The Sentence in Biblical Hebrew

JANUA L I N G U A R U M
STUDIA M E M O R I A E
N I C O L A I VAN WIJK DEDICATA
edenda curat
C. H. VAN S C H O O N E V E L D
Indiana University
Series Practica, 231

THE SENTENCE
IN
BIBLICAL HEBREW
by
FRANCIS L ANDERSEN
MOUTON PUBLISHERS THE HAGUE PARIS NEW YORK

First edition: 1974
Second printing: 1980
ISBN 90 279 2673 5
Copyright 1974 by Mouton Publishers, The Hague, The Netherlands

No part of this book may be translated or reproduced in any form, by print, photoprint,
microfilm, or any other means, without written permission from the publishers
Printed in Great Britain

PREFACE
Recent advances in linguistic theory and method make desir-

able a fresh examination of the sentence systems of natural
languages. Ancient Hebrew is practically a virgin field for
such research. The existing literature contains no system-
atic treatment of the subject.
I am grateful to the Church Divinity School of the Pacific
and to the National Endowment for the Humanities of the Uni-
ted States Government for supporting this project--the form-
er by its generous policy of sabbatical leave, the latter by
a grant (#R0-5068-72-155). The findings and conclusions pre-
sented here do not necessarily represent the view of the En-
dowment.
Thanks are due to many persons. To Dr. Anne Draffkorn Kil-
mer, Chairman of the Department of Near Eastern Languages,
University of California, Berkeley, and to Dr. William G.
Dever, Director of the William Foxwell Albright Institute of
Archaeological Research, Jerusalem, for hospitality in their
respective institutions. To the Reverend Esther H. Davis for
help in typing. Last and best, to my wife Lois for unfailing
encouragement and support.
It is a pleasure to dedicate this study to John Arthur
Thompson, my first Hebrew teacher, whose lifelong devotion
to biblical studies has been a constant inspiration.
Jerusalem
October, 1972
CONTENTS
PREFACE 5
SYMBOLS AND ABBREVIATIONS 14
THE TRANSLITERATION OF HEBREW 16
1. INTRODUCTION 17
1.0. Traditional Grammar 17
1.1. Discourse Grammar 18
Notes 19
2. THE SENTENCE IN HEBREW 21/
2.0. Definition of Sentence 21
2.1. The Clause as a Sentence Element 22
2.2. Sentence Types 24
2.3. Complex Sentences 26
2.4. Compound Sentences 27
2.5. Relationships Between Sentence Types 28
2.6. Surface Grammar and Deep Grammar 29
2.6.0. The English Relative Clause 29
2.6.1. Head is a Proper Noun 30
2.6.2. Head is a Pronoun 30
2.6.3. Head is a Definite Noun 30
2.6.4. Head is an Indefinite Noun 31
2.6.5. Coordination as Alternative Realization. . 31
2.6.6. An Exception 32
2.6.7. Naming Two Persons 32
2.6.8. The Use of the Nominalizer 34
2.6.9. Conclusions 34
Notes 35
3. APPOSITION SENTENCES 36
3.0. Deep Grammar of Apposition 36
3.1. Surface Grammar of Apposition 37
3.2. Verbal Repetition in Apposition 37
3.3. Synonymous Apposition 38
8 TABLE OF CONTENTS
3.4. Epic Repetition in Apposition 39

3.4.0. Introduction 39
3.4.1. Verb Patterns in Epic Apposition 39
3.4.2. Discourse Function of Epic Apposition. . . 40
3.4.3. Alternative Realization in Sequential
Clauses 42
3.5. Apposition for Emphasis 43
3.5.0. Formal Features 43
3.5.1. Antithesis in Apposition 43
3.5.2. Apposition in Another Perspective 44
3.5.3. Climactic Repetition in Apposition . . . . 44
3.5.4. Coordination of Emphatic Repetition. . . . 45
3.6. Resumption and Distribution in Apposition . . . . 45
3.6.0. Resumption 45
3.6.1. Distribution 45
3.7. Explanation in Apposition 46
3.7.0. The Semantics of Apposition 46
3.7.1. Specifying Apposition 47
3.7.2. Exposition in Apposition 49
3.7.3. An Example of Explication 50
3.7.4. A Summary in Apposition 53
3.7.5. Titles and Colophons 53
3.8. Curses and Blessings 54
3.9. Multiple Apposition 55
3.10. Apposition Instead of Other Constructions . . . . 56
3.10.0. Alternative Deep Relationships 56
3.10.1. Coordinate Commands in Apposition . . . . 56
3.10.2. Coordinated Declarative Clauses in
Apposition 57
3.10.3. Coordinate Questions in Apposition. . . . 57
3.10.4. Antithetical Clause in Apposition . . . . 57
3.10.5. Apposition Instead of Subordination . . . 58
3.10.6. Apposition Instead of a Relative Clause . 59
3.10.7. Apposition Instead of Sequence 59
Notes 59
4. COORDINATION 61
4.0. Kinds of Coordination 61
4.1. Upper-level Coordination 61
4.1.0. Heterogeneous Speeches 61
4.1.1. Coordination of Units of Narrative . . . . 62
4.1.2. Stories in Juxtaposition 62
4.1.3. Coordinated Stories 63
4.1.4. Story-level Episodes 63
4.1.5. Episode-level Paragraphs 64
4.2. Paragraph-level Coordination 64
4.2.1. Sequential Coordination 64
4.2.2. Paragraph-level Circumstantial Clause. . . 65
4.2.3. Paragraph-level Adjunctive Clause 66
4.2.4. Paragraph-level Surprise Clause 66
TABLE OF CONTENTS 9
4.3. Sentence-level Coordination 66

4.3.0. Sentence or Paragraph? 66
4.3.1. Conjunctive Coordination 67
4.3.2. Chiastic Coordination 67
4.3.3. Alternative (Disjunctive) Sentence . . . . 68
4.3.4. Contrastive Coordination 68
4.3.5. Antithetical Coordination 69
4.4. Inclusive and Exclusive Coordination 69
4.4.0. Deep Grammar 69
4.4.1. Inclusive Coordination 69
4.4.2. Exclusive Coordination 70
4.5. Inter-clausal Relationships in Precative and Pre-
dictive Discourse 71
4.6. Summary 73
4.7. Back-looping (Rank-shifting) 74
4.8. Alternative Surface Realizations 75
4.9. Empirical Testing 76
Notes 76
5. CIRCUMSTANTIAL CLAUSES 77
5.0. Introduction 77
5.1. Episode-marginal Circumstantial Clauses 78
5.1.0. Nucleus and Margin 78
5.1.1. Episode-initial Circumstantial Clauses . . 79
5.1.2. Episode-final Circumstantial Clauses . . . 80
5.1.3. Circumstantial Clause Beside an Episode. . 82
5.2. Sentence-level Circumstantial Clauses 86
5.2.0. Sentence versus Paragraph 86
5.2.1. Clauses Circumstantial to Time Margin. . . 86
5.2.2. Circumstance of a Circumstance 87
5.3. Pseudocircumstantial Sequential Clauses 87
5.4. Pseudosequential Circumstantial Clauses 88
5.5. Circumstantial Clauses as Alternatives to Noncir-
cumstantial Constructions 88
5.5.1. Circumstantial Form for Deep Subordination 89
5.5.2. Circumstantial Form for a Relative Clause. 90
Notes 91
6. ADJUNCTIVE CLAUSES 92
6.0. Structure 92
6.1. Function 92
6.2. Adjunctive Clauses Used Circumstantially. . . . . 93
6.3. Other Forms 93
Notes 93
7. SURPRISE CLAUSES 94
7.0. Form 94
7.1. Participant Perspective 94
7.2. Dream Reports 95
7.3. Other Uses 96
7.4. Other Forms 96
10 TABLE OF CONTENTS
8. CONJUNCTIVE SENTENCES 97
8.0. The Form of a Conjunctive Sentence 97
8.0.0. Optimum Realization 97
8.0.1. Double-duty Items 98
8.0.2. Multiple Coordination 98
8.1. Declarative Conjunctive Sentences 99
8.1.1. Reports of Accomplished Fact Using 'Per-
fect' Verbs 99
8.1.2. Conjoined Predictive Clauses 99
8.1.3. Conjoined Verbless Clauses 100
8.1.4. Reports of Present Facts Using Quasiverbal
Clauses 101
8.1.5. Circumstantial Conjunctive Sentences . . . 101
8.1.6. Conjoined Clauses in Poetry 101
8.1.7. Dissimilar Clauses Conjoined 102
8.1.8. Successive Events in Conjoined Clauses . . 103
8.2. Distributive Coordination 103
8.3. Conjoined Precative Clauses 104
8.3.0. The Variety of Combinations 104
8.3.1. Conjoined Verbless Precative Clauses . . . 107
8.3.2. Conjoined Imperative Clauses 108
8.3.3. Conjoined Jussive Clauses Ill
8.3.4. Conjoined Cohortative Clauses Ill
8.3.5. Mixed Linkages of Precative Clauses. . . . 112
8.4. Negation in Conjunctive Sentences 113
8.5. Conjoined Prohibitions 113
8.6. Coordination of Questions 114
8.7. Conjoining of Surprise Clauses 115
8.8. Coordination of Subordinate Clauses 115
8.9. Coordination of Relative Clauses 116
8.10. Conjunctive Sentences Instead of Other Construc-
tions 117
8.10.1. Hendiadys in Conjunctive Sentences. . . . 117
8.10.2. Coordination Instead of Apposition. . . . 117
8.10.3. Coordination Instead of Subordination . . 117
Notes 118
9. CHIASTIC SENTENCES 119

9.0. The Form of Inter-clause Chiasmus 119
9.1. The Surface Grammar of Chiastic Sentences . . . . 120
9.2. The Deep Grammar of Chiasmus 121
9.3. Chiasmus in Narrative Prose 122
9.3.0. An Illustration 122
9.3.1. Chiasmus in Poetic Discourse 123
9.3.2. Chiasmus in Epic Narrative 123
9.3.3. Three-clause Chiasmus 126
9.4. Grammatical Aspects of Chiasmus 127
9.4.1. Verbless Predicators 127
9.4.2. Perfect Verbs 127
TABLE OF CONTENTS 11
9.4.3. Subjects in Chiasmus 127

9.4.4. Objects in Chiasmus 128
9.4.5. Indirect Objects in Chiasmus 129
9.4.6. Other Clause-level Elements 129
9.4.7. Other Verbal Patterns 130
9.4.8. Chiasmus with Negation 130
9.4.9. Chiasmus not Involving the Verbs. . . . 131
9.5. Chiasmus in Predictive Discourse 131
9.5.0. Verb Patterns 131
9.5.1. Subjects in Chiasmus 131
9.5.2. Objects in Chiasmus 132
9.5.3. Mixed Constructions 132
9.5.4. Indirect Objects in Chiasmus 133
9.5.5. Other Clause-level Tagmemes in Chiasmus 133
9.5.6. Other Verb Forms in Chiasmus 133
9.6. Chiasmus in Precative Discourse 133
9.7. Chiasmus in Prohibition 135
9.8. Incompletely Formed Chiasmus 135
9.9. Discontinuous Chiastic Sentences 136
9.10. Chiasmus as a High-level Node 136
9.11. Chiastic Sentence as Nucleus 139
9.12. Chiasmus a Distortion 139
9.13. Sequential Clauses in Chiasmus 139
Notes 140
10. DISJUNCTIVE SENTENCES 141

10.0. Disjunctive Coordination 141
10.1. Phrase-level Disjunction 141
10.1.1. A or 141
10.1.2. Either A or 142
10.1.3. Either A and 142
10.1.4. whether A or 142
10.2. Disjunctive Coordination above Phrase Level. . 142
10.2.0. Transformations up and down the
Hierarchy 142
10.2.1. Disjunctive Sentences 146
10.2.2. Paragraph-level Disjunction 146
10.2.3. Disjunction of Paragraphs 146
10.3. Disjunctive Questions 147
10.3.0. Fully Formed Disjunction 147
10.3.1. Interrogated Disjunctive Sentence. . . 147
10.3.2. The Normal Construction 147
10.3.3. Phrase-level Disjunction of Questions. 148
10.3.4. Redundant Antithetical Tag Question. . 148
10.4. Disjunctive Realization of Conjunctive
Relationships ' 148
10.5. Coniunctive Realization of Disiunctive
12 TABLE OF CONTENTS
11. CONTRASTIVE SENTENCES 150

11.0. The Degree of Contrast 150
11.1. Contrastive Sentences and Other Constructions . 150
11.2. Contrast with Pronoun Subjects 151
11.3. Contrast with Nouns as Subjects 151
11.4. Contrast with Objects 152
11.5. Other Items in Contrast 152
11.6. Exceptions 152
11.7. Contrast Sentence with Asyndeton 153
12. INCLUSIVE SENTENCES 154

12.0. Inclusion and Addition 154
12.1. Phrase-level Coordination Using GAM 154
12.1.1. Inclusive Phrases 155
12.1.2. Compound Conjunction 155
12.1.3. Duals 155
12.2. Trans-sentence Inclusive Phrases 155
12.3. Double Coordination 157
12.4. Inclusive Coordination and Sentence Types . . . 157
12.4.1. Inclusive Chiastic Sentences 158
12.4.2. Inclusive Conjunctive Sentences . . . . 158
12.4.3. Inclusive 'Contrast' Sentences 159
12.4.4. Circumstantial Clauses 160
12.4.5. Surprise Clauses 161
12.4.6. Paragraph-level Inclusive Linkage . . . 161
12.4.7. Predictive Discourse 161
12.4.8. Other Constructions 162
12.4.9. Inclusive Complex Sentences 162
12.4.10. H p Equivalent to GAM 162
12.5. Inclusive Coordination and Negation 163
12.6. The Implication of Inclusive Coordination . . . 163
12.7. Noninclusive Uses of GAM 164
12.7.1. Coordination 164
12.7.2. Compound Inter-clause Conjunction . . . 165
12.8. GAM not a Conjunction 165
12.8.1. Appositive GAM 165
12.8.2. Emphasizing GAM 166
12.8.3. Focussing GAM 166
12.9. The Hierarchical Significance of GAM 166
Notes 167
13. EXCLUSIVE SENTENCES 168

13.0. Signals of Exclusive Relationships 168
13.1. The Form of the Exclusive Relationship 170
13.2. Phrase-level Exclusion 172
13.3. Trans-sentence Exclusive Phrases 172
13.4. Exclusive Sentences 173
13.5. Exclusive Forms used for Antithetical Relation-
ships 173
TABLE OF CONTENTS 13
13.6. Exclusive Forms Used for Coordination or Appo-

sition 174
13.7. Exclusive Relationships Realized by Antitheti-
cal Forms 174
13.8. Phrase 'Adverb* and Clause 'Adverb' 175
13.9. Limitative 'Adverbs' 175
13.10. Limitative Clause-modifier 177
13.11. Summary 177
Notes 178
14. ANTITHETICAL SENTENCES 179

14.0. Antithesis between Clauses 179
14.1. The Form of Antithetical Sentences 179
14.2. Antithesis with Implicit Negation 180
14.3. Antithesis by Means of Antonyms 181
14.4. Antithesis by Negation 181
14.5. Antithesis After Negation 183
14.5.1. Antithetical W- 183
14.5.2. Antithetical Sequential WW 183
14.5.3. Antithetical 183
14.5.4. Antithetical 1 1M 184
14.5.5. Antithetical 1 1M L"1 184
14.5.6. Antithetical ÎM. 184
14.5.7. Antithesis Using Exclusive Forms. . . . 185
14.6. Antithetical Questions 185
14.7. Antithesis in Apposition 185
Notes 185
15. SURFACE REALIZATIONS AND DEEP RELATIONSHIPS 186

15.0. Introduction 186
15.1. Alternative Surface Realizations 186
15.2. Limitations in Alternative Realizations . . . . 186
15.3. Juxtaposition and Concatenation 190
15.4. Coordination and Subordination 190
INDEX OF BIBLICAL REFERENCES 192

SYMBOLS AND ABBREVIATIONS
* Unattested form
+ Obligatory
+ Optional
11 Paragraph
< > (enclose) Symbol of a relationship or syntagmeme
A Apposition
Aj Adjunctive
Adv Adverb
Alt Alternative
Ant Antithes is
Antithetical
AV Authorized (King James) Version
Ch Chiasmus
Chiastic
Cir Circumstantial
Cj Conjunction
Conjunctive (Clause or Sentence)
CI Clause
Cn Contrastive
Co Coordination
D Discourse
Dc Declarative
Dj Disjunctive
Ep Episode
Eq Equative
Equivalent
Ex Exclamative
Exc Exclusive
Hb Hebrew
Infinitive Absolute
If Infinitive
Inc Inclusive
Int Interrogative
L Locative
Modification (Member -- Dik)
Mg Margin
Marginal
SYMBOLS AND ABBREVIATIONS 15
Noun
Nd Definite noun
Ni Indefinite noun
Nom Nominalized construction
Nominalizer
Np Proper noun
Ns Suffixed noun
Nuc Nucleus
0 Object
OC Object complement
OT Old Testament
Predicate, Predicator
Predication
Pc Precative
Ph Phrase
Pp Preposition
Pr Pronoun
Pt Participle
Q Quotation
QV Quasiverbal
Rel Relative
Res Resumption
S Subject
Se Sentence
Seq Sequence
Sub Subordinate
Subordination
Sur Surprise
Sus Suspended, Casus pendens
sy Syntagmeme
Tagmeme
Ti Time
V Verb
VC Cohortative verb
VI Imperative verb
VJ Jussive verb
VL Verbless (clause)
VP Prefixed (imperfect) verb
VS Suffixed (perfect) verb
Wd Word
WP Wv-consecutive with VP (sequential past)
WS Wv-consecutive with VS (sequential future)
THE TRANSLITERATION OF HEBREW
I apologize to Hebraists for not citing texts in Hebrew

characters. Inflation has been hard on books with exotic
scripts. The use of transcription should make this study
more useful to linguists who do not read Hebrew, and the
taxonomy of the Hebrew sentence system is not altogether
without interest for universal grammar.
For purposes of syntax the finer points of Massoretic
phonology are not often pertinent. The original can al-
ways be consulted. The system of transliteration used
here is a normalized quasi-phonemic representation of
the conventional orthography. Variations in spelling due
to the inconsistent use of matres lectionis have been ig-
nored. All long vowels are represented as v, no matter
what their historical development has been; for example,
long /o/ is whether <*u, <*aw, <*a, and whether spelled
with or without a mater lectionis. In fact the matres
lectionis are represented only by vowel length, never by
consonants, except in the case of a few historical spell-
ings like z">t this. This will be noticed particularly in
the omission of matres lectionis from the ends of words.
Thus n.T this is ze not zeh. Dgei is shown by doubling
the consonant when this is indicated, but the spirantiz-
ation of stops is ignored as subphonemic. The standard
equivalents of the consonants are used. The 1/atef vowels
are shown a, e, , but unfortunately "" was not avail-
able, and e has been regretfully used for shewa.
1
INTRODUCTION
1.0. TRADITIONAL GRAMMAR
The Hebrew SENTENCE (Se) as the domain of inter-clausal rela-

tionships has not been studied systematically on a full scale
since the volume on syntax in Eduard Knig's monumental gram-
mar. 1 The established approach is found too in several briefer
works. Thus A. B. Davidson's Hebrew Syntax (Edinburgh, 1894)
has brief sections on "The Conditional Sentence" (pp. 175-179),
"The Conjunctive Sentence" (pp. 184-185), etc. These writers
use 'sentence' where we would use 'clause.' The categories
are taken for granted, as if the criteria for classification
were obvious. Each clause type is illustrated by means of a
few selected examples. No attempt is made to establish a gen-
eral theory of inter-clausal relationships , and the evidence
of the texts is not presented in a comprehensive way.
Other presentations of Hebrew syntax above the level of PHRASE
(Ph) or CLAUSE (CI) are only sketches. Even Joon's Grammaire de
l'Hebreu biblique2 (Rome, 1947), w i t h its admirable overall
treatment of Hebrew syntax, often has but one page on sentence
types to which we devote a whole chapter. The best recent syntax
is Carl Brockelmann's Hebrische Syntax (Neukirchen, 1956). He
devotes Book Three (pp. 130-165) to Der Zusammengesetzte Satz.
The treatment is brief; for instance, he devotes only seven
lines ( 142, 159d) to clauses in apposition (Chapter 3 below).
The method is essentially a listing of an inventory of clause
types, more by reference to their inner structure than to dis-
course function. So far as the latter is concerned, the cate-
gories are assumed, and criteria for their identification are
drawn from logic, psychology or etymology. 2
Knig's work is still impressive, and a rich source of organ-
ized data. In the Second Part of the Syntax he studies first
individual 'sentences' (that is clauses) and then die Satzzu-
sammensetzungen under two headings --coordination (pp. 489-
546) and subordination (pp. 546-620). Clauses in apposition
he does not examine, except as asyndetic coordination. This
lack is made up partly in Ewald Khr, Die Ausdrucksmittel der
Konjunktionslosen Hypotaxe in der ltesten hebrischen Prosa:
Ein Beitrag zur historischen Syntax des Hebrischen: Beitrge
18 INTRODUCTION
zur Semitischen Philologie und Linguistik Heft 7 (Leipzig,

1929). Yet even here it is hypotactic clauses as such, not
the extended relationships between successive clauses in hypo-
taxis, that are examined. In effect what Khr does is (in our
terminology) to study the use of apposition and coordination
to realize subordinate relationships. In any case, the method
is based on Wundt's Vlkerpsychologie, and his interests are
strongly historical-comparative, as the title indicates.
In English-speaking scholarship the preeminence of S. R.
Driver as a Hebraist remains uneclipsed. His great work, A
Treatise on the Use of the Tenses in Hebrew2 (Oxford, 1891),
and all his commentaries still repay study. But many of the
concerns of nineteenth century scholarship no longer motivate
us, and many of its presuppositions no longer guide us. From
this distance it is possible to see the limitations in this
earlier work. In the light of twentieth century linguistics
the approach was too metaphysical (grammatical categories are
derived from 'ideas'), too physchological (a variety of clauses
may be gathered under a rubric like 'wish'), too etymological
(diachronic explanations take precedence over synchronic des-
criptions) . There are excursions into text criticism and some-
times regrettable attempts to correct difficult readings to
conform to the regularities of text-book grammar. Each of these
concerns is valid in its own way, but they can distract re-
search from the immediate task of describing what is there.
Comparison of traditional grammars soon reveals that an agreed
theoretical foundation was never laid. Without explicit and
methodologically rigorous definitions of basic units and re-
lationships the classification of a linguistic datum remains
whimsical, and the same clause will often be described dif-
ferently by different writers, with no discussion of the reason
for doing so. For such reasons we have not considered it prof-
itable to document everything that others have said on a par-
ticular construction, nor to enter into debate with existing
literature. This would only add to the size of the book, with-
out increasing its substance.
1.1. DISCOURSE GRAMMAR
The main stimulus for the present monograph has come from con-
temporary linguistics. The last decade in particular has seen
spectacular progress all along the line. Pertinent to the task
in hand is the emergence from tagmemic circles of a model of
language structure capable of handling the functions of clause,
sentence and PARAGRAPH (A) in terms of a hierarchy of syntag-
memes.
The groundwork of tagmemic theory was laid by Kenneth L.
Pike, 3 who has continued to develop the grammar of discourse
along hierarchical lines. Robert E. Longacre has forged
ahead, exploring the hierarchy of discourse in dozens of
languages." I am indebted also to Dr. Joseph Grimes for the
idea that alternative surface realizations are a matter of
'staging.'
IHTHODUCTION 19
The strong points of tagmemics are its empirical approach,

its respect for living language data, its concern for analyti-
cal description rather than generation, its search for units
and for classes, its interest in relationships within specific
constructions rather than abstract functions as such. All this
serves the present study, which is essentially a taxonomy of
Hebrew inter-clause constructions. To keep this aim in view has
required resistance to temptations to go off into all kinds of
theoretical discussion. Simon C. Dik's stimulating study of co-
ordination 5 shows how much work has to be done on even such a
familiar and apparently straight-forward word as "and." At this
stage of the investigation of Hebrew syntax, I have been con-
tent to point out a dozen different ways in which "and" is used
to join clauses together in that language.
This is only a beginning. An enormous amount of additional
work is called for, especially on three fronts. First, the the-
oretical basis must be strengthened by more careful work on
deep grammar. The present quest for language universals will
assist this, and open up the way for the investigation of com-
parative syntax within the Semitic family. Secondly, empirical
testing must be expanded to cover all known ancient texts. (Most
of our examples come from the Torah, and abundant evidence has
not been traced very far.) Thirdly, when the evidence is organ-
ized, we must look at the dynamics of structural changes in the
sentence repertoire, and higher up the hierarchy, in order to
write the history of Hebrew discourse and in particular the his-
tory of Hebrew conjunctions. When this work has been done we will
be able to return to problems of translation and exegesis, and
such knowledge will equip us for the task of literary criticism
and for the application of linguistic arguments to the identifi-
cation of sources and the dating of documents.
NOTES
1
Historisch-kritisches Lehrgebude der Hebrischen Sprache
mit comparativer Bercksichtigung des Semitischen berhaupt,
a u s g e a r b e i t e t v o n F r . E d u a r d K n i g . 2. H l f t e , 2 . ( S c h l u s s - )
Teil: Syntax (Leipzig, 1897).
2
For example fB lest is t r e a t e d as an i m p e r a t i v e v e r b j o i n e d
t o t h e f o l l o w i n g v e r b in a s y n d e t o n (p. 1 3 l ) , w i t h t r a n s l a t i o n s
t h a t r e f l e c t t h e o r i g i n a l m e a n i n g of t h e p u t a t i v e root . Quite
apart from the d u b i o u s e t y m o l o g y , t h i s o b s c u r e s the fact that
f r o m a s t r u c t u r a l p o i n t of v i e w |S is a c o n j u n c t i o n .
3
F o r t h e b i b l i o g r a p h y o f t a g m e m i c s s e e K e n n e t h L . P i k e , "A
G u i d e t o P u b l i c a t i o n s R e l a t e d to T a g m e m i c T h e o r y , " Current
Trends in Linguistics: Vol. Ill: Theoretical Foundations, Thom-
as A . S e b e o k , e d . (The H a g u e , 1 9 6 6 ) : p p . 365-39 1 *; R u t h M . B r e n d ,
" T a g m e m i c T h e o r y : A n A n n o t a t e d B i b l i o g r a p h y , " Journal of English
Linguistics, 1 ( 1 9 7 0 ) : p p . T - U 5.
M o s t r e c e n t l y in R o b e r t E . L o n g a c r e , Hierarchy and Univer-
sality of Discourse Constituents in Hew Guinea Languages. Vol.
I: D i s c u s s i o n ; V o l . I I : T e x t s . (Washington, Georgetown Univer-
20 INTRODUCTION
sity Press, 1972), where the bibliography will give sufficient

guidance to preceding discussion.
5
Coordination: Its implications for the theory of general
linguistics (Amsterdam, 1972).
2
THE SENTENCE IN HEBREW
2.0. DEFINITION OF SENTENCE
The term SENTENCE (Se) has been used in linguistics in several

different ways. The traditional definition of a sentence as "a
complete thought expressed in words" fell on evil times when
description of language data began with forms rather than ideas.
It was easier to observe a sound than a thought, and it was im-
possible to tell when a thought was 'complete.'
As a unit in the phonological component of a language, sen-
tence is the name for a set of elements whose distinctive iden-
tity is marked by phonological features. Each element in the set
is a sentence. The contrastive-distinctive features of a phono-
logical sentence, while language-specific in detail, are charac-
teristically melodies of tones and stresses, with certain contours
having particular importance as signals of sentence termination.
Segments of speech with such melodies are often uttered in com-
plete isolation from other speech, or in conversation as the al-
ternating contributions of individual speakers. Unfortunately we
know nothing whatever about these matters, so far as biblical He-
brew is concerned, for it is a dead language. There is a reason-
able presumption, however, that the complex Masoretic systems
(there are two quite distinct ones) 1 preserve some of the tradi-
tions of living speech, for the scriptures never ceased to be
read. The systems are unfortunately of little value, for they
are geared to written texts, they reflect liturgical artifici-
ality and the units of 'verses' and subverses bear only partial
correspondence to sentences as grammatical units.
As a unit in the grammatical component of a language, a sen-
tence may be defined as a construction that is grammatically com-
plete or self-contained; that is, the grammatical functions of
all the elements in a given sentence can be described in terms of
relationships to other elements within the same sentence.
Such definitions take us somewhere, but they do not take us
very far. The set of sentences for any language, identified as
units in speech by phonological criteria, is likely to include
a variety of construction types from the grammatical point of
view, ranging from a single word to an extended text. (We use
TEXT to refer to any given specimen of a language, spoken or
22 THE SENTENCE IN HEBREW
written.) Grammatical completeness, as a sine qua non of sen-

tence identity, may prove as hard to establish as completeness
of thought. If we can identify parts of a sentence as elements
of that sentence, then these ingredients, in their own way, will
have some measure of internal integrity that permits their iso-
lation. A WORD (Wd) , a PHRASE (Ph) , a CLAUSE (CI) has its own
internal structure or completeness. And few sentences, however
complete within themselves, are likely to be as entirely without
relationships to their context as the definition requires. Bloom-
field's famous definition--"a sentence is an independent linguis-
tic form, not included by virtue of any grammatical construction
in any larger linguistic form"--if applied strictly, would iden-
tify the unit we call DISCOURSE (D). Most sentences function
within larger discourse of some kind, to say nothing of the non-
linguistic behavioural context. Grammatical completeness is there-
fore a matter of degree, and cannot be made an absolute criterion
for the identification of sentences.
Nevertheless the definitions are not without value. Many phono-
logical sentences are able to stand by themselves in speech be-
cause they have some measure of grammatical completeness. Hence
the need to distinguish the kinds of grammatical relationship
that have their realization within the structure of a sentence
as such from the kinds of grammatical relationship that operate
between sentences in larger chunks of discourse.
While their hierarchies have their own characteristic struc-
tural features , the phonological and the grammatical components
of a language are likely to have some measure of congruence. To-
gether with the semantic component, they have such an ineluc-
tibly concomitant operation in the one thing that any language
is, that they cannot be separated, even though they can be talked
about separately. But the hierarchies are not isomorphous. Hence
sentence, as a unit in each component--a meaning complex, a gram-
matical construction, a sound pattern--needs an appropriate def-
inition in terms of the features of that component, and we should
know what we are doing when we bring together the results of
these definitions. 2
2.1. THE CLAUSE AS A SENTENCE ELEMENT
We can stalk the sentence from another direction by talking

first about the clause as a unit in the grammatical component of
a language. Many languages afford constructions which contain
two parts which can be described as the TOPIC and the COMMENT.
Again, many languages afford constructions which, in terms of
formal grammatical features, can be described as made up of a
SUBJECT (S) and PREDICATE (P). It often happens that the topic
is realized as the subject, while the predicate realizes some
comment on the topic. Readers of linguistic literature will know
at once that we are already on slippery ground, but we do not in-
tend to explore the familiar yet treacherous frontier between
meaning and form in this regard. We accept for the time being
subject and predicate as mutually self-defining co-occurring gram-
matical functions whose forms of realization are specific for
THE SENTENCE IN HEBREW 23
each language. When the language has the means of realizing a

relationship of PREDICATION () between a subject and a pred-
icate, then the construction in which S is realized is de-
finitive of a clause. We are careful not to define a clause as
+S +P, for this minimum NUCLEUS (Nuc) may have in addition
various MARGINAL (Mg) elements which are not part of either sub-
ject or predicate and also inter-clausal CONJUNCTIONS (Cj). Hence
CI = Cj +Nuc (Mg) n
More precisely, a construction in which predication is real-

ized once is a clause. A more rigorous definition would have to
protect this statement by restricting the form of realization
of predication to surface structure. John is crazy is an ascrip-
tive clause; crazy John is an attributive phrase. In each con-
struction we are talking about John and we are saying that he is
crazy. But the difference lies in the use of the explicit formal
(surface) equative predicator is for a clause, but not for a
phrase. So all kinds of 'deep' predication, as well as merged
or embedded clauses which realize various kinds of clause-level
tagmemes on the same level as the main predication, do not count
for purposes of the definition. A single clause may therefore
have more than one explicit predication realized in its surface
structure, but it is only one clause so long as the main (or
lead) predication can be distinguished from predication in clauses
which, by recursion, realize clause-level tagmemes within the
predicate of the lead clause, or which, by further recursion, are
included within an included clause, or which, by back-looping
(rank-shifting) are part of a clause-level noun phrase; etc.
A problem lurks here also, of course. For how firmly does a
second clause have to be attached to a lead clause before we say
that it is 'inside' it and not just 'alongside' it? We shall re-
turn to this question when we have developed the means for han-
dling it. Meanwhile we assume, without more ado, that the clause
is an identifiable unit in the grammatical hierarchy. If we then
identify in a given text the elements which are clauses, and com-
pare them with the elements identified as sentences, the set of
sentences is likely to consist of three subsets: a subset of sen-
tences, each of which is less than one clause; a subset of sen-
tences, each of which is just one clause; a subset of sentences,
each of which is more than one clause. Even if we continue to
call all of them sentences, we will still need to recognize that
there are three distinct kinds of sentence, so far as their in-
ner structure is concerned.3
In traditional grammar a one-clause sentence is called a sim-
ple sentence. This is unobjectionable; but we prefer to call such
a construction simply a clause, and to restrict the use of the
term '(grammatical) sentence' to a unit higher than clause in
grammatical rank. This can be done without confusion if we call
the set of 'complete' utterances PERIODS. The period is the min-
imum unit of discourse.
For a start we shall say that there are some constructions in
which two clauses are related to each other in certain ways and
which we propose to call sentences. This is not a definition of
2h THE SENTENCE IN HEBREW
sentence. It leaves open for study three questions: (i) What

kinds of two-clause constructions are sentences --all, or only
some? (ii) If only some, then what kinds of inter-clause rela-
tionships are sentence-level, and what kind are not? (iii) Are
other elements besides clauses related to each other in senten-
ces?
2.2. SENTENCE TYPES
Sentences containing two or more clauses have been tradition-

ally classified as COMPOUND or COMPLEX on the basis of a broad
distinction between COORDINATION (Co) and SUBORDINATION (Sub).
APPOSITION (A) was included under 'compound.' As we shall see,
this simple distinction cannot do justice to the range of in-
ter-clausal relationships provided for in the sentence reper-
toire of the natural languages. But the problem is not solved
simply by multiplying categories. On the one hand, an absolute
distinction between coordination and subordination is not easy
to nail down; on the other, there are different kinds of coor-
dination sentences and different kinds of subordination senten-
ces. The differential diagnosis of sentence types can be carried
on to any desired degree of delicacy, with both surface and deep
grammar helping to establish the typology. But if no limit is
set for the descent into semantic structure, a point may be
reached where each individual sentence has its own grammar. This
is what one must do eventually in exegesis. The present task
must strike a compromise, aiming at useful generalizations on
the basis, in the first place, of contrastive-distinctive fea-
tures in surface grammar. The price paid for standing back to
see the big picture is that finer details have to be left for
close-up viewing.
As a first step towards the classification of sentences, we
shall remove from the scene some construction types which do not
qualify as sentences according to our definition. There are some
clause-like constructions which do not qualify for recognition
as sentence-level clauses, and so are not constituents of sen-
tence structure as such. Hebrew has three kinds of verbal noun--
an INFINITIVE (If), a PARTICIPLE (Pt), and names of agents and
activities which have deep-structure affinities with their cog-
nate verbs. The participle is the most verbal of these, in that
it can form the predicate (or function as the head of the pred-
icate) in the nucleus of a well-formed clause, that is, a clause
with an explicit grammatical subject. Such participial clauses
do enter our purview when they are sentence constituents (#5.1.
3.1). A participle can also be used alone, that is unmodified,
or as the head of a predicate-like construction (a PARTICIPLE
PHRASE [PtPh]) in which other clause-level tagmemes, such as
OBJECT (0), LOCATION (L) , TIME (Ti), etc., are realized in re-
lationships typical of well-formed clauses. Such a participial
phrase (incomplete clause) generally functions as a NOUN EQUIVA-
LENT (NEq); the participle is verbal within the phrase, nominal
outside it. The noun-like functions of a participial phrase in-
clude the realization of such clause-level tagmemes as subject,
object, OBJECT COMPLEMENT (OC), etc. Some linguists call such in-
corporation of an incomplete clause into a larger clause 'merging'
or 'embedding.' By defining a sentence as a construction with at
least two well-formed clauses, we exclude merged sentences from
consideration.
The same applies to all infinitival constructions, even though
in some infinitival phrases an explicit subject is realized, the
infinitive being the predicate.'* Although such a construction might
seem to qualify as a well-formed clause, it is a clause type to be
studied in its own right. Its hierarchical functions are only mar-
ginal to our present discussion. An infinitival phrase rarely man-
ifests the clause property of occurring alone as a complete utter-
ance or as a simple sentence. (We are not talking about well-formed
clauses in which the predicator is an INFINITIVE ABSOLUTE [IAJ,
functioning as a universal verb.) And since an infinitival phrase
is not normally coordinated or subordinated in the usual way with
a regular clause, it does not figure as a sentence constituent.
On the contrary, infinitival constructions are typically clause-
level, realizing such tagmemes as Time, Purpose, Result, etc. For
this reason we do not need to go outside the domain of clause to
describe their grammatical functions. An infinitival phrasj can,
however, be the head to which a clause may be coordinated. See
#5.2.1.
Example: wayyism?^ êt-ql yhvh êlhlm
(+Cj +S+P):WP +0:PpPh
and they heard the sound of Yahweh God
mithallek baggn
+0C:PtPh
walking around in the garden (Ge 3)
Mithallek baggn is a predicate-1 ike construction (P = +V:Pt +L:
PpPh) which can be transformed into a surface predicate by 0 + S
and OC P. The resulting cl ause is (qol) yhwh êlShlm mithallek
baggn, (the sound of) Yahweh God Zisl walking around in the gar-
den. This has deep structure in common with Ge 3. In Ge 3 the
object of the verb is the (deep) subject of the participle. It is
a merging of the coordination sentence: They heard the sound of
Yahweh God fand) Yahweh God was walking around in the garden. We
do not include such alternative realizations of deep-structure
coordination in the present study.
In a QUOTATIVE (Q) clause the reported quotation is the object
of a quotative verb, and may be practically anything, including
a clause. Such a quoted clause simply realizes a clause-level tag-
meme, like any other object, and is integral to the total clause,
so that the whole construction cannot be compared with one in
which two distinct clauses are put together in coordination or
even subordination. Hence we do not call a quotative clause a
sentence.
A 'relative' clause may be used as an adjectival modifier of
a noun head. Such clauses, NOMINALIZED (Nom) in Hebrew by the use
of "lWK, exercise other noun-like functions, realizing such clause-
level tagmemes as subject or object, or in prepositional phrases
which realize various phrase-level or clause-level tagmemes. None
of the relationships into which these nominalized clauses enter

is inter-clausal, that is, sentence-level. So these are elimin-
ated from the present discussion. If is used as a subordin-
ating conjunction, an exception will have to be made. This prob-
lem can be shunted into the lexicon by recognizing in "iwk a pair
of homonyms, the nominalizer ('relative') and a conjunction.
2.3. COMPLEX SENTENCES
A complex sentence is a two-clause construction in which a

SUBORDINATE CLAUSE is dependent on a principal or LEAD CLAUSE.
Like many other languages, Hebrew has a set of SUBORDINATING
CONJUNCTIONS (SubCj) which help to signal the various dependent
relationships of subordinate clauses to lead clauses. There can
be no objection to calling such a construction a sentence, but
the primacy of the predication in the lead clause permits the
whole construction to be described as one entire clause, whose
nuclear predication is that of the lead clause while the sub-
ordinate clause realizes some marginal clause-level tagmeme or
other within the lead clause. The subordinate clause is clause-
level, not sentence-level. Such constructions can be advantageous-
ly classified with clauses which have only one predication in sur-
face structure rather than with sentences whose structure is above
clause level. Detailed justification of this tactical decision
need not be given here; a few remarks will give an idea of the
lines along which such a justification would proceed.
A typical clause-level but non-nuclear tagmeme like TIME (Ti)
can be realized by a variety of forms. Potential fillers of the
Ti slot are members of several parts of speech or form classes,
such as PRONOUN (for example, r tt, now), ADVERB (ymm, by day),
NOUN (layla, by night), NOUN PHRASE (ym ym, day by day), PRE-
POSITIONAL PHRASE, etc., all unquestionably clause-level and in-
tegral (albeit marginal) to the clause that they are in. There is
no great shift in construction type when a time tagmeme is real-
ized by a subordinate clause such as ki zrefc hassemes, when the
sun rose. To call the latter sentence-level means that the form
rather than the function of the time tagma has been used to class-
ify the construction.
There is no need to argue about the use of words. If a construc-
tion which includes a subordinate clause is called a (complex)
clause, it is a member of a distinct subclass of clauses that is
worth studying on its own. If it is preferred to call such a con-
struction a complex sentence, it is still a special subclass of
sentences distinct from other kinds of two-clause construction.
In so far as a subordinate clause is optional, that is, not part
of the nucleus of the main clause, the complete construction mi-
nus the subordinate clause is itself likely to be an acceptably
complete clause. Hence the subordinate clause can be described
as a second clause added to the first to make a two-clause con-
struction (sentence) rather than included in the first to be an
integral part of a whole which is still only one clause.
The point is a fine one, and we shall return to it again. Mean-

while we recognize a distinct construction type (call it clause
or sentence) which is the domain of such relationships as Cause
or Reason (the subordinate clause states the reason for what is
stated in the lead clause)(Cj:3, because), Effect or Result
(73V, therefore), Purpose (jyeV, so that), Negative Pur-
pose ( Tf lest, so as not), Concession ( 3, although), Condi-
tion (ox, 3, if), Time (3, when; in*, after; riy, until), etc.
The characteristic conjunctions shown signal these relationships.
2.4. COMPOUND SENTENCES
The traditional definition of a compound sentence recognizes

in it two 'principal' clauses. It does not matter whether they
are coordinated or stand in apposition. But the picture is not
nearly so simple. It would be better to make a distinction be-
tween COORDINATION (<Co>) and APPOSITION (<A>). In a coordina-
tion sentence, two (or more) clauses are joined by a coordin-
ating conjunction, typically 'and.' In an apposition sentence,
two (or more) clauses are placed in juxtaposition without a
conjunction.
There is more to be said than this, of course. The rest of
this book is devoted to this task. Here we make only three pre-
liminary remarks. First, it is a fact readily observed (in most
natural languages, we suspect) that many coordination sentences
remain equally acceptable if the coordinating conjunction is not
used, and many apposition sentences remain acceptable if a coor-
dinating conjunction is added. But this does not mean that the
categories are indistinguishable. If experimental limits are
found, and some coordination sentences do not survive the exci-
sion of the conjunction, or some apposition sentences do not sur-
vive the addition of a conjunction, then these constructions are
definitive of the sentence type. At the same time an apposition
sentence can be an alternative surface realization of a coordin-
ation relationship, and a coordination sentence can be an alter-
native realization of an apposition relationship (see #3.0).
Hence, in classifying such sentences, attention must be paid to
the deep relationships as well as to the surface features.
Secondly, the use of a conjunction or of NOTHING (0) to join
two or more clauses together does not tell the whole story. Un-
less the discourse is incoherent, some kind of signal joins clau-
ses together in a sentence, even if they stand in apposition;
otherwise it would not be possible to speak of the sentence as a
unit. Such signals are likely to operate on a coordination sen-
tence concomitantly with the conjunction. Because of these sig-
nals the conjunction can be dispensed with without dismantling
the construction. And these inter-clause signals are as much a
part of the grammar of the sentence as the conjunction.
Thirdly, the traditional definition of a compound sentence as
made up of two potentially independent clauses is not likely to
obtain in many cases. Witness the common sequence of noun fol-
lowed by anaphoric pronoun. Such clauses are likely to be depen-
dent without being subordinate. Everything that contributes to

the interdependence of clauses in a sentence has to be studied
as a part of sentence grammar.
2.5. RELATIONSHIPS BETWEEN SENTENCE TYPES
So far, using conjunctions as our chief clue, we have recognized

three main kinds of sentence.
Apposition <A>
Coordination <Co>
Subordination 
The connection between the constituent clauses is increasingly

tight as we descend the series, and we could continue into even
tighter attachment when an embedded clause is merged right into
the structure of the main clause, as in an infinitival construc-
tion that is used as an alternative realization of, say, a sub-
ordinate clause of Purpose. The connection is not uniformly tight
within each main type. In due course we shall distinguish several
kinds of apposition sentences, ranging from those in which the
clauses are closely linked to those in which the connection is
quite loose. Similarly, there may be varying degrees of depen-
dence in coordination sentences, and the most dependent coordin-
ated clauses might just as well be called 'subordinate.' We
have quantized the spectrum into three bands on the basis of
differences in gross surface features, namely, the use of con-
junctions. Formal criteria for distinguishing coordination from
subordination are discussed in Dik (pp. 34-41). There are three
tests, (i) If M! co co? M 2 then co? is not a coordinator. (ii)
The lead clause in coordination cannot have anaphora. (iii) If
Mi co M 2 then M 2 co M! is possible. The permutation and the an-
aphora test won't work on one and the same construction. As we
shall show in Chapter 12, test (i) is not decisive for Hebrew.
Furthermore, Hebrew does tolerate anaphora in the lead clause
of a coordination sentence. And not all Hebrew subordinate
clauses can switch order with their head.
Dik's criterion is admirably empirical. But the price paid
for this objectivity is that it is too language specific, and
too many surface features are arbitrary in natural languages
to be taken as a guide to universals. Attention to deep struc-
ture relationships would keep English 'for' and 'because' to-
gether. Dik's criterion separates them. We shall be interested
to discover for Hebrew to what extent the different parts of
the spectrum correspond to a continuum in deep structure.
The example we have chosen to illustrate this is particular-
ly interesting, because the dependent clause involved is attri-
butive to a preceding noun, and could be realized as a phrase-
level relative clause.
2.6. SURFACE GRAMMAR AND DEEP GRAMMAR
2.6.. The English Relative Clause

In English, information about a person's name can be supplied
in various ways. girl, Mary by name; a girl named Hary; a
girl they call Mary; a girl whose name is Mary; a girl--her
name is Mary; etc. We shall not attempt to discuss the English
grammar of all these constructions. They represent different
formal (surface) means of realizing essentially the same (deep)
relationships between Mary, name and girl. This aspect of lang-
uage has been extensively discussed in recent linguistics, es-
pecially in transformational-generative grammar. We shall not
attempt to review the question of how much the different sur-
face structures, each of which is a transform of any of the
others, secure differences of meaning. All we need to do, in a
general way, is to distinguish surface realizations (texts)
from underlying semantic materials and deep grammatical rela-
tionships. In particular, we shall not attempt to locate the
interface between semantic structures and deep grammar. We need
only the general ideas that the same deep grammar may be real-
ized in various surface constructions, and that the same sur-
face construction may sometimes realize more than one deep
grammar relationship.
Consider the phrases in English:
a man, whose name is/was Job
the king, whose name is David
Esau, whose name is Edom
you, whose name is Yahweh
These all have the structure
(antecedent)Head:NEq<A>ModifierTRelCl
They differ only in the class-membership of the various fillers
of the Head slot, which are, respectively, an indefinite noun,
a definite noun, a proper noun and a pronoun. The relative clause,
nominalized by 'whose', is in apposition with the Head nominal,
and realizes a phrase-level tagmeme. English relative clauses of
this kind are now confined to phrase-level functions, and an ex-
plicit antecedent is obligatory.
*whose name is David is king
He, whose name is David, is king
In view of the availability of a widely used 'relative' nwK, who,
which, etc., in Hebrew, one might have predicted the grammatical-
ity of literal translations. In Hebrew the corresponding construc-
tions would be
31 X 1WK vx*
low nwx i>an*
"iK iaw -iwx iwy*
' law i w *
In fact, no phrases with this structure are attested anywhere

in classical sources. They are, however, found in Aramaic.
Example: Daniyye^l di s?meh Belteâ"155 ar, Daniel, who his-name
(is) Belteshazzar (Da Z 2 6 , 4 8 , 1 '). In the Hebrew part of Daniel
there is a construction that looks like an attempt to translate
this Aramaic phrase into Hebrew: Dniyye^l ^Ser niqr^ 55m
Belt5?a^ssar, Daniel who is-called his-name Belteshazzar
(Da 10 1 ). This clause is without parallel in the rest of the
Bible. As we shall see in what immediately follows, in specify-
ing a name, the relative pronoun as used in both English and
Aramaic is avoided (except in #2.6.8). In Da 10 1 the structure
of the presumed Aramaic original has been followed, except for
the addition of the verb, even though verbless clauses are com-
monplace in Hebrew, and are otherwise used for such statements.
2.6.1. Head is a Proper Noun
An alternative proper name for a person or place is given as

in f esw h 1 êdm, Esau, he (is) Edom (Ge 36 1 ) , or 'en miSp^
hl 1 qdes, En Mishpat, she (is) Qadesh (Ge 14 7 ) , etc. This con-
struction is best construed as a verbless clause in parenthesis
in the discourse, or in apposition with the head noun.
NPh
Head N : N p ^ <A> \Modifier:Cl
S: Pr :Np
2.6.2. Head is a Pronoun
The only instance in the Old Testament in which the antecedent

is a pronoun is Ps 8 3 1 9 : âtt Jimk yhwh, you (who) your-name
(is) Yahweh. This has the same structure as #2.6.1 with S:Ns.
2.6.3. Head is a Definite Noun
The only construction which has a definite noun as antecedent

is hammelek yhwh 5?m, the king, Yahweh Sebaoth (is)
his name (Je 4 6 1 , 4 8 1 5 , 5 1 s 7 ) . This has the structure
NPh
2.6.4. Head is an Indefinite Noun
When the antecedent is an indefinite noun, two different con-

structions are available. Zee 6 1 2 resembles #2.6.3 above:
^ semah SSm5, a man. Branch (is) his-name.
Other examples: ISa 1 7 " 2 3 , 2Sa 20 2 1 , IKi 1 3 2 , Job l 1 , 2Chr 28 9 .
The sequence SP is used in the apposition clauses in #2.6.1

and #2.6.2; the sequence PS is used in the apposition clauses in
#2.6.3 and #2.6.4. The latter two invite comparison with the lit-
urgical yhvh fSm, his name is Yahweh, which comes at the end of
a list of titles. These compositions have been analysed gramma-
tically in various ways. Ex 15 3 is generally taken as two clau-
ses- -Yahweh is a man of war; Yahweh is his name. We suggest it
is more likely to be a recitation of three names (or a three-
fold name)--Yahweh, Man of War, Yahweh is his name. The hymns
in Am 4 1 3 , 5', 9 s are generally explained as CASUS P E N D E N S H e
who shapes the mountains...--Yahweh is his name. It is more
likely that they are of the same genre as Ex I S 3 , and that the
entire list is the name of God. See also Ex 3 4 ' V Yhvh qn 1 Smv
is correctly construed as He whose name is Yahweh ann'.
2.6.5. Coordination as Alternative Realization
Otherwise the information about the name is supplied in a cir-

cumstantial verbless clause (#5.1.3.3) with sequence SP as in
#2.6.2, that is the reverse of #2.6.4.
Example: velh siph misrlt semh hgr, and she had an Egyp-
tian slave, and her name was Hagar (Ge 1 6 ) .
The structure is
Antecedent:Ni <Co> CirCl
w?- S:Ns :Np
The antecedent head indefinite noun, which may be a phrase like

f
ebed misrl, Egyptian slave (IChr 2 i " ) , gives the class member-
ship of a new character.
man (Ge 3 8 1 2 , Jdg 13 2 , 1 7 l , ISa l 1 , 9 1 , 1 7 1 2 , 21,

2Sa 1 6 s , 2 0 1 , Est 2 5 )
H S s , woman (Ge 25 1 [compare Ge ll 2 '], 38, Jdg 16", Jos 2 1 ,
IChr 2 2 S )
ben, son (ISa 9 2 , 2 2 2 0 2Sa 4", 9 1 2 , 1 7 2 5 )
2
bat, daughter (2Sa 14 ')
''ah, brother (Ge 24 2 9 )
^ ljt , sister (2Sa 1 3 1 )
bafal, master (Je 3 7 1 3 )
mda^, acquaintance (Ru 2 1 )
r e f , friend (2Sa 1 3 3 )
f
ebed, male slave (2Sa 9 2 , IChr 2 3 1 t )
siph, female slave (Ge 1 6 1 )
piliegeS, concubine (Ge 22 2 *, 2Sa 3 7 )
32 THE S E N T E N C E IN HEBREW
The construction is highly favoured in narrative prose for intro-

ducing a character at the beginning of a story, or for bringing
in a new character along the thread. The preference for a circum-
stantial rather than a relative clause brings the construction tc
a higher level in the hierarchy and highlights the name. Compare
the similar circumstantial clauses in Ge 4 ? 1 and IChr 7 1 .
2.6.6. An Exception
Is 57 1 5 is usually interpreted as if it were a variant of #2.6.3

and #2.6.4 above, with, however, coordination rather than apposi-
tion. In this latter regard it resembles #2.6.5. But the sequence
PS is as in #2.6.3 and #2.6.4, not as in #2.6.5. Rm weniSS' S-
ken ^d veqd? f?m, the high and lofty one that inhabits eter-
nity, whose name is holy. This is dubious. The sequence is wrong
for comparison with #2.6.5; the use of the coordinating conjunc-
tion deviates from #2.6.3 and #2.6.4. The problem can be solved
in terms of established structures by recognizing another reci-
tation of the names of God as in Ex 15 3 , etc., except that here
the titles are coordinated and not just listed. There is but one
clause, with sequence PS. The predicate consists of four names,
coordinated in two pairs. He whose name is Elevated and Exalted,
Occupier of eternity and Holy One. In the light of this, the con-
structions cited in #2.6.3 above, together with Am 5 2 7 and the
like, are to be construed in the same way. This leaves #2.6.4 and
#2.6.5 as the normal ways of supplying a character's name, that
is, alternative realizations of relative clauses. They are syn-
tactically quite different; but there are no distribution pat-
terns that enable us to assign them to different historical per-
iods, to different regional dialects, or to different literary
genres, except that #2.6.4 is not used in the Pentateuch.
2.6.7. Naming Two Persons
It often happens that protagonists are introduced as a pair

whose parallel destinies will be played out in the following
narrative--two wives (Ge 4 1 9 , 1 1 , ISa l 2 Ru l1*) ; two daugh-
ters (Ge 29 1 6 , ISa 14*t9) ; two sons (Ge 1 0 2 i , IChr l 1 '); two men
(Nu l l 2 6 , 2Sa 4 2 ) , etc. The construction preferred for this kind
of material is illustrated by Ge 4 1 9 .
oWJ
my nnxn ow
^ niwn obi
two wives
the name of the one - Ada
and the name of the second - Zilla
Antecedent:NPh <A> Conjunctive Sentence
The differences between this construction and those in #2.6.4

and #2.6.5 are remarkable. Both of the conjoined clauses have
the same sequence (SP) which is the same as that used in #2.6.5.
But the conjunctive sentence is in apposition, as in #2.6.4. So
the result is not the same as either, while partly resembling
each. If #2.6.4 had been followed, we might have expected *sete
nslm <A> ''d vegill Jimtehen. If #2.6.5 had been followed,
we might have expected *sete nSIm Smtehen fd wesill. But
constructions of this sort are never found in the Bible, even
though there is no reason to suppose that either of them would
have been considered ungrammatical by native speakers. It is
not a frozen device. A variety of vocabulary is used for refer-
ring to the participants.
h^ ehd ..hêhd (Ex 183 . ISa 141*)

h^ eljd ..ha??lnr (Nu 1 1 2 , 2Sa 4 2 )
1
ah at ..haSSenlt (ISa l 2 )
h^ ahat ..hassenlt (Ge 4 1 9 , Ex l 1 5 , Ru 1")
h^ ehd ^ ahlv (Ge 10 2 5 , IChr l 1 9 , cp . Ge 4 2 1 )
haggdl . .haqqânna (Ge 2 9 l s )
habbSklr . .haqqStann (ISa 14 )
1
eSet-^ abrm . .">e?et-nl}r (Ge I I 2 9 )
Similar patterns are found in ISa 8 2 (Samuel's two sons), ISa

253 (Nabal and Abigail). Compare IChr 7 1 5 , and the names of
Job's three daughters in Job 421 **, where, however, the con-
struction is the object of a verb. A similar pattern of apposi-
tion followed by coordination is used for the names of the four
rivers of paradise in Ge 2 1 1 " 1 * , but the fourth clause switches
to the use of a SUSPENDED (Sus) followed by a RESUMPTIVE (Res)
subject. The names of Azel's six sons are given quite differ-
ently: Siss bnlm wSêlle semStm..., six sons, and these (are)
their names (the list follows) (IChr 8 3 , "''). Ru l 2 also shows
that #2.6.7 is not used for more than two names. Here three
clauses are coordinated as in #2.6.7, but the whole is coor-
dinated as in #2.6.5.
2.6.8. The Use of the Nominalizer
There are two passages in which information about two names is

connected to an antecedent by the relative "WK.
THE SENTENCE IN HEBREW
nnayn
nna tbx
(Ex l 1 5 ) nyio n*urn osn
to the Hebrew midwives

who the name of the one (was) Shipra
and the name of the second (was) Pua
Also Moses' two sons (Ex 18 3 ). It could be significant that the

antecedents here are definite, but this is not enough to explain
the use of ibx ; for the antecedent is definite in ISa 14 2 9 and
nwx is not used, and can be used in other constructions when
the antecedent is indefinite.
The use of the nominalizer brings the construction down the
hierarchy to the level of phrase, as the following diagram shows.
NPh
2.6.9. Conclusions
To sum up. The deep construction - his name is y can be real-

ized by several different surface structures in Hebrew.
For one person: +X <A> +C1:(+P:Y +S:55mo)(#2.6.4)

or +X <Co>:- +Cl:(+S:sem +P:) (#2.6.5)
For two persons: +X's <A> +Se: (+C1![SP] <Co> +C1 2 [SP]) (#2.6.7)
or +X's <Rel> +Se:(Cl![SP] <Go>+Cl 2 [SP])(#2.6.8)
There is both freedom and constraint in this system. For one per-
son, coordination of a SP clause is preferred, with apposition of
a PS clause as an acceptable alternative. For two persons, a con-
junctive sentence (both clauses with SP sequence) may be in appo-
sition or nominalized. For a larger cast (Ge 2 1 1 " 1 " , Ru l 2 , IChr
8 3 8 , 9****) other constructions again are used. The use of the same
semantic materials with the same grammatical functions (S:S?m,
P:Np) gives a methodological control, and highlights the arbi-
trary nature of the various surface realizations which are pre-
ferred for each number of participants.
A point to emphasize is the different placement in the gram-
matical hierarchy of the alternative realizations. A nominalized
('relative') clause (#2.6.8) is phrase-level. A coordinated ('cir-
cumstantial') clause (#2.6.5) is sentence-level. A clause ('spe-
cifying ')(# 2 . 6 . 4) or sentence ('expository') (#2.6.7) in apposition
is more detached, being virtually parenthetical on paragraph-level.
There is a further moral in this. An induction of the most gen-
eral kind would lead to a descriptive statement that all the con-
structions listed above 'occur' or are 'possible.' But as soon as
such a formulation is used generatively without constraints a false
impression is given. The formula will be overproductive. Thus SP

and PS sequences are both used; <Nom>, <Co>, and <A> are all
used. But the combinations actually found are restricted. Some
constructions produced by a general formula are not used at all,
and some are less common alternatives. And the latter cannot be
accounted for by statistics. Here the matter passes over to rhe-
toric, and questions must be asked about the different effects,
staging, for example, of the alternatives available.
The example also illustrates a practical problem in the organ-
ization of a grammar. Beginning with surface features as a basis
for classification, constructions with similar deep structure
would be scattered. Beginning with the same deep structure would
gather together under the same heading constructions which are
formally diverse. In the present study we prefer to describe
surface features systematically, making cross-references to oth-
er constructions which seem to be alternative realizations of
the same deep structure.
NOTES
1
William Wiekes, Treatise on the Accentuation of the three
so-called Poetical Books of the Old Testament, Psalms, Proverbs
and Job (l88l); and. A Treatise on the Accentuation of the Twenty-
one so-called Prose Books of the Old Testament (1887). Reprinted
in one volume by KTAV Publishing House (1970).
2
Harris' definition of 'sentence' is at once phonological and
grammatical: "sentences may be characterized as those segments of
speech (or writing) over which certain intonations occur or with-
in which certain structures occur" (String Analysis of Sentence
Structure: Papers on Formal Linguistics [The Hague: Mouton,
1962D: p. 7). The two alternative criteria yield two sets of
segments called 'sentences'. Even if the two sets largely over-
lap, as they'probably will, the fact remains that we are talking
about two distinguishable, albeit inseparable, components of
language.
3
Confusion can be diminished if we call a unit in discourse
that is 'complete' a PERIOD. A period may be non-predicative
(less than one clause); it may be one clause; or it may be a
'sentence' defined grammatically as an integral coirstruction
of two or more clauses.
* Or head of the predicate (predicator) . If predicate is de-
fined as all in the clause nucleus that is not the subject, then
a predicate without a subject is a contradiction in terms, at
least so far as surface structure is concerned. Infinitives
often realize deep-structure predication without surface-
structure clause formation.
3
APPOSITION SENTENCES
3.0. DEEP GRAMMAR OF APPOSITION
Sentence-level apposition resembles phrase-level apposition. A

typical phrase-level apposition places in juxtaposition two words
or phrases with a common referent, or at least some overlap in
their fields of reference. Examples: Dr. Livingstone, the explor-
er; red apple. Dr. Livingstone and the explorer are identical.
The phrase in apposition identifies the Head. In Dr. Livingstone,
an explorer, the apposition phrase classifies the Head. Red and
apple are both class names; the apposition phrase refers to the
overlap class of objects which are both red and apple.
Juxtaposition without a conjunction is found also in phrases
which may be called lists. These are alternative realizations of
some, kind of coordination. Conversely, when two items in apposi-
tion are joined by a coordinating conjunction, the resulting con-
struction embodies a figure called hendiadys. Thus we can disting-
uish the normal surface realization of a deep structure relation-
ship from aberrant, but not ungrammatical, alternative surface
realization.
Surface form Juxtaposition Concatenation
Deep relationship Apposition Coordination
An apposition sentence consists of two or more clauses in jux-

taposition (no conjunction). Such sentences may be arranged on a
spectrum depending on how much semantic overlap there is between
the constituent clauses. At one end of the spectrum the two clau-
ses may be completely identical (#3.2). At the other end, the two
juxtaposed clauses may have nothing in common at all. At this ex-
treme we have incoherent discourse, and there is no point in talk-
ing about a sentence when there is rio coherent structure. Most
apposition sentences lie somewhere between these two extremes.
APPOSITION SENTENCES 37
3.1. SURFACE GRAMMAR OF APPOSITION
The juxtaposition of two clauses is the simplest form of apposi-

tion sentence. Resumptive kl, yea'. may sometimes intervene (Ex
31 1 ).
Se
Although we have used clauses in apposition to illustrate the

construction, there is no constraint on the items in juxtaposi-
tion. The lead item and/or the apposed item could be clause, sen-
tence, paragraph, episode, etc. Two-clause apposition sentences
are, however, the easiest to talk about for a start.
The ubiquitous Hebrew 'and' is less common in vernacular speech
than in narrative and other more literary discourse. This may ac-
count for the rather jerky effect of the high level of asyndeton
in Pharoah's speech. Your father and your brothers have come to
you <A> the land of Egypt is at your disposal <A> in the best
part of the land settle your father and your brothers <A> let
them reside in the land of Goshen (Ge 4 7 s " 6 ) . The preamble to
the speech in De 5 contains four clauses (or sentences) in appo-
sition (De 5 2 " 5 ) . This is unusual. Such conjunctionless transi-
tion within a rather disjointed speech will not be considered as
sentence-level apposition, especially when a speech is a hetero-
geneous concatenation of discourses of different genres, such as
DECLARATIVE (Dc), INTERROGATIVE (Int), PRECATIVE (Pc). Examples:
Ge 2 2 3 , 4 S > 1 0 , 1 8 2 * , etc. But conjunctionless transition between
a curse (Pc) and a prediction (Dc) will be regarded as apposition
when the prediction expounds the curse (#3.8).
Conjunctionless juxtaposition which formally resembles apposi-
tion is also used to insert in a text extraneous remarks which,
in modern punctuation, would be placed in parenthesis, or gloss-
es which, in modern books, would be printed as footnotes. Exam-
ples: Ge 61*, 39 8 , De 3 9 . These do not belong in the present treat-
ment. Nor do we include the rare instances when a new paragraph
begins without conjunction. Examples: Ge 8 s , 221*.
A construction may present an appearance of apposition when
three clauses are coordinated. While 'and' may be used twice--
Cli and Cl 2 and C l 3 , it is equally acceptable to use 'and' only
once Clj, C l ? , and Cl 3 . It would be a mistake to say that Cl 2
is in apposition in the latter instance. The term 'asyndeton'
is properly used for such conjunctionless coordination (#3.10).
3.2. VERBAL REPETITION IN APPOSITION
The simplest kind of apposition is the repetition of an entire

clause. The nearest approach to this in Genesis is Ge 4 8 1 9 : y-
da^ti b?ni <A> yda^tl, I know, my son, I know. Note the repeti-
38 APPOSITION SENTENCES
tion of nelek in an apposition clause in Ex 10. In Ex 35 3 5

mille^ repeats waySmalle"1.
3.3. SYNONYMOUS APPOSITION

When two clauses in apposition are identical in both meaning and
grammatical structure, the result is the parallelismus membroram
so highly favoured in Hebrew, especially in poetry.
f
Example: d vesill semaân qli
n??e lemek ha^zenn îmrtl
Ada and Zilla, heed my-voice'.

Wives of Lemek, hear my- speech'. (Ge 4 2 3 )
Everything occurs exactly twice, and corresponding items come

in the same sequence.
Example: hSmlk timlk 'len

1im-ma?01 timfl bn
Will you really reign over us?
Will you really rule over us? (Ge 378)
A question may be repeated in apposition.
Example: ma-ppisÎ
ma hatt^tl
What is my offence?_
What is my failure? (Ge 3136)
The parallel items may be arranged in chiasmus.
Example: yaqnl'h 'bfzrlm

b?t0êbt yakîsh
They made him jealous with foreign gods:
With abominations they infuriated him. (De 32ls)
Example: hahdes hazze lkem r 1 ? hodsim

rl^fn h^ lkem lffhodse haS?n
This month for you (is) head of months

head it (is) for you of months of the year (Ex 122)
Other Examples: Ex 302 5 , Le 2 1 3 t , De 2 2 7 .
In Ge 41 3 3 311 Joseph reinforces his advice to Pharoah by repeat-

ing it in apposition, but with different vocabulary and differ-
ent grammatical constructions.
3.4. EPIC REPETITION IN APPOSITION
3.4.0. Introduction
In this construction the same (or very similar) semantic content

is expressed twice in two items in apposition. The items in appo-
sition may be single clauses; but entire sentences, paragraphs
or episodes might exercise the same functions. There may be some
measure of verbal repetition. Unlike the use of identical gramma-
tical sequences as in #3.3 above, however, similar grammatical
elements are characteristically placed in chiasmus. If coordin-
ation rather than apposition connected the clauses together, the
result would be what we call below a chiastic coordination sen-
tence (Chapter 9). The use or non-use of 'and' is not a matter
of chance. Chiastic apposition is a feature of high style in the
old sagas, and this is why we have called it 'epic.1 The result
is not a poetic bicolon, because the two parts can be of any
length, and may be quite disproportionate.
3.4.1. Verb Patterns in Epic Apposition
When the lead clause in epic apposition is a familiar paragraph-

level WP clause, the clause in apposition often uses the same
root in a VS.
Example: wayybS^ nh...êl-hatteb... <A> min-habbehem... b^

êl-nh êl-hatteb, and Soah went...into the ark...:<A> animals
...went in to Soah into the ark (Ge 7 7 " 8 ) ; wayyese^-nh... <A>
kol-hahayy...ys5^ min-hatteb, and Noah went out...: all the
animals...went out from the ark (Ge 8 l e ~ 1 9 ) .
The composition of these apposition sentences matches commands

and predictions made in other parts of the same story. Ge 7 7 " 8
obeys the command: b^ âtt... <A> mikkol habbShem...tiqqh-
lffk..., you will go in..., some of all animals.... you will take
to yourself... (Ge 7 1 " ) . Compare the similar prediction in Ge
6 1 9 " 2 0 . The verbs, however, are not the same. Similarly, Ge 8 1 "
19
obeys the command: se1 min-hatteb...: <A> kol-hahayy...
hawse^ îttfik, go out from the ark...: all the animals...bring
out'with you (Ge 8 1 6 " 1 7 ) . In both 8 1 7 and 8 1 9 the versions point
to wkl (coordination, not apposition). It is easy to see what
has happened here. Archaic chiastic apposition has been modern-
ized to classical chiastic coordination. Ge 7 2 and 7, however,
have escaped this, and support the authenticity of MT against
the versions in Ge 8 1 7 and 8 1 9 .
An apposition sentence is used to make a prediction in Ge 6 1 7 .
The lead clause is infinitival. lelahet kol-bsr ^ser-b rh
hayyim mittahat hassmyim <A> kl âSer b^res yigwa^, to
obliterate all flesh which (has) living spirit in it from under
the sky: all that (is) in the world will expire. The two parts
are synonymous. In the subsequent reported fulfilment, a simi-
lar apposition sentence is used, but the pattern of verbs is
different: vayyigva^ kol-bSr... <A> kl ^fer nismat ruh
hayylm b?âppyw mikkol ^?er behrb met, and expired all
1*0 APPOSITION SENTENCES
flesh...all that (had) breath of living spirit in its nostrils

from all that (was) in the dry land died (Ge 7 2 1 ' 2 2 ) . Besides
the chiasmus within each apposition sentence, there is also
some chiasmus between prediction and fulfilment.
Ge 4 6 6 * 7 has an apposition sentence exactly like Ge 8 1 8 " 1 9
in grammatical structure: wayyb^ mi^rym ya^Sqb...: <A>
bnyw... hebi^ îtt misrym, and-went to Egypt Jacob...:
his sons... he brought with him to Egypt. Compare Ex l 1 . This
is highly repetitious; the complete text also reiterates and
all his seed with him. The significance of this kind of con-
struction has generally escaped literary critics. Either they
assign parallel passages to different 'sources' as 'doublets',
thus destroying the fabric of the composition; or else they
speak disparagingly of its tedious redundancy. But if the text
is left as it is, and its grammatical structure is taken ser-
iously as serving artistic purposes, more positive conclusions
about the integrity of a passage and the solemnity of its style
are possible. Sentences from the Flood Epic used in the present
chapter cut across passages generally assigned to the 'J' and
'' documents. The same is true of the chiastic coordination
to be described in Chapter 9. This means that if the document-
ary hypothesis is valid, some editor has put together scraps
of parallel versions of the same story with scissors and paste,
and yet has achieved a result which, from the point of view of
discourse grammar, looks as if it has been made out of whole
cloth.
The epic of the rape of Dinah (Ge 34) is considered a patch-
work of J and E 1 . But if the text is left as it is, the climax
displays several examples of the chiastic epic repetition we
are describing here. In Ge 3 4 2 7 vayyb^ and b^ are in chi-
asmus. In Ge 3 4 2 6 wayyiqh and lqh are in chiasmus. And so
on. The clause-terminal VS is particularly striking, since
most Hebrew clauses begin with the verb. This extreme of chi-
asmus, which amounts to an inclusio for the entire sentence,
which then begins and ends with a verb with the same root, is
a hallmark of epic, whether apposition or coordination is used.
In Ge 411,8 the objects are chiastic with the verbs (vayyitten
... ntan). Other examples: Ex 8 1 3 (vay?hi... hy), Ex 8 2 0
(different verbs).
In prediction the corresponding sequence of verbs is WS...
VP.
Example: w ^ k e l êt-habblr ballayl hazze: <A> seli-

ges... y^kelh, and they will eat the meat this night: fire-
roasted... they will eat it (Ex 12). Similarly Ex 12 1 ".
3.4.2. Discourse Function of Epic Apposition
The rhetorical effect of this kind of epic repetition is to

slow down the pace of the narrative. It holds the picture a
little longer and enforces it on the mind. This may be done
at the beginning of a story to achieve a slow build-up. This
is illustrated by the opening clauses of Ge 14, a chapter that
APPOSITION SENTENCES Iii
is full of archaic epic devices. There is a preliminary state-

ment of the time setting which also anticipates some details of
the major battle to be described later (Ge 1 4 1 " 3 ) . This serves
as a kind of title for the whole epic. It consists of two clau-
ses in apposition. Then the narrative proper begins, again in
apposition, building up slowly by means of a traditional numer-
ical series.
In the 12th year they paid homage to Chedorlaomer,

and in the 13th year they refused,
and in the 14th year came Chedorlaomer. . . (Ge 141*"5)
Then the pace suddenly quickens, and a series of WP clauses

develops the narrative in a classical manner. A similar drawn-
out list of initial circumstances, using apposition, is found
in Ge 4 4 3 .
The Flood Epic has an elaborate and suspenseful prelude to
the onset of the deluge. In Ge 7 1 0 _ l e there are four references
to the date, all in apposition.
after a week (710)

in the 600th year... (7lla)
on that day (7llb)
on that exact day (713)
Compare the similar repetition of b e r e s e m hayym hazze, on that

very day, at the beginning of clauses in apposition in Ge 1721*
and 2 6 .
Again, the slackening of narrative pace by means of epic repe
tition in apposition can sustain tension at some high point in
a story. The climax of the Flood Epic is a majestic description
of the rising of the waters. Two whole and parallel paragraphs
are placed in apposition.
wShammayim gber me"* d me'Sd âl-h^res

wayekuss kol-hehrim haggebhim âser-taliat kol-hafs myim
hamei 'esre âmm milma^l gb?r hammyim
wayekuss hehrlm
and the waters prevailed very very over the earth
and were covered all the high mountains that are under
all the sky: <A>
15 cubits upwards prevailed the waters
and were covered the mountains (Ge 719-20)
There is chiasmus of SV... VS (the words are identical) between

the first and third clauses.
Ex 3 5 2 1 " 2 9 is a similar climactic passage full of repetition.
The ten verbs all have the same root bw^ , came, brought. The
clauses are either conjoined or in apposition.
The relaxation of the pace of narrative by means of epic re-
petition in apposition can also be used to wind down a story
at its end. Thus Ge 4 6 7 , already referred to above, marks the
h2 APPOSITION SENTENCES
end of a major episode. The apposition clause at the end of

Ge 25 l e similarly marks the end of the life of Ishmael. The
repetitive ken r, thus he did (Ge 6 2 2 , Ex 1 2 2 8 5 0 , 39 3 2 '
113
) marks the end of a paragraph. Contrast Ge 7 s .
3.4.3. Alternative Realizations in Sequential Clauses
Epic repetition in apposition halts the movement of a story

by giving two pictures of the same situation. Time stands
still. When the same thing is done by means of two WP clau-
ses in succession, violence is done to their normal function
of reporting events which come after each other in time (past
perfect sequential). The result is a striking alternative
presentation without apposition. Chiasmus is possible only
if a VS clause comes first.
At episode onset.
Example: weyhwh pqad êt-Sr ka^Sser ^ Smar

vayyaâl yhwh lesr ka^&ser dibber,
And Yahweh visited Sarah as he had said:

And-did Yahweh for Sarah as he had spoken (Ge 21 1 )
Compare Ge 47 1 3 .
At episode climax.
Example: vayyimah ^ et-kol-hay?qni. . .

wayyimmh min-h^re?,
And he wiped out everything that stood...

And they were wiped out from the earth (Ge 7 2 3 )
Compare Ge 4 s b . In Ge 1 4 1 1 " 1 2 there is epic repetition of con-

secutive verbs (WP) at the climax: vayyiqh... wayyelek way-
yiqty wayyelek. In Ge 343 there is a three-fold repetition
in WP clauses to underscore the intensity of Shechem's passion.
At episode end. After a note on his testamentary disposal
of his property (Ge 25 s " e ) and another on his age at death
(25 7 ) , the account of Abraham's life ends by piling up clau-
ses reporting his decease.
wayyigwa (
wayymot âbrhm b5Seb tb
zqen es\be f
wayye^sep êl-'ammyw,
And he expired
And died Abraham in good old-age
venerable and satisfied
And he was collected to his relatives (Ge 25 e )
Compare Ge 3 5 2 9 , 49 3 3 . There follows an elaborate description

of his burial, again with repetition in apposition. The resump-
10b
tive use of ?mm there in Ge 2 5 resembles the use of ken
thus in Ge 6 2 2 . A new story begins with 2 5 1 1 .
As in poetry, so in this epic prose, the repetition may in-
volve the use of the same vocabulary (repetitive parallelism),
or it may use conventional synonyms (synonymous parallelism),
which often come in an established sequence.
Example: vayesall5]jeh yhvh êlhim miggan-êden...

wayegre? êt-h^dm,
and Yahweh God expelled him from the garden of Eden . , ,

and he drove out the man (Ge 3 2 3 - 2 * )
This marks the end of the story of the Fall. Note that the lead
clause has the anaphoric him, while the following clause has
the explicit noun object the man. The same thing happens with
the subject Abraham in Ge 25 8 quoted above.
3.5. APPOSITION FOR EMPHASIS
3.5.0. Formal Features
Synonymous parallelism of two clauses in the course of a piece

of narrative does not necessarily mean that here we have a scrap
of poetry, proving, perhaps, that the present forms of Israel's
early traditions rest on a poetic substratum, as many have sur-
mised. The addition of an equivalent clause in apposition may
be used in prose to underscore an important point. The poetic
effect is real, but incidental. Although this construction may
not be very different from epic apposition (#3.4), in its typ-
ical form it uses neither the repetition of the same vocabulary
nor the chiastic arrangement of epic traditions.
3.5.1. Antithesis in Apposition
One device used to drive a point home is the redundant repeti-

tion of the same material by means of the negation of an anto-
nym.
Example: kenim ^natjn <A> 15^ hyin meraggellm,

we are honest men: we are not spies (Ge 42 3 1 )
Example: vattehl lray ^qr <A> ^ en lh wld,

Saray was sterile: she had no children (Ge l l 3 0 )
Example: wehabbr req <A>ên b myim,

and the cistern was empty: there was no water in
it (Ge 37 2 ")
Example: vayysar he^rb... <A> nisOar êljd,

and the flies departed...: not one remained (Ex 8 2 7 ;
similarly Ex 1 9 ' 2 3 , 1 4 2 " )
Example: wSêne yiSrêl kbed mizzqen <A> l^ ykal lir^t,

Now Israel's eyes were heavy with age: he could not
see (Ge 4 8 1 0 )
It is important to underscore this point, to bring it into line

with Ge 27 l .
Other examples: Ex 5 8 . 2 6 , 1 3 2 1 - 2 2 , 1 8 1 8 b , 1 9 1 2 b with 13ab
( d i s c o n t i n u o u s l y ) , 25 , D e 2 2 ' 3 * 3 6 (we took all the cities...

< A > there wasn't a city too strong for us < A > Yahweh gave us
everything), De 3", 4 2 2 , 4 2 6 (th is has an antithetical sentence
in apposition with the lead clause), 4 3 5 3 9 5 2 " 3 (antithetical
sentence in apposition), 6 1 3 " 1 * , 7 2 > 2 2 ' 2 " 2 *.
A command may be issued as a positive injunction and then re-

peated in apposition as a negative prohibition.
Example: zekr <A> "> al-tiSkh ,

remember', don't forget: (De 97)
Other examples: Ge 4 5 s ( c o m e down to me; don't stay there'.),

Ex 16 2 9 (rest...don't go out...), 29 3 *, De 5 2 9 .
3.5.2. Apposition in Another Perspective
What are essentially the same facts may be presented twice, from
two different points of view, in two clauses in apposition.
Example: ^nki êêrbenn <A> miyydl tebaq.senn,

I'll be responsible for him [ J u d a h ' s side]. from my
hand you will seek him [ J a c o b ' s s i d e ] (Ge 4 3 9 )
Similar apposition sentences in which the first clause states a

fact from the speaker's point of view and the second clause
states the same fact from the addressee's point of view are
found in Ge 3 1 $ , 1 5 1 , 3 1 s 9 .
In Ge 3911* Potiphar's wife accuses first her husband, then
Joseph in apposition. In Ge 8 2 1 " 2 2 two sentences are in appo-
sition: the first (a sentence of two conjoined clauses) in first
person; the second (with the verb in clause-final epic position
as episode terminus) in third person.
Other examples: Ex i 1 * , 14 3 , 2 2 2 6 , 2 3 1 3 , De 7 6 .
3.5.3. Climactic Repetition in Apposition
Several short clauses in apposition create tension and some-

times mount to a climax. Israel's terrible cry in Ge 3 7 3 3 con-
sists of three such clauses in apposition:
It's my son's cloak:

A wild animal has eaten him:
Joseph has s u r e l y been torn to pieces.
APPOSITION SENTENCES 1+5
The postponement of the name 'Joseph' heightens the effect.

The protest in Ge 4 2 1 0 " 1 1 consists of four short clauses in
apposition.
The commands in Ge 1 9 1 7 stand in discontinuous apposition.
f
himmlet al-nap?ek
hhr himmlet
Flee for your life'.
To the mountains flee I

This is also climactic, since the second clause is more speci-
fic than the first.
3.5.4. Coordination of Emphatic Repetition
Various kinds of coordination may also be used rather than ap-

position with this kind of emphatic repetition.
Example: lemaân yitab 1 ba^Sbrek weljayet naps! bigllek,

so that it may be good for me because of you and
Iso thatl my soul will live on your account (Ge 1 2 1 3 )
Example (discontinuous): wel^-nS^ ^tm h^re^ lsebet

yahdw...vel^ ykel lsebet yahdw,
and the land did not support them to live together
...: and they could not live together (Ge 13 )
3.6. RESUMPTION AND DISTRIBUTION IN APPOSITION
3.6.0. Resumption
A clause may be repeated in apposition, using deictic or ana-

phoric pronouns in order to continue discourse.
Example: And the sons of Noah...were Shem, Ham and Japheth

<A> these three were Noah's sons, and from these
all mankind spread out (Ge 9 1 8 " 1 , compare with
10 3 2 )
Note the resumptive time reference in Ge 1 5 1 8 , 172S.
3.6.1. Distribution
A distributive subject or other distributive item is often

placed in a separate apposition clause, with repetition of
the verb.
Example: vayyiptr-ln ^ et-halmten <A> ^ is kaljalm ptr,

and he interpreted to us our dreams: each according
U6 APPOSITION SENTENCES
to his dream he in terpreted (Ge 4 1 1 2 )
Note the WP...VS pattern of verbs with the same root, present
also in the preceding Ge 41 11 .
Other examples: Ex l l , 12*, 16 1 6 1 8 , 28 2 I b .
3.7. EXPLANATION IN APPOSITION

3.7.0. The Semantics of Explanation
In the forms of apposition studied so far there has been a

large measure of semantic overlap, even identity, between the
two juxtaposed clauses, and also a considerable degree of gram-
matical similarity between them. Often the same vocabulary is
used in each part. Such constructions doubtless contain redun-
dant material, but this also serves important rhetorical func-
tions .
We have moved along the spectrum from constructions with al-
most complete identity in apposition (#3.3) to constructions
in which the juxtaposed clauses match less closely (#3.6). If
the material in apposition adds as well as repeats there is
less overlap, and we have moved a little further along the
spectrum. The clause or sentence or paragraph in apposition
develops the theme a little further.
Example: weyitten-11 êt-me^rat hammakpel s er- âser
biqse sdeh <A> bSkesep mle^ yittenenn II
betkekem la^huzzat qber,
V:and let him sell IO:me 0:the Makpela Cave which
he owns, which is at the edge of his field: Price:
for full money V:let him sell 0:it IO:to me L:in
your midst OC:to be a freehold tomb (Ge 23)
All this could have been done in one clause, without repeating
let him sell it to me. But the clause would have been long, a
thing Hebrew does not like. But the placing of the reference to
Price at the beginning of the apposition highlights this point.
Grammatically everything in the first clause is repeated in the
second:
Cj V 10 0
<A> Price V 0 10 L OC
The second adds to this common nucleus the further details about
Price, Location, and the role of the object (OC). Ex 252 has
similar repetition and addition in apposition. Ex 30 10 repeats
and adds one item. De 8 2 0 repeats t'bedn and adds a simile.
These examples partly repeat, partly develop.
While there are many kinds of additive apposition--specify-
ing, explicating, augmenting, etc.--no attempt will be made here
to force all examples into rigidly defined categories. We shall
distinguish two-clause apposition in which the second is more

specific than the first from the apposition to a lead clause
of a more extended exposition in a juxtaposed sentence or para-
graph.
3.7.1. Specifying Apposition
The second clause makes explicit some detail lacking or present

only in a general way in the first clause.
Example: weâtt tb^ êl-^bteyk beslm <A> tiqqber

beseb tb,
and you, you will go to your ancestors in peace:
you will be buried in good old-age (Ge 15 )
To be buried is equivalent to joining one's ancestors; the cir-

cumstance of ripe old-age specifies what is meant by dying 'in
peace.'
Example: y eme sene megray 130 sn <A> meât werîm hy

yeme sene hayyay...,
the days of the years of my sojournings are 130
years: few and evil have been the days of the years
of my life... (Ge 4 7 9 )
The apposition clause adds the unexpected comment that Jacob's

life-span was brief. The common subject is repeated with a mi-
nor vocabulary alteration in chiasmus.
Example: wayyitten êl-hgr <A> sm âl-sikmh,

and he gave CitD to Hagar: he put tit 3 on her
shoulder (Ge 21 1 *)
Other examples: Ge l 2 8 (specifies the purpose of fruit and veg-

etables), 6 1 6 (specifies first, second and third), 6 1 9 (speci-
fies male and female), 8 5 (specifies the emergence of mountains
--note the chiasmus SVTi <A> TiVS), 9 6 (specifies capital pun-
ishment for murder), 1 7 1 3 (specifies that home-born and purchased
slaves are to be circumcised, as well as sons), 1 7 1 $ (specifies
that kings will descend from Sarah--comparison with Ge 17 6 shows
that it needs only the addition of we- to make this a well-formed
chiastic sentence [Chapter 9]), 1 7 2 0 (specifies the number of
promised children), 18' 1 (specifies menopause as the feature of
old-age that is pertinent to the narrative, Sarah's inability to
have children), 20 9 (specifies that Abraham's great crime is
something absolutely forbidden), 2 3 1 1 (specifies that the sale
must be formalized in public), 26 2 (specifies the imminence of
death as a result of extreme old age), 29 7 (specifies that the
early hour is not the time to assemble flocks), 30 3 3 (specifies
the exact nature of Jacob's responsibility for missing animals),
3 1 3 9 (similar), 3 4 1 2 (specifies that there is no limit to what
he will pay--compaTe Ge 23 9 ), 3 5 1 1 (specifies the nature of Ja-
cob's progeny), 39 9 (specifies Joseph's greatness), 3 9 2 3 (empha-
sizes total delegation--the repetition here is also climactic),
42 6 (specifies that Joseph was in charge of sales), 42

(specifies the aim of their spying), 48 (the apposition clause
names the two sons in question and specifies in what capacity
they will now be Jacob's sons), 48 6 (specifies the rights of
Joseph.'s children in the matter of inheritance), 481'* (a note
about Jacob's gesture), 5 0 2 3 (the apposition clause specifies
the great-grandchildren), Ex 6 2 6 2 7 (states who Moses and Aaron
are), 711* (specifies the outcome of hard-heartedness), 8 1 9 (spe-
cifies time), 12<*1:> ~ 5 (contains three successive clauses in appo-
sition each specifying a different rule about the Passover
lamb), 1 2 1 6 (explains what a miqr"1 qdes is), 1 2 1 7 " 2 0 (four
apposition clauses in succession explaining the eating of mas-
sot) , 13 2 (the rule applies to both man and beast), 14 2 (lo-'
cates the camp), 1 5 2 5 (an aetiological note), 1 6 2 (spells out
the pattern of manna), 1 6 3 5 (specifies the exact moment the man-
na ceased; this apposition clause, with its verbal repetition
and chiasmus, also provides heavy rhetorical underscoring in
apposition with a circumstantial clause that ends an episode,
marking a climax), 1 * (the second question is more specific
than the first), 1 9 1 2 (specifies the death penalty for encroach-
ment), 1 9 1 5 b (one specific ban), 21311 (prescribes money payment),
2 2 3 0 (torn flesh for dogs), 23 (rules for unleavened bread),
2 5 1 1 (inside and outside; note the chiasmus), 2 5 (gold plate;
this clause is inserted INTO the lead clause, a very rare event
in Hebrew), 2 5 1 9 ' 2 0 (chiastic with a circumstantial clause),
2 5 2 7 (placement of the rings), 2 5 2 9 (specifies pure gold), 2 5 3 1
(two such clauses in succession), 2 S 3 S (all one piece), 2 5 3 9
(specifies weight), 26 1 (material of the curtains),26 (number
of curtains), 26 (size of curtains), 2 6 3 1 (specifies embroi-
dery), 27 1 (a clause saying that the altar is square, which is
in fact redundant, is inserted parenthetically into the clauses
stating the three dimensions), 2 7 2 a B (horns continuous with
body) , 2 7 3 b (everything bronze), 27 (the altar is hollow) ,
2 7 1 7 (materials), 2 7 l e (dimensions), 2 7 2 1 (specifies Aaron's
duties), 283 2 b (binding), 283 7 b (exact location), 28" 2 b (exact
pattern), 2 9 2 b (ingredients), 293 * (it's not to be eaten.'),
2 9 3 5 ' 3 7 a (specifies period), 29 3 (holiness is contagious),
3 0 l b (material), 30 2 (square inserted in a conjunctive sen-
tence), 3 0 3 b (location), 3 0 2 9 b (holiness contagious), 303 3 > 3
(prescribes excommunication), 3031* (recipe), 3i Ilf bA,i5b (pre-
scribes death), 3 1 1 7 (the sabbath a sign), 3 2 1 5 (written on
both sides}, 3 4 1 (festival rules), 3 4 ? l b (keep sabbath in busy
times), 3 5 S b (death for sabbath breaking), 35 3 (an additional
rule about fire), 35 s (lists materials). 36 11*' (number), 3 6 3 5
(material), 3 7 7 b > e b . 1 * > 1 7 2 2 > 2 \ 3 8 2 . 3 > ' b , 39". \ 40 3 *, Le l 3 b ,
4 1 2 b (location), De 4 3 0 > 3 3 (specifies what 'this great thing 1
is), 9 1 6 (three clauses), etc.
In Ge 4 3 2 7 the second question (in apposition)--is he still
alive?--is more general than the first--Js he well?--which al-
ready assumes an affirmative answer to the second. The effect
is electric. The reply repeats the pattern.
Ge 6 l e shows that a specifying clause (with the dimensions of
the window) can be placed in coordination rather than in appo-
APPOSITION SENTENCES 1*9
sition (See #3.7.3). In Ge 17 1 2 specification of the age of cir-

cumcision is made in coordination, not apposition. This alterna-
tive realization is, however, rare.
3.7.2. Exposition in Apposition
The main feature here is a more extended exposition in apposi-

tion of material which is not necessarily implicit in the lead
clause. Instead of one clause in apposition, which we have called
'specification1 (#3.7.1), a sentence or paragraph of at least two
clauses is used for exposition.
Example: wehithatten ^tn <A>

bSntekem titt?n ln
w?êt-bSnten tiqeh lkem,
And you will inter-marry with us:

your daughters you will give to us,
and our daughters you will take to you ^(Ge 3 4 s )
This spells out the details of the clause of the proposed treaty
dealing with inter-marriage and restricts it also to the exchange
of women. The expository sentence is in the form we have called
'contrast.1 See Chapter 11.
Example: vayyippred îf me âl ^hiw <A>

1
abrm yfab be^ eres-k5nân
welt ysab be^re hakkikkr,
And they separated each from his brother:

Abram resided in the land of Canaan,
and Lot resided in the cities of the plain
(Ge 1 3 1 1 ~ 1 2 )
Example: vay?happes <A>

baggadSl hehel
baqqtn kill,
And he searched:
with the eldest he began
and with the youngest he ended (Ge 4412)
Joiion2 calls this 'circumstantial.' The construction invites

comparison with Ge 12.
wayyet ôholh <A>

bet- 1 el miyym
w?hây miqqedem,
And he pitched his tent:

Bethel (was) to the West
And Ay was to the East.
50 APPOSITION SEHTENCES
This too has been called 'circumstantial'--and he pitched his

tent with Bethel to the West and Ay to the East.3 A single
circumstantial clause is usually coordinated (#5.1.3.3), and
a circumstantial sentence also can be coordinated (#5.2.2). If,
however, Ge 12 8 contains a circumstantial sentence, the switch
from coordination to apposition is just like the difference be-
tween the one-clause coordination in #2.6.5 and the two-clause
conjunctive sentence in apposition in #2.6.7. Our interpreta-
tion of the construction as expository apposition detaches the
conjunctive sentence from the lead clause, making it more par-
enthetical than circumstantial.
Other examples: Ge 3 1 5 (gives details of the enmity between

the snake and the woman), 4 1 3 (details of Cain's expulsion),
9 5 (explains that both animals and men will be accountable for
murder), 1 6 1 2 (expands on Ishmael's wildness), 1 9 1 7 (two prohi-
bitions in apposition spell out the command to escape), 27
(gives detailed instructions; but the same thing in Ge 27" 3
uses coordination), 2 7 3 6 (recounts Jacob's two fraudulent acts)
2 8 1 7 (explains why the place is frightful), 311,1 (breaks the
twenty years into fourteen plus six), 3 1 5 2 (both monuments are
witnesses), 4 i 1 0 " 1 3 (this is in apposition with 4l' b , and ex-
pounds it), 4 1 1 3 (the respective fates of the two slaves), 4 1 2 6
~ 2 7 (both clauses enlarge on the fact that the two dreams have
the same meaning), 4 2 3 6 (three conjoined clauses list Jacob's
bereavements), 45 (lists Joseph's various presents to members
of his family), Ex 7 2 (explains in what sense Moses is 'god'
and Aaron is 'prophet'), 9 2 7 (exonerates Yahweh and incrimin-
ates Pharoah) , 10*11 (points out the scale of the disaster),
1 2 1 5 (liturgical details), 121*2 (two clauses in successive ap-
position enforce the solemnity of the vigil), 12" 3 - 1 , 9 (a string
of miscellaneous Passover regulations in apposition), 1 3 6 " 7
(Passover rules), 19 s (the respective roles of Moses and God),
2 0 1 9 (how to keep the Sabbath in detail), 2 2 2 9 (time schedule
for devoting the first-born}, 2 3 1 5 " 1 6 (three festivals--note
the inclusio with v . 1 7 ) , 25 (keep the poles in the rings),
2 5 1 0 and 30 2 (dimens ions), 3 0 7 ~ 8 a (a conjunctive sentence in
apposition with two specified times), 31 l , 3 4 2 1 , 35 2 (expounds
the detailed requirements of Sabbath keeping), 3 6 1 5 (dimensions
37 1 (dimensions). 37 6 (dimensions), 3 7 1 0 2 5 , 3 8 1 , 3 9 9 1 0 (the
four rows), De 3 f l , 6 1 3 , 7 3 , 7 5 , etc.
3.7.3. An Example of Explication
Ge 6 l l f " 1 6 gives a list of instructions for building the ark.

There are eleven clauses in all, joined together in various
ways by means of apposition and coordination. On first appear-
ance there seems to be no system in the formal patterns of link
age with a conjunction (vS- and), or conjunctionless juxtaposi-
tion. On closer study, however, the discourse proves to be
well structured.
The general command, using an IMPERATIVE VERB (VI)-- r aSe lek
tebat ^?e-gper, make yourself an ark of gopher-wood--is fol-
lowed, in apposition, by a paragraph which explicates the com-

mand, using predictive verbs. This paragraph consists of three
sub-paragraphs, each of which deals in turn with one structural
feature. Each successive feature is named by the first noun in
each sub-paragraph--qinnim, rooms (RSV) or reeds (NEB), ?har
window, petalj door. The three sub-paragraphs are coordinated
(conjoined--Chapter 8), but because Hebrew often coordinates
three items as A, and C, the second sub-paragraph ( 6 1 6 a A )
gives the appearance of being in apposition. That there are
three sections (sub-paragraphs) is further confirmed by the
repetition of an explicit reference to the teb ark at the
beginning of each sub-paragraph, followed by the use of an
anaphoric pronoun within each sub-paragraph. The stippled ar-
rows in the tree on page 52 trace these anaphoric connections.
The first sub-paragraph consists of two parts coordinated
together. The first part (6*" b ) is really a mini-paragraph,
since the second clause is a typical paragraph-level WP clause,
the only one in the passage. That is, wekpart, and you will
cover, is an immediate continuation of the prescriptive taâse,
you will make, not a noncontiguous continuation of the earlier
imperative âse, make'. The second part of the first sub-paragraph
( 6 s ) is not a new beginning, in spite of ze this (#3.7.5.1), be-
cause ^th it is anaphoric to hatteb, the ark (1"*"), and shows
that we are still in the same sub-paragraph. 6 1 5 gives the size
and shape of the ark by means of an expository apposition sen-
tence of the kind discussed in #3.7.2. The exposition consists
of three clauses with identical internal structure (PS). These
three clauses constitute a triple conjunctive sentence (#8.0.2)
with the coordination pattern A , and C. (Similar sentences in
Ex 2 5 1 0 2 3 have two and's and do not repeat the noun.) This con-
junctive sentence has its own inner unity marked by the use of
an explicit topic noun (hatteb) in the first clause, with ana-
phoric pronouns in the following two clauses.
The second sub-paragraph (6 1 6 ) consists of two coordinated
clauses, again with noun followed by anaphoric pronoun.
16aB-t>
The third sub-paragraph ( 6 ) is a specifying apposition
sentence (#3.7.1)."
The exposition is thus seen to be hierarchical, passing from
paragraph to sentence to clause as instructions become more
specific. 5 Each of the three sub-paragraphs divides into two
parts, the first being a general command to make rooms, win-
dow(s) and door(s) respectively, the second giving details and
dimensions. It is surprising, therefore, that apposition is
not used to connect the two parts (general and specific) of
each sub-paragraph. In fact coordination is used in the first
two, apposition only in the third. The result, however, is a
neat pattern which is probably not accidental.
Lead Clause <A> Sub-paragraph (11 and Se)

Co=0 Sub-paragraph (Cle and CI9)
and Sub-paragraph (Clio <A> CI11)
Another systematic pattern that unifies the whole is the steady

movement of the verb to the end of the clause. Recognizing the
seven paragraph-level clauses, namely, the lead clause of the
whole, the lead clauses of the three sub-paragraphs (Cl 2 , Cl 8 ,
Clio), and the first following clause in each sub-paragraph
( C I 3 , Cl 9 , e i n ) , we have the following
CI 1 V A
Cl 2 A V
CI 3 A V
Cl e A V
CI, A V
Cl 1 0 A V
Clu A BC V
Because of the frequent explications and the arrangement of

terms and conditions in lists, the Book of the Covenant is spar-
ing in the use of coordination. On the other hand 1 or is
used more often here than anywhere else in the Torah.
3.7.4. A Summary in Apposition
In G e 4 1 9 b t h e b u t l e r s a y s My faults 1 recollect today and pro-

ceeds to supply the details in apposition. Such a summary state-
ment can also come in apposition after the details are given.
Example: y o u have done evil by doing what you did (Ge 4 4 s ) sums
up in apposition the preceding accusations.
Other examples: Ge 42 3 6 (the obscure ^lay hy kulln seems

to sum up the preceding), Ex 2 6 e b (all curtains the same size),
2 6 1 7 b (all boards the same). 26 2 , , t , 31 1 1 , 34 2 3 (sums up like
the inclusio in 2 3 1 7 ) , De 9 " (summarizes verses 7 ~ 2 3 ) .
3.7.5. Titles and Colophons
3.7-5.0, Introduction
Hebrew commonly uses verbless clauses and sequence SP, S:Pr to

describe the content of a unit of literature. 6
Example: ^ e l l e h a d d ? b r l m . . . , these are the speeches which

Moses uttered... (De l 1 )
Without S:Pr, the predicate would be like one of our titles.

Debrlm speeches is in fact the conventional Hebrew name for
Deuteronomy.
Such a clause used as a title may precede the work it names,
or come at its end as a colophon.
5U APPOSITION SENTENCES
3.7.5.1. Titles
A construction in which a clause title is followed by a complete

literary piece is a special case of a general statement followed
by exposition (#3.7.2).
Example: z^t ^t-habberit... <A> êt-qastl ntattl be f nn...

this is the covenant-sign...: my bow I have placed in the clouds
(Ge 9 1 2 ; but in 9 1 1 the content of the covenant is coordinated).
Other
19 examples: Ge 5 1 , 6 9 ' 1 5 , 1 , l l 1 0 2 7 , 1710 , 201 3 , 2 5 1 2 '
, 34 1 5 , 36 l 9,10,12, etc. 4012,18) 4 1 1 > 4 2 3 3 ) 4 3 n f
45 ' 1 7 , Ex 3 1 2 , 1 " 1 6 , 7 1 7 , 9 1 * (note repeated ba f br), 12* 3 ,

21 1 , 29 ' 3 8 , 35 , etc.
3.7.5.2. Colophons
A summary statement at the end of a piece (#3.7.4) may likewise

function as a concluding title or colophon.
Examples: Ge 2 " % 10 2 0 ' 3 1 ' 32 , 22 2 S , 2S 1 , b > 1 6 , 3 5 2 6 b , 36sb,

Ex 6 Y 5 i ' 1 9 ' 2 , , b ' 2 5 6 , 19 , 38 2 i .
De 4 1 * 5-1,9 is either a colophon of extraordinary length that

rounds off De 1-4, or it is the title to De 5ff. Compare Ge 9 1 7 ,
which acts as inclusio with a similar clause in 9 1 2 . Similarly,
Ge 5 2 (bSym hibbre"lm) stands in apposition as an inclusio
with the similar construction at the beginning of the chapter.
In De 4**** the colophon is in coordination.
3.8. CURSES AND BLESSINGS 7
It is customary for a curse or blessing to take the form of a

general first statement in precative mood, followed, in apposi-
tion, by a detailed prediction, in indicative mood, of what the
curse or blessing will entail. It differs from the constructions
discussed in #3.7 in the switch in mood at the point of apposi-
tion.
Example: ^rr kenân <A> êbed. âbdimyihye leêhyw, cursed

(be) Canaan: slave of slaves he will be to his brothers (Ge 9 2 6 ) .
Other examples: Ge 3 l k > 1 7 , 4 1 1 .
The lead curse may itself be indicative (predictive), followed

in apposition by exposition. Examples: Ge 3 1 5 ' 1 6 , De 711*.
A conditional prediction may be followed, in apposition, by
an exposition of the alternatives using IF X,, THEN ; AND (NOT
">0 or) IF X2 , THEN Y 2 . Examples: Ge 13 9 , 31*, 34 1 5 , 43 ,( " 5 . In
legal texts, a series of such conditions may be listed in appo-
sition (Ex 2 1 3 " 6 ) . When we- is used, as in Ex 21 s , it means
but.
3.9. M U L T I P L E A P P O S I T I O N
W h e n three (or m o r e ) clauses (or similar elements) s t a n d in ap-

p o s i t i o n , the c o n s t r u c t i o n may be formally ambiguous.
CI <A>
C l 2 a n d C l 3 together c o u l d c o m p r i s e an a p p o s i t i o n sentence
w h i c h , as a u n i t , is in a p p o s i t i o n w i t h Cli.
Example: w S h ^ a n s i m s u l l e h h e m m w a h m r e h e m <A>
h e m y s e ^ êt-hÎr <A>
l^ h i r h l q ,
and the men were sent away, they and their asses:
they had left the city:
they had not gone far (Ge 4 4 3 )
The last clause specifies the s e c o n d l a s t , and b o t h together ex-

p l a i n the first. A n o t h e r e x a m p l e : Ge J 1 7 i ' " a .
Or Cli and C l 2 t o g e t h e r c o u l d c o m p r i s e an a p p o s i t i o n s e n t e n c e
w h i c h , as a u n i t , has CI3 in a p p o s i t i o n w i t h it.
Example: wayyibr^ ^ e l h l m ^ e t - h ^ d m b ? s e l m <A>

b ? s e l e m êlhim br^ ^ <>
zkr n e q e b br^ ^tni,
and created God man in his image:

in the image of God he created him:
male and female he created them (Ge l27)
The first two clauses are an excellent example of epic r e p e t i -

tion (note the c h i a s m u s ) . The third clause is a n i c e instance
of s p e c i f y i n g a p p o s i t i o n , in a p p o s i t i o n w i t h the p r e c e d i n g sen-
tence (compare Ge 6 1 S ; see Ge 5 also).
A g a i n it is p o s s i b l e for C l 2 and C l 3 to be s e v e r a l l y in appo-
s i t i o n w i t h CI . A s such they are c o o r d i n a t e d , and could be con-
n e c t e d by AND.
Example: nh "'i? saddlq <A>

t m i m h y b ? d 5 r t y v <A>
êt-hêlhim hithallek-nh ,
Noah Lwasl a righteous man:

perfect he was in his generations:
with God walked Noah (Ge 6 s )
There are two aspects of Noah's righteousness --his perfection

as a man (Cl 2 ), his relationship with God ( C I 3 ) . The two aspects
of Cain's curse stand in successive apposition with Ge 4 l 2 --his
failure as a farmer and his existence as a fugitive. In Ex 26 1 " 3
the details about the tabernacle are supplied in apposition.
2 6 l b B is a clause that specifies the material; 26 2 is an appo-
sition sentence (consisting of a conjunctive sentence and an
apposition clause) which, as a whole, specifies the dimensions;
26 is another clause in apposition. So each successive item in
apposition deals with a different aspect of the lead clause.
In Ex 32 7 " 8 there are three clauses in successive apposition,
each more specific than the preceding (reproduced in De 9 1 2 and
in De 9 1 6 but with verbs in different sequence in the latter).
The piling up of four clauses in apposition has a cumulative
effect in Ge 9*" 3 . The alternation of two pairs of identical
verbs gives a sequence ABA'B 1 . The clauses become more and more
and more specific. (Similar clauses occur in the reverse sequence
in Ge 12S.)
3.10. APPOSITION INSTEAD OF OTHER CONSTRUCTIONS
3.10.0. Alternative Deep Relationships
In all of the examples discussed so far there has been a connec-

tion between the two clauses or other elements in juxtaposition
which permits us to speak of a relationship of apposition. There
is some measure of semantic overlap between the two parts. When
there is no such connection, either the discourse is incoherent
or formal apposition may be used as an alternative realization
of a relationship normally realized by means of a conjunction of
some kind, or by means of the RELATIVE ^ i e r .
3.10.1. Coordinate Commands in Apposition
The juxtaposition of two imperative or cohortative verbs which

command the first of a series of actions results in an apposition
complex that could be called a VERB PHRASE, that is, a phrase
consisting of closely knit verbs and functioning as a single
predicator. The first verb is typically one of movement, usu-
ally derived from bv^ come, hlk walk, yq^ , go out, yrd, go down,
qwm, stand up, swb, return (do it again), sllj send.
Example: qum <A> qah ^ e t - ^ iJftek, get up: take your wife'.
(Ge 1 9 1 5 ) .
The first verb becomes semantically empty, functioning merely

as a hortatory particle. So in Ge 27 1 9 qm-n^, please stand up,
is followed by the contradictory sit down and eat: Or and come--
17
go '. (Ge 45 ) .
It is a token of the exclamatory function of the first verb
that it rarely receives additional clause elements, except for
n^; nothing intervenes between the two verbs.
A further indication of the drift of the first imperative verb

to the role of an exclamation is seen in the fact that the mas-
culine singular form may be used for all numbers and genders,
especially the long intensive foTm. So hb give (Ge l l 3 ' " ' 7 )
is purely hortatory. Compare lkh in Ge 1 9 3 2 (which Sam. reads
Iky) .
Examples: bw"1 - Ex 6 1 1
hlk - Ge 19 3 2 , 27 1 3 , 29 7 , 37 l,f , 4 5 1 7 , Ex 4 1 9 , 5 s
(apposition alternates with coordination in these
texts), S 1 1 ' 1 7 ' 1 8 , 8 2 1 , 10. 2 \ 12 3 , 1 9 2 \ 32 7 '
3
" , 33 , De 5 2 7
- Ex 17 s
yrd - Ex 1 9 2 1
ngs - 2Sa l 1 5
qvm - Ge 1 3 l " , 1 9 \ 21 1 8 , 27 1 9 . 1 * 3 , 28 2 , 31 1 3 , 3 S 1 ,
43 1 3 , 44*, Ex 1 2 3 l a , 3 2 1 , De 2 2 \ 9 1 2 , 1 0 1 1
Svb - Ge 4 3 2 , 4 4 2 5
- Ex 9 1 9
3.10.2. Coordinated Declarative Clauses in Apposition
In ##3.2-3.6 there is a considerable measure of semantic overlap

between two items in juxtaposition. In ##3.7-3.9 the juxtaposed
material develops the discourse with new material. When all the
following material is new, it is no longer appropriate to speak
of an apposition. The construction is better described as asyn-
detic coordination. Since there is usually some connection be-
tween successive clauses, the boundary between these two kinds
of construction remains indeterminate. Example: De 5 2 6 .
3.10.3. Coordinate Questions in Apposition

2
Example: Ex 17 .
3.10.4. Antithetical Clause in Apposition
Example: vePattem hSsabtem ^lay r' <A> êlhim hasbh le-

Jb, and you thought (of it) against me (for) evil: (but) God
thought of it for good (Ge 5 0 2 0 ; note use of ki in Ge 45 s ).
Example: âl-ti?p?k-dm ... hazze, don't shed blood: (on the

contrary) throw him into this cistern (Ge 3 7 2 2 ) .
Other examples of prohibition followed by opposite command:

Ge 2 1 1 2 , Ex 3 s , De 7 1 , 9 2 6 " 2 7 . See also Ex 8 2 3 , 2 3 3 0 , De 1 3 ,
4 3 5 (but you have been shown...'). Compare #3.5.1.
Deuteronomy likes to follow up the antithesis with a positive
statement.
Example: And requiting his enemies (sic'.) to his face <A>

He won't be behindhand to his enemy <A>
To his face he will requite him (De 7 1 0 )
Note the chiasmus. De g 2 6 " 2 7 has Neg-Pos-Neg.
3.10.5. Apposition I n s t e a d of Subordination9
Two clauses w h i c h s t a n d in some k i n d of causal relationship to

each o t h e r may be p l a c e d in simple j u x t a p o s i t i o n instead of us-
ing the usual s u b o r d i n a t i n g c o n j u n c t i o n s .
Example: eh kol-ha''me ^ y i h a q - l i , God

made laughter for me; [ t h e r e f o r e ] everyone who hears it will
laugh for me (Ge 2 1 6 ) .
E x a m p l e : w e n i k r e t hannepes hahl^ m e ' a m m e y h <A> êt-berltl

h e p a r , a n d that person will be cut off from his people Lbecausel
he has broken my covenant (Ge 17 11 *).
Example: 15^ tob h?yt h ^ d m lebadd <A> ^ e f e s e - l l f ezer

k e n e g d , It's no good for the man to be alone: CsoU I'll make
a suitable assistant for him (Ge 2 1 11) .
Other examples: Ge 2 4 s " (<A> = t h e r e f o r e ) . 4 2 l e b (Joseph's pi-

ety is the reason for his c l e m e n c y ) , Ex (the reason for
sheltering man and b e a s t ) , 14 111 (the reason for the s t r a t e g y ) ,
25
16 (it's sabbath : Zthereforel you won't find it in the field
today), 1 8 1 8 a (gives the r e a s o n for 1 8 1 7 ) , 2 1 5 (therefore).
Since this list is r e a s o n a b l y c o m p l e t e , it can be seen that the

s u b o r d i n a t i n g c o n j u n c t i o n is not o f t e n d i s p e n s e d w i t h . However
in p r i e s t l y m a t e r i a l s c e r t a i n c o m m o n 'reasons' appear alterna-
tively w i t h or w i t h o u t the c o n j u n c t i o n . E x a m p l e : (kl) pesah hu^
leyhvh (Ex 1 2 1 1 ) ; (ki ) t m e H ' , (kl) 1 ani yhwh . See Ex 29 1 " *
The r e l a t e d clauses may come in either s e q u e n c e : c a u s e - - ( T H E R E -
FORE) - - e f f e c t or e f f e c t - - ( B E C A U S E ) - - c a u s e . But sentences of this
k i n d should not be forced too c o n f i d e n t l y into temporal or log-
ical r e l a t i o n s h i p s , just b e c a u s e c o n j u n c t i o n s are r e q u i r e d in
t r a n s l a t i o n and therefore are 'understood' in H e b r e w . 1 0 To
t r a n s l a t e Pr 1 8 2 2 - 31B XXO K X O hat man ein Weib gefunden,
so hat man ein Gut gefunden obscures the equative r e l a t i o n s h i p ,
to say n o t h i n g of the p u n g e n c y - - h e found a wife = he found a
good. To add c o n j u n c t i o n s I F . . . T H E N . . . y i e l d s a c c e p t a b l e Eng-
lish. But the l a c o n i c H e b r e w shares w i t h o t h e r p o e t r y the art
of u s i n g a c r y p t i c s u r f a c e structure w h i c h in many instances
seems to intentionally r e a l i z e more t h a n one p o s s i b l e deep struc-
ture .
arx"'5 D'j?Dy c o
3.10.6. Apposition Instead of a Relative Clause
A c l a u s e p l a c e d in a p p o s i t i o n to a n o u n may f u n c t i o n as an alter-
n a t i v e r e a l i z a t i o n of a r e l a t i v e c l a u s e . This k i n d of c o n s t r u c -
t i o n has already b e e n d i s c u s s e d in #1.7 above. A further illus-
1u
tration is provided by Ge 1 0 : wP et-kasluhim ^ser yase^
missm pelistlm, and Kasluhim, where the Philistines came from.
In Ge 1 0 1 0 " 1 1 the same deep relationships are realized by means
of apposition: bePeres fin^r <A> min-h^res hahl^ ys^
^ asfs ilr.
Other examples: Ge 2 1 1 ' 1 3 ' 1 " , 4 2 0 ' 2 1 , 1 0 e ' 1 2 , i 4 *s7,, 7^

19 , 23 > , 3 5 s 1 9 . 2 0 2 7 , 361 , 4 8 \
37 39 2 19
It is a well-known fact that ^Sser is thus dispensed with

more often in poetry than in prose.
3.10.7. Apposition Instead of Sequence
This is rare. The only example I can find in Genesis is 42311 .

After the brothers have established their bonafides, Joseph
will return the hostage. This consequence is stated simply in
apposition.
NOTES
1
The state of the question is more complicated than this,
since numerous conflicting solutions of the problem have been
proposed. The details do not matter. The enterprise of doc-
umentary analysis becomes less plausible- to the extent that
repetitions and parallels in the texts can be accounted for
as grammatical devices serving rhetorical ends.
2
Paul Joon, Grammaire de l'Hebreu Biblique2 (Rome, 19^7):
p. U87.
3
A. B. Davidson, Hebrew Syntax3 (Edinburgh, 1901): p. 188;
Carl Brockelmann, Hebrische Syntax (Neukirchen, 1956): p. 138.
11
Ge is generally taken as describing decks, although
no noun is used. This would then be a fourth construction fea-
ture. Since petah door is usually masculine (it is, however,
apparently feminine in 2Sa 1 7 9 ) , the last pronoun suffix re-
fers to the ark, not to the door. U. Cassuto finds four dis-
tinct directions in this verse (A Commentary on the Book of
Genesis: Part II From Noah to Abraham CJerusalem, 196U1. pp.
6U-65). On such an analysis the design of the passage is even
more intricate; the specification of rooms, window and door
alternates with measurements of the ark as a whole. Cassuto
does not think that the one cubit measurement in 6 1 6 a p p l i e s
to the window. The problem still needs more work.
5
Rules about the three festivals (Ex 2 3 l l , _ 1 7 ) a r e broken
down in a similar way.
6
Francis I. A n d e r s e n , The Hebrew Verbless Clause in the
Pentateuch (Nashville-New York, 1970): pp. 52-5 ^ -
7
Herbert Chanan Brichto, The Problem of "Curse" in the
Hebrew Bible: JBL Monograph XIII (1963); W. Sibley Towner,
"Blessed be Yahweh.," Catholic Biblical Quarterly XXX (1968):
pp. 386-399; Willy Schottroff, Der altisraelitische Fluch-
spruch : Wissenschaftliche Monographien zum Alten und Neuen
Testament Band 30 (Neukirchen-Vluyn, 1969).
8
Brockelmann, Op. cit.:: 133a.
9
E w a l d K h r , Die Ausdrucksmittel der konjunktionslosen Hypo-
taxe in der ltesten hebrischen Prosa (Leipzig, 1929); Brock-
elmann, Op. cit.: pp. 139-l'3.
10
Brockelmann, Op. cit.: 13^c.
4
COORDINATION
4.0. KINDS OF COORDINATION

The ubiquity of the Cj :ife- and its allomorphs1 is a feature of
Hebrew discourse often remarked on. Literal translation by
AND makes biblical narrative monotonous to Western ears, es-
pecially when long strings of clauses begin with this word.
This endless coordination has been compared unfavorably with
the rich variety of inter-clause connectives found in other
languages, and especially Greek, with its ample stock of con-
junctions. This is hardly fair. Hebrew has its own kind of
versatility and realizes a considerable variety of relation-
ships by means of coordinations of various kinds. Only what
some languages do by means of distinctive conjunctions Hebrew
does by combining AND with distinctively patterned clauses.
Here we shall look at phrase-level coordination only in pass-
ing. Above phrase level we shall concentrate on inter-clause,
that is, sentence-level, coordination. At the same time, the
coordination of individual clauses with higher-level units,
such as sentences or even paragraphs, will be included. The
sequential coordination of clauses on paragraph-level by means
of WP and WS 'consecutive' verb constructions is only marginal
to sentence grammar.
4.1. UPPER-LEVEL COORDINATION

4.1.0. Heterogeneous speeches
When a single speech is composed of two or more sections, each

of which is a discourse of a different genre, the transition
from one kind of utterance to another is marked in various ways.
Simple apposition is often used, or simple coordination. Trans-
ition from declarative or interrogative to precative discourse
can be marked more impressively by weâtt (or we^lm), and
now. The present study is concerned primarily with transitions
62 COORDINATION
between clauses within homogeneous discourse, so we shall not

pursue the subject of heterogeneous discourse any further.
4.1.1. Coordination of Units of Narrative
We shall illustrate the various kinds of coordinating trans-

itions available in Hebrew by means of material from the story
of Pharoah and his two servants (Ge 40). The roles of the
three participants can be staged in many different ways. Any
one may be made central to the drama, or the differing ends
of the two servants can be brought into contrast, and so on.
The study of coordination by Dik referred to in #1.1 above
has shown that this construction is by no means as simple and
straight-forward as commonly supposed. Keeping this caveat
in mind, we shall begin with a simple representation of the
coordination of two units, and C, into a construction A.
The coordinator both links and C together and also marks the
boundary between them. Hebrew does not have any postpositive
conjunctions, although the so-called 'emphatic' we- and ki
could be cast in this role. So this simple picture is not
complicated by problems of sequence. But the tree above al-
ready bgs some questions. It supposes that the coordinator
does not belong to either of the units it joins, or that it
belongs equally to both. Coordination has been discussed as
if and C could stand on their own legs if they were not co-
ordinated, but this idea does not stand up to empirical test-
ing, as Dik has shown. To the extent that C is dependent on
and presupposes B, the conjunction is assigned to C rather
than to . can then be called the lead unit, and we say
that C is coordinated to rather than that and C are co-
ordinated together.
Many possibilities must be provided for. Coordinations
differ in several ways. They differ in the level of the gram-
matical hierarchy on which coordination takes place. They
differ in the formal means of securing coordination. They
differ in the degree of tightness in the junction. They dif-
fer in the deep relationships between the coordinated units.
4.1.2. Stories in Juxtaposition
The simplest way of putting two stories in juxtaposition with-

out any formal connection scarcely deserves to be called a
construction at all. To all intents, the stories are indepen-
dent. We are at the extreme of apposition described in #3.1.
COORDINATION 63
Example: par^5 hesib êt-masqeh 'al-kann <A>

par^ tl êt-^peh,
Pharoah returned his butler to his post.

Pharoah executed his baker.
4.1.3. Coordinated Stories
If there is any continuity at all between two stories, then

the signals of this continuity, such as the use of the same
characters in both, serves in lieu of formal conjunctions to
link them. We shall then speak of two stories in the same
saga or of two episodes in the same story. In English it is
not usual to begin a new story or even a new episode with AND.
In Hebrew, however, this is quite common, and some books begin
that way.
4.1.4. Story-level Episodes
In Hebrew, transition to a new episode in a story is charac-

teristically marked by wayehl, and it came to pass, followed
frequently by an episode-marginal time reference that secures
a time connection between successive episodes.
Example: *wayyseb par ^ êt-masqeh âl-kann Ep j
wayehl Trans' it ion
^ aljare-ken
'ahare haddebrim hêile Time
1
ahre hseb
par ^ "*et-masqeh âl-kann Margin
ka 1 s er]hellt
wayyitel par^ êt-^peh Ep ;
And-returned Pharoah his butler to his post
And it came to pass
after that
after these things
after
when returned Pharoah his butler to his post
as
and-executed Pharoah his baker
There are other options besides the ones shown here. If the
baker is a new dramatis persona, and his execution is the first
event in Ep2, the new episode could begin by introducing him
(#2.6.5): *l?par^ ^pe wayyitel ^t par^, now Pharoah had
a baker, and Pharoah executed him.
6k COORDINATION
4.1.5. Episode-level Paragraphs
In classical H e b r e w n a r r a t i v e prose the onset of a new p a r a -

g r a p h is o f t e n m a r k e d by u s i n g an explicit n o u n subject to re-
fer to the p r i m e p a r t i c i p a n t , w i t h o u t i n t e r r u p t i n g the sequence
of WP clauses. A break in the s e q u e n c e of WP clauses m a r k s
e i t h e r the onset of a new episode or the d e s c e n t to s e n t e n c e -
level c o o r d i n a t i o n (Chapter 5). Once the subject is i d e n t i f i e d
e x p l i c i t l y in the opening c l a u s e , a n a p h o r i c p r o n o m i n a l refer-
ence is likely to be p r e f e r r e d w i t h i n the b o d y of a p a r a g r a p h .
By this c r i t e r i o n the following clauses c o n s t i t u t e s u c c e s s i v e
p a r a g r a p h s , but w i t h i n the same e p i s o d e .
E x a m p l e : *wayyset par^ ^ e t - m a s q e h âl-kann

w a y y i t e l par' ^ e t - ^ p e h u ,
And-returned Pharoah his butler to his post.

And-executed Pharoah his baker.
H e r e there is m o r e c o n t i n u i t y t h a n f o u n d in #4.1.4, but h e r e ,

as t h e r e , the stage is o c c u p i e d by only two p a r t i c i p a n t s . a t a
time.
4.2. PARAGRAPH-LEVEL COORDINATION
4.2.0. Introduction
The o u t s t a n d i n g feature of H e b r e w p a r a g r a p h s t r u c t u r e , at
least in n a r r a t i v e p r o s e , is the e x p l o i t a t i o n of the tense-
a s p e c t contrasts of the verb s y s t e m to achieve d i f f e r e n t kinds
of staging. Two kinds of p a r a g r a p h - l e v e l c o o r d i n a t i o n are
available. O n e c o n s t r u c t i o n , u s i n g the 'consecutive' con-
j u n c t i o n , stages two events as if in s u c c e s s i o n ; the o t h e r ,
using the n o n c o n s e c u t i v e c o n j u n c t i o n , stages two events as if
contemporaneous.
4.2.1. Sequential Coordination
The c o n s t r u c t i o n type i l l u s t r a t e d in #4.1.4 implies a time

b r e a k or a time lapse (which c o u l d be e x a c t l y s p e c i f i e d in
the time margin) b e t w e e n the two e p i s o d e s . Two p a r a g r a p h -
level WP clauses in immediate s u c c e s s i o n imply that the events
they report o c c u r r e d in immediate s u c c e s s i o n in time. In con-
trast to the c o n s t r u c t i o n - t y p e i l l u s t r a t e d in #4.1.5, the lack
of an explicit n o u n s u b j e c t in the s e c o n d of two such consecu-
tively c o o r d i n a t e d c l a u s e s , w h e n there is no change in s u b j e c t ,
shows that w e are still in the same p a r a g r a p h . The c o m m o n sub-
ject does double duty for b o t h c l a u s e s , and u n i f i e s t h e m in
the fabric of the p a r a g r a p h .
COORDINATION 65
Example: *Vayy3eb par^ 1 et-magqehu âl-kann

wayyitel êt-^peh,
and-returned Pharoah his butler to his post

and-he-executed his baker.
Although the action is presented in two successive moments,

all three participants are on the stage. Pharoah is central
throughout; the others are brought to the centre with him in
turn. The episode-level paragraphs illustrated in #4.1.5. are
inbetween the story-level coordination of #4.1.4. and the para-
graph-level coordination of #4.2.1 in the matter of time se-
quence. #4.1.4 indicates a time interval; #4.2.1 implies im-
mediate succession; #4.1.5 is neutral in this regard.
4.2.2. Paragraph-Level Circumstantial Clause.
In order to stage two events as simultaneous or contemporan-

eous it is necessary to break the paragraph-level chain of
consecutive (WP) clauses. The commonest means of doing this
is the CIRCUMSTANTIAL (Cir) clause. A circumstantial clause is
dependent on a paragraph-level WP clause, which it may precede
or follow; or it may be marginal to a paragraph as a whole. The
circumstantial clause represents a second event as occurring at
the same time as that reported in the lead clause. The event
reported in the lead cl ause is in the centre of the stage; the
event reported in the circumstantial clause is to the side.
Example: *par^ hesib êt-mafqeh âl-kann

wayyitel êt-^peh,
and (when) Pharoah returned his butler to his post

(and) he executed his baker.
The anaphoric pronominal reference in the second (WP) clause

indicates that the same paragraph continues. The prominent ex-
plicit noun subject in the first clause is a kind of topicali-
zation. Such a circumstantial clause often marks paragraph on-
set, and functions as a time margin to the paragraph.
Example: *wayy5eb par êt-mafqeh âl-kann

weh 1 tl êt-^peh,
and-Pharoah returned his butler to his post

and Cat the same time1 he executed his baker.
Here the subject pronoun in the circumstantial clause is needed

to break the chain of WP clauses, for wtl is avoided as
ambiguous. (It can be either past or future). Such a circum-
stantial clause often marks the end of a paragraph. A repeti-
tion of the explicit noun subject par '" could have achieved
the same effect, but this might have suggested transition to
a new paragraph. In the first example above the baker is in
the centre of the stage; in the second his execution is sec-
66 COORDINATION
ondary. Considerations of tense and aspect show that several

different deep-structure temporal relationships are possible.
In a negative way the circumstantial clause achieves some kind
of simultaneous staging of two events if only by avoiding the
distinctively sequential staging of WP clauses. Even if the
events must have occurred in succession, they are represented
as if they had occurred at the same time. Since WP clauses
represent events as completed, they have a punctiliar effect,
whereas circumstantial clauses are often linear or durative.
This is why the translation of the circumstantial conjunction
by while is often appropriate.
4.2.3. Paragraph-Level Adjunctive Clause
Another way of throwing a piece of information alongside the

main thread of narrative is by means of an ADJUNCTIVE (Aj)
clause containing a suspended topic and a resumptive pronoun.
Example: *wayy?eb par' "'et-maSiqeh âl-kann

we'et- 1 peh ^t tl,
and Pharoah restored his butler to his post,

and, as for his baker, he hanged him.
4.2.4. Paragraph-Level Surprise Clause
Something unexpected can be brought in alongside the main stream

of narrative by using vShinne.
Example: *wayyseb par'" êt-masqeh âl-kann

wehinne par'" tl êt-^peh,
and Pharoah restored his butler to his job;

and-what-do-you-know '. --Pharoah executed his baker.
4.3. SENTENCE-LEVEL COORDINATION
4.3.0. Sentence or Paragraph?
Circumstantial, adjunctive, and surprise clauses could be con-

sidered sentence-level coordination with an adjacent clause;
but all of them, in another sense, are marginal to paragraph
structure.
The coordinations described in the present section are indu-
bitably sentence-level, integral two-clause constructions which
definitely imply simultaneity. Although we- and is used in all
of them, various combinations of clause-level sequence patterns
secure various relationships that bring out the similarity or
the contrast between two actions performed contemporaneously by
one or two participants. Once more a distinction has to be
made between the optimum realization of a deep relationship of
COORDINATION 67
concomitance or contrast and the alternative realization of the

same relationship in a less characteristic construction.
4.3.1. Conjunctive Coordination
A conjunctive sentence is the most neutral of all the coordin-

ation sentences. It is unmarked for sequence, so simultaneity
is implied, but not highlighted. It does not mark the events
in the two coordinated clauses for either similarity or differ-
ence, but, since a sentence is more likely to be marked to bring
out differences, similarity is generally implied by the very act
of coordinating the clauses. The similarity, however, is not
played up.
Example: *Par^ yiib "'et-maiqeh ''al-kann

weyitle "*et-^peh,
Pharoah will restore his butler to his post

and he will execute his baker.
This is quite matter of fact. The coincidence of the acts in

time is not explicitly marked; but the similar grammatical
structure of the two clauses and the common subject suggest
that the acts are somehow similar. This has the effect of
focussing on the importance of Pharoah and playing down the
contrast between the fates of the two slaves.
4.3.2. Chiastic Coordination
This construction achieves the most complete integration of

two clauses to represent actions of two participants as two
sides of a single event. The simultaneity of the two actions
is implied, and also their similarity; or rather, the facts
that they might have occurred in sequence, and that they were
indeed alike, are left out of the picture.
r
Example: *vayyeb par ^ et-ma?qeh al-kann
w?'et- "^peh tl,
and Pharoah restored his butler to his job,

while his butler he hanged.
This is like the construction actually used in Ge 4 0 2 1 " 2 2 .

The author is indifferent to the contrast in the fates of the
servants, and describes the events in a somewhat detached man-
ner as a dual act of Pharoah. Classical Hebrew does not have
any way o f s a y i n g Pharoah reinstated and executed the butler
and the baker respectively.
To link the fates of the two servants more closely, it w o u l d
be better to leave Pharoah out of the picture altogether, and
make the servants the subjects of the chiastic clauses.
68 COORDINATION
Example: *wayyab hamma'q.e âl-kann

weh^pe nitl,
and the butler was restored to his job

and the baker was executed.
4.3.3. Alternative (Disjunctive) Sentence
Two actions are only likely to be represented as alternatives if

they are in someway similar. Pharoah will do one of two things.
Example: *par^ ySIb 1 et-ma?q.eh âl-kann

"" yitle ^ t - ^ p e h ,
and Pharoah will either return his butler to his post

or he will execute his baker.
This is strained because the two acts are not necessarily mutu-
ally exclusive.
4.3.4. Contrastive Coordination
If it is desired to state two related events so as to bring out

a contrast in the involvement of the two participants, the words
referring to them are both brought into a position of prominence
before the verb.
Example: * 1 et-hammasqe heslb par^ âl-kann

weêt-ha^pe tl,
The butler he restored to his job,

and (but) the baker he executed.
This is like the construction actually used in Ge 41 1 3 . The

butler was obviously impressed by the difference!
4.3.5. Antithetical Coordination
The contrast between Pharoah's two actions is already secured

by the fact that heSIb. . .âl-kanno and tl are to some extent
antonyms. To bring out the antithesis more forcibly, some
kind of negation would probably be used.
Example: *wayySeb par^ êt-masqeh âl-kann

vre ' et--* peh l^ hesib
Iki tl ^t ]
wayyitel ^tj
and Pharoah restored his butler to his job;

(but) he did not restore his baker;
on the contrary, he executed him.
Compare the construction in Ge 4 0 2 3 .

COORDINATION 69
In Chapter 14 we shall examine in m o r e d e t a i l the v a r i o u s kinds

of a n t i t h e t i c a l r e l a t i o n s h i p s and their typical surface reali-
zations .
4.4. INCLUSIVE A N D E X C L U S I V E COORDINATION
4.4.0. Deep Grammar
The story of P h a r o a h and his two slaves does not c o n t a i n deep re-
lationships a p p r o p r i a t e l y r e a l i z e d by inclusive and exclusive
constructions. We shall b e g i n w i t h p h r a s e - l e v e l c o o r d i n a t i o n .
4.4.1. Inclusive Coordination
Simple c o o r d i n a t i o n resembles addition.
Example: kol-hfm (=) h ' n S i m w e h a n n s l m weha^âp,
all the people consist of the men and the women and
the infants.
The deep grammar of inclusive a n d exclusive c o o r d i n a t i o n can be

i l l u s t r a t e d by c o n s i d e r i n g a set of, say, three elements.
A = {, A 2 , A 3 >
Example: (kol-)h^m - O n s i m , n s l m , tap}
(all) the people consist of {men, women, children}
Inclusive c o o r d i n a t i o n s p e c i f i c a l l y includes an item implied by

the use of a c o m p r e h e n s i v e n a m e for the set. Thus k o l - h r m gam-
h a t t a p , all the people also the children, a l t h o u g h Strictly speak
ing r e d u n d a n t , makes it clear that c h i l d r e n are included. It
counteracts the p o s s i b i l i t y of restricting the m e a n i n g of k o l -
f
h m to adults b u t , in that very a c t , ironically drives kol-
h ^ m further into that s p e c i a l i z a t i o n . (In some t e x t s , the spe-
c i f i c a t i o n of w o m e n and c h i l d r e n restricts f am further to adult
m a l e s . ) The phrase thus becomes a regular c o o r d i n a t i o n - - a l l the
CadultJ people together with the children. Yet even in this con-
s t r u c t i o n gam h i g h l i g h t s the i n c l u s i o n of the children. For this
reason we study all the uses of gam together in Chapter 12 u n d e r
the h e a d i n g of inclusive c o o r d i n a t i o n .
It s h o u l d be added that e v e n in the d e f i n i t i v e inclusive coor-
d i n a t i o n c o n s t r u c t i o n , w ? - can alternate w i t h gam a n d s h o u l d be
t r a n s l a t e d even or the like. It is also p o s s i b l e that the in-
c l u d e d item is not just an element of the set (the children are
part of all the peopleJ, but comprises the whole set. The use of
a c o n j u n c t i o n is t h e n s t r i c t l y a p p o s i t i v e , as in h e n d i a d y s . Since
this is a phrase t y p e , w e do not study it further here. The m a i n
point to be e n f o r c e d is that simple c o o r d i n a t i o n a n d inclusion
70 COORDINATION
differ in deep structure, with we- the normal surface conjunc-

tion for the former, gam for the latter. But gam can be used
atypiclly in the former and we- can be used atypically in the
latter.
To call gam a conjunction, THE inclusive conjunction, applies
only to its use in a construction A gam-B. If, however, there
is no trace of anything corresponding to A, not even implicitly,
some other function must be ascribed to gam. By such a negative
test it may, in some occurrences, be called a phrase-level mod-
ifier, or a clause-level modifier ('adverb') 0 r emphasizer.
4.4.2. Exclusive Coordination
Exclusive coordination resembles subtraction and, in well-formed

constructions, represents a qualification of the comprehensive
use of kol.
Example: kol-haâm MINUS hatâp ,
all the people except the infants.
Quite a number of conjunctions are used to realize this exclu-

sive relationship-- 1 ak , 1 ak ki , "*im lo1 , 'epes, "'epes kl , bil-
fdl , bil^d.1 raq, biltl, biltl U i , zlti, ki îm, 15bad,
lebad min, lebad meâser (Est 4 1 1 ) , raq, raq 1 ak.
The length of this inventory is surprising. Finer distinctions

within the set, and constraints on interchangeability, await
investigation. Some are characteristically phrase-level rather
than inter-clause. The ones which are etymologically negative
label an exclusion after an assertion (all the people except
Ciut not] the children); others label an exception after a ne-
gation (none of the people except Lbut onlyl the children).
If the excluded item is contrasted, not with the complete set,
as in well-formed exclusive coordination, but with the residue
of the set after the exclusion, we have moved to antithetical
coordination--h 1 ansxm wehannslm wel^ h a ^ a p , the men and
the women but not the children. Some exclusive conjunctions
can be used as alternatives to we- in this kind of antithesis.
So here again, deep grammar, and not the conjunction as such,
must be diagnostic of coordination type. To call the 'parti-
cles' above conjunctions implies their use in a construction
A <Cj> B. If, however, there is no trace of anything corres-
ponding to A, not even implicitly, some other function must be
ascribed to any of these 'particles' which so occur. By such
a negative test it may be necessary to recognize these par-
ticles, in some occurrences, as phrase-level modifiers, or as
clause-level modifiers ('adverbs'). Thus raq only sometimes
has a restrictive or limitative function, without being a con-
junction.
COORDINATION 71
To sum up
Typical Anternative Not a Cj
we- <Cj > <Inc>, <Exc>, <Ant> Exegetical
gam <Inc> <Cj > Emphatic
raq <Exc> <Ant> Restrictive
4.5. INTER-CLAUSAL RELATIONSHIPS IN PRECATIVE

AND PREDICTIVE DISCOURSE
The examples used in 4.2 - 4.4 above, and most of the prime il-
lustrations used in the following treatment are drawn from nar-
rative prose which uses declarative clauses in which verbs are
indicative. The same relationships are found, mutatis mutandis,
in precative discourse which uses VI, VJ, or VC. But prescrip-
tive and predictive discourse are not easy to distinguish in
Hebrew, on account of the fact that any declarative clause in
future tense can be used precatively in the right context.
The use of the various devices of apposition and coordination
in an extended precative discourse is illustrated in the in-
structions given for making the tabernacle (Ex 25-30). Here the
same kinds of construction occur again and again, and not all
the examples are cited in this monograph.
Ex 25 1 -27 19 is reported as a single speech. It begins speak
( V I ) to the Israelites and let them get (VJ) materials...and
they will make (WS) me a shrine (Ex 25 l " e ). After that, each
successive instruction for each new item or for each new stage
in the work is given in a WS clause. There are 75 of these.
Most of them are precative, in continuation of the lead clause,
and their use implies the performance of the actions they des-
cribe in temporal sequence. A few of them, however, notably
2 5 2 2 2 8 t > 3 7 , predict standing results, and do not prescribe
specific acts. Apart from these divagations, the chain of
sequential WS clauses is broken only at 25 1 9 , where we- + VI
is an optional variant (va fase is equivalent to the usual we-
f
sit, but does not imply sequence), and at 26 1 , where ve-
to +VP:tarase is another optional variant^ (The VI's in 25"
are in a parenthetical exhortation; the we- +0 +VP clauses
in 26 29 serve to return to the topic of the frames.)
The line of WS clauses is thus clear and strong. It gives
the entire speech a texture that is almost uniform from begin-
ning to end. The division into paragraphs and sub-paragraphs
must be guided by content, by changes in topic, or grammati-
cal signals of explicit noun subject followed by anaphoric pro-
nominal reference, These topics are the ark (25 10 " 2 ), the
table (252 3 " 3 0 ), the lampstand (253 1 -1,), the tabernacle (26 "3 ),
the curtain (26 3 1 " 3 7 ), the altar (27 1 ' 8 ), the court (27 9 " i9 ).
The only break in the uniform series of WS clauses is at 2 6 1 .
Within each topic there may be several subtopics, such as the
72 COORDINATION
parts of the ark--the box, the rings, the poles, the cover-
ing, etc. But here too the discourse moves along from one to
the next without a break in the chain Of WS clauses. In other
words, the arrangement of the architectural topics is hier-
archical, but this is not reflected in the grammatical struc-
ture of the discourse.
Shrine furniture
Apart from the break at 261 a uniform string of WS clauses

handles all three levels of this hierarchy. But the lower
levels, dealing with materials and dimensions, are handled
by grammatical constructions below paragraph level, notably
apposition and coordination.
The commonest device for doing this is apposition. Nearly
20 individual clauses and 10 conjunctive sentences are thus
used to specify (#3.7.1) or expound (#3.7.2). There are three
summarizing clauses in apposition (#3.7.4), and five examples
of an apposition sentence used in apposition (#3.9). 25 2 0
even has a paragraph in apposition, for here the WS clause
is equivalent to a subclause of 'Result'.
Of more interest in the present context is the use of sen-
tence-level coordination. 25 1 2 and 3 8 have a conjunctive sen-
tence and an apposition sentence used circumstantially, al-
though such material is more often placed in apposition. But
most of the circumstantial constructions in the passage are
bonafide, whether clauses ( 2 5 2 0 a B > 3 2 , 26* b 2 e ) or sentences
( 2 5 35,3-39> 2 6 12-13)i
This leaves the few chiastic sentences for special notice.
Their use is not just an arbitrary stylistic variation. They
bring together pairs of related actions: 25 2 1 (put the kap-
preth on the box and put the redt in the box), 26 s 5 (the
symmetrical locations of the table and the lampstand)^ There
is a more intricate pattern of chiasmus in 261 2 2 , we^sit
the frames -- 20 frames for the S side
and for the side 20 frames
and for the W side taâse 6 frames
The fulfillment in Ex 3 6 2 3 " 2 7 has î in each of the chias-
tic clauses.
Other material is interspersed with this material. The re-
sult is complex but not chaotic. 26 1 8 is simultaneously and
independently the lead clause of two distinct sentences. With
26 1 9 it makes a circumstantial sentence (and all this is re-
peated in 2 6 2 0 " 2 J , but without verbs). With 2 6 2 2 , 26 1 makes
a discontinuous chiastic pattern, the one verb governing all
three objects. The further reference to the two corner-boards
in 26 2 3 is then chiastic, in another way, to 26 2 2 . There can
COORDINATION 73
hardly be a question of artistic decoration of such, dry facts.

The careful use of chiasmus in an architect's specifications
suggests that it has a precise grammatical meaning and a spec-
ial discourse function as well as a pleasing aesthetic effect.
That a chiastic sentence blends two activities into a single
scene is confirmed by the structure of the speech in Ex 40 1 "
1S
. It consists of a marginal time reference (40 2 a ) followed
by 34 clauses describing 34 acts to be performed in success-
ion. There are no apposition clauses and no circumstantial
clauses. The first clause is prescriptive, with VP. This is
followed by a string of WS clauses, 32 in all. These contin-
ue the prescription in the usual way. This chain is broken
only once, at 40 1 1 , a , which is chiastic with 4 0 1 2 a . The
effect is to stage the ordination of Aaron and his sons as
a single, two-sided, event. Otherwise all the events are
singular and sequential. In the fulfilment that follows the
same pattern is preserved. The time margin ( 4 0 1 7 a ) is fol-
lowed by a VS clause (40 17t> ) and then by a string of sequent-
ial WP clauses, broken only by a chiastic VS clause in 4 0 2 9 .
This presents the placement of the incense altar and the
holocaust altar as two aspects of a single arrangement.
A curious result follows. When the material in two co-
ordinated clauses contrasts, this relationship is realized
primarily in a sentence in which the clauses are arranged
with symmetrical patterns. When there is balance between
the content of two coordinated clauses, this relationship
is realized primarily in a sentence in which the clauses
are arranged a-symmetrically.
4.6. SUMMARY
The preceding discussion has recognized that the Hebrew

raconteur had at his disposal several different ways of
telling essentially the same story about Pharoah and his
two slaves. And it has also included under the broad
rubric of coordination various kinds of disjunctive,
inclusive, exclusive, and antithetical relationships,
making ten kinds of nonsequential interclausal coordin-
ation in all.
The system has two axes, polarized for similarity versus

dissimilarity and for succession versus simultaneity.
Tfc COORDINATION
C o n j u n c t i v e sentences are neutral in all r e s p e c t s . Further-

m o r e , interclausal r e l a t i o n s h i p s not p o s i t i v e l y m a r k e d for
time s u c c e s s i o n by the u s e of sequential c o n j u n c t i o n - v e r b
c o m b i n a t i o n s (WP a n d WS) stage events as c o n t e m p o r a n e o u s ,
or at the v e r y least as not s u c c e s s i v e , in all other sen-
tence types. S i m i l a r l y the staging of two events in
c o n t r a s t i v e , e x c l u s i v e , a n d antithetical sentences m a r k s
them for d i s s i m i l a r i t y and so, by d e f a u l t , implies that
events r e p o r t e d in other s e n t e n c e types are in some w a y
s i m i l a r , or, at least, n o t d i s s i m i l a r . The chart above
implies, h o w e v e r , that s i m u l t a n e i t y is m o r e to the fore
in c i r c u m s t a n t i a l , a d j u n c t i v e a n d surprise c l a u s e s , w h i l e
only i m p l i e d in the other n o n s e q u e n t i a l s e n t e n c e types.
S i m i l a r l y w e suppose that the m a t c h i n g pairs (chiastic-
contrastive, inclusive-exclusive, disjunctive-antithetical)
s t a n d at the extremes of the s i m i l a r - d i s s i m i l a r p o l a r i t y ,
all six sentence types in the m i d d l e c o l u m n being neutral
in this regard.
4.7. BACK-LOOPING (RANK-SHIFTING)
So far w e h a v e i l l u s t r a t e d the ten kinds of c o o r d i n a t i o n

s e n t e n c e in their simplest form. T h r e e (circumstantial,
a d j u n c t i v e , a n d surprise) are p a r a g r a p h - l e v e l , a n d w e
have i l l u s t r a t e d them by m e a n s of examples in w h i c h the
c o o r d i n a t e d element is a single clause. The r e m a i n i n g
s e v e n are s e n t e n c e - l e v e l , a n d w e have i l l u s t r a t e d them
by m e a n s of the simplest k i n d of t w o - c l a u s e c o n s t r u c t i o n s ,
emphasizing in p a r t i c u l a r the c o n t r a s t i v e - d i s t i n c t i v e
features of the s e v e n kinds of c o o r d i n a t e d c l a u s e , a n d
in some instances the total p a t t e r n that the c o o r d i n a t e d
c l a u s e m a k e s w i t h the lead clause. Fuller study w o u l d
n e e d to track d o w n all the c o m b i n a t i o n s , w i t h a list of
c l a u s e types that m a y be leads in each s e n t e n c e type as
w e l l as a list of c l a u s e types that m a y be c o o r d i n a t e d
w i t h them i n e a c h s e n t e n c e type. The following chapters
r e p r e s e n t a b e g i n n i n g of this w o r k , w i t h o u t being ex-
haustive .
But m o r e is involved t h a n simple t w o - c l a u s e s e n t e n c e s .
By b a c k - l o o p i n g (otherwise k n o w n as r a n k - s h i f t i n g ) a
s e n t e n c e , a p a r a g r a p h , or e v e n t h e o r e t i c a l l y a w h o l e
episode m i g h t f u n c t i o n just like a single c l a u s e in any
one of these s e n t e n c e types. We have a l r e a d y seen, es-
p e c i a l l y in #3.7.2 and # 3 . 7 . 3 , that a s e n t e n c e or p a r a -
g r a p h c a n be in a p p o s i t i o n w i t h a c l a u s e . In #3.7.3 w e
s h o w e d that a n entire p a r a g r a p h consisting of ten clauses
stands in a p p o s i t i o n w i t h a single lead c l a u s e . T h e same
k i n d of thing h a p p e n s in c o o r d i n a t i o n and s u b o r d i n a t i o n .
'cr <A> Cl'

Se <Co> Se
IT H
COORDINATION 75
This creates a terminological difficulty. Longacre surmounts

this by calling any filler of such a 'sentence1-level slot a
base.1 Furthermore, if such constructions are called 'sen-
tences', our initial definition of a sentence as a construc-
tion relating two clauses together works for only some of
them. This is why we began with an ostensive definition
(#2.1), and now that our list of illustrations is more
complete, we can enlarge the definition to include con-
structions in which any two bases are related together by
the kinds of apposition and coordination we have described.
A 'complex' sentence, which we prefer to describe as a
kind of clause (#2.3), is a construction in which any two
bases are related together in some subordinative way. So,
while constituent clauses may be diagnostic of sentence
type, and formal linkage by means of a conjunction or by
none is demarcative, it is really the deep-structure re-
lationships between the constituent bases that are defin-
itive of sentence types.
4.8. ALTERNATIVE SURFACE REALIZATIONS
This rich apparatus of sentence types equips the speaker

of Hebrew with diverse means for staging the details of
discourse in all kinds of ways. In narrative prose the op-
tions of succession or simultaneity and of similarity or
dissimilarity are available, together with neutrality in
both respects. These options are not completely open.
In telling a story there may be relationships inherent in
the deepest semantic structure that can be properly real-
ized in only a certain number of sentence types. Thus the
opposite fates of Pharoah's two slaves contrast in fact.
The raconteur may choose to highlight or to play down
this contrast, but there are constraints. To attempt to
stage antithetical facts as if they were similar, by using,
say, an inclusive sentence, imposes too great a strain on
the surface structure. Either the deep structure forces its
way into the surface, and twists it into another meaning
(as when exclusive sentence by form have antithetical mean-
ing by semantic content); or the surface structure holds
its own, filtering out or transmuting the semantic struc-
ture, frustrating the intention of the speaker.
In #4.4 above we have recognized three degrees of cor-
respondence between deep and surface structure. There is
an optimum surface realization of a deep relationship which
we have called 'typical'. Then there may be an alternative
realization by means of a surface structure which is com-
patible with that deep relationship but which is either
neutral or typical of another deep relationship. Thus the
form of a conjunctive sentence can be used as an alternative
albeit insipid realization of, say, an inclusive relationship.
Finally there are some sentence types which are incompatible
with certain deep relationships, and their use in an attempt
to realize such relationships would only betray the incompe-
76 COORDINATION
tence of the speaker.
4.9. EMPIRICAL TESTING
In keeping with the empirical thrust of this study, we be-

gin with surface structures in classifying constructions.
By using artificial sentences to illustrate the various
sentence types, we have been able to exhibit the contrast-
ive-distinctive features of each in optimum form. Not all
natural examples are so clear-cut. The validity of the
theory outlined above can be tested only by listing texts
from the sources to see if the distinctions made serve as
a guide to their correct interpretation.
Here there are traps for the unwary. While the incidence
of a construction of a given type is not an entirely reli-
able index of its grammatically, it does not follow that
a construction rarely used is necessarily less grammatical.
Chance factors are at work here. When atypical usage is
recognized as an alternative realization of a different
deep structure, the clash in usage is resolved, and the
atypical form cannot be adduced as evidence against a gen-
eral rule based on the use of the same form as the typical
realization of its deep structure.
A further untidiness in the results arises from another
cause. Many natural examples fall short of the maximum
contrast achieved in well-formed constructions. There may
be enough redundancy in the optimum realization to pre-
serve its categorical identity even when it is not fully
realized. This is why we regard form features as diagnostic
rather than definitive. When two constructions are con-
trasted by only one or two distinctive features, a falling
short of full realization can reduce both to formal identity.
The discovery of such ambiguity does not mean that only one
category should be used; it only means that some constructions
are indeterminate as to category. The present study admits the
existence of boundaries where categories merge. It helps to
explain the differences of opinion among Hebraists in such
matters as the labelling of individual clauses as 'circum-
stantial '.
NOTES
1
T h a t t w o m o r p h e m e s a r e i n v o l v e d is p r o v e d b y t h e c l a s s i c -
al t e s t of m i n i m a l c o n t r a s t . T h e d i f f e r e n c e in t e n s e b e t w e e n
weêq-fl, a n d I shall kill, a n d w ^ e q t l , and I killed, is
s e c u r e d s o l e l y by the c o n t r a s t i n g forms of t h e c o n j u n c t i o n s .
W e r e f e r t o t h e 'vv-consecutive' b y m e a n s of t h e s y m b o l W .
The sequential verb forms W +VP (past) we call b r i e f l y WP and
W +VS (future) we call WS.
2
Hierarchy and Universality..., p. xiv.
5
CIRCUMSTANTIAL CLAUSES
5.0. INTRODUCTION
The classic treatment of the circumstantial clause is found in

Driver's study of the tenses. 1 It is no derogation of this fel-
icitous work to point out its lack of the theoretical insights
of present-day linguistics. Previous studies of Hebrew syntax
have not approached the clause hierarchically. Hence Driver
did not perceive some of the discourse functions of circumstan-
tial clauses. For instance, the construction which we identify
as a formal marker of episode boundaries (#5.1) is not distin-
guished from a clause in chiasmus, and its adoption is traced
to "emphasis or the love of variety" (p. 200). He does recog-
nize, however, "the commencement of a new thread" (p. 201) by
means of a circumstantial clause. Since Driver defines the cir-
cumstantial clause rather vaguely as "a subordinate thought"
(p. 202), he includes under the same heading constructions with
quite diverse surface features, including apposition. On the
other hand this approach helped Driver to see what is now called
deep structure by means of correspondences or "transpositions"
(p. 249), a procedure remarkably like the transformations of
modern grammar.
The workhorse of Hebrew narrative prose is the vw-CONSECU-
TIVE WP clause. Each such clause marks one successive event
in a story. It is possible for a paragraph, even a book, to
begin with such a clause. WS clauses do the same thing in pre-
dictive discourse (#4.5). 2
WP and WS clauses cannot be negated. A negated clause in the
chain of events will use wel 1 +VS... (or wel^ +VP...). Such
a clause generally stands in the main stream of narrative (or
prediction). Examples: Ge 2 2 5 , 8 9 " 9 (clearly sequential), 8 1 2
(end of paragraph), 9 1 1 , 9 1 5 , 38 2 0 , etc. 3
Otherwise, any other kind of clause which breaks the chain
of WP (or WS) clauses interrupts the sequence of events, and
generally reports an event contemporaneous, concomitant, or
'circumstantial' to the main stream. The clauses used in the
main stream begin with V (unless negated), so if there is a
subject, the clause has sequence PS. The usual way to break
this succession is to reverse this PS sequence by inserting
78 CIRCUMSTANTIAL CLAUSES
some clause-level tagmeme between the conjunction and verb.

This inserted item is generally the subject, hence +w5- +S +P
is considered normal for circumstantial clauses. But other
items beside the subject can, on occasion, precede the pred-
icator. If the circumstantial clause is negated, some item
besides lo"1 has to be inserted between the conjunction and
to secure the same effect. Example: Ge 9 2 3 b B . (But the
clause in Ge 31 3 2--w?l'1 -yda ( y a f a q b . . . i s clearly cir-
cumstantial, even though it does not bring the subject as
well as the negator before the verb.]
Various kinds of predication are possible in circumstantial
clauses. In narrative prose, VS is compatible with WP, and
in predictive discourse, VP has the same tense as WS. But
participial, quasiverbal, and verbless clauses can also be
used circumstantially, especially in past-tense narrative.
Circumstantial clauses can be attached to other construc-
tions besides paragraph-level WP or WS clauses to refer to
some simultaneous event or accompanying circumstance.
5.1. EPISODE-MARGINAL CIRCUMSTANTIAL CLAUSES
5.1.0. Nucleus and Margin
A circumstantial clause is structurally marginal to a para-

graph. When the paragraph is episode-initial, the circum-
stantial clause may be marginal to the rest of the episode
(its nucleus) as a whole.
EpNuc Mg:CirCl
This tree implies nothing as to the physical sequence of nu-

cleus and margin. A circumstantial clause may come anywhere
in an episode, either as its opening clause (#5.1.1), or as
its closing clause (#5.1.2), or inserted somewhere along the
string of paragraph-level clauses. A circumstantial clause
at the beginning or end of an episode can be an important
signal of episode onset or episode closeout. It becomes a
matter of some importance, then, to distinguish circumstan-
tial clauses at episode boundaries from circumstantial clau-
ses which stand beside the body of an episode somewhere along
its length but do not dissect it.
In #5.2 we shall study circumstantial sentences, properly
so-called. These are strictly two-clause constructions,
consisting of a head clause and a dependent circumstantial
clause. The head clause in such a construction may, of course,
be paragraph-level, and through it the circumstantial clause
is attached to the paragraph as a whole. So the distinction
between an episode- (or paragraph-) level circumstantial clause
and a sentence-level circumstantial clause is a fine one.
CIRCUMSTANTIAL CLAUSES 79
Ep
Head: H <Cir> CirCl
Se
CI
But the point is worth making because circumstantial clauses at

episode boundaries mark major transitions in discourse, whereas
circumstantial clauses "which are integral to a sentence are
quite unobtrusive and often a mere aside.
A paragraph can consist of a single sentence. If it is a two-
clause circumstantial sentence, the distinction between a para-
graph-level and sentence-level circumstance vanishes.
5.1.1. Episode-Initial Circumstantial Clauses
The body of an episode in narrative prose is likely to consist

of a string of WP clauses. (Predictive discourse correspond-
ingly is built on chains of WS clauses.) Such a string is
likely to be preceded by some marginal material, such as a
time reference realized as adverb, noun phrase, prepositional
phrase, infinitive phrase, or clause (#4.1.4). Another way of
opening an episode is to state a preliminary circumstance, and
the usual way of doing this is by means of a circumstantial
clause.
The onset of a new episode is often marked by introducing a
new dramatis persona by means of a circumstantial clause. In
Ge 3 1 a new dramatis persona (the snake) is introduced, and a
new episode inaugurated, by means of a circumstantial clause:
wehannh? hy 'rm mikkl hayyat hassde... now the snake
was the cunningest animal... Other circumstantial clauses
which introduce a new character at the beginning of a new epi-
side are Ge 2 1 0 (Because a participle is used rather than a suf-
fixed verb, this could be circumstantial to preceding, but the
criterion is not sure-fire.), 231 0 , 24 2 9 (introduces Laban) ,
24 6 2 (change of scene), 2 6 1 5 ' 2 6 , 29 1 6 , 3 4 * ' 7 , . 3 6 1 2 , 4 1 1 2 ' 3 2
(moves on to the question of why the dream came twice), 4 5 1 6 ,
46 2 8 , 47 2 ' 1 3 , Ex l 7 , 2 1 6 , etc.
A new development in a story may be marked by a circumstan-
tial clause, even though the subject is not a new character.
So the story of Cain begins: vehâdm yda^ "'et-haww îitS,
and Adam knew Eve his wife (Ge 4 1 ) . After that, regular WP
clauses continue the story.
Other circumstantial clauses which mark a new point of de-
parture: Ge l 2 (NEB translation; but this is not desirable,
since there are counter-arguments that it is circumstantial
to the preceding--s.ee #5.2.1), 2 1 0 a (As usually interpreted,
with critics regarding it as an interpolation. From the gram-
matical point of view it could be conjoined to 2 9 b [Chapter 8]
making two coordinate circumstances of the garden, namely, the
special trees and the river system. It is also possible that
it is chiastic with 2 e so that the garden [or oasis] consists

of vegetation and water.), 2 2 0 b (This is usually taken as ad-
versative to the preceding [v5- but], but it is possible that
the repetition of êzer kSnegd links this clause with 2 1 .
2 l e b introduces the first attempt to supply Adam with a com-
panion; 2 2 0 introduces the second attempt. Neither solution
is entirely satisfactory. The problem is the inherent subject
of mas"1 . If it means But CAdamD did not find Lamong the ani-
malsl a suitable helper, then le^dm would have to be ethic,
which is hard to believe. Comparison with 2 l e b suggests that
Yhwh êlhim is the subject of ms 1 , but this breaks the con-
tinuity between 2 2 0 a and 2 2 0 b , for the change in subject does
not become explicit until 2 2 1 . This is all right if 2 2 0 b is the
preliminary circumstance to 2 2 '--And when he LGodl did not find
a suitable helper for the man, Yahweh God made a deep sleep
fall on the man... A third solution avoids the question of a
subject altogether by reading *muss^ [GPas] [synonym of nims"1
existed] so S:(ezer. This also points to a new paragraph at
20zk--now there was istilll no helper like an equal for Adam,
so Yahweh God made...), 7 6 (supplies a d a t e ) , 8 5 , 9 2 , 1 2 1 6 ,
llf
13 (Note the less usual positioning of marginal time after
the verb, where normal narrative would have *way5hl âl}ar hip-
pred..,wayy 1 mer yhwh... See also Ge 1 8 1 7 . ) , 1 4 1 0 1 8 , 1 6 1 ,
1 8 1 7 , 21 1 (The fact that the ensuing WP clause repeats the
names, rather than using anaphora, is atypical, as comparison
with other examples shows.), 24 1 (This begins the story with
the relevant circumstance that Abraham was old. In Ge 1 8 1 1 a
similar fact is brought out by means of a circumstantial clause
alongside the main stream [#5.1.3]; in Ge 27 l a similar datum
is cast in a time clause after wayShi.), 2 6 1 5 2 6 , 27 6 , 3 1 s ' 1 9 ,
32 2 , 333., 3 4 5 a > 7 , 373a , 39 l , 4 1 5 0 ' 5 6 , 42 2 0 (in prescriptive
discourse), 4 3 1 . 4 5 1 6 . Ex 3 1 , I I 1 0 * 1235. 1410&, I710^.12a,
19 3 , 241 > , 26 , 27 , 29" ' 15 ' 1 7 ' i1 , 33 , 36 3 (This leads on
to the deputation rather than being a further part of the pre-
ceding relative construction.), De 4 2 1 , 9 2 1 , 10 6 , etc.
This device is not always used when it might have been ex-
pected. The chain of WP clauses can continue right through a
natural break in the story. It is entirely an option of the
speaker (or writer) whether to highlight this new point of
departure by using a circumstantial clause, or whether to
smooth it over. Thus at Ge 2 1 1 1 , _ 1 5 there is a new development
--And when the water from the bottle was used up, she threw
the boy under one of the bushes. But the chain of WP clauses
is unbroken. A major new episode begins with a WP clause in
10
Ge 2 8 , and this kind of thing often happens. The means for
optimum marking of episode onset are not always availed of.
5.1.2. Episode-Final Circumstantial Clauses
It is usually supposed that a new paragraph begins at Ge 6 9 .

This leaves Ge 6 , a circumstantial clause, as the close-out
of the preceding material to which it is adversative--But
Noah found gxa.ce in the eyes of the Lord. On the other hand,
in view of Ge 37 1 " 2 , we must ask (#5.1.1) whether 6 8 hegins a
new paragraph in spite of the traditional punctuation. The
first episode in the story of Noah's drunkenness ends, after
an unbroken chain of eleven WP clauses, with a pair of con-
joined circumstantial clauses:
genehem âhrannit
weêrvat ^bihem l^ r^,
And their faces were backwards

and their father's nudity they did not see (Ge 9 2 3 )
This is repetitive, echoing the language of the two preceding

clauses. This is another device of Hebrew story-telling,
whether the grammatical form of a circumstantial clause is
used or not, to recapitulate some climactic point as an epi-
sode closeout.
Sometimes a marginal time reference at the opening of a
story is a note about someone's age at the time of the e-
vents, as in Ge 17 1 . Similar information may be supplied by
means of a circumstantial clause at the end of a story, to
round it off. Examples: Ge 1 6 1 6 , 17 21 - 25 (two clauses con-
joined), 25 2 6 , Ex 7 7 . The same kind of clause sometimes
comes in the middle of a story, but always at some juncture
between episodes (Ge 12 3 , 21 5 , 41"). It is sometimes unclear
whether such a clause ends one episode or begins the next
(Ge 7 6 ) .
Other examples of circumstantial clauses ending an episode
or paragraph are Ge 18 2 2 , 3 3 (Note the chiasmus.), 19,
between 1 9 and 1 1 (Note the chiasmus of êt-lS-fc and êt-
h^nslm.), 24 2 1 (The new episode begins with vayehl.),
27 s , 3 0 6 b (Note the repetition of the subject in the next
clause, whereas the anaphoric b in Ge 32 2 shows that
veyaâqob hlak. . . begins a new episode, even though its
chiasmus with the preceding clause suggests that it ends
the preceding episode. The same relationships obtain between
Ge 33 1 6 and . See also the extended discussion of Ge 18 2 2
below [#5.1.3.2 ()].), 3 2 2 2 , 37 1 1 (Note the chiasmus.
Furthermore, this clause makes an inclusio with veyiSriêl
^hab êt-ysep [37 3 ], the circumstantial clause which opens
the episode.), 37 3 6 (This clause is a detached and delayed
epilogue to the episode almost concluded in 37 2 e ; note
particularly that ^t5 in 37 3 6 is anaphoric to êt-ysep
in 37 2 8 .) , 415If, 5 7 , Ex 1 s t , 12" 0 , 16^ s , 36 7 (Repetitious
and redundant, but its effect is to mark majestically the
end of the episode.), etc.
The majority of these circumstantial clauses at paragraph
and episode boundaries have VS in narrative prose. In future
tense discourse (predictive or precative) VP has a similar
role. Example: Ge 46". Ge 47 1 3 has a quasiverbal circumstan-
tial clause at the onset of a new episode, followed by WP.
Since the same kind of circumstantial clauses can be used

to mark either paragraph or episode onset or paragraph or
episode end, the direction in which such clauses are at-
tached may sometimes be in doubt. Compare the discussion
of Ge 1 2 a and Ge 2 20) in #5.1.1 above. In most cases the
content shows whether the circumstantial clause gives
preliminary circumstantial information, such as the time
setting or something about a new dramatis persona, or
whether it gives a summary wrapup. When two such clauses
come in succession, the episode boundary is likely to be
between them. Thus Ge 13 , although it anticipates a
later theme, ends an episode because a new episode ob-
viously begins with the next clause. A similar sequence
of clauses marks a paragraph boundary between Ge 18 16
and 1 7 .
5.1.3. Circumstantial Clause beside an Episode
5.1.3.0. Circumstance as parenthesis
The best-known circumstantial clauses are those which

come alongside the main thread of discourse. They gen-
erally report some coetaneous event or state, hence the
name 'circumstantial'. For the same reason they are
sometimes described as subordinate or 'adverbial',
and not always distinguished from parenthetical inform-
ation placed in apposition. We pay more attention to
surface features, and distinguish circumstantial clauses
carefully from conjunctionless apposition.
It is a token of this standing alongside the main
time stream that predicators in such circumstantial clau-
ses are predominantly tenseless, neither past nor
future, even when the rest of the discourse is either
past or future. Verbless clauses, and clauses with
participles or quasiverbals as predicators, when used
circumstantially, take their tense from the lead clause
or from the paragraph as a whole.
5.1.3.1. Participial circumstantial clauses
Example: vayyitten lipnehem wehu^-^med âlehem tahat

h ^ e s , and he put [it] in front of them, and [=while]
he [is/was] standing beside them under the trees CGe 1 8 " ) .
Other examples: Ge l 2 t . 2 1 0 , 13 7 , 14J 2 , 1 3 a , 1 3, 15 2 ,

1 8 i iob 19i 246 2 , 25 , 28 12 , 32 3 2 , 372 5 , Ex 2 5 , 5 1 3 ,
9 2 * (like Ge l 2 b ). 132 1 , 148t>,*7 a S ,25 t , 18 i,,bB , 20 1 5 ,
25 32 , 371 , De 4 1 1 , 9 I 5 a , etc.
5.1.3.2. Quasiverbal circumstantial clauses
Since the commonest quasiverbal predicator used in circum-

stantial clauses is U n , is not, does not exist, something
should be said here about its syntax. It has three dintinct
functions.
Ci) It is the negative of existential yes,, (it) exists.
Example: yes-ll r b , exists to-me abundant, I have plenty
(Ge 33 9 ); Negative: ên lh wld, does-not-exist to-her
child, she had no children (Ge l l 3 0 ) . As an existential
predicator, ên or yes is clause-initial, and is used in
independent declarative clauses in apposition or conjoined
or else subordinated by kl (Ge 4 4 3 1 ) .
Cii) It is the negative of the locational-temporal
predicator hinne, it is present here and now. In particu-
lar it resembles the use of hinne in surprise clauses,
which are coordinated (Chapter 7). Negative surprise
clauses begin we 1 en... Example: w(P en mayim listt
h ^ m , and there was no water for the people to drink
(Ex 17 1 ). Other examples: Ge 3 9 1 1 , 41", 2 ", Le 26. It
is even possible to combine the two predicators, hinne
emphasizing the ingredient of surprise. Example: vehinne
1
e n y s e p b a b b r , and behold, Joseph was not in the cis-
tern (Ge 3 7 2 S ) .
(iii) It is the negator of verbless clauses, and, when
such a clause is used circumstantially, 'en comes after
the subject, in contrast with its clause-initial sequence
pattern in (i) and (ii) above. Example: weârba r -jne^t
7
'Is ^ i m m , and four hundred men [are] with him (Ge 3 2 ;
negative: wehannaâr ênenn îttn, and the lad is not
with us (Ge 4 4 3 0 ; note the redundant optional pronoun
subject suffix, and compare Ge 4 4 2 6 , 31*). Other examples:
Ge 1 9 3 1 , De 1 . This kind of clause can also realize
antithetical coordination (see Chapter 14).
The correlation of the sequence pattern we- +S + ên
+ ... with the functions of circumstantial clauses while
the sequence w e 1 e n +S +... is more independent explains
the subtle but important difference between the other-
wise similar clauses in the dream episodes of the Joseph
s t o r y . In Ge 40 a n d 4 1 1 5 p t e r ên ^ t S , and there was
no one who could interpret it, is c i r c u m s t a n t i a l b u t
matter-of-fact. After all, you would not expect to find
a qualified oneiromancer in a prison. But iii Ge 41
(compare 41 2 l t ) w i P e n - p t i r ^tm l e p a r ^ S , and no one
was able to interpret them for Pharoah, is c o n s e q u e n t i a l
but with a touch of surprise -- (ii) above. After all,
where else would you expect to find a good dream inter-
preter, if not in Pharoah's court?
(iv) The same patterns are met with the temporal quasi-
verbal predicator f 5d, still (or now) is.^Contrast the
independent declarative clause r d-ysep benl fcay, Joseph
28
my son is still alive (Ge 4 5 ) , with the circumstantial
clause weh^ ^denn ssun, and (while) he is (was) still
there (Ge 44 1 *), where the use of both the explicit sub-
ject pronoun h' and the subject pronoun suffix -enn
makes possible the circumstantial sequence we- +S +P.
Cv) With these patterns in mind, we can take a closer look

at Ge 18 22t) , already listed in #5.1.2 as a circumstantial
clause marking the close of an episode. There is discontin-
uity between 182 2 t and 1 8 2 3 a . The repetition of the subject
noun Abraham supports the conclusion that there is a major
break between these verses and that 18 2 2 rounds off the
preceding paragraph. But this is not the main problem.
Ge 18 2 2 also stands in chiasmus with 1 8 2 2 a , a pattern
compatible with its concomitant function of ending the
episode which began with the arrival of the three 'men'
in Ge 18 2 . Here the travellers part. Two go to Sodom
(Ge 19 1 ), one remains. Seen in this light, the obvious
subject for the circumstantial clause is the third member
of the party -- Yahweh, not Abraham. Furthermore, this
would secure chiasmus of the two subjects the men and
Yahweh within Ge 18 2 2 followed by a further chiasmus in
the subsequent treatment of these topics in the sequence
Yahweh ( 1 8 2 3 " 3 3 ) and the men (Ge 19 1 ).
Everybody knows that this is the first of the celebrated
Tikkun sopherim, allegedly altered in the interests of piety.
All the versions agree with MT and Ge 19 2 7 supports it. The
first hint that it was ever any different is given in Ber-
eshith Rabba, which quotes the opinion of Rabbi Simeon (2nd
century A.D.), that the original reading *vyhvh fvdnv...
had been altered by the scribes. How can we decide the cor-
rect reading? Hitherto discussion has relied on general ar-
guments. The authority of the (late) notices about the Tik-
kun sopherim is countered by arguments that the scribes had
too much reverence for the text to tamper with it, even
when motivated by concern for the dignity of God. But if
that be so, where did the competing reading come from? How
could the scribes decide between reverence for the text and
reverence for God? The piety that removed it as abhorrent
would never have invented it. R. Simeon must have known a
variant tradition if not text. And variants are known from
the earliest recensions which betray theological anxiety.
If the unanimity of the versions is not quite overwhelmed
by the principle of lectio difficilior, an argument from dis-
course grammar might tip the scales. So long as the two con-
tending clauses are looked at in isolation, their grammar
is identical and favours neither. But as soon as we stand
back and look at sentence structure and paragraph structure,
the odds are against the MT. The framework is provided by
the clauses
18 1 6 wayyqm missm h^nslm
182 2 a vayyipn missm hânsljn
wayyelek sedm
*veyhwh 'wdnw fmd lpny ^brhm
18 3 3 wayyelek yhvh
ve'atirhm sab limqSmS
The parting of the company is delayed by Yahweh's decision

(deliberated in 1 8 1 7 " --a circumstantial clause!) to confide
20 21
his plans to Abraham, which he does in 1 8 " . This episode
ends by stating what the travellers do in a chiastic sentence
in which Yahweh is matched more appositely with the other two
than Abraham is. Then Yahweh's tarrying with Abraham (a dif-
ferent idiom from the objectionable 'md lpny in the posture
of a servant) is matched by Abraham's initiative in approach-
ing with his intercessions (18 2 3 ). This does not prove the
point fully. But it shows how discourse grammar provides the
text critic with an additional tool.
5.1.3.3. Verbless circumstantial clauses'*
Example: nibne-lln îr migdl bassmayim, let us

build ourselves a city and a tower and (with) its head in the
sky (Ge 11") .
Other examples: Ge l 2 a B , 2 19 1 2 , 1 2 6 , 13 2 1 4 1 3 b 2, 16> 18 9 > 7

1 obB, 11 ,1 2% 13 > 2 7 192 20 * 24 '* 25 29 *7a3' 32
33 1 (note the unusual sequence PS. in contrast with 3 2 7 ) ,
3 4 1 9 ' 3 0 , 3 7 2 \ 46 3 2 , Ex 1 4 2 2 , 15 2 *, 1 6 3 1 , 17 9 , 2 1 2 8 , 2 4 i J ' 1 7 ,
2 5 2o,3-35,3 ( t h i s k i n d i s o t e n i n apposition), 26 1 *, 3 4 2 9 ,
3 7 2 0 , Le 23 1 0 , De 2 1 * , 9 l s b , etc.
5.1.3.^. Circumstantial clauses with perfect verbs
The circumstantial clauses described in #5.1.1 and #5.1.2 are

part of the chain of events , and VS is generally used in them
in narrative prose. By way of contrast, circumstantial clau-
ses outside the stream of events (##5.1.3.1 - 5.1.3.3) do not
generally use VS. But some do, and then it has special func-
tions .
(i) It may serve as a flash-back, translated by an English
pluperfect. Example: verhel lqSh..., now Rachel had stol-
en... (Ge 31 3 ). Other examples: Ge 3 4 % 39, Le 5 1 .
(ii) It may have Stative meaning, translated by an English
present. Example: wa^ni zqanti, and I am old (Ge 1 8 1 3 , com-
pare zqen in 1 8 1 2 ) . Other examples: Ge 2 4 3 1 5 6 , 2 6 2 7 , 2 9 I 7 ,
4 8 1 0 , Le 5 2 3 ' " .
The use of hy in such clauses raises the question of whe-
ther it should be translated had been or was. So, in Ge 1 5 1 7
and it came to pass, when the sun had set, and when darkness
had fallen (hy = np?l in 1 5 1 2 ) , that... But in Ge 2 9 1 6 and
Rachel was pretty... Does hy in Ge 34 5 mean that Jacob's
sons WERE in the field when their father heard about Dinah, or
that they HAD BEEN in the field when she was raped? The latter
is more probable, because contemporaneous circumstance is ade-
quately covered by a verbless clause. Hence it is more likely
that Ge l 2 a means the earth had become (or had come to be)...
as a circumstance prior to the first fiat recorded in Ge l 3 ,
than that it means the earth was... as a circumstance accom-
panying the first fiat. This disposes of the argument that
Ge l 2 a cannot be circumstantial because it is not verbless.

Compare Ge 7 6 > 1 0 and E x 1 s t a n d Joseph was already in Egypt
(iii^ It can be used with 'present' (contemporaneous) mean
ing: w e h e m l^ y a d e T u k l . . . and (all the time they were talk
ing) they (Pr for p r e v e r b a l S ) don't know that... (Ge 4 2 2 3 ) .
Other examples: Ex 1 2 3 3 (note gam), 1 3 j e , 1 4 2 9 a , l? 1 2 *,
19 1 8 .
5.1.3.5. Circumstantial clause with imperfect verb

A general concomitant future state of affairs can be placed
alongside a piece of predictive discourse by means of a cir
cumstantial clause using VP.
Example: wehâbnlm tihyeyn '"al-s emt bene-yiSr1 el. . . ,

and the stones will match the n a m e s of the sons of Israel...
(Ex 2 8 2 1 ) .
Other examples: Ex^4 1 2 (simple consequence could use we"1 ehye

or wehyltl: the we- +S +V sequence is distinctly circumstan
tial), 4 1 5 ' (close of speech^ 5 e , 7 1 S (0 before V), De 251
(a general truth; sequential wehy would be less appropri-
ate) .
5.2. SENTENCE-LEVEL CIRCUMSTANTIAL CLAUSES
5.2.0. Sentence versus Paragraph
So far the circumstantial clauses have been described as

structurally marginal to paragraphs, related in some way to
paragraph-level WP (or WS) clauses, or to a paragraph as a
whole. Circumstantial clauses can also be related to clau-
ses or phrases which are not paragraph-level.
5.2.1. Clauses Circumstantial to Time Margins
A paragraph or episode often begins with a marginal time ref

erence, realized as a prepositional phrase, or infinitival
construction, or a clause. An additional circumstance may
be attached to this head.
(i) After a PpPh. Example: Ge 22 1 , which reads literally,
after these events, and God tested Abraham, and he said to
him... Also Ge 41 1 , Ex 12 2 .
(ii) After an IfPh. The classic example is Ge 2* 6 ,
where the initial time is followed by a list of accompany-
ing circumstances in the six clauses of Ge 2 5 " 6 . The first
event follows in Ge 2 7 . Ge l 1 " 2 has the same structure. Ge 1
is a circumstantial sentence comprised of three conjoined
circumstantial clauses, the whole circumstantial to the open
ing time (Ge l 1 ). The first event is reported in Ge l 3 . An-
other example: Ge 44 3 0 (two clauses).
(iii) After a clause. Example: The time and circumstances
of the flood are r e p o r t e d in Ge 6 1 . When mankind began to
become numerous over the surface of the land, and daughters had
been born to them, the sons of the gods saw... Cj :ki governs
both clauses instead of coordinating *k.I-hehel...w?ki yullSd...
Other examples: Ge 1 5 1 2 > 1 7 , 19"> 2 3 , 2 7 3 0 , 29 s . 37 2 , 42 1 6 (in
precative mood), 42 1 9 (precative) , 5021*, Ex 10 3 b (the morning
and the wind came at the same time).
5.2.2. Circumstance of a Circumstance
A circumstantial clause may be conjoined with another circum-

stantial clause. Another way of describing this is to say that
a conjunctive sentence (see Chapter 8) may be used circumstan-
tially if the lead clause is circumstantial.
Example: vesray ^ elet âbrm l^ yleds l w?lh siph misrit

'Smh hgr, (i) and Sara y, Abram's wife, had borne him no
children, (ii) and she had an Egyptian slave, (iii) and her
name was Hagar (Ge 16 1 ). Here the third clause is circumstan-
tial to the second and these two as a whole are circumstantial
to the first which, along with them, is circumstantial (episode-
initial) to the ensuing narrative.
Other examples: Ge l 2 (three conjunctive clauses circumstantial

to the preceding) 2l,t'"6 (three circumstantial sentences of two
clauses each, the whole circumstantial to creation), 1 8 l 0 b ,
19 2 3 (three coincident circumstances of the over-throw of the
cities--tAe sun rose. Lot reached Zoar, Yahweh rained fire. The
three-clause complex is episode-initial.), 2 9 1 6 , 31 2 5 (Compare
this coordination with the apposition used in Ge 131 2 .),6 345 ,
etc.
It is possible that the clause in Ge 2 9 has three objects
(three kinds of tree). It is more likely that Ge 29t> is a cir-
cumstantial clause, the subject being the coordination phrase
'es hahayyim...v?'es hadda'at tob wr' realized discontinu-
ously--with the tree of life and the tree of the knowledge of
good and evil tbothl in the middle of the garden. These two
specified trees are included in the kol- f es of 2 9 a , not addi-
tional, because Ge 3 says that the tree of knowledge was good
for food. Ge 2 1 0 is equivocal. It could be"a new circumstantial
clause inaugurating a new paragraph. But we are inclined to
think that Ge 2 10 " Yl * is to be coordinated with Ge 2 9 b as a sen-
tence circumstantial to 2 9 a . The trees and the rivers go togeth-
er to make up the garden. (See above #5.1.1.)
5.3. PSEUDOCIRCUMSTANTIAL SEQUENTIAL CLAUSES
A string of WP clauses in narrative prose stages events as oc-

curring in a time sequence one after the other. It is implied
that one is finished before the next begins, so it is possible
to speak of the verbs as 'perfective* in aspect. So the succes-
sive generations are described in Ge 5 and in parts of Ge 10
and 11 by a series of clauses beginning wayyled, and he en-
gendered .
Such a time sequence is normally broken by a circumstantial

clause, which represents an event as contemporaneous; at least
it does not place it in sequence. It is therefore surprising
and unaccountable that Ge 4 l e uses clauses of the kind v?-X
ylad êt-Y, even though a sequence of generations is being
traced, and simultaneity is out of the question. These pseudo-
circumstantial clauses are restricted to the genealogies, and
are commonly ascribed to the J source. Examples: Ge 4 1 > * 6 ,
10 8,13,1S,2V,26 > 2 2 2 3 , 253 . The genealogies also alternate
wayglji x with - hay (ll 12 11 *) with no evident difference
in discourse function.
Apart from these examples confined to genealogical texts,
this happens rarely. Ge 22 l b could be an example. Ge 44" is
another. Here the first event in the new episode (and not a
preliminary circumstance) is vgysep ^Smar... instead of the
expected wayy^mer ysep. This is preceded by an elaborate
time margin of four clauses. In Ge 38 2 5 the sequence hi1 m-
s e U wShi 1 aiSh êl-hmih is a strange alternative to *w?hl1
mge^t vattislah... This is the more surprising in view of
Ge 38's impeccable use of WP clauses. The Flood Story con-
tains a possible example in Ge 7 6 --wehammabbl hy instead
of *way?hl hammabbl as the first event, rather than the cir-
cumstance of the deluge. Compare the similar but more cele-
brated probl em of Ge l 2 with v?h^res hay?ta rather than
*wattehl h^re, if this is the first event of creation.
Ge 4 2 2 is chiastic, and authentically circumstantial. Note
gam-hi1 (compare Ge 4"). In Ge 4 2 5 " 2 6 , 1 0 2 1 2 5 the passive
is used in the circumstantial clause.
5.4. PSEUDOSEQUENTIAL CIRCUMSTANTIAL CLAUSES
The converse of the phenomenon described in #5.3 is the use of

paragraph-level WP (or WS) clauses which normally place events
in sequence, to represent events which can hardly have occurred
in sequence but which obviously are contemporaneous. This is
an artistic option for a writer, but it strains the system. It
seems to have been rarely used. Thus the gift of Zilpah to Leah
as a slave took place at the time of her marriage, and we might
have expected this event to be staged as circumstantial to the
marriage. Instead it is presented by means of a WP clause
(Ge 292lt) as if it took place after the marriage, or even dur-
ing the bridal night. The same for Bilhah and Rachel in
Ge 29 2 9 .
5.5. CIRCUMSTANTIAL CLAUSES AS ALTERNATIVES TO

NONCIRCUMSTANTIAL CONSTRUCTIONS
5.5.0. Introduction
In most languages there seems to be a close connection in the

system of complex sentences between relationships of time and
relationships of cause and effect. The same conjunctions and
prepositions are often used with analogical meanings in both

systems. The same in Hebrew; ki has several meanings, includ-
ing when and because. Circumstantial clauses typically func-
tion as time margins, so we- is often fittingly translated by
a time conjunction such as when or while. Whenever the time
connection between two clausal statements is paramount to a
logical connection a circumstantial clause can be recognized
as an alternative realization of a subordinate clause condi-
tional, concessive, causal, etc. A circumstantial clause can
also function as an alternative realization of a relative clause
nominalized by means of n * <
When dealing with a dead language, the categorization of a
particular construction in a given text as an alternative sur-
face realization of a deep-structure relationship optimally
realized by different formal means is a risky business. For
lack of a living informant, the modern scholar may fall back
on his own competence. This can suggest, but it cannot decide.
Optimum realizations can be distinguished from alternative re-
alizations by testing the constraints on the interchangeability
of the contrastive formal features, especially such demarcative
items as conjunctions. Such a test can be addressed to extant
texts, but it is not enough to show that the alternative con-
structions 'exist 1 ; if we assume that all surviving texts were
acceptable to the scribes at some stage of their transmission
and that there were limits to the degree of gibberish they were
prepared to copy, and refrain from correcting their work be-
cause it seems gibberish to us, it does not follow that all the
evidence should be accepted uncritically as equally grammatical.
The test of substitutability of conjunctions must be aided by
deeper analysis of the semantic content before we can say that
two constructions are 'the same.' Fortunately the abundant use
of parallel texts supplies us with plenty of material amenable
to such control. Even so, when dealing with literature which
is the precipitate of hundreds of years of oral and scribal
transmission, factors of dialect and historical change compli-
cate the picture with other parameters. In the dynamics of
language evolution, alternative realizations may move from the
margins of grammaticality to the centre, and oust the former
optimum realization from its prestige. So far as Hebrew is con-
cerned, this history has yet to be written.
5.5.1. Circumstantial Form for Deep Subordination
Jeremiah protested his disqualification from the prophetic of-

fice by saying, X cannot speak, for (ki = because) I am a youth
(Je I s ) . Even here, in sequal to negation, ki could be adver-
sative, but, on the contrary. Moses expressed a similar pro-
test differently. The cause is stated as a fact, the effect
as a question: hen âni f aral S5ptayim weêk yisma f 1 elay
par^, look, I (am) uncircumcised of lips, and (=so) how will
Pharoah understand me? (Ex 6 3 0 ) . The question is rhetorical,
and tantamount to a negative declaration. In Ex 6 1 2 , however,
the question comes first and the cause is realized in the form
of a circumstantial clause: vSêk yismeênl par '" wa^ni raral

leptyim, and (=but) how will Pharoah understand me and (=see-
ing that) I (am) uncircumcised of lips? Other examples of a
question that is virtually a negation because of some vitiating
factor supplied in the form of a circumstantial clause are Ge
15 2 , 18 12 >'3 >17 243 1 , 262 7 , Jdg 13 l e . After negation or pro-
hibition: Ge 24 , De 9 2 9 .
If the circumstances conditioning the question are hypothet-
ical, the equivalent conjunction in the optimum subordinate
clause would be H i if. Example: For how can I go up to my
father and (=if) the boy is not with us? (Ge 44 31 *; question
1S
plus circumstance, compare Jdg 16 ); But we cannot see the
man's face and (if) our little brother is not with us (Ge 4 4 2 6 ;
negation plus circumstance). See also #13.7.
A declarative statement of fact may be followed by a state-
ment of the actual cause in the form of a circumstantial clause.
Example: You are in mortal peril and (=because) she is a man's
wife (Ge 2 0 3 ) ; And you could not realize that they had gone in-
side them, and (=because) their appearance was just as bad as
at first (Ge 4 1 2 1 ) .
Another function of the conjunction kl is to subordinate a
concessive clause.7 A circumstantial clause can realize the
same relationship. Example: Behold, I have presumed to speak to
Yahweh, and (=even though) I am dirt and dust (Ge 1 8 2 7 ) ; And
Israel stretched out his right hand and put it on Ephraim's
head and (even though) he was the younger, and his left hand
upon Manasseh's head (he crossed his arms) although () Man-
11
asseh was the firstborn (Ge 48 *).
5.5.2. Circumstantial Form for a Relative Clause.
In #2.6.5 it was shown that a clause circumstantial in form

could be used to supply the name of a person as an alternative
realization of an apposition relationship whose optimum reali-
zation is a relative clause. The same thing can be done with
other material. In introducing a new character, it is common
to supply his address along with his name. Example: A man of
the hill country of Ephraim (Jdg 1 7 1 ) ; a man from Bethlehem
(Ru I 1 ) In Ge 3 6 3 2 f the birth (or capital?) cities of the
Edomite kings are introduced in this way: Bela ben-Beor (and)
the name of whose city was Dinhaba. Ge 44 20 has three con-
joined circumstantial clauses equivalent to relative clauses:
We have an aged father and a young 'old-age' boy and his
(whose) brother is dead and he (who) is the only surviving
child of his mother and his (whose) father loves him. Deep
relationships concealed in alternative surface realizations
have not always been recognized by translators.
NOTES
1
S. R . D r i v e r , A Treatise on the Use of the Tenses in He-
brew3 ( O x f o r d , 1 8 9 2 ) : A p p e n d i x I (pp. 1 9 5 - 2 1 1 ) . See also
. J. Meek, "The Coordinate Adverbial Clause in Hebrew", Jour-

nal of the American Oriental Society 19 ( 1 9 2 9 ) : p p . 1 5 6 - 1 5 9 .
2
A c c o r d i n g t o W . H . B e n n e t t (.Hebraica 5 C1888-18893 : p.203).
t h e n a r r a t i v e p o r t i o n s of t h e O l d T e s t a m e n t , w h i c h h e d o e s not
l i s t , u s e t h e v e r b s as f o l l o w s :
Clearly past Not clearly past

VS 7,>*20 1+2
WP 13,11*9 It
Clearly future Not clearly future

WS -3,362 272
VP 6 ,U35 253
3
On w?l^ clauses as antithetical see Chapter lU.
11
F. I. A n d e r s e n , The Hebrew Verbless Clause..., pp. 69-76.
5
Driver, Tenses..., 16.
6
When the constituent clauses of a conjunctive sentence are
distributive, apposition rather than coordination may serve
for circumstantial linkage according to Driver on Ge 12 e (Ten-
ses...: p. U9). We call this exposition., (see p. k9 above).
7
Th. C. Vriezen, "Einige Notizen zur Ubersetzung des Bund-
eswortes KI", von Vgarit nach Qumran, Beihefte zur Zeitschrift
fr die Alttestamentliche Wissenschaft 77 (1958) CEissfeldt
Festschrift]; pp. 266-273; James Muilenburg "The Linguistic
and Rhetorical Usages of the Particle in the Old Testament",
Hebrew Union College Annual XXXII (1961): pp. 135-160.
6
ADJUNCTIVE CLAUSES
6.0. STRUCTURE
A well-formed adjunctive clause has a distinctive-contrastive

internal structure which correlates extensively with the real-
ization of a deep-structure coordination that is more tangent-
ial to the main stream of discourse than circumstantial coor-
dination. There is always explicit coordination by means of
we-, and, rarely gam or wegam. A marked break in the flow of
discourse is achieved by fresh topicalization (casus pendens)1
whose opening sequence of we- +S resembles a salient pattern
in circumstantial clauses. Because of this structural simil-
arity, adjunctive clauses have not hitherto been recognized
as a category distinct from other kinds of coordination; they
have generally been classified as 'circumstantial1. But ad-
junctive clauses differ from circumstantial clauses, as des-
cribed in this monograph, by the obligatory use of an explic-
it resumptive pronoun. While the suspended topic is commonly
resumed by the grammatical subject, it may correspond to some
other clause-level tagmeme, such as object.
Casus pendens with subsequent pronominal resumption can
also be used to announce a fresh topic, especially at the
beginning of a speech. Examples: Ge 28 l3t> , 34 s , Le 7 3 3 , Jos
9 , 2Ki l*. There is no coordinating conjunction, so we do
not regard such clauses as adjunctive.
6.1. FUNCTION
The information supplied in an adjunctive clause is gener-

ally less germane to the main discourse than the inform-
ation typically supplied in a circumstantial clause. It re-
sembles the kind of material which in English would be in-
troduced by means of By the way,... It is tangential rather
than marginal, and because of this loose attachment it is
not used as an alternative realization of subordinate re-
lationships, the way circumstantial clauses can be (#5.5).
Adjunctive clauses can, however, function circumstantially
(#6.2).
ADJUNCTIVE CLAUSES 93
Example: vehm h^ âbl kenân, and (as for) Ham -- he is

the father of Canaan (Ge 9 1 8 ) .
This clause sits loosely in the passage, which would flow

on quite smoothly without it.
Other examples: Ge 17 1 *, 21 13 (The adjunctive clause a-
voids giving Ishmael the same rank as Isaac, in spite of
gam, which does link the two sons together [Chapter 12]),
261 , 28 2 2 b (makes tithing incidental), 42* (The story re-
turns to the main thread after this as ide.), 431 , 47 1
(This is an impressive example; it has nothing to do with
the rest of the story.), Ex 16 3 , 321 6 , 3 9 s . D e 2 2 3 (The
note on the fAvwim is subsidiary.), 14 2 7 (The rule about
the Levite is an afterthought.)
6.2. ADJUNCTIVE CLAUSES USED CIRCUMSTANTIALLY
An adjunctive clause deviates momentarily down a little

side-track, to make a remark about somebody who does not
figure anywhere else in the story, like Nahor's concubine
Reumah (Ge 2221*), or to say something about a character
that does not contribute to the plot, like the scrap of
demographical information in Ge 47 2 1 . The judgment as to
whether a piece of information is relevant or not is ob-
viously a subjective one, so that the line between an
adjunctive and a circumstantial relationship to the con-
text is not always easy to draw. Some clauses, adjunctive
in form, bring in information that does contribute to the
development of the plot, and may be described as alterna-
tive realizations of circumstantial clauses.
Thus Jd 17s is an adjunctive clause that marks a new
beginning (#5.1.1). The adjunctive clauses in Ge 21 **^,
1915
, and 31"3 have a climactic effect (#5.1.2). Ge 15 i b
is simply circumstantial.
6.3. OTHER FORMS

15
In De 2 vegan introduces an extra comment without having
its usual inclusive function (Chapter 12).
NOTES
1
In his treatment of casus pendens (Tenses..., Appendix
V), Driver found the explanation of this kind of construction
largely in aesthetic considerations. Doubtless the lyrical
effect provided an additional motivation for its u s e , but the
g r a m m a t i c a l factor of topicalization is primary.
7
SURPRISE CLAUSES
7.0. FORM
An unexpected turn of events can be marked in Hebrew

by adding to a clause the deictic-exclamatory hinne,
traditionally translated lo or beholdi Its quasi-imper-
ative attention-rousing function is supported in Hebrew,
not by etymology, but by limited interchangeability
with VI:re^ 5 , look', and other such hortatory verbs.
Its usual LXX translation is a verb with similar
usage. The morphology of hinne is intriguing. It can
be inflected with personal pronoun suffixes which real-
ize the subject of the clause, not the object of hinne
unless one insists on the imperative verbal meaning of
the latter. But clauses with hinne are declarative.
Hinne predicates present and local existence. Hinneni
means I'm here! rather than Look at met Furthermore,
the subject of hinne as predicator can be a free-form
pronoun or noun, and in either case a concordant subject
pronoun suffix is optional, not obligatory as with finite
verbs. In some of its occurrences hinne is complemented
by a predicative participle, adjective, adverb, or
prepositional phrase, even VS (VP once), and for this
reason is often called an 'adverb'. But there is no
agreed nomenclature among Hebraists, and many are con-
tent to call it vaguely a 'particle'.
7.1. PARTICIPANT PERSPECTIVE
The difference between a clause like Ge 18 e b -- veh 1

^med âlehem, and he is standing beside them -- and
Ge 24 3 0 -- vehinne ^med âl-haggemalllm^ and behold he
is standing beside the camels is that the latter is
seen through Laban's eyes. This component, that something
comes into the view of one of the participants, is promin-
ent in the commonest idiom in which a vehinne clause is
used--wayyar^ vehinne..., and he looked, and behold... It
is this feature of the unexpected that we describe as 'sur-
prise' in such clauses.
SURPRISE CLAUSES 95
Examples: Ge l " 6 1 2 . 8 1 3 . 1 8 2 , 19 2 8 , 2 2 1 3 , 24 6 3 , 2 6 e , 29 2 ,
31 2 , 33 l . 37 , 40 , 4 2 * \ Ex 2 e , 3 2 , 1 4 , 0 t , 32* , 34 3 0 ,
39 ^ 3, De 9 1 3 > ' 6 , etc.
In Ge 31 2 the h i s t o r i a n reports: wayyar 1 y a r q b "'et-pgne

lbn w S h i n n e ^ e n e n n îmmS kitml , as if Jacob h a d
made a d i s c o v e r y . In reporting to his w i v e s , J a c o b says:
r5ê ^nSkl êt-p5ne ^biken k i - ^ e n e n n ^ elay kitml sil55m
(Ge 3 1 s ) , w h i c h is more m a t t e r of fact.
A w ? h i n n e clause c a n be u s e d to report a surprise develop-
ment after other verbs b e s i d e s r^, especially verbs of mo-
tion.
Examples: Ge 8 1 1 . 2 4 1 5 , 3 0 , ^ 5 . 2 5 2 \ 3 7 1 5 , 2 9 , 4 2 3 5 , 4321,
47 1 , 4 8 1 1 , Ex 2 , 4 6 > 7 , 6 , etc.
7.2. D R E A M REPORTS
The i m p l i c a t i o n of visual e x p e r i e n c e comes to the fore in

the d e s c r i p t i o n of d r e a m s . A n element of the m a r v e l l o u s is
also p r e s e n t , of course. The n a r r a t i v e ingredient is not
well d e v e l o p e d in the dreams r e p o r t e d in Genesis. There is
usually only one event clause; all the other i n f o r m a t i o n is
supplied by means of w e h i n n e c l a u s e s , w h i c h may accumulate
into quite a string. O t h e r w i s e they occur one at a time.
So J a c o b ' s d r e a m begins w i t h three w S h i n n e c l a u s e s , w i t h
God's s p e e c h as the d r e a m event (Ge 2 8 1 2 " 1 5 ) . Each clause
has the structure w e h i n n e +S:N +P:PtPh. The d r e a m in Ge 3 1 1 0
begins in the same way. The dreams in the J o s e p h story are
p a r t i c u l a r l y impressive in this regard. The d r e a m of the
sheaves has three w S h i n n e c l a u s e s f o l l o w e d by the event
clause (Ge 3 7 6 " 7 ) . The speech even b e g i n s w i t h w e h i n n e . T h e
third clause is the only instance in the entire Bible of
hinne followed by a n imperfect verb. The d r e a m report in
Ge 37 9 consists of one hinne clause a n d one w ? h i n n e clause.
Compare the single w ? h i n n e clause in the v i s i o n of Ge IS1*.
The b u t l e r ' s d r e a m consists of one w ? h i n n e clause followed
by a string of four c i r c u m s t a n t i a l clauses w h i c h set the
scene. This is followed by three event clauses.
Then I picked the grapes.

And I squeezed them into Pharoah's cup,
And I put the cup on Pharoah's palm.
In contrast to this the b a k e r ' s d r e a m consists of c i r c u m s t a n c e s

only -- one w ? h i n n e clause plus two c i r c u m s t a n t i a l clauses.
There is no event. In Pharoah's first d r e a m (the cows) there are
three w ? h i n n e clauses (with p a r t i c i p l e s ) . The only event is the
skinny cows eating the fat ones (41*"*). The s e c o n d d r e a m (the
wheat) consists of two w g h i n n e +S:N +P:PtPh clauses plus the
event clause in w h i c h the s h r i v e l l e d w h e a t devours the p l u m p
wheat ( 4 1 5 - 7 ) . Pharoah's reports repeat the p a t t e r n s , except
that the first is e m b e l l i s h e d w i t h some comments of his own
(Ge 4 1 1 7 " 2 " ) . The abundant use of w S h i n n e clauses is thus a
feature of d r e a m reports in classical H e b r e w .
96 SURPRISE CLAUSES
7.3. O T H E R U S E S
T h e r e are other uses of v e h i n n e clauses in w h i c h the f e a t u r e

of a n u n e x p e c t e d v i s u a l e x p e r i e n c e is not p r o m i n e n t or m a y b e
q u i t e absent. A n e w episode o f t e n begins w i t h v a y e h l + M a r -
ginal T i m e + W P C 1 for the first event. In a few instances a
time m a r g i n is f o l l o w e d by a v e h i n n e c l a u s e , w h i c h looks like
the first event c l a u s e , a n d n o t c i r c u m s t a n t i a l . Examples:
Ge 1 5 ' 2 > 1 7 , 29 2 5 , 3 8 2 7 ' 2 9 .
A v e h i n n e c l a u s e c a n also p r e d i c t a n impending event (Ge 6 1 3 ,
c o m p a r e Ge 6 1 7 " 1 9 ) .
7.4. O T H E R FORMS
S u p p l e m e n t a r y i n f o r m a t i o n c a n be s u p p l i e d by a c l a u s e b e g i n -
ning gam hinne (Ge 3 2 3 1 ) or v e g a m hinne (Ge 38 21 *). T h e latter
seems to express a s t o n i s h m e n t at s o m e t h i n g q u i t e s e n s a t i o n a l .
f
But gam hinne abdek ya r qfc ^ a h a r e n in Ge 3 2 2 1 is m e r e l y
a v a r i a n t of v e h i n n e gam-hiP ^ a h a r e n (Ge 3 2 l s ) .
8
CONJUNCTIVE SENTENCES
8.0. THE FORM OF A CONJUNCTIVE SENTENCE
8.0.0. Optimum Realization
A conjunctive sentence comes nearest to the ideal coordination

construction defined in the grammars. At the same time, it is
the least specialized of all the coordination constructions. It
is unmarked for sequence, contrast, or antithesis and, while
not specially marked for simultaneity or similarity, it is com-
patible with these relationships and generally implies them. It
does not, however, highlight similarity or simultaniety. Fur-
thermore, as the most neutral coordination construction, a con-
junctive sentence can be used as an alternative but insipid re-
alization of most of the deep coordination relationships (#4.6;
see also p. 189).
In a typical conjunctive sentence two clauses are joined by
we- and, and each constituent clause has (or could have) the
same grammatical function as the conjunctive sentence as a
whole. The 'sameness' of the two conjoined clauses is func-
tional, not necessarily semantic, although it may be semantic
also. The lead clause and the conjunctive clause usually have
the same kind of predication and, if the predicators are verbs,
they usually have the same tense-aspect and mood and often the
same subject. Furthermore, the two conjoined clauses often have
other clause-level tagmemes besides subject and predicator in
common, and the common tagmemes are realized in the same se-
quence in both clauses. The result is a kind of grammatical
rhyme.
We do not wish to insist on this high level of formal con-
gruence as necessary for well-formedness in a conjunctive sen-
tence, for the majority of conjunctive sentences fall short of
this ideal. Similarity of external function, rather than simi-
larity of internal content or structure, unites them. But ob-
viously two clauses must be compatible in order to have simi-
larity of external function. Hence it is unlikely to meet a
change of tense-aspect, mood, or subject person in the clearer
instancesof conjunctive sentence.
98 CONJUNCTIVE SENTENCES
As already stated in #4.3.1, a conjunctive sentence typically

links together two similar actions performed by one and the
same participant, generally realized as subject of both clau-
ses. It implies that the actions are contemporaneous (occur-
ring in the same general time), if not actually simultaneous;
at least, it does not represent the actions as occurring in
sequence. Similar actions by two participants can also be con-
joined, but such a relationship is more likely to be handled
by some other form of coordination.
8.0.1. Double-Duty Items
A common subject of two successive clauses is seldom repeated

on sentence level. It does 'double-duty' in both clauses and
helps unite them. Other items may similarly appear in one or
other of the clauses while functioning equally in both. While
this diminishes the measure of formal similarity between the
two conjoined clauses, it actually enhances the grammatical
integrity of the resulting sentence, since neither clause is
grammatically complete without material in a neighbouring
clause. This feature of Hebrew composition can be exploited
to a high degree of sophistication, especially in poetry.
Example: nilben lebenlm

veniSrep liSrep
V:let us brick 0: bricks
and V:let us burn Itheml 0C:to-burnt (Ge ll3)
The first clause supplies the object, the second the object
complement.
Each conjoined clause can be grammatically complete if the
double-duty items are marginal modifiers.
Example: k t^mar 15bet yaâqob

we- tagged libne yirêl
Thus you will say to the house of Jacob
and you will report to the sons of Israel (Ex 193)
k belongs in both clauses, not just the first. Its function

can be described as modifying the conjunctive sentence as a
whole.
8.0.2. Multiple Coordination
Unlike antithetical, chiastic, contrastive, alternative, exclu-

sive, and inclusive sentences, all of which are strictly binary
in form, conjunctive sentences are recursive. There is no theo-
retical limit to the number of clauses that may be coordinated.
With the fondness of Hebrew for 'and' it is common to have A
and and C and D and.. Sometimes, however, for three, A, B,
and C is used; and, for four, A and B, C and D, especially if
both A and B, C and D are natural pairs. These patterns are
met in phrases and sentences.
CONJUNCTIVE SENTENCES 99
Examples: Three clauses, A, B, and C, Ge 6 ll, b-i6 (features of

the ark: stories, windows, and doors [#3.7.3]), 6 l i b (d lmen-
sions: length, breadth, and height), 4 9 3 1 , Ex 1 2 1 1 .
Four clauses, A and B, C and D, Ge 9 7 (both pairs hendiadys).
In Ge 2 7 a e " 2 9 eight clauses are grouped in four conjoined pairs.
8.1. DECLARATIVE CONJUNCTIVE SENTENCES
8.1.0. Introduction
These are used to make statements of two (or more) related and
contemporary facts which are not contrastive nor antithetical.
The facts may be past (using 'perfect' verbs--VS), future (using
'imperfect* verbs--VP), or timeless (using quasiverbals--QV--or
verbless [VL] predication) .
8.1.1. Reports of Accomplished Fact using 'Perfect' Verbs
Examples: fm sm 1 hq. mispat

wesm nissh,
There he set a rule and a custom for him
and there he tested him (Ex I S 2 5 )
lehem lo"1 ^ kal

mayim l^ ft,
bread he did not eat
and water he did not drink (Ex 3 4 2 8 )
Another Example: Ge 3 1 3 8 .
8.1.2. Conjoined Predictive Clauses
Clauses referring to concomitant or parallel future events by

means of VP are likely to be conjoined. Since VP clauses rarely
begin with the verb, conjunctive sentences with VP often resem-
ble contrastive sentences (Chapter 11) , and whether such a sen-
tence is antithetical or not is determined more by semantic con-
tent alone, more in this case than others.
f
Example al-gehnek telek
wS'pr t^kal
Upon your belly you will walk,
and dust you will eat (Ge 3 1 )
The clauses have a similar pattern of a clause-level tagmeme

before VP, although it is Location in the first, Object in the
second. The marginal time: kol-ySme hayyeyk, all the days of
your life, modifies this conjunctive'sentence as a whole.
Other Examples: Ge 3 l 7 b B - i a (again kl ySme hayyeyk [in hinge

position] goes with both clauses), 9 s (same verb repeated), 1 6 Y 2 ,
2 2 1 7 (the conjunctive sentence is subordinated with kl), 27"
34 (correlative actions), 34 2 1 (chiastic in 34 1 S ), 3 5 1 2 (con-

junctive sentence is nuclear after suspended topic), 411*0 (same
thing from two different points of view), 42 2 (=43 , conjunc-
tion of a negated antonym), 42 3lf , 46lt (same S:Pr twice), 4 8 1 9 ,
Ex 9 2 9 b , 21 2 '32 , 23 (The common subject does double duty and
ki governs the conjunctive sentence. The VP's express a time-
less truth.), 23 1 1 ' 1 2 , 34 1 3 (three clauses), 35" , De 2 2 , 6 1 3
(three clauses), 7 5 (four clauses), 9 3 (S:Pr), 10 i o (four clau-
ses , effectively precative).
Ge 27 3 7 has three conjoined clauses, each with an item (not
subject) before VS. The three conjunctive clauses in Ge 4 5 2 2
each has an indirect object before VS, resembling a contrastive
sentence (Chapter 11).
8.1.3. Conjoined Verbless ' Clauses
Concomitant present or timeless facts can be expressed in verb-

less clauses and conjoined.
Example: Yhwh hassaddiq

waâni weâmmi hres ''im,
The one in the right is Yahweh
and the ones in the wrong are I and my people (Ex 9 2 7 )
The sequence is PS in each clause. 1
Example: habbnt bentay

w?habbnlm bnay
w?ha?^n
wekl âser-âtt rê li-h 1
The daughters are my daughters
and the sons are my sons
and the flock is my flock
and all that you see--it's mine (Ge 311*3)
The final summary is an alternative realization in coordination

of a construction usually in apposition (#3.7.4), but the clause
has the form of an adjunctive clause (Chapter 6). Compare Ge 211*.
The sequence SP is sustained throughout, which is abnormal for
classifying clauses of this kind. The whole four-clause construc-
tion must be seen as having the structure +SSus:N... +P.. . +SRes:
Pr. In other words, h"1 acts as resumptive subject for all four
suspended subjects, thus securing the normal PS sequence. Ge 41 2 6
and 2 7 are conjunctive sentences with SRes:Pr explicit in BOTH
clauses. In Ge 3 6 the three verbless clauses which describe the
virtues of the tree of knowledge are similar in grammatical
structure, at least in PS sequence.
kl ôb h f es lSma"1 kl
w?kl taâv- h^ le f enayim
we- nehmd hês l?haskil
The third clause is strange because it lacks kl and because

S:N:h r es is used after the progression S:N:h f es, SiPrih 1 .
But it sould not be discarded for these reasons! Because, on
the other hand, both patterns for coordinating ki clauses are
found, either coordinating two clauses each with ki, or govern-
ing a single conjunctive sentence of two clauses with a common
kl (see Ge 6 5 , 11, 22 1 2 " 1 3 > 1 6 > 1 7 , 39 3 ). In the light of Ge 1 8 2 0
and other passages, however, it is possible that here ki is not
a conjunction, but the elative--she saw how utterly good the
tree was, etc. In 2 s the res is both nehmd and tob, and the
idea that it was pleasant to look at is also found. Further-
more, the idea of wisdom is important in the story.
We have dwelt on this problem at length in order to make the
point that conjunctive sentences are the most general kind of
coordination and that clauses of differing patterns may be in-
volved.
Other examples: Ge 423 8 , Ex 9 3 1 , 1 6 3 1 , De 1 0 2 1 .
But in Ge 2 7 2 2 a similar sentence, conjunctive in form, is an-

tithetical in meaning.
.1.4. Reports of Present Facts using Quasiverbal Clauses
Example: ysep ênenn

w??im^n ênenn,
Joseph is no wore
and Simon is no more (Ge 4 2 3 6 )
8.1.5. Circumstantial Conjunctive Sentences
The function of such sentences in discourse has already been

discussed in #5.2.1. Since such successive circumstantial clau-
ses have the same function in discourse, they constitute a con-
junctive sentence. It remains to make a few remarks about their
internal structure.
The two conjoined circumstantial clauses in Ge 2 5 are closely
parallel in grammatical structure. The three circumstantial
clauses conjoined in Ge l 2 have different subjects and differ-
ent kinds of predication--VS, VL, and Pt. But each has sequence
SP.
Other examples of circumstantial conjunctive sentences are
Ge 7 6 , 9 2 3 , 17 2 " 2 5 , 241 (in the parallel Ge 243 5 the slave art-
fully emphasizes the fact that Abraham was rich rather than old).
8.1.6. Conjoined Clauses in Poetry
Apart from the distributive coordination of identical clauses

(#8.2) the closest similarity in conjunctive clauses is met in
poetry and poetic discourse. A bicolon often consists of two
similar lines (one clause each) joined by AND, not in apposi-
tion as in #3.3.
Example: ze-saml leClm

veze zikri led3r dr
This is my name for ever
and this is my title for eternity (Ex 31S)
Example^za.faqat
^qat se
sedSm wa^mr kl-ra'b'b
wehatt ^ tm
iâtâ^tm ki kbed me^d
The outcry of Sodom and Gomorra [is] very great
and their crime [is] very grave indeed
(Ge 182
Such parallel clauses may be noncontiguous.
Example: |anl yhwh..

seml yhwh..
I (am) Yahweh
and my name (is) Yahweh (Ex 6 2 )
Recognition of this parallelism solves an old problem. Misled

by the apparent Neg l1 and the sequence ">51 sadday, interpre-
ters translated and by my name (of) Yahweh I did not make my-
self known to them, and made inferences about the
history of
the word Yahweh, about sources, and other things.
All forced.
The Ariadne thread of discourse grammar can lead us through
the rhetorical maze. This tradition is rich with names of God
Yahweh, its rare variant Êhyeh, El Shadday, the God of our
ancestors, etc. The emphasis is on what God has done, not a
historical note about what he did not do. The parallelism with
1
Snl yhwh shows that semi yhwh is a conjunctive clause of iden-
tical structure, and not an apposition phrase--my name Yahweh.
In the following speech (6 6 ~ e ) the repeated "'nl yhwh makes an
inclusio.
There is similar parallelism between the interlacing clau-
ses .
I showed myself to Abraham, Israel and Jacob
I made myself known to them.
This compels recognition of l"1 as assertative. There is no hint

in Exodus that Yahweh was a new name revealed first to Moses.
On the contrary, the success of his mission depended on the use
of the familiar name for validation by the Israelites. Moses
interrogated the Revealer precisely to convince himself that
it really was the god of the ancestors who had called him. Pha-
roah, of course, had never heard of Yahweh (Ex 5 2 ), but the se-
cret name was certainly known to the Israelites and to God him-
self. It should be added that be- is essentiae, not instrumental,
and there is no warrant for supplying it to the following semi.3
8.1.7. Dissimilar Clauses Conjoined
So far we have emphasized the similarity of conjoined clauses,

especially in the matter of tense, and even in the kind of
1
predication. This is not always the case. In Ex 10 * a past
fact and a future prediction are joined.
lepnyw l^ hy ken ârbe kmh

viPahryv lo1 yihye-ken,
Before them there had not been so locusts like them
and after them there will not be so.
In Ge 241,1 the two conjoined clauses both have the same pro-
noun before the verb, but the tenses are different.
8.1.8. Successive Events in Conjoined Clauses
Events reported in a conjunctive sentence are usually contem-

poraneous. Some of the pseudocircumstantial sequence clauses
discussed in #5.3 come in coordinated chains, giving a clus-
ter of clauses in which a succession of events is represented
as concomitant. Thus in Ge l l 2 7 the use of vehrn hlld êt-
lot represents the birth of Haran and the birth of his son Lot
as'concomitant, telescoping two generations into one moment of
narrative. In Ge 14 11 " 5 a series of dates is given in this way.
12 sn ^bed...
we 13 sn mrd
be-14 sn b^...
The whole is in epic parallelism with ^s milhm in Ge 14 2

(p. 41).
Two dates are similarly conjoined in Ge 8 1 3 .
be-601 sn brlsn... hreb hammayim...

bahdes hassen!... ybes h^res
In the 601st year... the waters dried up
and in the second month...the earth dried up.
Some narrative WP clauses intervene. Each conjoined clause has

the same grammatical structure. The time sequence is indicated
semantically by the preverbal time references, not by the kind
of verb. The series of cohortative (or imperfect) verbs in
De 9 1 " record sequential acts.
8.2. DISTRIBUTIVE COORDINATION
Distributive constructions in Hebrew generally involve repeti-

tion. Identical constructions can be coordinated, instead of
being placed in apposition like a list.
Example: mizze êhd

mizze êhd
one on each side (Ex 1712)
* CONJUNCTIVE SENTENCES
Example: hmi5?im ll^t ta^Se bayeri f hêht wehmi^Tm

ll't ta^Se biqse hayerl^... (Ex 26 s , fulfillment des-
cribed identically in 36 1 2 ).
Example: yd5 bakkl

wSyad kl b,
His hand against everybody
and everybody's hand against him. (Ge 1612)
Here the chiasmus of - and kl overlays the grammatical con-

gruence of the two clauses. The conjoining of two coordinate
subjects list-wise is also accompanied by chiasmus in Ge 23 1 1 .
hassde ntatti lk
wehanme 'r ^ aser b5 lek netattlh
the field I have sold to you
and the cave which is in it to you I have sold it.
Note the chiasmus of verb and indirect object.
8.3. CONJOINED PRECATIVE CLAUSES
8.3.0. The Variety of Combinations
The coordination of precative clauses requires special treat-

ment for two reasons. First, a variety of clause types may be
strung together in precative discourse, which may consequently
become so heterogeneous in character that it might not be ap-
propriate to talk about sentences at all. Secondly, a speaker
has a variety of options in generating a precative text.
A piece of discourse is marked as precative by the use of one
or other of the distinctively precative verb forms--IMPERATIVE
(VI), exclusively second person, JUSSIVE(VJ), mainly third per-
son, and COHORTATIVE (VC), chiefly first person. If the first
clause in a piece of discourse is precative, it may be contin-
ued in three ways. (i) by coordinating another clause that is
formally precative; (ii) by coordinating a clause which uses
VP and so is formally future indicative; (iii) by means of a
WS clause, future sequential.
+{VI,VJ,VC}
{VI,VJ,VC}<^- vS- +VP
The top combination is homogeneously precative without equivo-

cation, and the simple coordination of similar clauses implies
contemporaneous actions. The bottom combination can also be
homogeneously precative, especially if there is no change of
person. (See the discussion already given in #4.5). The ac-
tions are staged in sequence; it is implied that a series of
commands are to be executed one after the other. The middle

construction is neutral in both respects, but the ambiguity
can be resolved in several ways. The use of a VP clause cir-
cumstantially (#5.1.3.5) or chiastically (Chapter 9) secures
simultaneity; the use of temporal adverbs or other semantic
signals can bring out succession in time.
An illustrat ion is afforded by the fiats of Ge 1. The atmos-
phere is created by means of two conjoined VJ clauses in syn-
onymous parallelism. The events are identical.
yehi raqr1" b?tk hammyim

wlhi mabdil ben-mayim lmyim,
let-be an expanse in the middle of the waters
and let-be a division between water and water (Ge Is)
The separation of water and land is the result of two conjoined

VJ clauses with different subjects.
yiqqw hammayim... "'el-mqSm êhd.

weterê hayyabbs,
let be gathered the w a t e r s to one place
and let appear the dry land (Ge l9)
The events are concomitant. The relationships between the two

events could have been brought about in other ways.
As contemporaneous: yiqqw hammayim... *wehayyabbs terê,

and Cat the same time3 the dry land will appear.
As sequential: yiqqw hammayim... *w?nir^t hayyabbs, and

Lsubsequentlyl the dry land will appear.
As cause and effect, the effect forecast as the purpose of the

first action: yiqqw hammayim... *lemaân terê hayyabbs,
so that the dry land may appear.
If it could be agreed that all four of these constructions have

the same deep structure (and this is a matter for continued
heated debate in linguistics, especially in the transformation-
al-generative tradition), then the choice of one or other of
them is purely a matter of staging.
The creatures of air and water are made on the fifth day by
means of two commands coordinated in chiasmus (Chapter 9).
yiSres hammayim ieres . . .

we^p yS^pep âl-h^res. . .
let-swarm CinD the water swarmers...
and flyers let-fly over the earth (Ge l 2 0 )
While y^pep is formally ambiguous (it could be VP) its pre-

cative role is proved by yireb in Ge l 2 2 . Note that the require
ments of chiasmus override the rule that a precative verb must
be clause-initial. In Gen l 2 0 the events are distinct but con-
comitant. There is no connection of cause and result between
them. They could have been placed in temporal sequence, but

this would have increased the tally of creative acts. They are
staged, not only as coincident in time but, by chiasmus, as
two sides of one event.
The heavenly bodies are created on the fourth day by means
of an utterance consisting of a VJ clause followed by two WS
clauses.
yehi m?^ rt . . .
wehy le^tt...
wShy lim"1 rt...
let-be lamps...
and-they-will-be (for) signs...
and-they-wi11-be (for) lamps (Ge 11""15)
It is likely that the precative mood carries over into the WS

clauses--and-let-them-be signs. It is also possible that the
subsequent clauses are predictive and merely matter of fact.
Again it is possible that the consequence of an action is its
purpose--Jet there be lamps...so that they may be signs...
In any case, the author has committed himself to staging
these two aspects in a time sequence. If he had imitated l 6 ,
*w?yihy ^tt, this would make signs simply a synomym of
lamps. If he had imitated l 2 0 , this would have implied the
simultaneous creation of two quite distinct things--lamps and
signs .
It is clearly inadequate to talk about these alternative
constructions as mere stylistic variants. The options avail-
able in precative discourse represent a delicate apparatus the
capability of which needs to be kept in mind when interpreting
sentences. Ge l 2 0 is committed to simultaneity; Ge i l l , ~ l s is
committed to sequence; Ge 1 6 and l 9 are neutral in both re-
spects. Exceptions, of course, are sometimes met. VP contin-
uation may be used atypically for an action which obviously
must follow the one just mentioned, and WS continuation can be
used atypically for an action which is obviously not subsequent
to the one just mentioned. Semantics settles such cases and
grammar is neutralized. Then it is more appropriate to talk
about 'style. '
These constructions are complicated by the fact that many
verb forms are ambiguous as to mood, yiqtl may be VJ or VP,
nibne may be VC or VP.
Since a precative verb usually comes first in its clause, it
is assumed in the present study that an ambiguous prefixed
verb is VJ or VC if it is clause-initial, but VP if it is not
clause-initial. A precative conjoined sentence usually has
two verb-initial clauses joined by we- and. In its present
form, the verbs in a precative conjoined sentence are both
of the same kind, for example, VI... weVI..., with common
subj ect.
But even when this difficulty can be surmounted, and pre-
fixed verbs can be disambiguated as VP or VJ or VC, an
additional ambiguity remains when VP or WS clauses are used

to continue a speech which began in the precative mood. The
precative mood, established at the onset by the use of un-
ambiguous VI, VJ, or VC in the opening clauses, may be sus-
tained in what follows, even when the speaker switches to
VP or WS clauses. But then the change to these other verb
forms might mark a change to indicative mood. If the speech
is not marked as precative in its opening stages by the use
of one or other precative verb or clause, then there is no
problem. Discourse consisting solely of VP or WS clauses is
simply predictive, although the remote possibility is al-
ways present that such discourse is intended to be prec-
ative. But when VP or WS clauses follow precative clauses
in continuous discourse, they could be either commands or
predictions. The same time relationships of coincidence or
sequence obtain whether the mood changes or not; a prediction
may be staged as concomitant or as consequent to the action
commanded in the lead precative clause.
We do not believe that Hebrew changes horses in midstream
to the extent of switching from precative to indicative mood
in the transition from clause to clause within a single sen-
tence, although in the nature of the case this is hardly
capable of proof. Such a change of mood, even if it occurs
after a single clause, we relegate to a much higher level of
the hierarchy, and describe the utterance as a heterogeneous
concatenation of discourses of diverse types. A systematic
investigation of the formal signals of inter-discourse trans-
ition on that level is greatly needed, but cannot be made
part of the present study. Meanwhile we must accept the fact
that there is an area of formal similarity between the two
kinds of transition.
There is an additional complication. For reasons which have
never been explained, some verbs which do differentiate VJ
(short) from VP (normal) still use the VP form sporadically
with evident jussive meaning. Thus tirade in Ge I s , which we
considered precative (let it appeari) because it is clause-
initial and because of the parallelism, is actually VP in
form (it will appear). Similarly some verbs which can dif-
ferentiate VC (long) from VP (normal) still use VP preca-
tively on occasion. How very near precative and indicative
utterances are to each other is illustrated by the free
fluctuation of âl +VJ and l 1 +VP in the realization of
peremptory prohibitions, with no apparent difference in
intention or meaning.'1
8.3.1. Conjoined Verbless Precative Clauses
Concomitant blessings or curses may be coordinated in a

conjoined sentence.
Example: brk âbrm...

ubrk êl êlyn...
Blessed (be) Abram...
and blessed (be) El Elyon... (Ge 1 4 1 9 " 2 0 )
Other examples: Ge 2 7 2 9 ^ (The different sequence suggests a

statement of fact [Ind], but the entire speech is precative),
52
31 (assuming it means, Let this cairn be a witness...).
8.3.2. Conjoined Imperative Clauses
See #3.1.0.1 for the use of apposition as an alternative re-

alization of sentences of this kind.
It is quite common to coordinate a string of commands is-
sued by means of VI clauses.
Two clauses example: hithallek 15pnay wehye tmlm, walk be-

fore me and be perfect (Ge 17 1 ). Compare Ex 2 4 1 2 . Essentially
the same construction occurs in Ge 1 2 1 " 2 : lek-lek...wehye
fcerk, you go...and be a blessing. Here, however, the sen-
tence is discontinuous, due to the insertion of a three-clause
promise. This accounts for the much discussed VI in the middle
of a string of VP clauses. The MT is by all means to be re-
tained. The same trick is played in Ge 2 0 7 , where the conjunc-
tive sentence: hseb ^ estet-h"1 is .. .wehye , return the man's wife
...and live', has two clauses inserted in it. There is a simi-
lar discontinuous coordination of two VI clauses in Ex 3 1 "--
lek...wehge^..., go...and bring out... This is confirmed by
the response: ml ^nkl êlek...weki ^^..., who am I that I
should go...and that I should bring out...? (Ex 3 1 1 ) .
Other examples: Ge 13 llf , 15 s , 16 (actually in succession in

time), 24 3 (two parts of one action), 24 1 2 , 2 7 ? a > 2 6 . 3 1 1 2 ' 1 3 ,
3711* (definitely two stages). 42 2 (two stages), 43 1 * (synchron-
ic acts - -see below). 4 4 1 . 4 5 % 50 s . Ex 2 79>, 811216.
9 1 ' 1 3 , 1 0 1 1 ' 1 7 , 12 , 1 4 i 2 ' 1 6 ; 16 , 1 7 5 3 ' 1 * , 23 2 1 , 3 2 2 ' 1 * (a
poetic bicolon), 3 4 1 " 2 (the lead clause is developed as a para-
graph) , De 3 2 7 , 10 1 .
Three VI clauses: Ge l 2 2 (followed by VJ in chiasmus), 9 1 (sim-

ilar, with following VP clause in chiasmus), 19 2 (followed by
WS), 22 2 (all in one complex, although time succession is in-
volved), 34 1 0 , 35 1 2 , 38* (synonyms), 4 3 1 6 , 4 5 1 7 " 1 8
Four VI clauses: Ge 4 2 3 3 (The first two have objects before

the verb, to make a contrast sentence.).
Five clauses: Ge l 2 8 , 27 3 (a definite succession of acts, where

you would expect WS clauses to be used).
The effect of conjoining VI clauses can be shown by a more de-

tailed analysis of Jacob's speech in Ge 43 1 1 " 1 ' 1 . In spite of
its brevity and apparent simplicity, it has an elaborate hier-
archical structure, and several different inter-clause con-
structions are used in a variety of ways. The speech is in
three parts: (i) A conditional command: if that's the wag it
is, then do this...; (ii) a prayer for the mercy of God on
the enterprise ( l l f a ); (iii) an act of resignation ( l l f b ). The
command consists of a general introduction, followed in apposi-

tion by detailed exposition (#3.7.2) in seven clauses. The de-
tailed instructions are in two parts: (i) qeh... (Clj); (ii)
weqm ?ub 'el-hÎs (CI6 7 ) , take...and get up return to the
man. We have recognized tliese two parts as constituents of a
conjunctive sentence, but each is actually not a single clause
but a complex of clauses. The second is an apposition sentence
of the kind discussed in #3.1.0.1. The first is in three parts
They are to get three things (i) some delicacies as a present,
(ii) money, (iii) Benjamin. Each of these three commands is
issued by using the same VI:q?h take.
(i) ^ q.5hu + Oi CI
(ii) we- + 0 2 + qSh Cl 3
(iii) we- + 0 3 + qh Cl 5
(i) continues as a well-formed conjoined clause ( C I 2 ) wffhrid

..., the sentence being integrated just like Ge ll 3 by having
the common object with the first verb and the common object
complement with the second. In (ii) and (iii) we have most un-
usual constructions, clauses which do NOT begin with VI. The
explanation is to be found in the use of chiasmus between (i)
and (ii) to link together as two sides of a single action the
taking of the present and the money. (ii) and (iii), however,
being identical in grammatical structure, are conjoined. So
both together make a conjunctive sentence. Hence Ge 43 1 1 aB-i3a
is a construction in which a VI...w?VI conjunctive sentence is
followed chiastically by a 0 + VI... w?-0 +VI conjunctive Se.
There is an additional clause ( C I O ( 1 2 i > ) built into (iii).
This has the structure we- + 0 + tslb, so this too is chias-
tic, at least so far as the objects are concerned. But the use
of VP rather than VI makes it a concomitant act, even though
semantic content shows it to be a subsequent action. So far
as time succession is concerned, the seven actions come in
three stages.
Stage 1 Stage 2 Stage 3
Clj Get delicacies C I 7 Return to the man

Cl 3 Get double money Cl 6 Arise C l 2 Take down the present
CI 5 Get Benjamine CI Return the original money
But the order in which the seven commands are given (shown by
the numbers above) does not correspond to the order in which
they are to be performed. The chiasm between Cli + CI2 and CI3
+ Cls integrates the preparations. The conjoining of CI 3 and
CI., brings out the feature that the money is in two parts. Fi-
nally a parenthetical comment (perhaps it was a mistake [12bB])
is made in apposition on the preceding CI*.
I SPEECH 1
I CONDITIONAL COMMAND K<Co>' I PRAYER

8.3.3. Conjoined Jussive Clauses
Example: _yehl rq.1 ^ betk hammayim

vihi mabdil ben-mayim lmyim,
Let-be expanse in-midst the-water
and-let-be division between water to water (Ge l6).
Each clause has identical structure, +P:VJ +S:N +L:PpPh. The

verbs are identical. The two nouns are synonyms, which suggests
that the second is functioning as a subject noun, not as a pred-
icate participle, which would make the second verb periphrastic
and its subject anaphoric to the first noun. Let there be an ex-
panse...and let [it] be dividing... No!5 The two clauses are
grammatically congruent. Not all conjunctive sentences are as
well-formed as this.
Other examples: Ge l 9 (As pointed out above, the two events are
concomitant, but the second could have been represented as sub-
sequent to or as the consequence of the first.), 9 2 7 , 27 2 9 (syn-
onymous poetic parallelism), 27 3 1 (Since âbl is surrogate for
'you', these VJ's are virtually VI.)(sequence is involved), 28 3
(The WS clause in the midst of these four VJ clauses is a con-
tinuation of VI in 28 2 [#8.3.2].), 4 1 3 3 (two such sentences in
apposition; note that the lead verbs are formally VP), Ex 5 2 1 ,
8.3.4. Conjoined Cohortative Clauses
A series of concomitant resolutions by the same speaker may be

expressed in the form of conjoined VC clauses.
Example: nis^ wenelek wenelek lenegdek, let us break camp

and let us set out, and let me go with you (Ge 3312).
As already mentioned, we assume that all clauses that begin with

a first person 'prefixed 1 verb are cohortative, even if the verb
lacks the cohortative Sx:-, whether because the morphology
does not provide for it (stem ends in a vowel and VP and VC are
homonyms), whether because the position is preempted by a pro-
noun Sx (even though this construction can always be avoided by
using nota accusativi or some other preposition for alternate
realization of a pronoun object), whether because VP is used
even when a contrastive VC is available. Thus there are three
cohortative conjunctive sentences in Ge ll 3 >'*11. Each is a
pair of conjoined VC clauses and each is preceded by hb. The
verbs are nilben, nisrep, nibne, naâse, nered and nbel,
and we call them all VC for reasons of syntax, even though nibne
and na f ase are formally ambiguous (they can be VP).
Other examples: Ge 12 2 , 17 2 , 1 8 2 1 , 1 9 3 2 (serial acts), 22 5 (ser-

ial acts), 2 4 5 7 (serial), 26 3 (two contemporaneous acts; it then
switches to WS clauses for subsequent acts), 27 7 (serial), 35 3 ,
3 7 2 0 (serial), 43" (serial), 4 3 , 4 6 3 1 , 50 s , Ex 3 3 1 8 , 4 i J , 5 3 ,
De 3 2 * .
8.3.5. Mixed Linkages of Precative Clauses
8.3.5.0. The combinations
VL, VI, VJ, and VC clauses which express various combinations

of precative curses, blessings, commands, wishes, resolutions,
etc. can be conjoined in order to represent the combination
as concomitant. We have not found examples of all imaginable
combinations.
8.3.5.1 Verbless plus jussive
Example: brk yhwh êlhe sem

wihi kena'an êbed lm ,
Blessed be Yahweh, Shem's god,
and let Canaan be a slave for him (Ge 9 2 6 )
8.3.5.2. Cohortative plus imperative
Example: ^Ssi^-nn 1 êthen "'alekem

wa^s lhen ka-ftb beênekem,
let me bring them out to you
and [you] do to them what you like (Ge 1 9 7 )
Other examples: Ge 1931* (followed again by VC).
8.3.5-3. Cohortative plus jussive
Example: na^se ^dm...

veyird...
let us make man...
and let them dominate... (Ge l 2 6 )
Another example: Ge 1 9 2 0 .
8.3.5^ Imperative plus cohortative
Example: hîêm êlen

wenede^ ^tm,
bring them out to us,
and let us know them (Ge 1 9 s )
The clauses in this combination are usually related in a tem-

poral and causal sequence. If you bring them out to us, then
we will know them. Or Bring them out to us so that we may
know them.
Other examples: Ge 183 * 5 , 19 3 ", 2 3 * > 1 3 , 24 1 " 5 6 , 27 7 (Se +

Se), 27 2 5 , 29 2 1 , 3 2 1 0 , 4 4 2 1 , Ex 1 4 1 2 , 1 7 2 , 2 0 1 6 , 24 1 2 , De 91"
(a series of WS clauses might have been expected).
8.3.5.5 Imperative plus jussive
Example: hyi leâlpe rebb

weyiras zarêk ^ et saâr sne^yw,
be thousands of myriads,
and let your seed possess his enemies' gate (Ge 2 4 6 0 )
Other examples: Ge 3821, Ex 7 2 6 . 811 (deep structure is cause

and effect), 8 1 6 , 9 3 2 2 , 10 3 7 > 1 2 ' 1 7 ' 2 1 , 14 2 (content of speech),
14 1 5 (speech), 1 4 2 6 , 1 9 1 0 , 25 2 .
8.3.5.6. Jussive plus cohortative
Example: tehi n^ ^l benten benen benek

wenikret berit îmmk
Let there be an oath between us (between us and you),
and let us cut a covenant with you (Ge 2 6 2 8 )
Other examples: Ge 27"2 (when...then...), 30 3 (parallel).
8.4. NEGATION IN CONJUNCTIVE SENTENCES
It is usual for negation to be repeated in each conjoined clause,

rather than negating a conjunctive sentence as a whole.
Example: l1 t ^ e l mimmenn
wel^ tigge^ bo,
You will not eat any of it
and you will not touch it (Ge 3 3 )
Example: ^l^-^slp leqallel f 5d...

wel^-^Ssip f d lehakkt...,
I will not curse again...
and I will not again smite... (Ge 8 ')
Note the chiasmus. Ge 9 1 1 is quite similar. Also De 4 2 , 7 2 b ~ 3 .

A single negation, however, can do double duty by modifying a
conjunctive sentence as a whole.
Example: ^ en ze ki ^ im bet êlhlm

we- ze safar hassmayim,
This is nothing but the house of God
and this is [nothing but] the gate of heaven (Ge 2 8 1 6 )
Recognition of this can sometimes make quite a difference to
interpretation.
8.5. CONJOINED PROHIBITIONS
It is usual for the prohibition particle to be repeated in each

clause, rather than making it modify a conjunctive sentence as
a whole.
llU CONJUNCTIVE SENTENCES
Example: âi-tefseh
we^.al-yi&ar b e f e n e k e m ...,
Don't be distressed
and don't be angry with yourselves... (Ge 45s)
Other examples: Ge 19 1 7 , 22 1 2 , De 2 9 ' 1 9 .
8.6. COORDINATION OF QUESTIONS
It is usual for each coordinated clause to have its inter-

rogative function separately marked, even if it means repeat-
ing the same interrogator.
Example: lmm hr lk
w e l m m npel p n e y k ,
Why are you angry?
and why has your face fallen? (Ge 4 6 )
The clauses are parallel.
Another example: Ge 16 8 .
When ha- is the interrogator in the lead clause, it is common

for 1 1m to be its parallel in the conjoined clause. This might
account for the apparently extraneous vePim-sr sandwiched
between two ha- questions in Ge 17 1 7 .
It is also possible to join two clauses together and then to
question the conjunctive sentence as a whole.
Example: hal^ ze âser yiste ^dnl b

wehtP nahes yenaês b,
Isn't this what my master drinks out of?
and (isn't) it (what) he divines with? (Ge 4 4 5 )
This might help to unravel the puzzle in Ge 4 7 .

hl^ îm-te'fclb seêt
w e U m l^ teÎti lappetafc hatta^t rSbe?
Isn't there acceptance if you do well,
and doesn't sin crouch at the door if you don't do well?
5
Another example: Ge 20 .
Recognition of the sustained long-range influence of an open-

ing interrogator can make a difference to the interpretation
of a passage. So in Ge 31 2 6 Laban asks eight questions using
only two explicit interrogators. There are three interroga-
tive paragraphs in apposition.
Why did you do it?

and [why] did you rob my mind?
and [why] did you drag away my daughters like prisoners of
the sword?
<A> Why did you sneak away?

and [why] did you deceive me?
and [why] didn't you tell me, so that I could send you
off with joy and song, with drum and harp?
and [why] didn't you let me kiss my grandsons and grand-
daughters?
Now [why] did you act so stupidly?
8.7. CONJOINING OF SURPRISE CLAUSES
As a kind of exclamation, hinne, Beholdl is clause-initial

and, when it is not primarily the predicator in a declara-
tive clause, is grammatically attached to the following
text in only a loose way. Hinne can also be used to carry
a resumptive subject after a suspended topic in an adjunc-
tive clause. But if the ensuing text consists of two con-
joined clauses, that is, a conjunctive sentence, the deictic
function of the initial hinne may embrace this whole con-
struction, especially when the conjoined clauses are closely
parallel in grammatical structure and so secure a well in-
tegrated conjoined sentence.
Example: hinne +S:Pr: 1 nkl +P:PtPh:nissb...

Cj : - +S:NPh:hen5t ""anse h f Ir
+P : PtPhiyss"1 t. . .
Behold I am standing...
and the women of the city are coming out...
(Ge 24 1 3 )
Similarly hen in Ge 27 3 7 carries three clauses.
Without the initial conjunction, such clauses are more
declarative than surprise. It is possible, however, that
successive clauses of this kind are conjoined with hinne
repeated in each. A vehinne clause following a hinne clause
is then formally like a surprise clause, but may be simply
declarative, as in dream reports (#7.2). Compare Ge IS 3 .
When the initial hinne is used only once, as in Ge 24 1 3
above, it further integrates the conjoined clauses. So
Abraham's slave draws attention to the situation at the well
as a single scene. If, however, hinne is repeated with each
clause, those clauses, although still conjoined, are some-
what more independent of each other. The effect is to paint
two pictures, by drawing attention to two items in succession.
Example: ^hinne haggal hazze

vehinne hammasseb...
Behold this heap
and behold this stone... (Ge 3 1 5 1 )
8.8. COORDINATION OF SUBORDINATE CLAUSES
When there are two clauses, two constructions are possible.

(i) Each clause may be marked for subordination and then
conjoined.
Example: kl âh.! ^blh- h"1

wekl ben-ribq h-1
That he is her father's brother
and that he is Rebecca's son CGe 2 9 1 2 )
Other examples: Ge 91* O a k is repeated), 31 5 0 (the H m

of disavowal in the oath formula is repeated for each
clause in the treaty), 31 s 2 (like the preceding), 33 1 1 .
In Ex 22 2 2 two conditional clauses are conjoined by kl
as the coordinating conjunction.
Um-ânne teânne ^t
ki yiâq êlay
If you degrade him
and if he protests to me.
(ii) The two clauses may be conjoined, and then one sub-
ordinating conjunction governs the conjunctive sentence
as a whole. Examples: Ge 6 5 , 22 1 7 , 286 (a rare coordination
of VS with past tense meaning).
The same patterns can be followed when there are more
than two conjoined clauses. But with three or more it is
possible to mix the patterns, as in Ge 3 6 (#8.1.3). The
pattern in Ge 45 2 6 is strange because the conjunction ki
is used only in the second clause. Perhaps it is assert-
ative: Joseph is still alive, and he surely is ruler...
8.9. COORDINATION OF RELATIVE CLAUSES
It is usual for the 'relative' âser to be repeated in each

of two conjoined relative clauses.
Example: vP et-ma^syv ^se

we 1 ser
we 1 ser ^s...
and his deeds which he did...
and which he did...
and which he did...
and which he did... (De ll 3 " 6 ) 8
Since a relative clause is a noun equivalent, the coor-

dination of two or more of them together makes a noun
phrase, not a sentence. It is possible, however, to coor-
dinate two clauses together and then to nominalize the
conjunctive sentence as a whole by means of ^ser.
Example: ser kmSh l nihyt

wekjnh 1 tslp
of which the like never has been
and the like never will be again (Ex ll6)
Other examples: De 8 3 , 9 2 .
8.10. CONJUNCTIVE SENTENCES INSTEAD OF OTHER CONSTRUCTIONS
8.10.0. Introduction
In #4.6 it was suggested that the conjunctive sentence was

available as an alternative realization of almost any other
kind of more specialized coordination. Examples of this are
noted, where appropriate, in the discussion of the various
sentence types. See, for example, #10.5. Here we shall note
only a few instances of the use of conjunctive sentences
as alternative surface realizations of non-coordinate deep
relationships.
8.10.1. Hendiadys in conjunctive sentences
Certain idiomatic sequences of coordinated verbs can amount

toâ composite description of a single action. Thus per
r e b , increase and multiply (Ge 1 ,2, 97, 3511, etc.)
m e a n s be abundantly fruitful.
8.10.2. Coordination instead of apposition
The availability of the so-called 'pleonastic' vw as a

marker of phrase-level apposition is familiar to Hebraists. 7
In a similar way, a conjunctive sentence may be used as an
alternative surface realization of some of the kinds of
sentence-level apposition described in Chapter 3. See #3.0.
Example: wehannaâr tbat marê me"1 d

betl
vePis yed^h
And the girl was extremely good-looking
[she was] a virgin
and no one had had sexual intercourse with her (Ge 2416)
Example: kl îm-mên âtt lesalleh

ve^dek mahziq bm
For if you still refuse to release them
and if you still hold on to them (Ex 9 2 )
For realizations in apposition of this kind of synonymous
parallelism by means of a negated antonym, see #3.5.1.
Other examples: Ge 1 6 1 1 , 38 8 (The explanation of the charac-
ter of the act is conjoined, not in apposition [#3.7]), Le 213
(repeated a third time in apposition).
8.10.3. Coordination instead of subordination
The deep relationships between clauses that are primarily

realized with the help of subordinating conjunctions are often
sufficiently indicated by the content of the two clauses that
are joined. The specialized subordinating conjunctions can
then be replaced by and without loss of clarity. Succession

in time is often enough to suggest a chain of cause and
effect. This alternative is often resorted to in Hebrew
composition, sometimes with effects that are subtle or coy.
It is likely that the deep relationships remain elusive,
or that the choice of coordination as a neutral linkage
amounts to double talk that leaves it open to the listener
to take it any way he likes.
Example: I know that you fear God and Cwe-3 you did not
withhold your son... (Ge 22 1 2 ).
Here and means either because or so that.
Example: Please don't let my lord become angry (VJ)

and I shall speak (VC)(Ge 18 3 0 ).
While it is possible that and here is conditional

(Please don't become angry IF I speak), the sequence in
which the clauses come, and the use of the cohortative
form, suggest rather that and is consequential (if you
won't be angry, THEN I would like to say something).
The interchangeability of subordinating and coordin-
ating conjunctions is illustrated by the use of lemaân
in Ge 27z% while Ge 2 7 5 , 1 9 , 3 1 uses ba r br. But Ge 28
uses w-.
NOTES
1
For discussion of this interpretation see F. I. Ander-
sen, The Hebrew Verbless Clause in the Pentateuch (Nash-
ville-New York, 1970): p. h2.
2
Either the mem is enclitic or the suffix -m is f.
dual, referring to the twin cities.
3
AV goes one further and supplies the name of to the
preceding El Shadday, tut inconsistently does not ital-
icize by in the following clause.
^ For evidence, and the conclusion that this leads to
the abolition of the jussive, see Alexander Sperber,
A Historical Grammar of Biblical Hebrew (Leiden, 1966):
p. U36^
5
*weyabdel (VJ) would be more likely than periphrasis
in that case. And if mabdll was intended as subject com-
plement, then le- would have been appropriate as in Ge
6
Other examples are conveniently accessible in
Mandelkern's Concordance.... p. 1308f.
7
The appositional ww needs to be distinguished from
the so-called emphatic ww. Considerable bibliography and
more examples are found in Anton C. M. Blommerde, North-
west Semitic Grammar and Job (Rome, 1969): p. 29.
9
CHIASTIC SENTENCES
9.0. THE FORM OF INTER-CLAUSE CHIASMUS
The importance of chiasmus in Hebrew composition has long

been appreciated by students of prosody and rhetoric. 1 The
name comes from the resemblance to the Greek letter of
the cross-over pattern that results when the sequence
A B B ' A' is arranged in two lines.
A
X
' A'
This is considered by aesthetes to be more pleasing than

the repetition of similar items in the same sequence.
A* B'
If A and are found in one clause, and A' and B 1 are found
in the next following clause, and if these two clauses are
related to each other in a sentence, then both patterns
are found in both apposition and coordination. Repetition
in apposition is discussed in #3.3; repetition in chiasmus
is discussed in #3.4. The repetition of the same pattern
in coordination is characteristic of conjunctive sentences
(Chapter 8) and contrast sentences (Chapter 11). The use of
chiasmus between coordinated clauses constitutes a distinct
type of sentence which we call 'chiastic.'
Hitherto chiasmus has been chiefly noticed on the level
of literary appreciation and hermeneutics. The present chap-
ter examines the strictly grammatical functions of inter-
clause chiasmus as the realization of a contrastive-distinct-
ive construction in the sentence system. Chiastic sentences
are a special type simply from the point of view of discourse
syntax.
120 CHIASTIC SENTENCES
9.1. THE SURFACE GRAMMAR OF CHIASTIC SENTENCES
Carl Brockelmann 1 s brief account of the grammar of chiasmus

is based on the statement
W i e in a l l e n s e m i t i s c h e n S p r a c h e n r u f t das S c h l u s s g l i e d
eines S a t z e s das e n t s p r e c h e n d e des f o l g e n d e n S a t z e s d u r c h
A s s o z i a t i o n z u e r s t ins B e w u s s t s e i n , so d a s s e i n e c h i a s -
tische Wortstellung zustande kommt.2
This is too narrow. The item that crosses over need not be
at the end of the first clause and at the beginning of the
second. Furthermore, he does not distinguish chiasmus in ap-
position from chiasmus in coordination, and even includes ex-
amples of chiasmus in a subordinate clause.
A chiastic sentence resembles a conjunctive sentence in
joining together two compatible and closely similar clauses.
It goes without saying that the two clauses satisfy the el-
ementary requirement of coordination by having the same gen-
eral external functions as each other and as the chiastic
sentence in which they are constituents. Furthermore, the
two clauses in a well-formed chiastic sentence have at least
two clause-level tagmemes in common, represented by A and
in the first and by A' and B 1 in the second. The difference
between a conjunctive sentence and a chiastic sentence lies
in the sequence of these similar elements when the structure
of the sentence is viewed as a whole. In a typical conjunc-
tive sentence the corresponding elements are in the same se-
quence in each clause; in a chiastic sentence at least two
of the matching elements have a sequence in the second clause
the inverse of their sequence in the lead clause. The common-
est elements to be arranged in this way are S and P, that is
S and V in most cases.
Example:
Cj V SC
wa- ttehf lhem halleben l1ben
+ +
ci s SC
ve-hahemr hy lahmer
and-was for-them the-brick for-stone

and-bitumen was for-them for-mortar (Ge II 3 )
Here the grammatical effect is the same as it is with a cir-

cumstantial clause, especially those which function on sen-
tence level (#5.2). The two actions are certainly staged as
contemporaneous or simultaneous. But more is involved than
time relationships. The construction is more intimately in-
terwoven and the result more integrated than one clause aug-
mented by a circumstantial clause or than two clauses joined
together in a conjunctive sentence. The latter are unified
CHIASTIC SENTENCES 121
mainly by the fact that they have similar external functions.

Clauses in chiasmus are unified by anetwork of relationships
WITHIN the sentence.
In contrast to a conjunctive sentence, where recursion ad-
mits no theoretical limit to the number of clauses that may be
coordinated, a chiastic sentence is strictly a two-clause con-
struction. Furthermore, apart from the banal fact that every
clause (by definition) has a subject and a predicate, there is
no need for the clauses in a circumstantial sentence or in a
conjunctive sentence to have similar elements in common. It is
not necessary for a conjunctive sentence to have similar ele-
ments in the same order; all that counts is that if they do have
similar elements they will be in the same order, if only to
avoid chiasmus. In other words, a conjunctive sentence is un-
marked, a chiastic sentence is marked.
While it does no harm to look on a chiastic clause as a spe-
cial kind of circumstantial clause, the resemblance between
them, for instance, the frequency of the sequence we- +S +V,
is coincidental and arises from different causes. A circum-
stantial clause throws S (or something else) in front of V in
order to break the chain of verb-initial clauses, and it is not
necessary for the lead clause to have an explicit subject in
order to secure this effect. A chiastic clause, on the other
hand, will use the sequence SV precisely to secure a chiastic
pattern with the sequence VS in the preceding clause.
9.2. THE DEEP GRAMMAR OF CHIASMUS
In a chiastic sentence, a chiastic clause combines with the

lead clause to give a single picture of two simultaneously oc-
curring aspects of the same situation or event. It would be
possible to ban idols by saying, 'You mustn't make yourselves
silver or gold gods with me.' What Ex 20 2 0 actually says is
l^ ta^sn îttl êlhe kesep

we^lhe zhb l^ ta^s lkem,
not you-will-make with-me gods of silver
and gods of gold not you-will-make for-you
The negated verbs are the same. The accomative (îttl) and the
benefactive (lkem) do double duty. The objects are in chias-
mus and are coordinated on sentence-level instead of in a
phrase. The construction has the effect of a single prohibi-
tion and each clause makes an equal contribution to the total
picture. The construction is balanced and symmetrical. Neither
clause can be said to be in any way dependent on the other.
Hence a chiastic sentence is never used as an alternative re-
alization of any kind of subordinating relationship, as in

conditional sentences or sentences of cause and effect. Even
if the two clauses are antithetical in content, the chiastic
arrangement subdues the antithesis. Proof of all this can be
given only by placing numerous natural texts on the interpre-
ter's workbench.
9.3. CHIASMUS IN NARRATIVE PROSE
9.3.0. An Illustration
The function of chiasmus in narrative may be illustrated from

the text of Ge 4 2 " 5 --the story of Cain and Abel. Abel's part
could be narrated:
A wayehi-hebel r ê s^n
vayybe^ (hebel) mibbekrt s^n mehelbehen minh leyhwh
C wayyisa^ yhvh êl-hebel vtP el-minht
Cain's part could be told separately:'
A' wayehl-qayin ^bed h^dm

' vayybe^ (qayin) mipperl h^dm minh leyhwh
C' wel^ s^ yhvh 1el-qayin wip el-minhto
The historian has not told his story this way. Instead he has
chosen to group the six events into three closely related
pairs of events, using three chiastic sentences. In each chi-
astic sentence a pair of similar events are joined together
as circumstances of each other, using the same verb, first
as WP, then as VS.
A wayehi-hebel rê s^n
A' veqayi^hy 'bed 'dm...
' weh^bl^hebl^^gam-h^ mibbekrt s^n mehelbehen
C wayyisa^ yhvh êl-hebel veêl-minht
C' w e P e l - q a y i n wep el-minht l^ s^
Since the story begins with the birth of Cain, then Abel, the
names alternate four times. The rhetorical aspects of this pat-
tern of names were first pointed out by Prof. D. N. Freedman
(unpublished). The three sentences are also linked by chias-

mus. When the verbs are the same, the subject changes; when
the verb changes, the subject remains the same. The unchanging
item is always switched chiastically. One result is a further
chiasmus in the vocabulary of the first four clauses--Abel,
Cain, Cain, Abel, and flock, earth, earth, flock. The even
clauses are, of course, formally circumstantial to the odd
clauses, but there are so many other ways in which it could
have been done, that the consciously motivated choice of re-
peated chiasmus must be taken seriously. What is achieved
throughout is an emphasis on the similarity and contemporane-
ity of the pairs of actions. Neither Cain nor Abel occupies
the centre of the stage. It is not until the end that the
mysterious partiality of Yahweh emerges.
The chiasmus is not achieved only by a VSSV pattern. It is
as participants and not just as grammatical subjects that the
names of Cain and Abel alternate. The last clause is not *ve-
yhvh l^ s f ..., which would certainly have been circumstan-
tial and chiastic, but more patently contrastive, changing the
focus to the opposite attitudes of Yahweh rather than keeping
it on the correlative involvement of the two brothers. It is
as indirect objects that the two names are chiastic in the
last sentence.
9.3.1. Chiasmus in Poetic Discourse
The two lines of a poetic bicolon often consist of one clause

each in some kind of parallelism (often synonymous) with a chi-
astic pattern between elements common to both clauses. This
phenomenon is so widespread and so familiar to all students of
Hebrev literature as to need no documentation. What the present
study contributes to this subject is the insight that such a
construction must be viewed as a whole, as one sentence, with
complex integrating grammatical relationships between them.
Chiasmus is a syntactic as well as an artistic device.
Example: r^ rîtl êt-^ âmml

wP et-?a^qtm sma^tl,
I have surely seen my people's suffering
and their crying I have heard (Ex 3 7 )
Example: wayyiddSm hassemes

v e y r e h ^md,
and-stopped the sun
and (the) moon stayed (JoS 1 0 1 3 )
There is a similar juxtaposition of sun and moon in Is 60 2 0 .
9.3.2. Chiasmus in Epic Narrative
As already shown in #9.3.0, chiasmus is often used in epic writ-

ing, which accordingly contains many sentences which have the
appearance of well-formed poetic bicolons. This has fostered the
12U CHIASTIC SENTENCES
impression that a book like Genesis is written in classical

verse, but considerable violence has to be done to the text
to force it into such a rigid mold. When that fails, the po-
etic fragments throughout the book are explained as the rem-
nants of an epic substratum in which, presumably,the poetic
form was more sustained than it is now. As a result of this
explanation, Genesis is not treated as an example of pristine
Hebrew composition in many works of Biblical criticism.
We want to suggest that the text of Genesis is in better
shape and has more artistic integrity than this. It is nei-
ther poetry nor prose, but epic composition containing both
poetic devices and extended rhetorical structures in which
chiasmus plays a key part. The Flood Story affordsmany illus-
trations .
Example : nibqe ^ kol-ma ^yent t ehm rabb

waârubbt hassmayim nipth,
AiVlsplit B:Slall the springs of the great abyss
and B':S :the water sluices of the sky A' '.V: opened (Ge 7 1 1 )
The preceding time reference is marginal to this sentence as a

whole.
The Flood was caused by the confluence of waters from the
two great storages--the waters above the firmament and the
waters below the firmament (Ge l 7 ). The whole complex of Ge 7 1 1
is sandwiched into another chiastic sentence.
me hammabbl hy âl-h^res Cverse 11J

vayehl haggesem âl-h^res,
and Slwaters of the flood V'.were Lion the earth...
and V:was S:the rain Lion the earth (Ge 7 1 0 " 1 2 )
Again the preceding time reference is marginal to this chias-

tic sentence as a whole. The world was inundated by the simul-
taneous action of a vast tidal wave and massive pluviation.
It is precisely the use of chiasmus that brings out these two
sides of a single event. In Ge 4 i _ s the sequence of verb forms
in a chiastic sentence is WP...VS. In Ge 7 i o " 1 2 this sequence
is reversed (compare Ge 2 1 E x 23 2 8 ). Such an attachment of
a WP clause chiastically to a preliminary Ep-marginal circum-
stantial VS clause makes the resulting chiastic sentence as a
whole circumstantial to the ensuing episode. In Ge 7 1 0 " 1 2 the
chiastic pattern between verse 10 and verse 12 shows that
verse 12 does not continue verse 11. Nor is the WP clause in
verse 12 the first moment in the epic with other clauses cir-
cumstantial and subsidiary, as in classical prose.
The episode actually begins with yb^ in verse 7. The whole
of Ge 7*" 1 6 is an elaborate piece of epic composition, in
which there are only two events both already stated at the
beginning, namely, the onset of the flood and the entry into
the ark. Descriptions of these events alternate--flood (verse
6), entry (verses 7-9), flood (verses 10-12), entry (verses 13
-16a), flood (verse 17). A third event (the closing of the
door) is mentioned only once (verse 16b).
CHIASTIC PATTERNS IN GENESIS 7 6 ' 1 7
6 FLOOD (Ti V-
V A. 1
ENTRY
A' V ^
Ti V
Ti V
FLOOD Ch Ch
' V-
Ch
V '
13 Ti V A-
Ch
14-15 ENTRY A' V-
16a V
16b The LORD shut the door
17 FLOOD ( V B-
Each of the other events has two aspects. There are two cau^
ses of the flood--the eruption of ocean waters (B) ajid the
massive downpour (B1). There are two kinds of passengers in
the ark--human CA) and animal (A1). These pairs as constantly
repeated grammatical subjects are arranged in five chiastic
patterns, interspersed with various time references. It is
important not to identify verses 10-12 as two successive bi-
colons with synonymous parallelism, or the pattern is lost.
Verse 11 is a fine chiastic sentence inserted into another
chiastic sentence.
The Shechem atrocities are described in epic style with
frequent chiasmus of WP and VS of the same root.
wayyiqehu...
wayyb'...
wayyaharg 0
and 0' hreg
wayyiqefcu. . .
wayyege 'G...
<A>
S b"1 G
wayybzz
<A> 0 lqly
and 0 sb
wayybzz 0
The scheme is admittedly not symmetrical, and the chiastic

clauses are mostly not contiguous. Some clauses are in epic
apposition (#3.4.1). But the repetitions and parallelism
are impressive all the same.3
Other examples: De 3 6 " 7 , 2 2 " 2 9 , 9 e > 2 0 .
9.3.3. Three-Clause Chiasmus
By its very nature, inter-clause chiasmus is restricted to a

two-clause construction. The discussion on p. 120 shows that
each new clause can be made chiastic to its precursor to cre-
ate a closely woven fabric, but the pattern is seen by taking
the clauses in pairs--Cl 2 is chiastic with Cli, Cl 3 is chias-
tic with Cl 2 , and so on.
If, however, Cl 2 and Cl 3 are both chiastic with Clj in the
same way, Cl 2 and Cl 3 will be congruent with each other and
therefore typical of a conjunctive sentence. It does not make
any difference, then, to call Cl 2 and Cl 3 together a conjunc-
tive sentence in chiasmus with Clj.
Example: laser hiqsapta êt-yhwh êlheyk bammidbr...

- beh.3reb hiqsaptem êt-yhvh. . .
G- betabêr...maqslpim heyitem êt-yhvh...
that you infuriated Yahweh your god in the wilderness...
and at Horeb you infuriated Yahweh...
and at Tabera...you were infuriating Yahweh
(De 9'.)
See also p. 109 above.
9.4. GRAMMATICAL ASPECTS OF CHIASMUS
9.4.0. Introduction
Chiasmus can occur between any pairs of clause-level tagmemes.

It occurs most often between nuclear tagmemes, the predicate
being usually one of them. Next to the predicate, the subject
is likely to be involved. Hence S W S is the commonest pattern.
But other items are eligible, and we have already seen exam-
ples of object, indirect object, or even marginal tagmemes
like 'location' making the cross-over.
9.4.1. Verbless Predicators
It is possible for the predicate of a verbless clause to be

placed in chiasmus with a verb.
Example: kl îm-tam hakkesep

miqne habbehem êl-^dnl
that V:finished S:the silver
and S:the stock of animals [is] '.to my lord (Ge 47 1
The two kinds of wealth are used up.

In priestly discourse the dimensions of various cult objects
are specified in conjoined verbless clauses with sequence ei-
ther SP or PS throughout (Ge 6 1 5 , Ex 2 S 1 0 " 2 3 , 27 1 , etc.).
Ex 38 l e is a rare chiastic variant of this.
we'esrlm âmm ^rek

weqm berhab hmes âramt,
and 20 cubits [is] [its] length
and height in width 5 cubits
Chiasmus is seen at its best in verbal clauses.
9.4.2. Perfect Verbs
The chiastic clauses are most alike when VS is used in each.

Examples: Ge 7 1 1 , Ex 3 7 (both quoted above), Ge 17 2 e , Ex 31
(identical verbs), 40 3 5 .
9.4.3. Subjects in Chiasmus
The commonest kind of chiastic sentence in narrative prose em-

bodies the pattern
+WP +S:N wayyigber hammayim

we- S:N +VS vehammayLm gber (Ge 718 19
)
Usually the subjects are different, but the verbs are the same
even having the same roots, when similar actions by two parti-
cipants are reported as in Ge 4 2 " 5 . The similarity of this con
struction to epic parallelism with chiasmus in apposition has
already been pointed out in #3.4. That construction, however,

is more likely to be used when the subject of both clauses is
the same.
Example: wayekas he^nn êt-^hel m '"ed

k?bd yhvh male 1 1 et-hammiskn ,
and-covered the cloud the tent of meeting
and the glory of Yahweh filled the tabernacle (Ex 4031*)
Other examples: Ge 7 1 5 " 1 6 (the extreme, in which VS is right

at the end of the chiastic clause), 1 4 l 6 b (VS at end. i g 3 7 " 3
(Lot's two daughters; note reinforcement with gam-hi'), 25 2 8
(Isaac and Rebekah; the second verb is a participle in MT, mak-
ing the clause circumstantial; but the roots are the same, and
archaic VS:*"*ahabat may be suspected.), 311*7 , 35 1 8 , 36 s " (like
1 9 3 7 ~ 3 S ) , 3 7 1 1 (this makes an inclusio with 37 3 and closes the
unit), 4 4 2 0 , 45 1 * (lead clause lacks explicit subject), Ex 9 2 3
(Concomitant actions of Moses and Yahweh are represented as
simultaneous, even though a time sequence of cause and effect
might be supposed. The effect of chiasmus is to abolish the
time interval and to highlight the immediate response.), 9 3 3
(thunder and rain), 1 0 1 ? (like 9 2 3 ) , 20 1 " (the people and Moses).
9.4.4. Objects in Chiasmus
Two objects can be arranged in chiasmus when two correlative

actions are performed on different objects, especially when
the actions are similar or when the same subject performs
them.
Example: wayyiqqah m5e has! haddm wayysem bâggnt

vahasl haddm zraq. âl-hammizbeh,
and Moses took half the blood and put it in the basins
and half the blood he threw over the altar (Ex 2 4 6 )
Other examples of objects in chiasmus: Ge 1 4 l s , 19 3 , 2 4 5 3 ,

1 22 52 22
39 u , 22 ) 40 - } 41 " , 4315,iE- * Ex 14 6 (mobile and
infantry units jointly mobilized), 36 (In the prediction
in Ex 26 3 the second verb is omitted, so chiasmus is not re-
alized and the sentence becomes a split-level clause.), 3 6 1 7 ,
23-2.,33-3. 3 7 2 e b - 2 7 a (fulfilment of 30 3 " 11 ) , 3 9 1 7 , De 3 1 2 ,
4 1 * , Jos l l i 0 , Jdg 7 2 5 , 9 * 5 , IKi 183 8 .
In Ex 9 2 5 there is a more complex picture in three clauses of

the joint activity of one subject on three objects.
Clj wayyak habbrd...êt kol-âser ballade

Cl2 v?'et kol- f eeb hallde hikk habbrd
Cl 3 we 1 et-kol-ês haslde Sibber,
and smote the hail...every man and beast in the field
and all the vegetation of the field smote the hail
and every tree of the field smashed (the hail)
CI2 and C I 3 , in which the subject is realized only once, are

conjoined (Chapter 8). This sentence as a whole, with its OV
(S) sequence, stands in chiasmus with Cli with its VSO sequence
9.4.5. Indirect Objects in Chiasmus
In #9.3.0 it was pointed out that in Ge 4 2 " 5 the names of the

participants are placed in chiasmus. But in Ge 4 l f ' ) " 5 a these
names function, not as subjects, but as indirect objects. Such
a chiastic sentence links together similar actions performed
by one participant (S) on two other participants (the IO's).
Example: wayyiqr^ êlhim l'r ym

wSlahS'ek qr^ lyl
WP S:N 10 OC
Cj 10' VS OC'
and-he-named God to-the-light day

and to-the-dark he-named night (Ge I s )
As in Ge 4 s , one realization of the subject as a noun does dou-

ble duty for both clauses.
Other examples of indirect objects in chiasmus: Ge l 1 0 , 17 2 0

(Isaac is the first topic, Ishmael the second), 2 0 l e , 2 3 1 1 ,
25 s " 6 , 3 3 1 7 .
Ge 3 9 " 1 9 contains an extended introversion. In conducting
the inquiry, Yahweh God deals with the man, then the woman,
then the snake. In pronouncing the curses, the sequence is
snake, woman, man. The three are interrogated one after an-
other, using WP clauses--wayy^mer...wayyo^mer...wayy 1 mer...
The curses come in a bundle, by placing the indirect objects
in chiasmus with the verbs. This chiasmus is realized only
between the first and second clauses. The second and third
are conjoined with identical patterns (compare the discussion
of Ex 9 i s at the end of #9.4.4), that is, a conjunctive sen-
tence is placed chiastically with the first clause. The A, B,
and C type coordination also helps to unify the three. (Ver-
sions point to w? 1 el in the second clause; this is understand-
able, because the asyndeton makes it look like apposition. But
MT is certainly to be preferred, and grammatical coordination
recognized.)
9.4.6. Other Clause-Level Elements
In Ge 18 s " 7 the elements that realize the goal of movement are

placed in chiasmus.
vay?maher âbrhm hôhel...

wS 1 el-habbqr ras âbrhm,
and hurried Abraham to the tent.,
and to the herd ran Abraham
Since a time s e q u e n c e is i n v o l v e d , it c o u l d be a r g u e d that

a n e w episode begins w i t h a c i r c u m s t a n t i a l c l a u s e (#5.1.1)
at Ge 1 8 7 . But w e think that the chiasmus links the two
a c t i o n s together in a single p i c t u r e , playing d o w n the time
succession.
Other e x a m p l e s : Ge 4 1 5 ^ (Location), D e 9 2 0 .
O c c a s i o n a l l y the elements p l a c e d in chiasmus h a v e different

g r a m m a t i c a l functions in the two clauses.
Example: l q a h y a ^ a q o b êt kol-^Sser l e P S b l n

me^ser l e ' b l n ^s êt k o l - h a k k b d hazze,
Took Jacob all that is our father's
and from what is our father's he made all this wealth
, ^ (Ge 3 1 1 )
T h e p h r a s e in c h i a s m u s p s e r le'bin) is o b j e c t in the
first c l a u s e , source in the second. S i m i l a r l y , âblken is
o b j e c t in Ge 3 1 s and c h i a s t i c subject in 3 1 7 .
9.4.7. Other verbal patterns
In c o n t r a s t to the u b i q u i t o u s pattern
WP X
and - VS
the r e v e r s e sequence
X VS
WP X' is rare.
Examples: Ge 7 1 0 w i t h 1 2 , 2 1 1 , Ex 7 2 1 .
It is also p o s s i b l e to have VS in chiasmus w i t h a n infini-
tive .
E x a m p l e : benogpS ^ e t - m i s r a y i m
wP e t - b t t e n
When he smote the Egyptians
and our families he rescued (Ex 1 2 2 7 )
9.4.8. Chiasmus with Negation
Example: wayebatter ^tm battwek
veêt-!-h.assipp3r l^ b t r ,
and he split them [the animalsJ down the middle
and the birds he did not divide (Ge 1 5 J 0 " 1 1 )
A l t h o u g h these c l a u s e s are a n t i t h e t i c a l , the c o n t r a s t is n o t

p l a y e d u p , as it m i g h t h a v e b e e n done by u s i n g , say, raq.
When two negated clauses are in chiasmus, there is, of course,

no antithesis.
Example: wel^- h.efdlp hammarbe

vreliammamît lo^ heh.sld,
And the one who gathered most had no surplus
and the one who gathered least had no shortage (Ex 161)
(The hip('il participles are elative.)
Another example: Ex 1621*.
9.4.9. Chiasmus not involving the Verbs
The verb is the usual pivot around which, other clause-level

items are arranged in various patterns. Sometimes clause-level
elements not verbs are arranged chiastically. Thus 'd is chi-
astic w i t h the infinitives in Ge 8 2 1 , and w i t h the subjects in
Ge 9 1 1 .
9.5. CHIASMUS IN PREDICTIVE DISCOURSE
9.5.0. Verb Patterns
In past-tense narrative the story marches down a m a i n road

paved w i t h WP clauses. A chiastic clause added in will produce
the sequence WP...VS. (See #9.3) The corresponding pattern
in predictions is WS...VP. It is also possible to use 'imper-
fect * verbs in both clauses in chiasmus.
Example: wa^brak mebrekeyk

meqallelk
and-I-will-bless those-who-bless-you
and-those-who-curse-you I-will-curse (Ge 123)
Other examples: Ge 2 5 2 3 b , Ex 15263213.
9.5.1. Subjects in Chiasmus
Example: wehy yhwh II le^lhlm

wehêben hazz^t..yihye bet êlhlm,
and-will-be Yahweh to me as God
and this stone will be [to me] God's house (Ge 2821)
(Two aspects of a single vow)
Other examples: Ge 1 5 1 3 (âtt is chiastic w i t h the preceding

^nki), Ex 19 s CS is the same), Le 5 9 1 6 (concomitant actions
of worshipper and priest, sequential in 5 2 6 ) .
132 CKIASTIC SENTENCES
9.5.2. Objects in Chiasmus
Example: Veireg-
ve^8t8k yeha.yy,
and they will slay me
and you they will spare (Ge 1212)
Other examples: Ex 1212, 2 1 3 5 , 2 6 3 5 (symmetrical arrangement

of lampstand and table), 28 2 1 f ~ 2 5 , 283e (materials contrast),
28 110 , 29 3 " (Aaron and his sons), 2 9 1 2 (two parts of blood
ritual), 2911( (two acts of burning flesh), 3 0 3 \ 3 0 2 6 - 3 0
(two acts of anointing}, Le l 8 " 9 (coordinate treatment of
parts to be was hed), l 1 5 (like l" 9 ), 4 7 (two actions with
blood), 4 1 7 " 1 ' (clusters several related liturgical acts by
chiasmus), 4 3 0 " 3 1 ' 3 " " 3 5 , De 5 2 1 , 1 0 1 6 .
9.5.3. Mixed Constructions
There are also some mixed constructions, in which subject and

object, or some such combination, are in chiasmus.
Examples: Ge 6 l e b ~ 1 9 a (WS + S + and + 0 + VP), 17 6 (OC and S),

4721* (0 and S) .
The verse divisions at Ge 3 4 9 " 1 0 disrupt a chiastic sentence.
vehithatten ^tn
viPittn teseb,
and you will intermarry with us
and with us you will settle.
A conjunctive sentence which expounds this sentence in appo-

sition is inserted into this. The next paragraph begins with
verse 10b, hot with 10a, as its circumstantial form shows.
The detailed provisions for intermarriage are stated three
times in this chapter.
In Hamor's proposal, the two clauses are conjoined (34 9 ).
bentekem titten-ln
w<P et-benten tiqh lkem
In the Israelites' counter-proposal they are chiastic (34 1 6 ).
ventann êt-ben5ten lkem

Vp et-bentekem niqqah-ln
When Hamor reports to his subjects he uses a conjunctive sen-

tence (34 21 ) ,
êt-bentm niqqalj-ln lensi.m

we 1 et-bentin nitten lhem
3
In De 7 the clauses which prohibit such, intermarriage are con-
joined.
There may be very little difference between these two pat-
terns. But, if we are right, the Shechemites are proposing a
mutual agreement in conjunctive sentences, while the Israel-
ites are proposing a reciprocal arrangement in a chiastic sen-
tence. The Israelites are thinking in terms of complete inte-
gration, to become a single people (34 1 6 ). Although Hamor re-
peats this phrase (verse 22) it is clear from verse 23 that he
expects the Shechemites to preserve their identity and to ab-
sorb the Israelites. The Shechemite proposal is not entirely
one-sided. There is chiasmus in the pronouns
-kem -n -m -n
-n -kem (verse 9) -n -hem (verse 21)
This was not obligatory. They could have said, "and you will
take our women-folk from us", rather than "for yourselves."
9.5.4. Indirect Objects in Chiasmus
21 2 1,0
Examples: Ex 25 , 28 with (linking Aaron and his sons).
9.5.5. Other Clause-Level Tagmemes in Chiasmus
13 16 11
Examples: Ge 1 5 " (Time), 3 5 (Origin), Ex 7 2 8 (Location),
s1 3 11 12
35 with * (Infinitives), De 2 " (Location ).
9.5.6. Other Verb Forms in Chiasmus
In contrast to the common pattern
WS X
and X' VP
the reverse sequence
X VP
WS X' is rare.
Examples: Ex 7 1 8 , 23 2 e .
It is also possible to have a VP in chiasmus with an infinitive,

as in Ge 3 , or with VS, as in Ex 5 2 , 7 1 , or with a participle,
as in Ex 1 3 l S (both are present tense).
9.6. CHIASMUS IN PRECATIVE DISCOURSE
The production of a precative chiastic clause is remarkable be-

cause chiasmus pulls a precative verb away from the clause ini-
tial position which it almost invariably occupies. A n example
already discussed in #8.3.2 is found in Ge 4 3 " . There are three
commands, to take a present, money, and Benjamin. The first has
CHIASTIC SENTENCES
normal sequence; <i?h muzzimrat h^res...; the next two are

chiastic with this {and therefore conjoined to each other):
wSkesep mi5ne q5h... weO e t h l k e m qh... Although the
brothers must have had to do tiiese things in some kind of se-
quence, the instructions are issued as a bunch of associated
or concurrent events. (The fulfilment uses the same kind of
chiasmus in Ge 43 I 5 .)
Another example using VI:
ten-ll hannepeif
wehrekus qah-lk,
give me the people
and the spoil keep for yourself (Ge 14zl)
Other examples: Ex 17 s , De 21*.
VP may be used in chiasmus with VI.
Example: w5slm kesep-Îs...

viP et-g?t>I ^ . . . ti? Im. . . ,
and put each man's money...
and my chalice... you will put... (Ge 441)
Other examples: Ge 3 2 1 7 , Ex 1 8 1 9 " 2 1 (heye att.. . we 1 att

teljSze . . .) .
Ge 44 3 3 has an example of jussive verbs in chiasmus.
yefeb-n^ âbdek tahat hannaâr

wehanna'ar yaâl îm-êhyw,
Let-remain please your slave instead of the youth
and let the youth go up with his brothers
Other examples: Ge l 2 0 (discussed in detail in #8.3.0), l 2 2

(but the preceding text does not have a V S:N clause to make
the chiasmus, although doubtless the command to fill the sea-
water is addressed to marine creatures).
Chiasmus can also be developed in precative discourse between

verbs of different kinds. In Ge 225 an ethical object stands
in chiasmus with a subject--wait you here with the ass and I
and the boy will go to there.
In Ge SS11* ""adnl = "'att. The chiastic pattern
VJ S:N
and S:Pr VC
is casual proposal that soft-pedals the divergence between

the actions. As will be shown in Chapter 11, the contrast
could have been brought out by
*weâtt âbr-n^ .. .
waânl êtnhXl. . .
9.7. CHIASMUS IN P R O H I B I T I O N
The p a r a l l e l p r o h i b i t i o n s in Ge 37 2 2 are s y n o n y m o u s , and this

may be w h y they are in chiasmus rather than conjoined.
^ al-tifpek-dm
veyad al-tiSle^-b ,
Don't shed blood
and a hand don't send against him
Compare w e y d e n al-tShl-b in Ge 3 7 2 7 .
9.8. INCOMPLETELY FORMED CHIASMUS
There are some sentences w h i c h throw two items into a sequen-

tial p a t t e r n that suggests c h i a s m u s , but there is no second
pair of closely c o r r e s p o n d i n g items to complete the chiasmus.
Example: ki l 1 himÎr yhwh 1 e l h l m ''al-h^res

w e ^ d m âyin la^bd ^ e t - h ^ d m ,
for not made-rain Yahweh God on the earth
and man did-not-exist to work the ground (Ge 2 5 )
This gives the reason for lack of v e g e t a t i o n in the world. Nei-

ther God n o r m a n was doing his a c c u s t o m e d w o r k in w h a t is essen-
tially a joint o p e r a t i o n - - G o d sends r a i n , m a n tills the soil.
The S:N y h w h 1 e l h l m and ^dm are p l a c e d for c h i a s m u s , but the
verbs do n o t correspond. 'Yahweh G o d d i d n ' t make it rain on
the e a r t h . ' But the n e x t clause d o e s n ' t s a y , 'and m a n didn't
w o r k the s o i l . ' It denies the existence of m a n a l t o g e t h e r , and
a c l a u s e to i n t e g r a t e w i t h this in a chiastic sentence w o u l d
say that there was n o G o d to w a t e r the earth.
In Ge 25 3 3 " 31 * s u c c e s s i v e clauses have s i m i l a r e l e m e n t s .
w a y y i m k r êt-bekrt leya^qb
w e y a ^ q b ntan leêw lehem...
The g r a m m a t i c a l structure is
Cj V 0 10
Cj S V 10 0
andr-he-sold LEsaul his birthright to Jacob

and Jacob paid to Esau bread...
The 0 and 10 are in c h i a s m u s . Subject and verb fail to achieve

chiasmus for lack of an explicit subject in the lead clause.
The p o i n t is of some m o m e n t , for it leaves the d i s c o u r s e func-
tion of this clause in some doubt. Many t r a n s l a t i o n s b e g i n a
new p a r a g r a p h at this p o i n t , ending the legal n e g o t i a t i o n s w i t h
v e r s e 33b, and making Jacob's act the first m o m e n t of Esau's
subsequent meal.
There is something to be said for this. Verse 34a certainly

has the form of a circumstantial clause that could begin a
new episode C'5,1.1). And the brachylogy in the next clause
C a n d he ate [them]) secures continuity w i t h the following, so
verse 34a is not a circumstantial clause used to end an epi-
sode. There are, however, reasons for preserving continuity
between verse 33 and verse 34 by recognizing > 3 B" 3 I , a a s a
chiastic sentence. The formal side of the business is com-
pleted with the oath in veTse 33a. What follows is an expo-
sition of this--Esau's side and Jacob's side. This is sup-
ported by the chiasmus already noted; supported further by
recognizing the lexical meaning of ntan in such a context
as paid (an agreed price). It spoils the story to say Jacob
'gave' him bread and lentil soup, and it is an added touch
that Esau did not say 'give' but 'gulp' me some of that red
stuff, for ten-li could mean also 'sell.' We conclude that
and he swore to him is the climax, and that the tension be-
gins to wind down after that point. Verse 33b is not paral-
lel to verse 33a, but to 34a, in a chiastic sentence. Jacob
in verse 34a is thus in chiasm with the implicit subject
(Esau) in verse 33a.
The same thing happens in Ge 2 7 l s where the act of disguis-
ing Jacob occurs.*
9.9 DISCONTINUOUS CHIASTIC SENTENCES
The preceding discussion included incidentally several exam-

ples of pairs of clauses in chiasmus which are separated by
intervening material. See especially the diagram on p. 125.
Here it remains to observe that widely separated clauses in
chiasmus mark the onset of two parallel or contemporaneous
paragraphs. The second clause will, of course, resemble a cir-
cumstantial clause as a mark of paragraph onset, and all that
was said in #5.1.1 applies to it. But its similarity to a chi-
astic clause further back lines the two ensuing paragraphs up
side by side, so far as sequence patterns are concerned. Thus
Moses' two correlative speeches to the people and Joshua are
reported in De 3 1 " 2 2 .
w^sav "'etkem bêt hahi^ le^mr "q"

wP et-yehos^ iwwetl bêt hahl^ le^mr "Q",
and I commanded y o u a t that time, saying "Q"
AND Joshua I commanded at that time, saying "Q"
9.10. CHIASMUS AS A HIGH-LEVEL NODE
On the m a i n thread of Hebrew discourse consecutive clauses

CWP or WS) come one after the other like beads on a string.
fWP| WPIWPIWPIWP |WP"|
The effect of a circumstantial clause of the type discussed

in #5.1.3 is to add on a bead beside the string.
I WP I WP | WP
WP|WP WP|WP|
CirCl
In a chiastic sentence both clauses are equally important, and

this is like incorporating two clauses side by side into the
string.
| WP I WP I WPWP WPIWPIWP1
ChCl
In Ge 4 2 5
a series of such sentences has the effect of synap-
sis
WP WP WP
|WP|WP WP1WP1
ChCl ChCl ChCl
It is quite possible for any such clause to serve as the start-

ing point for a considerable chain of clauses, so that the pair
of chiastic clauses make a fork.
The verb hissmer (VI), watch yourself, in De 8 1 1 is continued
by vezkart (WS), and you will remember in De 8 1 8 . All the in-
tervening material is governed by the conjunction pen, lest.
There are in fact two pen clauses in apposition. The first (8 11 )
is general (lest you forget); the second, in apposition (#3.7.2),
expounds in detail what this means. By means of chiasmus this
exposition branches into three threads. De 8 1 2 is a paragraph
headed by a VP clause that is chiastic with 8 l z a and continued
by a WS clause that is strictly sequential to 8 1 2 b A . 8 1 3 is a
conjunctive sentence of three VP clauses with the same verb, and
the same sequence pattern. This pattern is the same as that in
8 1 2 b A and so is chiastic with 8 l i a A . So 8 1 2 b A and 8 1 3 are each
chiastic with 8 1 2 a A , making a three-fold branch. This three-fold
branch reunites at 11*13, which is the sequel to all three of the
paragraphs which are united by chiasmus between the clauses at
their heads. This thread ends with 8 1 7 , which completes the lest
'clause1 that began with 8 1 2 . That is, the whole of 8 1 2 " 1 7 is in
apposition with 8 l l b , and the whole of 8 l l b ~ 1 7 is subordinate to
8 Hence 8 l e is not the sequel to 8 1 7 , but the antithesis of
giib-17 an( j t h e s e q u e i t 0 8ll--Benrare lest you forget and...
but (on the contrary) you will remember...
The connections secured by the hierarchical grammar of Hebrew
discourse are shown on p. 138. The example is instructive, be-
cause the results are different in translation when these struc-
tural signals to the higher levels of discourse structure are
not heeded. RSV, to look at only.one attempt, takes consider-
able liberties with the conjunctions, thus throwing the paragraphs
into quite different arrangements, breaking some close linkages,
and creating others where there should be a break. It omits the
conjunction at the beginning of 8 l e , making a break where there
is sequence. It misses completely the trio of clauses governed
by pen, which, as we have seen, are unified by chiasmus. Instead,
it adds a gratuitous beware lest to verse 17, severing its se-
quential connection with verse 14b.
HIERARCHICAL STRUCTURE IN DEUTERONOMY 8ll"le
11
Watch yourself (VI)
'--
~~Lest you forget Yahweh your God.
+
<A>
Lest you wil

eat (VP) <Ch>
\
and [lest] and [lest] your
fine houses herds and flocks will
you will multiply (VP)
+
build (VP)
<Cj>
and silver and

gold will multiply
(VP) for you
+
<Cj >
I
and all you- own
will multiply (VP)
<Seq> <Seq> <Seq> <Seq>
and you will and you will and your mind will
become re- dwell [in become superior
plete (WS) them] (WS) (WS)
<Seq>
and you will forget (WS)

Yahweh your God who...
<Seq>
and you will say (WS)

in your mind...
and you will remember (WS) Yahweh your God...

9.11. CHIASTIC SENTENCE AS NUCLEUS
Chiasmus usually takes place between nuclear tagmemes and the

result is a tightly knit structure which can function just like
a singular clause nucleus. It can be modified as a whole by a
marginal tagmeme, or governed as a whole by a subordinating
conjunction or by the 'relative.'
Example: vehinne nuttas mizbeh habbaâl
ve-h^ser ^ser-^lyv krt,
and behold, broken down was the altar of Baal
and the Asherah beside it was cut down (Jd 628)
Another example with hinne: ISa 18 2 2 .
Example: kl ksel yerslayim

wlhd npl,
For stumbled Jerusalem
and Judah fell (Is 3 8 )
Example: ^ser hehezaqtlk miqqest h^res

me^sileyh qer^tlk,
who I seized you from the ends of the earth
and (who) from its extremities I called you (Is 419)
9.12. CHIASMUS A DISTORTION
The formal requirements of chiasmus, so far as word order is

concerned, may cause a perturbation in normal syntax.
Example: ki kerem yhvh seb^t bet yiisrêl

v<Pls yehd neta' sa^s^yw,
For Yahweh Seba^ ot's vineyard is the House of Israel
and the men of Judah are his pleasant plantation (Is 5 7 )
The riddle is to identify the vineyard, so this is the gram-

matical subject of the solution. The parallel clauses are not
only synonymous; they have matching grammatical elements in
chiasmus. Hence the structure
Cj S
Cj S
Brockelmann analyses PSSP.5 The literal RSV gives structure

SPSP. But men of Judah must be the predicate of the second
clause.
9.13. SEQUENTIAL CLAUSES IN CHIASMUS
We have emphasized that a chiastic sentence represents two

events as contemporaneous. The construction is strained when
the two events must occur in sequence. Thus the instruction
of the next generation in De 4 1 < f t B is chiastic, not sequential.
This must be considered atypical.
lUo CHIASTIC SENTENCES
NOTES
1
Nils Wilhelm Lund, Chiasmus in the New Testament
(Chapel Hill, I9U2). This book reviews the history of
the study of patterns of introversion in biblical po-
etry. It contains an ample bibliography of the liter-
ature up to that date. Lund's illustrations include
many from the Old Testament. He had earlier published
papers on chiasmus in Hebrew poetry. More recent stud-
ies include William L. Holladay, "Chiasmus, the Key to
Hosea XII:3-6", Vetus Testamentum 1 6 ( 1 9 6 6 ) : pp. 53-
6b\ . T. Radday, "Chiasm in Samuel", Linguistica
Biblica Heft 9/10 (1971): pp. 21-31. These are prim-
arily literary investigations, and the currently bur-
geoning RHETORICAL CRITICISM can be expected to yield
a fresh harvest of observations. The grammar of chias-
mus presented here should add another dimension to this
research. It should be remembered, however, that these
authors use the term CHIASM(US) to refer to all kinds
of introverted patterns, whereas we restrict it to the
scheme AB-B'A' between coordinated clauses.
2
Hebrische Syntax, 138.
3
The assignment of verse 26 to J cuts across this
epic structure. See Note 1 p. 59 above.
This chiastic sentence has the familiar pattern
WP ... VS, where the verbs have the same stem
wattalbes ... hilbls, and the same subject. An addit-
ional feature arises from the fact that this verb nor-
mally has two direct objects -- the wearer and the garment.
In this passage each clause has only one object, the wear-
er in the first, the garments in the second. But the in-
tegration of successive clauses into sentences enables an
object in one clause to do double duty (brachylogy) in the
next. And Rebekah got Esau's clothes... And she made
Jacob Cwearer as object is explicit} wear (them) Cgarment
as object is present only by means of brachylogy from the
preceding clause] And the goat-skins Cthis object is chi-
astically in front of the VS verb] she made (Jacob) Cthis
object present only by means of brachylogy from the pre-
ceding clause] wear over his arms and neck.
5
Hebrische Syntax, 138.
10
DISJUNCTIVE SENTENCES
10.0. DISJUNCTIVE COORDINATION
Hebrew has a conjunction ^, whose main meaning is or and whose

primary function is disjunctive coordination. In ^ , X and
Y are alternatives, corresponding to mutually exclusive alter-
natives in semantic structure. We note also (in #10.3) the use
of the conjunction îm, which usually means if, as an alterna-
tive to "'S in disjunctive questions.
10.1. PHRASE-LEVEL DISJUNCTION
In most of its occurrences, ^ is phrase-level, linking two

nouns or two prepositional phrases. The phrase-level consti-
tuents linked in this way have, or could have, the same exter-
nal function. As in conjunctive coordination of prepositional
phrases, both constituents have the same preposition, since
Biblical Hebrew very rarely used one realization of a prepo-
sition to govern both nouns in a conjunctive or disjunctive
phrase.
10.1.1. A or
Examples of Ni or N2: ra f ^ tob, bad or good (Ge 24 50 ); also

Ge 245 5 , 44 s ' 1 9 , Ex 21" ' 2 1 ' 3 2 ' 3 3 ' 57 , 22 ,( ' 5 ' 6 ' 9 . 2 3 % Le 3 3 ,
5,7 7I6 - Q 3 2 ' 3 2 ' 32 12 6 ' 8 132 ' 2 ' 1 9 ' 2 "'29 3 ' 3 8 * 9 15 1 *'
29
202 7 ,'etc.
Examples of PpPhj 1 PpPh2 : fal-ymin 1 ''al-ISm^l, to

right or to left (Ge 24"9 ; also Ge Si" 3 , Ex 5 3 , 2 1 6 ' 1 8 ' 2 0 '
2 6 , 2 7 ,2 V,29 L e 1 0 , I H g2 , 2 , 2 , 1 1 , 2 1 , 2 t y21,21 12 6 7 13 2 '
2 9 , I 2 , 4 2 , 3 | ) 7 , I. 8 , 4 , ) , >t 8 , "> 9 , "f 9 , "t 9 , 9 ' e c '
Ik 2 DISJUNCTIVE SENTENCES
Examples of IfPhj "- IfPh 2 ; Ex 28" 3 , 30 2 0 , Le 5" > 2 1 2 1 r 23 >

23
> 2 \ 13 5 9 , etc.
10.1.2. Either A or
An alternative but rare phrase type uses ^ before both

disjunctively coordinated items. Compare the syntax of
gam (Chapter 12).
Example: ^3-ymayim ^-hdes ^-ymim, whether two

days or a month or a year (Nu 92 ).
10.1.3. Either A and
While the use of coordinating we- and as an alternative

for ^ is not uncommon (#10.5), the use of ve^ and
correlatively must be considered a curiosity.
Example: ^ ymam vlayl, whether by day or by night

(Nu 9 2 1 ) .
10.1.4. Whether A or
Example: "'im-behem îm-îs, whether animal or human

(Ex 1 9 1 3 ) .
Other examples: Le 27 2 6 , De 18 3 , Jos 24 1 5 .
10.2. DISJUNCTIVE COORDINATION ABOVE PHRASE LEVEL
10.2.0. Transformations up and down the hierarchy
Disjunctive coordination has fascinated logicians. But

disjunction as a binary operation in symbolic logic is
by no means simply related to the use of a conjunction
like or in natural languages.1
One property of clause-level disjunctive phrases often

remarked on is the expandability of the clause contain-
ing one of them to a sentence consisting of two clauses
of similar structure in disjunctive coordination. So
Ex 21lf weyled- bnlm ^ bnt, and she will bear
to^him^sons or daughters, J/iould be equivalent to
*weyled-l bnlm ^ yled- bnt, and she will bear
to him sons or she will2 bear to him daughters.
The matter is not as simple as that. Sentence-level dis-

junction seems to have moved further away from the fact
that phrase-level disjunctions are not necessarily ex-
clusive. There seems to be more room in the phrase-level
disjunction for the possibility that she might bear him
both sons and daughters. This has been recognized in
DISJUNCTIVE SENTENCES
English by the use of the barbaric "and/or", jneaning both

and and or or either and or or. Such a conjunction or dis-
junction of conjunctions puts a strain on English, although
constructions like "when and if" are tolerated, In Hebrew,
however, such constructions are totally unknown, (English
can also combine prepositions governing the same item,
like "under and over the rope", rather than coordinating
prepositional phrases -- "under the rope and over the rope".
But Hebrew never coordinates prepositions in this way.)
Keeping in mind that the Hebrew constructions adduced
above might be realizing more than one deep structure,
the transformation from clause to sentence represents a
movement of "or" up the grammatical hierarchy. Movement of
"or" down the hierarchy involves deletion of material re-
peated in the coordinated clauses. It is a curious feature
of Hebrew that it insists on full repetition in phrase-
level coordination, but avoids repetition in sentence-level
coordination by eschewing sentence-level disjunction for the
most part. Only rarely does function on sentence-level,
joining disjunctively two clauses which have the same verb
with the same subject. A clause of the type weki-yiggah sr
êt-îs "" êt-îss, and when a bull gores a man or a woman
(Ex 21 2 8 ) is more common than a sentence of the type "'-ben
yiggah ^-bat yiggah, whether it gores a son or it gores a
daughter (Ex 21 3T )."
#10.1.1 showed that alternative locations are readily joined

by ^ on phrase level. A rare example of the transformation
of such a construction to sentence level with repetition of
the verb is Nu 14 2 -- l-matn biPeres misrayim ' bammidbr
hazze l mtn, would we had died in the land of Egypt or
in this wilderness would we had died. Note the chiasmus.
Another of the rare instances of the repetition of a verb

when it is the subjects that are alternatives is Le 25"* 8 " 9 .
âhad meêhyw yig^lenn

dd ^-ben-dd yig^lenn
^-misseP er besr mimmispaht yig^lenn
^-hissig yd venig^l
[When an Israelite is forced by poverty to sell him-

self to a non-Israelite]
one [=any] of his brothers will redeem him
or his uncle or his uncle's son will redeem him
or any relative who is a member of his fratry will
redeem him
or, if he can raise the money, he will redeem himself.
The need for a distinct clause to express the final alterna-

tive is unavoidable, since the verb is different. Furthermore
the condition stated explicitly needs a verb to be attached
to subordinately. But the other provisions are divided among
DISJUKCTiyE SENTENCES
three clauses, each of which has the same verb, These, how-
ever, can hardly be reduced to one clause by joining all
the subjects together and using the verb only once. The
result would be grammatical, of course. But the listing of
a set of alternative subjects in one clause would suggest
that it is a matter of indifference which one does it.
The listing of a set of alternative actions in four clauses
in disjunction, as in the text, implies four grades of re-
sponsibility. First a member of his immediate family should
redeem him. If he is unable to do it (so it is implied],
then a near relative, such as an uncle, steps in. If that
fails, the chain of obligation passes to more distant rel-
atives, and only when that resort fails is the victim left
to raise his own ransom.
Nu 5 1 1 " 3 1 deals with the case of a man who suspects his
wife of infidelity. The alternatives are simply whether she
is guilty or not. In Nu 511* these alternatives are trans-
ferred to the husband. A paragraph of three clauses describes
his suspicions in either case, and two such paragraphs are
joined by "'. There is a lot of repetition. It is quite un-
usual to find "> pushed so high up the hierarchy, except in
Leviticus. See #10.2.3.
The tendency to keep ^ as far down in the hierarchy as
possible and to minimize sentence-level repetition results
in Le 5 1 in a clause which has one common subject for three
alternative verbs joined disjunctively, in one predicate.
weh^ êd
1 r^
^ yda ^
and he [is] witness

or saw
or knew
Priorities are implied -- first the formal witness, then the

eye-witness, and only then hearsay. It is also possible, how-
ever, that our translation is wrong. For if the construction
type shown in #10.1.2 is being used, whether he saw or whether
he knew is in apposition with witness and expounds it C#3.7.2).
There are two kinds of witness, firsthand and hearsay. We sus-
pect from this example that clause-level disjunction of pred-
icates with a common subject is the least favoured alternative
to sentence-level disjunction with repetition, that the use
of a disjunctive subject is more acceptable, and that the use
of a disjunctive phrase for object, location, etc., is defin-
itely preferred to spreading out such alternatives in disjoined
clauses. There are limits, however. In Ex 21 3 3 two clauses with
common subject and object are not reduced by having onl the
verbs in a disjunctive phrase: *wekl yiptah "'S yikre Îs. I3r.
Instead, each clause is fully generated, and even the conjunc-
tion is repeated:
DISJUNCTIVE SENTENCES 1^5
wek.1 yiptah H a b3r

^ ki-yikre' ^la b5r
and if someone takes the cover off a cistern

or if someone excavates a cistern
Other examples of ki ... ^ kl ...: Le 5 1 " 5 , 13 J 6 , 15 2 5 .
The conjunction H m , if, is not so prone to repetition

in alternative conditional clauses linked by ^. Nu 35 1 6 " 2 3
spells out the exceptions to the provision of refuge in
cases of accidental homicide. There are three kinds of
circumstance, and each has three alternative conditions.
(i) If the impliment used creates a presumption of

premeditation --
Verse 16 w P i m bikll barzel. . . but if with an iron weapon. . .
Verse 17 viPim begeben yd ... and if with a stone club ...
Verse 18 ^ bikll ês-yd ... or with a wooden club ...
These are murder. The 'redeemer' will kill the 'slayer'.
(ii) If there is evidence of malice aforethought or of pre-

meditation --
Verse 20 vP im-beSin1 yehdopenn and if in hatred he pushed

him
^ hisllk ^lyw bisdiyy or threw something at him
from a hiding place...
Verse 21 ^ bP eb hikkh beyd or in hostility struck him
with his fist...
The 'smiter' is guilty of a capital offence.
(iii) If the instrument is casual and there is no evidence

of animosity --
Verse 22 veîm-bepeta^.. hdp... and if on the spur of the

moment he pushed him
f
"'-hisllk lSyw kol-kell or threw something at him
without hiding
Verse 23 '3-bekol-"-eben, . . or made any stone fall on
him without seeing...
This was presumably an accident; there must be a proper trial.

The conditional conjunction îm is always used with we-, even
when the latter means or; but it is never "used with ^, even
when the latter links disjunctively another conditional clause.
The preference for coordinating nouns rather than verbs or
clauses disjunctively is seen when alternative relative clauses
lit6 DISJUNCTIVE SENTENCES
are involved. Instead of
"Is âser X Y
or Îs ^ser X ^ser Y

1
use Îs ^ser X ^ 1s ^ser Y
Examples: Le 17 3 , 2 1 1 9 , 22" ,5.
10.2.1. Disjunctive Sentences
In view of the discussion in #10.2.0 clauses which state al-

ternatives will be joined disjunctively on sentence level
when they do not have elements in common or at least when
the subject is the only clause-level tagmeme they have in
common.
Example: kl s q S l yissqel
^-yr yiyyre
for it will be stoned

or it will be shot (Ex 1913)
Other examples: with common subject - - E x 2 2 1 3 , Nu 11,

30 1 1 ; with change of subject -- Le 1 9 2 0 .
10.2.2. Paragraph-level Disjunction
When a series of future or conditional actions are described

sequentially by means of WS clauses, and there is a branch
in the chain due to alternative possibilities, ^ may be used
with VS and the tense is sequential future. In other words,
^ acts just like wv-consecutive in paragraph-level dis-
junction.
Examples: Le 4 2 3 , 5 21 , 251'9, Nu 11 (past frequentative).
In other contexts, however, VS after ^ is simply past

tense.
10.2.3. Disjunction of Paragraphs
Whole paragraphs which represent alternative possibilities

may be linked disjunctively by The commonest examples
are those in which the initial clause in each paragraph
states an alternative condition.
Examples: Ex 2 1 3 6 , Nu 5 1 1 " 3 1 , and often in Leviticus.
On this level of^the grammatical hierarchy, however, it is

common to find we-, used as an alternative realization of
the disjunctive conjunction. See #10.5.
DISJUNCTIVE SENTENCES ll7
10,3. DISJUNCTIVE QUESTIONS
20.3.0. Fully formed Disjunction
For the disjunctive coordination of two YES/NO questions

one might have expected that each clause would be marked
by the interrogator hS- and then these two independently-
interrogative clauses would be joined by But this
construction is rarely realized. The only example I have
found is Mai l e .
hayirseks
^ hyiisi1 pneyk
will he be favorable to you?

or will he lift up your face?
Compare Job 16 3 , where the second question (af t e r is

marked with ma, and Job 382 , where the second question
is marked with 11.
10.3.1. Interrogated Disjunctive Sentence
Alternatively, the two clauses may be coordinated dis-

junctively, using^, and the whole questioned with ha-.
Example: hateqasser ma^dannt kim

"-msekt kesll tepatteh
Can you bind the chains of the Pleiades?

Or the cords of Orion [can] you loose? (Job 3831)
In other pairs of questions in this passage the place of ">0

is occupied by either we- (#10.5) or îm (#10.3.2).
10.3.2. The Normal Construction
The common form for disjunctive questions is to have h-

on the first clause and îm on the second clause.
Example: hmlk timlSk flen

im-msQl ti_ms3l hn
Will you indeed reign over us?

[or] will you indeed rule us? (Ge 37e)
Other examples: 1 Ki 22 J 5 . Is 1 5 , Je 5 2 3 , Job 4 J 7 , 6 s ' 6 ,

1 0 S 5 , ll 7 , etc.
The exact role of "'im is not clear. If it substitutes for

as used in #10.3.1, it is a conjunction meaning or. If
it is parallel to ha- and synonymous with it, the clauses
are in apposition (#3.3) and no formal mark of disjunction
ll8 DISJUNCTiyE SENTENCES
is realized, That the second analysis is correct is supported

by the occasional use of veVtm to coordinate a second question
CJob 8 3 , II 2 , 211*, 22 3 ) , The analagous combination Îm like
ha- in #10.3.0 would clinch this, but it is not attested.
10.3.3. Phrase-level Disjunction of Questions
In accord with the principles outlined in #10.2.0 above, the

alternates being questioned are more lilcely to be realized
as an antithetical phrase in the predicate of a single clause.
Example: hameât h^ îm-rb
Is he little or big? CNu 13lebB)
Other examples: Nu 1 3 1 , a > i s t , Jos 5 1 3 . The phrase hehazSq...

harpe. is [he] strong or weak? is found as predicate in
Nu 1 3 l S t A . The alternatives are joined by ^3 in Jos 1 8 1 9 ,
2Ki 6 2 7 , and Ecc 2 1 9 . Finally Joel l 2 uses we 1 im in phrase-
level disjunctive coordination of questioned items.
10.3.4. Redundant Antithetical Tag Question
A yes/ho question may be followed up by "'im 15"1, or not,

which in effect restricts the YES possibility to h-, is
it true that?
Examples: Ge 24 2 1 , 27 2 1 , 3 7 3 2 , Ex 16", Nu I I 2 3 , De 8 2 .
The antithetical correlatives yes and ^Syin yield the dis-

junctive combination hyes... 'im-^yin, is there... or
isn't there? (Ex 17 7 , Nu 1 3 2 0 ) . Once the combination ha-...
""im-âyin is met (2Sa 17 6 ).
10.4. DISJUNCTIVE REALIZATION OF CONJUNCTIVE RELATIONSHIPS
The disjunctive conjunction is sometimes used when the two

items joined are not semantically exclusive alternatives,
when "and" would seem to be more appropriate. The priests
are to wash their hands and feet both when they enter the
meeting tent and also when they draw near to the altar.
These AND/OR conditions are joined by ^3 in Ex 281*3 and
3 0 2 0 . In Nu 15 s ^3 means and not or.
10.5. CONJUNCTIVE REALIZATION OF DISJUNCTIVE RELATIONSHIPS
Items which are clearly alternatives are sometimes joined by

"and" rather than by "or". Ex 21 3 7 recognizes two^things a
thief might do with a stolen animal: te.hh ^3 jneirS, and
he will butcher it or he will sell it. ButEx 2116 links the
alternative means of disposing of a stolen human being by
we-: mekr wenims 1 beyd, and he will sell him and Tor J
he will be caught in his possession.
DISJUNCTIVE SENTENCES ll+9
In studying disjunctive questions in #10,3 it was already

obseTvea that a variety of means of coordination can be used,
sometimes in such contiguity that questions of dialect can
not enter. The interchangeability of conjunctive and dis-
junctive conjunctions does not seem to correspond in any way
to the deep structure relationships between the linked quest-
ions, in particular whether the questions point to mutually
exclusive answers, and so are disjunctive, or whether they
all have the same answer, and so are conjunctive.
In seeking an explanation of Abraham's conduct, Abimelek

asks him two questions, only one of which can give the
right answer. But they are joined by "and" not "or" CGe 20 9 ) .
The questions Yahweh asks Moses in Ex 4 1 1 all have the same
answer, but they are joined by "or" not "and". Here, however,
there is both sentence-level and phrase-level disjunction,
complicated by the ambiguity that "or" can mean "and/or".
See also De 2 32 > 3 ".
NOTES
1
D i k , Coordination...: pp. 259-270.
2
VS can be consecutive future after ^.
11
CONTRASTIVE SENTENCES
ll.O. THE DEGREE OF CONTRAST
A mild contrast between two related clauses can be secured

without advancing to the outright opposition between anti-
thetical and exclusive clauses.
The participants in two parallel but in some ways differ-
ent activities are brought into prominence by realizing them
as grammatically similar items in preverbal positions. A
common way of doing this is to refer to the two participants
by means of explicit pronoun subjects.
Example: h^ yesgek r^s

wePatt tespenn ^qeb
He (on the one hand) will crush you head
and you (on the other hand) will crush him heel (Ge 3 1 5 )
Here the identical verbs point to similarity, while 'head'
versus 'heel' is contrastive. But neither of these pairs is
brought into prominence. The contrast is between 'he' and
'you'.
11.1. CONTRASTIVE SENTENCES AND OTHER CONSTRUCTIONS
The typical sequence (we- +S +V...) in a contrasting clause

is just like the sequence in many circumstantial clauses and
also in many chiastic clauses. It therefore needs to be em-
phasized that it is not the pattern in the coordinated clause
alone that determines its grammatical function. It is the to-
tal pattern of the two clauses together that determines the
total effect in the resultant sentence as a single construc-
tion. The contrastive sentence is marked for contrast in
that both clauses have features which are unusual from the
grammatical point of view, whereas the clauses in the neutral
(unmarked) conjunctive sentence are more ordinary. There is,
for instance, no obligation to use an explicit preverbal pro-
noun subject, as in the example above. Its very redundancy
focuses attention on it, and brings the two participants into
contrast.
CONTRASTIVE SENTENCES 151
In assessing a clause as 'unusual', allowance must be jnade for

the different sequence patterns which are 'usual' for different
verb forms. The verb is usually clause-initial when it is VSj
VI, VC, or VJ, so any item preceding such a verb is in a strik-
ing position. VP and Pt are rarely clause-initial, so a prever-
bal position in a clause with or Pt is less likely to be
contrastive, unless.contrast is secured by content. Many such
sentences are listed as neutral (conjunctive) in Chapter 8.
11.2. CONTRAST WITH PRONOUN SUBJECTS
Two balancing clauses joined in this way constitute a contras-

tive sentence.
Example: htP yihye-llek lepe

weâtt tihye- le^lhim,
He will be a mouth for you
and you will be God for him (Ex 416)
The same pattern is repeated in Ex 7 2 -^You fâttSj will

speak...and Aaron your brother will speak. These are Moses' and
Aaron's contrasting roles. But when the arrangement of Ex 4 1 6
is presented as God's act, the two parts are more closely inte-
grated by means of chiasmus.
netatti-k êlhim lepar^

wePaharn ^hik yihye nebiêk,
I have appointed you god for Pharoah
and Aaron your brother will be your prophet (Ex 71)
Another such mixture of noun and pronoun subjects is found in

Ex 21".
The respective duties of two partners in a contract or joint
enterprise can be spelt out in paragraphs which begin with the
same contrastive device.
Example: wa^nl... (Ge 6 1 7 )

wePatt. .. (Ge 6 2 1 )
Other examples: Ge 9 7 with 9 9 ; l?" with 17 i 0 ; Ex 2 5 , 1 4 1 6 " 1 7 ,

De 4 2 1 .
11.3. CONTRAST WITH NOUNS AS SUBJECTS
Explicit preverbal noun subjects can be used in coordination

to bring out a slight contrast.
Example: âbrfim ysab bP eres-kenân

velt ysab beware hakkikkr,
Abram settled in the land of Canaan
and (while, butJ Lot settled in the cities of the plain
(Ge 1 3 1 2 )
152 CONTRASTIVE SENTENCES
Other examples; Ge 27 17 (.Leah and Rachel), Ge 4 2 1 3 ? 3 2 (Joseph,

and Benjamin), 44 1 0 , 44 l f , Ex g 2 0 ' 2 1 (the heedful and unheed-
ful), 19 1 9 .
Noun subject followed by pronoun subject: Ge 27 1 1 , Ex 1411*,

211*, De l 3 9 "*, 3 9 , De 4 3 "* (both preposed topics are resumed).
In Ex 3 2 the same subject is repeated for purposes of contrast.
wehinne hassene b'er b1

we- hassene ênenn 1 ukkl,
and behold, the bush is ablaze with fire
and (but) the bush is not being eaten.
11.4. CONTRAST WITH OBJECTS
Example: ^tl hesib 'al-kannl

vtPt tl,
Me he returned to my post
and him he hanged (Ge 4 1 1 )
Contrast Ge 12 1 2 , where similar material is arranged chiasti-

cally, blending the two clauses into a single picture rather
than setting them against each other.
Two noun objects contrast in Ge 42 3 3 , Ex l 2 2 , 16 23 (three
objects), 1 8 2 2 a > 2 6 b , 21311 (compare 21 36 ), 293 9 (repeated in
29 1 to attach more details).
The contrasting rules for firstborn domestic animals, asses
and humans are coordinated in Ex 13 1 2 " 1 3 . De 3 1 2 " 1 3 (contrast-
ing allocation of tribal territory; in De 3 1 5 " 1 6 the recipients
are in contrastlve positions).
11.5. OTHER ITEMS IN CONTRAST
Very little contrast is involved in the listing of times.
Example: ben hfarbayim t^kel bsr,

babbqer tiSbe^ lhem,
In the evening you will eat flesh
and in the morning you will be sated with bread (Ex 1612)
Compare Ex 16 1 3 and the even more sophisticated construction

in Ex 16 6t> - 7a .
11.6. EXCEPTIONS
Contrast is not necessarily involved when the same item is re-
peated in preverbal position in two successive coordinated
clauses.
Example: ^ êred 'ijnmek misraym

ve^nSkl ââlk gam-^Sl,
I will go down with you to Egypt
and I will bring you up (Ge 46"*)
CONTRASTIVE SENTENCES 153
Compare De 3 2 8 , 9 3 . Here the focussed pronoun is exclusive and

emphatic.
Another example: kl beyd hazq yesallehem

beyd hazq yegrsem meârs,
For with a strong hand he will expel them
and with a strong hand he will drive them out from
his land (Ex 6 1 )
Correlative items like food and drink do not contrast in pre-

verbal position (De 2 28 , 9 9 1 8 ) . Similarly, heaven and earth
don't contrast in De 4 3 . Compare De 81*.
11.7. CONTRAST SENTENCE WITH ASYNDETON
A striking example is Ge 502 0 .
vtP attem hsabtem ^lay r^

êlhlm Ijsbh letb,
And you meant (it) against me (for) evil
[But] God meant it for good
Knig (490) thinks the asyndeton shows the complete disparity

between the two modes of operation, and regards the 'and' or
'but1 of the versions as secondary. But there is no evidence
that this surface structure realizes a high degree of antithe-
sis. Early Hebrew did not have a special antithetical conjunc-
tion. See Chapter 14.
12
INCLUSIVE SENTENCES
12.0. INCLUSION AND ADDITION
The Hebrew and English Lexicon of Brown, Driver, and Briggs

calls 1 ap a "conjunction denoting addition" and gam an "ad-
verb denoting addition." As we shall see, these two 'par-
ticles' are largely similar in their syntax. If anything,
gam is a conjunction used secondarily as an 'adverb', while
'ap is an 'adverb' used secondarily as a conjunction.
The primary function of gam is inclusive coordination, and
in this âp resembles it to some extent. In many of its oc-
currences gam functions as a coordinating conjunction mean-
ing and, and there is a considerable range of usage over which
gam and we- are interchangeable without perceptible difference
of meaning. But there are constructions containing we- which
do not accept gam as a substitute. Thus gam can be used instead
of we- in circumstantial, adjunctive, surprise, conjunctive and
chiastic sentences (#12.4). But, unlike we-, gam is not used to
link clauses in disjunctive, contrastive, exclusive, or anti-^
thetical sentences. Or, rather, if gam were substituted for we-
in such sentences, it would quite neutralize their adversative
character. Furthermore, when gam is substituted for we- as the
coordinating conjunction in circumstantial, adjunctive, surprise,
conjunctive, or chiastic sentences, it is no longer possible
to regard such constructions as alternative surface realizations
of deep inter-clausal relationships that are adversative in char-
acter. Gam is the synthetic or inclusive coordinator par excell-
ence, and when it is replaced by we-, which is not always poss-
ible, it loses some of its effect of adding and not just joining.
When two clauses are linked by gam in this distinctive way, we
have a construction which we call an INCLUSIVE SENTENCE (IncSe).
12.1. PHRASE-LEVEL COORDINATION USING GAM
That gam is primarily a conjunction is shown by its use to co-

ordinate two nouns in a phrase (Ge 7 3 , 2011) . ^ ap is not so used.
INCLUSIVE SENTENCES 155
12.1.1. Inclusive Phrases
The phrase-level coordination gam-X gam-Y means both X and Y.

Example: gam-^nahn gam-^bten, both we and our ancestors
(Ge 46 3 ") .
Other examples: Ge 242 5 , 32 2 0 , 43 8 , 44 1 6 , 4 7 3 ' 1 9 , SO 9 , Ex 4 1 0 ,

5 l ", 1231 , 12 32 , 18 18 , Nu 18 3 , De 32 2 5 , Jdg 8 22 , 191 9 , ISa 2 2 ,
12 ,2 5 , 173 6 , 202 7 , 219 , 251 6 , 286 1 5 , 2Sa 3 1 7 , 5 2 , 162 3 ,
IKi 3 1 3 ' 2 8 .
The first gam is almost never lacking when the phrase is con-
tinuous and NEVER replaced by we-. The following gam, however,
can be replaced by we-, although this is rare.
Examples: Ge 14 1 6 , Jdg 10 9 , 2Ki 17"
12.1.2. Compound Conjunction
In a rare variant of this phrase type, X wegam-Y is used.

I ^ ___
Examples: Ge 14 7 , De l 2 8 (, Y wegam-Z). But gam-X wegam-Y is
not found as a single continuous phrase.
If X we-Y, X gam-Y, and X wegam-Y are all X <Co> Y, this ren-
ders void Dik's "criterion for coordinators"1 as a language uni-
versal. The translation and also begs the question and explains
why gam has been called simply an 'adverb.' The double coordina-
tion analysed in #12.2 explains how two coordinators can function
together each in its own way. In the light of the discussion ther
the phrase-level use of vegan as one conjunction is abnormal, on
a par with the 'but however' that English purists object to.
12.1.3. Duals
The dual inclusion in gam-X gam-Y explains gam-senehem in Ge 27"5.

This cannot mean, 'why should I be bereaved of the two of them
ALSO?' because there has been no previous bereavement. We suggest
that it substitutes^for *gam-arqb gam-êsw. So in De 222
*gam-hÎs g a m - h ^ i s s = g a m - s e n e h e m (no one else is to die). Com-
pare De 2 3 , ISa 2 5 " 3 . Similarly Ex 4 9 means if they do not be-
lieve both signs (there were only two).
12.2. TRANS-SENTENCE INCLUSIVE PHRASES
An inclusive phrase of the form X gam-Y or gam-X gam-Y can be

realized discontinuously with (gam-)X in one clause and gam-Y in
another clause. Gam thus functions as a long range inter-clause
conjunction which does not link the two clauses as wholes, but
only single elements in each. X and Y have their own functions
in their respective clauses. This means that both simultaneously
realize two tagmemes. This can be avoided by replacing wegam-Y
156 INCLUSIVE SENTENCES
by w?-Y garn-h"* or the like. Y then realizes the clause-level

tagmeme, while htO functions in the inter-clause inclusive
phrase X...gam-hiO. Gam functions independently of, and side
by side with, any other conjunction which links the two clau-
ses together as wholes. The two clauses in which (gam-)X and
gam-Y occur need not be contiguous, and need not be joined to-
gether at all, except in so far as gam joins one part (X) of
one to one part (Y) of the other. Gam cannot be replaced by
we- when it is used in this way.
Example: wayeaw êt-hrl 1 son . . .

way?av gam et-hasisenl. . . ,
and he commanded the first...
and he commanded the second also... (Ge 3 2 1 8 - 2 0 )
The phrase et-hri 1 sn. . . gam 1 et-hasseni spans the interval

between two widely separated clauses. The clauses as such have
their own conjunctions.
Example: (we-) gam-Î? âl-yer^ bekol-hhr

garn-hasso"* wehabbqr âl-yirû "'el-ml hhr hah"* ,
both a man let not appear in all the\mountain
also the flocks and herds let not graze in front of
that mountain (Ex 34 3 )
The two clauses are in apposition. The second gam links hass^n
to ^; it does not link the second clause to the first. Gam-X
...gam-Y is an inclusive phrase linking man and flocks across
this sentence.
Gam coordinates pronouns in their free forms, not as suffixes.
*gm, he too, does not exist, only gam-h"1 .
When the item corresponding to Y is a pronoun affix, gam is
attached to the corresponding free-form pronoun in apposition.
Example: +VSt +S:(PrAx <A> gam-Pr)
wayya f a gam-h"1 mat âmmim

and he too prepared savoury (Ge 2731 )
Also Ge 22 2 ", 303 30 , 381 1 , Le 262>t (lap is used in the same

way in the preceding clause; compare De 2 1 1 2 0 ) , De 12 3 0 .
Example: +V + 0:(PrSx <A> gam-Pr)
brkenl gam- ^SnI,
Bless me, me too (Ge 27 3, *> 38 )
Example: +Pp +(PrSx <A> gam-p r )
^ lyw
Upon gam-h^
him, ,
him also (ISa 1 9 2 3 )
r 21
Contrast gam- leyh (Ge 26 ).
+N +(PrSx <A> gam-Pr)
beplw gam-h^,
in his mouth, his also (2Sa 17 5 )
Also IKi 21 l 9 .
Even when Y is a free form, gam is sometimes attached to a pro-
noun in apposition with it, rather than to Y itself.
Pr gam-Pr Ge 20
Pr g a m - N 2Sa 1710
gam-PrGe 411 . 2 2 , 2 6 , 10 2 1 , 19 3 ", 222 0 , Nu 4 2 \ Jdg l 2 2 ,
831
Note the unique gam-Pr <A>
gam-hem hartumme misrayim (Ex 7 1 1 )
12.3. DOUBLE COORDINATION
When an inclusive phrase (gam-)X... gam-Y is distributed between

two clauses, and gam-Y is clause-initial, and the second^clause
is coordinated, the second clause begins vegam-Y, Here vegam is
not a compound conjunction. Each conjunction operates indepen-
dently on a different level of the grammatical hierarchy, we- is
a sentence-level conjunction joining two clauses; gam is a phras
level conjunction joining Y to X in a phrase^that cuts across a
sentence. Furthermore, the clause to which we- links the ensuing
clause need not be the same as the clause containing X to which
gam links Y.
Example: wayyiSme^Tl h^m ^-al-hitnaddebm. .

vegam dwid hammelek lmalj iimh.5 gedl,
and the people rejoiced because they offered themselves
and David the king also rejoiced with great joy (lCh 29
Here we- joins the two clauses, while gam joins the two subjects
in an inclusive phrase.
Dik recognizes that a construction like this does not trans-
gress his criterion for coordinators. He says that the two con-
junctions in such a combination have different "scope". 2
12.4. INCLUSIVE COORDINATION AND SENTENCE TYPES
The calculus of sentence types in Hebrew makes provision for the

marking of succession or simultaneity in time sequences (tense-
aspect of verbs) in one dimension, and for the marking of simi-
larity or contrast in activities, in another dimension. The plac
ing of an inclusive phrase athwart any kind of sentence serves
as an additional and unequivocal mark of similarity. If the sen-
tence already expresses similarity, this is enhanced, as when

gam is used in a chiastic clause. If it implies similarity,
this is explicated. If it is equivocal, gam resolves the ambi-
guity in favour of similarity (circumstantial, alternative,
surprise). If the sentence is unmarked, gam marks it for sim-
ilarity (conjunctive sentence). If the sentence could imply con-
trast, gam cancels this (contrastive sentence). If the sentence
is marked for contrast, gam cannot be imposed on it (alternative,
exclusive, antithetical), as the following instance shows. In
comparing Ge 1 2 1 2 with 4 1 1 3 above (p. 68) it was pointed out
that the latter is contrastive. But the use of gam with a sim-
ilar sentence in Nu 22 3 3 can hardly overcome the antitheses in
the verbs.
gam-^tek hragti
we'th heheyeti,
I would have both slain you
and spared her
12.4.1. Inclusive Chiastic Sentences
It has been shown that a chiastic pattern between two coordin-

ated clauses links them very closely, and usually implies sim-
ilarity between the items in chiasmus. This similarity can be
placed beyond all doubt by the provision of an additional in-
clusive linkage between the items in chiasmus. Any two clauses
which are neither wholly identical nor totally antithetical are
likely to be partly similar, partly unlike. A contrastive sen-
tence plays up contrast; a chiastic sentence plays up similar-
ity. But a chiastic sentence can still be open to antithetical
interpretation. The chiastic sentence in Ge 43'b-'*a reports the
similar actions of Cain and Abel. But the gifts contrast, and
if the focus is on this contrast, the sentence is antithetical
--Cain brought vegetables BUT Abel brought animals. This inter-
pretation is banned by the use of gam to link Cain and Abel.
They both brought gifts.
Other examples: Ge 4 2 2 (Adah's and Zillah's actions similar),

4 2 6 (Seth like Adam), 1 0 2 1 (Shem like Japheth) , 1 4 1 6 (double
rescue), 1 9 3 8 (both sisters alike), De 7* (but nothing else is
sent--this is an extra), Jdg 8 3 1 .
Ge 5 0 2 3 has a chiastic clause in apposition; the expected v5gam

is not used.
12.4.2. Inclusive Conjunctive Sentences
While a conjunctive sentence is unmarked for contrast or simi-

larity, some similarity is generally implied. This can be re-
inforced by means of gam.
Example: we"1 Is l^-ya/le fimmk

wegam-Îs "'al-yer1 bekol-hhr,
and a man won't go up with you
and also a man mustn't appear in all the mountain (Ex 343)
gam would go better with the verb. "Nobody must either ascend
or appear."
Example: gam ^S ta^Se

wegain yk8l tkl,
you will BOTH do
and ALSO succeed (ISa 26 2 5 )
The use of gam... vegam... with the same pronoun subject in two
successive clauses bespeaks a conjunctive sentence which, as a
whole, is linked inclusively with the preceding text.
Example: gam-h"1 yihye-lle^m

wegam-h"* yigdl,
He too (i.e., Manasseh as well as Ephraim) will become
a people,
and he too will be great (Ge 4819)
12.4.3. Inclusive 'Contrast' Sentences
If Rebekah had said to Abraham's servant
*âtt sete
v e l i g m a l l e y k a êsP a b ,
you drink
and for your camels I will draw,
the two items in pre-verbal position would have been brought

into prominence and into contrast (Chapter 11).
(You look after yourself

and I'll look after the camels).
What she actually said was
gam-âtt sete
vegan ligmalleyk êsPb,
both you drink
and also for the camels I will draw (Ge 241*'*)
The imposition of gam... wegam... on the structure of a contrast-

ive sentence neutralizes the contrast and brings everything to-
gether.
In Ex 21 28 the rules for a goring ox and its owner are stated .
in chiasmus.
s q l yissq.el h a s s r . . .
u b a 'al h a s s r n q l ,
The ox will be stoned
and the ox's owner is not culpable
A l t h o u g h the fates c o n t r a s t , the c h i a s t i c s e n t e n c e plays d o w n

the a n t i t h e s i s a n d t r a n s l a t i o n 'but' is not indicated. That
w o u l d be done in a c o n t r a s t i v e sentence.
*hass5r y i s s q e l
b a ^ a l hassr nql
W h e n the owner is r e s p o n s i b l e b o t h are liable to the d e a t h p e n -

alty.
hassSr y i s s q e l >
wegam-be^lyw ymat,
The ox will be stoned
and his owners also will die (Ex 2125)
T h e a d d i t i o n of gam links 'owner' a n d 'ox' a n d draws a t t e n t i o n

to the s i m i l a r i t y of their fates. It thus r e v e r s e s the effect
of placing 'ox' a n d 'owner' in j u x t a p o s i t i o n in a regular con-
trast sentence. W i t h o u t gam, Ex 2 1 2 S w o u l d be contrasting the
d i f f e r e n t m o d e s of e x e c u t i o n .
In a regular c o n t r a s t s e n t e n c e h'1 ^mar... wehi 1 ^mer...
w o u l d be a n t i t h e t i c a l - - ' H e said this, but she said that.' The
u s e of gam w o u l d o b v i a t e the a n t i t h e s i s , as in Ge 2 0 s .
hal"1 h"1 ^mar-ll h5t hi"!

wehi^-gam-hl^ ^mer ^ hl h^
Didn't he say to me, "She's my sister"
and [didn't] she also say, "He's my brother"?
Other e x a m p l e s : Ge 13 2 w i t h 5 (The p a r t i c i p a n t s , Abram a n d Lot,

are in c l a u s e - i n i t i a l c o n t r a s t p o s i t i o n . [Abraham h a d silver
a n d gold, but Lot h a d tents.] But gam blocks this and brings out
the fact that they w e r e c o m p a r a b l y w e a l t h y , w h e r e a s Ge 13 1 ,
w h i c h lacks gam, contrasts A b r a m ' s a n d Lot's dwelling p l a c e s . ) ,
De 2 6 p k e l . . . w e g a m m a y i m ) .
This d i f f e r e n c e b e t w e e n a n inclusive s e n t e n c e a n d a c o n t r a s t -
ive s e n t e n c e (Chapter 11) is n i c e l y i l l u s t r a t e d by Ge 4 4 9 " 1 0 .
T h e b r o t h e r s insist o n their s o l i d a r i t y , u s i n g a n inclusive
sentence--the one of us it is found with will die and we also
(gam-"1 anahnu) will be slaves to you. But J o s e p h ' s slave re-
p l i e s w i t h a c o n t r a s t i v e s e n t e n c e - - t h e one it is found wih
will be my slave but you (weâttem) will be clear. In the
former gam u n i t e s d e a t h a n d slavery. In the s e c o n d the use of
gam w o u l d be i n c o m p a t i b l e w i t h the d i f f e r e n c e b e t w e e n slavery
a n d freedom.
Other examples: Ge 2 1 1 2 " 1 3 (links Isaac a n d Ishmael), Ge 241"'

''t.ite.'te (Abraham's slave a n d the camels).
12.4.4. Circumstantial Clauses
E x a m p l e : Ex 1 8 2 3 ; the safe arrival of the p e o p l e is c o n c o m i -

tant w i t h M o s e s ' survival; but the a d d i t i o n of gam implies
that M o s e s also will a r r i v e in p e a c e .
INCLUSIVE SENTENCES l6i
Another example: Ex 1 9 2 2 .
12.4.5. Surprise clauses
Surprise clauses ( C h a p t e r 7 ) are fairly detached. But gam-Y

in a surprise clause can tie it into context.
Example: I never expected to see your face again, and behold

[God has not only let m e see your face], God has let me see
your seed as well (gam êt-zarêk) (Ge 4 8 1 1 )
Other examples of gam with hinne: Ge 422228.
12.4.6. Paragraph-Level Inclusive Linkage
When X and Y are explicit in WP clauses, X... gam-Y is para-

graph-level .
Example: wattasqeyn êt-âblhen yayin ballayl h^...

wattasqeyn gam ballayl hah^ êt-'ablhen yyin, .
and they made their father drink wine that night...
and they made that night also their father drink
Wine (Ge 19 3 3 and 3i
)
gam links the two similar nights. Note the chiasmus.
Other examples: Ge 22 2 " (the concubine like the wife), 26 2 1

(They had already quarreled over one well; now they dug an-
other and they quarreled over it TOO.), 2731 (Rebekah had
already done the same--verse 14), 2 9 3 0 a A (as he had already
done to Leah. The following clause, which has another gam,
is c o n t r a d i c t o r y . He loved also Rachel rather than Leah [who
is called hated in the sequel]. But he had not previously
loved someone else more than Leah, gam suggests he loved Ra-
chel as well as Leah; hence its deletion in the versions.),
29 3 3 (she already had one son), 3 2 2 0 (as he has already in-
structed the first), 33 7 (the concubines had already approach-
ed), 3 8 1 0 (as he had already killed his brother), Ex and
11
" (Pharoah, like Yahweh, called his wise men and they, like
Moses and Aaron, performed tricks), Nu 11" (The Israelites
are like the camp followers.), De 3 3 (Og as well as Sihon),
De 919 and 1010 (on that occasion too, as previously), 920
(for Aaron as for the people linking wêtnappal in verse 18 .
with wêtpallel in verse 20).
12.4.7. Predictive Discourse
When X and Y are explicit in WS clauses X... gam-Y is a para-

graph-level linkage.
Examples: Ge 3 2 2 (he has already taken from one tree), 2 9 2 7

(Laban has already given him Leah, referred to by the preced-
ing z U ) , 3 0 1 5 (gam links to object of an identical verb),
\()Z INCLUSIVE SENTENCES
4 4 2 9 Can irony of the author; Jacob speaks as if he knows

that they have already removed Joseph), Nu 16 J 0 (the priest-
hood on top of other priveleges), 27 (like Aaron, the com-
parison follows), De 3 2 0 (like the eastern tribes).
12.4.8. Other Constructions
X and Y explicit in miscellaneous constructions.
Examples: Ge 19 3 * 0 ernes ...gam-hallayl), 24 1 9 (she had al-

ready looked after the man), 38 1 1 (his brother had already
died), Nu 12 2 (bemolie...gam-bn), De 12 3 0 (haggylm h 1 eile
...gam-^ nl) .
X implicit, with gam-Y in WP clause.
Examples: Ge 3 s (She had previously given some to herself.),

50 1 8 (This suggests that the deportation was in two stages.
If so, the first stage is now lost.)
X implicit before gam-Y in miscellaneous clauses.
Examples: Ge 19 2 1 (second boon like the first), 20 6 (I know

it just as well as you do; compare 2Ki 2 3 ' 5 ) , 30 3 0 (as well
as Leh), 3 5 1 7 (this time also; the previous time it was al-
so a son--Joseph), 4 8 1 9 (Manasseh as well as Ephraim), Ex 3 3 1 7
(second request to be granted like the first; compare Ge 1 9 2 1 ) ,
Nu 4 2 2 (like the other Levites already mentioned), 18 2 8 (Le-
vites to tithe like the others), 24 1 2 (the messengers as well
as you), De l 3 7 (twice)(Moses like the Israelites), 12 3 1
(they sacrifice animals [implied], they sacrifice their chil-
dren as well).
12.4.9. Inclusive Complex Sentences
An inclusive phrase can link together the two clauses in a

complex sentence.
Example: îm-ykal Îs limnt êt-âpar h^res gam-zarâk

y i m m n e , If anyone can count the dust of the earth, your seed
also will be counted (Ge 1 3 1 6 ) . g a m c o n n e c t s seed w i t h dust.
Other examples: Jdg 221 (j too will not keep my side of the
covenant.).
12.4.10 ^ a p Equivalent to gam
The use of âp as an inclusive conjunction indistinguishable

in meaning from gam is ullustrated by Ge 40 1 6 (x too, as well
as the butler), De 2 1 1 (they also, like the Anaqim), 2 2 0 (that
is, the land of Ammon, just like the land of Moab was consid-
ered a land of Eephaîm), De 15 1 7 (male and female slave
alike), Jdg 5 2 9 (as well as her ladies).
12.5. INCLUSIVE COORDINATION AND NEGATION
In negated clauses, (gam-)X.... gam-Y means neither X nor Y,

whether in a simple phrase or in a sentence-spanning phrase.
Example: gam-ll gam-lk l"1 yihye, it will be neither mine nor

yours (IKi 3 26 ) .
Rxample: gam-qb l^ tiqgobenn

gam-brek l^ tebrkenn,
Seither curse him
Nor bless him (Nu 2 3 2 5 )
Compare Is 48.
Example: vel5^-st libb gam-lzH, and he paid no thought to

this one EITHER (just as he had ignored the previous disaster)
(Ex 7 2 3 ; 8 29 is similar).
Example: wegam-âtt -higgadt II

wegam ^nkl l'1 sma 'tl biltl hayym,
(and) neither you told me
(and) nor I heard except today (Ge 21 2 e )
12.6. THE IMPLICATION OF INCLUSIVE COORDINATION
In a well-formed inclusive sentence the phrase (gam-)X... gam-Y

lies across the sentence. Attention is drawn to the similarity
of X's and Y's involvement. If X and Y are subjects they do the
same thing. If objects, they have the same thing done to them.
If locations, the same thing happens at both of them. And so
on. Even without the optional preliminary gam to mark it, X can
usually be identified once gam-Y comes up, even if there is
quite an interval between them.
Sometimes, however, X is quite lacking in the text preceding
gam-Y. The implication of gam-Y is that there has to be a sim-
ilar X in there somewhere, either implicit in the text, or pre-
supposed in the situation.
Example: wayyitten gam-liPlsh, and she gave it to her husband

also^(Ge 3 s ). There is no preceding indirect object to match
leÎsh, but it is implied that the woman is that indirect ob-
j ect.
Example: vensap gam-h^ ( al-snePen, and he too will join

our enemies (Ex l T o ) . This implies that someone else has al-
ready joined the enemy and now Israel will do so TOO. There is
nothing explicit or implicit in the preceding text about a
gathering of enemies, but use of gam makes that the presuppo-
sition of the incompletely formed inclusive phrase. There is a
similar effect in Ge 1 3 1 S , 1 6 1 3 , 19 2 1 .
Inclusive linkage by means of gam has its most powerful effect

when X and Y are both explicit with similar grammatical func-
tions, in contiguous clauses which are constituents of the

same sentence. This happens in well-formed chiastic, con-
junctive and constrastive sentences.
12.7. NONINCLUSIVE USES OF gam
There are instances when gam-Y does not link Y to any sim-
ilar X in the context, and when, moreover, it is not even
possible that such an X is implied or presupposed. Here gam
cannot mean "also" or "too."
12.7.1. Coordination
The use of gam as a coordinating conjunction virtually in-

distinguishable from we- is seen in its use as coordinator
in a variety of sentence types.
In inter-clause constructions, gam can be used in a cir-
cumstantial clause. This is rare.
Example: gam hÎs mSse gdl i P o d , while the man Moses

(was) very great (Ex ll3).
Or in an adjunctive clause--very rare.
Example: gam ^nkl hlll 111, as for me, a curse be on

me... (1 Sa 12 2 3 ).
In a surprise clause (for vehinne).
Example: gam hinne âbdek ya^qb âhren, behold, your

slave Jacob (is) behind us (Ge 32 2 1 , which seems to be gar-
bled from vehinne gam-h' 'ahSren in 321 9 ).
In a conjunctive sentence.
Example: êres r^s

gam-smayim ntp
gam-^blm nêp mayim,
The earth convulsed
and the sky collapsed
and the clouds dropped down as water (Jdg 51*)
In a chiastic sentence.
Example: vayyakkm lepl-hereb...

gam kol-he^rlm...silleh bês,
and they smote them to the mouth of the sword...
and all the cities... they sent into fire (Jdg 201*8)
Other examples: Ex 2 ", ISa 28 l s .

gam-âtt is occasionally used instead of wiâtt in high-level

discourse transitions.
Examples: Ge 4 4 1 0 , ISa 12 1 6 .
Similarly in ISa 2 1 5 gam links two paragraphs dealing with sim-

ilar abuses.
Gam can be placed between two clauses in apposition, as with
v5- also (#8.10.2).
Examples: De 23 3 ',, (specifying apposition, 286 1 , Jos 2 4 l e b

(makes an inclusio with the opening clause), ISa 28 2 0 , 2Sa
12 2 7 .
In Ge 30 s , gam is used as a post-positive inter-clause coordin-

ating conjunction.3
12.7.2. Compound Inter-Clause Conjunction
Just as in #12.1.2 it was shown that vegan was equivalent to a

single phrase-level conjunction, so, occasionally, this combin-
ation may be used to unite two clauses when there is no indica-
tion that gam is operating independently of we- to unite two
items in these clauses in a trans-sentence phrase. At least in
Ex 6 3 " 5 , where three clauses as wholes are coordinated by two
occurences of vegan, gam-^Snl, I too, should not be construed,
as the Massoretes were aware. For this would mean that someone
else besides God had heard their cry, which is not the case. In
each instance gam could go with the verb: I revealed myself...
and I also raised up my covenant... and I have also heard... But
in the light of the observations made in #12.8 below, it is
possible that here gam ânl means myself.
Other examples of Cli wggam-Cl2 are found in Ge IS 1 ", 1 7 l e ,
2 0 1 2 , 21 1 3 . For the construction gam-Clj wSgam-Cl 2 see
Ex 1 0 2 5 " 2 6 .
12.8. GAM NOT A CONJUNCTION
In #12.7.1 above gam is called a conjunction because it can be

replaced by v8-. When this is not possible, it might be neces-
sary to call gam an 'adverb', even though the name is hardly
suitable, because gam so used modifies other things besides
verbs.
12.8.1. Appositive gam
The use of v e g a n in Ge 15Ilf is not inclusive. There is no sim-

ilar preceding element to link êt-haggy with. Compare Ge 38 2 2
Ge 17' 6 is even more remarkable, for here v e g a n links two verbs
which are partly synonymous, and, furthermore, w?gam ntattl is
future (=WS). Compare Ge 30 where v?gam ima f beqll specifies

the meaning of dnannl êlhim, God has judged me and also he
heard my voice, a relationship normally realized in apposition.
12.8.2. Emphasizing gam"1
Gam is sometimes placed before an infinitive absolute after a

cognate verb, a thing which does not happen with w5-.
Examples: Ge 3 1 1 5 , 461*, Nu 1 6 1 3 , ISa l 6 . (Does this explain

the strange gm-brvx in Ge 2 7 3 3 ? )
1
The similarity of gam and a p as emphasizers is seen by compar-
1 1
ing g a m - ^ o m n , truly Indeed (Ge 20 ) with ap omnam, really
13
and truly (Ge 1 8 ) . That the particles are assertative is
seen by comparing haâp tispe, will you really destroy? (Ge 18
23 2lf
> ) with hSki in Ge 2 7 3 6 - - J s n ' t he truly called Jacob?
12.8.3. Focussing gam
A construction like milk gam-hl"1 , probably means Milkah her-

self, not Milkah she also, for there is no preceding partici-
pant for gam to link Milkah to (Ge 2 2 2 0 ) . Similarly, gam in
IKi 4 1 5 is not inclusive, for no one else had married her.
Other examples: Ge 20, 3 2 1 9 , Ex 1 2 3 2 b , Nu 2 2 1 9 , Jos 9*,

Jdg 9 " 9 , ISa 22 7 , 2Sa 1 2 1 3 ' 1 * .
Similarly, 1 ap-h^ means he himself in 2Ki 2 11 *, and ^ ap has

the same effect in its several occurrences in Le 2 6 1 6 " .
12.9. THE HIERARCHICAL SIGNIFICANCE OF GAM
Gam does not have a single function, but a range of functions

with associated meanings. Its various uses are not simply mis-
cellaneous, each with its own contrastive-distinctive features.
There are priorities. Its 'typical' function has first claim,
and only when that is eliminated as not applicable are less
likely functions tried out.
Gam-Y immediately suggests a phrase, X gam-Y, and if this is
found, especially in its form gam-X gam-Y (#12.1.1), the role
of gam as phrase-level inclusive coordinater is recognized.
Otherwise, one looks for X (or remembers it) explicitly in the
preceding context, or, failing that, implicitly, and gam is
recognized as a clause-spanning phrase-level inclusive conjunc-
tion. If that fails, one tries out gam as a conjunction coor-
dinating two clauses in a sentence (#12.7). Finally, when all
else fails, and no coordination is found, gam is interpreted
as an emphatic particle (#12.8.2), or 'adverb.'
NOTES
1
Coordination..., p.
2
Coordination. .. , p. 10. Dik's criterion is not flexible
enough, for it absolutizes the category of a given form by mak-
ing provision only for its classification as either a coordin-
ator or a subordinator, like the old lexicons. But a conjunc-
tion might vary in its role from text to text, as the frequent
examples of alternative surface realizations in the present
study show. In the case of wegam we must distinguish a com-
pound coordinator (#12.1.2, #12.3) from double coordination on
different levels (#12.2) and also from a residue (#12.8.2) in
which gam is a 'clause adverbal' or emphasizer.
3
F. I. Andersen, Journal of Biblical Literature 88 ( 1 9 6 9 ) :
p. 200.
" B. Jacob "Erklrung einiger Hiob-Stellen," Zeitschrift fr
die AlttestamentIiche Wissenschaft, 32 (1912) :pp.279-282 . The
long note on pages 281-282 contains valuable observations on
the similarity between gam and âp in this regard, and com-
plains that such particles have not been adequately studied.
Still true sixtyyears later.
13
EXCLUSIVE SENTENCES
13.0. SIGNALS OF EXCLUSIVE RELATIONSHIPS
A collation of 'particles' that have been called exclusive,

restrictive, or limitative conjunctions by one writer or
another yields the formidable list already given on page 70.
The items on this list, which includes some compounds, are
by no means all the same in their grammatical behaviour.
The realization of relationships of exclusion is a com-
plicated business in which the grammatical functions of con-
junctions, prepositions, and 'adverbs' are brought into play,
as the faltering treatment of this subject in the existing
Hebrew grammars and lexicons shows. Either the issue is ob-
fuscated by sweeping all such items under the elusive 'part
of speech' PARTICLE, or else the assignment of an item to
one part of speech or another is made so arbitrarily that
much disagreement results. Nothing is gained by berating
previous workers. The ensuing treatment will develop criteria
for distinguishing conjunctions, prepositions, and 'adverbs'
from one another by finding out the hierarchical level on
which each functions. In brief, a conjunction, by definition,
joins things, so its minimum manifestation is in a three-
item construction X-Cj-Y which is a sentence if X and Y are
clauses or a phrase if X and Y are phrases or words. A pre-
position, as its name indicates, comes in front of something,
and this might seem at first to be just a trivial structural
fact. But the construction Pp-X, a prepositional phrase, is
a relator-axis, which, as a whole, functions in relationship
to something else, say Z. Since Cj-Y is also a relator-axis
construction, there is a formal resemblance between prepos-
itions and conjunctions. But a preposition is always followed
by a nominal (noun, pronoun, noun equivalent [the last con-
ceals a circularity in the the definitions]), whereas a con-
junction always joins two things of the same kind. These
differential criteria do not settle all questions; for in-
stance, they do not tell us whether "except" is a conjunction
or a preposition in "noone except me". But at least they per-
mit us to put our finger on the reasons for the structural
ambiguity in such cases. A n 'adverb' is not a joiner and not
EXCLUSIVE SENTENCES 169
a relator, but a modifier. The name is misleading, for while
an adverb like me 1 d, very, can modify a verb, it can also
modify an adjective -- tb mP5d, very good. Furthermore
some so-called 'adverbs' modify a clause as a whole, and
there is no reason to attach them specifically to the verb
in the clause. It is in this regard that 'adverbs' and con-
junctions are sometimes indistinguishable, or rather, that
it is not easy in some particular occurrences of a given
word to tell whether it is functioning as an adverb or as
a conjunction.
In De 12 1 raq haddm l^ tS^kelu, only the blood you
shall not eat, begins with a conjunction if raq relates
the following clause as a whole to the preceding general
statement as a limitation to be imposed on it. you can
do anything you like, except that you mustn't eat blood.
But raq haddm, only the blood, could be a phrase (the
object of the verb) in which raq modifies haddm (haddm
is a noun, so raq should not be called an 'adverb') marking
it for exclusion from the list of permitted foods, YOU
can eat anything except (=only not) blood, where anything
except blood is a phrase in which except is a conjunction,
an excluding coordinator. If, finally, raq were an adverb
in the strict sense, it would modify the verb and so ex-
clude eat from the list of permitted activities, rou can
do what you like with blood, except eat it. The phrase-
level only blood is easier to distinguish from the clause-
level only eat and from the clause-modifying (sentence-
level) only-*(eat blood) than the last two are from each
other. Since the verb is the hub of the clause, a line
cannot always be drawn between a verb modifier ('adverb')
and a clause relator (conjunction).
The excluding relator lebad is strictly a preposition,
and so is the compound lebad min. 1 So lebad me^ser should
not be called a conjunction, as is usually done, for this
needlessly multiplies categories and flaunts the settled
function of âser as a nominalizer. Zltl except is usually
a preposition, but it is used once (Jos l l 1 3 ) as a con-
junction. Some of the allegedly exclusive conjunctions (or
'adverbs') are negatives, etymologically nouns, whose re-
semblance to adverbs or conjunctions is not only derivative
in the historical sense, but secondary from the functional
point of view, and not easy to disentangle from matters of
translation. They are êpes, biltl, bil^de. The deep-
structure relationships of negation, exclusion, and antithe-
sis shade into each other, and so their surface realizations
overlap to a considerable extent. Some conjunctions primarily
antithetical (Chapter 14) have a secondary use in alternative
realizations of exclusive relationships (#13.7). These are
^ im l1 and kl h a . This leaves âk 2 and raq as the exclusive-
restrictive 'particles' par excellence. These conversely have
secondary usage in alternative surface realizations of deep
antithetical relationships (#13.5). The reasons for present-
ing the material in this way have been explained on page 35.
170 EXCLUSIVE SENTENCES
^ ak kl is sometimes listed as an exclusive (compound) con-

junction. But the only occurrence is in ISa 8 where "*ak kl
is not a compound. Each word has its own function; âk is
exclusive, ki is assertative -- o.-ily you will certainly
warn them.
13.1. THE FORM OF THE EXCLUSIVE RELATIONSHIP
An exclusive or restrictive connection is a kind of coor-

dination between two items, one of which represents a whole,
the other a part which is excluded from that whole. Thus the
whole set of a man's sisters comprises three subsets.
(i) Full (same father and same mother)

(ii) Paternal (same father, different mother)
(iii) Uterine (same mother, different father)
In Hebrew nomenclature a member of sets (i) and (ii) is

(iv) bat-^bi, my father's daughter. A member of sets
(i) and (iii) is (v) bat-"1 imrni, my mother's daughter.
Members of set (i) are the overlap of sets (iv) and (v).
Members of set (ii} consist of set (iv) minus set (v) --
bat-^bi... âk l' bat-îmml, my father's daughter... but
not my mother's daughter (Ge 20 l 5 ). This construction is a
phrase-level exclusive coordination.
In a similar manner a clause which makes a blanket state-
ment can be subsequently limited by excluding some of what
it has comprehensively embraced.
Example: wayyiqen ysep êt-kol-âdmat misrayim lepar^...

raq âdmat hakkhanlm l^ qn
and Joseph bought all the land of Egypt for Pharoah...

except he did not buy the priests' land (Ge 47* )
There are three features to note in this fully formed ex-

clusive construction, (i) The first clause makes a compre-
hensive statement of which the explicit mark is kol all.
This feature is, however, often implicit; the land of Egypt
in the absence of any limitation, implies ALL the land of
Egypt. Furthermore, as soon as a limitation is applied, this
first statement is shown to be not strictly true if it is
left standing alone, (ii) The following exclusive clause
negates part of the lead clause. The explicit mark is l1
not. Negation may also be achieved by means of an antonym,
(iii) The use of an exclusive conjunction (âk or raq) to
coordinate the excluded part with the comprehensive lead
statement. Antithetical conjunctions may alternatively sub-
stitute for these (#13.7).
Already in this example a problem emerges. As will be
shown in #13.9, raq-X can be a phrase meaning only x.
If this is what raq is doing in Ge 47 2 0 , it is not a con-
junction joining the following clause as a whole to the
preceding, exclusively. It modifies "âdrnat hakkhSnim, and
the object of the verb buy is the priests' land only. In

clause-initial position, therefore, when the immediately
following item qualifies for phrase-level modification,
the function of rag is ambiguous. (The behaviour of "only"
in English is equally ambivalent, and made worse by its
greater mobility.) It could be a conjunction Cexcept, a-
long with the negation) or an 'adverb' (only). That raq
is here a conjunction is suggested by the similar clause
in Ge 47 26 : raq âdmat hakkShanim lebaddm l^ hyet
lepar^, except the land of the priestsâlone did not
come to belong to Pharoah. The use of lebaddm as a
restrictive 'adverb' modifying priests excuses raq from
a similar role, and virtually drives it into the camp of
antithetical conjunctions -- but. A similar use of "'ak
with lebadd in Ex 12 16 shows that there ">ak is an ex-
clusive conjunction. Raq and "'ak are interchangeable in
such constructions.
kol-mel^k l-y ^se bhem

^ ak ^ aser ye^kel leiol-nepes h1 lebadd yerSe lkem
all work will not be done among you

except what is eaten for everyone, that alone will
be done for you
The grammatical question is then whether raq or âk modifies

the verb as nucleus of the whole clause (a so-called 'clause-
adverb') and so governs the clause as a whole, which is tant-
amount to being a conjunction, since by modifying the clause
as a whole in this way it gives it an exclusive function in
the context, or whether it modifies merely one of the clause-
level elements, a noun or prepositional phrase, and so is
a more deeply embedded part of the clause. When raq or âk
precedes a noun or prepositional phrase and is not clause-
initial, it is certainly phrase-level (Ge 6 5 , 262). When
either of them is clause-initial and does not precede a noun
or a prepositional phrase that it might conceivably modify,
it is certainly an inter-clause conjunction (Ex 8 2 s ).
When it is clause-initial and also precedes a noun or a pre-
positional phrase that it might modify, it could be either
a conjunction or an 'adverb*. This structural ambiguity can
be resolved only by resort to semantic considerations, and
then only sometimes. Thus Ex 8 5 they will remain ONLY in the
Nile (phrase-level) is preferable to EXCEPT THAT they will
remain in the Nile (sentence-level). In Ge 24 BUT you won't
take my son back there (sentence-level) is preferable to
my son ALONE you won't take back there (phrase-level). But
raq in Ge 19 makes equal sense on either level, but is
better construed as antithetical (#13.5), since the prohib-
ition of abusing the men does not constitute an exception
to the permission to do what they like to the women.
The negation in the exclusive coordination construction
all x... excluding (only not) can be distributed different-
ly, using the surface features of antithesis. None of you
will enter the land that I promised to settle you in EXCEPT

(kl lim) Caleb... and Joshua... (Nu 1 4 3 0 ) . Compare Nu 26 6 5 .
With a prepositional phrase -- There is no expiation for
murder... EXCEPT (kl 'im) by the death of the one who com-
mitted it (Nu 3 5 3 3 ) .
13.2. PHRASE-LEVEL EXCLUSION
The preceding examples suggest that kl "'im is preferred

as the excluding conjunction after negation, but that raq
or âk is preferred when negation follows. Besides koi
everything, mePm anything can precede the exception.
M e P m kl ' i m - h a l l e h e m , anything EXCEPT the bread (Ge 396);
me^m ki H m - ^ t k i anything EXCEPT you (Ge 39). In Ge 14 2 *
negation is repeated with the excluding conjunction raq.:
I'll take nothing belonging to you EXCEPT (bil^dl r a q )
what the troops ate and the share of the men who accompanied
me.
We have seen that the exclusion of a subset Y from a set X
consisting of Y and can be stated in two ways, ONLY =
all EXCEPT or none of EXCEPT z. T o i d e n t i f y as not
ONLY is not exclusive, because the complete set is not
mentioned. But this antithetical construction has the same
surface form as the exclusive one, and this fluidity can
penetrate also into deep structure. Thus Nu 26 33 l^-hy
b n l m kl ^ i m - b n t , h e had no sons, BUT (only) daughters,
is just like the exclusive constructions cited above. But
h e r e sons (Y) a n d daughters (Z) c o m p r i s e t h e s e t children
(Z). Here the lack of a common gender in Hebrew places a
strain on the surface structure, for the use of the exclusive
construction means he had no sons except daughters, which
is a contradiction. But if the surface structure invades the
lexicon, it gives bnim in this context the generic meaning
(X) --he had no children except daughters.
13.3. TRANS-SENTENCE EXCLUSIVE PHRASES
In Chapter 12 there was considerable discussion of the spread-

ing of an inclusive phrase across two clauses which may be
otherwise joined together in a sentence. By analogy, the same
kind of construction might have been expected with exclusive
phrases. But this is rarely found. Unlike the double coordin-
ation realized by means of vegan, the combinations veraq and
weâk are almost totally unknown. Exclusion is secured either
within a continuous phrase or entirely between clauses as
wholes in a sentence. This constraint has doubtless something
to do with the component of negation in exclusive constructions,
but it would take more research into the deep semantic structure
of exclusion than we have room for here to lay this matter bare.
Example: vayyimalj et-kol-hayeqm. . .

vayyissêr âk-nh...
and he obliterated all that stood...
and survived ONLY Noah... (Ge 7 Z 3 " 2 1 f )
Since êpes has been called an "adverb of limitation"3 it

is interesting to notice how Nu 23 13 says You will see
nothing of him EXCEPT his edge.
êpes qseh tirê

ve-kull 1 tirê
only his edge you will see

and all of him you won't see.
13.4. EXCLUSIVE SENTENCES
In the commonest exclusive sentences the lead clause states

a general rule and the exclusive clause states a limiting
exception, with negation. The conjunction in these instances
is usually "'ak, and is equivalent to adversative however.
Example: I have given you everything/ HOWEVER Oak) flesh

in its life {its blood) you will not eat (Ge 93"l>).
Note, however, that at this point the speech changes from

declarative to what is virtually precative, and the use of
"'ak is comparable to its use, utterance-initial, in the
solemn premonitions studied in #13.10.
Other examples: Le 21 2 3 , 27 2 S , 2 8 , Nu l"",18 3 , 1 5 , 1 7 , 31 2 2 .
In Le ll 36 "'ak introduces the positive exception after a
negation; it can also introduce a positive qualification
after a positive statement (Nu 36, Jos 311). In Jos ll 13
zltl introduces the one positive exception to the pre-
ceding blanket negation (which is itself an exclusive clause
that excludes the act of burning from the general acts of
destruction just described, so that Jos ll* 3t is an exclus-
ion from an exclusion). That zltl is here a conjunction
is shown by the otherwise redundant use of l e b a d d h .
In contrast with âk as used in the above examples, raq
is not so obviously a conjunction when it is used in similar
constructions.
Example: W e r e n o t left Anakim in the land of the children

of Israel; r a q (but? however? except?) in Gaza, in Gath, and
in Ashdod they remained (Jos ll22).
But this is sufficiently explained as phrase-level limitation:

ONLY in Gaza, etc. See #13.9, and the discussion of the clause-
initial position in #13.1. The same applies to Ge 50*, where
clause-initial r a q t a p p m . . . follows a very comprehensive
statement that everybody went to Jacob's funeral, EXCEPT
( c o n j u n c t i o n ) or ONLY ('adverb') their infants... they left
in the land of Goshen.
13.5. EXCLUSIVE FORMS USED FOR ANTITHETICAL RELATIONSHIPS
The formal similarity between exclusive constructions and

antithetical constructions has already been remarked on sev-
eral times. In contrast with the preference for ak as

the exclusive conjunction in priestly writings of the Pent-
ateuch (#13.4), in the historical books and some prophets
âk signals an adversative relationship that falls short of
the outright opposition of full antithesis and is appropri-
ately translated however, nevertheless.
Examples: ISa 296 , 2Sa 2 1 0 , 3 1 3 , IKi 17 1 3 , 22"'* (In 2Ki 121*

raq replaces 1 ak in an identical construction.), 2Ki 12 1 *.
136 , 227 , 23 s , 26 , 35 , Is 141 5 , 4 3 2 \ Je 26 1 5 . 2 " , 28 7 , 30 l f ,
Ezk 46 1 7 , Jon 2 5 , Zee l 6 . Deuteronomy uses 'ak only once in
this way (De 18 2 ); once raq (De 10 l s ) once "Upes kl (De 15").
There is outright antithesis with âk in ISa 20 3 9 . See also

Nu 1 ", Ex 21 1 , Le ll* (= Del4 7 ).
13.6. EXCLUSIVE FORMS USED FOR COORDINATION OR APPOSITION
#13.10 contains numerous examples of utterance-initial 1 ak

which is assertative or an exclamation. When it comes be-
tween clauses that are in no way exclusive or antithetical
it may be described as simply coordinative (Nu 26 s 5 ) or
appositive (Nu 31 23 ). Also Ex 12 1 5 .
13.7. EXCLUSIVE RELATIONSHIPS REALIZED BY ANTITHETICAL FORMS
The antithetical conjunction kl 1 im except is used to mark

the exclusive condition that cancels a general rule.
Examples: Ge 32 2 7 , Le 21 2 , 226.
The variety of surface structures that can realize essent-

ially the same deep relationships is illustrated by the sev-
eral times that the exclusive condition for receiving aud-
ience with Joseph is stated.
Ge 42 1 S : im [= Neg . 1-tes?"1 mizze

kl îm-beb' ^hlkem haqqtn henn
You won't go out from here

EXCEPT when your youngest brother comes here
Ge 433 > 5 : l"1 tir"> pnay

bilti [= Neg.] ^ hikem îttekem
You won't see my face

EXCEPT your brother is with you
Ge 44 2 6 : kl-l^ nkal lir^St pene hîs

wePhin haqqtn ^ enenn îttn
For we cannot see the m a n ' s face

EXCEPT [literally a n d ] o u r youngest brother isn't with
In the last example the condition is simply coordinated.

In Ge 44 2 3 the condition is stated first.
îm-l^ yered âhlkem haqqtn îttekem

l^ tslpn lirôt pny
If your youngest brother does not come down with you

you won't see my face again
13.8. PHRASE 'ADVERB' AND CLAUSE 'ADVERB'
The combination raq-âk is listed as a "restrictive particle.""

Williams calls this "redundant."5 It occurs only in Nu 12 2 .
hraq-âk bemose dibber yhwh

hSl1-gam-bn dibber,
Did Yahweh only speak in Moses alone?
Did he not speak in us as well?
Here ^ ak bemose is a phrase of the type found in #13.9, and is

the correlate of phrase-level gam-bn. But raq is a clause-
modifier of the rest of the clause as a whole, as discussed in
#13.10. The two parts of the alleged compound thus function on
different levels of the hierarchy.
13.9. LIMITATIVE 'ADVERBS'
When âk and raq are used as modifiers of Y which they always
precede, and there is no trace of another item X from which Y
is to be excluded by the coordination X "âk/raq , X except Y,
then ^ ak/raq Y is a complete construction meaning only Y. To
call âkandraq 'adverbs' in such phrases is a misnomer, since
Y is not necessarily a verb.
s
Examples with raq: raqra^, nothing but evil (Ge 6 ), raq-t5b,
only good (Ge 262^, raq h a k k i s s e \ only the throne (=1)
110
(Ge 41 ), raq baye'r, they will remain only in the Nile
(Ex 8 5 7 ), Ex 9 2 6 (the land of Goshen was the only place where
there wasn't hail), 10 1 7 (raq not clause-initial), 102,(, De 2 3 5
(we took as spoil only the animals), 3 1 1 (the preceding ki
shows that raq is not a conjunction).
1
Examples with âk: ak-happa r am, just the once (Ge 1 8 3 2 , Ex 1 0 1 7 ,
J d g 6 3 9 , 1 6 2 8 ) , 3 4 1 5 ' 2 2 , E x 1 2 1 (on the very first day),
D e 1 6 1 5 ( n o t h i n g but joyous) , 2 8 2 ' ( n o t h i n g else but oppressed
and robbed), I S a 18 ( n o t h i n g less than the kingship), 21s
(merely from women), 2Sa 231" (with an infinitive, only to
strip the slain), 2Ki 18 20 (=Is 36 5 ) (mere words), Is 167 (ut-
11 1 15
terly stricken), 19 ( u t t e r l y foolish), 34 *' (even there,
not "yea, there. ."JTRSV]), 45 25 (only in Yahweh), Je 16 19 (noth-
ing but lies), 323 (definitely modifies a participle in mid-
clause position), Hos 122 ( n o t h i n g but vanity), Zeph l l e (could
be appositive, equivalent to gam), Ps 23 6 Cnothing but good-

ness and mercy', but usually interpreted as clause-level - <-
surely).
Example with bilti: Ge 21 2 6 (except today).
Examples with bil^dl: Ge 4 1 l e (only God). Compare Ge 41 **".
It is rarely that âk is adversative on phrase-level: ^ ak

l^ kol-hayymim, but^not forever (IKi ll 3 8 ).
In Je 10 âk bemispat probably means with nothing but
fairness.
The variety of means of securing this kind of modification
is illustrated by the several repetitions of Balaam's com-
mission from God. The word implies every word, only the word
or nothing but the word. Each of the six clauses is differ-
ent, but all have in common the suspended object and the re-
sumptive "*t. The option of marking the object with a 're-
strictive adverb' represents the point of greatest variabil-
ity in surface realization.
w(P epes -haddbr aser-^ adabber 'eleyk ''t tedabber

and no- the word that I speak to you it you will
thing but speak
(Nu 22 3 5 )
wePak êt-haddbr ^ser âdabber êleyk ^t taâSe
and only the word that I speak to you it you will
do
(Nu 22 2 0 )
hal'* et âser yslm yhwh bepi Is
t ^ esmr
is it not that puts Yahweh in my it I shall
mouth watch
ledabber
to speak
(Nu 23 1 2 )
haddbr âser ysim êlhim bepi ^t ^ adabber

the word that puts God in my it I shall
mouth speak
(Nu 22 3 8 )
kl âser-yedabber yhwh ^ t
all that speaks Yahweh it I shall
do
(Nu 23 2 6 )
âser-yedabber yhwh ^t âdabber
that speaks Yahweh it I shall
speak
(Nu 24 1 3 )
13.10. LIMITATIVE CLAUSE-MODIFIER
The exclusive particles raq and âk are inter-clause conjunc-

tions in an exclusive sentence when the second clause restricts
the scope of the first (#13.4). When this sentence relationship
is not realized, raq or âk may be a restrictive modifier of
the clause it is in.
Example: raq "'et-benl 1 tseb samm (Ge 24). Here raq "'et-
beni is not a phrase, as in #13.9. It does not mean: The only
person you won't take back there is my son. It m e a n s : The only
thing you mustn't do is take my son back there.
Example: raq ê^ber b e r a g l y , I JUST want to go through on

foot [that's all] (De 22').
Another example: De 4 9 .
In the case of âk, this function as a clause 'adverb' is the

ground for recognizing an assertative or emphatic meaning, es-
pecially when it is utterance-initial and the clause is preca-
tive.
Examples: Ge 23 1 3 , 26 9 , 2 7 1 3 3 \ 29 1 * (.surely), 34 2 3 (let's

28 13
just agree with them), 44 , Ex 3 1 (be sure to...; it can
hardly be restrictive), Le 2 3 2 7 (like Ex 3 1 1 3 ) , Nu 14 9 , De 1 2 2 2 ,
19
Jdg 3 * (not " o n l y " [RSV]), 7 (they had just set the guard
[compare Ge 2 7 3 0 ] ) , 1 0 1 5 , 2 0 3 9 , ISa l 2 3 , 1 2 2 0 2 " , 1 6 1 6 (surely/),
1 8 l 7 (just be a hero!), 2 5 2 1 , 2Ki 5 7 (Just realize...), IKi 2232
(rt m u s t be the King of Israeli), 2 K i 2 4 3 , Is 6 3 e ( c o m p a r e
Ge 2911*), Je 2 3 5 , 3 i 3 , 5", 1 0 1 9 , 12 1 , 3 4 \ Hos 4", 12 9 ("Ah"
[RSV]), Zeph 3 7 .
In one instance kl "".im seems to have the same effect: kl îm-

z e k a r t a n l , Just remember me [that's all] (Ge 40 1 1 1 ).
13.11. SUMMARY
The inclusive and exclusive 'particles' are ambivalent in gram-

matical function because they act differently on different lev-
els of the hierarchy.
(i) As inclusive or exclusive conjunctions.
Phrase-level XAY or X...AY

Trans-sentence phrase Cli- ...X...
Cl2-> ...AY...
Sentence-level Cli A CI2
(ii) As additive or restrictive 'adverbs'.
Phrase-level A <M> Y
Clause-level A <M> CI
1T8 EXCLUSIVE SENTENCES
NOTES
1
lebad be- should be added, as its parallelism with
zlt- in Is 26 1 3 shows.
2
. H. Snaith "The Meaning of the Hebrew ^SS Vetus
Testamentum (196U): pp. 221-225.
3
Brown, Driver, and Briggs, Hebrew and English
Lexicon' p. 67
11
Davidson, Hebrew Syntax: p. 202.
5
Ronald J. Williams, Hebrew Syntax: An Outline
(Toronto, 1967): P. 67.
6
Brown, Driver, and Briggs (p. 36) has a good dis-
cussion of asseverative versus restrictive âk,
with examples where it could be either.
14
ANTITHETICAL SENTENCES
14.0. ANTITHESIS BETWEEN CLAUSES
Ancient Hebrew had no one specialized conjunction meaning

but. ^Sbl, heard so frequently in modern Hebrew, had only
begun to acquire this function toward the end of the bib-
lical period. 1 But in English covers several kinds of
antithesis, including exclusive except = but not. In the
Hebrew sentence system we have distinguished a mild con-
trast (jbut - on the other hand) (Chapter 11) from outright
antithesis (but = per contra)(Chapter 14). Between these
extremes come 'adversative' sentences, for which Hebrew
has no distinctive sentence type. For hut - nevertheless,
however, exclusive or antithetical sentences are used.
14.1. THE FORM OF ANTITHETICAL SENTENCES
The structure of an antithetical sentence involves a

RELATIONSHIP between two clauses in antithesis (the fol-
lowing clause is the antithesis of the lead clause) and
a CONJUNCTION joining them. Antithesis involves not just
contrast (Chapter 11) but contradiction or opposition.
The setting of opposites in contradiction usually involves
a deep structure negation of some kind. This may find ex-
pression in several ways.
(i) Negation is implicit in any assertion. This makes
it possible for an antithetical clause to be used alone,
as a denial of the contrary statement, especially in the
repartee calculus. An utterance beginning But... implies
antithesis, even if its contrary is not explicit in the
context.
(ii) Antithesis may be achieved by placing two clauses
containing antonyms in opposition.
(iii) Antithesis may take the form of negation of the
lead clause, or of some element in it, by means of not.
(iv) Or a negation in the lead clause may be followed
by a contrary assertion in the antithetical clause.
In all these constructions the antithetical relation-
ship between two clauses coordinated in antithesis is
180 ANTITHETICAL SENTENCES
sufficiently signalled by their semantic content. This explains

why the neutral coordinating conjunction ve-r> and can be used to
join them and properly translated but. Other more distinctively
antithetical conjunctions are available, and are characterist-
ically used when there is explicit negation, especially in the
lead clause, to introduce the antithetical clause. Here kl but
is to the fore, but other conjunctions may be substituted for
it.
It needs to be emphasized that it is the total pattern of the
sentence, made up of the semantic content and sequence patterns
of the constituent clauses as well as the choice of the con-
junction, that realizes the distinctive syntax of the antithet-
ical sentence. A n antithetical clause in isolation will often
resemble formally a circumstantial, conjunctive, or chiastic
clause. In such cases the total sentence structure will often
decide what it is; but a sentence that falls short of the max-
imum realization of the distinctive-contrastive features of
any one sentence type may have to be considered indeterminate
as to type. The occurrence of such structural ambiguity, how-
ever, does not invalidate the categorizations defined in terms
of optimal realizations.
14.2. ANTITHESIS WITH IMPLICIT NEGATION
While a well-formed antithetical sentence is a two-clause

construction, an antithetical relationship may be realized
between a single clause and its general context, as when an
antithetical clause is placed alongside an entire paragraph
to bring out some kind of contrast. Since we- is the conjunc-
tion preferred in this construction, and since the item to
be contrasted is generally in the position of prominence at
the beginning of the clause, such an antithetical clause has
a formal resemblance to circumstantial, chiastic, or contrast-
ive clauses. But it is not on a side track like a circumstnat-
ial clause (which does not involve any contradiction); it does
not develop the ABBA pattern of the chiastic sentence (which
highlights similarity, not opposition); it does not develop
the ABAB pattern of the contrastive sentence (which usually
is not so contrastive as to involve negation). Such a margin-
al antithesis usually leaves the negation implicit.
Example: wiPetkem lqah yhwh, but O:you V'.acquired S'.Yahweh

(De 4 2 0 ) . This is in antithesis with Yahweh's relationship
to other nations; it is implied that he did not acquire them.
The same effect could have been secured by saying you alone.
Example: wenh ms^ hen beêne yhwh, but Noah found grace
in the eyes of Yahweh (Ge 6'). This has the form of a circum-
stantial clause rounding off an episode (#5.1.2); but it is
also antithetical, implying that Noah was not evil like the
rest of mankind as just described.
Other examples: Ge 1 7 2 1 (but Cwe-D my covenent I will estab-

lish with Isaac implying that he will not establish his
ANTITHETICAL SENTENCES 181
covenant with Ishmael [Ge 6 1 8 does not develop a similar

antithesis]), Ge 27 2 2 (The voice is Jacob's voice/ but Cve^l
t h e hands a r e Inot Jacob's hands butJ Esau's hands -- i n
Ge 31" 3 similar clauses with no such implied antithesisâre
simply conjoined), 31 s (but Uwe-] God...), 31 2 9 Cbut Cwe-J
God...), 4115 (but I have heard...), 421 0
, etc.
In Ge 28 1 9 and 4 8 1 9 the antithetical conjunction is wePlm.
14.3. ANTITHESIS BY MEANS OF ANTONYMS
Example: waêhl heek-âpel bekol-êres^mijrayim...

lekol-bene yisrêl hy ^r bemseb3tSm
And there was total darkness in all the land of Egypt

and all the Israelites had light where they lived
CEx 10 2 3 )
This is a well-formed chiastic sentence, and the conjunction

could be translated while as a signal that the actions are
contemporaneous. The contrast between darkness and light
secures antithesis, with but as the appropriate conjunction.
The Hebrew use of and enables both these deep-structure re-
lationships to be simultaneously realized in one sentence.
Other examples: Ge 29 3 1 . Ex 20 1 0 (but the seventh day is

Sabbath), 35 2 , De 2 1 1 2 6 , S 1 3 " 1 " , etc.
14.4. ANTITHESIS BY NEGATION
A negated clause following a matching assertative clause

is generally antithetical in some way and the conjunction
we-, which is the one usually used, requires but.
Example: wePkel ^tm ^Xser kuppar bhem...

wezr l^-y^kal,
And those who were atoned for with them shall eat them
BUT a foreigner will not eat [them] (Ex 2 9 s 3 )
In the optimum realization of such an antithetical clause, the

item to which the antithesis applies is placed in the focal
position before "1, even if this requires an explicit fand
redundant) subject pronoun.
Example: weniggas m5.se lebaddm êl-yhvh

wehem viggs
weh^m l^ yaâl f imm,
And Moses alone will draw near to Yahweh

BUT they [the elders] will not draw near
and the people will not go up with him ( 242)
Here the second and third clauses have the same structure
and function; they are conjoined in a sentence which, as a
whole, is antithetical with the first clause.
Other examples of antithetical clauses with structure we-X

l 1 V...: Ge 2 1 7 2 0 (but see #5.1.1), 3 3 , 4 5 (see #9.3.0),
19
19 (S:Pr does not specifically contrast with anything: its
position secures the pattern of general antithesis), 20 ,
16b
28 . 301*2. SS 1 " CS:Pr), 38 2 3 (S:Pr), 408 . 421,'e (S:Pr),
Ex 5 . 9 s * 1 " 3 2 , 19 2 " (a prohibition), 24 1 1 , 3 3 1 J l B > 2 3 ,
Le 2 1 2 , Nu l1*7, De 4 1 2 (this also exploits the pattern of
a contrastive sentence for greater antithesis by placing
ql versus temn in the focal positions), De 4* 2 (but
he did not hate him...), etc.
This prime pattern for negation in antithesis probably ex-

plains the use of Conjunction + Infinitive Absolute +
Negative + Verb in Ex 5 2 3 rather than Conjunction + Negative
+ Infinitive Absolute + Verb, which would keep the infinitive
next to the verb. But we do not wish to insist that the con-
trast between these two sequence patterns is all-important.
Negation itself frequently implies some kind of antithesis,
so the sequence Conjunction + Negation + Verb will sometimes
be found in an antithetical clause, without the additional
feature of bringing the contrasted item into focal position
before the negator.
Example: vel^-zkar Sar-hammasqim êt-ysep
But the chief butler did not remember Joseph (Ge 4 0 2 3 )
Other examples: Ge 22", 8 s , 3 1 7 t , 3 2 b , 38 2 6 , 391 0 b , De l 2 6 . - 5 .
Occasionally the sequence pattern that is optimum for an

antithetical clause (Conjunction + Contrasted item + Negator
+ Verb) is used when there is no antithesis. In Ex 9 2 9 the
position of hatbrd the hail before lo"1 serves the different
purpose of bringing it into a conjunctive pattern (Chapter 8)
with haqqSlSt thunder, while the negation is parallel to the
verb cease. Similarly, if the preceding clause is also neg-
ative, the pattern ve-X "1 V may be simply chiastic with it
and not antithetical at all (#9.4.8) (Ex 16 2 ", 30 1 0 ): or
two such clauses may be joined conjunctively (Ex 222 ) . in
Ex 12 3 9 gam cancels the antithetical force of such a clause,
making it inclusive; compare Ex 34 3 . In the Book of the Cov-
enant clauses of the form ve-X l1 V... are not antithetical
because there is no deep-structure contrast. The preverbal
position serves rather for topicalization, as each new
subject in a series of prohibitions is introduced (Ex 23 a e >
is sometimes used as the antithetical conjunction in

clauses of this kind (Ge 4 4 2 6 , Ex 33 s , De 3 2 7 } On the use of
exclusive conjunctions in this kind of antithesis, see #13.5.
14.5, ANTITHESIS AFTER NEGATION
When an assertative antithetical clause follows a negative

clause, several conjunctions can be used, interchangeably
it would seem.
14.5.1. Antithetical ve-
Example: vel^ yiqqre^ êt-simk âbrm

vehy simk âbrhm,
And your name will no longer be called Abram
but (Lit. and) your name will be Abraham (Ge 1 7 s )
Other examples: Ge 2e (There was no rain... but (hit. and)

a stream used to come up from the underworld...), Ex 1 2 1 0 ,
21 13 . Compare Ge 42 10 .
14.5.2. Antithetical Sequential wv
It is also possible to use a sequential clause antithetic-

ally after negation.
Example: vel^ ^S kaâser dibber âlehen melek misrayim
wattehayyeyn êt-hayeldim,
And they did not do as the King of Egypt told them.
And (BUT) they saved the boys' lives (Ex l17)
Another example: Ge 40 23 . Compare Ge 4 1 2 , 281'2.

14.5.3. Antithetical kl
Example: Sray ^ istek l^ tiqr^ ^ et-semh iray

kl Sr semb,
Saray your wife -- you won't call her name Saray
But Sarah is her name (Ge 1 7 1 5 )
Other examples: Ge 3 s , 24 3 " 1 , Ex 4 1 0 , 23*5, 34 13 (after a

virtual prohibition). De 4 2 2 2 6 , 5 s , 8 3 (note the repetition
of both verb and subject), 9 s , etc.
Such an antithetical clause often follows a simple negation
consisting of ^ so', which contradicts a statement previous-
ly made by another speaker.
Example: l^ kl ?hqte
Sol I You're lyinglj But you did laugh (Ge 18 1 s )
Other examples: Ge 192, 42 12 , 481'.

Such an antithetical sentence is not always clear for recogni-

t i o n , because ki has a variety of m e a n i n g s , and sometimes more
than one of them is compatible w i t h the context. In p a r t i c u l a r ,
the m e a n i n g for or because, governing a subordinate c l a u s e , may
give the reason for the n e g a t i v e statement just made.
Example: The last clause in De 2 s gives both the antithesis of

and the reason for the p r e c e d i n g statement--i won't give you
any of his land as an inheritance because/but I have given Ar
to the descendants of Lot as an inheritance. Compare De 5 s .
The antithetical possibilities of ki have n o t always been suf-

ficiently recognized by t r a n s l a t o r s . 2 Sometimes an antithetical
meaning fits b e t t e r than a causal one. So Ge 3 1 1 " 5 does not mean
you certainly won't die FOR CRSVD God knows,... The statements
a r e a n t i t h e t i c a l - - y o u certainly won't die, but ion the contrary1
(as) God knows, when you eat some of it, your e y e s will open...
can be used antithetically without a preceding n e g a t i o n ,
but w h e n so used it implies a n e g a t i o n . So Ex 1 6 s : [We did not
die w i t h full stomachs in E g y p t , as w e w o u l d have p r e f e r r e d ]
but you brought us out into this wilderness to kill this whole
community with famine. C o m p a r e Ge 1 8 l s , 3 8 1 .
1
The correlatives l" . . .ki can also realize p h r a s e - l e v e l an-
tithesis: l"1 attem. . .ki h ^ e l h i m , not you...but God (Ge 45).
Compare Ex 12not raw...but roasted.
14.5.4. Antithetical ki U m
Example: lo"1 yiqqre"1 fimk yaâqb

kl îm y i s r ^ e l yihye iimelcl,
Your name will no longer be called Jacob
but Israel will be your name (Ge 3 5 1 0 )
In Ge 3 2 2 9 the second clause is incomplete, making the conjunc-

tion phrase-level.
Other examples: Ge 1 5 " , Le 211*, Nu 1 0 3 0 , 2 4 2 2 , De 7 5 (after a

series of p r o h i b i t i o n s ) , I Z 5 ' 1 " ' 1 8 , 1 6 s .
1
14.5.5. Antithetical im l"1
This is rare. Ge 24 3 8 uses ""im l"1 antithetically w h e r e the

p a r a l l e l Ge 24'* uses ki. A s i m i l a r relationship is achieved by
means of apposition in Ge 28 1 " 2 .
1
14.5.6. Antithetical im
Example: Ge 30 3 1 (you will pay me nothing, BUT [hm] you will

do this for me; but conditional subordination is possible--
you need not give me anything IF you will do this for me).
14.5.7, Antithesis Using Exclusive Forms
See #13.5.
14.6. ANTITHETICAL QUESTIONS
An antithetical question is introduced by we^ and in Ge 227 :

Here's the flint and the fuel; BUT (Lit. and) there's the
kid for the burnt offering?
In Ge 31 3 0 the antithetical question stands in apposition
(#8.6).
14.7. ANTITHESIS IN APPOSITION3
Example: Don't look on it as an evil, BUT (=0) everything

that Sarah tells you, heed her voice (Ge 2 1 1 2 ) . Another
example: Ex 14 1 3 .
NOTES
1
According to Knig (Syntax: p. 53*0 , "'bl is used
adversatively only in Dan 1 0 7 ' 2 1 , Ezra 1 0 1 3 , 2Ch 1* , 1 9 3 ,
3 2 1 7 , but becomes frequent in the Mishnah.
2
Brockelmann, Hebrische Syntax: 13Ub.
3
An utterance-initial'adverb' whose clause-leyel function
is to mark the following statement as strongly asseverative
takes on the additional role of an inter-clause conjunction
when it introduces in the course of an utterance a statement
that is strongly antithetical to the preceding. The syntax of
1
k e n is exactly like that of ki and 'ak, so that etymolog-
ical connections have been suspected. If, in such a construc-
tion, they are translated surely, certainly, rather than but,
however, nevertheless, just as when they are utterance-init-
ial, then they are invariably 'adverbs', and the asseverative
antithesis stands in apposition with the preceding. Distinc-
tions between different kind of surface structure thus become
arbitrary in instances like this, "'ken is utterance-initial
asseverative in Ge 2 8 1 6 , Ex 2 1 " , ISa 15 32 , Is Uo 7 , U5 1 5 ,
Je 3 2 3 > 2 3 (but these could resume the antithesis in 3 2 0 ) ,
10, 8 e . It is inter-clause, meaning nevertheless, in spite
of that, introducing an unexpected fact the opposite of what
had been wrongly supposed, in Is 191', 531*, Je 3 2 0 , Zep 3 7 ,
Ps 31 2 3 , 66 1 9 , 8 2 \ Job 32 s . But it has no exclusive function.
Causal asseverative antithetical exclusive
1 ken
1 ak
raq
15
SURFACE REALIZATIONS AND DEEP RELATIONSHIPS
15.0. INTRODUCTION
In the preceding discussion distinctions between sentences

were made in two different ways (see #4.8). In terms of the
relationship secured in the sentence between the constitu-
ent clauses, which may be alternate to each other, concom-
itant with each other, and so on, there is a sentence con-
struction by means of which this relationship is distinct-
ively signalled. Each such construction in its most dis-
tinctive form is definitive of a sentence type. Each sen-
tence type functions as the prime surface realization of
one of these deep-structure relationships.
RELATIONSHIP TYPE
Consecutive Sequential (para-

graph-level) clause
Extraneous Adjunctive clause
Incidental Surprise clause
Simultaneous Circumstantial clause
Concomitant Chiastic sentence
Additive Inclusive sentence
Neutral Conjunctive sentence
Alternative Disjunctive sentence
Opposite Contrastive sentence
Contradictory Antithetical sentence
Subtractive Exclusive sentence
15.1. ALTERNATIVE SURFACE REALIZATIONS
The one-to-one correlation between deep relationship and sur-

face realization shown in #15.0 is not the whole story of the
Hebrew sentence system. Some relationships can be realized by
more than one sentence type and conversely the same sentence
type can often realize more than one deep relationship.
On first sight this might seem to threaten the validity of
the categorizations on both levels, reducing the neat system
S U R F A C E R E A L I Z A T I O N S AND DEEP R E L A T I O N S H I P S 187
to chaos. But this f l e x i b i l i t y w i t h i n the system does not

m e a n that there is no system. The a v a i l a b i l i t y of m o r e t h a n
one w a y of saying e s s e n t i a l l y the same thing gives a speaker
o p t i o n s in saying w h a t he w a n t s to say. He decides how to
say it. But this f r e e d o m is quite d i f f e r e n t from the formal
a r b i t r a r i n e s s of any language as a code. Once the code is
g i v e n , the speaker m u s t adhere to the c o n v e n t i o n s , or else
c o m m u n i c a t i o n breaks down. H o w e v e r , the c o n v e n t i o n s include
important c o n s t r a i n t s . Not every s e n t e n c e type can be u s e d
to r e a l i z e every deep r e l a t i o n s h i p . A n immense p r o b l e m
lurks in the w o r d "can" as we have u s e d it here. It sounds
legislative. We c o u l d be empirical a n d say that not every
sentence type is f o u n d to be a c t u a l l y u s e d to realize every
deep r e l a t i o n s h i p ; but this creates other p r o b l e m s . For how
do we k n o w that the m i s s i n g usage m i g h t not be f o u n d tomor-
row? Both p r o b l e m s c a n be h e d g e d by using a q u a l i f i c a t i o n
like " p r o p e r l y " or " a c c e p t a b l y " or " n o r m a l l y " u s e d , but this
involves a j u d g e m e n t o n g r a m m a t i c a l i t y that we are not real-
ly a b l e to m a k e .
The c o r r e l a t i o n in #15.0 has s e l e c t e d one sentence type
as the p r i m e (normal, p r e f e r r e d , optimum) r e a l i z a t i o n of
one r e l a t i o n s h i p , w i t h the i m p l i c a t i o n that the use of an-
other sentence type as a n a l t e r n a t i v e surface r e a l i z a t i o n
of that r e l a t i o n s h i p is somehow secondary or a b n o r m a l .
A j u d g e m e n t of this kind is d i f f i c u l t to m a k e , especially
w h e n w e are dealing w i t h a d e a d language a n d do not have
access to the intuitions of native c o m p e t e n c e . In the
absence of such a t o u c h s t o n e , w e m u s t do our best w i t h
other m e a n s . O n the basis of o b s e r v e d o c c u r r e n c e in extant
texts, all we can say is that for any g i v e n r e l a t i o n s h i p
there is one s e n t e n c e type w h i c h is its p r i m e r e a l i z a t i o n ,
there m a y be one or m o r e other sentence types w h i c h are
(apparently) u s e d as a l t e r n a t i v e (but secondary) r e a l i z a t i o n s
of that r e l a t i o n s h i p , and that some other sentence types are
never (so far as w e know) u s e d as a l t e r n a t i v e r e a l i z a t i o n s
of that r e l a t i o n s h i p .
The w r i t e r is fully aware that f r e q u e n c y of o c c u r r e n c e in
extant texts is not to be t r u s t e d as a safe m e a s u r e of the
n o r m a l c y of any g i v e n c o n s t r u c t i o n type. A c o r p u s of an-
cient literature like the O l d T e s t a m e n t offers no a s s u r a n c e
that it is g r a m m a t i c a l l y h o m o g e n e o u s , so inductions are
p e r i l o u s . A fact a b o u t incidence (this c o n s t r u c t i o n type
is f o u n d f i f t y - t h r e e times, for example) m a y be u s e f u l l y
s t a t e d as a fact, but any statistical inference is inter-
d i c t e d . The extant texts do, h o w e v e r , offer some m e t h o d -
ological controls. D e a d w r i t e r s m a y still be u s e d as in-
formants. A c o n s t r u c t i o n f o u n d a b u n d a n t l y m a y be c o n s i d e r e d
a r e p r e s e n t a t i v e sample of the language, and p r o n o u n c e d
" n o r m a l " a n d even c o n s i d e r e d n o r m a t i v e . This is w h a t w e
imply by calling a c o n s t r u c t i o n type "primary". But it does
not follow that a c o n s t r u c t i o n rarely m e t is, for that rea-
son, less g r a m m a t i c a l than one that is o f t e n used. It m i g h t
be, of course; w e have no g u a r a n t e e s that the text contains
no solecisms. But a d e c i s i o n to d i s c a r d evidence b e c a u s e it
does not satisfy our text-book grammar should never be m a d e
h a s t i l y , and the e m e n d a t i o n of texts has no p l a c e in lin-
188 SURFACE REALIZATIONS AND DEEP RELATIONSHIPS
guistics, even though linguistics has a great deal to do

with text-criticism. It may be that, in some of its oc-
currences, a conjunctive sentence is a misbegotten at-
tempt to realize antithesis, but the discovery of such
evidence, however interesting in itself, does not war-
rant the conclusion that all observed means of real-
izing an antithetical relationship could be used inter-
changeably to achieve an identical result. If one had
access to a living informant, one could test the inter-
changeability of conjunctive and antithetical sentences
by transforming one into the other and trying them out
on a native speaker. In the simplest cases, one could sub-
stitute a coordinating conjunction for an antithetical
conjunction, and vice versa. The risks attached to such
discovery procedures are well known. In our case, all we
can do is to see whether the extant texts yield such pairs
of sentences which seem to have the same function in sim-
ilar contexts. The frequent use of repetition with minor
variations in our texts makes this test available quite
frequently, and we have often applied it in the preceding
discussion.
15.2. LIMITATIONS IN ALTERNATIVE REALIZATIONS
There are limits to the interchangeability of sentence

types as alternative realizations of the same deep re-
lationships. A general relationship like simultaneity
has many possible realizations, since this factor is in-
volved in the deep relationships of all sentence types
that are not disjunctive or adversative, and the cir-
cumstantial clause, which we have identified as the prime
realization of simultaneity, should be described more ex-
actly as realizing simultaneity in the rather negative
sense of not being in temporal sequence. An 'unmarked'
sentence type like the conjunctive sentence can be used
to realize many different relationships, but its neutral
form means that the underlying relationships are often
left indeterminate. Distinctive relationships like add-
ition, subtraction, or alternation have few formal real-
izations, and a highly marked sentence type, such as a
chiastic sentence, secures a highly specialized staging
of the relationship of concomitance and cannot properly
be used for anything else.1
The material in. a chiastic sentence can, of course,
be transformed into a conjunctive sentence that is quite
grammatical. But this acceptability does not qualify a
conjunctive sentence as an alternative realization of the
kind of concomitance that is highlighted in a chiastic
sentence. When a conjunctive sentence is used, the con-
comitance is played down, although not necessarily ex-
cluded as an implicit possibility. The ubiquity of and
as an inter-clause conjunction in Hebrew is due to the
readiness with which conjunctive sentences function as
SURFACE REALIZATIONS AND DEEP RELATIONSHIPS 189
CHART OF PRIMARY AND ALTERNATIVE SURFACE REALIZATIONS OF
DEEP-STRUCTURE RELATIONSHIPS
SENTENCE TYPE
RELATIONSHIP Seq Adj Sur Cir Ch Inc Cj Dj Cn Ant Exc
Consecutive A A
Extraneous A
Incidental A A A(#12.4 .5)
Simultaneous A A A A A A
(#5 .4)(#6.2) (12.4.4)
Concomitant A(#12.4 1)
Additive A
Neutral A A (#10. 4)
(#12.,4.2)
Alternative A
(#10.5)
Opposite A A
Contradictory A
Subtractive A
(#13.5)
at the point of intersection means that the sentence type

above is the primary realization of the relationship to the
left. Thus a disjunctive sentence is the primary realization
of the relationship of alternation. A indicates that the
sentence type above it is an alternative realization of the
relationship indicated at the left. Thus simultaneity can be
realized in several ways, and the conjunctive sentence, as
the most neutral coordination, realizes more relationships
than any other type. There is clearly a spectrum with simil-
arity at one end and difference at the other, and alternative
realizations are largely restricted to the neighboring sen-
tence type on the spectrum.
190 SURFACE REALIZATIONS AND DEEP RELATIONSHIPS
alternative realizations of other coordinative relationships

as well as appositive and subordinative relationships. We-
can substitute for any other Hebrew conjunction, as used
in a primary realization sentence, and produce an alternat-
ive surface realization of the same deep relationship. But
the reverse is not true. One of the other conjunctions can
replace we- used as a sentence-level conjunction only if the
deep relationship between the clauses in that sentence is
one for which that other conjunction is used as a signal.
Thus and can replace or in any Hebrew disjunctive sentence,
to yield the form of aconjunctive sentence as an alternative
realization of the disjunctive relationship. But or cannot
replace and in a conjunctive form of sentence unless the deep
structure relationship between the clauses in that sentence
involves alternation. There are even more stringent constraints
on the interchangeability of the other conjunctions among
them-selves. Thus , gam, and raq are absolutely noninter-
changeable. See also the diagram of the interchangeability
of ki, âk, and ^ken in note 3 on page 185.
Viewed in this light, the spectrum of coordination sen-
tences shown on page 189 can be divided into four bands.
Within each band the several sentence types function as
alternative realizations of the deep relationships for which
the others are primary. Adjunctive, surprise, and circum-
stantial clauses are similar in discourse function and share
deep features, such as being nonconsecutive. Inclusive and
chiastic sentences have much in common. The disjunctive sen-
tence is unique. Contrastive, antithetical, and exclusive
sentences constitute the adversative band of the spectrum,
with considerable interchangeability between the sentence-
types. All four of these groups have the conjunctive sen-
tence as an insipid alternative.
15.3. JUXTAPOSITION AND CONCATENATION
Besides the interchangeability of coordination sentences

among each other, discussed in #15.2, some apposition re-
lationships primarily realized by means of conjunctionless
juxtaposition (Chapter 3) may sometimes be realized as a
surface coordination -- sequential (#3.4.3) or conjunctive
(#8.10.1, 8.10.2). And coordination and subordination re-
lationships of various kinds may sometimes be realized in
apposition, that is without the use of the appropriate
conjunctions (#3.10, 11.7, 14.7).
15.4. COORDINATION AND SUBORDINATION
Because of the specialized conjunctions involved, some

of the sentence-types grouped above under the rubric of
'coordination' might have been described as involving
subordination. This is particularly so with the advers-
ative sentences in which the antithesis is dependent on the
thesis, but not vice versa. Or, again, a concessive use of
SURFACE REALIZATIONS AND DEEP RELATIONSHIPS 191
kl, although, which we have omitted along with'causal for,

because, as subordinative, might have been added to its list
of coordinative functions along with the appositive and
antithetical functions that are included here. In #2.5 we
admitted that it was hard, and probably unnecessary to
draw a boundary line between coordination and subordination.
Although we have not included a detailed study of subor-
dination in this monograph, we have noted from time to
time the use of coordination sentences to realize subordin-
ation relationships (#5.5.1, 8.10.3).
NOTES
1
N o t e , h o w e v e r , f r o m # ll*. U t h a t c h i a s m u s is c o m p a t i b l e
w i t h t h e r e a l i z a t i o n of b o t h c o n c o m i t a n c e a n d a n t i t h e s i s in
the same sentence.
INDEX OF BIBLICAL REFERENCES
Genesis 21" 93
217 182
1 105 2 18 58,80
l1"2 86 21 9 93
l1 86 2 20 80,82,182
l2 79,82,85,86 221 80
87,88,101 223 37
l3 85,86 22* 182
l5 129 22 5 77
l6 105,106,107 31 79
111 33 113,182
l7 124 35 183,184
l9 105,106,111 36 87,100,116,
l10 129 162,163
21 -1 5 106 38 25
l1" 118 J9-19 129
l15 118 31" 54,99
l20 105,106,134 315 50,54,150
122 105,108,117 3 16 44,54
134 317-18 99
l26 112 31 7 54,55
l27 55 318 55
l28 47,108,117 319 133
l29 56 32 2 161
l31 86,87,95 32 3 43
2*-6 86,87 32" 43
2" 54 41 79
25"6 86 42 " 5 122,124,127
25 101,135 137
2s 183 43--. 158
27 86 4*~ 5 129
28 80 4" 88,157
29 79,85,87, 45 42,129,182
101 4s 114
2 10 79,82,87 47 114
211"1" 33,34,59 49 37
212 85 410 37
INDEX OF BIBLICAL R E F E R E N C E S 193
4 11 54 71 J- I 6 124
412 56,183 7 13 41
4 13 50 7 1 5- 1 6
128
4 18 88 7 16 124
419 32,33 71 7 124
420 719-19
59 127
421 32,33,59 719-20 41
422 88,157,158 721-22 40
42 3 38 723-21 172
425 ~ 2 6 88 72 3 42
42 6 88,157,158 85 37,47,80
5 87 8 8" 9 77
5 1" 2 55 89 182
51 54 811 95
52 54 812 77
61 86 813 95,103
1* 37 8 16 " 1 7 39
65 101,116,171, 817 39
175 8 18" 1 9 39,40
68 80,81,180 8 18 39
69 54,55,80 8 19 39
61 2 95 8 2 1" 2 2 44
6 13 96 821 113,131
1 6
61* 50,99 91 108
1" 57 92 " 3 56
6 15 51,54,99,127 92 80
9 3-
616 47,48,51,59 173
6 17 39,151 9" 116
6 1 7" X9 96 95 50,99
618 - 19 132 96 47
618 181 97 99,117,151
619- 20 39 99 151
619 47,55 9 11 54,77,113,131
62 1 151 912 54
62 2 42,43 9 15 77
7i-% 39 91 7 54
72 39 gie-19 45
73 154 918 93
7s 42 92 3 78,81,101
7 6" 1 6 124,125 92 8 54,112
76 80,86,88, 92 7 111
101,124 10 87
7 7- 8 39,124 101 54
7' 124 108 59,88
78 39 1 0 i o - n 59
79 124 1012 59
y 1 0 -1 2 124,126 1013 88
7 1 0 -1 6 41 101" 59
y 10 41,86,124, 1015 88
130 IO20 54
11
7 41,124,126, 1021 88,157,158
127 IO2" 88
712 124,130 1025 88
19 1 INDEX OF B I B L I C A L REFERENCES
25 32,33 IS- 95,184

26 88 IS5 108
31 54 IS10"11 130
32 45,54 151 2 85,87,96
11 87 IS13"16 133
II3 57,98,109,120 1513 131
II" 57,85 IS1" 165
II7 57 1515 47
IIs 101 15 1 7 85,87,96
1 1 " 54 1510 45
II12 88 161 31,80,85,87
II1* 88 16 s 114
II27 54 , 1 0 3 169 108
II29 31,32,33 1611 117
II30 43,83 1612 50,99,104
121 " 2 108 1613 163
122 111 1616 81
12 3 81,131 171 81,108
126 85 172 111
128 49,50,91 17* 151
12 1 2 132 ,152 ,158 17s 183
12 1 3 45 17s 47,132
12 1 6 80 17 1 0 54,151
132 85,160 171 2 49
13s 160 171 3 47
13 6 45 171" 58,93
137 82 171 5 183
139 54 17 16 47,165
1311"12 49 17 1 7 114
131 2 87,151,160 1720 47,129
1313 82 172 1 180
131 ^ 57,80,108 j^ 2 - 2 5 81,101
131 5 162,163 172" 41
14 40 172 6 41,45
14 1 ~ 3 41 18 1 82
142 59,103 18 2 84 , 9 5
14 3 85 183 _ 5 112
14*"5 41 ,103 18 6 " 7 129
14 7 30,85,155 187 130
14 8 59 18" 94
14 1 0 80,128 18 9 85
1411 ~1 2 42 18 10 82,85,87
141 2 82 1811 47,80,85
14 1 3 82,85 18 1 2 85,90
141 s 128,155,158 18 1 3 85,90,166
14 1 7 59 18 1 6 82,84
14 1 8 80,82 18 1 7 80,82,90
1 4 9 - 2
107 1817" 19 85
142 1 134 18 1 8 82
142* 172 18 1 9 184
15 1 44 18 2 0 101,102
152 82,90,93 18 2 0 " 2 1 85
153 115 18 2 2 ~ 2 3 84
1822 81,84 221 86,88

182 3 84,85,166 222 108
1 8 2 11 37,166 22" 37
18 2 7 85,90 22s 111,134
18 3 0 118 227 185
18 3 2 175 2212" 13 101
183 3 81.84 221 2 114,118
19 1 82,84 2213 95
192 108,183 2216 101
19 3 128 221 7 99,101,116
19" 87 22 2 0 157,166
19 5 112 222 3 54,88
19 6 81 2 2 2 ** 31,93,156,161
19 7 112 232 59
19' 171 231* 112
1910 81 239 46,47
19 1 1 81 2310 79
191 * 57 2311 47,104,129
191 5 56 2313 112,177
191 7 45,50,114 231 9 59
19 1 9 182 241 80,101
1920 85,112 243 " 183
192 1 162 ,163 243 108
19 2 3 87 24* 184
192 7 84 248 171,177
192 8 95 2410 85
19 3 1 83 241 2 108
19 3 2 57,111 2413 115
19 3 3 161 24111 103,112,160
193 112,162 2415 95
1935
161 241 s 117
29 3 7 - 3 128 24 1 9 162
19 3 7 59 242 1 81,148
193 8 157,158 242 5 155
19 3 9 59 2 42 9 31,79
20 3 90 243 94,95
20" 85,154 ,182 24 3 1 85,90
20s 114,157,160 243 5 101
20s 162,166 24 3 8 184
207 108 24"" 159,160
209 47,149 24 1 * 5 85
20 1 2 165,166,170 24"' 160
2013 54 24 1 * 9 141
201 6 129 2450 58,141
211 42,80,124,130 24 5 3 128
21s 81 24 5 141
216 58 2456 85,90,112
2112-13 24* 7 111
160
211 2 57,185 246 0 113
211 3 93,165 24 6 2 79,82
2 1 1 '*" 1 5 80 24s3 95
211" 47 251 31
211 8 57 253 88
2126 163,176 25" 54
196 INDEX OF B I B L I C A L REFERENCES
25s- 42,129 27-3 50,57

257 42 27-5 155
25 8 42,43 28 1 " 2 183
25 1 0 43 28 2 57,111
25 1 1 43 28 3 111
25 1 2 54 286 116
25 1 54 28 8 118
25 1 8 42 281 2 " 1 5 95
25 1 9 54 281 2 82
25 2 3 131 2813 92
2 5 2 11 95 28 1 6 113,182,185
252 6 81,82 28 1 7 50
25 2 8 128 28 1 9 181
252 9 85 28 2 1 131
25 3 3 " 3 " 135 28 2 2 93
25 3 3 135,136 292 85,95
25 3 136 29 7 47,57
26 2 47 29 9 87
26 3 111 291 2 116
26 8 95 291 * 177
26s 177 29 1 6 32,33,79,85,87
26 1 5 79,80,93 2917 85
26 2 1 156,161 29 2 1 112
26 2 6 79,80 292* 88
26 2 7 85,90 29 2 5 96
26 2 8 113 29 2 7 161
26 2 9 171,175 29 2 9 88
27 1 44,80 29 3 0 161
27 3 . 108 29 3 1 85,181
27s 81,118 29 3 3 161
27 s 80 303 113,156
27 7 108,111,112 306 166
27 9 50 308 165
27 1 1 152 301 5 161
271 3 57,177 303 0 156,162
27 1 5 136 3031 184
27 1 7 152 30 3 3 47
27 1 9 56,57,118 303 6 81
272 1 148 30"2 182
27 2 2 101,181 311 130
Z72S 112 31 2 95
27 2 6 108 315 95,181
272 8 ~2 9
99 316 54,80,130
27 2 9 108,111 31 7 130,182
27 3 0 86,177 3110 95
27 3 1 111,118,156,161 311 2 108
27 3 3 166 3113 57
27 3 " 156 311 3 108
27 3 6 50,166 311 5 166
27 3 7 100,115 311 9 80
27 3 8 156 31 2 5 87
27" 99 31 2 6 114
27"2 113 31 2 9 181
31 3 0 185 35 2 108
31 3 2 78,182 35 3 111
31 3 ^ 85 35 6 59
31 3 6 38 3510 184
31 3 6 99 3511 47,117,1;
31 3 9 44,47 35 1 2 100
31*1 50 3517 162
31 113 93,100,141,181 35 1 8 128
31 **7 128 35 1 9 59
31 5 0 116 35 2 0 59
3151 115 35 2 s 54
315 2 50,108,116 35 2 7 59
32 2 80,81 35 2 9 42
32 7 83,85 36 1 30,54,59
3210 112 36s 54
3217 134 36 9 54
3 2 1 2 0 156 36 1 2 54,79
321 9 96,164,166 36 3 2 90
3220 155 36 3 ^ 128
32 2 1 96,164 37 1 " 2 81
32 2 2 81 37 2 54,87
32 2 7 174 37 3 128
32 2 9 184 37 t _ 7 95
32 3 0 161 37 8 38,147
32 3 1 96 37 9 95
32 3 2 82 37 1 1 128
331 85,95 37 1 ^ 57,108
33 3 80 37 1 5 95
33 7 161 37 2 0 111
33s 83 37 2 2 57,135
33 1 1 116 372" 43,85
3313 111 37 2 5 95
33 1 * 134 37 2 7 135
33 1 81 37 2 8 81
331 7 81,129 37 2 9 83,95
34 40 37 3 2 148
34 3 42 37 3 3 44
34 5 79,80,85,87 37 3 5 82
34 7 79 37 3 6 81
34 8 92 38 1 31
34 9 49,100,132 38 2 31
34 1 0 108,132 38 6 31
34 1 2 47 38" 1)8,117
34 1 5 54,175 38 1 0 161
34 1 6 100,132 38 1 1 156,162
34 1 9 85 381" 182
34 2 1 100,132 38,s 184
34 2 2 175 38 2 0 77
34 2 3 177 38 2 2 165
34 2 7 40 38 2 3 182
34 2 8 40 38 2 11 96,113
34 3 0 85 38 2 5 88
35 1 57,108 38 2 6 182
198 INDEX OF B I B L I C A L REFERENCES
38 2 7 96 42 182
38 2 9 96 42 6 48,93
391 80 42" 182
39 3 101 42s 48
39" 128 42 1 0 " 1 1 45
39 6 172 421 0 181,183
39" 37,85 421 2 183
39s 47 4213 152
391 0 182 421 5 174
39 1 1 83 421 6 87
391" 44 4218 58
3922 128 421 87
392 3 47 422 0 80
40s 95 422 2 161
40 83,182 422 3 86
401 2 54 422 7 95
401 k 177 422 8 161
401 6 162 42 3 1 43
401 8 54 42 3 2 152
4021 " 2 2 128 42 3 3 54,108,152
402 3 68,182,183 42 3 59,100
411-" 95 42 3 5 95
411 86 42 3 8 50,53,101
415-7 95 42 3 8 101
41' 83 43 1 80
41' 50,53 432 57
1 3
411 50 43 3 174
4111 46 43 1 * 111
4112 46,79 43*-5 54
4113 50,68,152,158 43 5 174
4115 83,181 438 100,155
411 6 176 43' 44
4 1 1 7 ~ 21 95 43 1 1 _ 1 108,109
41 2 1 90 4311 54,133
2
4 1 " - 95 43 1 2 108,109
41 "2 83 4313 57
41 2 6 " 2 7 50 431 ^ 93,108,109
41 2 6 100 43 1 5 128,134
41 2 7 100 431 6 108
41 2 8 54 4321-22 128
413 2 79 432 1 95
4 1 3 3 " 3 > 38 432 7 48
413 3 111 44 1 108,134
41" 100,175 44 3 41,55
41** 176 44- 57,88
41 * 8 81 44 5 53,114
41-" 40 44 8 141
41 5 0 80 449-10 160
41 5 52 128 152,165
44 1 0
415- 81,130 44 1 2 49
41 5 6 80 441 * 83
41 5 7 81 44 16 155
42 2 100,108 44 1 7
152
INDEX OF BIBLICAL R E F E R E N C E S 199
4 4 1 3 141 50s 108,111
4420
90,128 50" 173
44 2 1 112 50 155
4 4 2 3
175 501 8 162
4425
57 50 2 0 57,153
44" 8 3 , 9 0 , 1 7 4 , 182 50 2 3 48,158
4428
177 502" 87
4429
162
44 3 0 83,86
44 3 1 83 Exodus
44 3 3 134
4 4 3 - 83,90 1 1
40,46
114 5
45s 81,86
57 1 7
45" 79
45s 184 110 163
44,108 1
459 44
4510 54 11 5 33
4S1" 128 11 7 183
451 6 79,80 12 2 152
451 7 " 1 8 108 25 82
45 1 7 54,56,57 26 95
452 2 50,100 29 108,151
452 6 116 2 13 95
45 2 8 83 21* 185
46- 81,100,152 ,166 21 6 79
46s-7 40 22 2 116
46 7 42 31 80
462 2 85 32 95,152
462 9 79 33 111
4631 111 3s 57
463" 155 37 123,127
47.1 95 3 10 108
47 2 79 3 11 108
47 3 155 312 54
4 7 5 - 6
37 315 102
47s 47 318 111
47 1 3 42,79,81 46 95
47 1 s 127 47 95
47 1 9 155 49 155
47 2 9 170 410 155,183
47 2 1 93 4 1 1 149
47 2 ^ 132 412 86
47 2 6 171 41 5 86
48 5 48 41 6 151
48s 48 417 86
48 7 59 4 18 111
48 1 0 44,85 4 19 57
48 1 1 95,161 52 102,133
48 1 * 48,90 53 111,141
48 1 8 183 58 44,57,86
48 l s 37,100,159 ,162, 5 11 57
181 5 13 82
49 31 99 5 111 155
49 33 42 517 57
200 INDEX OF BIBLICAL REFERENCES
518 57,182 922 113

52 1 111 92 3 128
52 3 182 9 2 If 82
61 153 92 5 128,129
62"3 102 g2 6 175
6 3"5 165 92 7 50,100
66 -6 102 92 9 100,182
611 57 931-32 182
612 90 931 101
61* 54 933 128
615 54 103 113
616 54 107 113
61 54 108 57
62* 54 109 37
62 5 85 1011 108
626 48 1012 113
62 7 48 1013 87,128
630 90 10l,( 50,103
71 133,151 1017 108,113,175
72 50,151 1019 44
77 81 1021 113
79 108 1023 44,181
711 157,161 102* 57,164,175
1 it 48 1025-26
165
71 5 86 1026 44
717 54 ll3 164
718 133 ll6 116
719 108 ll10 80
72 1 130 122 38
72 3 163 12 " 5 48
72 6 113 12" 46
J 2 133 12 6 171
81 108 12 8 40
8* 113 12 9 184
85 171,175 12 1 0 183
87 175 12 1 1 58,99
812 108 12 1 2 132
8 13 40 12 1 40
gli 108,113 12 1 5 50,174,175
819 48 12 1 6 48
82 0 40 12 1 7 " 2 0 48
82 1 57 12 2 1 108
82 3 57 12 2 7 130
82* 171 1228 42
82 s 171 12 2 9 86
828 163 12 3 1 57,155
91 108 12 3 2 155,166
92 117 12 3 5 80
93 113 12 3 6 86
96 182 12 3 8 86
91 3 108 12 ^ 0 81
916 54 12 1 * 2 50
915 57,58,152 12 3 " " 9 50
g20-21 152 12"3 54
3.2 5 0 42 181* 48,82

132 48 18 1 7 58
136 7 50 18 1 8 58,165
13 86 1819 21
134
I312-13 152 182 2 152
131 5 133 18 2 3 160
1 3 2 1 - 2 2 44 18 2 6 152
132 1 82 19 3 80,98
142 48,111,113 19 6 54,131
143 44 19s 50
146 128 19 1 0 113
148 82 191 2 44,48
1410 80,95 1913 44 , 4 8 , 1 4 2
141 2 108,112 146
1413 185 1915 48
141- 58,152 1918 86
1415 113 19 2 1 57
1416 "1 7
151 19 2 2 161
141 6 108 19 2 * 57,182
14 2 2 85 20 1 0 181
142 6 113 201 5 82
14 2 7 82 20 1 6 112
142 8 44 2018 128
14 2 9 82,86 2019 50
IS3 31,32 202 8 121
152 5 48,99 211 54
152 6 131 212 100
152 7 85 213 6
54
16 3 184 21" 141,142,151,152
16* 148 21s 54,58
166" 7 152 21 6 141
1610 95 2113 183
16 1 2 152 211 6 148
16 1 3 152 211 8 141
1616"1 8
46 21 2 0 141
1618 131 21 2 1 141,174
16 2 3 152 21 2 6 141
16 2 * 131,182 21 2 7 141
162 5 58 21 2 8 85,141,143,159
162 6 48 21 2 9 141,160
162 9 44 2131 143
16 3 1 85,101 213 2 100,141
16 3 3 108 213 3 141,144
16 3 5 48,81 213* 48,152
16 3 6 93 213 5 132
171 83 213 6 146,152
172 57,112 213 7 141,148
175 108,134 22" 141
17 7 148 22s 141
17s 57,85,108 22s 141
17 1 0 80 22 9 141
17 1 2 80,86,103 22 1 3 146
17 1 * 108 22 2 6 44
18 3 33,34 22 2 7 182
222' 50 253 9 48
22 3 0 48 25" 71
23 3 182 261 "3 0 71
231* 141 261 ~ 3 56
238 100,182 261 48,56,71
23 1 1 100 262 56
23 1 2 100 263 56,128
2313 14 26* 72
231""17 60 26s 104
2315 "1 6 50 26 7 48
23 1 s 48 26 8 48,23
231 7 50,53 2612 " 13 72
232 1 108 26 1 1 1 85
232 5 183 261 7 53
23 2 8 124,133 2618 "2 2 72
23 3 0 57 26 1 8 72
24 1 80 2619 72
24 2 181 2620-21 72
24 s 128 262 2 72,80
24 1 0 85 26 2 3 72
24 1 1 182 262" 53
24 1 2 108,112 2628 72
2413 183 262 9 71
24 1 ^ 80 263 1 "3 7 71
24 1 7 85 26 3 1 48
25-30 71 26 3 5 72,132
25 1 - 27 1 9 7
> 27 1 " 8 71
25 1 ~ 8 71 27 1 48,127
25 2 46,113 27 2 48
2510 "2 2 71 27 3 48
2510 "2 3 127 27 8 48
25 1 0 50,51 27 9 " 1 9 71
25 1 1 48 27 1 7 48
25 1 2 72 27 1 8 48
25 1 5 44,50 272 0 80
25 1 8 48 27 2 1 46,48
25 1 9 48,71 28 2 133
25 2 0 48,72,85 28 2 1 86
25 2 1 72,133 28 2 - 2 5 132
25 2 2 71 28 3 2 48
25 2 3 " 3 0 71 28 3 7 48
25 2 3 51 28 3 9 132
252 7 48 28" 132,133
25 2 8 71 28"2 48
2529 48 28*3 142,148
2 531 ~ 0 71 29 1 54
25 3 1 48 292 48
25 3 2 72,82 293-.. 132
25 31* 85 29s 80
25 3 5 72,85 291 2 132
25 3 6 48 29 1 ^ 58,132
25 3 7 71 291 5 80
25 3 8 72,85 291 7 80
25 3 8 " 3 9 72 29 2 3 181
INDEX OF B I B L I C A L REFERENCES 203
29 3 1 80 35 3 48
29 3 ^ 44,48 36* 100
29 3 5 48 35s 48
29 3 7 48 35 2 1 ~ 2 9 41
29 3 8 54 35 3 1 133
29 3 152 353" 133
291*1 152 35 3 5 37
301 48 363 80
30 2 48,50 367 81
3 0 3 " 11 128,132 3610 128
30 3 48 36 1 2 104
30 7 " 8 50 361 * 48
3010 46,182 3615 50
30 2 0 142,148 361 7 128
30 2 5 38 3621 50
30 2 6 _ 3 0 132 362 3 "2 7 72
30 2 3 48 3 6 2 3 -21* 128
30 3 3 48 36 3 3 " 3 ^ 128
303* 48 363 5 48
30 3 8 48 37 1 50
31 6 127 37 6 50
3111 53 37 7 48
311 3 177 37 8 48
311 37,48 37 1 0 50
3115 48,50 37l* 48
3117 48 37 1 7 48
32 1 57 37 1 8 82
32 2 108 37 2 0 85
32 7 " 8 56 37 2 2 48
32 7 57 37 2 48
32s 95 37 2 5 50
2 6 - 2 7
32 1 2 108 3 7
128
32 1 3 131 38 1 50
32 1 5 48 38 2 48
32 1 6 93 38 3 48
32 3 " 57 38 7 48
33 1 57 38 2 1 54
33 3 182 39" 48
33 6 80 39 s 93
3311 182 39 9 48,50
331 7 162 3910 50
33 2 6 182 39 1 93
34 1 - 2 108 39 1 7 128
34 3 156,159,182 39 3 2 42
34 1 3 100,183 39*3 95
341- 31 40 1 " 1 5 73
34 8 48 40 2 73
34 2 1 48,50. 40 1 2 73
34 2 3 53 40 1 73
34 2 8 99 40 1 7 73
34 2 9 85 40 2 9 73
34 3 0 95 403 2 48
35 1 54 403" 128
35 2 48,50,181 403 5 127
20k INDEX OF BIBLICAL REFERENCES
Leviticus 151* 141

152 5 145
3
l 48 152 9 141
132 17 3 146
l10 141 192 0 146
l13 132 2027 141
l1" 141 21 2 174
23 85 211" 184
210 85 2119 146
212 182 21 2 3 173
2 13 38,117 22'* 146
36 141 22 s 146
47 132 22 174
412 48 2327 177
17-Ji 132 2 5 " <1 9
4
143
42 3 146 251* 9 146
430-31 132 26 s 83
4 i<* -3 s 132 26-2 166
5 1"5 145 262" 156
51 85,144 27 2 8 142,173
52 85,141 27 2 8 173
53 85
5* 85,142
5s 141 Numbers
57 141
5s 131 l*7 182
5 11 141 l"9 173
518 131 422 162
52 1 141,142,146 42t 157
52 3 142 511 "31 144,146
52* 141,142 51" 144
52 8 131 921 142
7 16 141 g 22 142
721 141 103 184
7 33 92 11" 161
11" 174 ll8 146
ll32 141 ll23 148
ll36 173 ll28 32,33
12 6 141 12 2 162,175
12 7 141 13 1 8 148
12" 141 13 1 9 148
13 2 141 1320 148
13 1 s 145 14 2 143
13 1 9 141 14 9 177
132* 141 14 3 0 172
132 9 141 IS8 148
133 0 141 163 166
13 3 8 141 1610 162
13*2 141 161* 174
13"3 141 18 3 155,173
13*7 141 18 1 5 173
13* 8 141 18 1 7 173
13"9 141 18 2 8 162
13 5 9 142 21 2 2 173
INDEX BIBLICAL REFERENCES
22 1 9 166 2 32 149
2220 176 23" 44,149
22 3 3 158 23s 175
22 3 5 176 23 6 44
22 3 8 176 33 161
23 1 2 176 3* 44
2313 173 36 "7 126
23 2 5 163 39 37,152
23 2 6 176 311 50,175
24 1 2 162 3 1 2 ~ 13 152
2413 176 31 2 128
24 2 2 184 315"1 6 152
26 3 3 172 3 1-22 136
26 5 5 174 320 162
26 6 5 172 325 111
3011 146 32 7 108
31 2 3 174 328 153
3516 " 2 3 145 42 113
35 1 s 145 4- 152
35 1 7 145 49 177
35 1 8 145 410 139
35 2 0 145 411 82
35 2 1 145 412 182
35 2 2 145 41- 128
35 2 3 145 420 180
35 3 3 172 421 80,151
36 s 173 422 44,183
42 s 44,183
430 48
Deuteronomy 433 48
435 44
1-4 54 436 153
l1 53 439 44
12S 182 4*12 182
l28 155 41.". 54
l32 83 4 "t 5 - it 9 54
l35 182 5 54
l37 162 52 ~5 37
l38 57 5 2 "3 44
J39-0 152 S3 183
22 112 5s 184
2* 134 513-1- 181
2s 160 521 132
2s 114,184 s2S 57
2 i 1-12 133 527 57
211 156,162,181 529 44
21* 85 613"111 44
215 93 613 50,100
2 19 114 g22- 23 126
220 156,162,181 72-3 113
22 3 93 72 44
22" 57 73 50,133
227 38,44,182 7s 50,100
22a 100,153,177 78 44
7 57 14 7 174
7 1 ** 54 14 2 7 93
7 " 57 15" 174
2 0
158 15 1 7 162
2 2
44 16 s 184
72* 44 1615 175
7 25
44 18 3 142
82 148 18 2 0 174
83 116,183 22 2 2 155
8" 153 23 3 165
gll-18 138 23" 165
811 137,138 2319 155
81 2 " 1 7 137 2515 86
81 2 137,138 28 2 9 175
813 137,138 28 6 1 165
81* 137,138 32 1 6 38
81 7 137,138 32 2 5 155
81 8 137,138 34 1 6 133
8 20
46 34 2 1 133
92 116 34 2 2 133
93 100,153 34 2 3 133
95 183
g7-23 53 Joshua
97 44,126
98 126 21 31
9s 153 3" 173
91 2 56,57,90 513 148
g1 3 95 9" 166
gl- 112 912 92
91 5 82,85 1013 123
g16 48,56,95
g1 8 ll10 128
153,161 ll13 169
919 161 ll32 173
g20 126,130,161 18 1 9 148
92 1 80 24 1 5 142
922 126 24 1 8 165
92* 53
g2 6 - 2 7 57,58
10 1 108 Judges
10 6 80
J22 157
1010 161
10 1 1 57 221 162
1015 174 3 2 ** 177
1016 132 5" 164
102 100 529 162
1021 101 62 8 139
ll3"6 116 639 175
12 5 184 719 177
12 6 169 72 5 128
12 1 * 184 82 2 155
12 1 184 8 31 157
12 2 2 g s
177 128
12 3 0 156,162 g 9
166
12 3 1 162 10 9 155
INDEX OF B I B L I C A L R E F E R E N C E S 2 07
51
10 177 25-3 155
13 2 31 262 5 159
13 1 8 90 28 6 155
16" 31 281 5 155
16 1 5 90 28 1 9 164
16 2 8 175 28 2 0 165
17 1 31,90 296 174
17 s 93
19 1 * 155
20 3 9 177 2 Samuel
20"8 164
l15 57
2s 31
1 Samuel 210 174
37 32
l1 31 313 174
l2 32,33 31 7 155
l6 166 42 32,33
l23 177 4* 31
22 6 155 52 155
82 33 92 31
8 170 91 2 31
91 31 12 1 3 166
92 31 121* 166
12 1 4 155 12 2 7 165
12 1 5 165 131 31
12 1 6 165 133 31
12 2 0 177 14 2 7 31
12 23 164 162 3 155
12 2 - 177 17s 157
12 25 155 17s 148
14" 33 17 9 59
14 2 9 32,34 171 0 157
14-9 33 1 7 25
31
15 3 2 185 202 1 31
16 1 s 177 23 10 175
1711 31
17 1 2 31
17 23 31 1 Kings
17 3 6 155
18" 175 313 155
18 1 7 177 3 26 155,163
18 2 2 139 415 166
19 23 156 ll 3 8 176
20 27 155 132 31
20 3 9 174 18 3 8 128
21 8 31 2119 157
21 9 155 221 5 147
22 7 166 22 3 2 177
22 2 0 31
25 3 33
25 s 175 2 Kings
25 1 6 155
25 2 1 177 1" 92
23 162 26 1 5 174
25 162 26 2 * 174
21" 166 28 7 174
57 177 30 1 1 174
62 7 148 32 3 0 175
12" 174 341* 177
12 1 * 174 37 3 31
13 6 174 46 1 8 30
17*1 155 48 1 5 30
18 2 0 175 51 5 7 30
22 7 174
23 9 174
23 2 6 174 Ezekiei
23 3 5 174
24 3 177 46 1 7 174
Isaiah Hosea
3 139 4* 177
57 139 12 2 175
10 1 5 147 12 9 177
14 1 5 174
16 7 175
19 1 1 175 Joel
34 1 * 175
34 1 5 175 l2 148
35 6 175
40 7 185
41 9 139 Amos
43 2 " 174
413
45 1 5 185 31
45 2 * 175 59 31
48 8 163 52 7 32
49* 185 9s 31
53" 185
57 1 5 32
60 2 0 123 Jonah
63 177
25 174
JeTemiah
Zephaniah
l6 89
23 5 177 l18 175
3 13 177 37 177
520 37 185
185
32 3 185
5" 177
529 147 Zechariah
10 1 9 177
10 2 * 176 l6 174
12 1 177 612 30
16 1 9 175
INDEX OF BIBLICAL REFERENCES
Malachi Ecclesiastes
147 219 148
Psalms Esther
23s 176 4 11 70
312 3 185
661 9 185
827 185 Daniel
83 1 9 30
226 30
48 30
Proverbs 41 9 30
101 30
18' 58 107 185
10 21 185
Job
Ezra
31
147
65 147 10 1 3 185
66 147
83 148
" 147 1 Chronicles
10s 147
ll2 148 1" 32.33
ll7 147 22 6 31
163 147 23* 31
21* 148 715 33
22 3 148 716 32
32" 185 836 33.34
38 2 9 147 g<t 33,34
38 3 1 147 299 157
421 * 33
2 Chronicles
Ruth
l1 90 1" 185
l2 33,34 19 3 185
1* 32,33 289 31
21 31 321 7 185
'^

The Sentence in Biblical Hebrew

Uploaded by

Copyright:

Available Formats

The Sentence in Biblical Hebrew

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Sentence in Biblical Hebrew

Uploaded by

Copyright:

Available Formats

JANUA L I N G U A R U M

Series Practica, 231

MOUTON PUBLISHERS THE HAGUE PARIS NEW YORK

ISBN 90 279 2673 5

Copyright 1974 by Mouton Publishers, The Hague, The Netherlands

Printed in Great Britain

Recent advances in linguistic theory and method make desir-

SYMBOLS AND ABBREVIATIONS 14

THE TRANSLITERATION OF HEBREW 16

3.4. Epic Repetition in Apposition 39

4.3. Sentence-level Coordination 66

9. CHIASTIC SENTENCES 119

9.4.3. Subjects in Chiasmus 127

10. DISJUNCTIVE SENTENCES 141

11. CONTRASTIVE SENTENCES 150

12. INCLUSIVE SENTENCES 154

13. EXCLUSIVE SENTENCES 168

13.6. Exclusive Forms Used for Coordination or Appo-

14. ANTITHETICAL SENTENCES 179

15. SURFACE REALIZATIONS AND DEEP RELATIONSHIPS 186

INDEX OF BIBLICAL REFERENCES 192

I apologize to Hebraists for not citing texts in Hebrew

1.0. TRADITIONAL GRAMMAR

The Hebrew SENTENCE (Se) as the domain of inter-clausal rela-

zur Semitischen Philologie und Linguistik Heft 7 (Leipzig,

1.1. DISCOURSE GRAMMAR

The strong points of tagmemics are its empirical approach,

sity Press, 1972), where the bibliography will give sufficient

THE SENTENCE IN HEBREW

2.0. DEFINITION OF SENTENCE

The term SENTENCE (Se) has been used in linguistics in several

written.) Grammatical completeness, as a sine qua non of sen-

2.1. THE CLAUSE AS A SENTENCE ELEMENT

We can stalk the sentence from another direction by talking

each language. When the language has the means of realizing a

More precisely, a construction in which predication is real-

sentence. It leaves open for study three questions: (i) What

2.2. SENTENCE TYPES

Sentences containing two or more clauses have been tradition-

of the relationships into which these nominalized clauses enter

2.3. COMPLEX SENTENCES

A complex sentence is a two-clause construction in which a

The point is a fine one, and we shall return to it again. Mean-

2.4. COMPOUND SENTENCES

The traditional definition of a compound sentence recognizes

dent without being subordinate. Everything that contributes to

2.5. RELATIONSHIPS BETWEEN SENTENCE TYPES

So far, using conjunctions as our chief clue, we have recognized

The connection between the constituent clauses is increasingly

2.6. SURFACE GRAMMAR AND DEEP GRAMMAR

2.6.. The English Relative Clause

In fact, no phrases with this structure are attested anywhere

2.6.1. Head is a Proper Noun

An alternative proper name for a person or place is given as

Head N : N p ^ <A> \Modifier:Cl

2.6.2. Head is a Pronoun

The only instance in the Old Testament in which the antecedent

2.6.3. Head is a Definite Noun

The only construction which has a definite noun as antecedent

2.6.4. Head is an Indefinite Noun