expected nor a precise knowledge of their
number and gender features. As a consequence, when they function as subject or
object pronoun, the agreement option is typically avoided, even in Cicero and Caesar.
In (m), the resumptive anaphoric pronoun eam has the number/gender features of
contentio in the preceding autonomous relative clause. The relative pronouns in (n)–
(p) agree with their head noun in the superordinate clause.
(m) Sed quae ex statu contentio efficitur, eam Graeci ŒæØ (sc. vocant) . . .
(‘The Greeks call the debate which arises from the issue ŒæØ . . . ’ Cic. Top. 95)
(n) Nec minus admirandum quod fit in floribus, quos vocant heliotropia . . .
(‘No less remarkable is the behaviour of the flowers which they call “heliotropes” . . . ’
Var. R. 1.46.1)
(o) . . . ad eum locum qui appellabatur Palaeste . . . milites exposuit.
(‘ . . . he disembarked his troops at a place that was called Palaeste . . . ’ Caes. Civ. 3.6.3)
(p) . . . est sinus Euboicus quem Coela vocant suspectus nautis.
(‘ . . . the Euboean gulf, which they call Coela, is dangerous to mariners.’ Liv. 31.47.1)
Supplement:
Foreign words: . . . cohibere motus animi turbatos quos Graeci πάθη nominant appe-
titionesque quas illi ὁρμάς . . . (Cic. Off. 2.18); Est etiam genus radicis . . . quod appel-
latur chara quod admixtum lacte multum inopiam levabat. (Caes. Civ.
3.48.1); . . . unus erat toto naturae vultus in orbe, / quem dixere Chaos . . . (Ov. Met.
1.6–7); Facit idem . . . et radix, quam Graeci σίλφιον vocant . . . (Col. 6.17.7);
Geographical names: Est genus quoddam hominum quod Hilotae vocatur . . . (Nep.
Paus. 3.6); Inde Phocaeam petentes ad insulam quam Bacchium vocant—imminet
urbi Phocaeensium—adpulerunt . . . (Liv. 37.21.7); Sponsalia in ea parte Italiae quae
Latium appellatur hoc more atque iure solita fieri scripsit Servius Sulpicius . . . (Gel. 4.4.1);
Comparable uses of adjectival words functioning as subject of the clause with a noun
phrase functioning as the subject complement are illustrated by (q) and (r).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(q) . . . ut qui virtutem praemio metiuntur nullam virtutem nisi malitiam putent.
(‘ . . . therefore those who measure virtue by the reward it brings believe that
nothing is a virtue except vice.’ Cic. Leg. 1.49)
(r) Nam si omnia facienda sint quae amici velint non amicitiae tales sed
coniurationes putandae sint.
(‘For supposing that we were bound to do everything that our friends
desired, such relations would have to be accounted not friendships but
conspiracies.’ Cic. Off. 3.44)
It is obvious that the agreement and independent options may be available at the
same time. Many of the so-called exceptions to back agreement are of this type, as in (s).
Hoc refers to the event of departure mentioned before, and has the correct n. sg. form
for cross-clausal reference to events (see § 13.27). Livy might have written haec, which
would constitute clause-internal agreement with the subject complement.
(s) Si hoc profectio et non fuga est.
(‘If, indeed, this is a departure and not rather a flight.’ Liv. 2.38.5)
Essentially the same rules and tendencies mentioned above hold for agreement across
the border of clauses and sentences.
use) or independently (their so-called substantival use). In the latter case they derive their
number/gender properties from a noun or noun phrase in a preceding clause or sentence
to which they refer. Alternatively, they refer to the content of a preceding clause or
paragraph as a whole. In that case a n. sg. form of the pronoun is used. These three
possibilities are illustrated by anaphorical(ly used) pronouns in (a)–(c). In (a), the old
man Hegio has noted the parasite Ergasilus before and refers to him with hunc hominem
(hunc agrees with hominem). In (b), hanc independently refers to the young woman
mentioned in the preceding context. In (c), id independently refers to the story about
Amphitruo’s brave deeds known from the preceding context.
(a) Perlubet hunc hominem conloqui. Ergasile!
(‘I really want to speak with him. Ergasilus!’ Pl. Capt. 833)
(b) Quid ergo hanc dubitas conloqui?
(‘Then why do you hesitate to address her?’ Pl. Mil. 1008)
(c) An etiam id tu scis? # Quipp’ qui ex te audivi, ut urbem maxumam /
expugnavisses regemque Pterelam tute occideris.
(‘Do you know about that? # Of course! I heard from you how you conquered the
great city and slew King Pterelas yourself.’ Pl. Am. 745–6)
The relative pronoun in an adnominal relative clause most often derives its number
and gender from its head. In (d), qui agrees with its head homo; in (e), qua with urbe;
in (f ), quibus with castra.
(d) O Libane, uti miser est homo qui amat.
(‘Oh Libanus! How miserable a chap is when he’s in love.’ Pl. As. 616)
(e) Videsne igitur vel in ea ipsa urbe in qua et nata et alta sit eloquentia quam ea
sero prodierit in lucem?
(‘You see thus . . . even in that city in which eloquence was born and grew to maturity,
how late it stepped forth into the light of day . . . ’ Cic. Brut. 39)
(f ) Inde iam duxit ad Alcen urbem, ubi castra Celtiberorum erant, a quibus
venerant nuper legati.
(‘From there he next led his troops to the town of Alce, where the camp of the
Celtiberians lay from which the ambassadors had recently come.’ Liv. 40.48.1)
In the case of apposition, there are two possibilities: the relative agrees either with the
appositive, as in (g), or with the head noun, as in (h) (the appositives in both these
examples are restrictive, for which see § 11.81). An example of agreement with a non-
restrictive appositive is (i).
(g) . . . undique loci natura Helvetii continentur. . . . tertia (sc. parte) lacu Le-
manno et flumine Rhodano, qui provinciam nostram ab Helvetiis dividit.
(‘ . . . the Helvetii are confined on all sides by the nature of their territory . . .
on the third [side, they are confined] by the Lake of Geneva and the river
Rhone, which separates our province from the Helvetii.’ Caes. Gal. 1.2.3)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(h) Hostes protinus ex eo loco ad flumen Axonam contenderunt, quod esse post
nostra castra demonstratum est.
(‘The enemy hurried immediately from their station to the river Aisne,
which, as has been shown, was behind our camp.’ Caes. Gal. 2.9.3)
(i) Duo propugnacula belli Punici, Cn. et P. Scipiones, qui Carthaginiensium
adventum corporibus suis intercludendum putaverunt, quid?
(‘What of the two bulwarks of the Punic War, Gnaeus and Publius Scipio,
who deemed it their duty to bar with their own bodies the onset of the
Carthaginians?’ Cic. Parad. 1.12)
24
For Cic. Tusc 2.9 (needlessly emended), see Bohlin (2005). For more examples, see Sz.: 431–2. For
Cato, see Adams (forthc.).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(ii) If the members of a compound noun phrase are all inanimate entities or if they
are mixed animate/inanimate entities, the anaphoric or relative pronoun that refers to
them is usually neuter plural. Examples are (d)–(g). In (f ) and (g), the members are all
feminine, but nevertheless a neuter plural pronoun is found in the next clause.
(d) Aurum atque vestem muliebrem omnem habeat sibi, / quae illi instruxisti.
(‘All the jewellery and clothes you fitted her out with, let her keep.’ Pl. Mil.
1099–1100)
(e) Quoi quom . . . omnia . . . parerent, domi otium atque divitiae, quae prima
mortales putant, adfluerent, fuere tamen cives qui . . .
(‘The whole world . . . rendered obedience to her; at home there was peace and an
abundance of wealth which mortal men deem the chiefest of blessings. Yet there were
citizens who . . . ’ Sal. Cat. 36.4)
(f ) (sc. civis) . . . ita iustitiae honestatique adhaerescet ut dum ea conservet
quamvis graviter offendat . . .
(‘ . . . but he will surely cleave to justice and honour so closely that he will submit to
any loss, however heavy, provided that he preserve them . . . ’ Cic. Off. 1.86)
(g) Conciliantur autem animi dignitate hominis, rebus gestis, existimatione
vitae; quae facilius ornari possunt, si modo sunt, quam fingi, si nulla sunt.
(‘Now feelings are won over by a man’s merit, achievements, or reputable life,
qualifications easier to embellish, if only they are real, than to fabricate where non-
existent.’ Cic. de Orat. 2.182)
Supplement:
Anaphoric(ally used) pronouns: . . . tam natura putarem hominis vitam sustentari
quam vitis, quam arboris. Haec enim etiam dicimus vivere. (Cic. Tusc. 1.56); . . . deinde
tuam domum tuosque agros, eaque remoto, salubri, amoeno loco. (Cic. Fam. 7.20.2);
. . . incedunt per ora vostra magnifici, sacerdotia et consulatus, pars triumphos suos
ostentantes. Proinde quasi ea honori, non praedae habeant. (Sal. Jug. 31.10);
Relative pronouns: Delectabatur cereo funali et tibicine, quae sibi nullo exemplo
privatus sumpserat. (Cic. Sen. 44); Vicit disciplina militaris, vicit imperii maiestas,
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
quae in discrimine fuerunt an ulla post hanc diem essent. (Liv. 8.35.4); Attali regis
legatus naves captivosque quae ad Chium navali proelio capta essent . . . pro incor-
ruptis restitui (sc. postulabat). (Liv. 32.33.5—NB: captivos are human beings);
Instrumenti eius et supellectilis parsimonia apparet etiam nunc residuis lectis
atque mensis, quorum pleraque vix privatae elegantiae sint. (Suet. Aug. 73.1—NB:
both quorum [quarum mss.!] and pleraque are neuter);
There is an exceptional instance of a plural relative clause following a disjunctive
coordination with aut, (h). The text is usually emended in such a way that the relative
clause agrees with the first member of the coordination (puerum).
(h) Mater ancillas iubet, / . . . / . . . quaerere / puerum aut puellam qui supponan-
tur (supponatur cj. Acidalius) mihi.
(‘Mother told the maids . . . to seek out a boy or girl whom I could pass off as
my own.’ Pl. Truc. 401–4)
(iii) For relative pronouns referring to a compound noun phrase there exists an
alternative, which is more common, viz. agreement with one of the members of the
compound noun phrase. The member involved is more often the nearest than the
farthest one, as in (i)–(k). Note in (i) that, besides the relative pronoun, nulla and erit
are also singular. Ex. (l) shows agreement with the first member.
(i) Virgo atque mulier nulla erit quin sit mala / quae praeter sapiet quam placet
parentibus.
(‘There won’t be a girl or woman who isn’t bad, if she knows more than her parents
like.’ Pl. Per. 365–6)
(j) . . . omnia aliorum causa esse generata, ut eas fruges atque fructus quos terra
gignit animantium causa . . .
(‘ . . . everything else . . . was created for the sake of some other thing; thus the corn
and fruits which the earth produces were created for the sake of animals . . . ’ Cic. N.D.
2.37—NB: the determiner of the entire noun phrase agrees with fruges)
(k) Ipsos oppida vicosque quos incenderant restituere iussit.
(‘He also ordered them to restore with their own hands the towns and villages which
they had burnt.’ Caes. Gal. 1.28.3)
(l) Tu, si me diligis, fruere isto otio tibique persuade praeter culpam ac pecca-
tum, qua semper caruisti et carebis, homini accidere nihil posse quod sit
horribile aut pertimescendum.
(‘If you care for me, enjoy your present leisure, and persuade yourself that no matter
for fear or trembling can befall a man apart from fault and wrongdoing, of which you
are free and always will be.’ Cic. Fam. 5.21.5)
Supplement:
Grammatical agreement with the last member: Mihi enim numquam venerat in
mentem furorem et insaniam optare vobis in quam incidistis. (Cic. Pis. 46); . . . (sc.
experimentum) factum etiam nuper in Italia servili tumultu, quos tamen aliquid usus
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
ac disciplina, quam (v.l. quae) a nobis accepissent, sublevarent. (Caes. Gal. 1.40.5);
(sc. data sunt) . . . oppida vici castella agri qui ad Pisidiam vergunt . . . (Liv. 37.56.6);
Grammatical agreement with the first member: . . . SI QUAE LEX PLEBEVE SC(ITUM) EST
QUAE MAG<ISTRATUM> QUEM EX H(ACE) L(EGE) <DE ALIQUA RE DECERNERE OPORTET DE EA RE /
2
DECERNERE VETET . . . > (CIL I .585.41 (Lex Agr., Rome, c.111 BC)); Provinciae Galliae
M. Fonteius praefuit, quae constat ex iis generibus hominum et civitatum qui, ut
vetera mittam, partim nostra memoria bella cum populo Romano acerba ac diuturna
gesserunt, partim . . . (Cic. Font. 12); Feriarum festorumque dierum ratio in liberis
requietem litium habet . . . Quas conpositio anni conferre debet ad perfectionem ope-
rum rusticorum. (Cic. Leg. 2.29); . . . omnibus vicis aedificiisque, quos (v.l. quo) adire
potuerant, incensis . . . (Caes. Gal. 2.7.3); At ubi febres sunt, satius est eius rei causa
cibos potionesque assumere qui simul et alant et ventrem molliant. (Cels. 2.12.1);
Notional agreement of the verb with the subject of its clause is relatively common in
Early Latin (especially comedy), in particular with collective nouns. It is avoided by
Cicero and Caesar, though not entirely absent; it is more common in Sallust and quite
common in Livy (but his subject matter contributes to this relative frequency), in
poets, and in later prose authors.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(i) The collective nouns that are most common in controlling the number (and, for
the nominal part, gender) of the verb are those referring to an unordered collective
entity, such as multitudo ‘multitude’, pars ‘part’, and vis ‘force’, iuventus ‘youth’,
nobilitas ‘nobility’, and plebs ‘people’. Less common are collectives that refer to an
organized institution, such as classis ‘fleet’, cohors ‘cohort’, and exercitus ‘army’.25
Examples of the former type are (a)–(d). Note that the concept of plurality is often
reinforced by a plural noun phrase in the genitive or by the preceding context. (For
collectively used miles ‘the soldiers’, see (f ) in § 13.32(i).)
(a) Atque <ea> haud est fabri culpa, sed magna pars / morem hunc induxerunt.
(‘And this is not the builder’s fault at all, but most people have adopted this custom.’
Pl. Mos. 114–15)
(b) Tum meretricum numerus tantus quantum in urbe omni fuit / obviam
ornatae occurrebant suis quaeque / amatoribus.
(‘And then the harlots—as many as were in the entire city—all decked out, were
running up to meet their own special lovers.’ Pl. Epid. 213–15—NB: secondary
predicate ornatae feminine because of meretricum; also quaeque)
(c) Mortalium autem pars in hominum, pars in bestiarum genere numerantur
(v.l. numeratur).
(‘Of those concerned with mortal beings, part are reckoned as belonging to human
beings, and part to beasts.’ Cic. Inv. 1.35)
(d) Nam cum tanta multitudo lapides ac tela coicerent, in muro consistendi
potestas erat nulli.
(‘For when so vast a host hurled stones and missiles, no one had the ability to stand
firm on the wall.’ Caes. Gal. 2.6.3)
Supplement (in alphabetical order by noun):
Tunc enimvero omnis aetas currere obvii . . . (Liv. 27.51.1—NB: secondary predicate
in the plural); HUIC / DECURIA·EX·AERE·CONLATO·IMAGINEM / DECREVERUNT (CIL
VI.4421.4–5 (Rome, end I BC)); Omnis enim multitudo oppidanorum . . . advenienti
Caesari occurrerunt seque ei dediderunt. (B. Alex. 32.3); . . . multitudo clamore in-
genti nunc consulum, nunc patrum fidem inplorant . . . (Liv. 3.21.1); Igitur primum
correpti qui fatebantur, deinde indicio eorum multitudo ingens haud proinde in
crimine incendii quam odio humani generis convicti sunt. (Tac. Ann. 15.44.4);
Conveniunt a Tyro et ex aliis regionibus multitudo, etiam transmarina (Tert. Marc.
4.13.6); Pars ludicre saxa / iactant, inter se licitantur (Enn. Ann. 73–4V=69–70S);
Nimioque hic pluris pauciorum gratiam / faciunt pars hominum . . . (Pl. Trin. 34–5);
Motae sonuere colubrae, / parsque iacent umeris, pars circum pectora lapsae / sibila
dant . . . (Ov. Met. 4.492–4—NB: secondary predicate in the plural); . . . MAIOR PARS
2
DECURIONUM . . . CENSUERINT . . . (CIL I .593.150 (Lex. Jul. Munic., 45 BC)); Equorum
pars magna nantes loris a puppibus trahebantur . . . (Liv. 21.27.9—NB: secondary
25
Instances of notional agreement are regularly recorded in dictionaries in the lemmata of the
controlling nouns. See especially TLL s.v. multitudo 1601.67ff. and s.v. pars 454.68ff. See also Sz.: 436–7.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
predicate in the plural); Pars navium haustae sunt . . . (Tac. Ann. 2.24.2—NB: haustae
feminine because of navium); . . . pars Armeniae, ut cuique finitima, . . . Antiocho
parere iussae sunt. (Tac. Ann. 14.26.2); Haec fremunt plebes. (Liv. 3.38.10); Omnium
primum avidum novae libertatis populum, ne postmodum flecti precibus aut donis
regiis posset, iure iurando adegit neminem Romae passuros regnare. (Liv. 2.1.9);
Simul ex minoribus navibus magna vis eminus missa telorum multa nostris . . .
vulnera inferebant. (Caes. Civ. 2.6.3); . . . magna vis hominum simul immissi corbi-
bus fudere in Tiberim . . . (Liv. 2.5.3—NB: secondary predicate in the plural);
Examples of notional agreement with the nouns of the second type (organized
institutions) mentioned above are (e)–(g). A few other nouns are given in small type.
(e) Cetera classis praetoria nave amissa quantum quaeque remis valuit fugerunt.
(‘The rest of the fleet, when their flagship was lost, fled as fast as the oars could drive
them.’ Liv. 35.26.9)
(f ) Cuius conscientia ac fama ferox exercitus nihil virtuti suae invium et pene-
trandam Caledoniam inveniendumque tandem Britanniae terminum con-
tinuo proeliorum cursu fremebant.
(‘Upon realizing this and hearing the report, the army began to cry that nothing
could bar the way before its courage, that Caledonia must be penetrated, that the
farthest shores of Britain must once for all be discovered in one continuous cam-
paign.’ Tac. Agr. 27.1)
(g) Tunc inhonora cohors laceris insignibus aegras / secernunt acies . . .
(‘Then, an inglorious troop with battered standards, they separate their failing
lines . . . ’ Stat. Theb. 10.8–9)
Supplement (in alphabetical order by noun):
At chorus aequalis dryadum clamore supremos / implerunt montis. (Verg. G.
4.460–1); Parthyaeorum deinde gens . . . claudebant agmen. (Curt. 4.12.11); Hic
genus antiquom Terrae, Titania pubes, / fulmine deiecti fundo volvontur in imo.
(Verg. A. 6.580–1—NB: secondary predicate in the plural); Populus aestuat diversa
tendentes . . . (Apul. Met. 2.29.6—NB: secondary predicate in the plural); . . .
catulorum blanda propago / discutere et corpus de terra corripere instant . . . (Lucr.
4.997–8); . . . cum interea totum eius servitium hilares sunt atque epulantur, nec in ea
re quicquam efficit condicione. (Apul. Flor. 23.5—NB: Hilares is the subject com-
plement. Totum may have favoured this form of agreement.); Sed et vulgus iam
sciunt (scit Vulg.) Christum (Tert. Apol. 21.3);
NB: Sed quid huc tantum hominum incedunt? (Pl. Poen. 619); Di vostram fidem, /
quid turbae’st apud forum! Quid illi hominum litigant! (Ter. An. 744–5);
There are a few instances of notional agreement with plural collective nouns in which
natural gender (or sex) is involved: legiones ‘legions’ understood as ‘soldiers’, etc.
Most instances are cross-clausal (see § 13.32). Examples are (h) and (i).
(h) . . . aliquot turmae cum levi armatura impetu facto loco sunt deiecti et . . . protriti.
(‘ . . . several squadrons of ours with some light-armed troops were attacked
and dislodged from their position . . . and completely crushed.’ B. Hisp. 14.2)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(ii) There are virtually no instances of subjects referring to countries, cities, etc. in
which the verb takes the plural as if agreeing with the inhabitants as the subject
instead. An almost unique example is (m), where, omnis may have favoured notional
agreement. (For instances of cross-clausal notional agreement see § 13.33.)
(m) Leonides Laco, qui simile apud Thermopylas fecit, propter eius virtutes
omnis Graecia gloriam atque gratiam praecipuam claritudinis inclitissimae
decoravere monumentis . . .
(‘The Laconian Leonidas, who performed a like exploit at Thermopylae—because of
his valour all Greece marked their unexampled praise and gratitude with memorials
of the highest distinction . . . ’ Cato hist. 83)
(iii) Various singular pronouns imply a plurality and may therefore trigger notional
agreement. Uterque ‘each of the two’ is the example that is most often cited. Notional
agreement can occur with this pronoun both when it is used independently and when
it is used as a determiner, as is shown in (n)—note its plural modifier eorum—and (p),
respectively. Another example is quisque, both independent, as in (o), and as a
determiner, as in (q) and (r)—in combination with a noun and a plural modifier and
26
For exx. (a) and (b), see Baehrens (1912: 455–7); for ecclesiastical authors, see Hoppe (1932: 49–51);
Sz.: 437.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(n) . . . eodemque die uterque eorum ex castris stativis a flumine Apso exercitum
educunt, Pompeius clam et noctu Caesar palam atque interdiu.
(‘ . . . and on the same day they each led out their forces from their permanent camps,
quitting the River Apsus, Pompeius secretly by night, Caesar openly by day.’ Caes.
Civ. 3.30.3)
(o) Iamque cognito casu noxiae potionis varie quisque praesentium auctores
insimulabant extremi facinoris.
(‘When they discovered the tragedy of the poisoned drink, everyone present began to
accuse different people of responsibility for the monstrous crime.’ Apul. Met. 10.5.2)
(p) Nam forte utraque acies simul conclamavere . . .
(‘For, as it happened, both armies shouted at once . . . ’ Liv. 33.9.2)
(q) Nam etiam in concordia suum quaeque plebs morem figurandi ceras fin-
gendique servant.
(‘For even if they live in harmony together, each community keeps to its own manner
of shaping and constructing its waxen cells.’ Col. 9.15.7)
(r) . . . intimus quisque amicorum et plerique militares . . . multique etiam ignoti
vicinis e municipiis, pars officium in principem rati, plures illos secuti, ruere
ad oppidum Brundisium . . .
(‘There was a rush of people to Brundisium . . . Every intimate friend was present;
numbers of military men; . . . even many strangers from the local towns, some thinking
it respectful to the emperor, the majority following their example.’ Tac. Ann. 3.1.2)
(s) . . . siquis quid alter ab altero peterent, si ambo pares essent . . .
(‘ . . . if anyone claim anything one from another, and both are equal . . . ’ Cato orat. 51)
Supplement (in alphabetical order by pronoun):
Independent use: Ab his, propter quae creati erant, comitia consularia habita, quia
neuter consulum potuerant (cj. potuerat) bello abesse. (Liv. 9.44.2); Furente deinde
Antonio simulque Lepido, quorum uterque, ut praediximus, hostes iudicati erant . . .
(Vell. 2.66.1); . . . missi confestim honoratissimus quisque e patribus. (Liv. 2.15.1);
Cetera multitudo sorte decimus quisque ad supplicium lecti. (Liv. 2.59.11); Crescebat
interim in dies Sullae exercitus confluentibus ad eum optimo quoque et sanissimo.
(Vell. 2.25.2); Tum intimus quisque libertorum vincti abreptique. (Tac. Ann.
15.35.3); . . . ex iuventute utriusque ordinis profligatissimus quisque . . . famosi iudicii
notam sponte subibant. (Suet. Tib. 35.2);
27
For additional Late Latin instances, see Sz.: 437.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
In § 13.30 (i) several examples are included in which a subject or object complement
shows notional agreement with the subject or object of its clause. A few instances of
secondary predicates are noted there as well. Some further examples of plural
secondary predicates are (a)–(d).
(a) Imitari deinde eos iuventus simul inconditis inter se iocularia fundentes
versibus coepere . . .
(‘Next the youth began to imitate them, at the same time exchanging jests in uncouth
verses . . . ’ Liv. 7.2.5)
(b) Qui scelere vaecors, simul vulgi rumore territus robur et saxum aut parrici-
darum poenas minitantium cessit urbe.
(‘ . . . who, maddened by his crime and terrified also by the comments of the multi-
tude, threatening him with the dungeon and the rock or the penalties of parricide, left
Rome.’ Tac. Ann. 4.29.2)
(c) . . . seditiosissimum quemque vinctos trahunt ad legatum legionis primae
C. Caetronium . . .
(‘ . . . they dragged the most prominent of the mutineers in chains to Gaius Caetro-
nius, legate of the first legion . . . ’ Tac. Ann. 1.44.2)
(d) . . . per aliquot dies incerti rerum omnium suspensique de statu alterius
uterque consul ageret . . .
(‘ . . . for some days each consul was acting in a state of uncertainty about all things
and in a state of suspense about the other’s state.’ Liv. 9.43.4—NB: see also § 13.11)
Supplement:
. . . cum subito pars equitatus Caesaris cum levi armatura contra Gaetulos iniussu ac
temere longius progressi paludemque transgressi multitudinem hostium pauci sustinere
non potuerunt . . . (B. Afr. 61.3); Hic manus ob patriam pugnando volnera passi . . .
(Verg. A. 6.660); Stare omnem multitudinem ad portas viam hinc ferentem prospec-
tantes certum habeo. (Liv. 7.30.22—NB: unless prospectantes goes with portas);
NB: Notional agreement of gender with a plural subject: . . . legiones duae X et VIIII
ex Sicilia navibus onerariis profectae cum iam non longe a portu Ruspinae
abessent conspicati naves . . . ad Caesarem perveniunt. (B. Afr. 53.1); Servitia regum
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Notional agreement in number (and, for its nominal part, in gender) of a verb with a
noun phrase that is grammatically singular but conceptually plural also occurs across
the border of clauses (coordinate clauses, relative clauses, and argument and satellite
clauses) and sentences. This form of agreement is found with nouns, names of cities
and countries (rarely), and pronouns.
(i) In (a), plural imperant agrees with sg. quisquam.28 In (b), putant in the reason
clause is plural, although it refers to sg. multitudinem in the main clause. In (c), pl.
exirent in the object clause refers to sg. civitati in the main clause. In (d), pl. videant in
the relative clause refers to sg. equitatum in the main clause. In (e), pl. decrevere in the
second coordinate clause refers to senatus in the first clause (note also the pl. anaphoric
pronoun ii). In (f ), pl. sternunt refers to collective sg. miles in the preceding sentence.
(a) Superas facile, ut superior sis mihi quam quisquam qui imperant (cj. Ussing
instead of impetrant).
(‘You easily gain the upper hand so that for me you are above any of those who
command me.’ Pl. Men. 192)
(b) . . . multitudinem haec maxime allicit, quod ita putant . . .
(‘ . . . what most attracts the crowd is the fact that they believe . . . ’ Cic. Fin. 1.25)
(c) (sc. Orgetorix) . . . civitati persuasit ut de finibus suis cum omnibus copiis
exirent.
(‘ . . . he persuaded the community that they should march out of their territory in full
force.’ Caes. Gal. 1.2.1)
(d) Idem facit Caesar equitatumque omnem . . . praemittit, qui videant quas in
partes hostes iter faciant.
(‘Caesar did likewise, sending forward the whole of his cavalry . . . to observe in which
direction the enemy were marching.’ Caes. Gal. 1.15.1)
(e) . . . senatus a Bestia consultus est placeretne legatos Iugurthae recipi moeni-
bus iique decrevere . . .
(‘ . . . the senate was consulted by Bestia as to whether they would consent to
receive Jugurtha’s envoys within the walls. The members thereupon decreed . . . ’
Sal. Jug. 28.2)
28
For more instances of quisquam, see Lindsay (1907: 5).
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
Notional agreement in gender of the verb with a plural collective noun across the
border of clauses is occasionally found, as in (g).
(g) . . . diversae duae legiones undecima et octava profligatis Viromanduis qui-
buscum erant congressi ex loco superiore in ipsis fluminis ripis proeliabantur.
(‘ . . . two detached legions, the Eleventh and the Eighth, having broken the Viroman-
dui with whom they had engaged, left the higher ground, and continued the fight on
the very banks of the river.’ Caes. Gal. 2.23.2)
Supplement:
Servitia urbem ut incenderent distantibus locis coniurarunt populoque ad opem passim
ferendam tectis intento ut arcem Capitoliumque armati occuparent. (Liv. 4.45.1); . . .
vitam ipsam in extremum adductam a clientelis et servitiis Octaviae, quae plebis sibi
nomen indiderint, ea in pace ausi quae vix bello evenirent. (Tac. Ann. 14.61.2);
(ii) Notional agreement of the verb with names referring to countries, cities, etc.
across the border of clauses is rare.
(h) Alexander devicta perdomitaque Thracia petens Asiam veritus ne post ipsius
discessum sumerent arma reges eorum secum velut honoris causa traxit . . .
(‘When Alexander had conquered and subdued Thrace and was setting out for Asia,
fearing that after his departure the Thracians would take up arms, he took with him,
as though by way of conferring honour, their kings and officials . . . ’ Fron. Str. 2.11.3)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(iii) Notional agreement of the verb with singular pronouns (and the like) across the
border of clauses can be seen in examples like (i)–(k).
(i) Nam numquam quisquam meorum maiorum fuit / quin parasitando pave-
rint ventris suos.
(‘For never was there one of my ancestors who didn’t feed his belly by being a
hanger-on.’ Pl. Per. 55–6)
(j) Hic cum uterque me intueretur seseque ad audiendum significarent paratos . . .
(‘Upon this they both looked at me and signified their readiness to hear me . . . ’ Cic.
Fin. 2.1)
(k) Trebellium dum uterque dedignatur, supra tulere.
(‘While each of them disdained Trebellius, they allowed him to surpass them both.’
Tac. Ann. 14.46.2)
Supplement (in alphabetical order by pronoun):
. . . cum eorum nemo, qui ad lacum Averni se adissent, aut ipsi venirent aut nuntium
litterasve mitterent . . . (Liv. 24.20.14); Adeo quidem artatum angustiis temporum ut
nemo memoria dignus alter ab altero videri nequiverint (nequiverit cj. Aldus). (Vell.
1.16.5); Ut quis ex longinquo revenerat, miracula narrabant . . . (Tac. Ann. 2.24.4);
Quotus enim quisque est qui hanc in re publica sectam sequatur . . . cum illam viam
sibi videant expeditiorem ad honores et ad omnia quae concupiverunt? (Cic. Flac.
104); Exite, agite exite, ignavi, male habiti et male conciliati, / quorum numquam
quicquam quoiquam venit in mentem ut recte faciant . . . (Pl. Ps. 133–4); Hoccine’st
credibile aut memorabile / tanta vecordia innata quoiquam ut siet / ut malis
gaudeant . . . (Ter. An. 625–7); Quorum utrique mortem est minitatus, nisi sibi
hortorum possessione cessissent . . . (Cic. Mil. 75);
Anaphoric(ally used) pronouns and relative pronouns often show notional agreement
with a constituent in a preceding clause or sentence. Ex. (a) shows the plural
masculine anaphoric pronoun eorum referring to the singular feminine collective
noun aetas in the main clause. In (b), masc. pl. eorum refers to the inhabitants of
Liburnia (fem. sg.). In (c), pl. masc. is refers to the people implied by the masc. sg.
indefinite pronoun quisquam. In (d), there is notional agreement between plural
earum and the ‘women’ inherent in the adjective muliebri in the noun phrase de
ingenio muliebri. Something similar is the case in (e), where illum refers to the person
Precius, suggested by the adjective Preciana.
(a) Maxime autem haec aetas a libidinibus arcenda est . . . ut eorum . . . vigeat
industria.
(‘And this time of life is above all to be protected against sensuality . . . so that the
diligence of men of this age may be strong . . . ’ Cic. Off. 1.122)
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
(b) . . . cum in Liburniam venisses, te vidisse matres familias eorum adferre ligna . . .
(‘ . . . that when you were in Liburnia, you saw mothers carrying logs . . . ’ Var. R. 2.10.8)
(c) Si quisquam’st qui placere se studeat bonis / . . . / in is poeta hic nomen
profitetur suom.
(‘If there’re any who are eager to please as many worthy citizens as possible . . . the
playwright wishes to enrol in their number.’ Ter. Eu. 1–3)
(d) Profecto ut ne quoquam de ingenio degrediatur muliebri / earumque artem
et disciplinam optineat colere.
(‘Under no circumstances is she to depart from her womanly ways, and she is to
continue to practise their art and discipline.’ Pl . Mil. 186–7)
(e) De hereditate Preciana, quae quidem mihi magno dolori est (valde enim
illum amavi)—sed hoc velim cures . . .
(‘About the Precius bequest (I am really sorry to get it—I was very fond of him),
would you please see that . . . ’ Cic. Fam. 14.5.2)
Supplement:
Collective nouns: Eodem impetu cohortes sinistrum cornu circumierunt eosque a
tergo sunt adorti. (Caes. Civ. 3.93.8); . . . subito ex oppido erupit multitudo atque
equitatus subsidio uno tempore eis casu succurrit . . . (B. Afr. 6.1); Sed iuventutem,
quam, ut supra diximus, inlexerat, multis modis mala facinora edocebat. Ex illis testis
signatoresque falsos commodare. (Sal. Cat. 16.1–2); . . . Hannibal, quia fessum militem
proeliis operibusque habebat, paucorum iis dierum quietem dedit . . . (Liv. 21.11.3);
NB: Plural: Germanica vexilla diu nutavere invalidis adhuc corporibus et placatis
animis, quod eos a Nerone Alexandriam praemissos atque inde rursus longa naviga-
tione aegros impensiore cura Galba refovebat. (Tac. Hist. 1.23.3); NB: Porro eunu-
chum dixti velle te, / quia solae utuntur is reginae. (Ter. Eu. 167–8);
Geographical names: Athenis aiunt cum quidam apud eos qui sancte graviterque
vixisset . . . (Cic. Balb. 12); Postquam inde expulsus Thebas venerit, adeo studiis
eorum inservisse ut . . . (Nep. Alc. 11.3); Ad eam oppugnandam novos exercitus scribi
Romae, nec ullius urbis defectioni magis infensos eorum animos esse. (Liv. 24.12.2);
Adjectives: Nos postquam populi rumore[m] intelleximus / studiose expetere vos
Plautinas fabulas, / antiquam eius edimus comoediam . . . (Pl. Cas. 11-3); Ad hi-
rundininum nidum visa est simia / ascensionem ut faceret admolirier. / . . . — /
neque eas eripere quibat inde. (Pl. Rud. 598–600); . . . Carneadea nobis adhibenda
divisio est, qua noster Antiochus libenter uti solet. Ille igitur vidit non modo quot .
. . (Cic. Fin. 5.16); Adice obsonatores, quibus dominici palati notitia subtilis est, qui
sciunt, cuius illum rei sapor excitet . . . (Sen. Ep. 47.8);
(f ) Sed quisnam est iste tam demens? De tuisne—tametsi qui magis sunt tui
quam quibus tu salutem insperantibus reddidisti?—an ex eo numero qui una
tecum fuerunt?
(‘But who is that lunatic of yours? One of your friends?—and yet who are more
truly your friends than those to whom you have restored security though they
were without hope?—or is it one of those who were once on your side?’ Cic.
Marc. 21)
(g) . . . factum (sc. periculum) etiam nuper in Italia servili tumultu, quos tamen
aliquid usus ac disciplina, quam a nobis accepissent, sublevarent.
(‘We have made further trial of late in Italy in the revolt of the slaves, whom,
however, the practice and training they had received from us supported to some
degree.’ Caes. Gal. 1.40.5)
(h) Pro Cyrenenses populares! Vostram ego imploro fidem, / agricolae, accolae
propinqui qui estis his regionibus . . .
(‘Citizens of Cyrene! I implore your protection, you who are farmers and neighbours
living close to these regions . . . ’ Pl. Rud. 615–16)
(i) Egon’ ab lenone quicquam / mancipio accipiam, quibus sui nihil est nisi una
lingua / qui abiurant si quid creditum est?
(‘Am I to take anything legally from a pimp, those people who have nothing of their
own except for the bare tongue with which they swear off if anything’s been entrusted
to them?’ Pl. Cur. 494–6)
Supplement:
. . . estque ex eo numero qui semper apud omnis sancti sunt habiti itaque dicti . . .
(Cic. Arch. 31); Igitur unus ex eo numero qui ad caedem parati erant paulo
inconsultius Massivam adgreditur. (Sal. Jug. 35.6); . . . Phidippumque cursorem
eius generis qui hemerodromoe vocantur Lacedaemonem miserunt . . . (Nep.
Milt. 4.3);
Syracusanam quidem civitatem ut abs te adfecta est, ita in te esse animatam
videmus, apud quos etiam Verria illa flagitiosa sublata sunt. (Cic. Ver. 4.151); Veiens
bellum exortum, quibus Sabini arma coniunxerant. (Liv. 2.53.1);
Contra caprile (sc. semen) mobilius esse, de quarum velocitate in Originum libro
Cato scribit haec. (Var. R. 2.3.3);
. . . cum tu discessu ceterorum nostra tamen qui remansissemus caede contentum
te esse dicebas? (Cic. Catil. 1.7); . . . uti corpora nostra ab iniuria tuta forent, qui . . .
plerique patriae, sed omnes fama atque fortunis expertes sumus. (Sal. Cat. 33.1);
Quid ad hanc (= horum) clementiam addi potest, qui ne hoc quidem permiserint
ut . . . (Cic. Tul. 55);
Hic si Peripateticus fuisset, permansisset, credo, in sententia, qui dolorem malum
dicunt esse . . . (Cic. Fin. 5.94); L. Cantilius scriba pontificis, quos nunc minores
pontifices appellant . . . (Liv. 22.57.3); His informatum manibus iam parte polita /
fulmen erat, toto genitor quae plurima caelo / deicit in terras, pars inperfecta
manebat. (Verg. A. 8.426–8);
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
Deviations from regular agreement occur especially in the domain of the abstract
noun res ‘thing’ and the neuter pronouns (quod and quae). Illustrations from Early
Latin are (j) and (k).29 A remarkable instance of variation is (l).
(‘The sum which I paid for estates in Italy was about six hundred million
sesterces, and the amount which I paid for lands in the provinces was about
two hundred and sixty million.’ Aug. Anc. 16.1)
To be distinguished from the instances of non-agreement above are instances found
in later texts and inscriptions, which are either errors or early testimonies of the
changes in the paradigm of relative and interrogative pronouns that took place in
later Latin and in Early Romance.30 An early example is (m).
Poets and occasionally also prose writers sometimes assign to geographical names the
gender of hyperonyms like mons ‘mountain’, oppidum ‘town’, and urbs ‘city’, with
which they often function as appositives. For modifiers this means that they show
deviant agreement. An example is (a). Neuter Pelion is modified by masculine altus,
which takes its gender from that of mons.31 Another form of deviant agreement is
shown in (b), where eas can be understood if the synonym nubes is inferred from
nubila. Also remarkable is (c).
29
References, also to later instances, can be found in Sz.: 431–2. The instances of ‘concordantia
anomala’ in Pl. are collected by Lodge s.v. qui 460A/B, § 3. See also Bailey ad Lucr. 3.185.
30
See Väänänen (1981: 125); Schafroth (1993: 67–72).
31
For examples, see K.-H.: 271–2.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Agreement
32 33
For further examples, see K.-St.: 26–7. For more details, see OLD s.v. partim § 1.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
(a) (sc. Isocrates) cuius e ludo tamquam ex equo Troiano meri principes exie-
runt. Sed eorum partim in pompa, partim in acie inlustres esse voluerunt.
(‘Isocrates . . . from whose school, as from the Horse of Troy, none but leaders
emerged, but some of them sought glory in ceremonial, others in action.’ Cic. de
Orat. 2.94)
Something similar is the case with alter alter ‘the one the other’, as in (b), where quis
‘someone’ is interpreted as plural by the addition of alter ab altero.
(b) Atque ego a maioribus memoria sic accepi, si quis quid alter ab altero
peterent . . . illi, unde petitur, ei potius credendum esse.
(‘And I have learnt this from the tradition of our ancestors: if anyone claim anything
from another . . . the one from whom the claim is made ought rather to be credited.’
Cato orat. 206)
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
BIBLIOGRAPHY
A. Abbreviated references
Bibliography
Bibliography
B. Linguistic studies
Aalto, Pentti (1949). Untersuchungen über das lateinische Gerundium und Gerundivum.
Helsinki: Academia Fennica.
Abel, Fritz (1971). L’adjectif démonstratif dans la langue de la Bible. Tübingen: Niemeyer.
Acosta, Diego de (2011). ‘Rethinking the genesis of the Romance periphrastic perfect’,
Diachronica 28: 143–85.
Adam, Eugene (2002). ‘Review of Oldsjö 2001’, Bryn Mawr Classical Review 2002.06.36.
Adamik, Béla (2015). ‘Periodization of Latin: An old question revisited’, in Gerd Haverling
(ed.), 638–50.
Adams, James N. (1971). ‘A type of hyperbaton in Latin prose’, Proceedings of the Cambridge
Philological Society 17: 1–16.
Adams, James N. (1973a). ‘The substantival present participle in Latin’, Glotta 51: 116–36.
Adams, James N. (1973b). ‘The vocabulary of the speeches in Tacitus’ historical works’,
Bulletin of the Institute of Classical Studies 20: 124–44.
Adams, James N. (1976). The Text and the Language of a Vulgar Latin Chronicle (Anonymus
Valesianus II). London: Bulletin of the Institute of Classical Studies.
Adams, James N. (1977). The Vulgar Latin of the Letters of Claudius Terentianus. Manchester:
Manchester University Press.
Adams, James N. (1984). ‘Female speech in Latin comedy’, Antichthon 18: 43–77.
Adams, James N. (1990). ‘The latinity of C. Novius Eunus’, Zeitschrift für Papyrologie und
Epigraphik 82: 227–47.
Adams, James N. (1991). ‘Some neglected evidence for Latin habeo with infinitive: The order of
the constituents’, Transactions of the Philological Society 89: 131–96.
Adams, James N. (1994a). ‘Wackernagel’s law and the position of unstressed personal pro-
nouns in classical Latin’, Transactions of the Philological Society 92: 103–78.
Adams, James N. (1994b). Wackernagel’s Law and the Placement of the Copula esse in Classical
Latin, Cambridge Philological Society, Suppl. Vol. 18. Cambridge: Cambridge Philological
Society.
Adams, James N. (1995a). ‘The language of the Vindolanda Tablets: An interim report’,
Journal of Roman Studies 85: 86–134.
Adams, James N. (1995b). Pelagonius and Latin Veterinary Terminology in the Roman Empire.
Leiden: Brill.
Adams, James N. (1996). ‘Interpuncts as evidence for the enclitic character of personal
pronouns in Latin’, Zeitschrift für Papyrologie and Epigraphik 111: 209–10.
Adams, James N. (1999). ‘Nominative personal pronouns and some patterns of speech
in republican and Augustan poetry’, in James N. Adams and Roland G. Mayer (eds),
97–133.
Adams, James N. (2000). ‘The poets of Bu Njem: Language, culture and the centurionate’,
Journal of Roman Studies 89: 109–34.
Adams, James N. (2003a). ‘Petronius and new non-literary Latin’, in József Herman and
Hannah Rosén (eds): 11–23.
Adams, James N. (2003b). Bilingualism and the Latin Language. Cambridge: Cambridge
University Press.
Adams, James N. (2005a). ‘Neglected evidence for female speech’, Classical Quarterly 55: 582–97.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Adams, James N. (2005b). ‘The Bellum Africum’, in Tobias Reinhardt, Michael Lapidge, and
James N. Adams (eds), 73–96.
Adams, James N. (2006). ‘Greek interference in Egyptian Latin’, Oxford University Working
Papers in Linguistics, Philology & Phonetics 11: 1–4.
Adams, James N. (2007). The Regional Diversification of Latin, 200 BC–AD 600. Cambridge:
Cambridge University Press.
Adams, James N. (2013). Social Variation and the Latin Language. Cambridge: Cambridge
University Press.
Adams, James N. (forthc.). ‘Informal’ Latin: An Anthology of Fifty Texts with Commentary, 200
BC to AD 900. Cambridge: Cambridge University Press.
Adams, James N. and Mayer, Roland G. (eds) (1999). Aspects of the Language of Latin Poetry,
Proceedings of the British Academy 93. London: The British Academy.
Adema, Suzanne (2007). ‘Temporal bases and the use of the narrative infinitive in the Aeneid’,
in Gérald Purnelle and Joseph Denooz (eds), 7–18.
Adema, Suzanne (2008). ‘Discourse modes and bases: A study of the use of tenses in Vergil’s
Aeneid ’. Dissertation, VU University, Amsterdam.
Ágel, Vilmos, Eichinger, Ludwig M., Eroms, Hans Werner, Hellwig, Peter, Heringer, Hans
Jürgen, and Lobin, Henning (eds) (2006). Dependenz und Valenz: Ein internationales
Handbuch der zeitgenössischen Forschung. Berlin: De Gruyter
Agostiniani, Luciano (1983). Le ‘iscrizioni parlanti’ dell’ Italia antica. Firenze: Olschki.
Agud, Ana, Fernández Delgado, José A., and Ramos, Agustín (eds) (1996). Las lenguas de
corpus y sus problemas lingüísticos. Madrid: Ediciones Clásicas.
Ahlquist, Helge (1909). Studien zur spätlateinischen Mulomedicina Chironis. Uppsala: Berling.
Aili, Hans (1979). The Prose Rhythm of Sallust and Livy. Stockholm: Almqvist & Wicksell.
Alföldi, Geza (1986). Die römische Gesellschaft. Stuttgart: Steiner.
Allen W. Sidney (1973). Accent and Rhythm. Cambridge: Cambridge University Press.
Álvarez Huerta, Olga (1996). ‘The connecting relative in the oratio obliqua in Latin’, in
Hannah Rosén (ed.), 567–75.
Álvarez Huerta, Olga (2001). ‘Sur le subjonctif de citation en latin’, in Claude Moussy (ed.),
171–83.
Álvarez Huerta, Olga (2005). ‘¿Accusativus pendens en latín?’, in Gualtiero Calboli (ed.), 433–42.
Álvarez Huerta, Olga (2007). ‘El acusativo proléptico en latín’, in Gérald Purnelle and Joseph
Denooz (eds), 19–29.
Álvarez Huerta, Olga (2010a). ‘La expresión del temor en latín: El marco predicativo de timeo’,
Actas del XII Congreso Español de Estudios Clásicos, II.241–50.
Álvarez Huerta, Olga (2010b). ‘Acusativo y ablativo para la expresión del tiempo en latín’,
Revue de linguistique latine du centre Alfred Ernout: De Lingua Latina 4: 17–30.
Álvarez Huerta, Olga (2011). ‘Quelques réflexions sur l’emploi du subjonctif dans les relatives
latines’, Les Études Classiques 79: 105–22.
Álvarez Huerta, Olga (2014). ‘El acusativo exclamativo en Latin’, in Concepción Cabrillana
and Christian Lehmann (eds), 183−92.
Álvarez Rodríguez, Adelino (2001). El futuro de subjuntivo del latín al romance. Málaga:
Universidad de Málaga.
Amacker, René (1986). ‘Entre syntaxe et sémantique: L’ambiguité par alternat en latin’, Cahiers
Ferdinand de Saussure 40: 61–74.
Amenta, Luisa (2003). Perifrasi aspettuali in greco e latino. Milan: Franco Angeli.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
André, Aurélie (2010). ‘La concurrence entre is et ille dans l’évolution de la langue latine: Étude
comparative de Cicéron à saint Augustin’, Latomus 69: 313–29.
André, Aurélie and Fruyt, Michèle (2012). ‘Le rôle de is dans les changements de l’endophore
et de la deixis en latin’, in Frédérique Biville et al. (eds), 291–307.
Andrewes, M. (1937). ‘Caesar’s use of tense sequence in indirect speech’, Classical Review 51:
114–16.
Andrewes, M. (1951). ‘The function of tense variation in the subjunctive mood of oratio
obliqua’, Classical Review 1: 142–6.
Andrews, Avery D. (2007). ‘Relative clauses’, in Timothy Shopen (ed.), Language Typology and
Syntactic Description. Cambridge, Cambridge University Press, 206–36.
Anreiter, Peter and Kienpointner, Manfred (2010). Latin Linguistics Today, Innsbrucker
Beiträge zur Sprachwissenschaft 137. Innsbruck: Institut für Sprachen und Literaturen der
Universität Innsbruck.
Anscombre, Jean-Claude and Pierrot, Alain (1984). ‘Y a-t-il un critère de performativité en
latin?’, Linguisticae Investigationes 8: 1–19.
Anscombre, Jean-Claude and Pierrot, Alain (1985). ‘Sulla performatività in latino’, Paideia 40:
35–47.
Arcos Pereira, Trinidad, Fernández López, Jorge, and Moya del Baño, Francisca (eds) (2009).
‘Pectora mulcet’: Estudios de retórica y oratoria Latinas. Logroño: Instituto de Estudios
Riojanos.
Arias, Carmen (1996). ‘La sustantivación del adjetivo latino: El caso de los adjetivos derivados’,
in Hannah Rosén (ed.), 231–40.
Arias, Carmen (1999). ‘Innovaciones sintácticas en el latín cristiano: Participio de presente en lugar
de forma personal del verbo’, in Hubert Petersmann and Rudolf Kettemann (eds), 195–207.
Arias, Carmen (ed.) (2006). Latin vulgaire—Latin tardif VII. Sevilla: Universidad de Sevilla.
Arjoca-Ieremia, Eugenia (2008). ‘L’expression des modalités déontiques: Le cas de la conju-
gaison périphrastique passive’, in Roger Wright (ed.), 138–47.
Asensio, Paloma (2002). ‘Superare aliquem aliqua re. El comparativo de exceso en latín’, in
Empar Espinilla et al. (eds), 11–37.
Ashdowne, Richard (2002). ‘The vocative’s calling: The syntax of address in Latin’, Oxford
University Working Papers in Linguistics, Philology & Phonetics 7: 143–62.
Ashdowne, Richard (2008). ‘E-vocative invocation: On the historical morphosyntax of Latin
oaths’, in Roger Wright (ed.), 13–25.
Axelson, Bertil (1933). Senecastudien. Lund: Universitets Årsskrift.
Axelson, Bertil (1939). Neue Senecastudien. Lund: Universitets Årsskrift.
Axelson, Bertil (1945). Unpoetische Wörter. Lund: Gleerup.
Babcock, Clinton Leroy (1901). A Study in Case Rivalry, Being an Investigation Regarding the
Use of the Genitive and Accusative in Latin with Verbs of Remembering and Forgetting. New
York: Macmillan.
Bach, Joseph (1888). De usu pronominum demonstrativorum apud priscos scriptores Latinos
particula prima. Strassburg.
Baehrens, Wilhelm Adolf (1912). Beiträge zur lateinischen Syntax, Philologus Supplementband
12. Leipzig: Dieterich.
Bagordo, Andreas (2001). Beobachtungen zur Sprache des Terenz. Mit besonderer Berücksich-
tigung der umgangssprachlichen Elemente. Göttingen: Vandenhoeck & Ruprecht.
Bakkum, Gabriël C. L. M. (2009). The Latin Dialect of the Ager Faliscus, 2 vols. Amsterdam:
Amsterdam University Press.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Baldi, Philip (1977). ‘Morpho-syntax and the Latin genitive’, Folia Linguistica 11: 93–108.
Baldi, Philip (1983). ‘Speech perception and grammatical rules in Latin’, in Harm Pinkster
(ed.), 11–26.
Baldi, Philip (1989). ‘Negative prefixation in Latin’, L’information grammaticale 43: 3–6.
Baldi, Philip (1999). The Foundations of Latin. Berlin: De Gruyter.
Baldi, Philip and Cuzzolin, Pierluigi (eds) (2009). New Perspectives on Historical Latin Syntax,
I: Syntax of the Sentence. Berlin: De Gruyter.
Baldi, Philip and Cuzzolin, Pierluigi (eds) (2010a). New Perspectives on Historical Latin Syntax,
II: Constituent Syntax: Adverbial Phrases, Adverbs, Mood, Tense. Berlin: De Gruyter.
Baldi, Philip and Cuzzolin, Pierluigi (eds) (2010b). New Perspectives on Historical Latin Syntax,
III: Constituent Syntax: Quantification, Numerals, Possession, Anaphora. Berlin: De Gruyter.
Baldi, Philip and Cuzzolin, Pierluigi (eds) (2011). New Perspectives on Historical Latin Syntax,
IV: Complex Sentences, Grammaticalization, Typology. Berlin: De Gruyter.
Baldi, Philip and Nuti, Andrea (2010). ‘Possession’, in Philip Baldi and Pierluigi Cuzzolin
(eds), III.239–387.
Baloira, Manuel Adolfo (1981–3). ‘La lengua de Terencio: Genitivo y ablativo de cualidad’,
Estudios Clásicos 25: 103–20.
Bammesberger, Alfred and Heberlein, Friedrich (1996). Akten des VIII. internationalen Kollo-
quiums zur lateinischen Linguistik. Heidelberg: Winter.
Banniard, Michel (1995). ‘Ille et son système: Chronologie du développement’, in Louis
Callebat (ed.), 313–21.
Banniard, Michel (1997). Du latin aux langues romanes. Paris: Nathan.
Baños, José Miguel (1995). ‘El “doble dativo” en latín’, in Esperanza Torrego et al. (eds), 11–34.
Baños, José Miguel (1996a). ‘Sobre el infinitivo de futuro pasivo en latín (I): Consideraciones
generales’, in Ana María Aldama et al. (eds), De Roma al siglo XX. Madrid: SELat.-UNED,
Universidad de Extremadura, 31–41.
Baños, José Miguel (1996b). ‘Sobre el infinitivo de futuro pasivo en latín (II): Su desaparición
en latín tardío’, Cuadernos de Filología Clásica (Lat) 10: 9–20.
Baños, José Miguel (1996c). ‘Litteras Neroni / ad Neronem mittere: ¿Alternancia dativo / ad +
acc.?’, in Ana Agud et al. (eds), 217–35.
Baños, José Miguel (1997). ‘Sujeto, transitividad, agente y pasiva: La diátesis verbal en latín’,
Tempus 16: 61–90.
Baños, José Miguel (1998). ‘Sintaxis y semántica del dativo objeto indirecto en latín clásico’, in
Esperanza Torrego (ed.), 11–41.
Baños, José Miguel (1999). ‘La consecutio temporum y la expresión de la posterioridad en latín’,
in Empar Espinilla et al. (eds), 57–84.
Baños, José Miguel (2000). ‘Vulgarismos sintácticos en Plauto (II): Quae ad patrem vis nuntiari
(Cap. 360)’, in Benjamín García Hernández (ed.), 1–15.
Baños, José Miguel (2003). ‘Paenitet y los verbos impersonales en latín’, in José Miguel Baños
et al. (eds), 51–77.
Baños, José Miguel (2007). ‘Estructuras predicativas de los verbos de sentimiento en latín: la
complementación nominal de gaudeo y laetor’, in Esperanza Torrego et al. (eds), 11–37.
Baños, José Miguel (2012). ‘Verbos soporte e incorporación sintáctica en latín: El ejemplo de
ludos facere’, Revista de Estudios Latinos 12: 37–57.
Baños, José Miguel (2013). ‘Sobre la manera de ‘hacer la guerra’ en latín: bellum gero, belligero,
bello’, in José Antonio Beltrán et al. (eds), 27–39.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Baños, José Miguel (2014a). ‘Dos tipos de intransitividad en latín: sintaxis y semántica’, in Jesús
de la Villa et al. (eds), Ianua Classicorum: Temas y formas del mundo clásico, Actas del XIII
Congreso Español de Estudios Clásicos. Madrid: Ediciones Clásicas, I.759–90.
Baños, José Miguel (2014b). Las oraciones causales en latín: Su evolución diacrónica. Madrid:
Escolor y Mayo Editores.
Baños, José Miguel (coordinador) (2009). Sintaxis del latín clásico. Madrid: Liceus.
Baños, José Miguel et al. (eds) (2003). Praedicativa: Complementación en griego y latín.
Santiago: Servicio de publicaciones de la Universidad de Santiago de Compostela.
Baratin, Marc (1981). ‘Remarques sur l’emploi des temps et des modes dans le système
conditionnel latin’, Bulletin de la Société de Linguistique de Paris 76: 249–73.
Barbelenet, Daniel (1913). De l’aspect verbal en latin ancien, et particulièrement dans Térence.
Paris: Champion.
Barrios-Lech, Peter (2014). ‘Quid ais and female speech in Roman comedy’, Hermes 142: 480–6.
Bauer, Brigitte L. M. (2003). ‘The adverbial formation of mente in vulgar and late Latin:
A problem in grammaticalization’, in Heikki Solin et al. (eds), 439–58.
Bauer, Brigitte L. M. (2008). ‘Nominal apposition in Vulgar and Late Latin: At the crossroads
of major language changes’, in Roger Wright (ed.), 42–50.
Bauer, Brigitte L. M. (2009). ‘Strategies of definiteness in Latin’, in Vit Bubenik et al. (eds),
Grammatical Change in Indo-European Languages. Amsterdam: Benjamins, 71–87.
Becker, Maria (2005). ‘Syntaktische Varietas im Altlatein: Verbote bei Cato’, Glotta 81: 1–24.
Beckmann, Gustav (1963). Die Nachfolgekonstruktionen des instrumentalen Ablativs im Spät-
latein und im Französischen. Tübingen: Niemeyer.
Beikircher, Hugo (1969). Kommentar zur IV. Satire des A. Persius Flaccus. Vienna: Böhlhaus.
Bejarano, Virgilio (1983). ‘Las proposiciones completivas en la Peregrinatio Egeriae’, Helman-
tica 34: 91–101.
Bejarano, Virgilio (1994). ‘Las proposiciones completivas e causales en dieciocho sermones de
San Agustín’, Augustinus 39: 75–83.
Beltrán, José Antonio (1994). ‘Una nota sobre el infinitivo de narración en Quintiliano y
Prisciano’, Cuadernos de Filología Clásica 7: 41–56.
Beltrán, José Antonio (1996). El infinitivo de narración en latín. Zaragoza: Universidad de
Zaragoza.
Beltrán, José Antonio et al. (eds) (2013). Otium cum dignitate: Estudios en homenaje al profesor
José Javier Iso Echegoyen. Zaragoza: Universidad de Zaragoza.
Bendz, Gerhard (1941). ‘Sprachliche Bemerkungen zu Petron’, Eranos 39: 27–55.
Bendz, Gerhard (1948). ‘La construction du type “rebus in arduis” ’, Eranos 46: 42–53.
Beneš, Pavel (1971). Phrases à agents indéterminés dans le Nouveau Testament. Brno: Opera
Universitatis Purkynianae Brunensis.
Bennett, Charles E. (1910/14). Syntax of Early Latin, 2 vols. Boston: Allyn and Bacon (repr.
Hildesheim: Olms, 1966).
Benveniste, Émile (1960). ‘ “Être” et “avoir” dans leurs fonctions linguistiques’, Bulletin de la
Société Linguistique de Paris 55: 113–34.
Berdasco, Juan Carlos (2009). ‘Caracterización funcional del sintagma preposicional de +
ablativo en latín clásico’, in: Trinidad Arcos Pereira, Jorge Fernández López, and Francisca
Moya del Baño (eds), 171–80.
Bergh, Birger (1975). On Passive Imperatives in Latin. Stockholm: Almqvist & Wiksell.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bergh, Birger (1990). ‘De imperativis Latinis passivis’, in Sven-Tage Theodorsson (ed.), Greek
and Latin Studies in Memory of Cajus Fabricius. Göteborg: Acta Universitatis Gothobur-
gensis, 165–70.
Bertocchi, Alessandra (1994). ‘Occurrences of is as grammaticalizations of the speaker’, Papers
on Grammar 4: 1–27.
Bertocchi, Alessandra (1996a). ‘Some properties of ipse’, in Hannah Rosén (ed.), 539–52.
Bertocchi, Alessandra (1996b). ‘Some semantic and pragmatic properties of paene’, in Alfred
Bammesberger and Friedrich Heberlein (eds), 456–72.
Bertocchi, Alessandra (1999). ‘Ipse, un intensifieur’, in Alessandra Bertocchi, Mirka Maraldi,
and Anna Orlandini, ‘L’argumentation en latin’, Lalies 19: 153–62.
Bertocchi, Alessandra (2000). ‘Ipse as an intensifier’, Papers on Grammar 5: 15–30.
Bertocchi, Alessandra (2002). ‘Scalarity and concession: The case of quamvis’, in Lea Sawicki
and Donna Shalev (eds), 39–50.
Bertocchi, Alessandra (2005). ‘Indefinite pronouns in conditional clauses’, Papers on Grammar
9.1: 453–66.
Bertocchi, Alessandra and Maraldi, Mirka (1990). ‘Nominalization and possessives in Latin’,
Papers on Grammar 3, 69–123.
Bertocchi, Alessandra and Maraldi, Mirka (2005). ‘Indefinite pronouns in conditional clauses’,
Papers on Grammar 9.1: 453–66.
Bertocchi, Alessandra and Maraldi, Mirka (2006). ‘Menaechmus quidam: Indefinites and
proper nouns in Classical and Late Latin’, in Carmen Arias (ed.), 88–107.
Bertocchi, Alessandra and Maraldi, Mirka (2008). ‘Universal quantifiers in Latin: Some
remarks on ambiguity’, in Ghislaine Viré (ed.), 9–21.
Bertocchi, Alessandra and Maraldi, Mirka (2010). ‘Mid-scalar quantifiers in Latin’, in Peter
Anreiter and Manfred Kienpointner (eds), 193–206.
Bertocchi, Alessandra and Maraldi, Mirka (2011a). ‘Les adverbes de répétition rursus, iterum,
denuo’, in Claude Moussy (ed.), Espace et temps en latin. Paris: Presses de l’Université de
Paris-Sorbonne, 61–75.
Bertocchi, Alessandra and Maraldi, Mirka (2011b). ‘Conditionals and concessives’, in Philip
Baldi and Pierluigi Cuzzolin (eds), IV.93–193.
Bertocchi, Alessandra, Maraldi, Mirka, and Orlandini, Anna (2010). ‘Quantification’, in Philip
Baldi and Pierluigi Cuzzolin (eds), III.19–173.
Bertocchi, Alessandra and Orlandini, Anna (1994). ‘La quantificazione distributiva in latino:
singuli e i numerali distributivi’, in Gianluigi Borgato (ed.), Teoria del linguaggio e analisi
linguistica. Padova: Unipress, 47–62.
Bertocchi, Alessandra and Orlandini, Anna (1996). ‘Quelques aspects de la comparaison en
latin’, Indogermanische Forschungen 101: 195–232.
Bertocchi, Alessandra and Orlandini, Anna (2002). ‘Impossible n’est pas latin’, in Michèle
Fruyt and Claude Moussy (eds), Les modalités en latin’. Paris: Presses de l’Université de
Paris-Sorbonne, 9–23.
Bertrand-Dagenbach, Cécile (1995). ‘L’attraction modale en latin’, in Dominique Longrée
(ed.), 19–30.
Bisang, Walter (2011). ‘Word Classes’, in Jae Jung Song (ed.), The Oxford Handbook of
Linguistic Typology. Oxford: Oxford University Press, 280–302.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bischoff, Bernhard (1981). ‘Eine Beschreibung der Basilika von Saint-Denis aus dem Jahre
799’, Kunstchronik 34: 97–103.
Bitter, Norbert (1976). Kampfschilderungen bei Ammian. Bonn: Habelt.
Biville, Frédérique (1998). ‘Le statut linguistique des noms propres en latin’, in Benjamín
García Hernández (ed.), 825–39.
Biville, Frédérique (2008). ‘Entre substantif et adjectif: La “graduabilité” du nom propre
anthroponymique’, in Ghislaine Viré (ed.), 22–36.
Biville, Frédérique, Lhommé, Marie-Karine, and Vallat, Daniel (eds) (2012). Latin vulgaire—
Latin tardif IX. Lyon: Maison de l’Orient et de la Méditerranée.
Blaise, Albert (1955). Manuel du latin chrétien. Strasbourg: Le Latin Chrétien.
Blake, Barry J. (2001). Case. Cambridge: Cambridge University Press.
Blasco Ferrer, Eduardo (1986). La lingua sarda contemporanea. Cagliari: Edizioni della Torre.
Blase, Heinrich (1894). Geschichte des Plusquamperfekts im Lateinischen. Gießen: Ricker.
Blase, Heinrich (1896a). ‘Der Konjunktiv des Präsens im Bedingungssatze’, Archiv für Latei-
nische Lexikographie und Grammatik 9: 17–45.
Blase, Heinrich (1896b). ‘Amabo’, Archiv für Lateinische Lexikographie und Grammatik 9: 48–91.
Blase, Heinrich (1898). ‘Zur Geschichte der Futura und des Konjunktivs des Perfekts im
Lateinischen’, Archiv für Lateinische Lexikographie und Grammatik 10: 313–43.
Blase, Heinrich (1903). ‘Tempora und modi’, in Gustav Landgraf (ed.), Historische Grammatik
der lateinischen Sprache. Leipzig: Teubner, III.97–312.
Blatt, Franz (1952). Précis de syntaxe latine. Lyon: IAC.
Blomgrén, Sven (1937). ‘De sermone Ammiani Marcellini quaestiones variae’. Dissertation,
Uppsala University.
Blümel, Wolfgang (1979). ‘Zur historischen Morphosyntax der Verbalabstrakte im Latei-
nischen’, Glotta 57: 77–125.
Bodelot, Colette (1987). L’interrogation indirecte en latin. Paris: Société pour l’information
grammaticale.
Bodelot, Colette (1990). Termes introducteurs et modes dans l’interrogation indirecte en latin de
Plaute à Juvenal. Paris: Bibliothèque de Vita Latina.
Bodelot, Colette (1995). ‘Complétives appositives en latin: Forme, sens, syntaxe’, in Dominique
Longrée (ed.), 45–58.
Bodelot, Colette (1996). ‘Is, hic, ille, iste coréférentiels d’une proposition complétive: Étude
distributionnelle’, in Hannah Rosén (ed.), 525–38.
Bodelot, Colette (1998). ‘Tempto si, subsisto si . . . ou les propositions hypothétiques à sens final
chez les historiens latins’, in Bruno Bureau and Christian Nicolas (eds), 91–100.
Bodelot, Colette (1999). ‘L’indicatif dans l’interrogation indirecte en latin: Auteurs anciens vs.
auteurs tardifs’, in Hubert Petersmann and Rudolf Kettemann (eds), 213–22.
Bodelot, Colette (2002). ‘Réflexions sur la modalité à propos de ut “final” ’, in Michèle Fruyt
and Claude Moussy (eds), 249–63.
Bodelot, Colette (2003). ‘L’interrogation indirecte’, in Colette Bodelot (ed.), 193–333.
Bodelot, Colette (2007). ‘Syntaxe liée vs. syntaxe non liée ou de l’utilité de distinguer une
macro-syntaxe de la (micro-)syntaxe latine’, in Colette Bodelot (ed.), 15–33.
Bodelot, Colette (2010a). ‘Quam: Marqueur de degré interrogatif et/ou exclamatif ’, in Michèle
Fruyt and Olga Spevak (eds), 335–51.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Bolkestein, A. Machtelt (2001a). ‘Possessors and experiencers in Latin’, in Irène Baron, Michael
Herslund, and Finn Sørensen (eds), Dimensions of Possession. Amsterdam: Benjamins, 269–84.
Bolkestein, A. Machtelt (2001b). ‘Random scrambling? Constraints on discontinuity in Latin
noun phrases’, in Claude Moussy (ed.), 245–58.
Bolkestein, A. Machtelt (2002). ‘Anaphoric ablative absolutes in Caesar’, in Lea Sawicki and
Donna Shalev, (eds), 51–63.
Bolkestein, A. Machtelt, Kroon, Caroline H. M., Pinkster, Harm, Remmelink, H. Wim, and
Risselada, Rodie (eds) (2002). Theory and Description in Latin Linguistics. Amsterdam: Gieben.
Bonnet, Max (1890). Le latin de Grégoire de Tours. Paris: Hachette.
Boon, Pieter (1978). ‘Obliquitätskonjunktiv oder Konjunktiv der indirekten Rede?’, Indoger-
manische Forschungen 83: 324–44.
Bork, Hans Dieter (1990). ‘Der Mythos vom Schwächetod des lateinischen Futurs’, in Ewald
Könsgen (ed.), Arbor amoena comis: 25 Jähre Mittellateinisches Seminar in Bonn. Stuttgart:
Steiner Verlag, 15–22.
Bortolussi, Bernard (2002). ‘Ne velit et l’expression de la défense en latin’, in Michèle Fruyt and
Claude Moussy (eds), 121–30.
Bortolussi, Bernard (2005). ‘Subordination seconde du relatif: Contraintes de l’emploi’, in
Gualtiero Calboli (ed.), 479–92.
Bortolussi, Bernard (2012). ‘Évolution des indéfinis existentiels et de quisque en latin tardif ’, in
Frédérique Biville et al. (eds), 309–22.
Bortolussi, Bernard and Sznajder, Lyliane (2008). ‘Is vs zero-anaphora in predicative esse
sentences’, Papers on Grammar 10: 39–56.
Bourgery, Abel (1922). Sénèque prosateur: Études littéraires et grammaticales sur la prose de
Sénèque le philosophe. Paris: Les Belles Lettres.
Bourova, Viara (2005). ‘À la recherche du “conditionnel latin”: Les constructions infinitif +
forme de habere examinées à partir d’un corpus électronique’, in Claus D. Pusch, Johannes
Kabatek, and Wolfgang Raible (eds), Romanistische Korpuslinguistik II, Korpora und dia-
chrone Sprachwissenschaft. Tübingen: Narr, 303–16.
Bourova, Viara (2007). ‘Les constructions latines infinitif + habebam vs. infinitif + habui et le
développement du conditionnel roman’, in David Trotter (ed.), Actes du XXIVe Congrès
International de Philologie et Linguistique Romanes. Tübingen: Niemeyer, II.461–74.
Bourova, Viara (2008). ‘Les participes futurs en -urus / -undus combinés avec esse au passé en
latin tardif: Un conditionnel non abouti?’, in Roger Wright (ed.), 271–80.
Bourova, Viara and Tasmowski, Liliane (2007). ‘La préhistoire des futurs romans—ordre des
constituants et sémantique’, Cahiers Chronos 19: 25–41.
Bräunlich, Alice Freda (1920). ‘The indicative indirect question in Latin’. Dissertation, Uni-
versity of Chicago.
Brea, Mercedes (1985). ‘Las preposiciones, del latín a las lenguas románicas’, Verba 12: 147–82.
Brenous, Joseph (1895). Étude sur les hellénismes dans la syntaxe latine. Paris: Klincksieck.
Brightbill, Jeremy and Kosch, Branden (2012). ‘Case and Preposition Usage in Select Latin
Texts’. Ms., University of Chicago.
Brink, Charles Oscar (1969). ‘Horatian notes: Despised readings in the manuscripts of the
Odes’, Proceedings of the Cambridge Philological Society 15: 1–6.
Briscoe, John (2010). ‘The fragments of Cato’s Origines’, in Eleanor Dickey and Anna Chahoud
(eds), 154–60.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Brugmann, Karl (1925). Die Syntax des einfachen Satzes im Indogermanischen. Berlin: De
Gruyter.
Brunet, Claude (2008). ‘Accipere, verbe support et élément du passif nominal chez Sénèque et
Valère-Maxime’, in Ghislaine Viré (ed.), 167–79.
Bruyne, André de (1961). ‘Reflexief en pseudo-reflexief aspect van het gebruik der pronomina
“sui, sibi, se, en suus” ’, L’Antiquité Classique 30: 45–66.
Bulhart, Vinzenz (1955). ‘Ausdrucksweisen für das prädikative Verhältnis im Lateinischen’,
Wiener Studien 68: 47–64.
Bulhart, Vinzenz (1957). ‘De Sermone Tertulliani’, in Tertulliani Opera Pars Quarta, CSEL 76.
Vienna: Hoelder-Pichler-Tempsky, ix–xlviii.
Bulhart, Vinzenz (1965). ‘Futurus venire’, Wiener Studien 78: 180–1.
Bureau, Bruno and Nicolas, Christian (eds) (1998). Moussyllanea. Louvain-Paris: Peeters.
Burkard, Thorsten (2006). ‘Zum Modus im lateinischen Nebensatz’, Glotta 82: 1–24.
Burkard, Thorsten and Schauer, Markus (2000). Lehrbuch der lateinischen Syntax und Seman-
tik. Darmstadt: Wissenschaftliche Buchgesellschaft.
Burkhardt, Friedrich (1971). ‘Zur doppelten Enallage’, Gymnasium 78, 407–21.
Bybee, Joan L., Perkins, Revere, and Pagliucci, William (1994). The Evolution of Grammar:
Tense, Aspect and Modality and the Languages of the World. Chicago: University of Chicago
Press.
Cabrillana, Concepción (1996). ‘El vocativo latino: Una revisión’, Tempus 13: 5–40.
Cabrillana, Concepción (1997a). ‘Expresión casual de complementos de venio y eo’, Habis 28:
325–36.
Cabrillana, Concepción (1997b). ‘Complementos direccionales en el marco predicativo de
venio’, Minerva 11: 117–37.
Cabrillana, Concepción (1998). ‘Structures of identification and attribution with sum’, in
Benjamín García Hernández (ed.), 217–33.
Cabrillana, Concepción (1999). ‘La consecutio temporum en las oraciones consecutivas latinas’,
in Empar Espinilla et al. (eds), 91–113.
Cabrillana, Concepción (2002). ‘Semantic and syntactic characterization of the “Genitive +
esse” construction’, Papers on Grammar 8: 43–68.
Cabrillana, Concepción (2003). ‘Estudio de rasgos diferenciales en las estructuras de Genitivo y
Dativo “posesivos” en latín’, in José Miguel Baños et al. (eds), 79–109.
Cabrillana, Concepción (2004). ‘Forma lingüística y tipo de personaje en la comedia terenci-
ana’, in Antonio López Eire and Agustín Ramos (eds), 11–31.
Cabrillana, Concepción (2006). ‘Factores condicionantes en la expresión de la “posesión” en
latín clásico y tardío’, in Carmen Arias (ed.), 137–52.
Cabrillana, Concepción (2007). ‘La elipsis verbal: Un rasgo differenciador de sum?, in Espe-
ranza Torrego et al. (eds), 39–67.
Cabrillana, Concepción (2008). Vocativo y participantes en el acto de habla en la comedia
latina y la novela de Petronio’, in Roger Wright (ed.), 69–77.
Cabrillana, Concepción (2010a). Consideración sintáctico-semántica de esse: Un estudio a
través de Livio. Santiago de Compostela: Universidade de Santiago de Compostela.
Cabrillana, Concepción (2010b). ‘Praedicativum and subject complement’, in Peter Anreiter
and Manfred Kienpointner (eds), 233–45.
Cabrillana, Concepción (2011). ‘Purpose and result clauses’, in Philip Baldi and Pierluigi
Cuzzolin (eds), IV.19–92.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Carney, T. F. (1964). ‘The words sodes and quaeso in Terentian usage’, Acta Classica 7: 57–63.
Carvalho, Paulo de (1991a). ‘Sur les “impersonnels météorologiques” du latin’, in Michel
Maillard (ed.), L’impersonnel: Mécanismes linguistiques et fonctionnements littéraires.
Actes du Colloque de Grenoble (1990). Grenoble: CÉDITEL, 36–45.
Carvalho, Paulo de (1991b). ‘Structure morpho-sémantique de l’adjectif latin ou: La présence
de l’absent’, in Robert G. Coleman (ed.), 251–67.
Carvalho, Paulo de (1992). ‘La grammaire des temps en latin’, Orphea Voce 4: 101–67.
Carvalho, Paulo de (1996). ‘Cas et préposition en linguistique latine et en linguistique
théorique’, in: Rosén, Hannah (ed.), 241–58.
Castillo, Marina del (2010). ‘Las formas de futuro perfecto y de perfecto subjuntivo en la
versificación latina: ¿Una cuestión métrica o una cuestión morfológica?’, Cuadernos de
Filología Clásica (Lat) 30: 249–61.
Cennamo, Michela (1999). ‘Late Latin pleonastic reflexives and the unaccusative hypothesis’,
Transactions of the Philological Society 97: 103–50.
Cennamo, Michela (2001). ‘Classi verbali e cambiamento sintattico: La interpretazione passiva
del construtto riflessivo’, Atti del XXII Congresso internazionale di linguistica italiana
(Budapest 1998). Rome: Bulzoni, 225–42.
Cennamo, Michela (2005). ‘Passive auxiliaries in Late Latin’, in Sándor Kiss et al. (eds),
177–94.
Cerra, Gabriela (2009). Linguistic Questions in Cicero’s poetic Translations. Bahia Blanca:
Editorial de la Universidad Nacional del Sur.
Chafe, Wallace (1994). Discourse, Consciousness, and Time. Chicago: University of Chicago
Press.
Champion, James Joseph (1978). The Periphrastic Futures Formed by the Romance Reflexes of
vado (ad) plus Infinitive. Chapel Hill: University of North Carolina Press.
Charney, Ben L. (1943). ‘Ellipsis of the verb in Seneca’s Epistulae Morales’, Classical Philology
38: 46–8.
Chausserie‐Laprée, J. R. (1969). L’expression narrative chez les historiens latins. Paris: De
Boccard.
Clackson, James (2004). ‘The word order pattern magna cum laude in Latin and Sabellian’, in
John Penney (ed.), 391–404.
Clackson, James (2011). ‘The social dialects of Latin’, in James Clackson (ed.), 505–26.
Clackson, James (ed.) (2011). A Companion to the Latin Language. Oxford: Blackwell.
Codoñer, Carmen (1968). ‘El sistema de los indefinidos latinos’, Emerita 36: 7–24.
Coleman, Robert G. (1975). ‘Greek influence on Latin syntax’, Transactions of the Philological
Society 74: 101–56.
Coleman, Robert G. (1985). ‘The Indo-European origins and Latin development of the
accusative with infinitive construction’, in Christian Touratier (ed.), 307–42.
Coleman, Robert G. (1991). ‘Latin prepositional syntax in Indo-European perspective’, in
Robert Coleman (ed.), 323–38.
Coleman, Robert G. (1999). ‘Poetic diction and poetic register’, in James N. Adams et al. (eds),
21–93.
Coleman, Robert G. (ed.) (1991). New Studies in Latin Linguistics. Amsterdam: Benjamins.
Comrie, Bernard (19892). Language Universals and Linguistic Typology. Oxford: Blackwell.
Conradie, C. Jac (1999). ‘Preterite loss in Early Afrikaans’, Folia Linguistica 23: 19–37.
Corbett, Greville (2006). Agreement. Cambridge, Cambridge University Press.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Cornell, Timothy J. (ed.) (2013). The Fragments of the Roman Historians, 3 vols. Oxford:
Oxford University Press.
Corrodi, Hans (1925). ‘Das Subjekt der sog. unpersönlichen Verben’, Zeitschrift für Ver-
gleichende Sprachforschung 53: 1–36.
Courtney, Edward (2004). ‘The Greek accusative’, Classical Journal 99: 425–31.
Cova, Pier Vincenzo (1986). ‘L’ablativo assoluto nella Naturalis Historia’, in Pier Vincenzo
Cova et al. (eds), 13–142.
Cova, Pier Vincenzo et al. (eds) (1986). Studi sulla lingua di Plinio il Vecchio. Milan: Università
Cattolica.
Crawford, Michael (1996). Roman Statutes, 2 vols. London: Institute of Classical Studies.
Crevels, Mily (2000). ‘Concession: A typological study’. Dissertation, University of Amsterdam.
Cugusi, Paolo (2009). ‘Una silloge recente di testi latino-volgari e la storia del latino (volgare e
non volgare)’, Rivista Italiana di Filologia Classica 137: 221–48.
Cugusi, Paolo and Sblendorio Cugusi, Maria Teresa (2008). Carmina Latina Epigraphica
Moesica, Carmina Latina Epigraphica Thraciae. Bologna: Pàtron.
Cuzzolin, Pierluigi (1991). ‘Sulle prime attestazioni del tipo sintattico dicere quod’, Archivio
Glottologico Italiano 76: 28–78.
Cuzzolin, Pierluigi (1994). Sull’origine della costruzione dicere quod: Aspetti sintattici e
semantici. Firenze: La Nuova Italia.
Cuzzolin, Pierluigi (2008). ‘Some remarks concerning the consecutive clause’, Papers on
Grammar 10: 73–90.
Cuzzolin, Pierluigi and Haverling, Gerd (2009). ‘Syntax, sociolinguistics, and literary genres’,
in Philip Baldi and Pierluigi Cuzzolin (eds), I.19–64.
Dahlén, Eric (1964). Études syntaxiques sur les pronoms réfléchis pléonastiques en latin.
Göteborg: Acta Universitatis Gothoburgensis.
Dahlén, Eric (1977). Remarques syntaxiques sur certains verbes pronominaux en latin et en
langues romanes. Göteborg: Acta Universitatis Gothoburgensis.
Dalès, Jacques (1976/7). ‘Voix, temps et aspects dans les périphrases médiopassives -us sum,
-us fui’, Revue des Études Anciennes 78–9: 129–44.
Daube, David (1956). Forms of Roman Legislation. Oxford: Oxford University Press.
Deckman, Alice Anna (1920). ‘A study of the impersonal passive of the ventum est type’.
Thesis, University of Pennsylvania, Philadelphia.
Delatte, Louis, Evrard, Étienne, Govaerts, Suzanne, and Denooz, Joseph (1980/1). Lucius
Annaeus Seneca: Opera philosophica, Index verborum: Listes de fréquence, relevés gramma-
ticaux, 2 vols. Hildesheim: Olms.
D’Elia, Mario (1976). ‘Sull’uso di quod con il senso di si nel latino giuridico’, in Scritti in onore
di Giuliano Bonfante. Brescia: Paideia, I.191–204.
Denizot, Camille and Dupraz, Emmanuel (eds) (2012). Anaphore et anaphoriques: Variété des
langues, variété des emplois. Mont-Saint-Aignan: Publications des Universités de Rouen et
du Havre.
Denooz, Joseph (1980). Lucius Annaeus Seneca: Tragoediae, Index verborum: Listes de
fréquence, relevés grammaticaux. Hildesheim: Olms.
Denooz, Joseph (1988). ‘Quelques observations sur la fréquence des substantifs et des préposi-
tions dans la littérature latine’, Revue Informatique et Statistique 24: 151–62.
Desbordes, Françoise (1984). ‘Actes de langage chez Varron?’, in Sylvain Auroux et al. (eds),
Matériaux pour une histoire des théories linguistiques. Lille: Université de Lille, III.1–8.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Detges, Ulrich (1998). ‘Echt die Wahrheit sagen: Überlegungen zur Grammatikalisierung von
Adverbmarkern’, Philologie im Netz 4: 6–12.
Detges, Ulrich (2000). ‘Time and truth: The grammaticalization of resultatives and perfects
within a theory of subjectivization’, Studies in Language 24: 345–77.
Devine, Andrew M. and Stephens, Laurence (2006). Latin Word Order: Structured Meaning
and Information. Oxford: Oxford University Press.
Devine, Andrew M. and Stephens, Laurence (2013). Semantics for Latin. Oxford: Oxford
University Press.
Dickey, Eleanor (2000a). ‘O egregie grammatice: The vocative endings of Latin words ending in
-ius’, Classical Quarterly 50: 548–62.
Dickey, Eleanor (2000b). ‘O dee ree PIE’, Glotta 76: 32–49.
Dickey, Eleanor (2006). ‘The use of Latin sis as a focus-marking clitic particle’, Oxford
University Working Papers in Linguistics, Philology & Phonetics 11: 21–5.
Dickey, Eleanor (2012a). ‘The rules of politeness and Latin request formulae’, in Philomen
Probert and Andreas Willi (eds), 313–28.
Dickey, Eleanor (2012b). ‘How to say “please” in Classical Latin’, Classical Quarterly 62: 731–48.
Dickey, Eleanor and Chahoud, Anna (eds) (2010). Colloquial and Literary Latin. Cambridge:
Cambridge University Press.
Dietrich, Wolf (1973). ‘Der periphrastische Verbalaspekt im Griechischen und Lateinischen’,
Glotta 51: 188–228.
Dik, Simon C. (1997). The Theory of Functional Grammar, 2 vols. Berlin: Mouton de Gruyter.
Dittenberger, Wilhelm (1869). ‘Zu Caesar’, Hermes 3: 375–81.
Dixon, Robert W. M. (1977). ‘Where have all these adjectives gone?’, Studies in Language 1:
19–80.
Dognini, Cristiano (2007). ‘Forme di agentività in latino: Il caso di per + accusativo’, Giornale
Italiano di Filologia 59: 137–50.
Dokkum, Thomas (1900). De constructionis analyticae vice acc. c. inf. fungentis usu apud
Augustinum. Dissertation, University of Groningen.
Domínguez Domínguez, Juan Francisco and Martín Domínguez, Antonio María (1993). ‘Dare
con infinitivo en latín clásico: Puntualizaciones sobre un problema de sintaxis y de traduc-
ción (Caes. Civ. 1.6.3)’, Cuadernos de Filología Clásica 4: 9–22.
Donati, Margherita (2004). ‘Le nozioni di agentività e strumentalità in latino e greco: Ca-
ratteristiche semantiche e codifiche grammaticali’, Studi e saggi linguistici 42: 121–51.
Draeger, Anton August (1882). Über Syntax und Stil des Tacitus. Leipzig: Teubner.
Dressler, Wolfgang (1965). ‘Die Funktion des historischen Infinitivs im lateinischen Verbal-
system’, Kratylos 10: 191–6
Dressler, Wolfgang (1968). Studien zur verbalen Pluralität. Iterativum, distributivum, durativum,
intensivum in der allgemeinen Grammatik, im Lateinischen und Hethitischen. Vienna: Böhlau.
Dressler, Wolfgang (1970). ‘Comment décrire la syntaxe des cas en latin?’, Revue de Philologie
44: 25–36.
Dressler, Wolfgang (2002). ‘Latin inflection classes’, in A. Machtelt Bolkestein et al. (eds),
91–110.
Drexler, Hans (1962). ‘Über Gerundium und Gerundivum’, Gymnasium 69: 429–45.
Dupraz, Emmanuel (2012). ‘Ille comme démonstratif anamnestique en latin classique’, in
Camille Denizot and Emmanuel Dupraz (eds), 73–115.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Dutsch, Dorota M. (2008). Feminine Discourse in Roman Comedy. Oxford: Oxford University
Press.
Ebeling, Carl L. (1957). ‘On case theories’, Museum 62: 129–44.
Eckert, Günter (2003). Spezifizität. Münster: Nodus Publikationen.
Edwards, George Vail (1902). ‘Ingenium in the ablative of quality and the genitive of quality’,
in Studies in Honour of B. L. Gildersleeve. Baltimore: Johns Hopkins Press, 301–14.
Eeden, Willem van (1985). Învǎtǎturi preste toate zilele (1642): Édition et étude linguistique, 2
vols. Amsterdam: Rodopi.
Ehrismann, Gustav (1900/1). ‘Dutzen und Ihrzen im Mittelalter’, Zeitschrift für deutsche
Wortforschung 1: 117–49.
Ehrismann, Heinrich (1886). De temporum et modorum usu Ammianeo. Straßburg: Truebner.
Eklund, Sten (1970). The Periphrastic, Completive and Finite Use of the Present Participle in
Latin. Stokholm: Almqvist & Wiksell.
Encinas, Jorge (1994). ‘Oraciones interrogativas indirectas con infinitivo en latín’, Actas del
VIII Congreso Español de Estudios Clásicos I. Madrid: Ediciones Clásicas, 473–80.
Eriksen, Pål, Kittilä, Seppo, and Kolehmainen, Leena (2010). ‘The linguistics of weather: Cross-
linguistic patterns of meteorological expressions’, Studies in Language 34: 565–601.
Eriksson, Nils (1934). ‘Studien zu den Annalen des Tacitus’. Dissertation, Lund University.
Ernout, Alfred (1908–9). ‘Recherches sur l’emploi du passif latin à l’époque républicaine’,
Mémoires de la Société de Linguistique de Paris 15: 273–333.
Espinilla, Empar, Quetglas, Pere, and Torrego, Esperanza (eds) (1999). La ‘consecutio tem-
porum’ latina. Madrid and Barcelona: Universidad Autónoma de Madrid / Universidad de
Barcelona.
Espinilla, Empar, Quetglas, Pere, and Torrego, Esperanza (eds) (2002). La comparación en latín.
Madrid and Barcelona: Universidad Autónoma de Madrid / Universidad de Barcelona.
Evans, Trevor Vivian and Obbink, Dirk D. (eds) (2010). The Language of the Papyri. Oxford:
Oxford University Press.
Fanelli, Valentina (2010). ‘Le costruzioni reciproche nella lingua latina’, in Peter Anreiter and
Manfred Kienpointner (eds), 257–65.
Farron, Stephen George (1973). ‘The imperfect indicative in Latin narrative’. Dissertation,
Columbia University, New York.
Feltenius, Leif (1977). Intransitivizations in Latin. Stokholm: Almqvist & Wiksell.
Ferri, Rolando (2008). ‘Linguistic politeness in Roman comedy’, Materiali e Discussioni
61: 12–26.
Ferri, Rolando (2012). ‘How to say no in Latin: Negative turns, politeness and pragmatic
variation’, in Martti Leiwo, Hilla Halla-aho, and Marja Vierros (eds), 115–35.
Feuillet, Jack (1998). ‘Typologie de “être” et phrases essives’, in Jack Feuillet (ed.), 663–751.
Feuillet, Jack (ed.) (1998). Actance et valence dans les langues de l’Europe. Berlin, Mouton de
Gruyter.
Finke, Annerose (1987). ‘Zur direkten Satzfrage im Französischen und Lateinischen’, in
Wolfgang Dahmen et al. (eds), Latein und Romanisch. Tübingen: Narr, 270–92.
Fleck, Frédérique (2008). Interrogation, coordination et subordination: le latin quin. Paris:
Presses de l’Université Paris-Sorbonne.
Fleischman, Suzanne (1982). The Future in Thought and Language. Cambridge: Cambridge
University Press.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Fridh, Åke (1971). Der sogenannte prospektive Konjunktiv im Lateinischen. Göteborg: Kungl.
Vetensaps- och Vitterhets-Samhället.
Froehlich, Heinrich (1889). De grammaticae Latinae locis aliquot controversis. Hagenau:
Buchdr. F. Gilardone.
Fruyt, Michèle (1987). ‘Interprétation sémantico-référentielle du réfléchi latin’, Glotta 65: 205–21.
Fruyt, Michèle (1990a). ‘La plurivalence des noms d’agent latins en -tor: Lexique et séman-
tique’, Latomus 49: 59–70.
Fruyt, Michèle (1990b). ‘La formation des mots par agglutination en latin’, Bulletin de la
Société de Linguistique de Paris 85: 173–209.
Fruyt, Michèle (1995). ‘L’accusatif et les noms en -tio chez Plaute’, in Dominique Longrée (ed.),
131–42.
Fruyt, Michèle (2001). ‘Réflexions sur la notion de mot en latin: Les verbes du type calefacio’, in
Claude Moussy (ed.), 81–94.
Fruyt, Michèle (2002a). ‘Réfléchi indirect et énonciation en latin’, in Lea Sawicki and Donna
Shalev (eds), 123–32.
Fruyt, Michèle (2002b). ‘La négation comme modalité originale: Quelques exemples latins’, in
Michèle Fruyt and Claude Moussy (eds), 37–52.
Fruyt, Michèle (2003). ‘Anaphore, cataphore et deixis dans l’Itinerarium d’Égérie’, in Heikki
Solin et al. (eds), 99–119.
Fruyt, Michèle (2005). ‘La corrélation et la proposition relative dans l’Itinerarium d’Égérie’, in
Sándor Kiss et al. (eds), 337–52.
Fruyt, Michèle (2008). ‘Focalisation des pronoms personnels et des adjectifs possessifs en latin’,
in Claude Brunet (ed.), Des formes et des mots. Besançon: Presses universitaires, 75–89.
Fruyt, Michèle (2011). ‘Cas et évolution linguistique en latin’, in Michèle Fruyt et al. (eds),
235–57.
Fruyt, Michèle (2012). ‘Quominus à l’interface entre négation et subordination’, in Linguarum
Varietas 1: 65–77.
Fruyt, Michèle and Moussy, Claude (eds) (1996). Structures lexicales du latin, Recherches
linguistiques du Centre Alfred Ernout 3. Paris: Presses de l’Université de Paris-Sorbonne.
Fruyt, Michèle and Moussy, Claude (eds) (2002). Les modalités en latin. Paris: Presses de
l’Université de Paris-Sorbonne.
Fruyt, Michèle and Orlandini, Anna (2008). ‘Some cases of linguistic evolution and grammat-
icalisation in the Latin verb’, in Roger Wright (ed.), 230–7.
Fruyt, Michèle, Mazoyer, Michel, and Pardee, Dennis (eds) (2011). Variations, concurrence et
évolution des cas dans divers domaines linguistiques / Grammatical Case in the Languages of
the Middle East and Europe, Studies in Ancient Oriental Civilization 64. Chicago: Oriental
Institute of the University of Chicago, <http://oi.uchicago.edu/pdf/saoc64.pdf>.
Fruyt, Michèle and Spevak, Olga (eds) (2010). La quantification en latin. Paris: Harmattan.
Fugier, Huguette (1973). ‘L’apposition en latin’, La Linguistique 9: 97–113.
Fugier, Huguette (1974). ‘Y a-t-il des pronoms personnels en latin?’, Revue des Études Latines
52: 384–409.
Fugier, Huguette (1985). ‘Le vocatif dans la phrase latine’, in Christian Touratier (ed.), 105–18.
Fugier, Huguette (1989). ‘La temporalisation de l’énoncé en latin’, in Marius Lavency and
Dominique Longrée (eds), 139–48.
Fugier, Huguette (1994). ‘Le verbe latin “incorpore”-t-il ses compléments?’, in József Herman
(ed.), 75–90.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Bibliography
Hahn, E. Adelaide (1950). ‘Genesis of the infinitive with subject accusative’, Transactions of the
American Philological Association 81: 117–29.
Hahn, E. Adelaide (1953). ‘Vestiges of partitive apposition in Latin syntax’, Transactions of the
American Philological Association 84: 92–123.
Hahn, E. Adelaide (1965). ‘Was there a nominative gerund? (Iterum de gerundio et gerundivo)’,
Transactions and Proceedings of the American Philological Association 96: 181–207.
Hajdú, István (1999). ‘Pro consule oder proconsul’, Museum Helveticum 56: 119–27.
Hale, William Gardner (1891). Die cum-Konstruktionen: Ihre Geschichte und ihre Funktionen.
Leipzig: Teubner.
Hall, Jon (2005). ‘Politeness and formality in Cicero’s letter to Matius’, Museum Helveticum
62: 193–213.
Hall, Jon (2009). Politeness and Politics in Cicero’s Letters. Oxford: Oxford University Press.
Halla-aho, Hilla (2009). The Non-literary Latin Letters: A Study of their Syntax and Pragmatics.
Helsinki: Societas Scientiarum Fennica.
Halla-aho, Hilla (2010). ‘Linguistic varieties and language level in Latin non-literary letters’, in
Trevor V. Evans and Dirk Obbink (eds), 171–83.
Halla-aho, Hilla (2011). ‘Epistolary Latin’, in James Clackson (ed.), 426–44.
Halla-aho, Hilla (2012). ‘A historical perspective on Latin proleptic accusatives’, Revue de
Linguistique Latine du Centre Ernout: De Lingua Latina 7.
Hamp, Carl (1888). ‘Die zusammengesetzten Präpositionen im Lateinischen’, Archiv für
Lateinische Lexikographie und Grammatik 5: 320–68.
Hand, Ferdinand G. (1829/45). Tursellinus seu de particulis Latinis commentarii, 4 vols.
Leipzig: Weidmann.
Händel, Paul and Szantyr, Anton (1954). ‘Induo und Verwandtes’, Museum Helveticum 11: 43–6.
Handford, Stanley A. (1946). The Latin Subjunctive. London: Methuen.
Hannay, Mike and Vester, Elseline (eds) (1990). Working with Functional Grammar: Descrip-
tive and Computational Applications. Dordrecht: Foris.
Happ, Heinz (1967). ‘Die lateinische Umgangssprache und die Kunstsprache des Plautus’,
Glotta 45: 60–114.
Happ, Heinz (1976). Grundfragen einer Dependenz-Grammatik des Lateinischen. Göttingen:
Vandenhoeck & Ruprecht.
Harrison, Stephen (2006). ‘Some textual problems in Apuleius’ Metamorphoses’, in Wytze
Hette Keulen et al. (eds), Lectiones Scrupulosae. Groningen: Bakhuis, 59–67.
Hartmann, Markus (2005). Die frühlateinischen Inschriften und ihre Datierung. Bremen: Hempen.
Haspelmath, Martin (1987). Verbal Noun or Verbal Adjective? The Case of the Latin Gerundive
and Gerund. Köln: Institut für Sprachwissenschaft.
Haspelmath, Martin and Michaelis, Suzanne (2008). ‘Leipzig fourmille de typologues—Genitive
objects in comparison’, in Greville G. Corbett and Michael Noonan (eds), Case and
Grammatical Relations. Amsterdam: Benjamins, 149–66.
Haug, Dag (2004). ‘Review of Oldsjö (2001)’, Eranos 102: 125–32.
Hausburg, Bianca C. (2014). ‘Nachklassisches Griechisch bei Plautus? Zu Pseudolus 483–4 und
Bacchides 1162’, Rheinisches Museum für Philologie 157: 16–26.
Häusler, Sabine (2000). ‘Parenthesen im Lateinischen am Beispiel der Pliniusbriefe’, Glotta
76: 202–31.
Haverling, Gerd (1988). Studies of Symmachus’ Language and Style, Studia Graeca et Latina
Gothoburgiensia 49. Göteborg: Acta Universitatis Gothoburgensis.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Haverling, Gerd (1989). ‘On the use of absque in Cicero and Quintilian’, Eranos 87: 133–8.
Haverling, Gerd (1995). ‘On the “illogical” Vos in Late Latin epistolography’, in Louis Callebat
(ed.), 336–53.
Haverling, Gerd (1998). ‘On the development of the perfect and imperfect tenses in Late Latin’,
in Benjamín García Hernández (ed.), I.363–78.
Haverling, Gerd (2000). On Sco-verbs, Prefixes and Semantic Functions. Göteborg: Acta
Universitatis Gothoburgensis.
Haverling, Gerd (2002). ‘On the semantic functions of the Latin perfect’, in A. Machtelt
Bolkestein et al. (eds), 153–67.
Haverling, Gerd (2005). ‘Sur les fonctions de l’imparfait dans le latin tardif ’, in Sándor Kiss
et al. (eds), 165–76.
Haverling, Gerd (2006a). ‘Tempo e aspetto nel sistema verbale latino’, Athenaeum 94: 275–9.
Haverling, Gerd (2006b). ‘On cases and prepositions in vulgar and literary Latin’, in Carmen
Arias (ed.), 345–60.
Haverling, Gerd (2010a). ‘Actionality, tense, and viewpoint’, in Philip Baldi and Pierluigi
Cuzzolin (eds), 277–523.
Haverling, Gerd (2010b). ‘On tense, viewpoint and modality from Early to Late Latin’, in Peter
Anreiter and Manfred Kienpointner (eds), 143–58.
Haverling, Gerd (2010c). ‘On absolute-relative future, viewpoint and mood’, Papers on Grammar
11: 131–45.
Haverling, Gerd (2011). ‘Le flottement entre les cas en latin tardif ’, in Michèle Fruyt et al.
(eds), 275–96.
Haverling, Gerd (2012). ‘Latin tardif littéraire et latin tardif parlé’, in Frédérique Biville et al.
(eds), 91–102.
Haverling, Gerd (2013). ‘On tense and mood in conditional clauses from Early to Late Latin’,
in Folke Josephson and Ingmar Söhrman (eds), Diachronic and Typological Perspectives on
Verbs. Amsterdam: Benjamins, 13–56.
Haverling, Gerd (ed.) (2015). Latin Linguistics in the Early 21st Century. Uppsala: Acta
Universitatis Upsaliensis.
Havers, Wilhelm (1911). Untersuchungen zur Kasussyntax der indogermanischen Sprachen.
Strassburg: K. J. Trübner.
Havers, Wilhelm (1925). ‘Der sogenannte nominativus pendens’, Indogermanische Forschun-
gen 43: 207–57.
Heberlein, Friedrich (1986). ‘Über “Weglaßbarkeit” und “Notwendigkeit” in einer lateinischen
Valenzgrammatik’, in Peter Krafft and Hans J. Tschiedel (eds), Concentus Hexachordus.
Regensburg: Pustet, 33–77.
Heberlein, Friedrich (1989). ‘Gerundivkonstruktion: Interne Struktur und Diathese’, Gymna-
sium 96: 49–72.
Heberlein, Friedrich (1996). ‘Über “enge” Appositionen’, in Hannah Rosén (ed.), 343–60.
Heberlein, Friedrich (1998). ‘Relationalität und relationsverändernde Operationen beim
Verbum infinitum’, in Benjamín García Hernández (ed.), 379–95.
Heberlein, Friedrich (2001). ‘Zur Entwicklung des lateinischen Konsekutivsatzes’, Acta
Antiqua Academiae Scientiarum Hungaricae 41: 41–54.
Heberlein, Friedrich (2002). ‘NCI und Modalverb’, in Lea Sawicki and Donna Shalev (eds),
172–83.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Heberlein, Friedrich (2011). ‘Temporal clauses’, in Philip Baldi and Pierluigi Cuzzolin (eds),
IV.235–371.
Heerdegen, Eugen Gottfried Ferdinand (1913). Das Wort vivere in phraseologischem Gebrauch
bei Horaz und im älteren Latein. Universität Programm, Erlangen.
Heick, Otto William (1936). The Ab Urbe Condita Construction in Latin. Lincoln: University
of Nebraska.
Heidrich, Georg (1892). Der Stil des Varro. Melk: Selbstverlag des Gymnasiums.
Heilig, Ernst (1978). ‘Untersuchungen zur Valenz und Distribution lateinischer und deutscher
Verben’. Dissertation, University of Tübingen.
Heine, Bernd (1997). Possession: Cognitive Sources, Forces, and Grammaticalization. Cam-
bridge, Cambridge University Press.
Heine, Rolf (1990). ‘Review of Harm Pinkster: Lateinische Syntax und Semantik. Franke
Verlag, Tübingen 1988’, Göttingische Gelehrte Anzeigen 242: 1–14.
Hellegouarc’h, Joseph (1964). Le monosyllabe dans l’hexamètre latin. Paris: Klincksieck.
Helttula, Anne (1985). ‘Post depositum militiae munus’, Arctos 2: 41–56.
Hengst, Daniel den (2007). ‘Die Kausalkonjunktionen quod, quia und quoniam in der Historia
Augusta’, in Giorgio Bonamente and Hartwin Brandt (eds), Historiae Augustae Colloquium
Bambergense. Bari: Edipuglia, 165–73.
Hentschel, Elke (1998). Negation und Interrogation. Tübingen: Niemeyer.
Herman, József (1957). ‘Cur, quare, quomodo: Remarques sur l’évolution des particules
d’interrogation en latin vulgaire’, Acta Antiqua Academiae Scientiarum Hungaricae 5,
369–77 (= 1990: 289–97).
Herman, József (1963). La formation du système roman des conjonctions de subordination.
Berlin: Akademie Verlag.
Herman, József (1987). ‘La disparition de -s et la morphologie dialectale du latin parlé’, in
Herman, József (ed.), 97–108.
Herman, József (1989). ‘Accusativus cum Infinitivo et subordonnée à quod, quia en latin tardif:
Nouvelles remarques sur un vieux problème’, in Gualtiero Calboli (ed.), 133–52.
Herman, József (1990). Du Latin aux langues romanes: Études de linguistique historique.
Tübingen: Niemeyer.
Herman, József (1991). ‘On the grammatical subject in Late Latin’, in Robert Coleman (ed.),
415–25 [= 2006: 55–64].
Herman, József (1995). ‘Les ardoises wisigothiques et le problème de la différenciation
territoriale du latin’, in Louis Callebat (ed.), 63–76.
Herman, József (1996a). ‘À propos du si interrogatif: Évolutions achevées et évolutions
bloquées’, in Alfred Bammesberger and Friedrich Heberlein (eds), 296–307 [= 2006:
65–75].
Herman, József (1996b). ‘Remarques sur l’histoire du futur latin—et sur la préhistoire du futur
roman’, in Rodie Risselada et al. (eds), 57–70.
Herman, József (2003). ‘Notes syntaxiques sur la langue de Trimalchion et de ses invités’, in
József Herman and Hannah Rosén (eds), 139–46.
Herman, József (2006). Du latin aux langues romanes II. Tübingen: Niemeyer.
Herman, József (ed.) (1987). Latin vulgaire—Latin tardif I. Tübingen: Niemeyer.
Herman, József (ed.) (1994). Linguistic Studies in Latin. Amsterdam: Benjamins.
Herman, József and Rosén, Hannah (eds) (2003). Petroniana. Gedenkschrift für Hubert
Petersmann. Heidelberg: Winter.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Hernández Cabrera, Tomás (2002). ‘La comparación con quam como criterio de caracteriza-
ción funcional’, in Empar Espinilla et al. (eds), 105–27.
Herslund, Michael (1988). Le datif en français. Louvain and Paris: Peeters.
Hessen, Bernd (1984). Der historische Infinitiv im Wandel der Darstellungstechnik Sallusts.
Frankfurt am Main: Lang.
Hettrich, Heinrich (1990). ‘Rektionaler und autonomer Kasusgebrauch’, in Heiner Eichner and
Helmut Rix (eds), Sprachwissenschaft und Philologie. Wiesbaden: Reichert Verlag, 82–98.
Hettrich, Heinrich (1992). ‘Die Entstehung des lateinischen und griechischen AcI’, in Robert
Beekes et al. (eds), Rekonstruktion und relative Chronologie. Innsbruck: Innsbrucker Bei-
träge zur Sprachwissenschaft, 221–34.
Hettrich, Heinrich (1993). ‘Nochmals zu Gerundium und Gerundivum’, in Gerhard Meiser
(ed.), Indogermanica et Italica: Festschrift für Helmut Rix zum 65. Geburtstag. Innsbrucker
Beiträge zur Sprachwissenschaft 72. Innsbruck: Institut für Sprachwissenschaft der Uni-
versität Innsbruck, 190–208.
Hettrich, Heinrich (1994). ‘Semantische und syntaktische Betrachtungen zum doppelten
Akkusativ’, in George E. Dunkel, Gisela Meyer, Salvator Scarlata, Christian Seidl (eds),
Früh-, Mittel, Spätindogermanisch. Wiesbaden: Reichert Verlag, 111–34.
Hettrich, Heinrich (1997). ‘Syntaktische Rekonstruktionen bei Delbrück und heute: Nochmals
zum lateinischen und griechischen AcI’, in Emilio Crespo and José García Ramón (eds),
Berthold Delbrück y la sintaxis indoeuropea hoy. Wiesbaden: Reichert Verlag, 219–38.
Hetzron, Robert (1978). ‘On the relative order of adjectives’, in Hansjakob Seiler (ed.),
Language Universals. Tübingen: Narr, 165–84.
Hey, Oskar (1906). ‘Zur enallage adiectivi’, Archiv für Lateinische Lexikographie und Gram-
matik 14: 105–12, 268.
Heyde, Klaas van der (1930). ‘L’ablatif de comparaison en latin’, Revue des Études Latines 8:
230–41.
Heyworth, Stephen J. (2007). Cynthia. Oxford: Oxford University Press.
Hillen, Michael (1989). Studien zur Dichtersprache Senecas: Abundanz, explikativer Ablativ,
Hypallage. Berlin: De Gruyter.
Hilton, John (1997/8). ‘The role of discourse and lexical meaning in the grammaticalisation of
temporal particles in Latin’, Glotta 74: 198–210.
Himmelmann, Nikolaus (1986). Morphosyntactic Predication: A Functional-Operational
Approach, AKUP 62. Köln: Institut für Sprachwissenschaft.
Hine, Harry M. (2009). ‘Subjectivity and objectivity in Latin scientific and technical literature’,
in Liba Taub and Aude Doody (eds), Authorial Voices in Greco-Roman Technical Writing.
Trier: Wissenschaftlicher Verlag, 13–30.
Hintzen, Beate (1993). Das Partizip Präsens in Ciceros Reden unter Einbeziehung des Partizip
Perfekt der Deponentien. Münster: Waxmann.
Hjelmslev, Louis (1935, 1937). La catégorie des cas: Étude de grammaire générale. Aarhus:
Universitetsforlaget.
Hoecke, Willy van (1996). ‘The Latin dative’, in Willy van Langendonck and William van Belle
(eds), I.3–37.
Hoff, François (1983). ‘Interrogation, interrogation rhétorique et exclamation en latin’, in
Harm Pinkster (ed.), 123–9.
Hoffmann, Maria E. (1987). Negatio contrarii. Assen: van Gorcum.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Hyart, Charles (1960). ‘La concordance des temps du subjonctif dans le style indirect chez César’,
Hommages à Léon Herrmann. Bruxelles-Berchem: Latomus, Revue des Études Latines, 454–63.
Ihm, Max (1890). ‘Römische Spieltafeln’, in Aufsätze aus der Altertumswissenschaft, Reinhard
Kekulé zur Erinnerung an seine Lehrthätigkeit in Bonn gewidmet von seinen Schülern.
Festschrift Reinhard Kekulé von Stradonitz, Bonner Studien. Berlin: W. Spemann, 223–9.
Iliescu, Maria (1999). ‘L’origine des syntagmes à/a + SN (COI) dans les langues romanes’, in
Hubert Petersmann and Rudolf Kettemann (eds), 267–75.
Iliescu, Maria and Marxgut, Werner (eds) (1992). Latin vulgaire—Latin tardif III. Tübingen:
Niemeyer.
Iordache, Roxana (1999). ‘Le complément de concession en latin: Définition du complément
de concession en latin. L’oxymoron en latin’, Živa Antika 49: 67–100.
Iovino, Rossella (2012). La sintassi dei modificatori nominali in latino. Munich: Lincom.
Iso, José Javier (1994). ‘Sobre la desaparición de la flexión en el latín tardío’, Actas del VIII
Congreso Español de Estudios Clásicos. Madrid: Ediciones Clásicas, I.343–62.
Jacob, Daniel (1995). ‘À propos de la périphrase habeo + participe parfait passif ’, in Louis
Callebat (ed.), 367–81.
Jacobs, Joachim (1994). Kontra Valenz. Trier: Wissenschaftlicher Verlag.
Jacquinod, Bernard (1992). ‘Le double accusatif dans les Métamorphoses d’Apulée’, Revue de
Philologie 66: 81–92.
Janson, Tore (1979). Mechanisms of Language Change in Latin. Stockholm: Almqvist &
Wiksell.
Jensen, Paul J. (1941). Det latinske perfektsystem: Studier i kategorierne modus og tempus i
Ciceros sprog. Copenhagen: Munksgaard.
Jocelyn, Henry David (1993). ‘Sprache, Schriftlichkeit und Charakterisierung in der römischen
Komödie’, in Gregor Vogt-Spira (ed.), Beiträge zur mündlichen Kultur der Römer. Tübin-
gen: Narr, 125–39.
Joffre, Marie-Dominique (1995). Le verbe latin: Voix et diathèse. Louvain-la-Neuve: Peeters.
Joffre, Marie-Dominique (1997). ‘De l’existence de véritables verbes pronominaux en latin
classique’, Revue de Philologie 71: 55–64.
Joffre, Marie-Dominique (1998). ‘Les possessifs des personnes 1 et 2: Un moyen de faire surgir
dans l’énoncé l’un des acteurs de l’énonciation’, in Benjamín García Hernández (ed.), 409–17.
Joffre, Marie-Dominique (2002). ‘Réflexions sur le fonctionnement de la périphrase ab domo
abeundum est mihi (Pl. Aul. 105)’, in Michèle Fruyt and Claude Moussy (eds), 163–72.
Joffre, Marie-Dominique (2007a). ‘Le participe passé latin et la diathèse’, in Frédéric Lambert,
Catherine Moreau, and Jean Albrespit (eds), Les formes non finies du verbe 2. Rennes:
Presses Universitaires, 37–51.
Joffre, Marie-Dominique (2007b). ‘Ocelle mi (Plaut. Trin. 245), un emploi hors structure?’, in
Colette Bodelot (ed.), 123–35.
Joffre, Marie-Dominique (2007c). ‘Ipse, anaphore et deixis’, in Gérald Purnelle and Joseph
Denooz (eds), 97–110.
Joffre, Marie-Dominique (2011). ‘Que montrent les déictiques?’, Revue des Études Latines 88:
53–71.
Joffre, Marie-Dominique (2012a). ‘La concurrence hic/iste dans les Métamorphoses d’Apulée:
Réflexions sur leur emplois exophoriques’, in Frédérique Biville et al. (eds), 335–46.
Joffre, Marie-Dominique (2012b). ‘Signification et emplois de hic: Réflexions sur l’endophore et
l’exophore dans un corpus littéraire’, in Camille Denizot and Emmanuel Dupraz (eds), 57–72.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Kiss, Sándor, Mondin, Luca, and Salvi, Giampaolo (eds) (2005). Latin et langues romanes
(Études de linguistique offertes à József Herman). Tübingen: Niemeyer.
Klingner, Friedrich (1955). ‘Beobachtungen über Sprache und Stil des Tacitus am Anfang des
13. Annalenbuches’, Hermes 83: 187–200.
Klug, Wolfgang (1992). Erzählstruktur als Kunstform: Studien zur künstlerischen Funktion der
Erzähltempora im Lateinischen und Griechischen. Heidelberg: Manutiusverlag.
Köhler, Helga (1999). ‘Der historische Infinitiv in den Briefen des Sidonius’, in Hubert
Petersmann and Rudolf Kettemann, 409–18.
Koller, Hermann (1951). ‘Praesens historicum und erzählendes Imperfekt: Beitrag zur Aktions-
art der Praesensstammzeiten im Lateinischen und Griechischen’, Museum Helveticum
8: 63–99.
König, Ekkehard (2010). ‘Le latin ipse et ses équivalents dans d’autres langues: Essai de
typologie des intensifieurs’, in Michèle Fruyt and Olga Spevak (eds), 21–37.
Kooreman, Marion (1989). ‘The historical development of the ablative of the gerund’, in
Marius Lavency and Dominique Longrée (eds), 219–30.
Kooreman, Marion (1995). ‘The expression of obligation and necessity in the works of
Tertullian: The use of habere + infinitive, -urus esse, and the gerundive’, in Louis Callebat
(ed.), 383–94.
Kooreman, Marion (1996). ‘The use of the active periphrastic future in some biblical transla-
tions’, in Hannah Rosén (ed.), 323–32.
Kraggerud, Egil (2008). ‘Fors et in Vergil, Horace, and Propertius’, Eranos 105: 36–9.
Kravar, Miroslav (1969). ‘Nochmals zum historischen Präsens im Lateinischen’, Živa Antika
29: 25–9.
Kravar, Miroslav (1980). Pitanja glagolskoga vida u latinskom jeziku. Skopje: Živa antika/
Posobni izcania 6.
Kraz, Heinrich (1862). Über den Modus der rhetorischen Frage in der lateinischen oratio
obliqua: Die sogenannte unwillige oder missbilligende Frage mit dem Conjunctiv, mit ut
und dem Conjunctiv, mit dem Accusativ und Infinitiv. Stuttgart: Druck der k. Hof- und
Canzleibuchdruckerei von G. Mäntler.
Krenn, Herwig (1993). ‘Vom lateinischen zum romanischen Adverb’, in Christian Schmitt
(ed.), Grammatikographie der romanischen Sprachen. Bonn: Romanistischer Verlag,
302–18.
Kroon, Caroline H. M. (1989a). ‘Rarum dictu: The Latin second supine construction’, Glotta
67: 198–227.
Kroon, Caroline H. M. (1989b). Causal connectors in Latin: The discourse function of nam,
enim, igitur, ergo’, in Marius Lavency and Dominique Longrée (eds), 231–43.
Kroon, Caroline H. M. (1995). Discourse Particles in Latin. Amsterdam: Gieben.
Kroon, Caroline H. M. (1998). ‘Discourse particles, tense, and the structure of Latin narrative
texts’, in Rodie Risselada (ed.), 37–61.
Kroon, Caroline H. M. (2000). ‘Het ene verhaal is het andere niet. Een taalkundige kijk op
teksttype’, Lampas 33: 211−38.
Kroon, Caroline H. M. (2002). ‘How to write a ghost story? A linguistic view on narrative
modes in Pliny Ep. 7.27’, in Lea Sawicki and Donna Shalev (eds), 189–200.
Kroon, Caroline H. M. (2007). ‘Discourse modes and the use of tenses in Ovid’s Metamor-
phoses’, in Rutger J. Allan and Michel Buijs (eds), The Language of Literature: Linguistic
Approaches to Classical Texts. Leiden: Brill, 65–92.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Kroon, Caroline H. M. (2011). ‘Latin particles and the grammar of discourse’, in James
Clackson (ed.), A Companion to the Latin Language. Chichester and Malden: Wiley-
Blackwell, 176–95.
Kroon, Caroline H. M. (forthc.). ‘Textual deixis and the “anchoring” use of Latin hic’
(submitted to Mnemosyne).
Kroon, Caroline H. M. and Risselada, Rodie (1998). ‘The discourse functions of iam’, in
Benjamín García Hernández (ed.), 417–33.
Kroon, Caroline H. M. and Risselada, Rodie (2002). ‘Phasality, polarity, focality: A feature
analysis of the Latin particle iam’, Belgian Journal of Linguistics 16: 63–78.
Kroon, Caroline H. M. and Rose, Paula (1996). ‘Atrociter corruptus? The use of the “narrative”
tenses in Ammianus Marcellinus’ Res Gestae’, in Rodie Risselada et al. (eds), 71–89.
Kümmel, Martin Joachim (2013). ‘Zur Verwendung des Perfektfuturs im Altlatein’, Glotta 89:
130–45.
Kuntz, Friedrich (1962). ‘Die Sprache des Tacitus und die Tradition der lateinischen Histo-
rikersprache’. Dissertation, Heidelberg University.
Kuryłowicz, Jerzy (1949). ‘Le problème du classement des cas’, Bulletin de la Société polonaise
de linguistique 9: 20–43.
Kurzová-Jedličková, Helena (1961). ‘Die Demonstrativa im Vulgärlatein’, Acta Antiqua Aca-
demiae Scientiarum Hungaricae 11: 121–43.
Kurzová-Jedličková, Helena (1981). Der Relativsatz in den indogermanischen Sprachen. Ham-
burg: Buske.
Kurzová-Jedličková, Helena (1989). ‘Subordination in Latin’, Eirene 26, 23–32.
Laer, Sophie van (2007). ‘L’apposition et la construction d’un énoncé: Prédication seconde et
stratégie discursive’, in Gérald Purnelle and Joseph Denooz (eds), 203–13.
La Fauci, Nunzio (1991). ‘La continuità nelle diversità formale: Aspetti di morfosintassi
diacronica’, in Vincenzo Orioles (ed.), Innovazione e conservazione nelle lingue. Pisa:
Giardina, 135–58.
Laing, Gordon L. (1920). The Genitive of Value in Latin and Other Constructions with Verbs of
Rating. Chicago: University of Chicago Press.
Lambert, André (1946). ‘Die indirekte Rede als künstlerisches Stilmittel des Livius’. Disserta-
tion, University of Zürich.
Lambertz, Thomas (1996). ‘Zur syntaktischen Kategorie der Verbalabstrakta auf -tu im
Lateinischen’, in Alfred Bammesberger and Friedrich Heberlein (eds), 337−53.
Landfester, Manfred (1997). Einführung in die Stilistik der griechischen und lateinischen
Literatursprachen. Darmstadt: Wissenschaftliche Buchgesellschaft.
Landgraf, Gustav (1881). ‘De figuris etymologicis linguae Latinae’, Acta Seminarii Philologici
Erlangensis 2, 1–69.
Landgraf, Gustav (1888). ‘Substantivische Parataxen’, Archiv für Lateinische Lexikographie und
Grammatik 5: 161–91.
Landgraf, Gustav (1893). ‘Der Dativus commodi und der Dativus finalis mit ihren Abarten’,
Archiv für Lateinische Lexikographie und Grammatik 8: 39–76.
Landgraf, Gustav (1898). ‘Der Akkusativ der Richtung’, Archiv für Lateinische Lexikographie
und Grammatik 10: 391–402.
Landgraf, Gustav (1898/9). Beiträge zur historischen Syntax der lateinischen Sprache. Pro-
gramm des K. Wilhelms-Gymnasiums München.
Langen, Peter (1870). ‘Beiträge zur Kritik und Erklärung des Ammianus Marcellinus’, Philo-
logus 29: 469–87.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Langendonck, Willy van and Belle, William van (eds) (1996/8). The Dative, 2 vols. Amster-
dam: Benjamins.
Langer, Stefan (2004). ‘A linguistic test battery for support verb constructions’, Linguisticae
Investigationes 27: 171–84.
Langslow, David R. (2000). Medical Latin in the Roman Empire. Oxford: Oxford University
Press.
Langslow, David R. (2005). ‘Linguistic “highs” and “lows” in Late Latin medical texts’, in
Sándor Kiss et al. (eds), 313–25.
Larjavaara, Meri (2000). Présence ou absence de l’objet: Limites du possible en français con-
temporain. Helsinki: Academia.
Laughton, Eric (1960). ‘Observations on the Style of Varro’, Classical Quarterly 10: 1–28.
Laughton, Eric (1964). The Participle in Cicero. Oxford: Oxford University Press.
Lauric, Eva (2001). ‘Parties du corps actives et passives’, Revue de Linguistique Romane 65:
145–68.
Lavency, Marius (1975). ‘Les valeurs de la “conjonction” cum en latin classique’, Les Études
Classiques 43: 376–86.
Lavency, Marius (1986). ‘Le paradigme syntaxique de l’ablatif absolu’, in Freddy Decreus and
Carl Deroux (eds), Hommages Jozef Veremans. Bruxelles: Latomus, 184–91.
Lavency, Marius (1991). ‘Pour une description fonctionnelle du syntagme nominal’, in
Charles-Marie Ternes (ed.), La nominalisation. Luxembourg: Centre Universitaire de Lux-
embourg, 55–77.
Lavency, Marius (1994). ‘Sujet de verbe ou sujet de proposition’, in József Herman (ed.),
103–11.
Lavency, Marius (19972). Usus. Louvain-la-Neuve: Peeters.
Lavency, Marius (1999). ‘La période conditionnelle du latin’, Les Études Classiques 67: 361–90.
Lavency, Marius (2003a). ‘À propos de la conjonction cum en latin classique’, Živa Antika 53:
41–57.
Lavency, Marius (2003b). ‘La proposition infinitive (AcI)’, in Colette Bodelot (ed.), 97–192.
Lavency, Marius (2005). ‘Il congiuntivo nella proposizione relativa nominalizzata’, in Gual-
tiero Calboli (ed.), 591–5.
Lavency, Marius and Longrée, Dominique (eds) (1989). Actes du Ve colloque de linguistique
latine: Louvain-la-Neuve/Borzée, 31 mars–4 avril 1989. Leuven: Peeters.
Lease, Emory B. (1902). ‘On the use of neque and nec in Silver Latin’, Classical Review 16: 212–14.
Lease, Emory B. (1913). ‘Neve and neque with the imperative and subjunctive’, American
Journal of Philology 34: 255–75, 418–36.
Lease, Emory B. (1919). ‘The use and range of the future participle’, American Journal of
Philology 40: 262–85.
Lease, Emory B. (1928). ‘The ablative absolute limited by conjunctions’, American Journal of
Philology 49: 348–53.
Lease, Emory B. (1931). ‘Addendum’, American Journal of Philology 51: 175.
Le Bourdelles, Henri (1966). ‘Pluit, tonat’, Revue des Études Latines 44: 377–406.
Lebreton, Jules (1901a). Études sur la langue et la grammaire de Cicéron, Paris (= Hildesheim:
Olms, 1979).
Lebreton, Jules (1901b). Caesariana syntaxis quatenus a Ciceroniana differat. Paris: Hachette.
Lehmann, Christian (1973). Latein mit abstrakten Strukturen. Munich: Fink.
Lehmann, Christian (1983). ‘Latin preverbs and cases’, in Harm Pinkster (ed.), 145–61.
Lehmann, Christian (1984). Der Relativsatz. Tübingen: Narr.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Lehmann, Christian (1986). ‘On the typology of relative clauses’, Linguistics 24: 663–80.
Lehmann, Christian (1991). ‘The Latin nominal group in a typological perspective’, in Robert
G. Coleman (ed.), 203–32.
Lehmann, Christian (1998). ‘Regiones espaciales en perspectiva tipológica’, in Benjamín
García Hernández (ed.), 455–65.
Lehmann, Christian (2002). ‘Latin valency in a typological perspective’, in A. Machtelt Bol-
kestein et al. (eds), 183–203.
Lehmann, Christian (2005). ‘Sur l’évolution du pronom possessif ’, in Sándor Kiss et al. (eds), 37–46.
Lehmann, Christian (forthc.). ‘Latin causativization in typological perspective’, in Murielle
Lenoble and Dominique Longrée et al. (eds), Actes du 13ème Colloque International de
Linguistique latine. Louvain: Peeters.
Leiwo, Martti (2010). ‘Petronius’ linguistic resources’, in Eleanor Dickey and Anna Chahoud
(eds), 281–91.
Leiwo, Martti, Halla-aho, Hilla, and Vierros, Marja (eds) (2012). Variation and Change in
Greek and Latin. Helsinki: Foundation of the Finnish Institute at Athens.
Lemaire, Maryvonne (1983). ‘Un aspect de l’échange des arguments du verbe: Les verbes à
double régime de l’objet en latin’, Bulletin de la Société de Linguistique de Paris 78: 283–324.
Leo, Friedrich (1878). De Senecae tragoediis observationes criticae. Berlin: Weidmann (repr.
1963).
Leo, Friedrich (1896). Analecta Plautina: De figuris sermonis I. Göttingen: Dieterich.
Leumann, Manu (1962). ‘Lateinisch habere mit Infinitiv’, Museum Helveticum 19: 65–71.
Leumann, Manu (1977). Lateinische Grammatik. Erster Band: Lateinische Laut- und Formen-
lehre. Munich: Beck.
Levy, Harry L. (1958). ‘Andria 560–565’, American Journal of Philology 29: 173–8.
Lindsay, William M. (1907). Syntax of Plautus. Oxford: Clarendon Press.
Lindskog, Claes (1896). Beiträge zur Geschichte der Satzstellung in Latein. Lund: Gleerup.
Löfstedt, Bengt (1961). Studien über die Sprache der Langobardischen Gesetze. Stockholm:
Almqvist & Wiksell.
Löfstedt, Bengt (1963a). ‘Bemerkungen zum Problem Genus : Sexus im Lateinischen’, Symbo-
lae Osloenses 38: 47–68 (= 2000: 24–46).
Löfstedt, Bengt (1963b). ‘Zum lateinischen possessiven Dativ’, Zeitschrift für vergleichende
Sprachforschung 78: 64–83 (= 2000: 3–23).
Löfstedt, Bengt (1965/6). ‘Die subjektlose dritte Person im Lateinischen’, Kungliga Humanis-
tiska Vetenskapssamfundet i Uppsala Årsbok 80–107 (= 2000: 47–74).
Löfstedt, Bengt (1966). ‘Die Konstruktion c’est lui qui l’a fait im Lateinischen’, Indogerma-
nische Forschungen 71: 253–77 (= 2000: 75–100).
Löfstedt, Bengt (1967). ‘Bemerkungen zum Adverb im Lateinischen’, Indogermanische For-
schungen 72: 79–109.
Löfstedt, Bengt (1976). ‘Zum spanischen Mittellatein’, Glotta 54: 117–57 (= 2000: 169–211).
Löfstedt, Bengt (1985a). ‘Sprachliches zur Vulgata’, Studia Classica 23: 73–84 (= 2000:
289–301).
Löfstedt, Bengt (1985b). ‘Lexikalisches zur Vulgata’, in Studia in honorem Iro Kajanto, Arctos
Supplement 2: 99–107 (= 2000: 310–18).
Löfstedt, Bengt (2000). Ausgewählte Aufsätze zur lateinischen Sprachgeschichte und Philologie.
Stuttgart: Hersemann.
Löfstedt, Einar (1911). Philologischer Kommentar zur Peregrinatio Aetheriae. Stockholm:
Almqvist & Wiksell.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Löfstedt, Einar (1936). Vermischte Studien zur lateinischen Sprachkunde und Syntax. Lund:
Gleerup.
Löfstedt, Einar (19422/1933). Syntactica, 2 vols. Lund: Gleerup.
Löfstedt, Leena (1966). Les expressions du commandement et de la défense en latin et leur survie
dans les langues romanes. Helsinki: Société Néophilologique.
Longacre, Robert E. (1983). The Grammar of Discourse. New York: Plenum.
Longacre, Robert E. (2007). ‘Sentences as combinations of clauses’, in Timothy Shopen (ed.),
Language Typology and Syntactic Description. Cambridge: Cambridge University Press,
II.372–420.
Longrée, Dominique (1990). ‘À propos du concept d’“Apposition” ’, Information Grammati-
cale 45: 8–13.
Longrée, Dominique (1991). ‘La relative en apposition, un cas de nominalisation?’, Études
Classiques (Centre universitaire de Luxembourg) 2: 77–100.
Longrée, Dominique (1996). ‘La concurrence entre “rallonge” et “parataxe” chez Tacite:
Conditionnements linguistiques, choix stylistiques’, in Hannah Rosén, (ed.), 483–98.
Longrée, Dominique (2002). ‘Verba sentiendi et modalisation de l’énoncé chez les historiens
latins’, in Michèle Fruyt and Claude Moussy (eds), 77–93.
Longrée, Dominique (2007). Mirum dictu: Les appositions de phrase chez les historiens latins’,
in Colette Bodelot (ed.), 233–47.
Longrée, Dominique (2008). ‘Ici et là chez les historiens latins’, Cahiers Chronos 20: 147–65.
Longrée, Dominique (ed.) (1995). De usu: Études de syntaxe latine offertes en hommage à
Marius Lavency. Louvain-la-Neuve: Peeters.
López Eire, Antonio and Ramos, Agustín (eds) (2004). Registros lingüísticos en las lenguas
clásicas, Aquilafuente 68. Salamanca: Ediciones Universidad.
López Fonseca, Antonio (2000). ‘Predicados no factivos y posterioridad. Possum + infinitivo en
latín tardío’, in Benjamín García Hernández (ed.), 139–56.
López Moreda, Santiago (1987). Los grupos lexemáticos de ‘facio’ y ‘ago’ en el latín arcaico y
clásico. Estudio estructural. León: Universidad de León.
Lücht, Wilko (2011). ‘Das Lateinische als pro-drop-Sprache: Der Gebrauch von Subjektpro-
nomina in der römischen Komödie aus sprachtypologischer und empirischer Sicht’, Glotta
87: 126–54.
Luck, Georg (1964). Über einige Interjektionen der lateinischen Umgangssprache. Heidelberg:
Winter.
Ludewig, Antonius (1891). Quomodo Plinius maior, Seneca philosophus, Curtius Rufus, Quinti-
lianus, Cornelius Tacitus, Plinius Minor particula ‘quidem’ usi sint. Prague: Verlag Dominicus.
Lühr, Rosemarie (2011). ‘ “Focus pronouns” in Old Latin reflexive constructions’, Studien zur
Typologie und Universalienforschung 64: 114–25.
Lundström, Sven (1961). Abhinc und ante. Studien zur Geschichte der lateinischen Zeitdiffe-
renzbestimmungen. Lund: Gleerup.
Lundström, Sven (1964). Vermeintliche Glosseme in den Tusculanen. Stockholm: Almqvist &
Wiksell.
Lundström, Sven (1982). Ein textkritisches Problem in den Tusculanen. Stockholm: Almqvist &
Wiksell.
Lundström, Sven (1986). Zur Textkritik der Tusculanen. Stockholm: Almqvist & Wiksell.
Lundström, Sven (1993). ‘Vindikationen zu Senecas Controversiae und Suasoriae’, Eranos 91: 106–19.
Luque, Jesús (1978). ‘En torno al sintagma “haber impersonal + sustantivo” y sus orígenes
latinos’, Revista Española de Lingüística 8: 125–47.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Luque, Jesús (2006). Puntos y commas: La grafía de la articulación del habla. Granada:
Editorial Universidad de Granada
Luque, Jesús (2014). ‘Deus . . . totus est . . . animae: Plin. Nat. 2.14’, Estudios Clásicos 145: 55–70.
Luraghi, Silvia (1995). ‘The pragmatics of verb-initial sentences in some ancient Indo-
European languages’, in Pamela Downing and Michael Noonan (eds), Word Order in
Discourse. Amsterdam: Benjamins, 354–86.
Luraghi, Silvia (1997). ‘Omission of the direct object in Latin’, Indogermanische Forschungen
102: 239–57.
Luraghi, Silvia (2004). ‘Null objects in Latin and Greek and the relevance of linguistic typology
for language reconstruction’, in Karlene Jones-Bley et al. (eds), Proceedings of the Fifteenth
Annual UCLA Indo-European Conference (2003). Journal of Indo-European Monograph
Series 49. Washington, DC: Institute for the Study of Man, 234–56.
Luraghi, Silvia (2005). ‘Prepositions in cause expressions’, in Gualtiero Calboli (ed.), 609–19.
Luraghi, Silvia (2010). ‘Adverbial phrases’, in Philip Baldi and Pierluigi Cuzzolin (eds),
II.19–107.
Lyons, John (1968). Introduction to Theoretical Linguistics. Cambridge: Cambridge University Press.
Mackay, Christopher S. (1999). ‘Expressions to indicate goal of motion in the colloquial Latin
of the Early Empire’, Zeitschrift für Papyrologie und Epigraphik 126: 229–39.
Magni, Elisabetta (1999). ‘La significazione del possesso in latino’, Archivio glottologico italiano
84: 44–66.
Magni, Elisabetta (2010). ‘Mood and modality’, in Philip Baldi and Pierluigi Cuzzolin (eds),
II.193–275.
Mal-Maeder, Danielle van (1994). ‘Sens et fonction du pronom/adjectif quidam dans les
Métamorphoses d’Apulée’, Museum Helveticum 51: 214–25.
Malchukov, Andrej (2000). Dependency Reversal in Noun–Attribute Constructions: Towards a
Typology. Munich: Lincom.
Malosti, Sandra (1967). ‘Uno stilema virgiliano: L’ablativo di estensione’, in Alfonso Traina
(ed.), Studi sulla lingua poetica latina. Rome: Ateneo, 19–101.
Maltby, Robert (2005). ‘Gerund and gerundive usage in Early Latin and Cicero ad familiares’.
Paper presented at 13e Colloque International de Linguistique Latine, Facultés universitaires
Saint-Louis Bruxelles.
Manfredini, Adriana M. (ed.) (2010). Debates en lenguas clásicas, I: Lengua. Buenos Aires: EFFL.
Maraldi, Mirka (1983). ‘New approaches to accusative subjects: Case theory vs. raising’, in
Harm Pinkster (ed.), 167–76.
Maraldi, Mirka (1986). ‘The proleptic accusative: Problems of structural analysis’, Papers on
Grammar 2: 86–103.
Maraldi, Mirka (1994). ‘Some remarks on the historical development of the gerund in Latin’,
Papers on Grammar 4: 141–64.
Maraldi, Mirka (2002). ‘Free-choice quantification and concession in Latin’, in Lea Sawicky
and Donna Shalev (eds), 243–53.
Marichal, Robert (1992). Les ostraca de Bu Njem. Libya Antiqua, Supplement 9. Tripoli:
Grande Jamahira Arabe, Libyenne, Populaire, et Socialiste, Departement des Antiquites.
Mariani, Manuela (2002). ‘The Latin and Greek “Dativus agentis”: Inheritance, convergence,
or interference’, Glotta 78: 116–35.
Mariani, Manuela (2004). Raggionando di grammatica e interferenza nelle lingue classiche:
Sguardo sulla sintassi del nome. Alessandria: Edizioni dell’Orso.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Marouzeau, Jules (1910a). La phrase à verbe ‘être’ en latin. Paris: Paul Geuthner.
Marouzeau, Jules (1910b). L’emploi du participe présent latin à l’époque républicaine. Paris:
Champion.
Marti, Heinrich (2005). ‘Lateinische Predigten zwischen Mündlichkeit und Schriftlichkeit’,
Museum Helveticum 62: 102–25.
Martín Puente, Cristina and Santos Guzmán, Patricia (2002). ‘El verbo muto y la expresión del
“resultado” ’, in A. Machtelt Bolkestein et al. (eds), 234–47.
Martín Puente, Cristina and Santos Guzmán, Patricia (2003). ‘El marco predicativo de verto y
el complemento direccional’, Revista Española de Lingüística 33: 409–27.
Martín Rodríguez, Antonio M. (1986). ‘Algunas observaciones sobre el dativo en Plauto’,
Estudios Humanísticos 8: 87–96.
Martín Rodríguez, Antonio M. (1996a). ‘Dare, auxiliaire lexical en latin’, in Michèle Fruyt and
Claude Moussy (eds), 49–64.
Martín Rodríguez, Antonio M. (1996b). ‘Verbes à double régime et contraintes de genre
littéraire. Aperçu diachronique sur donare’, in Alfred Bammesberger and Friedrich Heber-
lein (eds), 213–21.
Martín Rodríguez, Antonio M. (2002). ‘Debeo, verbe de modalité’, in Michèle Fruyt and
Claude Moussy (eds), 149–61.
Massaro, Matteo (2007). ‘Una coppia affittata: CLE 959’, in Peter Kruschwitz (ed.), Die
metrischen Inschriften der römischen Republik. Berlin: De Gruyter, 271–97.
Matthews, Peter Hugoe (1972). Inflectional Morphology: A Theoretical Study Based on Aspects
of Latin Verb Conjugation. Cambridge: Cambridge University Press.
Maurach, Gregor (1983). Enchiridion Poeticum. Darmstadt: Wissenschaftliche Buch-
gesellschaft.
Mayer, Roland G. (1979). ‘Pharsalica damna’, Mnemosyne 32: 338–59.
Mayer, Roland G. (1999). ‘Grecism’, in James N. Adams and Roland G. Mayer (eds), 157–82.
Meader, Clarence L. (1900). The Latin Pronouns ‘is’, ‘hic’, ‘iste’, ‘ipse’: A Semasiological Study.
New York: Macmillan.
Meader, Clarence L. (1910). ‘The usage of idem, ipse and words of related meaning’, in
Clarence L. Meader (ed.), Latin Philology. New York: Macmillan, 1–112.
Meiser, Gerhard (1993). ‘Das Gerundivum im Spiegel der italischen Onomastik’, in Frank
Heidermanns (ed.), Sprachen und Schriften des antiken Mittelmeerraums. Innsbruck: In-
stitut für Sprachwissenschaft, 252–68.
Meiser, Gerhard (2003). Veni, vidi, vici: Die Vorgeschichte des lateinischen Perfektsystems.
Munich: Beck.
Meister, Karl (1916). Lateinisch-Griechische Eigennamen. Leipzig: Teubner.
Meisterfeld, Reinhard (2000). ‘Die unbestimmte Bestimmung: Zur Entstehung des unbe-
stimmten Artikels in den Romanischen Sprachen’, in Bruno Staib (ed.), Linguistica Roma-
nica et Indiana: Festschrift für Wolf Dietrich zum 60. Geburtstag. Tübingen: Narr, 303–32.
Mellazzo, Lucio (2005). ‘Latin object and subject infinitive clauses’, in Katalin É. Kiss (ed.),
Universal Grammar in the Reconstruction of Ancient Languages. Berlin: Mouton de Gruyter,
339–72.
Mellet, Sylvie (1987). ‘Aiebam, aiebas ou les soi-disant “imparfaits aoristiques” ’, in Sylvie
Mellet (ed.), 355–68.
Mellet, Sylvie (1988). L’imparfait de l'indicatif en latin classique: Paris and Louvain-la-Neuve:
Peeters.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Mellet, Sylvie (1989). ‘À propos du futur: Temps et modalité’, in Marius Lavency and Domi-
nique Longrée (eds), 269–78.
Mellet, Sylvie (ed.) (1987). Études de linguistique générale et de linguistique latine offertes en
hommage à Guy Serbat. Paris: Société pour l’information grammaticale.
Mellet, Sylvie, Joffre, Marie-Dominique, and Serbat, Guy (1995). Grammaire fondamentale du
latin: Le signifié du verbe. Louvain-la-Neuve: Peeters.
Melo, Wolfgang D. C. de (2002). ‘The sigmatic future in Plautus’, in A. Machtelt Bolkestein
et al. (eds), 75–90.
Melo, Wolfgang D. C. de (2005). ‘The sigmatic subjunctive in Plautus and Terence’, in
Gualtiero Calboli (ed.), 45–56.
Melo, Wolfgang D. C. de (2006). ‘If in doubt, leave it in: Subject accusatives in Plautus and
Terence’, Oxford University Working Papers in Linguistics, Philology & Phonetics 11: 5–20.
Melo, Wolfgang D. C. de (2007a). The Early Latin Verb System. Oxford: Oxford University
Press.
Melo, Wolfgang D. C. de (2007b). ‘Zur Sprache der republikanischen carmina Latina epigra-
phica: Satzumfang, Satzkomplexität und Diathesenwahl’, in Peter Kruschwitz (ed.), Die
metrischen Inschriften der römischen Republik. Berlin: De Gruyter, 97–120.
Melo, Wolfgang D. C. de (2007c). ‘Latin prohibitions and the origins of the u/w perfect and the
type amāsti’, Glotta 83: 43–68.
Melo, Wolfgang D. C. de (2007d). ‘The present tense with future meaning in the accusative and
infinitive construction in Plautus and Terence’, Transactions of the Philological Society 105:
105–25.
Melo, Wolfgang D. C. de (2010). ‘Possessive pronouns in Plautus’, in Eleanor Dickey and Anna
Chahoud (eds), 71–99.
Melo, Wolfgang D. C. de (2012). ‘Kurylowicz’s first “law of analogy” and the development of
passive periphrases in Latin’, in Philomen Probert and Andreas Willi (eds), 83–101.
Melzani, Graziano (1975). ‘Studio sull’infinito esclamativo in latino’, Aevum 49: 94–9.
Melzani, Graziano (1986). ‘Un aspetto della lingua d’uso nella Naturalis Historia: Interiezioni
ed espressioni esclamative’, in Pier Vincenzo Cova et al. (eds), 201–33.
Mesa, Juan Francisco (1998a). El deseo y el subjuntivo. Alicante: Publicaciones de la Universi-
dad de Alicante.
Mesa, Juan Francisco (1998b). ‘Estudio pragmático de utinam + subjuntivo’, in Benjamín
García Hernández (ed.), 541–54.
Michaelis, Susanne (1998). ‘Antikausativ als Brücke zum Passiv: fieri, venire und se im
Vulgärlateinischen und Altitalienischen’, in Wolfgang Dahmen et al. (eds), Neuere Beschrei-
bungsmethoden der Syntax der romanischen Sprachen. Tübingen: Narr, 69–98.
Milham, Mary Ella (1959). ‘An inventory of the double accusative in Apicius’, Classical
Philology 54: 40–2.
Milham, Mary Ella (1961). ‘Case and prepositional usage in Apicius’, Glotta 39: 276–302.
Miniconi, Pierre-Jean (1951). Causa et ses dérivés. Paris: Les Belles Lettres.
Mohr, Paul (1889). ‘Zum Konjunktiv nach Komparativ mit quam’, Archiv für Lateinische
Lexikographie und Grammatik 6: 418.
Molinelli, Piera (1988). Fenomeni della negazione dal latino all’italiano. Firenze: La Nuova
Italia Editrice.
Molinelli, Piera (1989). ‘Double negation from Latin to Italian’, in Gualtiero Calboli (ed.),
612–33.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Molinelli, Piera (1996). ‘Casi e preposizioni in latino: Lo sviluppo del genitivo e del dativo. Una
prospettiva (anche) sociolinguistica’, Linguistica e Filologia NS 3: 73–125.
Molinelli, Piera (2008). ‘Tra oralità e scrittura: rogo nelle lettere private in latino’, in
Romano Lazzeroni, Emanuele Banfi, Giuliano Bernini, Marina Chini, and Giovanna
Marotta (eds), Diachronica et Synchronica: Studi in onore di Anna Giacalone Ramat. Pisa:
ETS, 365–78.
Molinelli, Piera (2010). ‘From verbs to interactional discourse markers: Latin rogo, quaeso’,
Papers on Grammar 11: 181–92.
Molinelli, Piera, Cuzzolin, Pierluigi, and Fedriani, Chiara (eds) (2014). Latin vulgaire—Latin
tardif X, 3 vols. Bergamo: Bergamo University Press.
Monserrat Roig, Catalina (2010). ‘Los vocativos metafóricos en las comedias de Plauto’,
Cuadernos de Filología Clásica (Lat) 30: 7–26.
Moralejo, José-Luis (1989). ‘Temps absolu et temps relatif en latin’, in Marius Lavency and
Dominique Longrée (eds), 297–308.
Moralejo, José-Luis (1996). ‘Procesos de transitivación: Una approximación estructural’, in
Alfred Bammesberger and Friedrich Heberlein (eds), 141–53.
Moralejo, José-Luis (1998). ‘Problemas de la consecutio temporum’, in Benjamín García
Hernández (ed.), 571–82.
Moralejo, José-Luis (1999). ‘La consecutio temporum: Cuestiones de concepto y límites, in
Empar Espinilla et al. (eds), 133–49.
Morris, Edward Parmelee (1889/90). ‘On the sentence-question in Plautus and Terence’,
American Journal of Philology 10: 397–436, 11: 16–54, 145–81.
Moure Casas, Ana (1998). ‘Sum + genitivo y su herencia en castellano’, in Benjamín García
Hernández (ed.), 195–208.
Moussy, Claude (1966). Gratia et sa famille. Paris: Presses Universitaires Françaises.
Moussy, Claude (1987). ‘Les complétives en quin, quominus et le jeu des négations’, in Sylvie
Mellet (ed.), 279–91 (= 2010: 208–20).
Moussy, Claude (1998). ‘Négation et lexique en latin’, in Benjamín García Hernández (ed.),
583–95 (= 2010: 221–32).
Moussy, Claude (2002). ‘Nequeo est-il un verbe modal?’ in Michèle Fruyt and Claude Moussy
(eds), 133–47.
Moussy, Claude (2010). Synonymie et antonymie en latin. Paris: Harmattan.
Moussy, Claude (2012a). ‘Les “postpositions” servant à l’expression de la cause’. Paper
presented at Colloque biennal du centre Alfred Ernout, Université de Paris-Sorbonne
(Paris IV), 4–6 June.
Moussy, Claude (2012b). ‘Négation et lexique en latin’, Linguarum Varietas 1: 80–90.
Moussy, Claude (2012c). ‘Les emplois de proprius comme substitut de l’adjectif possessif au
cours de la latinité’, in Michèle Biraud (ed.), (Dis)continuité en linguistique latine et
grecque: Hommage à Chantal Kircher-Durand. Paris: L’Harmattan, 208–20.
Moussy, Claude (ed.) (2001). De lingua Latina novae quaestiones: Actes du Xe Colloque
International de Linguistique Latine, Paris-Sèvres, 19–23 avril 1999. Louvain: Peeters.
Moussy, Claude and Mellet, Sylvie (eds) (1992). La validité des catégories attachées au verbe.
Paris: Presses de l’Université de Paris-Sorbonne.
Muhr, Oton (1971). Die Präposition ‘per’ bei Sallust. Vienna: Notring.
Mulder, Hotze (1991). ‘Object continuity and elliptical objects in Latin’, Romance Linguistics
and Literature Review 4: 12–26.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Müller, Carl Friedrich Wilhelm (1895). ‘Zu Caesars Bellum civile’, Festschrift Ludwig Fried-
länder. Leipzig: Hirzel, 543–54.
Müller, Carl Friedrich Wilhelm (1908). Syntax des Nominativs und des Akkusativs. Leipzig:
Teubner.
Müller, Carl Werner, Sier, Kurt, and Werner, Jürgen (eds) (1992). Zum Umgang mit fremden
Sprachen in der Griechisch-Römischen Antike. Stuttgart: Steiner.
Müller, Roman (1997). Sprechen und Sprache. Dialoglinguistische Studien zu Terenz. Heidel-
berg: Winter.
Müller, Roman (2001). Sprachbewusstsein und Sprachvariation im lateinischen Schrifttum der
Antike. Munich: Beck.
Müller, Roman (2012). ‘Sit autem sermo vester est est non non: Klassisches und nichtklas-
sisches “Ja”’, in Frédérique Biville et al. (eds), 111–20.
Müller, Rudolf Wolfgang (1964). ‘Rhetorische und syntaktische Interpunktion: Untersuchun-
gen zur Pausenbezeichnung im antiken Latein’. Dissertation, Universität Tübingen.
Müller-Lancé, Johannes (1994). Absolute Konstruktionen vom Altlatein bis zum Neufranzö-
sischen. Tübingen: Narr.
Müller-Wetzel, Martin (2001). Der lateinische Konjunktiv. Hildesheim: Olms.
Murphy, Paul R. (1954/5). ‘A Terentian pattern of negation’, Classical Weekly 48: 203–5.
Murphy, Paul R. (1956). ‘The origins and survival of a Latin negative pattern’, American
Journal of Philology 77: 396–407.
Murphy, Paul R. (1959). ‘Terence, Andria 560–5: A reply to Professor H. L. Levy’, American
Journal of Philology 80: 306–9.
Murphy, Paul R. (1990). ‘On questions introduced by non and nonne’, Classical Journal 86:
226–32.
Naas, Valérie (2002). Le projet encyclopédique de Pline l’Ancien. Rome: École française.
Nägelsbach, Karl Friedrich von and Müller, Iwan von (1905). Lateinische Stilistik. Nürnberg:
Konrad Geiger (repr. Darmstadt: Wissenschaftliche Buchgesellschaft, 1966).
Napoli, Maria (2009). ‘Impersonal passivization and agentivity in Latin’, in Stephanie
W. Jamison, H. Craig Melchert, and Brent Vine (eds), Proceedings of the 20th Annual
UCLA Indo-European Conference (2008). Bremen: Hempen Verlag, 163–78.
Nausester, Walter (1907). ‘Beiträge zur Lehre vom Deponens und Passivum des Lateinischen’,
in Novae Symbolae Joachimicae. Halle: Verlag der Buchhandlung des Waisenhauses,
135–68.
Nausester, Walter (1935). Hören und Lesen II. Kirchhain: Max Schmersow.
Nelson, Hein L. W. (1981). Überlieferung, Aufbau und Stil von Gai Institutiones. Leiden: Brill.
Nesselrath, Heinz-Günther (1992). Ungeschehenes Geschehen: Beinahe-Episoden im grie-
chischen und lateinischen Epos. Stuttgart: Teubner.
Nicholas, Barry (1964). ‘Videbimus’, in Antonio Guarino and Luigi Labruna (eds), Synteleia
Vincenzo Arangio-Ruiz, 2 vols. Napoli: Jovene, I.150–4.
Nicholas, Barry (1971). ‘Videbimus II’, in Luigi Aru et al. (eds), Studi in onore di Edoardo
Volterra, 4 vols. Milan: Giuffrè, II.577–604.
Nocentini, Alberto (1990). ‘L’uso dei dimostrativi nella Peregrinatio Egeriae e la genesi dell’
articolo romanzo’, in Atti del convegno internazionale sulla Peregrinatio Egeriae. Arezzo:
Accademia Petrarca, 137–58.
Nocentini, Alberto (2005). ‘Du latin aux langues romanes: La contribution de la typologie’, in
Sándor Kiss et al. (eds), 411–18.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Norberg, Dag (1943). Syntaktische Forschungen auf dem Gebiete des Spätlateins und
des frühen Mittellateins. Uppsala and Leipzig: A. B. Lundeqvistka Bokhandeln and
O. Harassowitz.
Norde, Muriel (2009). Degrammaticalization. Oxford: Oxford University Press.
Núñez, Salvador (1991). Semántica de la modalidad en latín. Granada: EUG.
Núñez, Salvador (1995). ‘Materiales para una sociología de la lengua latina. Terencio y los
modificadores de imperativo’, Florentia Iliberritana 6: 347–66.
Núñez, Salvador (2001). ‘Los adverbios modales asertivos latinos y la modalidad evidencial’,
in Claude Moussy (ed.), 511–24.
Núñez, Salvador (2002). ‘Les adverbes modalisateurs profecto et certe: Remarques sur la
subjectivité épistémique en latin’, in Michèle Fruyt and Claude Moussy (eds), 187–99.
Nutting, Herbert C. (1925). The Latin Conditional Sentence. Berkeley: University of California
Press.
Nutting, Herbert C. (1927). ‘Quid hoc homine facias’, American Journal of Philology 48:
10–17.
Nutting, Herbert C. (1928a). ‘The utor fruor group’, University of California Publications in
Classical Philology 10: 1–16, 17–36, 37–62, 63–150, 151–68.
Nutting, Herbert C. (1928b). ‘Quid me fiet’, University of California Studies in Classical
Philology 8: 331ff.
Nutting, Herbert C. (1930). ‘The ablative absolute and the stenographic ablative’, University of
California Publications in Classical Philology 10: 203–17.
Odelman, Eva (1972). ‘Études sur quelques reflets du style administratif chez César’. Disser-
tation, Stockholm University.
Odelstierna, Ingrid (1926). De vi futurali ac finali gerundii et gerundivi Latini observationes:
Accedunt de verbo imputandi adnotationes. Stockholm: Almqvist & Wiksell.
Oldsjö, Fredrik (2001). Tense and Aspect in Caesar’s Narrative, Studia latina Upsaliensia 26.
Uppsala: Acta Universitatis Upsaliensis.
Önnerfors, Alf (1955). ‘Zur Vermischung der Ausdrücke für Ortsruhe und Bewegung im
Lateinischen’, Eranos 53: 148–65.
Önnerfors, Alf (1956). Pliniana: In Plinii Maioris Naturalem Historiam studia grammatica,
semantica, critica. Stockholm: Almqvist & Wiksell.
Önnerfors, Alf (1989). ‘Dare und auris/auricula im Spätlatein’, Symbolae Osloenses 64:
130–57.
Önnerfors, Alf (2002). ‘Über den dativus possessivus mit esse bzw. habere mit Akkusativobjekt
in den Büchern 7–11 der Naturalis Historia des Plinius’, in Pol Defosse (ed.), Hommages
Carl Deroux. Bruxelles: Latomus, II.336–46.
Orlandini, Anna M. (1980). ‘Unius figurae crudelis eventus or On rhetorical questions’, Papers
on Grammar 1: 103–40.
Orlandini, Anna M. (1991). ‘La négation dans la phrase simple en latin’, Bulletin de la Société
de Linguistique de Paris 86/1: 195–210.
Orlandini, Anna M. (1993). ‘Contexte pour la négation explétive: Une analyse pragmatique’,
Bulletin de la Société de Linguistique de Paris 88: 163–85.
Orlandini, Anna M. (1994). ‘Le nom “chose”: Analyse d’un phénomène latin et roman’, in
Gualtiero Calboli (ed.), 165–76.
Orlandini, Anna M. (1995). Il riferimento del nome: Un analisi semantico-pragmatica del nome
latino. Bologna: Clueb.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Penney, John H. W. (1999). ‘Archaism and innovation in poetic syntax’, in James N. Adams
and Roland G. Mayer (eds), 249–68.
Penney, John H. W. (ed.) (2004). Indo-European Perspectives. Oxford: Oxford University
Press.
Pepe, Luigi (1975/6). ‘Lat. homo indefinito’, Helikon 15–16: 408–16.
Perdicoyanni-Paléologou, Hélène (2006). ‘Les emplois de ecce, eccum, eccistum, eccillum chez
Plaute’, Faventia 28: 41–52.
Perl, Gerhard (1978). ‘Doppelte Negation in Pompejanischen Inschriften’, Philologus 112:
111–19.
Perotti, Pier Angelo (1987). ‘À propos des génitifs latins dits du grief et de la peine’, Les Études
Classiques 55: 377–84.
Perotti, Pier Angelo (1989). ‘L’infinito perfetto in Tibullo e nel Corpus Tibullianum’, Orpheus
10: 141–9.
Petersmann, Hubert (1977). Petrons urbane Prosa. Untersuchungen zu Sprache und Text
(Syntax). Vienna: Verlag der Österreichischen Akademie der Wissenschaften.
Petersmann, Hubert (1995). ‘Zur mündlichen Charakterisierung des Fremden in der Komödie
des Plautus’, in Lore Benz, Ekkehard Stärk, and Gregor Voigt-Spira (eds), Plautus und die
Tradition des Stegreifspiels. Tübingen: Narr, 123–36.
Petersmann, Hubert and Kettemann, Rudolf (eds) (1999). Latin vulgaire—Latin tardif V.
Heidelberg: Winter.
Pettersson, Oscar (1930). ‘Commentationes Livianae’. Dissertation, Uppsala University.
Pezzini, Giuseppe (2011). ‘Contraction of est in Latin’, Transactions of the Philological Society
109: 327–43.
Pfister, Raimund (1936). ‘Zum Aspekt der Verba des Sehens bei Plautus’. Dissertation, Uni-
versität München.
Pieroni, Silvia (2000). ‘Agents in Latin impersonal passives’, Mnemosyne 53: 288–301.
Pieroni, Silvia (2001). ‘Subject properties, zero-anaphoras and semantic roles’, in Claude
Moussy (ed.), 539–52.
Pieroni, Silvia (2010a). ‘Deixis and anaphora’, in Philip Baldi and Pierluigi Cuzzolin (eds),
III.389–501.
Pieroni, Silvia (2010b). ‘Nos as expression of the “ego” in Cicero’s letters to Atticus (book
I–IV)’, in Peter Anreiter and Manfred Kienpointner (eds), 599–611.
Pieroni, Silvia (2014). ‘A note on the functional distribution of ille in Late Latin’, Indogerma-
nische Forschungen 119: 1–20.
Pieroni, Silvia (2015). ‘On the agreement pattern Varium et mutabile semper femina’, in
Gerd Haverling (ed.), 351–61.
Pinkster, Harm (1972). On Latin Adverbs. Amsterdam: North-Holland (reprint on demand
from Amsterdam University Press).
Pinkster, Harm (1982). ‘The use of the so-called ablativus qualitatis in the function Praedica-
tivum’, in Jan den Boeft and A. H. M. Kessels (eds), Actus: Studies in Honour of
H. L. W. Nelson. Utrecht: Klassiek Instituut, 247–56.
Pinkster, Harm (1983). ‘Tempus, aspect and Aktionsart in Latin (recent trends 1961–1981)’, in
Hildegard Temporini and Wolfgang Haase (eds), Aufstieg und Niedergang der römischen
Welt, II: Prinzipat, Sprache und Literatur (Sprachen und Schriften) 29.1. Berlin: De Gruyter,
271–319.
Pinkster, Harm (1984). ‘Het Latijnse passief ’, Lampas 17: 420–32.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Pinkster, Harm (1985a). ‘The discourse function of the passive’, in A. Machtelt Bolkestein et al.
(eds), Syntax and Pragmatics in Functional Grammar. Dordrecht: Foris, 107–18.
Pinkster, Harm (1985b). ‘The development of future tense auxiliaries in Latin’, Glotta 63:
186–208.
Pinkster, Harm (1985c). ‘Latin cases and valency grammar: Some problems’, in Christian
Touratier (ed.), 163–89.
Pinkster, Harm (1986a). ‘Three notes on the Latin subjunctive’, Revue, Informatique et
Statistique dans les Sciences Humaines 22: 147–56.
Pinkster, Harm (1986b). ‘Ego, tu, nos. Opmerkingen over het gebruik van subjektpronomina,
in het bijzonder in Cicero de Oratore II’, Lampas 19: 309–22.
Pinkster, Harm (1987a). ‘The strategy and chronology of the development of future and
perfect tense auxiliaries in Latin’, in Martin Harris and Paolo Ramat (eds), The Historical
Development of Auxiliaries. Berlin: De Gruyter, 193–223.
Pinkster, Harm (1987b). ‘The pragmatic motivation for the use of subject pronouns: The case
of Petronius’, in Sylvie Mellet (ed.), 193–223.
Pinkster, Harm (1988). ‘Sicut unus pater familias. Opmerkingen over het gebruik van ille
en unus in verband met de ontwikkeling van de Romaanse talen’, in A. Maria van
Erp Taalman Kip et al. (eds), Propemptikon. Amsterdam: University of Amsterdam,
109–16.
Pinkster, Harm (1989). ‘Some methodological remarks on research on future tense auxiliaries
in Latin’, in Gualtiero Calboli (ed.), 311–26.
Pinkster, Harm (1990). ‘The development of cases and adpositions in Latin’, in Harm Pinkster
and Inge Genee (eds), Unity in Diversity. Dordrecht: Foris, 195–209.
Pinkster, Harm (1991). ‘Evidence for SVO in Latin?’, in Roger Wright (ed.), Latin and the
Romance Languages in the Early Middle Ages. London: Routledge, 69–82.
Pinkster, Harm (1992a). ‘The Latin impersonal passive’, Mnemosyne 45: 159–77.
Pinkster, Harm (1992b). ‘Notes on the syntax of Celsus’, Mnemosyne 45: 513–24.
Pinkster, Harm (1996). ‘Review article of Selig, Maria (1992)’, Mittellateinisches Jahrbuch
31: 143–49.
Pinkster, Harm (1998a). ‘Is the Latin present the unmarked, neutral tense in the system?’, in
Rodie Risselada (ed.), 63–83.
Pinkster, Harm (1998b). ‘Narrative tenses in Merovingian hagiographic texts’, in József
Herman (ed.), La transizione dal latino alle lingue romanze. Tübingen: Niemeyer,
229–35.
Pinkster, Harm (1999). ‘The present tense in Virgil’s Aeneid’, Mnemosyne 52: 705–17.
Pinkster, Harm (2004). ‘Attitudinal and illocutionary satellites in Latin’, in Henk Aertsen,
Mike Hannay, and Rod Lyall (eds), Words in Their Places: A Festschrift for J. Lachlan
Mackenzie. Amsterdam: Vrije Universiteit Amsterdam, 191–8.
Pinkster, Harm (2005a). ‘Non certe in Cicero de Divinatione 2.94’, Mnemosyne 58: 268–9.
Pinkster, Harm (2005b). ‘The language of Pliny the Elder’, in Tobias Reinhardt et al. (eds),
239–56.
Pinkster, Harm (2005c). ‘The use of is and ille in Seneca Rhetor’, in Sándor Kiss et al. (eds),
57–64.
Pinkster, Harm (2010). ‘Les adverbes de fréquence en latin’, in Michèle Fruyt and Olga Spevak
(eds), 83–192.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Pinkster, Harm (2011). ‘The use of the dative with Latin compounds’, STUF—Language
Typology and Universals / Sprachtypologie und Universalienforschung 63: 32–43.
Pinkster, Harm (ed.) (1983). Latin Linguistics and Linguistic Theory. Amsterdam: Benjamins.
Pitkäranta, Reijo (1978). Studien zum Latein des Victor Vitensis. Helsinki: Societas Scientiarum
Fennica.
Poirier, Michel (1989). ‘Le fonctionnement du réfléchi latin, témoignage sur la pertinence
linguistique de l’opposition sujet / prédicat?’, in Marius Lavency and Dominique Longrée
(eds), 345–54.
Porzio Gernia, Maria Luisa (1974). ‘Contributi metodologici allo studio del latino arcaico’, in
Memorie della Accademia Nazionale dei Lincei, Cl. di Sc. morali, storici e filologiche, Ser.
VIII, 17/4: 113–337.
Posner, Rebecca (1996). The Romance Languages. Cambridge: Cambridge University Press.
Prat, Louis C. (1975). Morphosyntaxe de l’ablatif en latin archaïque. Paris: Les Belles Lettres.
Probert, Philomen (2002). ‘On the prosody of Latin enclitics’, Oxford University Working
Papers in Linguistics, Philology & Phonetics 7: 181–206.
Probert, Philomen and Willi, Andreas (eds) (2012). Laws and Rules in Indo-European. Oxford:
Oxford University Press.
Pruvost-Versteeg, Géraldine (2015). ‘Dativus sympatheticus et tour possessif chez Plaute:
Concurrence ou complémentarité?’, in Gerd Haverling (ed.), 264–75.
Pultrová, Lucie (2011). The Latin Deverbative Nouns and Adjectives, Acta Universitatis
Carolinae: Philologica monographia 162. Prague: Éditions Karolinum.
Purnelle, Gérald and Denooz, Joseph (eds) (2007). Ordre et cohérence en latin, Bibliothèque de
la Faculté de philosophie et Lettres de l’Université de Liège 293. Genève: Droz.
Quetglas, Pere (1979). ‘Anàlisi de l’exclamació’, Faventia 1: 181–94.
Quirk, Randolph, Greenbaum, Sidney, Leech, Geoffrey, and Svartvik, Jan (1985). A Compre-
hensive Grammar of the English Language. London: Longman.
Raible, Wolfgang (1992). ‘The pitfalls of subordination’, in Bela Brogyanyi and Reiner Lipp
(eds), Historical Philology. Amsterdam: Benjamins, 299–337.
Raiskila, Pirjo (1990). ‘Periphrastic use of habere in Tertullian’, in Gualtiero Calboli (ed.), 209–17.
Ramat, Anna Giacalone and Sansò, Andrea (2010). ‘L’emploi indéfini de homo en latin tardif:
Aux origines d’un “européanisme” ’, in Michèle Fruyt and Olga Spevak (eds), 93–115.
Ramat, Paolo (1999). ‘Negation’, in Keith Brown and Jim Miller (eds), Concise Encyclopedia of
Grammatical Categories. Amsterdam: Elsevier, 241–8.
Ramos, Agustín (1998). Consideraciones sobre la expresión de la posesión externa en latín’, in
Benjamín García Hernández (ed), 637–88.
Ramos, Agustín (2007a). ‘La complementación de dicere. Sobre causatividad, incorporación de
instrumento y metalenguaje’, in Esperanza Torrego et al. (eds), 151–78.
Ramos, Agustín (2007b). ‘Ante diem V Kalendas Sextilis. Problemas sintácticos de una fórmula
de datación’, in Gregorio Hinojo and José Carlos Fernández Corte (eds), Munus quaesitum
meritis: Homenaje a Carmen Codoñer. Salamanca: Ediciones Universidad de Salamanca,
747–56.
Raviolo, Francesca (2002). ‘Elementi di latino volgare nei sermoni 1–5 di Agostino’, Orpheus
23: 90–104.
Reichenkron, Günther (1957). ‘Die Umschreibung mit occipere, incipere und coepisse als
analytische Ausdrucksweise eines ingressiven Aorists’, in Günther Reichenkron et al.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
(eds), Syntactica und Stilistica: Festschrift für Ernst Gamillscheg zum 70. Geburtstag. Tübin-
gen: Niemeyer, 451–80.
Reinhardt, Tobias (2010). ‘Syntactic colloquialism in Lucretius’, in Eleanor Dickey and Anna
Chahoud (eds), 203–28.
Reinhardt, Tobias, Lapidge, Michael, and Adams, James N. (eds) (2005). Aspects of the
Language of Latin Prose, Proceedings of the British Academy. Oxford: Oxford University
Press.
Reinhold, Heinz (1956). ‘Zum lateinischen Verbalaspekt’, Zeitschrift für Vergleichende Sprach-
forschung 74: 1–44.
Renzi, Lorenzo (2000). ‘La storia di IPSE’, in József Herman and Anna Marinetti (eds), La
preistoria dell’italiano. Tübingen: Niemeyer, 181–203.
Revuelta, Antonio (2015). ‘On com compounds’, in Gerd Haverling (ed.), 158–69.
Ricca, Davide (2010). ‘Adverbs’, in Philip Baldi and Pierluigi Cuzzolin (eds), II.109–91.
Riemann, Othon (1885). Études sur la langue et la grammaire de Tite-Live. Paris: Thorin.
Ripoll, Arthur (2007). ‘Les emplois des formations adverbiales sur thèmes d’adjectifs en latin’.
Dissertation, Paris X Nanterre.
Ripoll, Arthur (2010a). ‘Le syntagme nominal composé d’un substantif et d’un adverbe en
latin’, in Olga Spevak (ed.), 139–62.
Ripoll, Arthur (2010b). ‘À propos de la quantification en latin’, in Peter Anreiter and Manfred
Kienpointner (eds), 301–12.
Ripoll, Arthur (2013). ‘Le changement invisible ou “que tout reste identique pour que tout
change” ’, in Alain Christol and Olga Spevak (eds), Les évolutions du latin. Paris: L’Harmat-
tan, 299–323.
Ripoll, Arthur (2014). ‘Scin ut dicam?: L’interrogatif latin ut’, Latomus 73: 60–79.
Ripoll, Arthur (forthc.). ‘Latine scire, fidibus scire, litteras scire: À propos de certaines construc-
tions du verbe scire’, in Muriel Lenoble et al. (eds), De linguae Latinae usu. Louvain: Peeters.
Risch, Ernst (1984). Gerundivum und Gerundium. Berlin: De Gruyter.
Risselada, Rodie (1984). ‘Coordination and juxtaposition of adjectives in the Latin NP’, Glotta
62: 202–31.
Risselada, Rodie (1989). ‘Latin illocutionary parentheticals’, in Marius Lavency and Domi-
nique Longrée (eds), 367–78.
Risselada, Rodie (1993). Imperatives and Other Directive Expressions in Latin: A Study in the
Pragmatics of a Dead Language. Amsterdam: Gieben.
Risselada, Rodie (1994). ‘Modo and sane, or what to do with particles in Latin directives’, in
József Herman (ed.), 319–42.
Risselada, Rodie (1998a). ‘The discourse functions of sane: Latin marker of agreement in
description, interaction and concession’, Journal of Pragmatics 30: 225–44.
Risselada, Rodie (1998b). ‘Tandem and postremo: Two of a kind?’, in Rodie Risselada (ed.),
85–116.
Risselada, Rodie (2000). ‘The end of time: A perfect future in Latin?’, in Caroline Kroon and
Daniel den Hengst (eds), Ultima Aetas. Amsterdam: VU University Press, 25–39.
Risselada, Rodie (2004). ‘Actes de langage et rapports sociaux: les directifs dans Plaute Asinaria
649–732’, in Antonio López Eire and Agustín Ramos (eds), 268–80.
Risselada, Rodie (2005). ‘Particles in questions’, in Gualtiero Calboli (ed.), 663–79.
Risselada, Rodie (ed.) (1998). Latin in Use: Amsterdam Studies in the Pragmatics of Latin.
Amsterdam: Gieben.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Risselada, Rodie, Jong, Jan R. de, and Bolkestein, A. Machtelt (eds) (1996). On Latin: Linguistic
and Literary Studies in Honour of Harm Pinkster. Amsterdam: Gieben.
Roberts, W. Rhys (1912). ‘A point of Greek and Roman word order’, Classical Review
26: 177–9.
Roby, Henry J. (1882). A Grammar of the Latin Language from Plautus to Suetonius, 2 vols.
London: Macmillan.
Roca, María José (1999). ‘La posterioridad en las completivas conjuncionales del latín tardío’,
in Empar Espinilla et al. (eds), 155–70.
Roca, María José (2003). ‘El marco predicativo de accuso en latín tardío’, in José Miguel Baños
et al. (eds), 167–83.
Rochette, Bruno (2007). Éléments hors structure dans la Cena Trimalchionis
(26.7–78.8). Remarques sur la structure de la phrase de Pétrone’, in Colette Bodelot
(ed.), 265–94.
Roesch, Sophie (2001). ‘Les emplois de verbum et sermo dans les expressions à verbes
support verba facere, verba habere et sermonem habere’, in Claude Moussy (ed.),
859–74.
Roesch, Sophie (2005). ‘L’échec des clôtures du dialogue dans les comédies de Plaute’, in
Gualtiero Calboli (ed.), 921–32.
Romero, María Isabel (1996). ‘The partitive reading of summus mons’, in Hannah Rosén (ed.),
361–76.
Ronconi, Alessandro (1959). Il verbo latino. Firenze: Le Monnier.
Ros, Hilke (2005). ‘Gratias agere quod: Causal or complement clause’, Mnemosyne 58: 416–23.
Rosellini, Michela (2005). ‘Giulio Valerio e il ‘futuro del congiuntivo’, Rivista di Filologia e di
Istruzione Classica 133: 452–62.
Rosellini, Michela (2009). ‘Prisciano e il futuro del congiuntivo’, Philologus 153: 300–9.
Rosellini, Michela (2010). ‘Le costruzioni verbali nel l. XVIII di Prisciano’, Materiali e
Discussioni 32: 69–94.
Rosén, Haiim B. (1998). ‘Quam quisque norit artem, in hac se exerceat and the typology of
relative clauses’, in Benjamín García Hernández (ed.), 705–21.
Rosén, Haiim B. and Rosén, Hannah (eds) (1980). On Moods and Tenses of the Latin Verb.
Munich: Fink.
Rosén, Hannah (1981). Studies in the Syntax of the Verbal Noun in Early Latin. Munich: Fink.
Rosén, Hannah (1983). ‘The mechanisms of Latin nominalization and conceptualization in
historical view’, in Hildegard Temporini and Wolfgang Haase (eds), Aufstieg und Nieder-
gang der römischen Welt II: Principat, Sprache und Literatur (Sprachen und Schriften) 29.1.
Berlin: De Gruyter, 178–211.
Rosén, Hannah (1986). ‘One instance of the partitive subject in literary Latin’, Glotta 64:
83–90.
Rosén, Hannah (1992). ‘Die Arten der Prolepse im Lateinischen in typologischer Sicht’, in
Oswald Panagl and Thomas Krisch (eds), 243–62.
Rosén, Hannah (1994). ‘The definite article in the making’, in József Herman (ed.), 129–50.
Rosén, Hannah (1995). ‘The Latin infinitivus historicus revisited’, Mnemosyne 48: 536–64.
Rosén, Hannah (1996a). ‘Grammatical equivalence and choice in Ancient Latin translations’,
in Alfred Bammesberger and Friedrich Heberlein (eds), 533–50.
Rosén, Hannah (1996b). ‘Eam vitam vivere quae est sola vita nominanda: Reflections on
cognate complements’, in Rodie Risselada et al. (eds), 127–49.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Rosén, Hannah (1998). ‘Latin presentational sentences’, in Benjamín García Hernández (ed.),
723–42.
Rosén, Hannah (1999). Latine loqui. Trends and Directions in the Crystalization of Classical
Latin. Munich: Fink.
Rosén, Hannah (2002a). ‘Equating and comparing: Remarks on apposition, adjective gra-
dation, and equation sentences in Latin’, Papers on Grammar 8: 209–28.
Rosén, Hannah (2002b). ‘Complementarity within the pool of cohesion devices?’, in
A. Machtelt Bolkestein et al. (eds), 333–46.
Rosén, Hannah (2003). ‘Immo: Its atypical use in Petronius’, in József Herman and Hannah
Rosén (eds), 169–81.
Rosén Hannah (2005). ‘Particles: Hypercharacterization and status shift in Latin and
Romance’, in Sándor Kiss et al. (eds), 227–36.
Rosén, Hannah (2009). ‘Coherence, sentence modification, and sentence-part modifica-
tion: The contribution of particles’, in Philip Baldi and Pierluigi Cuzzolin (eds),
I.317–441.
Rosén, Hannah (2012a). ‘The Late Latin coepi + infinitive construction: Evidence from
translated texts’, Classica & Medievalia 63: 189–215.
Rosén, Hannah (2012b). ‘Coepi + infinitif dans une sélection de traductions en latin’, in
Frédérique Biville et al. (eds), 365–75.
Rosén, Hannah (2012c). ‘Two phasal verbs: Lat. Coepi and Gk. ƺ/ƺÆ, ººø/Çø/
ø’, Indogermanische Forschungen 117: 119–72.
Rosén, Hannah (2013). ‘About non-direct discourse: Another look at its parameters in Latin’,
Journal of Latin Linguistics 12: 231–68.
Rosén, Hannah (2015). ‘Dativus ethicus’, in Gerd Haverling (ed.), 240–63.
Rosén, Hannah (ed.) (1996). Aspects of Latin: Papers from the Seventh International Collo-
quium on Latin Linguistics. Innsbruck: Institut für Sprachwissenschaft.
Rosumek, Peter and Najock, Dietmar (1996). Concordantia in C. Plinii Secundi Naturalem
Historiam. Hildesheim: Olms-Weidmann.
Rubio, Gonzalo (2009). ‘Semitic influence in the history of Latin syntax’, in Philip Baldi and
Pierluigi Cuzzolin (eds), I.195–239.
Ruwet, Nicolas (1989). ‘Weather verbs and the unaccusative hypothesis’, in Carl Kirschner
and Janet Decasaris (eds), Studies in Romance Linguistics. Amsterdam: Benjamins,
313–54.
Sánchez Lafuente, Ángela and García de la Calera, Roberto (1997). ‘Acerca de la consecutio
temporum en las proposiciones consecutivas’, Myrtia 12: 47–61.
Sánchez Manzano, María Asunción (2002). ‘Définition de la modalité impressive en latin’, in
Michèle Fruyt and Claude Moussy (eds), 111–20.
Sánchez Manzano, María Asunción (2008). ‘Censere y arbitrari en textos dialogados’, in
Ghislaine Viré (ed.), 180–93.
Sánchez Martínez, Félix (1996). ‘La construcción personal activa: A propósito de Propercio
2.1.41–2’, Cuadernos de Filología Clásica (Lat) 11: 17–27.
Sánchez Martínez, Jesús (2000). Morfosintaxis latina coordinative. Murcia: Universidad de
Murcia.
Santos, Patricia (2009). ‘Retórica y pragmática: La expresión del complemento agente en pro
Milone’, in Trinidad Arcos Pereira, Jorge Fernández López, and Francisca Moya del Baño
(eds), 231–46.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Sas, Louis Furman (1937). The Noun Declension System in Merovingian Latin. Paris: André.
Sasse, Josephus (1889). ‘De numero plurali qui vocatur maiestatis’. Dissertation, Leipzig
University.
Sawicki, Lea and Shalev, Donna (2002). ‘Perfects and exclamatives’, in Lea Sawicki and Donna
Shalev (eds), 299–305.
Sawicki, Lea and Shalev, Donna (eds) (2002). Donum Grammaticum. Leuven: Peeters.
Schafroth, Elmar (1993). Zur Entstehung und vergleichenden Typologie der Relativpronomina
in den romanischen Sprachen. Tübingen: Niemeyer.
Schenk, Richard (1892). De dativi possessivi usu Ciceroniano. Programm Bergedorf; Jena:
Fromannsche Buchdruckerei.
Scherer, Anton (1975). Handbuch der lateinischen Syntax. Heidelberg: Winter.
Schlicher, John J. (1914a). ‘The historical infinitive I: Its simple form (Infinitivus impotentiae)’,
Classical Philology 9: 279–94.
Schlicher, John J. (1914b). ‘The historical infinitive II: Its literary elaboration’, Classical
Philology 9: 374–94.
Schlicher, John J. (1915). ‘The historical infinitive III: Imitation and decline’, Classical Phil-
ology 10: 54–74.
Schlicher, John J. (1931). ‘Historical tenses and their functions in Latin’, Classical Philology
26: 46–59.
Schmid, Josef (1988). Untersuchungen zum sogenannten freien Dativ in der Gegenwartssprache
und auf Vorstufen des heutigen Deutsch. Frankfurt am Main: Peter Lang.
Schmid, Josef (2006). ‘Der Dativ’, in Vilmos Ágel et al., 951–65.
Schøsler, Lene (2001). ‘The coding of the subject–object distinction from Latin to Modern
French’, in Jan Terje Faarlund (ed.), Grammatical Relations in Change. Amsterdam: Benja-
mins, 273–302.
Schøsler, Lene (2003). ‘Grammaticalisation of valency patterns? An investigation into valency
patterns and support verb constructions, based on diachronic corpora’, Forum for Modern
Language Studies 39: 400–13.
Schøsler, Lene (2008). ‘Argument marking from Latin to modern Romance languages’, in
Thórhallur Eythórsson (ed.), Grammatical Change and Linguistic Theory. Amsterdam:
Benjamins, 411–38.
Schøsler, Lene (2012). ‘The development of the declension system’, in Deborah L. Arteaga
(ed.), Research on Old French: The State of the Art. Berlin: Springer, 167–86.
Schøsler, Lene (forthc.). ‘How useful is case morphology: From Latin to French’.
Schrickx, Josine (2011). Lateinische Modalpartikeln: ‘Nempe’, ‘quippe’, ‘scilicet’, ‘videlicet’ und
‘nimirum’. Leiden: Brill.
Schunck, Julius (1900). Quantum intersit inter dativi possessivi usum Ciceronis et Plauti.
Programm Zweibrücken; Zweibrücken: Kranzbühler.
Schwartz, Francis Robert (2002). Lucans Tempusgebrauch. Textsyntax und Erzählkunst.
Frankfurt am Main: Lang.
Seefranz-Montag, Ariane von (1983). Syntaktische Funktionen und Wortstellungsveränderung:
Die Entwicklung ‘subjektloser’ Konstruktionen in einigen Sprachen. Munich: Fink.
Seldeslachts, Herman (2001). Études de morphologique historique du verbe latin et indo-
européen. Leuven: Peeters.
Selig, Maria (1991). ‘Inhaltskonturen des “Dativs”: Zur Ablösung des lateinischen Dativs durch
ad und zur differentiellen Objektmarkierung’, in Peter Koch and Thomas Krefeld (eds),
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Solin, Heikki (2013). ‘Zu pompejanischen Wandinschriften’, in Werner Eck, Bence Fehér, and
Péter Kovács (eds), Studia epigraphica in memoriam Géza Alföldi. Bonn: R. Habelt, 327–50.
Solin, Heikki, Leiwo, Martti, and Halla-aho, Hilla (eds) (2003). Latin vulgaire—Latin tardif VI.
Hildesheim: Olms-Weidmann.
Sonnenschein, Edward A. (1910). The Unity of the Latin Subjunctive: A Quest. London: Murray.
Sörbom, Gunnar (1935). Variatio sermonis Tacitei aliaeque apud eundem quaestiones selectae.
Stockholm: Almqvist & Wiksell.
Sornicola, Rosanna (1997). ‘L’oggetto preposizionale in siciliano antico e in napoletano antico.
Considerazioni su un problema di tipologia diacronica’, Italienische Studien 18: 66–80.
Sornicola, Rosanna (2002). ‘Diacronia e pancronia nella sintassi e semantica dei pronomi
soggetto tra latino e lingue romanze’, in Valeria Viparelli (ed.), 29–66.
Sornicola, Rosanna (2005). ‘Sintassi e stilistica dei pronomi soggetto nel Saint-Alexis’, in
Sándor Kiss et al. (eds), 529–46.
Sornicola, Rosanna (2012). ‘Multifunzionalità di ipse in latino e polimorfismo degli esiti
romanzi’, in Frédérique Biville et al. (eds), 347–62.
Speidel, Michael Alexander (1992). ‘Roman army pay scales’, Journal of Roman Studies 82: 87–106.
Speidel, Michael Alexander (1996). Die römischen Schreibtafeln von Vindonissa. Windisch: Pro
Vindonissa.
Spevak, Olga (2002). ‘La concessive en quamquam, quamvis et licet: Subjonctive ou indicative’,
in A. Machtelt Bolkestein et al. (eds), 347–59.
Spevak, Olga (2005a). ‘Itinerarium Egeriae: L’ordre des constituants obligatoires’, Mnemosyne,
58: 235–61.
Spevak, Olga (2005b). La concession en latin. Bruxelles: Latomus.
Spevak, Olga (2006). ‘Les enclitiques en latin’, Indogermanische Forschungen 111: 249–74.
Spevak, Olga (2010a). Constituent Order in Classical Latin Prose. Amsterdam: Benjamins.
Spevak, Olga (2010b). ‘Le temps et l’espace dans la correspondance de Cicéron. Le concept de
“cadre” et de “spécification” ’, Revue de Linguistique Latine du Centre Ernout: De Lingua
Latina 4: 68–83.
Spevak, Olga (2014a). The Noun Phrase in Classical Latin Prose. Leiden: Brill.
Spevak, Olga (2014b). ‘Noun valency in Latin’, in Olga Spevak (ed.), 183–210.
Spevak, Olga (forthc.). ‘A propos de l’aspect verbal en grec, en slave et en latin’, Bulletin de La
Société Linguistique de Paris.
Spevak, Olga (ed.) (2010). Le syntagme nominal en latin: Nouvelles contributions. Paris:
L’Harmattan.
Spevak, Olga (ed.) (2014). Noun Valency. Amsterdam: Benjamins.
Stassen, Leon (1997). Intransitive Predication. Oxford: Oxford University Press.
Stassen, Leon (2009). Predicative Possession. Oxford: Oxford University Press.
Steele, R. B. (1921). ‘Ablative of the efficient’, Classical Philology 16: 354–61.
Stegmann, Carl (1885). ‘Zur lateinischen Schulgrammatik I’, Neue Jahrbücher für Philologie
und Pädagogik 132: 225–49.
Stegmann, Carl (1887). ‘Zur lateinischen Schulgrammatik II’, Neue Jahrbücher für Philologie
und Pädagogik 136: 252–69.
Stempel, Reinhard (1997). ‘Das lateinische Gerundium und Gerundivum in historischer und
typologischer Perspektive’, Glotta 72: 235–52.
Stempel, Reinhard (2002). ‘Movetur—se movet: On the development of the middle voice in
Latin’, in Lea Sawicky and Donna Shalev (eds), 329–36.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Szantyr, Anton (1970). ‘Bemerkungen zum Aufbau der vergilischen Ekphrasis’, Museum
Helveticum 27: 28–40.
Szantyr, Anton (1974). ‘Zu Sall. (?) Rep. 2.13.6’, Gymnasium 81: 41–8.
Sznajder, Lyliane (1981). ‘Y a-t-il “un” réfléchi en latin? Étude sur les conditions d’emploi de se
et suus’, L’Information Grammaticale 10: 17–22.
Sznajder, Lyliane (1987). ‘Un procédé d’art chez César: Les complétives paratactiques’, in
Sylvie Mellet (ed.), 293–306.
Sznajder, Lyliane (1996). ‘Sur la concurrence entre présents historiques et infinitifs de narra-
tion chez les historiens latins’, in Hannah Rosén (ed.), 307–22.
Sznajder, Lyliane (1998). ‘Verbes transitifs sans objet en latin’, in Benjamín García Hernández
(ed.), 791–808.
Sznajder, Lyliane (2001). ‘Discours indirect et dépendance syntaxique’, in Claude Moussy
(ed.), 609–26.
Sznajder, Lyliane (2003). ‘Complétives au subjonctif sans conjoncteur’, in Colette Bodelot
(ed.), 13–95.
Sznajder, Lyliane (2012a). ‘Dixit autem serpens ad mulierem / mulier quoque dixit: La double
expression de l’allocutaire dans les propositions introductrices de discours direct dans la
Vulgate’, in Frédérique Biville et al. (eds), 271–88.
Sznajder, Lyliane (2012b). ‘Considérations sur la prolepse en hébreu biblique et son traitement
dans la traduction biblique latine’, parts I–II, Revue de Linguistique Latine du Centre Ernout.
De Lingua Latina 7.
Târa, George Bogdan (2012). ‘Observations sur l’ambiguïté de la tournure habeo + participe
parfait passif chez Grégoire de Tours’, in Frédérique Biville et al. (eds), 377–87.
Târa, George Bogdan (2014). Les périphrases verbales avec habeo en latin tardif. Paris:
Harmattan.
Tarriño, Eusebia (2007). ‘Sobre la construcción de suadeo y persuadeo’, in Esperanza Torrego
et al. (eds), 211–32.
Ternes, Charles-Marie (1998). Oratio soluta—Oratio numerosa, Études luxembourgeoises
d’histoire et de littérature romaines, I. Luxembourg: Centre Alexandre-Wiltheim.
Tesařová-Nováková, Drahomíra (1987). ‘Ein Beitrag zur Geltendmachung des Agens in
lateinischen Passivkonstruktionen’, Listy Filologické 110: 94–6.
Tesařová-Nováková, Drahomíra (1988). ‘Die Kategorie des Genus verbi und der Deagentität
im Lateinischen’, Listy Filologické 111: 145–54.
Tesnière, Lucien (1959). Éléments de syntaxe structurale. Paris: Klincksieck.
Théoret, Michel (1982). Les discours de Cicéron. La concurrence du tour casuel et du tour
prépositionnel. Montreal: Presses de l'Université de Montréal.
Thesleff, Holger (1960). Yes and No in Plautus and Terence. Helsingfors: Akademiska
Bokhandeln.
Thielmann, Philip (1885a). ‘Habere mit dem Infinitiv und die Entstehung des romanischen
Futurums’, Archiv für Lateinische Lexikographie und Grammatik 2: 48–89, 157–202.
Thielmann, Philip (1885b). ‘Habere mit dem Part. Perf. Pass.’, Archiv für Lateinische Lexiko-
graphie und Grammatik 2: 372–423, 509–49.
Thomas, François (1938). Recherches sur le subjonctif latin: Histoire et valeur des forms. Paris:
Klincksieck.
Thomas, François (1956). ‘Dominance et résistance dans la syntaxe latine’, in Hommages à
Max Niedermann. Bruxelles: Latomus, 315–23.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Tillmann, Heinrich (1881). ‘De dativo verbis passivis linguae Latinae subiecto qui vocatur
Graecus’, Acta Seminarii Philologici Erlangensis 2: 71–140.
Torrego, Esperanza (1987). ‘Las construcciones pasivas de nominativo con infinitivo en latín
clásico. Estudio sintáctico’, Emerita 55: 71–84.
Torrego, Esperanza (1988). ‘Restricciones de la categoría verbal aspecto sobre los elementos de
duración en latín’, Minerva 2: 259–78.
Torrego, Esperanza (1989a). ‘Distribución de los adverbios de frecuencia en latín’, Minerva 3:
231–8.
Torrego, Esperanza (1989b). ‘Les notions temporelles “temps dans lequel”, “temps depuis
que”, “temps jusqu’à ce que”, et durée: Valeur fonctionnelle’, in Marius Lavency and
Dominique Longrée (eds), 423–34.
Torrego, Esperanza (1991). ‘Genitive with verbal nouns in Latin’, in Robert G. Coleman (ed.),
281–93.
Torrego, Esperanza (1994a). ‘Sobre el imperfecto y perfecto latinos: ¿tiempo o aspecto?’, Actas
del VIII Congreso Español de Estudios Clásicos I, Madrid, Ediciones Clásicas, 775–80.
Torrego, Esperanza (1994b). ‘The syntax of adhuc with verbal predicates’, in József Herman
(ed.), 345–55.
Torrego, Esperanza (1996a). ‘Principios y métodos para una descripción funcional de la
sintaxis de los casos en latín’, in Ana Agud et al. (eds), 199–216.
Torrego, Esperanza (1996b). ‘Conditions syntaxiques pour la formation des noms d’agent en
latin’, in Hannah Rosén (ed.), 181–92.
Torrego, Esperanza (1998). ‘Praeter + accusativo: Descripción funcional y semántica’, in
Esperanza Torrego (ed.), 111–45.
Torrego, Esperanza (1999a). ‘Algunas reflexiones sobre la consecutio temporum’, in Empar
Espinilla et al. (eds), 189–215.
Torrego, Esperanza (1999b). ‘Cohésion et rupture dans les propositions irréelles latines’, Les
Études Classiques 67: 391–411.
Torrego, Esperanza (2002a). ‘Los SN comparativos: El segundo término de la comparación’, in
Empar Espinilla et al. (eds), 251–79.
Torrego, Esperanza (2002b). ‘Statim, paulatim, repente/subito: Scope and lexical aspect’, in
A. Machtelt Bolkestein et al. (eds), 379–92.
Torrego, Esperanza (2003). ‘La estructura de complementación del latino aspicio (“mirar”)’, in
Baños et al. (eds), 185–214.
Torrego, Esperanza (2005). ‘Grammar and pragmatics: The textual uses of repente and subito’,
in Gualtiero Calboli (ed.), 763–73.
Torrego, Esperanza (2007). ‘Esquemas de complementación de los verbos “ver” en latín’, in
Esperanza Torrego et al. (eds), 233–64.
Torrego, Esperanza (2008). ‘Analisis funcional de los elementos que espresan la medida del
espacio en latín’, Papers on Grammar 10: 277–99.
Torrego, Esperanza (2009). ‘Coordination’, in Philip Baldi and Pierluigi Cuzzolin (eds), I.443–87.
Torrego, Esperanza (2010). ‘Les ablatifs de temps en latin: Facteurs pour leur interprétation’,
Revue de Linguistique Latine du Centre Ernout. De Lingua Latina 4: 1–16.
Torrego, Esperanza (2014a). ‘Genetivus and ablativus pretii in Latin: Concepts, functions and
formal marks’, Journal of Latin Linguistics 13: 145–61.
Torrego, M. Esperanza (2014b). ‘Esquemas de complementación del verbo disto’, in Concep-
ción Cabrillana and Christian Lehmann (eds), 437–52.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Torrego, Esperanza (ed.) (1998). Nombres y funciones. Estudios de sintaxis griega y latina.
Madrid: Ediciones Clásicas.
Torrego, Esperanza, Baños, José Miguel, Cabrillana, Concepción, Méndez Dosuna, Julián (eds)
(2007). Praedicativa II: Esquemas de complementación verbal en griego antiguo y en latín.
Monografías de Filologia Latina 15. Zaragoza: Universidad de Zaragoza.
Torrego, Esperanza, Quetglas, Pere J., and Espinilla, Empar (eds) (1995). Sintaxis del dativo
latino. Madrid and Barcelona: Universidad Autónoma de Madrid / Universitat de
Barcelona.
Torrego, Esperanza and Villa, Jesus de la (2009). ‘De dónde y en dónde: Cruces espaciales en
latín y griego’, in Antonio Cascón et al. (eds), Estudios en homenaje al Professor Vicente
Picón. Madrid: Ediciones UAM, 187–204.
Touratier, Christian (1977). ‘Valeurs et fonctionnement du subjonctif latin I’, Revue des Études
Latines 55: 370–406.
Touratier, Christian (1980a). ‘L’accusatif proleptique en latin’, Lalies 1: 53–6.
Touratier, Christian (1980b). La relative. Essai de théorie syntaxique. Paris: Klincksieck.
Touratier, Christian (1991). ‘Adjectif et syntagme nominal’, in Robert G. Coleman (ed.),
233–49.
Touratier, Christian (1994). Syntaxe latine. Louvain-la-Neuve: Peeters.
Touratier, Christian (1995). ‘Le cas est-il une unité fonctionnelle?: Le génitif latin’, in Dom-
inique Longrée (ed.), 307–28.
Touratier, Christian (1996). ‘Troisième personne de singulier: Analyse morphématique et
analyse syntactique’, in Hannah Rosén (ed.), 333–42.
Touratier, Christian (1998). ‘L’imparfait, temps anaphorique?’, in Bruno Bureau and Christian
Nicolas (eds), 265–79.
Touratier, Christian (2007). ‘La négation: Essai de définition et portée’, in Christian Touratier
and Charles Zaremba (eds), 11–32.
Touratier, Christian (2010). ‘Qu’est ce qu’un SN dans une langue sans article comme le latin?’,
in Olga Spevak (ed.), 121–37.
Touratier, Christian (ed.) (1985). Syntaxe et latin: Actes du IIème congrès international de
linguistique latine, Aix-en-Provence, 28–31 mars 1983, Aix-en-Provence, Université de
Provence.
Touratier, Christian and Zaremba, Charles (eds) (2007). La négation. Aix-en-Provence: Pu-
blications de l’Université de Provence.
Traina, Alfonso (1963). ‘Malo me paenitere’, Latinitas 11: 279–80.
Traina, Alfonso (1981). ‘Seneca, Thyest. 713 Mactet sibi o sibi dubitat? Un recupero esegetico’,
Maia 33: 151–3.
Traina, Alfonso (1999). Forma e suono: Da Plauto a Pascoli. Bologna: Pàtron.
Unceta, Luis (2008). ‘Incidencia de factores pragmáticos en la evolución semántica del verbo
rogare’, in Ghislaine Viré (ed.), 244–55.
Unceta, Luis (2009). La petición verbal en latín: Estudio léxico, semántico y pragmático.
Madrid: Ediciones Clásicas.
Unceta, Luis (2010). ‘La expresión del agradecimiento en la comedia latina’, in Peter Anreiter
and Manfred Kienpointner (eds), 625–36.
Unceta, Luis (2014). ‘La politesse linguistique en latin. Bilan d’une recherche en cours’,
Dictionnaire Historique et Encyclopédie Linguistique du Latin, Partie encyclopédie linguis-
tique, <http://www.linglat.paris-sorbonne.fr>.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Vaan, Michiel de (2008). Etymological Dictionary of Latin and the Other Italic Languages.
Leiden: Brill.
Väänänen, Veikko (19663). Le latin vulgaire des inscriptions pompéiennes. Berlin: Akademie-
Verlag.
Väänänen, Veikko (1977). Ab epistulis . . . ad sanctum Petrum: Formules prépositionnelles
latines étudiées dans leur contexte social. Helsinki: Suomainen Tiedeakatemia.
Väänänen, Veikko (19813). Introduction au latin vulgaire. Paris: Klincksieck.
Väänänen, Veikko (1987). Le journal Épître d’Égérie (Itinerarium Egeriae): Étude linguistique.
Helsinki: Suomainen Tiedeakatemia.
Vairel-Carron, Hélène (1975). Exclamation, ordre et défense. Paris: Les Belles Lettres.
Vairel-Carron, Hélène (1980). ‘La personne du nom: 3e personne et absence de personne’,
Bulletin de la Société de Linguistique de Paris 75: 267–83.
Vairel-Carron, Hélène (1981a). ‘Un modèle d’analyse linguistique des conditionnelles:
Latin si di sunt, si di sint, si di essent’, Bulletin de la Société de Linguistique de Paris
76: 275–326.
Vairel-Carron, Hélène (1981b). ‘The position of the vocative in the Latin case system’,
American Journal of Philology 102: 438–47.
Vairel-Carron, Hélène (1986). ‘Le problème du cas de tu’, Revue de Philologie 70: 31–57.
Vallat, Daniel (2002). ‘Les noms propres métaphoriques: Aspects référentiels et sémantiques’,
in A. Machtelt Bolkestein et al. (eds), 405–19.
Vallat, Daniel (2005). ‘Pluriel et nom propre: Problèmes de référence’, in Gualtiero Calboli
(ed.), 785–95.
Vandvik, Eirik (1942). Genitivus und Ablativus qualitatis im Latein. Oslo: Norske Videnskaps-
Akademie.
Vaughan, Alden G. (1942). Latin Adjectives with Partitive Meaning in Republican Literature.
Baltimore: Waverly (repr. New York: Kraus, 1966).
Vecchio, Tommaso del (2008). ‘Sull’uso di iste negli scenici latini arcaici’, Papers on Grammar
10: 93–108.
Vester, Elseline (1983). Instrument and Manner Expressions in Latin. Assen: Van Gorcum.
Vester, Elseline (1985). ‘Agentless passive constructions’, in Christian Touratier (ed.),
227–40.
Vester, Elseline (1989). ‘Relative clauses: A description of the indicative–subjunctive oppo-
sition’, in Gualtiero Calboli (ed.), 326–50.
Vester, Elseline (1990). ‘The satellite status of gerunds and gerundives in Latin’, in Mike
Hannay and Elseline Vester (eds), 103–13.
Vester, Elseline (1991). ‘Reflections on the gerund and gerundive’, in Robert G. Coleman (ed.),
295–309.
Vester, Elseline (1994). ‘The internal structure of adverbial ut-clauses’, in József Herman (ed.),
269–79.
Victor, Benjamin (2013). ‘A syntactical Hellenism at Horace, Satires 1.3.120–1, and a possible
imitation in Livy’, Classical Quarterly 63: 430–3.
Vidén, Gunhild (1984). The Roman Chancery Tradition: Studies in the Language of
Codex Theodosianus and Cassiodorus’ Variae. Göteborg: Acta Universitatis
Gothoburgensis.
Viljamaa, Toivo (1983). The Infinitive of Narration in Livy: A Study in Narrative Technique.
Turku: Annales Universitatis Turkuensis.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Bibliography
Westman, Rolf (1961). Das Futurpartizip als Ausdrucksmittel bei Seneca. Helsinki: Societas
Scientiarum Fennica.
Westman, Rolf (1983). ‘Fore, fore ut und futurum ut bei Seneca’, in Studies in Classical and
Modern Philology Presented to Y. M. Biese on the Occasion of his Eightieth Birthday.
Helsinki: Suomalainen Tiedeakatemia, 163–9.
Wetzel, Martin (1884). ‘Die Zulässigkeit as Konjunktius der Nebentempora nach Nichtpräteritis
im Lateinischen’, Gymnasium 2: Nr. 21, 22.
Wetzer, Harry (1996). The Typology of Adjectival Predication. Berlin: De Gruyter.
Wharton, David (1996). ‘Attributive prepositional phrases in Latin prose’, in Alfred Bammes-
berger and Friedrich Heberlein (eds), 165–75.
Wharton, David (2009). ‘On the distribution of adnominal prepositional phrases in Latin
prose’, Classical Philology 104: 184–207.
Wheeler, Arthur Leslie (1906). ‘The syntax of the imperfect indicative in Early Latin’, Classical
Philology 1: 357–90.
Wiberg, Lars (1956). ‘Étude sur les expressions du type “la fondation de Rome par Romulus” ’,
Studia Neophilologica 28: 148–217.
Wiesthaler, Franz (1956). Die oratio obliqua als künstlerisches Stilmittel in den Reden Ciceros.
Innsbruck: Wagner.
Wingo, Elvis O. (1972). Latin Punctuation in the Classical Age. The Hague: Mouton.
Wirth-Poelchau, Lore (1977). AcI und ‘quod’: Satz im lateinischen Sprachgebrauch mittelal-
terlicher und humanistischer Autoren. Erlangen-Nürnberg: J. Hogl.
Wisse, Jakob (1996). ‘The presence of Zeno: The date of Philodemus’ On Rhetoric and the use
of the “citative” and “reproducing” present in Latin and Greek’, in Rodie Risselada et al.
(eds), 173–202.
Wisse, Jakob (2014). ‘A popular consul’s claptrap and the sequence of tenses: The text and
rhetorical significance of Cicero, De lege agraria 2.7’, Mnemosyne 67: 911–29.
Wistrand, Erik (1933). Vitruviusstudier. Göteborg: Elanders boktryckeri aktiebolag.
Wistrand, Erik (1941). Über das Passivum. Göteborg: Elanders boktryckeri aktiebolag.
Wistrand, Erik (1958). Horace’s Ninth Epode and its Historical Background. Stockholm:
Almqvist & Wiksell.
Wistrand, Erik (1961). ‘Das lateinische Pronomen is in deiktischer Funktion’, Eranos 59: 101–15.
Wistrand, Erik (1963). ‘Nachträgliches zum situationsanaphorischen Gebrauch von is’, Eranos
61: 94–7.
Wölfflin, Eduard (1886). ‘Der substantivierte Infinitiv’, Archiv für Lateinische Lexikographie
und Grammatik 3: 70–91.
Wölfflin, Eduard (1887a). ‘Ut quid?’, Archiv für Lateinische Lexikographie und Grammatik 4:
617–18.
Wölfflin, Eduard (1887b). ‘Usque mit Accusativ’, Archiv für Lateinische Lexikographie und
Grammatik 4: 52–67.
Wölfflin, Eduard (1904). ‘Die adiectiva relativa’, Archiv für Lateinische Lexikographie und
Grammatik 13: 407–14.
Wolterstorff, Gottfried (1917). ‘Artikelbedeutung von ille bei Apuleius’, Glotta 8: 197–226.
Woodcock, Eric Charles (1959). A New Latin Syntax. London: Methuen.
Wright, Roger (2002). A Sociophilological Study of Late Latin. Turnhout: Brepols.
Wright, Roger (ed.) (2008). Latin vulgaire—Latin tardif VIII. Hildesheim: Olms-Weidmann.
Zawadowski, Leon (1959/60). ‘Criteria of Latin periphrastic forms’, Eos 50: 213–20.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
C. Commentaries
Austin, Roland Gregory (1960). M. Tulli Ciceronis Pro M. Caelio Oratio. Oxford: Clarendon Press.
Austin, Roland Gregory (1971). P. Vergili Maronis Aeneidos Liber Primus. Oxford: Clarendon
Press.
Austin, Roland Gregory (1977). P. Vergili Maronis Aeneidos Liber Sextus. Oxford: Clarendon
Press.
Bailey, Cyril (19492). Titi Lucreti Cari De Rerum Natura Libri Sex, 3 vols. Oxford: Clarendon Press.
Bömer, Franz (1969/1986). P. Ovidius Naso Metamorphosen, 6 vols. Heidelberg: Carl Winter.
Briscoe, John (1989). A Commentary on Livy, Books 31–33. Oxford: Oxford University Press.
Briscoe, John (2008). A Commentary on Livy, Books 38–40. Oxford: Oxford University Press.
Briscoe, John (2012). A Commentary on Livy, Books 41–45. Oxford: Oxford University Press.
Brix, Julius, Niemeyer, M., and Köhler, Oskar (19164). Plautus IV: Miles Gloriosus. Leipzig:
Teubner.
Conway, Robert Seymour (ed.) (1901). Livy Book II. Cambridge: Cambridge University Press.
Dougan, Thomas Wilson (1905/35). M. Tulli Ciceronis Tusculanarum Disputationum Libri
Quinque, 2 vols. Cambridge: Cambridge University Press.
Dyck, Andrew R. (2004). A Commentary on Cicero, De Legibus. Ann Arbor: University of
Michigan Press.
Dyck, Andrew R. (2013). Cicero Pro Marco Caelio. Cambridge: Cambridge University Press.
Dziatzko, Karl and Hauler, Edmund (19134). Ausgewählte Komödien des P. Terentius Afer.
Leipzig: Teubner.
Enk, Pieter-Jan (1932). Plauti Mercator. Leiden: Sijthoff.
Fedeli, Paolo (1980). Il Primo Libro Delle Elegie. Florence: Leo S. Olschki Editore.
Fedeli, Paolo (2005). Properzio elegie libro II. Cambridge: Francis Cairns.
Flach, Dieter (1997). Marcus Terentius Varro: Gespräche über die Landwirtschaft, Buch 2.
Darmstadt: Wissenschaftliche Buchgesellschaft.
Furneaux, Henry (ed.) (1896/1907). The Annals of Tacitus, with Introduction and Notes, 2
vols. Oxford: Clarendon Press.
Goodyear, Frank R. D. (1972/81). The Annals of Tacitus, Books I and II, 2 vols. Cambridge:
Cambridge University Press.
Heubner, Heinz and Fauth, Wolfgang (1982). P. Cornelius Tacitus: Die Historien, Band V.
Heidelberg: Carl Winter.
Hijmans, Benjamin Lodewijk et al. (1981). Apuleius Madaurensis Metamorphoses, Books VI
25–32 and VII: Text, Introduction and Commentary. Groningen: Bouma.
Horsfall, Nicholas (2000). Virgil, Aeneid VII: A Commentary. Leiden: Brill.
Keulen, Wytse (2007). Apuleius Metamorphoses I. Groningen: Forsten.
Kleywegt, Aad J. (2005). Valerius Flaccus, ‘Argonautica’, Book I: A Commentary. Leiden: Brill.
Koestermann, Erich (1963/8). Cornelius Tacitus, Annalen, 4 vols. Heidelberg: Carl Winter.
Lancel, Serge (1972/91). Actes de la conférence de Carthage 411, 4 vols. Paris: Les Éditions du
Cerf.
MacCary, Thomas and Willcock, M. M. (1976). Plautus Casina. Cambridge: Cambridge
University Press.
Manuwald, Gesine (2007). Cicero Philippics 3–9, 2 vols. Berlin: De Gruyter.
Mynors, Roger (1990). Virgil Georgics. Oxford: Clarendon Press.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
Bibliography
Nipperdey, Karl and Andresen, Georg (191511). P. Cornelius Tacitus, erklärt von Karl Nipper-
dey, Erster Band. Berlin: Weidmannsche Buchhandlung.
Oakley, Stephen (2005). A Commentary on Livy, Books VI–X, Volume IV: Book X. Oxford:
Clarendon Press.
Petersmann, Hubert (1973). T. Maccius Plautus Stichus. Heidelberg: Carl Winter.
Poortvliet, Harm Marien (1991). C. Valerius Flaccus Argonautica Book II. Amsterdam: VU
University Press.
Powell, J. G. F. (1988). Cicero Cato Maior de Senectute. Cambridge: Cambridge University
Press.
Richardson-Hay, Christine (2006). First Lessons. Book I of Seneca’s ‘Epistulae Morales’:
A Commentary. Bern: Lang.
Rizzo, Silvia (1991). M. Tulli Ciceronis Pro A. Cluentio Habito Oratio. Milan: Mondadori.
Ruggiero, Fabio (1991). Atti dei martiri Scilitani, Atti della Accademia Nazionale dei Lincei
388, Memorie V Serie IX, Volume I, Fascicolo 2, 67–139.
Schuster, Mauriz and Hanslik, Rudolf (19583). C. Plini Caecili Secundi Epistularum libri
novem; Epistularum ad Traianum liber; Panegyricus. Leipzig: Teubner.
Shackleton Bailey, D. R. (1965). Cicero’s Letters to Atticus, Volume I (Books I–II). Cambridge:
Cambridge University Press.
Shackleton Bailey, D. R. (1977). Cicero Epistulae ad Familiares, 2 vols. Cambridge: Cambridge
University Press.
Skutsch, Otto (1985). The Annals of Quintus Ennius. Oxford: Clarendon Press.
Speidel, Michael Alexander (1996). Die Römischen Schreibtafeln von Vindonissa: Lateinische
Texte des militärischen Alltags und ihre Geschichtliche Bedeutung. Brugg: Gesellschaft Pro
Vindonissa.
Tarrant, Richard (2012). Virgil Aeneid Book XII. Cambridge: Cambridge University Press.
Weissenborn, Wilhelm and Müller, H. J. (19115). Titi Livi Ab Urbe Condita Libri. Buch XXVI.
Berlin: Weidmannsche Verlagsbuchhandlung.
Woodman, Anthony J. and Martin, Ronald H. (1996). The Annals of Tacitus, Book 3.
Cambridge: Cambridge University Press.
Woytek, Erich (1982). T. Maccius Plautus, Persa: Einleitung, Text und Kommentar. Vienna:
Verlag der Österreichischen Akademie der Wissenschaften.
OUP CORRECTED PROOF – FINAL, 5/8/2015, SPi
INDEX LOCORUM
This Index contains all the examples that are discussed in the Syntax. The additional examples that are
given in the Supplements are not included. Abbreviations wherever possible follow the Oxford Latin
Dictionary and (for Late Latin authors) Blaise’s Dictionnaire.
Graecus 244, 267, 1076 to mark subject of infinitive 736, 1186; see also
adverbial accusative 85, 330, 334, 697, 827, 828, accusative and infinitive clauses
844, 847, 865, 885; see also expansion of with compound verbs of motion 102
arguments with impersonal verbs of emotion 132
as ‘base form’ in Late Latin 1216 with judicial verbs 157
cognate accusative 86, 102, 213 with verbs of concealing 168
double accusative 163 with verbs of dressing 264
with compound verbs 171, 1185 with verbs of filling 150
with verbs of asking 165, 1185 with verbs of producing a sensation 91
with verbs of teaching 167, 1185 with verbs of remembering and forgetting 117
Greek accusative 242, 264 with weather verbs 193
in calendar expressions 1229 achievements (state of affairs), defined 23
in exclamations 363 actions (state of affairs), defined 22
internal accusative 86, 102, 213 active voice 54, 230–305
object accusative 101, 137, 140, 146, 153, 163, ad (adverb) 1229
183, 186, 754 ad (preposition)
of extent in degree 330, 334, 697, 885, 1080, 1215 as alternative to dative case 1027
of extent of space 827 to mark arguments of adjectives 222, 227
of extent of time 844, 847 beneficiary arguments 218
of inner object 86, 102, 213; see also expansion to mark arguments of verbs 123
of arguments to mark beneficiary adjuncts 897
of limit of motion 92, 178 to mark direction and goal adjuncts 812
of local object 92, 178 to mark extent of space adjuncts 829
of person or thing affected 101, 137; see also to mark instrument adjuncts 880
object: affected objects to mark manner adjuncts 868
of respect 242 to mark optional attributes of nouns 1030
of result produced 101, 137; see also object: to mark position in time adjuncts 837
effected objects to mark purpose adjuncts 907
replacing nominative in Late Latin non-literary to mark respect adjuncts 917
texts 1186 with adjectives of desire 222
retained with autocausative passives 264 with adjectives of helpfulness 218
retained with passive participles 242 with judicial verbs 157
synecdochical accusative 242, 264, 915 with verbs of preparing 896
to mark arguments of verbal nouns with verbs of sending letters 142
1039, 1215 ad sensum construction, see agreement: notional
to mark cognate object 86, 102, 213 agreement
to mark direction and goal adjuncts 810, 819 address (semantic function), of vocatives 1224
to mark direction and goal arguments 178 address, forms of, modified by appositions 1068
to mark extent of space adjuncts 827 addressee (semantic function), defined 27
to mark extent of time adjuncts 844, 847 addressee arguments 27, 143, 1192
to mark items in a list 1216 adhuc 856
to mark manner adjuncts 865 adjective phrases 17, 215–29, 1074–85; see also
to mark modifiers of adverbs 1215 valency: of adjectives
to mark non-finite clauses as object 1216 arguments of adjectives 215–29, 1074, 1204
to mark object complements 186, 787, 1211 optional modifiers of adjectives 1074–85
to mark optional attributes of nouns 1028 degree modifiers 1079–85
to mark optional modifiers of adjectives 1215 adjectives
to mark price adjuncts 885 as a word class 45
to mark pseudo-object in proleptic adverbial use, see secondary predicates
construction 631, 759, 1188 arguments of, see adjective phrases; valency: of
to mark quantity and degree adjuncts 330, 334, adjectives
697, 885 as attribute of noun 216, 993
to mark quantity and degree modifiers of rarely expanded by optional modifiers 1075
adjectives 1080 as object complement 790
to mark respect adjuncts 915 as subject complement 215, 769
to mark respect expressions with comparative degree 46, 1075
adjectives 1075 descriptive adjectives 45, 1047
to mark second arguments (accusative combined with alius 947
objects) 101, 137, 140, 146, 153, 163, enallage or transference of 1051
183, 186, 754 evaluative adjectives 64, 208, 362, 369, 1047
to mark so-called locative object 92 modifying proper name 938
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
annalistic use (of historic present), see present diachronic developments in marking of 1238
tense: present indicative: historic present direction and goal arguments 28, 124, 142, 175,
annon, see an (question particle) 176, 178, see also direction and goal
ante (adverb) 841, 1229 arguments
with ablative indicating extent of time 842 experiencer arguments 27, 116, 132
ante (preposition) of adjectives, see valency: of adjectives
in calendar expressions 841 of nouns, see noun phrases: adnominal
to mark direction and goal adjuncts 812 arguments; valency: of nouns
to mark position in time adjuncts 840 of verbs, see also valency: of verbs
antequam clauses first arguments 81, 100, 139, 1181, 1186
use of moods in 638 second arguments 100, 139, 1181
use of tenses in 611 use of cases and prepositions to
anteriority 383 mark 1189–95
anticausative 272, 275 third arguments 139, 1181
antonyms, occurring in similar constructions 109, use of cases and prepositions to
110, 118, 140, 146, 622, 1191 mark 1195
apo koinou construction 1275 patient arguments 27
apodosis, see conditional sentences place arguments 28
apposition 1053–74 recipient arguments 27, 141, 142, 1192
agreement of relative pronoun with 1283 source arguments 28, 125, 178, 1196; see also
agreement of verb with subject modified source arguments
by 1259 statistical data on frequency 1181
clausal apposition 1070 asking, verbs of 164, 165, 1185
compared with genitive of definition 1023 performative use of 354, 630
distributive apposition 1067 aspect 380
modified by adverbs 1037 distinguished from Aktionsart 380
nominal apposition 1054–66 perfective aspect of perfect tense 444
non-referring noun phrase as 1117 assentio/assentior 283
non-restrictive apposition 1055, 1061 assertion (illocutionary force) 308, 309
of noun phrase with first person subject 738 asseverative expressions 375, 427, 676, 677
partitive apposition of measure assimilation 1278
expressions 1069 associative (semantic function), defined 27
restrictive apposition 1055, 1056 associative adjuncts 28, 897
with toponyms 1058, 1260 associative arguments 27, 119, 121, 161
terminology for 1056 with adjectives 227
to forms of address 1068 asyndetic coordination 1249
to unexpressed first and second person at 12, 33
subjects 1068 atque
approaching and befalling, verbs of 104, 1192 as coordinator of negative declarative
arbitror, in answers to questions 373 clauses 715
arceo, with dative beneficiary adjunct 894 with locally negated constituents 689
argument clauses attendant circumstances 854
imperative clauses (indirect commands), see attitudinal adverbs, see disjuncts: attitudinal
subordinate imperative clauses disjuncts
interrogative clauses (indirect questions), see attitudinal disjuncts, see disjuncts
subordinate interrogative clauses attraction
ne clauses, see ne clauses: as arguments of infinitive to match perfect tense governing
quia clauses, see quia clauses verb 540
quod clauses, see quod clauses of moods 666
use of indirect reflexive pronoun and possessive of pronouns to agree with subject or object
adjective in 1126, 1130 complement 1278
use of moods in 621 of verb to agree with subject complement 1261
use of tenses in 566 attributes, see noun phrases
ut clauses, see ut clauses: as arguments audeo, perfect subjunctive ausim 491
argumentative texts, see text types audio, with subject complement 209
arguments 11, 24, 72, 736–87, 797, 1176–98 aut, with locally negated constituents 691
addressee arguments 27, 143, 1192 autocausative passives, see passive voice
agent arguments 26; see also agent arguments auxiliary verbs 210–15
argument clauses, see argument clauses auxiliary phrase 211
associative arguments 27, 119, 121, 161, 227 distinguished from full verbs 210
cause arguments 27, 245 with passive infinitives 254
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
as a test to distinguish types of adjuncts 799 temporal cum clauses 611, 641
asyndetic coordination as factive or non-factive 641
distinguished from apposition 1056 compared with participles functioning as
comitative coordination with cum 899 secondary predicates 643
conjunction reduction 339, 749, 757, 1253, 1277 compared with temporal ubi, simulac, and
conjunctive coordination 713, 715 postquam clauses 613
agreement with subjects joined by 1249 with anterior meaning 613
disjunctive coordination 713, 716 with habitual or iterative meaning 611
agreement with subjects joined by 1253, with simultaneous meaning 613
1267, 1286 use of moods in 641, 644
emphatic et . . . et and neque . . . neque 1252 use of tenses in 589
of adjectives 45, 993 with historic present 605
adjective of amount and descriptive cunctus 804, 985
adjective 982 co-occurring with a determiner 971
of adjuncts 797 substantival use 1172
of events within a single temporal clause 616 cupido 22, 1039
of locally negated constituents 688 cupidus 22, 215, 222, 1039
of negative declarative clauses 713 cupio 21, 215
of negative imperative clauses 717 with dative or accusative 129
of negative interrogative sentences 340 cur 336
of nouns or noun phrases curo, imperative used as metadirective 351
agreement of modifiers with 1273
agreement of pronouns with 1284 damno 157
of prepositions governing same noun 1228 dative case 1179, 1183, 1206, 1216–19
zero coordination (asyndetic) archaic forms of 1236
agreement with subjects joined by 1249 dativus
coordinators 14, 68 auctoris 247, 269, 1217
copular verbs, see verbs commodi/incommodi 892, 1192
coram (adverb) 1229 ethicus 895, 931, 1203
coram (preposition) 806 finalis 778, 895
core 11, 19, 25 iudicantis 926, 1203
crassus 228 possessivus 107
crebro 852 sympatheticus 917, 919
credo, compared with qualifed truth disjuncts 926 double dative construction 780
cuias 336 ethic dative 931, 1203
cum (preposition) local dative 813
in result expressions 163 of advantage/disadvantage 141, 892, 1192
to mark accompanying circumstance of agent 248, 245, 269, 296
adjuncts 901 of direction and limits of motion 813
to mark arguments of adjectives 227 of indirect object 29, 104–10, 138, 141, 162, 763,
to mark associative adjuncts 897 813, 1192, 1195
to mark associative arguments 119, 144, 151, 161 of interest 141, 892, 1192
to mark manner adjuncts 866 of local standpoint 926, 1203
to mark means and instrument adjuncts 878 of person judging 926, 1203
to mark optional attributes of nouns 898, 1030 of possessor 107, 301
compared with ablative of description 1031 of purpose 77, 186, 780, 778, 793
to mark position in space adjuncts 806 of reference 141, 892, 1192
to mark position in time adjuncts 837 of separation 106, 1191
to mark subject complements 786 predicative dative
with verbs of changing 151 as object complement 186, 793
cum clauses as subject complement 77, 778
causal cum clauses 578, 644 redundant use of sibi with suus or se 980, 1136
combined with idcirco, ideo, and propterea 912 replaced by prepositional phrases with ad 1027
concessive cum clauses 578 to mark addressee arguments (indirect
cum identicum 611, 613, 641, 643 object) 141, 763, 1192, 1195
cum inversum construction 420 to mark agent adjuncts in non-passive
with historic infinitive 608 clauses 248
with historic present 605 to mark agent arguments 245, 269
cum narrativum 611, 613, 641, 643 in gerundive + sum construction 296, 301
cum temporale 611, 613, 641, 643 of adjectives in -bilis 248, 296
reason cum clauses 578, 644 to mark associative arguments 120
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
preverbs pronominals 50
relationship with prepositions and pronouns
adverbs 1228 as a word class 48
relationship with prepositions and cases 1234 anaphoric pronouns, see anaphoric(ally used)
price adjuncts 881–5 pronouns
as subcategory of means and instrument demonstrative pronouns, see demonstrative
adjuncts 881 pronouns
compared with arguments 882 determinative pronouns, see idem; ipse; is
use of ablative case to mark 1200 indefinite pronouns, see indefinite pronouns
price or value arguments 113, 158 interrogative pronouns, see interrogative
compared with adjuncts 882 pronouns
pridie personal pronouns, see personal pronouns
in calendar expressions 841, 1229 relative pronouns, see qui (relative)
use of moods with 640 prope (adverb) 1229
primary tenses, defined 554 as position in space adjunct 808
primo (adverb) 841 prope (preposition)
primus 1050 to mark direction and goal adjuncts 812
priusquam clauses to mark position in space adjuncts 806
use of moods in 618, 638 proper names 42
use of tenses in 611 as head of noun phrase 937
pro (preposition) as secondary predicate 940
to mark beneficiary adjuncts 897 combined with aliquis 1106
to mark cause adjuncts 904 definiteness of 1087
to mark position in space adjuncts 806 modified by appositions 938, 1056
to mark subject complements 786 modified by evaluative adjectives 938
probe, in answers to questions 373 proprius, compared with suus 980
process adjuncts 857–99, 1201; see also beneficiary propter (adverb) 1229
adjuncts; manner adjuncts; means and propter (preposition)
instrument adjuncts to mark position in space adjuncts 806
processes (state of affairs), defined 22 to mark purpose adjuncts 907
procul 1229 to mark reason adjuncts 911
producing a sensation, verbs of 91 propterea 909
profecto 309, 348, 924 prosodic unit 67, 69, 732
proficiscor, statistical data on tense usage 394 prospective value, see participles: future participle:
prohibeo, with dative beneficiary adjunct 894 prospective value of
prohibitions 349–58, 497–504; see also prospicio
illocutionary force: directive with dative beneficiary adjunct 893
illocutionary force; imperative sentences with dative or accusative 129
commands and prohibitions of the past 503 protasis, see conditional sentences
containing metadirectives 351 provideo
coordination of 717 with dative beneficiary adjunct 893
inhibitive use of present imperative 516 with dative or accusative 129
modulation of 350 proviso clauses 652
ne + future imperative 519 pseudo-final ut/ne clauses 566, 930
ne + perfect subjunctive 352, 498, 683 pseudo-object 631, 759, 1188
ne + present imperative 516, 682 pseudo-subject 761
ne + present subjunctive 351, 499, 682 pudet 132
referring to internal mental activities 516 agreement of complements and secondary
ne quis velit + perfect infinitive 538 predicates 1270
noli + infinitive 352, 513, 683 punctuation 15, 68
nolito and nolitote + infinitive 519 in antiquity 16
non + perfect subjunctive 683 to mark non-restrictive apposition 1055
responses to commands and prohibitions 376 purpose (final) adjuncts 905–9
statistical data on frequency 513 compared with cause adjuncts 903
use of tenses in subordinate clauses after main compared with purpose disjuncts 905
clause containing a prohibition 562 compared with reason adjuncts 909
proin(de), used to modulate directive purpose (final) clauses 28, 703
expressions 355 as illocutionary disjuncts 930
prolative infinitive 164, 183, 522, 526 combined with idcirco, ideo, and propterea 909
proleptic constructions 631 use of indirect reflexive pronoun and possessive
promising, verbs of 525, 567 adjective in 1127, 1131
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
nulla res supplying missing forms of nihil 710 of modifiers within noun phrase 1047
respect adjuncts 914–18 of negation 674, 676, 715, 731, 797, 1165
compared with qualified truth disjuncts 914 disjuncts excluded from 798, 925
respect expressions modifying adjectives 1077 ipse as scope of negation 1154
responses 368–78 of sentence questions 318, 326
responses to directive utterances 376 constituent scope 318
responses to questions 370 repeated in answer 371
restrictive apposition, see apposition sentence scope 318
result (consecutive) clauses with particle -ne 323
quin result clauses 706 scribo, valency of 143
use of moods in 653 second person, see personal pronouns; subject
use of tenses in 566, 574 secondary predicates 30, 46, 187, 206
ut result clauses 623 adjectives of amount as 982
with verbs of causation 567, 624 agreeing with implicit object 757
resulting state interpretation agreement with constituent they relate to 1264,
of perfect indicative 446 1265, 1267
of perfect participles 541 agreement with non-subject arguments 1270
of pluperfect indicative 610 agreement with unexpressed subject of
retrospective value accusative and infinitive 1265
of future perfect tense 467 ambo as 987
of perfect tense 445, 573 binary quantifiers as 989
rhetorical plural 1119–20 compared with cause adjuncts 903
rhetorical questions 319, 343, 374, 600 compared with clausal apposition 1074
becoming accusative and infinitive clauses in compared with distributive apposition 1067
indirect speech 510 compared with manner adjuncts 859, 861
negated with indefinite pronouns 1107 compared with temporal cum clauses 643
repudiating questions 344 containing question words 337
use of tenses and moods in 486, 489, 493 distinguished from attributive participles 994
use of indefinite pronouns in 732 distinguished from object complements 789
rogo 164, 1185 future participle as 546
followed by indicative indirect question 630 gerundives as 289, 293
in imperative sentences 353 in infinitive clauses 746
route adjuncts, see path adjuncts in vocative case 1225
ruling, obeying, and serving, verbs of 104, 116, ipse as 1161
1192 medius as 1050
rursus 853 modified by adverbs 1037
rus notional agreement with collective
ablative as source adjunct 816, 819 subject 1292
accusative as direction or goal adjunct 811, 819 omnis and cunctus as 985
locative as position in space adjunct 803, 819 prepositional phrases as 1033
present participle as 542
sacrificing, verbs of 185 proper name as 940
saepe 852 quisque as 987
sane 1081 transference of adjective to 1053
as degree adjunct 888 uterque as 990
in answers to questions 373 verb agreement with quisque as 1260
in concessions 361 secundum (adverb) 1229
satellite clauses 566; see also adjunct clauses; secundum (preposition) 806
disjuncts sedeo, as copula 207
satellites 12, 25, 72, 797; see also adjuncts; semantic functions 26
disjuncts with three-place verbs 137
diachronic developments in marking of 1240 with two-place verbs 99
satis semantic negation 673, 1107, 1167
satin (ut) used to make questions more semantic passives 231, 236–58
emphatic 341 semantic reflexive 272, 273
with genitive of quantity 1022 semel 852
saying, verbs of, see communication, verbs of semideponent verbs 55
scilicet 310, 924 semper 852
scio 448, 552 sending letters, verbs of 142
scope 318, 674 sentence questions, see interrogative sentences
of ipse 1153 sentence scope 318
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi
thinking, verbs of, see cognition, verbs of to mark direction and goal adjuncts 812
third person, see subject to mark position in space adjuncts 806
three-place verbs, see verbs umquam 841
time adjuncts 833–56 unde (interrogative adverb) 336, 818
time of speaking 380, 383 unde (relative adverb) 818
freedom of speaker in determining 399 undecumque 818
time within which adjuncts 849–51 universal quantifiers 985
combined with frequency expressions 852 unus
timeo combined with aliqui(s) 942, 970
with dative beneficiary adjunct 893 combined with quisque 987, 1173
with dative or accusative 129 combined with zero quantifier 973
tmesis 1231 developing into indefinite article 1114
topic (pragmatic function) 31, 773 non-numerical uses of 1115
contrastive topic 676 unusquisque 987, 1173
of accusative pseudo-object 759 urbs, in apposition to name 1058
of nominative pseudo-subject 761 agreement of verb with 1260
topic shift 271 usquam 808
in choice between active and passive 251 usque (adverb) 848, 1229
toponyms usque (preposition) 837
grammatical agreement with 1281 usus est 116, 622
modified by restrictive apposition 1058 with passive participles 256
agreement of verb with 1059, 1260 ut (interrogative adverb) 336
totiens 852 ut (particle)
totus 804, 991 in wishes 359
co-occurring with a determiner 971 with subjunctive commands
neuter totum modifying infinitive 943 similarity to ut clauses 500
with meaning ‘all’ 992 ut clauses
towns and small islands, in space expressions 819 adverbial ut clauses 566, 930
trans (adverb) 1229 announced by cataphoric determiner 1099
trans (preposition) announced by cataphoric pronoun 1145
to mark direction and goal adjuncts 812 as adjuncts
to mark position in space adjuncts 806 purpose (final) ut clauses 566
transcategorial parallelism 21, 1191, 1204, use of subjunctive mood in 651
1215, 1218 use of tenses in 573
transeo 172 result (consecutive) ut clauses 623
transfer and giving, verbs of 137, 140, 763, more fequently following main clause with
1192, 1195 perfect tense 445
transmitto 172 use of tenses in 574
treatment, verbs of 137, 146, 1195 temporal ut clauses, use of tenses in 609, 610
tremo, valency of 85 as arguments 621
trivalent, see verbs: three-place verbs imperative clauses (indirect commands) 567,
true passives 231 236–58 619, 621
true reflexive 272, 273 use of moods in 625
trying, verbs of 627 use of tenses in 567
tu, see personal pronouns ut non with negated expressions of
tum 841 necessity 705
correlated with temporal cum clause 641 with facio and efficio 190
two-place verbs, see verbs with impersonal est 95
with impersonal verbs denoting addition 624
ubi (interrogative adverb) 336 with impersonal verbs denoting
ubi (relative adverb) 808 existence 624
ubi clauses 610 with metadirective fac 351
ubi temporal clauses, use of tenses in 609 with verbs denoting transition to another
ubicumque (adverb) 808 topic 624
ubicumque clauses, use of moods in 638 with verbs of causation 567, 624
ubique 808 with verbs of fearing 623, 702
ullus 972, 1107 with verbs of happening and befalling 624
ultimus 1050 with verbs of manipulation 183
with partitive genitive 948 as illocutionary disjuncts 566, 930
ultra (adverb) 1229 as indignant questions 347
ultra (preposition) pseudo-final ut clauses 566, 930
OUP CORRECTED PROOF – FINAL, 8/8/2015, SPi