Beginning Syntax

A coherent introduction to generative syntax by a leader in the field,

this textbook leads students through the theory from the very beginning,
assuming no prior knowledge. Introducing the central concepts in a sys-
tematic and engaging way, it covers the goals of generative grammar, tacit
native-speaker knowledge, categories and constituents, phrase structure,
movement, binding, syntax beyond English and the architecture of gram-
mar. The theory is built slowly, showing in a step-by-step fashion how
different versions of generative theory relate to one another.
Examples are carefully chosen to be easily understood, and a
comprehensive glos-sary provides clear definitions of all the key terms
introduced. With end of chapter exercises, broader discussion questions,
and annotated further reading lists,

beginning syntax is the ideal resource for instructors and beginning

undergraduate students of syntax alike. Two further textbooks by Ian
Roberts, Continuing Syntax and Comparing Syntax, will take students to
intermediate and advanced level.

IanRoberts_9781316519493_C000.indd 1 10-09-2022 16:54:50

 CA M BR I DGE T EXTBOOKS I N LI NGU IST ICS

General editors: u. a n s a l d o , p. au s t i n, b . c o m r i e , r . l a s s ,
d. l i g h t f o o t, k . r i c e , i . r o b e rt s , s . r o m a i n e , m . s h e e h a n,
i. tsimpli

Beginning Syntax
An Introduction to Syntactic Analysis

IanRoberts_9781316519493_C000.indd 2 10-09-2022 16:54:50

 In this series:

r . ca n n Formal Semantics
j. l av e r Principles of Phonetics
f. r . pa l m e r Grammatical Roles and Relations
m . a . j on e s Foundations of French Syntax
a . r a d f o r d Syntactic Theory and the Structure of English: A Minimalist
Approach
r . d. va n va l i n j r a n d r . j. l a p o l l a Syntax: Structure, Meaning and
Function
a . d u r a n t i Linguistic Anthropology
a . c ru t t e n d e n Intonation Second edition
j. k . c h a m b e r s a n d p. t ru d g i l l Dialectology Second edition
c . lyon s Definiteness
r . k ag e r Optimality Theory
j. a . h o l m An Introduction to Pidgins and Creoles
g . g . c o r b e t t Number
c . j. e w e n h . va n d e r h u l s t The Phonological Structure of Words
f. r . pa l m e r Mood and Modality Second edition
b . j. b l a k e Case Second edition
e . gu s s m a n Phonology: Analysis and Theory
m . y i p Tone
w. c r o f t Typology and Universals Second edition
f. c ou l m a s Writing Systems: An Introduction to Their Linguistic Analysis
p. j. h o p p e r a n d e . c . t r au g o t t Grammaticalization Second edition
l . w h i t e Second Language Acquisition and Universal Grammar
i . p l ag Word-Formation in English
w. c r o f t a n d a . c ru s e Cognitive Linguistics
a . s i e w i e r s k a Person
a . r a d f o r d Minimalist Syntax: Exploring the Structure of English
d. bü r i n g Binding Theory
m . bu t t Theories of Case
n. h o r n s t e i n, j. n u ñ e s a n d k . g r o h m a n n Understanding Minimalism
b . c . lu s t Child Language: Acquisition and Growth
g . g . c o r b e t t Agreement
j. c . l . i n g r a m Neurolinguistics: An Introduction to Spoken Language
Processing and Its Disorders
j. c l ac k s on Indo-European Linguistics: An Introduction
m . a r i e l Pragmatics and Grammar
r . ca n n, r . k e m p s on a n d e . g r e g o r o m i c h e l a k i Semantics: An
Introduction to Meaning in Language
y. m at r a s Language Contact
d. b i b e r a n d s . c on r a d Register, Genre, and Style
l . j e f f r i e s a n d d. m c i n t y r e Stylistics
r . h u d s on An Introduction to Word Grammar
m . l . m u r p h y Lexical Meaning
j. m . m e i s e l First and Second Language Acquisition
t. m c e n e ry a n d a . h a r d i e Corpus Linguistics: Method, Language and
Practice
j. s a k e l a n d d. l . e v e r e t t Linguistic Fieldwork: A Student Guide
a . s p e n c e r a . lu í s Clitics: An Introduction
g . c o r b e t t Features
a . m c m a h on a n d r . m c m a h on Evolutionary Linguistics
b . c l a r k Relevance Theory
b . l on g p e n g Analyzing Sound Patterns
b . da n c yg i e r a n d e . s w e e t s e r Figurative Language
j. by b e e Language Change
s . g . t h o m a s on Endangered Languages: An Introduction
a . r a d f o r d Analysing English Sentences Second edition
r . c l i f t Conversation Analysis

IanRoberts_9781316519493_C000.indd 3 10-09-2022 16:54:50

 r . l e v i n e Syntactic Analysis
i . p l ag Word-Formation in English Second edition
z . g . s z a b ó a n d r . h . t h o m a s on Philosophy of Language
j. p u s t e j ov s k y a n d o. bat i u kova The Lexicon
d. b i b e r a n d s . c on r a d Register, Genre, and Style Second edition
f. z ú ñ i ga a n d s . k i t t i l ä Grammatical Voice
y. m at r a s Language Contact Second edition
t. h o f f m a n n Construction Grammar
w. c r o f t Morphosyntax: Constructions of the World’s Languages
i . r o b e rt s Beginning Syntax

Earlier issues not listed are also available.

IanRoberts_9781316519493_C000.indd 4 10-09-2022 16:54:50

 Beginning Syntax
An Introduction to Syntactic Analysis

I A N RO BE RT S
Downing College, University of Cambridge

IanRoberts_9781316519493_C000.indd 5 10-09-2022 16:54:50

 Shaftesbury Road, Cambridge CB2 8EA, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467

Cambridge University Press is part of Cambridge University Press & Assessment,

a department of the University of Cambridge.
We share the University’s mission to contribute to society through the pursuit of
education, learning and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781316519493
DOI: 10.1017/9781009023849
© Ian Roberts 2023
This publication is in copyright. Subject to statutory exception and to the provisions
of relevant collective licensing agreements, no reproduction of any part may take
place without the written permission of Cambridge University Press & Assessment.
First published 2023
Printed in <country> by <printer>
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
ISBN 978-1-316-51949-3 Hardback
ISBN 978-1-009-01058-0 Paperback
Cambridge University Press & Assessment has no responsibility for the persistence
or accuracy of URLs for external or third-party internet websites referred to in this
publication and does not guarantee that any content on such websites is, or will
remain, accurate or appropriate.

IanRoberts_9781316519493_C000.indd 6 10-09-2022 16:54:51

 Contents

Preface page xi
List of Abbreviations xiii

Introduction: What Linguistics Is About and

What Syntax Is About 1
What Is (a) Language? 3
What Is Language? vs What Is a Language? 3
Natural Language vs Artificial Languages 4
Language and Languages 7
Natural Language 9
Exercises 12
For Further Study and/or Discussion 13
Further Reading 13

1 Tacit Knowledge (or: Several Things You Didn’t Know

You Knew about English) 15
1.1 Introduction 15
1.2 Some Fish 15
1.3 More Fish and Some Relative Clauses 22
1.4 Really Challenging Fish 25
1.5 Conclusion: What the Fish Have Taught Us 31
Exercises32
For Further Study and/or Discussion 33
Further Reading 34

2 Constituents and Categories 35

2.1 Introduction and Recap  35
2.2 Categories  36
2.2.1 Introduction: Lexical and Functional Categories  36
2.2.2 Morphological Diagnostics for Categories  38
2.2.3 Syntactic Diagnostics for Categories  39
2.2.4 Phonological Diagnostics for Categories  42
2.2.5 Semantics and Categories  42
2.2.6 Conclusion 44
2.3 Constituent Structure  44
2.3.1 Phrasal Constituents  44
2.3.2 Hierarchical Relations among Constituents  45
2.4 Conclusion  47

vii

IanRoberts_9781316519493_C000.indd 7 10-09-2022 16:54:51

 viii contents

Exercises  48
Further Reading  49

3 Phrase-Structure Rules and Constituency Tests 51

3.1 Introduction  51
3.2 Phrase-Structure Rules  52
3.3 Recursion  55
3.4 Constituency Tests  59
3.5 Summary and Conclusion  75
Exercises  76
For Further Discussion  77
Further Reading  77

4 X′-theory 78
4.1 Introduction  78
4.2 Possible and Impossible PS-Rules  78
4.3 Introducing X′-theory  80
4.4 X′-theory and Functional Categories: The Structure of the Clause  87
4.5 Adjuncts in X′-theory  95
4.6 Conclusion  96
Exercises  97
For Further Discussion  98
Further Reading  98

5 Movement 99
5.1 Introduction and Recap  99
5.2 Types of Movement Rules  100
5.3 Wh-Movement: The Basics  104
5.4 Subject Questions, Indirect Questions and Long-Distance
Wh-Movement  112
5.5 The Nature of Movement Rules  118
5.6 Limiting Wh-Movement  122
5.7 Conclusion  126
Exercises  127
For Further Discussion  128
Further Reading  128

6 Binding 130
6.1 Introduction and Recap  130
6.2 Pronouns and Anaphors  130
6.3 Anaphors  132
6.4 Pronouns  142
6.5 R-expressions  143
6.6 The Binding Principles  144
6.7 Variables, Principle C and Movement  146
6.8 Conclusion  153

IanRoberts_9781316519493_C000.indd 8 10-09-2022 16:54:51

 Contents ix

Exercises  154
For Discussion  154
Further Reading  155

7 Syntax beyond English 156

7.1 Introduction and Recap  156
7.2 Approaching Universal Grammar  156
7.3 Verb Positions in French and English  157
7.4 Word Order and X′-theory  160
7.5 A Case Study: German  165
7.6 The Wh-Movement Parameter  171
7.7 Conclusion  174
Exercises  176
For Further Discussion  178
Further Reading  178

8 The Architecture of Grammar 179

8.1 Introduction and Recap  179
8.2 A Model of Grammar  179
8.3 The Lexicon  180
8.4 D-structure  187
8.5 S-structure  190
8.6 Phonological Form (PF)  195
8.7 Logical Form (LF)  203
8.8 Conclusion  211
Exercises  212
Further Reading  214

Conclusion 215

Glossary 219
Index 234

IanRoberts_9781316519493_C000.indd 9 10-09-2022 16:54:51

IanRoberts_9781316519493_C000.indd 10 10-09-2022 16:54:51
 Preface

The purpose of this book is to present the essential elements of the theory of
syntax, presupposing no prior knowledge of either syntax or linguistics more
generally. The first chapter introduces the thinking behind modern formal
linguistics, defining the core background concepts. The second chapter is an
extended demonstration of linguistic competence, revealing to an English
speaker their tacit, untutored knowledge of the intricacies of the syntax of the
language. The next four chapters introduce the core technical concepts of syn-
tax: phrase structure, constituency, movement rules and construal rules. Chapter
7 is a very brief introduction to comparative syntax, illustrating how the theory
developed in the earlier chapters can apply to languages other than English.
Finally, Chapter 8 discusses the overall architecture of grammar, introducing
several levels of representation, including the interface levels of Phonological
Form and Logical Form.
The book is based on a first-year lecture course I have taught at Cambridge for
most of the past ten years, called ‘Structures’. That course has successfully laid
the groundwork for more advanced study; several professional syntacticians had
their first exposure to the subject in that context.
I do not attempt to review either the history or the state of the art in contem-
porary generative syntax. Instead, the book is intended to provide a solid basis
from which the student can move on to more advanced topics. In the decades
since its inception, generative theory has yielded a range of core insights and
concepts which can be studied independently of the details of a given framework
and which form the core of any formal approach to syntax (and core background
to the formal study of other areas of linguistics, in particular semantics). These
are the focus of this introduction, where the theoretical orientation of the book is
most clearly set out. This volume is intended as the first in a series; future vol-
umes will build on the material presented here, as well as both adding an explic-
itly comparative and typological dimension and engaging more directly with the
minimalist programme for linguistic theory. The overall aim of the series is to
provide a complete course in syntax, taking the student from beginner level to
being able to engage with contemporary research literature.
This volume is intended for students starting out in syntax and assumes no
prior knowledge. As mentioned above, it corresponds to a first-year course
taught at Cambridge. It can be taught over a single term or semester with tutorial

xi

IanRoberts_9781316519493_C000.indd 11 10-09-2022 16:54:51

 xii preface

support (as I have done at Cambridge) or over an entire year as a lecture course.
The book tries to introduce the main concepts of syntactic theory in an engaging
way with a minimum of technicalities. Each chapter features exercises, some
straightforwardly testing the material covered in the chapter, others raising fur-
ther questions for reflection and discussion (e.g. in tutorials), as well as sugges-
tions for further reading. Model answers are provided separately for the first type
of exercise. The second type presupposes student participation in discussions.
Chapter 1 is somewhat different from the others in that it introduces some fairly
complex data; the point at this stage of the exposition is to demonstrate tacit
knowledge, i.e. native-speaker competence; the technical issues are, as appropri-
ate, revisited and introduced more gradually later. Instructors may wish to skip
over some of this material and/or revisit it later (for example, the Italian data can
supplement the material covered in Chapter 7).
I would like to thank the other people who have taught this course in whole or
in part over the years: Dora Alexopoulou, Theresa Biberauer, Craig Sailor and
Jenneke van der Wal, as well as the many doctoral students, postdocs and others
who have given tutorials for the course. Most of all, I’d like to thank the students
themselves: you ultimately make everything happen. Thanks also to Bob Freidin
and Dalina Kallulli for comments on early drafts. And of course, thanks to the
two Helens at Cambridge University Press: Helen Barton and Helen Shannon,
whose tolerance of missed deadlines fully matches my capacity to bring them
about.

IanRoberts_9781316519493_C000.indd 12 10-09-2022 16:54:51

 Abbreviations

AAVE African-American Vernacular English

acc accusative case
Agr agreement
A(P) Adjective (Phrase)
Adv(P) adverb (phrase)
ASL American Sign Language
Aux(P) auxiliary (phrase)
BSL British Sign Language
C(P) Complementiser (Phrase)
CNPC Complex NP Constraint
Comp complementiser
Conj conjugation
D(P) determiner (phrase)
gen genitive case
LBC Left Branch Condition
LF Logical Form
M mood
Mod modifier
Neg(P) negation (phrase)
nom nominative case
N(P) noun (phrase)
NSL Nicaraguan Sign Language
num number
OV object-verb
OSV object-subject-verb
OVS object-verb-subject
PF phonological form
Pl plural (preceded by an Arabic numeral, denotes the relevant
person, e.g. 3Pl = third-person plural)
P(P) preposition(al phrase)
Perf(P) perfect (phrase)
Pres present
Prog(P) progressive (phrase)
pst past
S sentence

xiii

IanRoberts_9781316519493_C000.indd 13 10-09-2022 16:54:51

 xiv list of abbreviations

SASL South African Sign Language

Sg singular (preceded by an Arabic numeral, denotes the relevant
person, e.g. 3Sg = third-person singular)
SOV subject-object-verb
SVO subject-verb-object
Spec specifier
top topic (marker)
T(P) tense (phrase)
UG Universal Grammar
VO verb-object
V(P) verb (phrase)
VOS verb-object-subject
VSO verb-subject-object
WALS World Atlas of Language Structures (Haspelmath et al. 2013)

IanRoberts_9781316519493_C000.indd 14 10-09-2022 16:54:51

 Introduction: What Linguistics Is About
and What Syntax Is About

This book is an introduction to one part of the general subject of l­ inguistics: syn-
tax (words which are boldfaced are either technical terms or names of languages
and language families which may not be familiar; at the end of the book you can
find a glossary where these terms are defined). Syntax is the study of the struc-
ture of sentences, how these sentences arise and how speakers of a language are
able to use and understand sentences by having a mental representation of their
structure. In this book, the language we focus on is English, not because there
is anything particularly special about English, but just because that’s the one
language I can be sure you know, since you’re reading this book.
Syntax is one of the central sub-disciplines of modern linguistics. The other
central areas of linguistics are phonology (the study of sounds and sound sys-
tems) and semantics (the study of meaning). There are many other aspects to
linguistics (e.g. morphology, the study of the form of words), but these three
form the core of language: the fundamental thing about language is that it con-
nects sound and meaning through sentences. Phonology deals with the sounds,
semantics deals with the meanings and syntax deals with the sentences. Syntax
is the bridge between the sound and the meaning of sentences. We’ll have plenty
more to say about this in the chapters to follow, but first let’s look at linguistics
more broadly.
Modern linguistics is the scientific study of language. The education systems
in many parts of the world make a major distinction between sciences (like biol-
ogy, physics and chemistry) and arts or humanities (like history, music or lit-
erature). Languages are almost always classified on the arts/humanities side,
so it might seem odd to talk about looking at language and languages scientifi-
cally. ‘Science’ conjures up visions of men (mostly) in white coats carrying out
experiments using various kinds of specialised equipment, all of which seems a
far cry from reading Shakespeare, learning French verbs or trying out Spanish
conversation.
But really, looking at something scientifically means two main things. First,
it means observing things as part of nature (rocks, stars or slugs, for exam-
ple). Language is so much a part of us that it can be difficult to think of it as a
natural object, the way rocks, stars and slugs obviously are. But imagine you
were a Martian observing planet Earth: you’d see rocks, slugs, plenty of insects
and a featherless biped building machines (including spacecraft), in control of
everything, and making noises with its breathing and eating apparatus. As an

IanRoberts_9781316519493_Introduction.indd 1 05-09-2022 11:17:14

 2 introduction

intelligent Martian, you’d soon recognise that there is some connection between
the bipeds’ noises, their ability to build machines and their control over all the
other species. You’d also notice that only the featherless bipeds make these
noises and that the young of the species spontaneously start making the noises
when they’re very small and still dependent on their parents for survival. After
abducting some bipeds and looking inside their brains, you’d realise that there’s
just a couple of areas of the biped brain which seem to control the noise-making
ability. The noises are language (Martians would probably first notice phonol-
ogy, but being super-intelligent they’d quickly realise that there was more to
language than just that). By looking at ourselves through Martian ‘eyes’ (which
might of course be infrared heat sensors attached to their feet, but never mind),
we can appreciate language as a natural object.
The second thing about science is that it’s not just about observing things;
it’s also about putting the observations together to make a theory of things. One
of the most famous of all scientific theories is Einstein’s Theory of Relativity.
This is hardly the place to go into that, but the point is that Einstein and other
physicists wanted to construct a general understanding, based on observation, of
certain aspects of nature: the Theory of Relativity is mostly about space, time
and gravity. A good theory, like relativity, makes sense of the observations we
have already made and also predicts new ones that we should be able to make.
So saying linguistics is scientific doesn’t necessarily mean that we have to put
on white coats and start a laboratory. But it does mean that we should try to make
observations about language and put those observations together to make a theory
of language. Since we’re going to focus on syntax, we’re going to make obser-
vations about syntax and make a theory of syntax. So the goal of this book is to
present a particular theory of syntax, a scientific theory of what it is that links sound
and meaning (as you can imagine, there are separate but connected theories of pho-
nology and semantics). The theory that I will introduce here is called generative
grammar. The reasons for that name will emerge over the next couple of chapters.
Approaching language, or syntax, in this way means that there are aspects
of the more traditional, humanities-style approaches to language that we leave
behind, since they do not contribute to the goal of making observations and
building theories. First, it implies that the goal of linguistics is not to set stand-
ards of ‘good’ speaking or writing. This is not to imply that such standards
should not be set, or that linguists, or others with particular expertise in lan-
guage, should not be those responsible for setting them. Instead, it means that
setting these standards falls outside the scientific study of language, since doing
this doesn’t involve making observations about what people say, write or under-
stand but involves recommending that people should speak or write in a cer-
tain way. Zoologists don’t tell slugs how to be slimy; linguists don’t tell people
how to speak or write. Ideas of ‘good’ syntax have no place here. Prescriptive
grammatical statements of a kind once common in the teaching of English in
schools in most of the English-speaking world (see below for some examples)
are similarly irrelevant to our concerns here.

IanRoberts_9781316519493_Introduction.indd 2 05-09-2022 11:17:14

 What Is (a) language? 3

A second consequence of leaving behind the traditional humanities-style

approach to the study of language and adopting a scientific approach instead
is that we do not evaluate language aesthetically. This does not mean that the
aesthetic appreciation of language, in particular literary language, is without
interest or value as a study in itself, merely that this represents a different way of
studying (and enjoying) language from the scientific one. Botanists may appre-
ciate the beauty of flowers as much as anybody else, and zoologists no doubt
admire the beauty of slugs, but the scientific study of flowers and slugs leaves
aside any aesthetic response to them. Similarly, linguists may well appreciate
the beauty and power of great works of literature, but the goal of modern lin-
guistics is not to explicate our reaction to such works; that is for our colleagues
in the literature departments.
There is also a third aspect of a scientific approach to a field of investigation, a
really basic one. That is to begin by defining our terms, so let us start in this way.
If linguistics is the scientific study of language, then we should start by defining
language itself. That, as we shall see, is not as simple as might first appear.

What Is (a) language?

What Is language? vs What Is a Language?
First of all, it is useful to distinguish language as a general concept
from individual languages. Our definition of linguistics makes reference to this
general concept and, as such, tells us that linguistics is concerned with this gen-
eral concept rather than with the study of individual languages (and hence tells
us that linguistics is not about acquiring proficiency in individual languages; this
is another aspect of the traditional humanities approach to language that modern
linguistics leaves aside). It is very difficult to give an adequate general definition
of language, especially in advance of the first steps in constructing any kind of
theory of language or its subparts. For the moment, let us define language as the
communicative and cognitive capacity that, as far as we know, most clearly dis-
tinguishes humans from other animals; this is what the Martians saw when they
observed terrestrial featherless bipeds making noises. All human societies and
cultures that have ever been discovered have language, most commonly spoken
language; current estimates are that there are over 7,000 languages spoken in
the world (www.ethnologue.com), although individual languages are hard to
count, and we will see below that there are reasons to question whether English,
French, etc., are really natural objects. On the other hand, no non-human animal
has a linguistic or communicative capacity comparable to ours. This is not to
deny that communication systems of various kinds are easy to observe in other
species, but none of these systems seem to have the structural complexity we
can observe in human language; as we shall see, the nature of the syntax of
human language may lie behind the gulf between human language and animal
communication. Indeed it could be argued that language is not just unique to our

IanRoberts_9781316519493_Introduction.indd 3 05-09-2022 11:17:14

 4 introduction

species, but a defining property of it. The official biological name of our species
(in the Linnean taxonomy) is Homo sapiens, Latin for ‘wise man’. Since a
glance at a history book or a newspaper leads one to question the wisdom of our
species, one could think that Homo loquens (‘talking man’) might have been a
better term to define us.
As just mentioned, the general concept of language, roughly defined above,
should be distinguished from individual languages. An individual language
can be taken to be a specific variant of the uniquely human capacity defined
above, usually part of a given culture: English, Navajo, Warlpiri, Basque, etc.
Concepts such as ‘English’ are thus, at least in part, cultural concepts. The dis-
tinction between language (in general) and languages (individual languages as
partly cultural entities) distinguishes two variants of a single word in English.
This is rather like the way we can distinguish the general concept of cheese
from individual cheeses such as Brie, Cheddar, Parmigiano Reggiano and so on.
Language and cheese are mass nouns; they denote general concepts. Individual
cultures in different parts of the world produce their local variants of the gen-
eral thing, associated with countable versions of the same nouns, i.e. individual
languages and cheeses. In several European languages, different words are used
to make the distinction between language and languages. French distinguishes
langage (language in general) from langue (individual languages); in Italian lin-
guaggio is distinguished from lingua in a similar way, and Spanish distinguishes
lenguaje from lengua.

Natural Language vs Artificial Languages

A very important distinction in the context of defining the object of
study of scientific linguistics is that between natural languages and artificial lan-
guages. Natural languages are natural objects (like slugs, etc., as we have seen),
and so they can be the object of scientific study. Natural languages are languages
spoken as native languages (they are almost always somebody’s mother tongue)
and are intrinsically capable of fulfilling a full range of communicative func-
tions (making statements, asking questions, giving orders, etc.). Moreover, the
origins of natural languages are obscure in that we are unable to fix a specific
time or place for where they began. For example, handbooks on the history of
the English language typically date the beginning of English to the arrival of the
Angles, Saxons and Jutes in the British Isles starting in the fifth century CE. The
Angles, Saxons and Jutes spoke closely related Germanic dialects which origi-
nated where they themselves originated, in what is now Southern Denmark and
Northern Germany. As far as we know, nothing changed about their language
when it was transplanted, along with its speakers, from the mainland to the off-
shore islands. Hence, the date attached to the ‘beginning of English’ reflects
a random aspect of external history and almost certainly does not reflect any
linguistic change at all. English is a direct continuation of the Germanic dialects
spoken on the continent before the fifth century by the Angles, Saxons and Jutes.

IanRoberts_9781316519493_Introduction.indd 4 05-09-2022 11:17:14

 What Is (a) language? 5

These dialects were in turn a continuation of one dialect of Indo European,

which was originally spoken (as far as we know) three or four millennia earlier
in either the Pontic-Caspian steppe region of what is now Ukraine or in Central
Anatolia (modern Turkey). Again, the beginnings and endings of ‘English’,
‘Germanic’ and ‘Indo European’ are arbitrarily chosen historical events. The
languages themselves simply continue; what came before Indo European is
uncertain, and this further illustrates the point that natural languages have an
essentially unknown origin. All we can be sure of is that our capacity for lan-
guage must have evolved at some stage in the history of our species, but this
event is lost in the mists of time many millennia before Indo European or its
precursors.
Artificial languages, on the other hand, are languages designed for some
specific purpose and restricted in terms of their functions. We can usually say
when and by whom they were invented. As such, they are not natural objects.
So, for example, logic, which can thought of as an artificial language, was
invented by Aristotle ca 400 bce, and the various computer-programming lan-
guages in use today and in recent years were all invented after roughly 1950,
when modern computers themselves were invented. Semaphore, Morse code
and other signalling systems are also artificial languages of this general kind,
and also highly restricted in function. Artificial languages are of less interest for
our purposes here, since they are not natural objects. So I will leave them aside
in what follows.
In this context, we should mention other kinds of language and languages.
Sign language is of great interest. In fact, sign language is not a single linguistic
entity: there are various different Sign Languages, e.g. British Sign Language
(BSL), American Sign Language (ASL), Nicaraguan Sign Language (NSL),
South African Sign Language (SASL) and many others in many parts of the
world. Sign languages are primarily used by Deaf communities. It is now an
established finding of modern linguistics that sign languages are natural lan-
guages in precisely the sense defined above: they are natural objects. Sign lan-
guages serve the full range of communicative functions and their origins are
obscure. Most importantly, sign languages are not manual versions of the spo-
ken or written language used by the hearing communities around them. For
example, BSL is the creation of the British Deaf community. Its history goes
back to accounts in the fifteenth century describing deaf people using signs. The
first description of those signs appears in the Marriage Register of St Martin’s,
Leicester, in 1576, describing the vows signed by Thomas Tillsye, who ‘was
and is naturally deafe and also dumbe, so that the order of the forme of mar-
riage used usually amongst others which can heare and speake could not for his
parte be observed’ (www.ucl.ac.uk/british-sign-language-history/beginnings/
marriagecertificate-thomas-tillsye).
The only difference between sign language and familiar languages such
as English is that sign languages use a different medium, the gestural/visual
medium, while familiar languages use the oral/aural medium (and of course

IanRoberts_9781316519493_Introduction.indd 5 05-09-2022 11:17:14

 6 introduction

writing). One of the striking features of natural languages, both spoken and
signed, is that their salient structural properties are independent of their medium
of transmission: sign-language syntax, for example, has the same structural
properties as English, French and other spoken languages.
Constructed languages, or conlangs, are also relevant here. Conlangs are of
two main kinds. First, there are languages which were deliberately constructed
as ‘ideal’ languages intended to serve for more efficient communication than
could be afforded by seemingly imperfect already existing languages. Esperanto
is probably the best known, although far from the only, language of this kind.
Esperanto was invented by L. L. Zamenhof in 1887, at a time when there was
no clear international lingua franca in the Western world (since 1945, English
has played this role; at earlier times Latin and French did). The vocabulary of
Esperanto is a mixture of Romance, Germanic and Slavic elements, and the
grammatical system is a highly simplified and artificially regularised. The sys-
tem is, however, closely based on various European, principally Romance,
languages. As such it is arguably a natural language, which happens to have a
particular origin and purpose. It is debatable, however, whether Esperanto has
any true native speakers, although it is certainly used as a second language by
up to 2 million people all around the world (compare this with the estimated 2
billion second-language speakers of English). Many other invented languages
(Interlingua, Volapük, etc.) have the same status as Esperanto, although these
days they are hardly spoken at all.
The other principal kind of constructed languages are those invented for
fictional purposes, in order to give linguistic realism to invented worlds and
their denizens. Among the best-known examples are the languages invented by
J. R. R. Tolkien in his extensive mythological writings (The Lord of the Rings,
The Hobbit, The Silmarillion, etc.). Tolkien was a professional philologist, and
so the languages he constructed have an air of authenticity. Like Esperanto,
however, they are largely based on existing, mainly European, languages. They
are thus natural languages which are unusual in their origin and purpose; they
are also barely spoken at all and certainly have no native speakers. Above all for
this last reason, we leave such languages aside here; our interest is in languages
which are acquired naturally.
Two other kinds of ‘language’ should also be mentioned. First, there is ‘body
language’: frowning, shaking or nodding one’s head, crossing one’s arms or
legs, etc. Body language can certainly communicate various attitudes or emo-
tional states (friendliness, aggression, etc.), but it is not a language in the sense
that it lacks the structural properties of natural, spoken and signed, languages;
it has no discernible syntax, for example. Second, there is music. Music is often
said to be a language, and indeed it has been shown to have structural, syntac-
tic properties which are akin to, perhaps even identical to, natural language.
However, although music may have a profound emotional impact, it lacks a clear
propositional semantics, in that it cannot communicate true or false statements.
It may be that music is a cognitive capacity which shares some, but not all, of

IanRoberts_9781316519493_Introduction.indd 6 05-09-2022 11:17:14

 Language and Languages 7

the properties of natural languages. Certainly, music appears to be both univer-

sal across human cultures and unique to humans; in these important respects, it
resembles natural language.
To sum up, our focus here is on natural language. Modern linguistics is really
about natural language in the sense described at the beginning of this section
and, henceforth, I will restrict the discussion along these lines. Linguists thus
believe that it makes sense to study natural language in general and not just
individual languages. If natural language forms a coherent object of study, and
if individual languages are specific variants of language in general, then this
implies that all individual languages must have something in common. Hence, a
major focus of linguistic research, particularly since the middle of the twentieth
century, but with much older historical roots, has been the following question:
what are the common properties of natural languages not shared by other sys-
tems of communication?1 One of the central goals of modern linguistic theory
is the attempt to answer this question, and we will address this question as it
applies to syntax in much of what follows.

Language and Languages

Once we define natural language in the way we have done so far,
its study can still be approached from various perspectives. As we mentioned
earlier, one approach, which comes from the traditional humanities, is the pre-
scriptive one. This involves defining precepts of ‘good’ English, in principle
independently of how native speakers might actually speak or write the lan-
guage. This has led to the formulation of precepts such as ‘Don’t use two neg-
atives, since that makes a positive.’ In many varieties of non-standard English
all over the English-speaking world, it is easy to observe people saying things
like I don’t like no-one or Mick can’t get no satisfaction, and so on. Here two
negatives are used, but the result is not a positive. Invoking an artificial kind of
logic, which does not correspond to the facts we can observe about the syntax
and the meanings of these sentences, is not in line with the scientific approach
to language we are adopting here. Such artificial ‘rules’ mostly originate in the
eighteenth and nineteenth centuries and reflect little of substance about the true
nature of English or any other language.2 So I will say no more about ‘rules’ of
this kind (although negative sentences of the kind shown above are very com-
mon in many languages of the world, including French; they illustrate a phe-
nomenon known as negative concord).

1
See R. Robins (1967), A Short History of Linguistics from Plato to Chomsky, London: Long-
man, and V. Law (2003), The History of Linguistics in Europe, Cambridge: Cambridge Univer-
sity Press, for discussion.
2
See R. Freidin (2020), Adventures in English Syntax, Cambridge: Cambridge University Press,
for a very illuminating discussion.

IanRoberts_9781316519493_Introduction.indd 7 05-09-2022 11:17:14

 8 introduction

Another perspective is a teleological one. This is certainly relevant for the

study of individual languages with the goal of attaining proficiency in speaking
and/or writing them, another important aspect of the traditional way of studying
language and languages. If I say I am learning to speak Catalan, then ‘speaking
Catalan’ here is defined as a goal I am striving towards. It does not correspond
to any currently existing knowledge or ability I have. But this is clearly some-
thing distinct from studying Catalan, or any other individual language, with the
goal of understanding its structure; in particular, understanding its structure in
relation to the general question of the possible structural commonalities of all
natural languages, as discussed in the previous section. Hence, the teleological
approach to the study of individual languages, although of course worthwhile,
falls outside of modern linguistics as we understand it here.
Still another perspective on language is the sociopolitical one. One very
important distinction in this connection is that between a dialect and a ‘stand-
ard’ language. Despite the social and political importance of this distinction, if
we consider individual languages and dialects to be variants of the more general
notion of language, then we are led to consider all natural-language systems as
equal, whatever their social or political status (or that of the people who speak
them). If we consider language to be a single coherent entity, then we expect to
find the same general structural characteristics in all individual languages and
dialects; this, of course, is the initial hypothesis that we intend to investigate and,
if possible, substantiate.
Moreover, all natural languages and dialects appear to be of roughly the same
level of complexity. Although this point is hard to assess and verify in detail and
in a fully satisfactory way, there does not appear to be any reason to assert that
one or another language or dialect is ‘simpler’ or ‘more basic’ or ‘more primi-
tive’ than another. In structural terms, then, there is no reason to privilege one
language or dialect over any other; the fact that one particular variety may have
emerged as a standard is usually just a historical accident. For example, Standard
(British) English emerged in the fifteenth and sixteenth centuries as English
became the main language of commerce and correspondence. Standard English
came to predominate as the written form of the language and came to be taught in
schools as the ‘correct’ or ‘standard’ form. The emergence of Standard English
was a consequence of social and political factors; it had nothing to do with the
structural features of that variety of English, and there is accordingly no reason to
think that that particular variety is in any way intrinsically superior to any other.
So, from the point of view we adopt here, the distinction between dialects and
standard languages is an artificial one, extrinsic to the structural features of the
varieties in question. When we look at language variants as natural objects, we
see that the language vs dialect distinction is spurious; it comes from culture, not
nature. Back in 1945, the American sociolinguist Max Weinreich captured this
point by saying that ‘a language is a dialect with an army and a navy’. Political,
economic and military power confer prestige on standard languages, but from a
purely linguistic point of view, they are not distinct from dialects.

IanRoberts_9781316519493_Introduction.indd 8 05-09-2022 11:17:15

 Natural Language 9

We can take this line of thought further. The notion of a standard language is
really a type of idealisation, one specially constructed by language planners and
educationalists often interested in telling people what to do. In many countries,
one of the goals of teaching the standard language is teleological, in the sense dis-
cussed above. In reality, nobody really speaks the ideal standard language; every-
body speaks some form of dialect. One very clear result of the scientific study
of the social aspects of language (the field known as sociolinguistics) is that no
language is homogenous. Everybody, including people who consider themselves
speakers of ‘Standard English’, actually speaks a slightly different variant of the
language, approximating an ideal standard in slightly different ways (which, in
practice, one may hardly ever have cause to notice). As pointed out well over a
century ago by the German historical linguist Hermann Paul, ‘we must in reality
distinguish as many languages as there are individuals’. No two speakers of a
given language actually know precisely the same things about that language or
use it in precisely the same way. Alongside dialects, then, we recognise idiolects,
the variety of a language employed by a particular person. We may also want
to recognise sociolects (varieties associated with particular social classes), eth-
nolects (varieties associated with particular ethnic groups) and so on.
All of this may leave you feeling perplexed. If all speakers have different idio-
lects, is there a notion of an individual language that can really be usefully defined
at all? Atkinson (1992:23) gave the following definition of an English speaker:
The person in question has an internal system of representation … the overt
products of which (utterance production and comprehension, grammatical-
ity judgements), in conjunction with other mental capacities, are such that
that person is judged (by those deemed capable of judging) to be a speaker
of English.3

We can see from this that defining ‘a language’ isn’t as simple as we

might have at first thought.
Leaving aside the social, cultural and political dimensions (which are evident
in Atkinson’s definition), we can try to understand what ‘language’ is so that we
can then define ‘a language’ as a specific variant (token) of this more general
entity (type). In this way, we take the everyday terms for individual languages
(English, French, etc.) to be primarily sociocultural. Strictly speaking, these
terms are not, in fact, part of scientific linguistics, since they designate cultural
rather than natural objects.

Natural Language
Leaving aside the sociocultural dimensions, then, we concentrate on
looking at natural language from a scientific perspective. More precisely, we

3
M. Atkinson (1992), Children’s Syntax, Oxford: Blackwell.

IanRoberts_9781316519493_Introduction.indd 9 05-09-2022 11:17:15

 10 introduction

will look at language from a cognitive perspective. This approach treats lan-
guage as a form of knowledge; the central idea is that our uniquely human lan-
guage capacity is part of our cognitive make-up. Language is really a kind of
instinct humans, and no other species, are pre-programmed with (in fact, The
Language Instinct is the title of an influential and important book by Stephen
Pinker in which this approach is explained in detail; see Further Reading at the
end of this Introduction). This is the view that forms the basis of the theory of
generative grammar and so this is the view I will adopt here. This approach, and
most of the concepts introduced in this section, are due to Noam Chomsky (see
Further Reading at the end of the Introduction for more details).
The cognitive approach treats our capacity for language as an aspect of human
psychology which is ultimately rooted in biology. In this way, it clearly treats
language, and our capacity for language, as natural objects. On this view, there
are three factors in language design. The first is contributed by genetics: an
innate aptitude for language that is unique to our species, our ‘language instinct’.
The second factor is experience, particularly in early life. The language we are
exposed to as children represents the crucial linguistic experience; when I was
growing up in England, everyone around me spoke English and so I acquired
English. My great-grandfather was surrounded by Welsh speakers when he
was growing up in nineteenth-century North Wales, so he acquired Welsh.
Whichever language, or languages, we are exposed to as small children, our
innate language-learning capacity is brought to bear on this experience in such a
way as to result in our competence as native speakers of our first language (I will
say more about language acquisition below). Third, general cognitive capac-
ities, not specific to language and perhaps not specific to humans, clearly play
a role in shaping our knowledge and use of language, although the exact role
these capacities play and how they interact with (and can be distinguished from)
the language-specific aspects of our linguistic capacities are difficult questions.
Together, these three factors constitute the human language faculty. It can be
extremely difficult to distinguish the specifically linguistic aspects that contrib-
ute to forming the language faculty from more general cognitive abilities, but
the distinction can certainly be made in principle and is of course very important
for our general cognitive theory of language.
In these terms, Universal Grammar (UG) is the theory of the first factor
which makes up the language faculty: our innate genetic endowment for lan-
guage. As already mentioned, UG is assumed to play a central role in language
acquisition. It is also vital in helping us to understand how we can make sense
of the idea that specific languages are actually variants of a single type of entity,
language. Our goal here, then, is to look at one part of UG: how words combine
to form sentences. In so doing, we will construct the theory of syntax.
We can now make an important distinction. Language can be seen from an
internal, individual perspective, I-language, or it can be seen as external to the
individual, E-language. Here we are going to focus on I-language, which arises
from the interaction of the three factors just introduced. This is a natural approach,

IanRoberts_9781316519493_Introduction.indd 10 05-09-2022 11:17:15

 Natural Language 11

given our interest in how syntactic structures arise in the mind. E-language is
in fact a more complicated notion, involving society, culture, history and so on.
Concepts like ‘English’ and ‘French’ in their everyday senses are E-language
concepts; it is for this reason that they do not, strictly speaking, form part of our
object of study. I-language is a natural object; E-language is not.
Taking I-language as the central notion treats language as part of individ-
ual psychology. This is a cognitive theory of language, because it intrinsically
involves the human mind. In fact, this theory of language could be part of an
overall theory of the mind. However, we would really like our theory of language
to be part of an overall theory of the brain; being an obviously physical object,
the brain is a natural object, and it is a clearer notion than ‘mind’ (in fact, philos-
ophers have worried for centuries about whether ‘mind’ is something physical,
but we don’t need to). Unfortunately, though, at present our understanding of
how brain tissue supports cognitive processes such as language, thought and
memory is extremely limited, and so we are unable to say very much about the
relation between the physical brain and cognitive processes. That’s why I will
continue to use the older, strictly speaking vaguer, term ‘mind’.
What does it mean to say our theory is a formal theory? A formal approach to
any kind of problem or phenomenon assumes discrete, systematic ways of form-
ing complex things out of simple things. For example, the alphabet is formal: its
twenty-six letters can be combined in various different, but more or less system-
atic, ways to form a very large number of words. Arithmetic is a formal system
combining numbers of various kinds and functions such as addition, subtraction,
multiplication, etc.
In formal syntax and formal semantics, we combine simple elements (roughly
words and their meanings) to form more complex elements (sentences and their
meanings). The modes of combination must be systematic and precise; as we
will see, this is a major part of the challenge of constructing such a theory.
A further natural question to ask is why we want a formal and cognitive theory.
The answer to this lies in certain general trends of thought both in psychology
and in philosophy of mind (at least in the English-speaking world). It is widely
believed that the best way to understand the mind is to think of it as a kind of
computer. Computers manipulate symbols according to formal instructions, i.e.
algorithms and programs. We can think of I-language in these terms as a piece
of cognitive software, a program run on the hardware of the brain (this raises
the intriguing question, which I will not go into here, of who or what wrote the
program). Therefore, our theory of I-language must be a formal theory. What we
are interested in is how our knowledge of I-language is represented in our mind.
If we can get an idea of this, which I believe we can from studying syntax, then
we gain a very important insight into the human mind; what it is to be human,
what it is to be you.
To conclude this general discussion, we will concentrate here on developing
one aspect of a formal, cognitive theory of I-language: the theory of syntax.
As already mentioned, syntax is concerned with how relatively simple units,

IanRoberts_9781316519493_Introduction.indd 11 05-09-2022 11:17:15

 12 introduction

words, are combined to form more complex units, sentences. This is taken to
be a cognitive capacity all humans have, as a reflex of their genetic endowment
which includes UG. UG interacts with linguistic experience in early life, and
with domain-general ‘third factors’ as described above, to give rise to mature
adult competence in one’s native language. This competence manifests itself
in the ability to produce and understand an unlimited number of sentences, and
to make judgements regarding both the syntax and the semantics of those sen-
tences. In the next chapter, we will look in detail at a concrete example of this
competence in action, for native speakers of English.
Just a final note: given what I’ve said, strictly speaking I shouldn’t talk about
‘English’, as it is not a scientific term since it does not designate a natural object.
Instead of talking about ‘English speakers’, I should really say something like
‘individuals who identify themselves and are identified in their cultural milieu
as possessing an I-language which corresponds to the E-concept “English”’. For
brevity I will gloss over this more accurate formulation and continue to talk about
‘English speakers’; the same goes for other E-language names (Italian, etc.).
More generally, the term ‘language’ will refer to that aggregate of I-languages
whose speakers recognise themselves and each other as belonging to the same
E-language community.
Now we can move from the rather general matters we have been considering
here to actually starting out on the study of syntax. As we delve more and more
into the detailed and intricate nuts and bolts of the theory of syntax in the chapters
to follow, the background issues we have discussed here should be kept in mind
as they form the overall conceptual underpinning to the theory we’ll develop.
But now for the nuts and bolts. Or, actually, the fish.

Exercises
1. Write a short paragraph (not more than half a page as an absolute
maximum; ten–twelve lines per concept would be ideal) to explain
in your own words what the following notions mean to linguists:
• The language faculty
• Formal approaches to the study of language
• ‘Language’ vs ‘languages’
• English

2. You are Martian Space Cadet λxCloverx231057. Write a report

to Martian Mission Control describing your observations having
abducted a featherless terrestrial biped with the aim of understanding
more about its noise-making capacities.
3. ‘Children are very good imitators and are able to learn when their
mistakes are corrected by their elders.’ How could you go about
refuting these claims about language acquisition?

IanRoberts_9781316519493_Introduction.indd 12 05-09-2022 11:17:15

 Further Reading 13

4. Martian Space Cadet λxCloverx231057 has discovered that the ter-

restrial featherless bipeds can communicate by means additional to
the noises produced by their vocal apparatus. Report these discover-
ies to Mission Control.

For Further Study and/or Discussion

1. No two people speak the same language.
2. Could we understand an alien language?
3. Are conlangs ‘real’ languages?

Further Reading
Adger, D. 2019. Language Unlimited. Oxford: Oxford University Press.
Pinker, S. 1994. The Language Instinct. New York: Harper Perennial Modern Classics.
Roberts, I. 2017. The Wonders of Language, or How to Make Noises and Influence
People. Cambridge: Cambridge University Press.

These three books all provide general introductions to language and linguistics assum-
ing no prior knowledge on the reader’s part. Each book has its own perspective: Adger
concentrates on how syntax is fundamental to human language and cognition, and so is
perhaps most in line with our concerns here. Pinker is a classic; it is a witty and engaging
introduction to linguistics with the emphasis on the relation between language and mind.
Roberts offers a comprehensive introduction to several different subfields of linguistics,
ranging from phonetics to historical linguistics; syntax is covered in Chapter 4.
Larson, R. 2010. Grammar as Science. Cambridge, MA: MIT Press, Unit 1. This book
is an introduction to syntax whose central idea is to present the field as an exercise in the
construction of a scientific theory. In this first unit, the central ideas regarding knowledge
of language and Universal Grammar are introduced in an attractive and original way.
Carnie, A. 2013. Syntax: A Generative Introduction. Oxford: Blackwell, Chapter 1.
Another very sound and well-written introduction to generative grammar. This first
chapter presents the central ideas behind syntactic theory, along with a section on dif-
ferent approaches to syntax (something I do not attempt here). Isac, D. & C. Reiss.
2008. I-language. Oxford: Oxford University Press, Chapters 1 and 3.As the title sug-
gests, the focus of this book is on the I-language approach to linguistics and syntax.
The first chapter starts with a presentation of linguistic data in order to illustrate the
approach. The third chapter considers different notions of language, similar to what has
been presented here but with a different overall slant; reading that chapter will comple-
ment this one nicely.
Freidin, R. 2012. Syntax: Basic Concepts and Applications. Cambridge: Cambridge
University Press, Chapter 2.This chapter contains an excellent introduction to I-language,
the nature of grammaticality, native speakers’ ability to recognise deviant sentences as
an indication of knowledge of language, language acquisition and the argument from
the poverty of the stimulus, and, in particular, language production and comprehension.

IanRoberts_9781316519493_Introduction.indd 13 05-09-2022 11:17:15

 14 introduction

Adger, D. 2003. Core Syntax: A Minimalist Approach. Oxford: Oxford University

Press, Chapter 1.This chapter also gives an excellent introduction to the central concepts
behind generative grammar, including the notions of grammaticality and acceptability,
tacit knowledge, recursion, poverty of the stimulus and I-language.

Other good general introductions to the goals and assumptions of generative

grammar include the following:
Anderson, S. & D. Lightfoot. 2002. The Language Organ. Cambridge: Cambridge
University Press, Chapters 1 and 2.
Radford, A. 2016. Analysing English Sentences. Cambridge: Cambridge University
Press, Chapter 1
Smith, N. 2005. Chomsky: Ideas and Ideals. Cambridge: Cambridge University Press,
Chapter 1.

IanRoberts_9781316519493_Introduction.indd 14 05-09-2022 11:17:15

 1 Tacit Knowledge (or: Several Things You
Didn’t Know You Knew about English)

1.1 Introduction
In this chapter, we begin our study of the theory of syntax. As we will
see later, there is no end to syntax. It is also rather difficult to discern where it
begins. Here we will make a start by doing two things. First, you’ll find out some
rather surprising facts about your knowledge of English, things you didn’t know
you knew. This will give you a concrete illustration of your competence in English.
Second, by carrying out a little translation exercise, we’ll get a first glimpse of what
makes languages similar to one another and what makes them differ.
We begin with very simple sentences (see (1)) and build up to rather strange
and complex ones. It’s not necessary to follow every detail of the discussion
here; all the technical ideas are presented again in later chapters. Our goal here
is to demonstrate the nature of your tacit knowledge of English syntax. We can
do it with just one word: fish. This chapter can also be skipped and returned to
later (e.g. after Chapter 7).

1.2 Some Fish

Let’s start with a single, rather banal-looking English word. We’ll
see that we can build some quite striking and intriguing sentences with it. The
word is fish.1 So our first sentence is:
(1) Fish!

Like quite a few basic words in English, fish is actually ambiguous. It can be
understood either as a verb or as a noun. As a noun, it refers to a class of aquatic
animals; as a verb, it refers to the activity of hunting those aquatic animals.
We indicate this ambiguity in the standard way, by surrounding the word with
square brackets, and writing the ‘category label’ (Noun/Verb) as a subscript to
the left bracket. So, the sentence in (1) actually has two distinct representations:

1
Thanks to my good friend and colleague Professor Robert Freidin of Princeton University for
these examples. The implications of the fish sentences are discussed and explored in more detail
in Chapter 1 of Freidin (2012), . Freidin’s discussion is more detailed than here, although it
focuses exclusively on English. The exercises given there are also well worth trying.

15

IanRoberts_9781316519493_C001.indd 15 06-09-2022 11:48:49

 16 1 tacit knowledge

(2) a. [Noun Fish ]

b. [Verb Fish ]

The brackets are just a way of saying ‘what is inside here is a Noun/Verb’. A
representation like that in (2) is called a labelled bracketing.
The fact that fish, like many other words including cook, book, police, report,
promise and many others, is ambiguous between a verb and a noun is a conse-
quence of the fact that English has very few inflectional endings to mark gram-
matical information of various kinds. If we compare English with a more richly
inflected language such as Italian, we see that the two versions of fish correspond
to two differently inflected words:
(3) a. [Noun fish ] = Italian pesce
b. [Verb fish ] = Italian pescare

It is quite easy to see that the Italian words share the root pesc- and distinct
inflections: -e, indicating a singular noun; and -are, indicating the infinitive of
a first-conjugation verb (the infinitive is the basic form of a verb with no tense
marking; ‘first conjugation’ refers to an arbitrary morphological class of verbs
in Italian: there are four conjugations altogether, as we will see below). To cut
a very long historical story short, English has largely lost its equivalents of -e
and -are, and so we are left with the equivalent of the ambiguous root pesc- (you
may also note a family resemblance between pesc- and fish; this is because both
are ultimately derived from an ancient root in the common ancestor language of
English and Italian, Indo European).
The ambiguous single word in (1) and (2) constitutes an entire sentence on its
own. More precisely, each interpretation of fish constitutes a sentence, so really
there are two distinct sentences here. We can represent them as follows (here,
again following standard conventions, ‘Noun’ is abbreviated as N and ‘Verb’ as
V, and S stands for ‘Sentence’):
(4) a. [S [N Fish ]]
b. [S [V Fish ]]

Interpreted as in (4a), the sentence draws attention to the presence of a single fish or
group of fish (note that fish is one of a relatively small group of English nouns that
is identical in the singular and the plural: the plural form fishes exists, but it denotes
several species of fish, not several individual fish of the same species). Interpreted
as in (4b), it is an imperative, indicating an order given to go fishing. Here there
is an implicit second-person pronoun, since an order is naturally understood as
being addressed to an interlocutor, so we could elaborate (4b) as in (5):
(5) [S [N You ] [V fish ]]

Here the pronoun (a kind of noun, hence N) is ‘understood’; we take that to

mean that it is present in the syntactic and semantic representations of the sen-
tence, but not in the phonological representation, what you actually hear or say.
This is our first indication that there is more to syntax than meets the ear.

IanRoberts_9781316519493_C001.indd 16 06-09-2022 11:48:49

 1.2 Some Fish 17

Two occurrences of the word fish also make up a grammatical sentence:

(6) Fish fish.

The natural interpretation of this sentence is to treat the first fish as a noun and
the second one as a verb, the combination again forming a sentence. We repre-
sent this as in (7):
(7) [S [N fish ] [V fish ]]

In this sentence, the noun fish is the subject, understood as carrying out the
action performed by the verb. The verb fish indicates the action the subject car-
ries out. So the sentence means “Fish fish stuff”. As this rough gloss indicates,
there is an implicit direct object here, indicating, somewhat vaguely, what is
being fished for, what undergoes the action of fishing.
Alternatively, we can understand the first fish as a verb, and the second fish
as a noun. This is easier to see if we add an exclamation mark to (6) (Fish fish!),
corresponding to a different intonation pattern in speech. Then the sentence has
the structure in (8):
(8) [S [V fish ] [N fish ]]

Here, as in the verbal interpretation of the single-fish example in (1), the verb
is understood as an imperative and so there is a deleted second-person pronoun
(you) as subject:
(9) [S YouN [V fish ] [N fish ]]

The second fish, the noun, is an explicit direct object, indicating that fish are
what is fished.
In Italian, where the ambiguities of the English roots are clarified by inflec-
tional endings, the two interpretations of (6) can be rendered as in (10):
(10) a. I pesci pescano. (= (7))
b. Pesca pesci! (= (8/9))

In addition to disambiguating the English sentence in (6), these translations

show us that there might be more silent material in the English examples than
just the deleted pronoun in (8/9). The first word of (10a), i, is the masculine
plural form of the definite article, i.e. ‘the’. It is obligatory in this context
in Italian. What it shows us is that the subject is more than just a noun, but a
phrasal category of which the noun is the most important member: a Noun
Phrase, or NP. The basic function of the definite article in both English and
Italian is to indicate that the noun refers to some unique known existing entity,
in this case what philosophers would call the ‘natural kind’ fish. Since there
is no fundamental meaning difference between (6) with the structure and inter-
pretation in (7) and (10a) (in other words, since (10a) is a reasonable translation
of the English sentence), the noun fish in (7) also denotes the unique, known,
existing, natural kind fish. So we could posit a silent “the” in the subject of (7)

IanRoberts_9781316519493_C001.indd 17 06-09-2022 11:48:49

 18 1 tacit knowledge

and, correspondingly, an NP rather than just an N. Then (7) lines up exactly

with (10a):
(11) [S [NP the fish ] [V fish ]]
[S [NP I pesci ] [V pescano ]]

Why go to this trouble to line up English and Italian in this way? There are two
reasons. First, the English and the Italian sentences mean the same thing. Although
we tend in our everyday lives to pay more attention to cultural differences than
to cognitive similarities, it seems reasonable to assume that English speakers and
Italian speakers are cognitively alike (we could perhaps say that English speakers
and Italian speakers are I-alike but E-different, thinking of the distinction made
in the previous section between I-language and E-language). In that case, the
semantic representations of the two sentences ought to be the same. It therefore
simplifies the connection (more technically, the ’mapping’) between the syntax
and the semantics if the respective syntactic representations ‘line up’, as in (11).
Secondly, we are attempting to develop a theory of Universal Grammar (UG, see
the Introduction); we therefore want our syntactic representations of comparable
(i.e. semantically matching, or at least highly similar) sentences in different lan-
guages to be as uniform as possible. Of course, it is not always possible to achieve
a parallel as straightforward as that depicted in (11), but we should strive towards
this goal as part of the UG project. Ideally, then, we would want (a) the semantic
representations of sentences which mean the same thing to be identical across
languages, (b) the syntactic representations of such sentences to be as uniform as
possible across languages while, of course, recognising (c) that their phonologi-
cal and phonetic shapes, including the linear order of the words in a sentence, are
different. The syntax and semantics are ‘externalised’ with different phonologies
in different languages, but the syntax is (near-)uniform and the semantics entirely
uniform. We will come back to these issues in Chapters 7 and 8.
The Italian sentence in (10a) tells us two more interesting things. First, the
noun pesci is plural: it has the plural ending -i (compare the singular pesce in
(3a)). In the English version, fish is plural too; we have already noted that fish
has no overt plural ending. If we substitute a more regular noun, one which
forms its plural with -s, we can see this:
(12) Boys fish.

Second, the verb pescano consists of the root pesc- and the ending -ano (again
compare the infinitive ending -are in (3b)). This is not the place to go into the
full details of Italian verbal morphology, but this ending contains the infor-
mation that the verb is first conjugation, present tense, indicative mood (it
makes a statement of fact) and third-person plural (‘they’). That’s quite a bit
of information in just three overt phonemes. The -a- part of the ending recurs
in the infinitive ending, and this can be seen as the marker of first conjuga-
tion. The -no part of the ending shows up in many other tenses and is clearly
marking third-person plural. Present tense and indicative mood are arguably

IanRoberts_9781316519493_C001.indd 18 06-09-2022 11:48:49

 1.2 Some Fish 19

just ‘understood’. As we have already seen, saying something is ‘understood’

is a way of saying that it is part of the semantic representation of the sentence,
where there is no pronounced element that represents it. In order to simplify
the syntax–semantics connection, we want the syntactic representation to reflect
the semantic interpretation as fully as possible. Since we have silent syntactic
elements, perhaps we can also have silent morphological elements. So we are
led to think that pescano has the structure in (13):
(13) pesc-a- Indicative - Present -no

Furthermore, we can, and for the sake of consistency must, put labelled brackets
around each part of the structure:
(14) [V [V pesc- ] [Conj -a- ] [M Indicative ] [T Present ] [Agr -no ]]

The labels are fairly straightforward: ‘V’ indicates that both the root and the
whole thing form a verb; ‘Conj’ indicates the conjugation-class marker ‘M’ indi-
cates mood (the traditional grammatical category indicating whether a sentence
describes a state of affairs believed to be true or not), ‘T’ indicates tense and
‘Agr’ indicates agreement: the verb is third-person plural because the subject, i
pesci, is third-person plural.
Now, if the Italian verb has a structure like (14), and the Italian sentence in
(10a) is an accurate translation of the English sentence in (7) and our assumption
that English and Italian verbs should be minimally different means that we want
structures to ‘line up’ across languages, our logic leads us to posit something
like (15) as the structure of the verb fish in (7):
(15) [V [V fish- ] [Conj ?? ] [M Indicative ] [T Present ] [Agr 3Pl ]]

Here, Conj, M, T and Agr are all silent. But our logic leads us to conclude that
they are all structurally present. Now we come up against the limits to this lin-
ing-up-in-the-name-of-UG approach. It seems reasonable to take the verb fish in
(7) to be indicative in mood, present in tense and third-plural in agreement, but
the conjugation-marking seems to impose an idiosyncracy of Italian morphology
onto English, moreover an idiosyncracy that appears to have no semantic corre-
late. So perhaps we should eliminate ‘Conj’ from (15). The other silent endings
are, however, justified by the semantics and our universalising methodology.
Both tense and agreement endings do audibly appear on English verbs (mood
may too, but this is a little trickier and so I’ll leave it aside). If we make the sub-
ject of our example in (12) singular, but keep the present tense, we have:
(16) The boy fishes.

Here the verb has the ending -es. If we make the verb past tense, we have:
(17) The boy fished.

Here the verb has the ending -ed. So we could give the verbs in (16) and (17) the
structures in (18a) and (18b) respectively:

IanRoberts_9781316519493_C001.indd 19 06-09-2022 11:48:49

 20 1 tacit knowledge

(18) a. [V [V fish- ] [M Indicative ] [T Present ] [Agr -s ]]

b. [V [V fish- ] [M Indicative ] [T -ed ] [Agr 3Sg ]]

Here, we have dropped the Italocentric Conj, for the reasons given above; Mood
remains silent, but T is realised as -ed in the past tense and Agr is realised as -es in
the third-person singular (agreeing with the third-person singular subject the boy)
in the present tense. So there is justification for lining up these aspects of English
verbal inflection with Italian. The fundamental difference between English and
Italian has already been mentioned: in the history of English many markers of
inflection have lost their overt phonological representation, while their Italian
counterparts have retained most of theirs. So English has more silent inflections
than Italian. Semantics and UG together lead us to postulate those inflections
nonetheless.
We can also observe another interesting parallel between English and Italian
brought out in (16). Here, unlike in (12) where boys is plural, an article must
appear with boy. Compare:
(19) *Boy fishes.

The article doesn’t have to be definite. In fact, a range of determiners of various

kinds can appear here:
(20) A/This/That/One/Every/Each/No/Some boy fishes.

The fact that where boys is plural as in (12) no audible determiner is required,
combined with the unique, known, existing ‘natural kind’ interpretation assigned
to boys here, supports the idea that there is a silent determiner. This in turn sup-
ports the lining-up of English and Italian seen in (11).
We can now replace (11) with a fuller lined-up, (nearly) uniform English and
Italian structure, in which the determiners, silent and overt, are represented by
the category D:
(21)
[S [NP [D the ] [N [N fish-[Num Pl ]]] [V [V fish- ] [M Indicative ] [T Present ]
[Agr 3Pl ]] ]
[S [NP [D I ] [N [N pesc-[Num i ]]] [V [V pesc- ] [Conj -a- ] [M Indicative ] [T Present ]
[Agr -no ]] ]

Here I’ve added the specification of number on the noun (the ending Num);
plural number is silent in the case of fish in English as we have observed. We
could add further specifications: Person on the noun (3rd here), number marking
on the Italian article i (implying, by our now-familiar reasoning, that English
the has a silent plural ending). Furthermore, the Italian articles show gender
marking: article i is masculine. So we should add that specification to the Italian
sentence. Should we add it to the English one? The Italian gender is grammati-
cal, not semantic. Pesce is a masculine noun, but not all fish are male (or there
would be no more fish, since they reproduce sexually). In Italian, all nouns have
masculine or feminine gender, and this has no semantic basis at all in many
cases: ‘sun’ is masculine (il sole); ‘moon’ is feminine (la luna), ‘sincerity’ is

IanRoberts_9781316519493_C001.indd 20 06-09-2022 11:48:49

 1.2 Some Fish 21

feminine, ‘feminism’ is masculine, ‘violence’ is feminine, ‘socialism’ is mascu-

line and so on. Try as one might, it is hard to find the slightest metaphorical or
cultural basis for this: the words are just arbitrarily assigned to two distinct mor-
phological classes. Hence gender-marking has no consequences for the seman-
tic representation. So perhaps, as with the Conj-ending on verbs, it would be
Italocentric of us to impose Italian genders on English nouns or articles.
One last point on the structure in (21): are M, T and Agr part of the verb V,
as indicated there, or, analogously to D, part of the Verb Phrase, VP? We could
think that V and N each have their ‘satellite categories’ which carry grammatical
information often realised as inflection or as ‘little words’ like the articles the
and i: the satellites of N are D and Num, and the satellites of V are M, T and
Agr. Semantically, though, Mood and Tense arguably modify the sentence, by
indicating whether it makes a true statement and relating the time of the situation
described to the speech time (this is what Tense does). So M and T look like
they belong to S. Agr seems to connect the subject NP and the verb; as such, this
too looks like a sentence-level, rather than a verb-level, property. We leave the
question of the status of M, T and Agr open for now; later we will see that M
and T are indeed parts of S, while Agr may be a relation rather than a category.
Let us now go back to the structure in (9), which I repeat here for convenience:
(9) [S YouN [V fish ] [N fish ]]

We saw the Italian translation of (9) in (10b):

(10b) Pesca pesci!

The full representations of both examples, with silent endings indicated, are
given in (22):
(22) a. [S [NP you ] [VP [V [V fish- ] [M Imperative ] [T Present ] [Agr 2 ]] [[D some ]
[N [N fish-[Num Pl ] ]
b. [S [NP tu ] [V [V pesc- ] [Conj -a- ] [M Imperative ] [T Present ] [Agr 2Sg ]] ]
[NP [D i ] [N [N pesc-[Num i ]]]

Both examples have an understood second-person subject. English you can be

either singular or plural, and so (8/9) can be interpreted as giving an order to a
single person or to a group of people. The Italian subject is singular, as the ver-
bal ending shows (the plural is pescate), and the understood subject is the singu-
lar pronoun tu. M is specified as Imperative: these sentences do not make factual
statements but give orders. Imperative is the ‘ordering Mood’, while, as we saw,
Indicative is the ‘stating Mood’. The object NPs are similar to the subject NPs
in (21), except for the Ds. The Italian example does not have a definite article
and the English silent D is ‘some’. Both of these points are related to the fact
that the object here has a subtly different interpretation from the subject in (21).
In (21), the subject names a unique, known, existing ‘natural kind’; this inter-
pretation is known as the generic reading of the NP. The object does not denote
a natural kind; it merely denotes some arbitrary, not necessarily known and not

IanRoberts_9781316519493_C001.indd 21 06-09-2022 11:48:49

 22 1 tacit knowledge

necessarily existing group of fish (the fishing expedition could be completely

unsuccessful with no fish at all actually being caught). This is an indefinite
interpretation of the NP. Hence the definite D, silent in English, overt in Italian,
does not appear. I have glossed the silent indefinite D as ‘some’ in English.
From the above, we can see that our simple two-fish sentence in (6) has a
great deal more to it than meets the eye, or ear. The Italian translations are very
revealing, especially when we try to line them up as closely as we can to their
English counterparts. We see that both verbs and nouns have satellite cate-
gories which may be realised as inflectional endings or as articles such as the
and i, and are frequently silent. We also see that not all satellite categories are
universal: Italian has grammatical gender on nouns and Conj-marking on verbs,
but there is no reason to posit counterparts to these in English. Of course, a very
natural, and, it turns out, very difficult question is whether the other categories
such as D, M, T and Agr are universal. We are in no position to answer this
question at the moment, but we can observe that it seems reasonable to take
these categories to be shared by Indo-European languages such as Italian and
English.
So we see that simple sentences like English (6) and Italian (10) can tell us
a lot. In the next section, we’ll take this line of thought a bit further, with some
more complicated sentences (but still using the same English word).

1.3 More Fish and Some Relative Clauses

In this section, we’ll look at three-, four- and five-fish sentences.
These sentences are simultaneously silly and serious; they’re silly because they
describe rather unlikely situations (which you can nonetheless imagine), and
they’re serious because they can reveal a lot about our linguistic competence.
So let’s look at what happens when we combine three occurrences of fish:
(23) Fish fish fish.

The Italian translation of (23) is (24):

(24) I pesci pescano pesci.

Both of these are full sentences where the first ‘fish’ is the subject, the second the
verb and the third the direct object. Clearly, what is added to the two-fish exam-
ples discussed above is an explicit direct object, compared to (7), and an explicit
subject, compared to (9). Traditional grammars tell us that a complete sentence
consists of a subject, what the sentence is about, and a predicate, which says
something about the subject. The predicate contains the main verb and the direct
object. Thus we could represent (23) as follows (leaving aside for the moment
the representation of the silent inflections):
(25) [S [N Fish ] [Predicate [V fish ] [N fish ]]]

IanRoberts_9781316519493_C001.indd 22 06-09-2022 11:48:49

 1.3 More Fish and Some Relative Clauses 23

Since the verb is the most important part of the predicate, we can give a more
strictly categorial representation of (25) as follows:
(26) [S [N Fish ] [VP [V fish ] [N fish ]]]

Here VP, the Verb Phrase, is the category which functions as the predicate.2
The ‘full’ representations for (23) and (24), complete with specification of
silent inflections and other satellite categories, are shown in (27):
(27) a. [S [NP [D the ] [N [N fish-[Num Pl ]]] [VP [V [V fish - ] [M Indicative ] [T Present ]
[Agr 3Pl ]] [NP [D some ] [N [N fish-[Num Pl ]]]]
b. [S [NP [D I ] [N [N pesc-[Num i ]]] [V [V pesc- ] [Conj -a- ] [M Indicative ] [T Present ]
[Agr -no ]] [NP [D i ] [N [N pesc-[ Num i ]]]]

These representations are almost exactly the expected combination of (21) with
an explicit direct object as in (22). Again, the subject and object have differ-
ent structures and interpretations, with the subject being generic and the object
indefinite.
Now let’s try four fish:
(28) Fish fish fish fish.

This sentence is both grammatical and interpretable, but, unlike the two-fish and
three-fish examples it requires a little thought. The intonation with which it is to
be read is important: it should be ‘ish FISH fish [short pause] FISH’, with stress
on the second fish and the strongest stress on the final fish. The structure here
is significantly more complex than our earlier examples. For this reason, from
now on I will mainly give simplified representations of these more complex
examples, leaving aside the silent inflectional material (although we should not
forget that it is there).
What does (28) mean and what is its structure? Again, the Italian translation
can give us an important clue:
(29) I pesci che i pesci pescano pescano.

We have seen that pesci is unambiguously a noun and pescano is unambigu-

ously a verb. At the absolute minimum, then, (29) has the structure N che N V
V. The English example in (28) lines up with this, except that there is no coun-
terpart to che. The Italian word che usually translates as that. We can add that
to (28), to get:
(30) Fish that fish fish fish.

The markers that and che introduce a relative clause, a clause which modifies a
noun; for our purposes, the only real difference between that and che is that che
appears obligatorily in this context, while that can be ‘dropped’ (i.e. silent). So

2
We will look in detail at the relation between grammatical categories like NP and VP and gram-
matical functions like subject and predicate in Chapter 2 of Volume II.

IanRoberts_9781316519493_C001.indd 23 06-09-2022 11:48:49

 24 1 tacit knowledge

the second two fish, the sequence noun-verb, constitute a relative clause mod-
ifying the first noun fish (known as the head of the relative). The fourth fish is
the verb of the main clause, with an implicit direct object just as in the two-fish
example with the structure in (7). We can thus make a first pass at a structure for
(28) as follows:
(31) [S [N fish ] [Relative clause fish fish ] [VP [V fish ]]]

The relative clause is a kind of subordinate clause. In fact, we can tell that the first
fish inside the relative clause is the subject and the second fish is the verb of the
predicate. So we have an NP-VP structure here (remember that the presence of the
overt definite article in Italian shows us that the subject is an NP, not just a noun):
(32) [S [N fish ] [Relative clause [NP [N fish ]] [VP [V fish ]]] [VP [V fish ]]]

The main clause, i.e. the independent clause containing the relative clause, indi-
cated as S here, must also have a subject NP, not just N. This NP includes the
relative clause, since the relative clause modifies the noun, so we have (33):
(33) [S [NP [N fish ] [Relative clause [NP [N fish ]] [VP [V fish ]]]] [VP [V fish ]]]

Both the main clause and the relative have a subject and a predicate. Identifying
predicates with VP again, (33) tells us that each VP contains just the verb fish. In
the case of the main clause, this may be correct, since the object here is implicit
and has a rather vague interpretation (roughly ‘stuff’). But the object of the rel-
ative has a completely different, and very precise, interpretation: ‘fish’. Here
we see an important characteristic of relative clauses: there is always something
missing, often referred to as a ‘gap’, which corresponds semantically to the head
of the relative (which we could then call the ‘filler’). Since we know that syntac-
tic representations contain silent elements (actually, we are beginning to see that
they mostly contain silent elements) and we want to keep the syntax-semantics
mapping as straightforward as possible, the best way to account for the interpreta-
tion of the gap is that it is a silent copy of the filler. So we replace (33) with (34):
(34) [S [NP [N fish ] [Relative clause [NP [N fish ]] [VP [V fish ] [NP [N fish ]]]]] [VP [V fish ]]]

We said above that English that, unlike its Italian counterpart che, can be
‘dropped’. This implies that sentences with and without that are synonymous: that
and che are just meaningless syntactic markers (saying ‘what comes next is a rel-
ative clause’ in our examples). However, (30) has another interpretation that (28)
doesn’t have. In (28) the gap inside the relative clause is the direct object, as (34)
shows; it means ‘fish that get fished fish stuff’. But (30) allows a further interpre-
tation, where the gap can be interpreted as the subject of the relative clause: ‘fish
that fish other fish fish stuff’. This example has the following structure:
(35) [S [NP [N fish ] that [Relative clause [NP [N fish ]] [VP [V fish ] [NP [N fish ]]]]] [VP [V fish ]]]

Here it is the subject of the relative that is the silent copy of the head, and
the marker that is obligatory. The Italian translation of this English sentence is
unambiguous and should be compared to the one in (29):

IanRoberts_9781316519493_C001.indd 24 06-09-2022 11:48:49

 1.4 Really Challenging Fish 25

(36) I pesci che pescano pesci pescano.

It is now easy to interpret a five-fish example:

(37) Fish fish fish fish fish.

As we saw, the four-fish example in (28) had an implicit direct object in the main
clause (‘Fish fish fish fish stuff’). The five-fish example in (37) has an explicit
direct object in the main clause. We can therefore give the structure as in (38),
and the Italian translation in (39):
(38) [S [NP [N fish ] [Relative clause [NP [N fish ]] [VP [V fish ] [NP [N fish ]]]]] [VP [V fish ]
[NP [N fish]]]]

(39) I pesci che i pesci pescano pescano pesci.

By now we may begin to wonder just how many repetitions of the word fish (in
the phonology, the subpart of the syntactic structure that is actually pronounced)
can give rise to a grammatical, interpretable sentence. It is clear that the step
from three fish to four fish, involving as it did the introduction of a relative
clause into the structure, places a burden on comprehension. But this burden can
be readily overcome and the sentence is fully comprehensible; in fact we can
recognise that it describes a rather odd state of affairs, the very silliness of the
sentence proves that we can understand it. We see, then, that the comprehension
burden does not increase linearly as we add extra words; the structural complex-
ity of the four-fish sentence is significantly greater than the three-fish one, and
this is reflected in our initial hesitation in understanding it.
In the next section, we’ll see some really challenging examples as we increase
the number of fish still further.

1.4 Really Challenging Fish

Here we’re going to look at six- and seven-fish sentences. These are
really hard to understand, especially the six-fish one (it’s harder than seven, as
we’ll see). You don’t have to follow every detail of the analysis of these sen-
tences here, as long as you get the main point, which comes after (54). Skip ahead
to that discussion if the fish tire you out. Or, fish-like, just go with the flow.
The non-linear increase in comprehension difficulty becomes very sharp
indeed when we move from five fish to six fish. Here is the six-fish example:
(40) Fish fish fish fish fish fish.

Most people are completely stumped by this sentence at first sight; it appears to
be simply a random repetition of the word fish with no meaning (and so, perhaps,
no structure) at all. But in fact this isn’t true, as we’ll see below. What is striking
though is that seven fish is easier to interpret than six:
(41) Fish fish fish fish fish fish fish.

IanRoberts_9781316519493_C001.indd 25 06-09-2022 11:48:49

 26 1 tacit knowledge

We know from (28) that a sequence of three fish can be a relative clause with
an object gap, ‘fish that are fished by other fish’. In (28), the relative clause is in
the subject position. But relative clauses can be in object position too (in fact,
(28) can be interpreted that way too, a point that was left aside above). So (41)
has the approximate structure in (42a), given in fuller detail in (42b) (where RC
stands for relative clause, and the structure is presented as a tree diagram, as this
makes it easier to see the relations among the fish; we’ll look at tree diagrams
more systematically in the next chapter):
(42) a. [ [ Fish fish fish ] [ fish [ fish fish fish ]] ]
b.

Pronounced with the right intonation (‘Fish FISH fish [pause] fish fish FISH
fish’), this example is fairly easy to understand, especially in the light of the
four- and five-fish examples. But none of this makes (40), with just six fish, any
easier to understand.
In order to see what the structure, and hence the interpretation of (40) is, we
must go back to our earlier, simpler examples. Here, once again, is the two-fish
example with noun-verb interpretation, repeated from (7):
(7) [S [N fish ] [V fish ]]

Of course, we now know that N and V here are contained in their respective
NPs and VPs. Relative clauses are really a kind of sentence, as we pointed out
above. So let us label them as S from now on. We go from two fish to four fish
by putting the structure in (7) inside the subject NP:
(43) [S [NP [N fish ] [S [N fish ] [V fish ]] ] [VP [V fish ]]]

Actually it’s not quite accurate to say we put (7) inside the subject NP; really, we
put the three-fish sentence, with the structure seen in (26), inside the subject NP
and make the object a silent version of the head of the relative:

IanRoberts_9781316519493_C001.indd 26 06-09-2022 11:48:50

 1.4 Really Challenging Fish 27

(44) [S [NP [N fish ] [S [NP [N fish ]] [VP [V fish ] [NP [N fish ]]]]] [VP [V fish ]]]

On the basis of (44), we can add two more pronounced fish by inserting a relative
clause inside the subject NP of the relative clause, i.e. after the second [N fish ].
This gives the string of six fish in (41), and shows us what the structure is:
(45) [S [NP [N fish ] [S [NP [N fish [S [NP [N fish ] [VP [V fish ] [NP [N fish ]]]]]]] [VP [V fish
] [NP [N fish ]]]]] [VP [V fish ]]]

If we add some relative markers, preferably which (another relative marker in

English alongside that), we can just about start to make sense of (45):
(46) Fish which fish which fish fish fish fish.

So (41) is, despite initial appearances, a grammatical and interpretable sentence:

the string is N N N V V V. It might be easier to see this if we use some different,
and more varied, vocabulary. Let’s start with (47):
(47) The mouse the cat chased died.

By now, we know a relative clause when we see one. The essential structure of
(47) has a relative clause inside the subject containing an object gap:

(48) [S [NP the mouse [S [NP the cat [VP chased [ the mouse ]]]]] [VP died ]]

(48) is quite easy to understand. The analogue to (41), but with different words, is (49):

(49) [S [NP the mouse [S [NP the cat [S [NP the dog [VP bit [NP the cat ]]]]]]] [VP chased
[NP the mouse ]]]]] [VP died ]]

Again, the jump from (48) to (49), caused by simply adding another relative
clause modifying the cat, poses severe difficulties of comprehension. Adding
some relative markers (that again) makes (49) considerably easier to understand:
(50) The mouse that the cat that the dog bit chased died.

This phenomenon is known as centre-embedding; it seems that embedding

more than one relative clause inside another without an overt relative marker
creates real comprehension difficulties. Nonetheless, such sentences are in fact
grammatical and interpretable. The rules of English syntax allow us to embed
relatives inside one another and, under the right conditions, to ‘drop’ relative
markers. The fact that embedding more than one relative clause inside another
causes comprehension difficulties is a matter of performance; our short-term
memory seems unable to ‘keep track of’ the structure. But since these sentences
are generated by the rules of English syntax they are grammatical; so we desig-
nate them as grammatical, but, because of the comprehension load they impose,
unacceptable. Grammaticality is a competence notion, while acceptability is a
performance notion. Most of the time, grammaticality and acceptability coin-
cide, but in cases like centre-embedding we can see the distinction. Another
famous example where the two notions come apart is the sentence Colourless

IanRoberts_9781316519493_C001.indd 27 06-09-2022 11:48:50

 28 1 tacit knowledge

green ideas sleep furiously; most people find this sentence unacceptable for the
perfectly good reason that it doesn’t make any sense. It is, however, syntac-
tically well-formed, as its exact formal counterpart Revolutionary new ideas
spread quickly shows. Both sentences are grammatical in that they conform to
the rules of English syntax; the first one is semantically anomalous (in fact, it
is self-contradictory) and so unacceptable. This contrast also shows that syntax
and semantics are distinct in that a sentence can be syntactically well-formed but
semantically ill-formed; we will see more examples of this type in Section 8.3.
Of course, examples with centre-embedding like (50) become still more
difficult to understand when the same phonological word occurs six times over,
as in (40). There is, in principle, no difference in grammaticality between (40)
and (50) but a real difference in acceptability (they just go from bad to worse).
But now we can line (40) up with the variant of (50) without the relative mark-
ers, and we can see how (40) works, and see that it is in fact grammatical:
(40) Fish fish   fish fish fish fish.

(51) The mouse the cat the dog bit chased died.

Finally, here are the easier versions complete with the relative pronouns which
and that inserted in the appropriate places:

(46) Fish which fish which fish fish fish fish.

(50) The mouse that the cat that the dog bit chased died.

These examples have the structures shown in (49) and (45) respectively.
Relative clauses can be put one inside another without limit. They can just go
on and on. Consider (52):
(52) This is [NP1 the guy [ S1 who loved [NP2 the girl [S2 who befriended [NP3 the boy
[S3 who lived ]]]]]].

For those familiar with the Harry Potter stories, each NP here describes one
of the central protagonists by means of modification by a relative clause, each
relative clause is inside the NP it modifies, and each NP – except NP1 – is
embedded in the next relative clause. NP3 describes Harry Potter himself, the
boy who lived. NP2 describes Hermione Granger, the girl who befriended the
boy who lived, and NP1 describes Ron Weasley, the guy who loved the girl who
befriended the boy who lived. Relative clauses give language a great deal of its
expressive power, by making it possible to modify nouns using modifiers that
can be as complicated as we want: in principle, there is no upper limit to the
number of times one relative clause can be embedded inside another. This is
reflected in nursery rhymes like ‘The House that Jack Built’: the dog who bit
the cat who ate the rat who lived in the house that Jack built. Centre-embedded
relatives can also, in principle, be constructed without limit. The problem is
that, in practice, as we have seen, they quickly become very hard to understand.

IanRoberts_9781316519493_C001.indd 28 06-09-2022 11:48:50

 1.4 Really Challenging Fish 29

Compare the following centre-embedded sequence with the excerpt from ‘The
House that Jack Built’ just given:
(53) [ The rat [ the cat [ the dog [ bit ]] [ ate ]] [ lived in the house that Jack built ]

All of this leads us to three questions:

(54) a. What is the highest number of repetitions of the word fish alone that
­constitutes a grammatical English sentence?
b. Have you seen these sentences before?
c. Where did you learn these facts about English?

To answer (54a), we could keep on trying longer and longer fish sentences.
There is certainly no reason to take seven, the most we have seen here (see (41)
and the structures in (42)), to be an upper limit. If we look at the examples with
even numbers of fish, (28) and (40), we see that they have the general form in
(55):
(55) a. N1 + N2 + V2 + V1 (four fish, (28))
b. N1 + N2 + N3 + V3 + V2 + V1 (six fish, (40))

(This is not their exact structure, as we have seen, but these simplified representa-
tions make our point here.) Each pair Nn + Vn, for n = n forms a relative clause
modifying Nn-1. Hence, as Freidin (2012:13) points out ‘any sentence containing
an even number of fish from four onwards will have at least one grammatical
representation’. Furthermore, although each relative clause (i.e. each pair Nn +
Vn in (55) for n > 1) has an object gap, as we saw above (see (44) and (45)), V1
does not have an object. But of course it could have, and this object could be fish.
This is what got us from four fish to five fish and from six fish to seven fish. To
quote Freidin (2012:13) again: ‘As this procedure [adding an object to V1, IR]
can be applied to any sentence with an even number (n) of fish, there will be a
corresponding sentence with an odd number (n+1) of fish … Hence any number
of fish will correspond to a sentence of English.’ In principle, then, a sentence
containing any number of repetitions of the word fish is grammatical. Our lin-
guistic competence allows us to understand such sentences, although more than
five iterations, with the exception of seven, give rise to sentences that are difficult
to understand in practice, i.e. sentences which may be unacceptable to varying
degrees. We are able to assign syntactic representations, more technically known
as structural descriptions, to the sentences, and from there recover the meaning.
The full representations of these sentences, as we saw in our discussion of the
Italian counterparts of the two- and three-fish examples, are quite complex. The
full representation of the seven-fish example in English would be (56):
(56) [S [NP [D the ] [N [N fish-[Num Pl ]]] [S [NP [D the ] [N [N fish-[Num Pl ]]] [VP [V [V
fish- ] [M Indicative ] [T Present ] [Agr 3Pl ]] [NP [D the ] [N [N fish-[Num Pl ]]]]]]
[VP [V [V fish- ] [M Indicative ] [T Present ] [Agr 3Pl ]] [NP [D some] [N [N fish-
[Num Pl ]]] [S [NP [D the ] [N [N fish-[Num Pl ]]] [VP [V [V fish- ] [M Indicative ]
[T Present ] [Agr 3Pl ]] [NP [D the ] [N [N fish-[Num Pl ]]]]]]

IanRoberts_9781316519493_C001.indd 29 06-09-2022 11:48:50

 30 1 tacit knowledge

You don’t need to pick your way through all the details of this representation in
order to see two things. First, most of the structure is silent. Second, sentences
of this kind are highly complex objects. But remember, and this is the really
important thing, we quite unconsciously compute such complex sentence struc-
tures all the time.
Concerning question (54b), it is highly unlikely that anyone would come
across sentences of this kind (outside of a linguistics class). But, as we have
seen, they can be fairly readily understood. It is extremely difficult to see how
any account of linguistic knowledge based on simply imitating, learning and
generalising routines (perhaps through processes of stimulus, response and rein-
forcement, as in behaviourist psychology) could account for our general capac-
ity to produce and understand, mostly instantaneously, sentences we have never
heard before and which may never have been uttered before.
Question (54c) is the question that goes to the heart of the matter. As native
speakers of English, we have very rich tacit knowledge, of even the silliest and
most awkward sentences, which we were almost certainly never ‘taught’ in any
meaningful sense of the word. Our experience in early childhood, combined with
our cognitive capacities (both language-specific and not), has made it possible
for us to perform the kinds of mental computations required to extract structure
and meaning from what appear to be merely strings of repetitions of the same
word. Of course, this ability is not restricted to native speakers of English: native
speakers of Italian can just as easily make sense of sentences like (24), (29), (36)
and (39), and the same exercise can be repeated, in principle, with any speaker
of any language.
The fish sentences clearly illustrate the distinction between competence
and performance. As we have mentioned, competence refers to a state of
knowledge, knowledge of an I-language, that a native speaker of a given lan-
guage (in roughly the ‘E’-sense as discussed in the Introduction) possesses.
Performance refers to the application of I-language in a given situation: com-
bined with other cognitive and social capacities, performance makes normal
speaking and comprehension possible. Freidin’s conclusion that ‘any number
of fish will correspond to a sentence of English’ clearly concerns competence;
the rules of English syntax are such that sentences of this kind can exist, but
of course, owing to performance limitations, no actual speaker could produce
or understand a sentence consisting of an infinite number of occurrences of
the word fish and nothing else. Since placing an upper bound on the number
of times one relative clause could be embedded in another would be arbitrary,
we regard the rules of English (and other languages) as allowing for this in
principle. Of course, the same applies to numbers: no individual can ever write
out an infinite number, although our mathematical competence tells us, and
this can of course be proved, that the series of natural numbers is infinite. In
Chapters 3 and 4, we will see the precise structure-building mechanisms that
achieve this result.

IanRoberts_9781316519493_C001.indd 30 06-09-2022 11:48:50

 1.5 Conclusion: What the Fish Have Taught Us 31

1.5 Conclusion: What the Fish Have Taught Us

In this chapter we have tried to outline and illustrate, with the con-
crete help of the fish sentences, what linguistic theory investigates. The central
goal is to elucidate what it means to be a competent speaker of a language. In
other words, linguistic theory is primarily about knowledge: what is it that a
person described as a native speaker of English (for example) knows? What
are the properties of individual I-languages and what are the properties of UG
that underlie them? Clearly this must involve the operations that are capable
of producing the kinds of structures we saw in relation to the fish examples. A
further central question is where this knowledge comes from: how do children
acquire their first language? Related to this is the question of how similar and
different languages are. We saw that we can ‘line up’ English and Italian quite
well, especially if we assume that English has a number of silent inflectional
‘satellite’ categories many of whose Italian counterparts are overtly pronounced.
In language-acquisition terms, this suggests that children are looking for evi-
dence – semantic, syntactic, morphological or phonological – for the presence
of these categories. Many complex issues arise as we try to pursue these ideas,
but we saw in connection with the Italian fish examples that the combination
of semantic evidence and UG predispositions of various kinds can lead us to
significantly limit the differences between English and Italian. As we said there:
semantics is universal, syntax is near-universal and languages differ primarily in
their morphophonologies. We will return to these ideas repeatedly in the chap-
ters to follow.
This is a formal, cognitive theory of linguistic knowledge. Each individual
over the age of five or so has a specific I-language which makes possible their
competence in their native language. This I-language corresponds more or less to
a sociocultural E-language construct such as English, Italian and so on. We also
saw that English and Italian speakers are I-alike but E-different: the I-languages
may turn out to be quite similar to each other, while many aspects of the cultures
borne by the corresponding E-languages are strikingly different (in ways that
people generally find fascinating). Given how little we understand about the
brain, as well as ethical restrictions on what kinds of experiments we can do on
healthy children or adults, we can only find out about what is inside the mind,
how an individual’s grammatical system works, from the observed outputs of
the I-language. This includes, quite simply, the things we say and hear (language
production and comprehension) along with judgements about grammaticality,
interpretation and so on.
We saw in the Introduction that there are three factors that make an I-language.
These are, first, the genetic endowment. Here we take this to be UG, containing
the various kinds of rules and systems of rules that build and interpret syntactic,
semantic and phonological representations. The details of what constitutes UG
will occupy us for much of what follows. The second factor is experience, in

IanRoberts_9781316519493_C001.indd 31 06-09-2022 11:48:50

 32 1 tacit knowledge

particular linguistic experience: exposure to a language as a small child. This

happens to almost everyone, and the nature of the early linguistic experience
will determine which E-language group an individual belongs to, in part deter-
mined by the I-language which develops from the interaction of the linguistic
experience, UG and the third factors. The third factors cover a range of different
things, but among the most important are general cognitive abilities, not specific
to language. As we mentioned in the Introduction, it can be very difficult in cer-
tain cases to distinguish general cognitive abilities from language-specific abili-
ties (which we would attribute to UG unless they clearly stem from experience).
All the above remarks apply to linguistic theory, conceived as here, in gen-
eral: to semantics, syntax, morphology and phonology. The goals of syntactic
theory in particular are above all to investigate the nature of the adult’s tacit
competence in the native language, i.e. to investigate I-language. This is done
by discovering and describing grammatical and ungrammatical structures
(which, as we saw, mostly but not always correspond to acceptable and unac-
ceptable sentences according to native-speaker judgements); these observa-
tions are systematized as a model for that knowledge. In the chapters to follow,
we will see how this is done. This model is then tested and adapted against
further language data; if we are looking at UG, we may look at data from a
range of languages. The rule systems are grounded in the three factors dis-
cussed above. So three really fundamental, and in this approach, intertwined
questions are:
(57) a. how does a given I-language relate to UG?
b. how does adult I-language result from acquisition?
c. how is I-language related to general cognition and the third factors?

Each of these questions relates I-language to one of the three factors in language
design that underlies it. In this connection, our fish sentences are very important.
They show us that syntactic structures are potentially infinite and yet built out
of very simple elements. In particular, the structures are formed by repeating
the same operations again and again. The next few chapters, as well as much of
Volume II, are devoted to demonstrating this in full detail. The fish have given
us a first inkling of the nature of the syntactic component of I-language, but now
it is time to look more systematically into the details.

Exercises
1. Consider again the four-fish example in (28), repeated here as (i):
(i) Fish fish fish fish.
We assigned the structure in (34), here (ii), to this sentence:
(ii) [S [NP [N fish ] [Relative clause [NP [N fish ]] [VP [V fish ] [NP [N fish
]]]]] [VP [V fish ]]].

IanRoberts_9781316519493_C001.indd 32 06-09-2022 11:48:50

 1.5  For Further Study and/or Discussion 33

We also observed that inserting the relative marker that between the
first two fish made possible a different interpretation, which gave in
(35) (= (iii)):
(iii) [S [NP [N fish ] that [Relative clause [NP [N fish ]] [VP [V fish ] [NP [N
fish ]]]]] [VP [V fish ]]].
In all of (i–iii) there is an implicit main-clause object, meaning
something vague like ‘stuff’. It is, however, possible to interpret (i)
as having an overt object. Try to see that interpretation, indicate its
structure (in a rough way, along the lines of (ii) and (iii)), its interpre-
tation and, if you can, its approximate intonation. (HINT: look again
at the ambiguity of the two-fish example discussed in (7–9)).
2. Now look again at the five-fish example from (37) (= (i)):
(i) Fish fish fish fish fish.
We said that (i) is the four-fish example plus an overt direct object
for the main-clause verb, giving the structure in (38) (= (ii)):
(ii) [S [NP [N fish ] [Relative clause [NP [N fish ]] [VP [V fish ] [NP [N fish
]]]]] [VP [V fish ] [NP [N fish]]]]
But in fact (i) has a further interpretation, where the main-clause
object is a relative clause. Give the structure for this interpretation
of (i).
Furthermore, the relative clauses in both interpretations can be
marked with that, and in each case this gives rise to a further ambi-
guity inside the relative clause. Explain the ambiguity and give the
relevant structures.
3. Give the structure for the six-fish example in (40) (= (i)):
(i) Fish fish fish fish fish fish.
Now try to add a further level of embedding to (40), using any lexical
items you like (except perhaps fish). This will correspond to an eight-
fish sentence; give the structure for this one.

For Further Study and/or Discussion

1. If you speak a language other than English or Italian, translate the fish
sentences into your language as best you can; if you don’t speak any
other language, ask a friend who does (preferably a patient friend).
Try to assign structures along the lines of (27) to your sentences.
What difficulties arise in the translations? How similar to or different
from English or Italian are your translations? Which of the satellite
categories we identified do you need? Do you need other ones?

IanRoberts_9781316519493_C001.indd 33 06-09-2022 11:48:50

 34 1 tacit knowledge

2. Google Translate gives (i) as the Italian translation of the three-fish

English sentence:
(i) Pesce pesce pesce.
This is not a grammatical sentence but rather obviously just the
Noun pesce repeated three times. As we saw in (24), the correct Ital-
ian translation of the three-fish English sentence is (ii):
(ii) I pesci pescano pesci.
What does an Italian speaker have that Google Translate lacks?

Further Reading
R. Freidin. 2012. Syntax: Basic Concepts and Applications. Cambridge: Cambridge
University Press.

IanRoberts_9781316519493_C001.indd 34 06-09-2022 11:48:50

 2 Constituents and Categories

2.1 Introduction and Recap

In the last chapter we saw, using the fish sentences, that syntactic
structures are potentially infinite and are built out of very simple elements by
repeating the same operations on structure again and again. We also defined the
goals of linguistic theory, construed as a formal, cognitive theory, as being pri-
marily to discover what it is that people know when they know a language, i.e.
what is I-language competence? A further central question is how this knowl-
edge is acquired.
Our focus henceforth will be on elucidating the nature of I-language in rela-
tion to syntax. As we have seen, syntax is a central aspect of language, in that
it relates sound and meaning over an infinite domain, so clearly any theory of
syntax will be a central part of UG. We continue to take UG to be universal
innate knowledge, one of the three factors making up the language faculty. We
will therefore develop the theory of I-language syntax, taking this, like the rest
of the language faculty, to arise from the three factors in language design: UG,
exposure to language in early life and domain-general third factors (as described
in the Introduction).
The fundamental notion in syntax is constituent structure, i.e. the way in
which words group together into intermediate units (or phrases) of various cat-
egories, ultimately forming whole sentences. Some groupings of words, which
we hear as different ordered strings of words, are ill-formed, or ungrammatical;
others are well-formed, or grammatical. We saw in Chapter 1 that grammatical-
ity and acceptability generally coincide: almost all of the examples we consider
from now on are of this kind. So when I use the terms ‘grammatical/ungram-
matical’ (or ‘well-formed/ill-formed’), these will refer to examples which are
correspondingly acceptable or unacceptable, i.e. native-speaker judgements
consistently correspond to what is generated, or not, by the rules of grammar.
Consider, for example, the following fairly simple five-word English sentence:
(1) Goona hopes that Vurdy will be ready for his dinner.

There are ten! = 3,628,800 possible orders for this ten-word sentence, the vast
majority of which are ungrammatical, such as the following:
(2) a. *Hopes Goona that Vurdy will be ready for his dinner.
b. *Hopes that Goona Vurdy will be ready for his dinner.

35

IanRoberts_9781316519493_C002.indd 35 03-09-2022 06:53:57

 36 2 constituents and categories

Our theory must make the distinction between grammatical and ungrammat-
ical strings and structures; it must tell us how and why (1) is grammatical
while the examples in (2) are ungrammatical. Furthermore it should tell us
when two sentences consisting of different words (and therefore meaning
different things) have the same structure. This is true for (1) and the exam-
ples in (3):
(3) a. Ron said that Hermione would be happy with her homework.
b. Boris thinks that Emile must be angry with his behaviour.
c. Alex believes that Eric should be worthy of a medal.

The basic notions we use to do this are those of categories and constituents,
which we now look at in turn.

2.2 Categories
2.2.1 Introduction: Lexical and Functional Categories
Like many approaches to morphology and syntax, including tradi-
tional grammar, we make a first distinction between two main types of catego-
ries, which we will call lexical categories and functional categories.
The principal lexical categories correspond to some of the ‘parts of speech’
of traditional grammar: they are noun, verb, adjective, adverb and preposition,
abbreviated standardly as N, V, Adj, Adv and P (we saw nouns and verbs in
our fish examples in Chapter 1). These lexical categories are open: people can
and do invent nouns and verbs all the time, and even the most up-to-date online
dictionaries struggle to keep up. Words belonging to these categories have clear
non-linguistic semantic content: as we saw, fish, as a noun, denotes a certain
class of aquatic animals, while fish, as a verb, denotes the activity of hunting
these animals. New adjectives and adverbs are less readily coined, but they do
arise, e.g. downloadable, googlable, etc. Prepositions are the one lexical class it
is difficult to add to; in this and certain other respects, prepositions are more sim-
ilar to functional categories. We will return below to the possibility of defining
syntactic categories semantically.
Furthermore, lexical categories do not vary greatly across languages. The dis-
tinction between nouns and verbs, in particular, seems to recur in all known
languages (although there is some debate about this in certain cases). Adjectives
are not found in every language; some indigenous languages of North America
may lack them, for example. In other languages, e.g. the West African languages
Hausa and Igbo, they form a small, closed class. In languages where the class
of adjectives is small or non-existent, the semantic work of adjectives, roughly
describing qualities, is usually done by relative clauses (a bus which reds). It
is probable that prepositions are not universal: the World Atlas of Language
Structures (WALS) lists thirty languages lacking adpositions (a cover term
for pre- and postpositions), including several indigenous languages of North

IanRoberts_9781316519493_C002.indd 36 03-09-2022 06:53:57

 2.2 Categories 37

America such as Blackfoot, Cree and Kiowa.1 English is comparatively rich

in prepositions (although they do not really constitute an open class, unlike
the other lexical categories; to some degree, prepositions straddle the border
between lexical and functional categories).
Functional categories include the ‘satellite words’ we introduced in the discus-
sion of the Italian fish sentences in Chapter 1; some of these, e.g. determiner (or
article), correspond to traditional parts of speech, but many of the others would be
treated in traditional grammar as parts of words rather than as syntactic categories
(a consequence of the fact that the traditional parts of speech were designed for
the description of Greek and Latin, both highly inflected languages). There we
saw D, Num, Mood, Tense and Agr(eement). Functional categories tend to con-
stitute small classes of, typically, rather small words; they may also be realised
as inflectional endings or, indeed, have no overt realisation at all, as we saw for
several cases in English in Chapter 1. English auxiliaries (e.g. must, be, do, have,
etc.) represent several functional categories, notably Mood and Tense. English
is particularly rich in auxiliaries and they play a central role in many aspects of
English syntax, as we will see. Further functional categories of English include
determiners (D: the, a, this, my, etc.) and Number (represented by the plural
inflection -s). There are other functional categories, e.g. complementisers, or C,
the words which introduce subordinate clauses of various kinds: if, that, for, etc.
(these are known in traditional grammar as subordinating conjunctions).
Functional categories are closed classes: it is seemingly impossible to con-
sciously invent new ones. Furthermore, their semantic content is either highly
abstract, as in the case of Mood, Tense and determiners, or perhaps non-ex-
istent: for example, what is the meaning of the complementiser that in (1)? It
seems that it can be left out without affecting the meaning of the sentence at
all. Furthermore, functional categories appear to vary quite a lot from language
to language. Here we have given examples from English. Even in languages
fairly closely related to English, the auxiliary system is less rich, and various
semantic notions indicated by auxiliaries in English (e.g. future tense will) may
be indicated by inflections instead, as in several of the Romance languages for
example. In Chapter 1 we saw that Italian is much more richly inflected than
English: Mood, Tense and Agreement (M, T and Agr) are mostly realised by
verbal inflections in Italian and the other Romance languages but often not real-
ised at all or realised by auxiliaries in English (although English does have a lit-
tle bit of tense and agreement inflection, as we saw). In a radically non-inflecting
language such as Chinese or Thai, Mood and Tense are indicated by invariant
(non-inflecting) ‘sentence particles’ and agreement is entirely absent. One of
the major issues in developing the theory of UG is understanding the extent to
which functional categories can vary, both in their realisation (as inflections,
separate words, or silence) and in their existence (recall our discussion of Conj-
marking in English and Italian in Chapter 1). We will gradually develop this

1
WALS Map/Feature 85A, https://wals.info/feature/85A#2/16.3/153.1.

IanRoberts_9781316519493_C002.indd 37 03-09-2022 06:53:57

 38 2 constituents and categories

theme over the coming chapters, and return to it armed with more empirical and
theoretical knowledge in Volume III.
How can we distinguish the various categories? This is quite a tricky matter,
and there are no absolute hard-and-fast diagnostic tests. For lexical categories in
particular, we can make use of four kinds of properties, relating to the main areas
of linguistic structure. Hence we can distinguish categories on the basis of their
morphology, their syntax, their phonology and aspects of their meaning. Let us
look at each of these in turn.

2.2.2 Morphological Diagnostics for Categories

The morphological criteria are arguably the easiest to see. We can
distinguish lexical categories in terms of the typical inflections and other end-
ings they can have. So, for example, regular count nouns in English can have
the plural marker -s (as in boys but not fish, as we saw in Chapter 1). There is
a handful of nouns with irregular plurals: children, mice, men, feet and a few
more. Mass nouns such as milk, water, air, bread denote uniform quantities of
stuff rather than countable individuals, and as such do not have plurals. Many of
them can, however, be ‘coerced’ into acting like count nouns and thereby form-
ing plurals: for example, beer is a mass noun, but one can say I drank two beers,
meaning two units (glasses, pints or types) of beer.
English verbs inflect for past tense and third-person singular (3Sg) agree-
ment in the present, as we saw in Chapter 1. The regular form of the past tense
is -ed, although there are about 150 irregular verbs in Modern English: sang,
brought, drove, took, etc. All verbs (as opposed to auxiliaries) take the regular
3Sg present-tense ending: the only irregularities are says, which is pronounced
‘sez’ (IPA /sɛz/) rather than ‘says’ by most people and does, pronounced /dʌz/
not /du:z/ whether used as a main verb or an auxiliary.
As we saw in Chapter 1, languages with richer inflections offer more morpholog-
ical ways of distinguishing categories. We saw that in Italian fish the noun is pesce,
while fish the verb is pescare. This distinction reflects the prevalent pattern in Italian,
making verbs and nouns very easy to distinguish in terms of their inflections (and all
but eliminating fully ambiguous forms comparable to English fish). Nouns mainly
fall into one of three classes, which can be distinguished by their singular endings:
one class ends in -a, forms the plural in -e and is feminine in gender (e.g. mela/mele
‘apple(s)’); a second class ends in -o, forms the plural in -i and is masculine in gen-
der (e.g. gatto/gatti ‘cat(s)’) and the third class, illustrated by pesce, ends in -e in the
singular and forms its plural in -i; this last class may be either masculine or feminine.
Verbs largely fall into one of four conjugation classes marked by a characteristic
vowel immediately following the root. The first conjugation is marked by -a-, as in
pescare (this is the truly open, productive conjugation; for example, a fairly recent
innovation is googlare, ‘to google’); the second conjugation is marked by unstressed
-e-, as in leggere (‘to read’), the third by stressed -e- as in vedere (‘to see’) and the
fourth by -i- as in dormire (‘to sleep’). In all cases, the infinitive ending is -re. Italian
verbs have a large number of tense and agreement endings, far too many to list here.

IanRoberts_9781316519493_C002.indd 38 03-09-2022 06:53:57

 2.2 Categories 39

All of this makes it very easy to identify verbs in terms of their morphology. What
we see in Italian is typical of the Romance languages, and, with variations, com-
mon across the Indo-European languages as a whole. Some languages show case
marking on nouns, which indicates the function of the noun (or really of the NP) in
the clause: nominative case marking typically marks subjects, accusative direct
objects, dative indirect objects, etc. So in Latin dominus (‘master’) has the -us
ending for nominative singular (in this class of nouns, known as the second declen-
sion), -um for accusative singular, -o for dative singular and in fact three more cases.
In languages of this type, nouns can be identified by case marking (adjectives agree
with the nouns they modify, so they can too).
In Modern English, common adjectives can be identified by their ability to
form the comparative with the ending -er (tall – taller) and the superlative in -est
(tallest). This applies only to adjectives of two syllables or less; hence there is no
comparative beautifuller or superlative beautifullest. This constraint applies to
the root form of adjectives, so unhappier exists, since the root is the two-syllable
happy. As usual, there is a handful of irregular adjectives, e.g. good – better –
best, bad – worse – worst.
Regular adverbs are formed from the corresponding adjective by adding -ly,
e.g. beautifully, happily, badly. There are some irregular adverbs, e.g. well, and
some which simply do not add -ly, e.g. fast. Conversely, there are some adjec-
tives which end in -ly, e.g. friendly. Prepositions are invariant: they do not inflect
at all in English.
So morphological criteria can identify lexical categories in English. In most
of the Romance and the other Germanic languages, where there is typically more
inflection than in English, these criteria are correspondingly more useful and
reliable. As we saw with our fish sentences in Chapter 1, English allows many
apparently inflectionless, and hence morphologically ambiguous, forms (we
suggested that this may be due to a large amount of silent inflection in English).

2.2.3 Syntactic Diagnostics for Categories

Given the paucity of overt inflection in English, syntactic criteria
may provide better diagnostics for categories, especially functional categories
(which tend either to show irregular inflection or to be invariant). Typical syn-
tactic tests rely on the distribution of categories, identifying positions in the
sentence which are reserved for words of a given category; each category thus
has its characteristic distribution. For example, only auxiliaries invert in direct
yes/no questions, so alongside the declarative in (4a) we have the yes/no inter-
rogative (so-called because it invites the answer ‘yes’ or ‘no’) in (4b):
(4) a. Cambridge will flood as a consequence of global warming.
b. Will Cambridge flood as a consequence of global warming?

In (4a), the declarative has the order Subject – Auxiliary – Verb … . In (4b),
the interrogative shows the ‘inverted’ order of subject and auxiliary. Here the
auxiliary is will, indicating (roughly) future tense. If there is no auxiliary in the

IanRoberts_9781316519493_C002.indd 39 03-09-2022 06:53:57

 40 2 constituents and categories

declarative sentence, the seemingly meaningless, ‘dummy’ auxiliary do appears

in the ‘inverted’, pre-subject position in the interrogative. This can be seen in (5):
(5) a. Cambridge flooded after all the heavy rain last spring.
b. Did Cambridge flood after all the heavy rain last spring?

In the declarative (5a), there is no auxiliary and past tense is marked by the -ed
ending on the verb. In (5b), the auxiliary do appears, marked for past tense as
did, and precedes the subject, while the verb has no ending. Placing the inflected
verb in front of the subject is ungrammatical (indicated by the asterisk * preced-
ing the sentence):
(6) *Flooded Cambridge after all the heavy rain last spring?

This way of forming interrogatives is characteristic of Modern English. Until

roughly the seventeenth century, verbs could also ‘invert’ with the subject. So
we find examples like the following in Shakespeare:
(7) Looks it not like the king? (Hamlet Act One, Scene I)

For this reason, readers familiar with Shakespeare and other writers from the
seventeenth century or earlier may find that examples like (6) have a certain
Shakespearean ring to them, but they are not really part of Modern English.
Following traditional grammar, we have said that clauses are divided into a
subject and a predicate. The predicate is usually a Verb Phrase, VP. Other cate-
gories are able to function as predicates, though, providing we include the aux-
iliary be. We can see this in examples like the following, where the predicative
category is labelled in the way we saw for the fish sentences in Chapter 1 (AP
stands for Adjective Phrase, PP for Prepositional Phrase):
(8) a. John is [AP nice ].
b. John is [AP interesting ].
c. John is [PP in a bad mood ].
d. John is [NP a nice person ].
e. John is [VP sleeping ].

In (8e), the verb sleep appears in its ‘progressive’ form marked with -ing, indi-
cating an ongoing situation.
In addition to following be, predicative categories can also follow certain
verbs, such as seem.
An important distributional property of VPs is that, unlike predicative APs,
PPs or NPs, they are unable to directly follow verbs like seem, as the ungram-
maticality of (9e) shows:
(9) a. John seems [AP nice ].
b. John seems [AP interesting ].
c. John seems [PP in a bad mood ].
d. John seems [NP a nice person ].
e. *John seems [VP sleeping ].

IanRoberts_9781316519493_C002.indd 40 03-09-2022 06:53:57

 2.2 Categories 41

So this test picks out VPs, since only VPs are ungrammatical following seem.
One important distributional test which picks out NPs is that only NPs can be
subjects. Thus only an NP (which may consist of just a single noun or pronoun)
can appear in the blank in (10):
(10) ___ can be a pain in the neck.

From (10), we can form the sentences in (11a) by inserting a single noun or pro-
noun (representing a whole NP), or those in (11b) where we see more complex
NPs, but not those in (11c), where the words inserted are not nouns:
(11) a. You/kids/injections/syntax/Dave can be a pain in the neck.
b. Professors of Linguistics/other people’s kids/injections which go wrong/
fish fish fish can be a pain in the neck.
c. *Walk/tall/in can be a pain in the neck.

Similarly, only VPs (which can be just a single verb) can appear between an
auxiliary and a manner adverb, i.e. in the slot in (12):
(12) Students can ___ quickly.

In (13a), we have single-verb VPs in that position and in (13b) more com-
plex VPs. In (13c) the words inserted are not VPs and so the sentences are
ungrammatical:
(13) a. Students can talk/write/learn/understand quickly.
b. Students can dissolve in sulphuric acid/get married/conclude that you’re not
worth listening to/fish fish quickly.
c. Students can *Olly/*kids/*injections/*syntax/*tall/*in quickly.

The distributional tests we have seen so far allow us to identify NPs, VPs and
auxiliaries.
Here is a distributional test for APs. Only (gradable) APs can appear in the
blank slot in (14):
(14) The students are very ____ .

In (15a) we see simple, one-adjective APs in that slot, in (15b) we have complex
APs of various kinds, while (15c) shows that other categories cannot appear there:
(15) a. The students are very intelligent/diligent/nice/eager.
b. The students are very much more intelligent than I expected/diligent in
handing in their essays/nice to talk to/eager to please.
c. The students are very *from/*walk/*Olly.

Finally, a test for PPs is that they can be identified by their characteristic intensifiers
such as straight and right. Thus only PPs can appear in the blank slot in (16):
(16) John walked straight ____ .

In (17), we have the same pattern as in (11), (13) and (15). Example (17a) gives
simple PPs (intransitive PPs, containing just a P); (17b) illustrates PPs of various

IanRoberts_9781316519493_C002.indd 41 03-09-2022 06:53:57

 42 2 constituents and categories

kinds, almost always with an NP object, and (17c) shows that APs, NPs and VPs
cannot appear in this slot:
(17) a. John walked straight out/on/up/in.
b. John walked straight out of the room/on to his destiny/up the hill/into the
pub.
c. John walked straight *red/*talk/*Olly.

We see that syntactic, distributional tests can isolate the main lexical categories,
NPs, PPs, VPs and APs. In an inflectionally poor language like English, these
are probably the most reliable category diagnostics available.

2.2.4 Phonological Diagnostics for Categories

Phonological tests are rather limited in scope in English. However,
there are some. For example, there is a small class of disyllabic words in English
which, like fish, are ambiguously nouns or verbs in their written form. However,
the categorial distinction is marked by stress, as the following examples
illustrate:
(18) a. Apple wants to increase its profits.
b. Apple wants an increase in its profits.

In (18a), increase is a verb and has stress on the second syllable (shown by the
bold letters). In (18b), increase is a noun and is stressed on the initial syllable.
Another case where phonology indicates an aspect of syntactic structure concerns
the difference between word stress and phrasal stress. Hence blackbird, with
stress on the first syllable, is a word, in fact a compound noun. This noun denotes
a particular species of bird. On the other hand, black bird, with stress on bird, is an
NP (or part of an NP; strictly speaking there should be a determiner or plural mark-
ing). This NP denotes any bird which happens to be black, following the usual rules
for attributive modification in English (so, red bus denotes any bus which happens
to be red, and so on). Thus one can say That black bird is not a blackbird without
self-­contradiction, which would be impossible if the different stress patterns did not
indicate different structures (NP vs N) and therefore different meanings.

2.2.5 Semantics and Categories

Turning now to semantic criteria, these are typically invoked in
traditional definitions of syntactic categories (often called ‘parts of speech’).
Nouns are defined as naming a person, place or thing, while verbs name actions
and adjectives qualities. Here are some typical definitions of verbs: ‘A verb
describes what a person or thing does or what happens’ (https://en.oxforddic-
tionaries.com/grammar/word-classes/verbs); ‘A verb is simply a doing or action
word’ (www.bbc.com/bitesize/articles/zpxhdxs); ‘Verbs are words that show
an action (sing), occurrence (develop), or state of being (exist)’ (www.merri-
am-webster.com/dictionary/verb).

IanRoberts_9781316519493_C002.indd 42 03-09-2022 06:53:57

 2.2 Categories 43

There is something to these definitions, although they are rather vague and lim-
ited. On their own, they scarcely suffice to identify categories. Take, for example,
‘what someone does’ as part of the first definition just given, or ‘action’ in the
second and third definitions (after all, actions are what people do). The word
action clearly denotes ‘action’, but it’s a noun. Similarly for occurrence, also a
noun, or ‘state of being’ (existence is a noun). Or consider the following example:
(19) The economy worries John.

Here worries is a verb, but it’s neither an action nor a ‘state of being’ (either
of the economy, or of John). Is it an occurrence? That seems a rather difficult
question to answer, but it’s not clear that answering either way would shed much
light on things. Similarly, nothing is a noun, but it doesn’t denote a thing, and it
certainly doesn’t denote a place or a person.
Despite the vagueness of these definitions, we can discern some semantic
component in category distinctions. This can be seen relatively clearly in English
where, as we have repeatedly noted by now, words can belong to different cate-
gories without changing form. For example, consider the different meanings of
round in the following examples:
(20) a. the round church
b. Round the rugged rock the ragged rascal ran.
c. These cars round corners very nicely.
d. Time for another round.

In (20a), round is an adjective, making up an AP on its own, and, much in line

with the traditional semantic definition of adjectives, it denotes a property of
the noun it modifies, its shape. In (20b), round is a preposition. It denotes a path
in relation to the NP following it; again, prepositions are typically defined as
denoting locations or paths. In (20c), round is a verb. Here again, it denotes a
path in relation to its object, corners, traversed by its subject, these cars; so we
can roughly think of it as denoting an action or a characteristic situation involv-
ing the subject and the object. In (20d), we have round as a noun. Here it denotes
a particular cultural convention (characteristic of English pubs) metaphorically
based on a circular shape. In each case, there is the same core of meaning in
round, involving a circular shape or path; how that notion is interpreted in rela-
tion to the other elements in the sentence depends on the word’s syntactic cate-
gory and, as a consequence of that, its syntactic relations with the other words.
So the traditional idea that adjectives denote qualities, prepositions locations or
paths, verbs actions or situations and nouns “things” has some purchase here. We
saw the same with fish: as a noun, it denotes a category of animals (a ‘thing’), as
a verb it denotes the action of hunting the thing in question. The adjective fishy,
formed by adding the characteristic adjectival suffix -y, denotes the quality of
being ‘fish-like’ (and has various metaphorical extensions which are perhaps
more commonly used than the basic sense). There is no prepositional variant of
fish, and it is rather difficult to imagine what that might be.

IanRoberts_9781316519493_C002.indd 43 03-09-2022 06:53:57

 44 2 constituents and categories

2.2.6 Conclusion
We have now seen several different ways of identifying syntactic
categories. None of them is absolutely foolproof, but taken together, they are
fairly reliable. Arguably, the distributional syntactic criteria are the most reliable
for English. In languages with richer inflectional morphology, such as Italian,
morphology can be more informative than in English.
There is little doubt that the diagnostics for categories vary from language
to language. But do the categories themselves vary? Does every language have
the same category inventory? In discussing this question at the beginning of
this section, we pointed out that the noun–verb distinction may be universal,
while adjectives and prepositions probably are not. Moreover, there appears to
be variation in the inventory of functional categories a language can have. We
might want to include Gender as a functional category, a ‘satellite’ of N, in
Italian given the presence of grammatical gender in that language, but we would
probably not include it in English, since there is no grammatical gender. This
reasoning might lead us to exclude quite a few familiar functional elements,
determiners for example, from our grammar of Chinese. These questions remain
open and, as we said above, we will revisit them in Volume III, when we have
more facts and more theory to bring to bear on them. For now, we will assume
that English has the lexical categories N, V, Adj, Adv and P and at least the func-
tional categories Aux(iliary), D(eterminer) and C(omplementiser). Furthermore,
our conception of the nature and inventory of the functional categories will be
successively revised as we go on.

2.3 Constituent Structure

2.3.1 Phrasal Constituents
Now that we have an idea of the inventory of categories (at least for
English) and how to find them using various morphological, syntactic, semantic
or phonological tests, we can begin to look at how categories combine to form
constituents and constituent structure. This is really the core notion of syntax;
everything else flows from our conception of the structure of complex syntactic
units such as phrases and sentences, and how these are made up of words (or
smaller units) of various categories.
Let us start with some very simple two-word sentences:
(21) a. Clover slept.
b. Vurdy grew.
c. Charlie laughed.
d. Fish fish.

Taking our cue from how we analysed (21d) in (7) of Chapter 1, we can take
each of these sentences to consist of at the very minimum a noun and a verb, and
assign them the labelled bracketings in (22):

IanRoberts_9781316519493_C002.indd 44 03-09-2022 06:53:57

 2.3 Constituent Structure 45

(22) a. [Noun Clover ] [Verb slept ]

b. [Noun Vurdy ] [Verb grew ]
c. [Noun Charlie ] [Verb laughed ]
d. [Noun Fish ] [Verb fish ]

But we have seen that nouns belong in NPs. So we can substitute more complex
NPs for the simple nouns in (22):
(23) a. [NP The fat cat ] [Verb slept ]
b. [NP The boy who lived ] [Verb smiled ]
c. [NP The owner of the famous dog] [Verb laughed]
d. [NP Fish fish fish ] [Verb fish ]

These more complex categories are Noun Phrases, NPs. NPs consist of at least
a noun, along with other words and phrases that depend on and/or modify that
noun (adjectives/APs, PPs, relative clauses, etc.) as well as satellite functional
categories such as Num and D, as we saw in Chapter 1.
Similarly, we have seen that verbs belong in VPs, so we can substitute more
complex VPs for the simple verbs in (22) and (23):
(24) a. [NP Clover ] [VP ate a mouse ].
b. [NP Kate] [VP hopes that chocolate will not get more expensive ].
c. [NP Beth ] [VP ate her dinner with a nice Chianti ].
d. [NP Fish ] [VP fish fish fish fish ].

So here we see another kind of complex category: the Verb Phrase, VP. VPs
consist of a verb and other words and phrases that depend on/modify that verb
(objects, adverbs, adverbial phrases, subordinate clauses, PPs, etc.). We will
come back to the question of the status of satellite functional categories con-
nected to VP such as T, M and Agr in Chapter 4.

2.3.2 Hierarchical Relations among Constituents

As we saw in Chapter 1 with the fish examples, we can represent syn-
tactic structure generally with labelled bracketings, such as the following for (24a):
(25) [S [NP [N Clover ]] [VP [V ate ] [NP [D a ] [N mouse ]]]]

In technical terms, the labelled bracketing provides a structural description of

the sentence. Another way to give the structural description of a sentence is with
a tree diagram (or phrase marker):
(26) S

NP1 VP

N1 V NP2
Clover ate
D N2
a mouse

IanRoberts_9781316519493_C002.indd 45 03-09-2022 06:53:58

 46 2 constituents and categories

(Here NP1, N1, NP2 and N2 are numbered simply to keep them distinct; the num-
bers have no theoretical significance.) It is important to see that tree diagrams
and labelled bracketings present exactly the same information in typograph-
ically different ways. The choice between them is a matter of convenience.
However, most people find trees easier to work with, since the relations among
the elements of the tree are more readily visible than in labelled bracketings.
But this is just a matter of personal preference; nothing theoretical depends on
it. In what follows, I will mostly use trees to represent structural descriptions.
We can now introduce some terminology that relates to the parts of the tree
diagram. S, NP, VP, etc. are the nodes of the tree; the nodes are linked by
branches. Branches never cross and all emanate from S, which is often referred
to as the ‘root’ node (in this sense the tree is in fact upside down in relation to
botanical trees, since the root is depicted as being at the top; again this is purely a
matter of convention). The words are terminal nodes; the category symbols (S,
NP, VP, etc) are non-terminal nodes. Pursuing the tree comparison, one could
think of the terminal nodes as the leaves.
The most important relations in the tree are the vertical, hierarchical ones.
The two fundamental relations are dominance and constituency. A category A
dominates another category B just where A is both higher up in the tree than B
and connected to B. We can put this more precisely as follows:
(27) A node A dominates another node B just where there is a continuous path of
branches going down the tree from node A to node B.

In terms of (27), we can define immediate dominance:

(28) Node A immediately dominates node B just where A dominates B and no
node C intervenes on the downward path from A to B (i.e. there is no node
C such that C dominates B and C does not dominate A).

Constituency is the reverse of dominance. A given category, call it B, is a con-

stituent of another category A just where B is both lower down in the tree than
A and connected to A by a continuous path of branches going up the tree from
node B to node A. In parallel with immediate dominance, we now define imme-
diate constituency as follows:
(29) B is an immediate constituent of A just where B is a constituent of A and no
node C intervenes on the upward path from B to A (i.e. there is no node C
such that B is a constituent of C and A is not a constituent of C).

It is easy to see that (immediate) dominance and (immediate) constituency are

inverse relations: dominance ‘looks downward’ in the tree, while constituency
‘looks upward’. More generally, we can say:
(30) A node A (immediately) dominates a distinct node B if and only if B is an
(immediate) constituent of A.

Let us now apply these definitions to the tree diagram in (26), which I repeat
here for convenience:

IanRoberts_9781316519493_C002.indd 46 03-09-2022 06:53:58

 2.4 Conclusion 47

(26) S

NP1 VP

N1 V NP2
Clover ate
D N2
a mouse

We can observe the following dominance relations in (26):

(31) a. NP1 and VP are immediately dominated by S.
b. V and NP2 are immediately dominated by VP.
c. V, NP2, D, N2 and N1 are dominated, but not immediately dominated, by S.
d. D and N2 are immediately dominated by NP2.
e. D and N2 are dominated, but not immediately dominated, by VP.
f. N1 is immediately dominated by NP1.
g. NP1 does not dominate VP or anything VP dominates.
h. VP does not dominate NP1 or anything NP1 dominates.

From (31a) and (31c), and the fact that immediate dominance entails dominance,
we see that S dominates all the nodes in the tree. In fact, this is a good definition of
the root node: the root node dominates all the other nodes in a tree; all the nodes in a
tree are constituents of the root. Similarly, all non-terminal nodes dominate at least
one distinct non-terminal or terminal node, and terminal nodes dominate nothing.
The dominance statements in (31) are exactly equivalent to the constituency
statements in (32):
(32) a. NP1 and VP are immediate constituents of S.
b. V and NP2 are immediate constituents of VP.
c. V, NP2, D, N2 and N1 are constituents, but not immediate constituents, of S.
d. D and N2 are immediate constituents of NP2.
e. D and N2 are constituents, but not immediate constituents, of VP.
f. N1 is an immediate constituent of NP1.
g. NP1 is not a constituent of VP or of any constituent of VP.
h. VP is not a constituent of NP1 or of any constituent of NP1.

The relations of constituency and dominance are the most fundamental syntactic
relations.

2.4 Conclusion
The central goal of the theory of syntax is to account for how words
are grouped into larger units, phrases and sentences. Categories and constituents
are the building blocks for analyses. In this chapter, we have seen how to test for
and isolate various categories of English. None of these tests is foolproof, espe-
cially when taken alone, but taken together they do allow us to identify and dis-
tinguish categories. The second thing we have seen here are the ways in which
we can present the structural description of a sentence: labelled bracketings and

IanRoberts_9781316519493_C002.indd 47 03-09-2022 06:53:58

 48 2 constituents and categories

tree diagrams (or phrase markers). In these terms (using a tree diagram because
it is more convenient), we defined the fundamental relations of dominance and
immediate dominance, constituency and immediate constituency.
We take the notions of category and constituent to be part of UG (they are prob-
ably not primitives, but defined in terms of more abstract aspects of UG; nonethe-
less, their existence and nature are determined, perhaps indirectly, by UG). Other
languages may have categories English lacks or may lack categories English has
(although we have mentioned that the verb–noun distinction seems to hold up pretty
well everywhere). But the notion that the words of a language fall into distinct cat-
egories, isolable by some battery of tests (which, as we saw, may relate to various
other parts of the structure of language: phonology, morphology, semantics), is a
hypothesis about the nature of UG which has certainly not been disproved to date.
Similarly, other languages have constituent structures that differ from those
of English; not all tree diagrams represent universal structures. But, again, the
fundamental ideas of (immediate) dominance and constituency are a hypothe-
sis about the nature of UG, as are the proposals that branches never cross, all
branches emanate from the root node and words are terminal nodes. As hypoth-
eses about UG, these are ultimately hypotheses about genetically given aspects
of human cognition, as we mentioned in Chapter 1.
In the next chapter, we will see ways to justify the particular constituent struc-
tures we have proposed here.

Exercises
A. Categories
1. Assign the following words to a syntactic category (noun,
verb, adjective, preposition), using the tests discussed in
this chapter along with any others you know of. In your
answer, show not only the category, but also how you
apply the tests:
cat, moon, sing, fish, to, possibly, well, dispute, round
2. Now do the same with the following words: noble, enno-
ble, near, asleep and around
B. Constituents
3. Give the labelled bracketings and tree diagrams for the
following examples:
a. Boris snorted.
b. The onx splooed the blarrg.
c. Fish fish fish fish.
d. Vurdy saw Goona.

IanRoberts_9781316519493_C002.indd 48 03-09-2022 06:53:58

 Exercises 49

e. Clover ate the mouse.

f. Harry met Hermione.
4. Prepositions (invariant words typically denoting spatial
or temporal relations such as on, about, with, to, behind,
etc.) form Prepositional Phrases, or PPs, by combining
with an NP, as follows:
(i) [PP [P on ] [NP the bus ]]
Give the tree structure and the labelled bracketing for the following
sentences, all of which contain PPs with a structure like (i):
a. Boris talked about himself.
b. Boris talked to the reporter about himself.
c. Boris sent some poison to Emile.
What challenges do you encounter in the last two examples?
C. For further discussion:
5. What do you think might be the structure of NPs like the
following? Present your hypothesis as a tree diagram.
a. fat cat
b. the invisible visible stars
6. Specify all the (immediate) dominance relations in the
following tree:
S

NP VP

N V PP
Ron thought
D NP

about D N

the problem

6. How can we define terminal and non-terminal nodes in

terms of constituency?
7. S is the root of the tree. How can we define the root in
terms of constituency?

Further Reading
All of the references below discuss the ways in which we can distinguish categories,
using tests broadly similar to those discussed here and, as here, concentrating largely (but

IanRoberts_9781316519493_C002.indd 49 03-09-2022 06:53:59

 50 2 constituents and categories

not exclusively) on English. All of them therefore provide useful backup, and plentiful
further examples, to the ideas introduced here.

Carnie, A. 2011. Modern Syntax: A Coursebook. Cambridge: Cambridge University

Press, Chapters 2 and 3.
Carnie, A. 2013. Syntax: A Generative Introduction, Oxford: Blackwell, Chapters 2, 3
and 4.
Haegeman, L. & J. Guéron. 1999. English Grammar: A Generative Perspective.
Oxford: Blackwell, Sections 1.2.2 and 1.2.3.
Koeneman, O. & H. Zeijlstra. 2017. Introducing Syntax. Cambridge: Cambridge
University Press, Chapter 1.
Larson, R. 2010. Grammar as Science. Cambridge, MA: MIT Press, Unit 9.
Radford, A. 2016. Analysing English Sentences. Cambridge: Cambridge University
Press, Chapter 2.
Tallerman, M. 1998. Understanding Syntax. London: Routledge, Chapter 2.

IanRoberts_9781316519493_C002.indd 50 03-09-2022 06:53:59

 3 Phrase-Structure Rules and Constituency
Tests

3.1 Introduction
In the previous chapter, we saw the basics of the structural descrip-
tions of sentences, the essential features of syntactic representations. Structural
descriptions can be presented as labelled bracketings, as in (1a), or as tree dia-
grams, as in (1b):
(1) a. [S [NP [N Clover ]] [VP [V ate ] [NP [D a ] [N mouse ]]]].
b. S

NP VP

N V NP
Clover ate
D N
a mouse

We also introduced basic tree terminology: node, root node, t­erminal node,
non-terminal node, branch, etc., and the very important mutually definable rela-
tions of (immediate) dominance and (immediate) constituency. In addition to
constituent structures, as represented in trees and labelled bracketings (i.e. struc-
tural descriptions), categories are the other fundamental notion. We illustrated
the main categories of English, as well as the very important distinction between
lexical and functional categories, and showed how categories can be fairly relia-
bly isolated by a battery of syntactic, semantic, morphological and phonological
tests (which vary somewhat from language to language).
The goals of this chapter are to start to describe (and explain) the difference
between grammatical and ungrammatical sequences. In order to do this, we
introduce the mechanisms that generate structural descriptions, i.e. the rules that
build trees and labelled bracketings. These are the Phrase-Structure rules. The
exact Phrase-Structure rules and the structural descriptions they generate can be
justified by independent tests designed to isolate and distinguish the constituents
in a tree or labelled bracketing; these are the constituency tests. We look first
at the Phrase-Structure rules and then at some constituency tests which, for the
most part, work quite well in English.

51

IanRoberts_9781316519493_C003.indd 51 05-09-2022 19:49:18

 52 3 phrase-structure rules and constituency tests

3.2 Phrase-Structure Rules

Here are some examples of ungrammatical sentences of English (in
the sense that they use English words and are just about intelligible to English
speakers; strictly speaking, these sentences are part of neither I-language English
or E-language English):
(2) a. *Spoke John.
b. *Hopes Goona that Vurdy will be on form.
c. *Loves the Leader of the Opposition his wife.

The question is: how do we, as speakers of English, know that these sen-
tences are ungrammatical? Clearly we don’t merely store this information
for each individual sentence or we wouldn’t have grammaticality judgements
about novel sentences such as the fish sentences we saw in Chapter 1. What
is needed is some general schema determining which are the grammatical
sentences of English (that is, of the I-language of the typical native speaker
of English).
To put it more technically, we need a mechanism to generate the well-
formed structural descriptions of English sentences. The mechanism in ques-
tion is the set of Phrase-Structure rules, PS-rules for short. PS-rules are the
formal devices which generate constituent structure, by specifying all and
only the possible ways in which categories can combine. An example of a
basic PS-rule of English (and probably many other languages) is given in
(1), where S stands for Sentence;

(3) i. S → NP VP

This is read as ‘Rewrite the symbol S as the sequence NP VP in that order.’

More generally, the rules are instructions to replace the symbol on the left of
the arrow with the symbol or symbols on the right of the arrow in the linear
order given. In (4) we see some more PS-rules of English:
(4) ii. VP → V (NP) (PP)
iii. NP → (D) N (PP)
iv. PP → P NP

The brackets indicate optional categories. Strictly speaking (4ii), for exam-
ple, abbreviates VP → V, VP → V NP, VP → V PP and VP → V NP PP, but
we collapse these rules under (4ii) using the bracket notation. These rules can
generate the structural description for the sentence Clover ate a mouse, given
as a tree diagram in (1a) and a labelled bracketing in (1b), and similar ones.
Here is another example of a tree diagram (which we saw in the Exercises
sections in the previous chapter):

IanRoberts_9781316519493_C003.indd 52 05-09-2022 19:49:18

 3.2 Phrase-Structure Rules 53

(5) S

NP VP

N V PP
Ron thought
P NP

about D N

the problem

This tree can be generated by the PS-rules in (6), which combines the rules
in (3) and (4):
(6) i. S → NP VP
ii. VP → V (NP) (PP)
iii. NP → (D) N (PP)
iv. PP → P NP

Each rule generates a small piece of tree, typically (but not always) a branch-
ing node and the constituents which that branching node immediately domi-
nates. As such, each rule specifies a set of immediate dominance/constituency
relations, such that the category to the left of the arrow is the label of the node
which immediately dominates the constituent(s) to the right of the arrow, and
the category or categories to the right of the arrow label the node(s) which
is/are immediate constituent(s) of the category to the left of the arrow. So,
PS-rule (6i) generates the structure in (7):
(7) S

NP VP

PS-rule (6ii) has two categories in brackets on the right of the arrow. The
brackets indicate that the categories in question are optionally immediate con-
stituents of VP (another way to say this is to say that they are optionally part
of the ‘expansion of VP’, given that PS-rules generally expand the category
to the left of the arrow by specifying more than one category on the right).
Strictly speaking, (6ii) really collapses the four distinct PS-rules seen in (8):
(8) a. VP → V
b. VP → V NP
c. VP → V PP
d. VP → V NP PP

Option (8c) of rule (6ii) generates the piece of structure in (9):

IanRoberts_9781316519493_C003.indd 53 05-09-2022 19:49:19

 54 3 phrase-structure rules and constituency tests

(9) VP

V PP

Rule (6iv) generates the structure in (10):

(10)  PP

P NP

Like rule (6ii), rule (6iii) collapses four rules using the bracket notation.
These are given in (11):
(11) a. NP → N
b. NP → D N
c. NP → N PP
d. NP → D N PP

Rule (6a) generates (12a) and rule (6b) generates (12b):

(12) a. NP

N

b. NP

D N

The pieces of structure in (7), (9), (10) and (12a,b) combine to form the tree
in (5). In (13), (5) is repeated with each part of the structure annotated giving
the PS-rule which generates it:
(13) S
    rule (6i)
NP VP
  
rule rule (8c) (one subcase of (6ii))
  
(11a) N V PP
   Ron thought rule (6iv)
   P NP
    rule (11b) (one subcase of (6iii))
    about D N
     
the problem
This illustrates how the structural description of the sentence is generated by
the PS-rules. The tree diagram in (13) gives the structural description of the
grammatical English sentence in (14):

(14) Ron thought about the problem.

Let us now return to the ungrammatical sentences, repeated here:

(2) a. *Spoke John.
b. *Hopes Goona that Vurdy will be on form.
c. *Loves the Leader of the Opposition his wife.

IanRoberts_9781316519493_C003.indd 54 05-09-2022 19:49:21

 3.3 Recursion 55

In (2a) the verb and the subject are in the wrong order for English. PS-rule (6i),
S → NP VP states “Rewrite the symbol S as the sequence NP VP in that order.’
Whatever expansion of VP we choose from the options in (6ii) (see (8)), V is
always the first immediate constituent of VP. It follows from the combination of
rules (6i) and (6ii) that the subject NP (the NP in (6i)) will precede the verb. Hence
the sentences in (2) cannot be generated by our PS-rules where the immediately
postverbal NP is the subject; if it is interpreted as the object, then the subject is
missing and again the rules in (6) cannot generate the sentence. Our PS-rules are
examples of the PS-rules of English, specifying all and only the well-formed struc-
tural descriptions of English sentences, and so the sentences in (2) are ill-formed.
It is important to see that the notion ‘ill-formed’ here means ‘not generated by
the syntactic rules of English’. Our aim is to make that notion coincide as far as
possible with native speakers’ intuitive judgements regarding the grammaticality
of English sentences, which we take to reflect their I-language competence. Here
it is important to remember what we saw in our discussion of centre-embedding in
Chapter 1: native-speaker judgements reflect performance, i.e. acceptability, rather
than competence, i.e. grammaticality. As we mentioned there, though, most of the
time grammaticality and acceptability coincide, and so most of the time matching
native-speaker judgements is a good benchmark for how well our theory is doing.
As we can see from our account of the ungrammaticality of (2), PS-rules
give information about the linear (left-to-right) order of nodes, including the
words at the terminal nodes. They also give information about hierarchical
structure, in that they specify immediate dominance and constituency rela-
tions. Finally, since the symbols they use are category symbols (S, N, V, P,
etc.), they give information about the category labels of nodes. So PS-rules
specify three kinds of information about structural descriptions simultane-
ously. We will see in Volume II that these three rather distinct kinds of
information can be teased apart.

3.3 Recursion
One of the most important things our discussion of the fish sentences
in Chapter 1 showed us was that the length of natural-language sentences is in
principle unbounded. If infinitely long sentences are grammatical (but maybe
not acceptable, as they can never be ‘performed’), they should be well-formed.
In other words, they should be generated by the PS-rules. Let us see how.
In fact, the rules we already have can generate infinite structures. Consider
again rules (6iii) and (6iv):
(6) iii. NP → (D) N (PP)
iv. PP → P NP

It is easy to see that NP appears on the left of the arrow in (6iii) and on the right
of the arrow in (6iv). Given the way PS-rules work, this means NPs can appear
inside other NPs, i.e. NPs can be constituents of other NPs, as illustrated in (15):

IanRoberts_9781316519493_C003.indd 55 05-09-2022 19:49:21

 56 3 phrase-structure rules and constituency tests

(15)  NP1
generated by (6iii)
D N1 PP
generated by (6iv)
P NP2

(Again, the subscript numbers on the Ns and NPs serve merely to keep the two
occurrences of this category distinct; they have no theoretical significance.)
Since nothing requires us to apply rules (6iii) and (6iv) in that order, after
applying both rules so as to generate the structure in (15), we can go back and
apply rule (6iii) again so as to expand NP2 identically to NP1, thereby intro-
ducing a second PP, which we can expand by rule (6iv) to give a third NP,
which we can expand by rule (6iii) to give a fourth NP, which we expand by
rule (6iv) to give a third PP, and so on. In principle, there is no limit to how
many times we can keep applying rules (6iii) and (6iv) to their own output.
The result of iterated application of these rules is complex recursive NPs of
the following kind:
(16) [NP [D the] height [PP of [NP [D the] lettering [PP on [NP [D the ] covers [PP of [NP [D the
] manuals [PP on [NP [D the ] table [PP in [NP [D the ] corner [PP of [NP [D the ] room …

The ability of rules to apply to their own output is known as recursion, a con-
cept that originates in mathematics and logic. Rules (6iii) and (6iv) are not indi-
vidually recursive, but together they form a recursive rule system. Recursion is
an extremely important concept, as it gives rise to the possibility of sentences
of unlimited length and underlies the fact that human languages are able to
make ‘infinite use of finite means’. As we have just seen, we can construct an
infinitely long NP using just rules (6iii) and (6iv). If our minds contain a recur-
sive rule system of this kind for generating sentences in our native I-language,
then our finite brains have infinite capacity. This is clearly a very important
and interesting claim about human cognition, one which justifies looking at
language from the formal and cognitive perspective adopted here.
In fact, we already saw recursion in action in the fish sentences in Chapter 1.
Consider the structural description we gave there for the seven-fish example:
(17) [S [NP [N fish ] [RC [NP [N fish]] [VP [V fish] [NP [N fish ]]]]] [VP [V fish ] [NP [N fish ] [RC
[NP [N fish]] [VP [V fish] [NP [N fish ]]]]]]]

The relative clauses introduce NPs inside other NPs and the object relative
clause introduces a VP inside the main-clause VP. Since, as we said just after
introducing the representation in (17), relative clauses are really sentences,
i.e. of category S, we should substitute S for the RC labels in (17). Then
we see that relative clauses introduce Ss inside Ss. So the reason we can
have infinite fish sentences is that the PS-rules generating relative clauses
are recursive. They feature occurrences of the symbols S, NP and VP on both
sides of the arrow, just as rules (6iii) and (6iv) do for NP and PP.

IanRoberts_9781316519493_C003.indd 56 05-09-2022 19:49:22

 3.3 Recursion 57

Relative clauses are one kind of subordinate clause. I won’t give the PS-rules
that generate them here as certain details of their structure remain uncertain and
controversial. Another kind of subordinate clause whose structure appears to be
much more straightforward, however, are complement clauses. A simple com-
plement clause is illustrated in (18):
(18) Goona hopes [ that Vurdy arrived].

Here we see a finite clause functioning as the complement, essentially a

clausal direct object, of the verb hope, introduced by the complementiser
(subordinating conjunction) that. It is easy to observe that what follows that
in (18), Vurdy arrived, is a complete sentence on its own. We introduce the
PS-rules in (19) to generate the complement-clause structure of (18):
(19) i. VP → V S′
ii. S′ → Comp S

Rule (19i) adds to the various possible expansions of VP seen in (8). We could add
‘(S′)’ to rule (6ii); this would subsume (19i) and predict that the sequence V NP S′
is possible (it is, as in persuade Mary that John arrived), the sequence V PP S′ is
possible (it is, as in say to Mary that John arrived) and the sequence V NP PP S′
is possible (this may not be correct, but I will leave this complication aside here).
In (19ii) ‘Comp’ abbreviates ‘complementiser’. Rule (19ii) states that Comp and S
are immediate constituents of the subordinate clause S′. S′ and S are not the same
category: S is a clause and constitutes the root node of a tree, while S′ is the label
of a subordinate clause introduced by rule (19i). Thanks to rule (19ii), we have a
further case of S-recursion: S appears on the left of the arrow in rule (6i) and to
the right of the arrow in rule (19ii). These rules will together generate structural
descriptions with S inside another occurrence of S.
We can expand S introduced by rule (19ii) as NP VP (in fact, this is the only
expansion of S our PS-rules allow as we have formulated them so far). We can
then expand VP using rule (19i) and introduce S again by rule (19ii) and expand
S again as NP VP, reapply rule (6i), reapply rules (19i) and (19ii), and so on
without limit. Once again we see rules applying to their own output, the basic
property of recursion. Recursive application of these rules in this way gives rise
to unlimited sequences of subordinate clauses embedded inside one another.
This is observed in English examples like (20):
(20) Mary hopes that John expects that Pete thinks that Dave said that …

As far as competence is concerned, just as with the fish sentences and the
complex NP in (16), there is no limit to the length, or to the depth of embed-
ding, of sentences like (20). The usual performance restrictions mean that
infinite sentences cannot be uttered or written down, but three PS-rules ((6i),
(19i) and (19ii)) suffice to generate them.
The structural description of (18) in tree format is given in (21):

IanRoberts_9781316519493_C003.indd 57 05-09-2022 19:49:22

 58 3 phrase-structure rules and constituency tests

(21) S

NP VP
generated by rule (19i)
N V S′
Goona hopes generated by rule (19ii)
Comp S
that
NP VP
Vurdy
V
arrived
There are other complementisers in English, which introduce different kinds
of subordinate clauses. As we can see in (18) and (21), that introduces a
finite, declarative subordinate clause. In (22a), whether introduces an indi-
rect question, and in (22b) for introduces an infinitival clause:
(22) a. Goona wonders [S′ whether Vurdy will arrive on time ].
b. Goona arranged [S′ for Vurdy to arrive on time ].

As indicated here, both subordinate clauses are S′s, with whether and for in C.
What follows whether in (22a) can stand alone as a complete sentence (Vurdy will
arrive on time), and so there is no difficulty with applying rule (19ii) here. The
difference with the that-complement in (18) is that the whole subordinate S′ stands
for the direct interrogative Will Vurdy arrive on time? What we observe in indirect
interrogatives is the absence of subject-aux inversion (briefly discussed in Chapter
2, see examples (3) and (4) there): in the S′ in (22a) Vurdy precedes will just as in
a declarative. We also observe that the presence of whether contributes what we
could call ‘interrogative force’, i.e. it makes the declarative into an interrogative.
In (22b), what follows the complementiser for cannot stand alone as a com-
plete sentence: *Vurdy to arrive on time. We clearly have a subject (Vurdy) and
a predicate (arrive on time) as in (18) and (22), but what is missing is tense: there
is no finite verb or auxiliary. ‘Complete’ clauses, those able to stand alone, must
have a finite verb or auxiliary to mark tense (although the tense marker may be
silent in English, as in the following complete sentences: I/you/we/they/the boys
arrive). The particle to marks non-finiteness; it shows up in almost all infinitives
in English. It looks as though we cannot generate *Vurdy to arrive on time with
rule (6i), since there is no place for to. But it seems clear Vurdy is the subject NP
and arrive on time the VP. Assuming that to is not part of the subject NP, which
seems highly unlikely, there are two possible structural descriptions for *Vurdy
to arrive on time, which are shown in the labelled bracketings in (23):
(23) a. [S [NP Vurdy] [? to ] [VP arrive on time ]]
b. [S [NP Vurdy ] [VP to arrive on time ]]

Rule (6i) cannot generate (23a), but it could generate the NP VP structure in
(23b). But now rule (6ii) (with its various expansions) cannot generate the VP
there, which has to as its first constituent. In order to accommodate infiniti-
val subordinate clauses, one of the two basic PS-rules in (6) will have to be

IanRoberts_9781316519493_C003.indd 58 05-09-2022 19:49:23

 3.4 Constituency Tests 59

modified. I will leave this question open here, but we will come back to it in
Chapter 4 (see Section 4.4).
Combining (6) and (19), we now have the following set of PS-rules:
(24) i. S′ → Comp S
ii. S → NP VP
iii. VP → V (NP) (PP) (S′)
iv. NP → (D) N (PP)
v. PP → P NP

These rules can generate a large number, in fact, given their recursive nature,
an infinite number of English sentences. However, they do not generate all
the grammatical sentences English: we have seen that there is no place for the
infinitive to here, so we cannot generate (22b).
As we said at the end of the previous section, PS-rules give us three kinds of infor-
mation: (i) information about hierarchical structure (‘vertical’ information, in terms
of tree diagrams), information about linear precedence (‘horizontal’ information)
and information about the category labels of nodes. These rules are very powerful
formal devices which are capable of generating structural descriptions in a precise
way. But how do we know that structural descriptions of the kind we have been
looking at in this chapter are the right ones for English? The descriptions make very
clear claims about constituent structure, but are they correct? We have also seen at
least one example where it is not clear where to place a constituent in a structure: the
question of where to put to in infinitival clauses like (22b). Which of the structures in
(23) is the correct one, and why is the possibility that to is a constituent of subject NP
‘highly unlikely’ as I said above? In the next section we will turn to questions of this
kind, by showing how there are various ways of testing constituent structure, just as
there are various diagnostics for categories as we saw in Section 2.2.

3.4 Constituency Tests

Let us look once more at PS-rule (6i):

(6i) S → NP VP

This rule applies to give us the constituent structure (25b), rather than (25c),
for (25a):
(25) a. John ate the cake.
b. [S [NP John ] [VP ate the cake ]]
c. [S [VP John ate ] [NP the cake ]]

But why do we write the rule like this? What would be wrong with writing the
rule as (6i′), which would give the constituent structure we see in (25c) for (25a)?
(6i′) *S → VP NP

IanRoberts_9781316519493_C003.indd 59 05-09-2022 19:49:23

 60 3 phrase-structure rules and constituency tests

One obvious answer comes from the traditional idea that clauses consist of a
subject and a predicate. But this venerable idea does not really tell us anything
about phrase structure (despite what we said in Chapter 1): phrase structure rep-
resents hierarchy, order and categories, as we have seen. It does not represent
grammatical functions or relations such as subject and predicate. From what we
have seen up to now, these notions have no place in our theory; we may wish to
build them in somehow, and we will do this in Section 5.2 (since these notions
seem to be so useful for informal discussion we might want to have a theoretical
way to understand them, but the fact remains that we have not actually said any-
thing about this so far). We will look at the question of the theoretical status of
grammatical functions in full detail in Chapter 1 of Volume II.
So, coming back to choosing between (6i) and (6i′), what this really amounts
to is choosing between the structural description in (25b) and that in (25c) for
(25a). In (25b), which is the analysis we have been assuming up to now, the
verb and the direct object form a constituent that does not include the subject,
while in (25c) the subject and the verb form a constituent that does not include
the object. The question is: why should we prefer (25b) over (25c)? To put the
question another way, what is the evidence that ate the cake is a constituent and
John ate is not a constituent? The evidence is not directly audible on the basis
of what we hear as (25a); this is because constituent structure, like most of syn-
tax, is silent. The linear order in (25a) is clearly compatible with either (25b)
or (25c). So we must find a way to determine what the hierarchical structure is.
The same question arises in relation to the NP P sequence in (26a) or the N A
sequence in (26b):
(26) a. They gave [NP the book ] [P to ] Mary.
b. They gave [NP Mary ] [AP sweet ] cookies.

We would naturally treat to in (26a) as a Preposition forming a PP with Mary.

PS-rule (6iv) will generate this structure (and one version of (6ii) can generate
the sequence V NP PP seen in gave the book to Mary). We could justify this on
the grounds of the tight semantic connection between to and Mary, or on the
grounds that to marks the indirect-object function of Mary here. But these are
arguments based on semantics and grammatical functions, and this is not the
kind of information the structural descriptions generated by PS–rules carry (at
least not on the basis of anything we have seen so far). Similar considerations
apply to (26b): semantic and functional considerations, i.e. the fact that sweet
must be interpreted as modifying cookies and cannot be interpreted as modifying
Mary, would lead us to group the AP sweet with the noun cookies to form an
NP (which has the direct-object function). But can we find constituency-based
arguments for this conclusion? If we can, then not only do we establish the
constituency and have clear indications as to which set of PS-rules is the right
one, we also confirm the intuition (or hunch, really) that constituent structure is
closely connected to semantics, and that grammatical functions are also some-
how linked to constituency relations (see Section 5.2).

IanRoberts_9781316519493_C003.indd 60 05-09-2022 19:49:23

 3.4 Constituency Tests 61

Constituency tests of various kinds can show us to a large extent what the cor-
rect constituent structures are. These tests are manipulations of sentences which
are sensitive to phrasal categories such as NP, VP, etc. There are several kinds
of constituency tests. These involve: (a) manipulating the order of elements in
such a way as to show that certain sequences of words must be manipulated
together, i.e. that they constitute a phrase (these are clefting, wh-movement
and fronting); (b) substituting a sequence of words with ‘pro-forms’ of various
kinds; (c) ellipsis, deleting a sequence of words in such a way that its interpreta-
tion is recoverable from the linguistic context; (d) coordination, conjoining two
phrases of the same category with and, and (e) fragments, whether a sequence
can stand alone and be in an intuitive sense ‘complete’, even if it is not a com-
plete sentence. We will now look at each kind of test in turn. These operations
are cases of a class of rules distinct from PS-rules, transformational rules, one
type of which (wh-movement) we will focus on in Chapter 5.
Let us begin with wh-movement. Fronting a phrase containing a wh-word (i.e.
the interrogative pronouns and determiners who, what, which, etc.; see Chapter
5) also targets constituents. This operation, known as wh-movement, is a very
important one for syntactic theory, and we will introduce it fully in Chapter 5.
Again, there is a gap in the sentence where the questioned phrase was, which we
mark with a ‘t’ (and which can also be taken as a silent copy):
(27) a. Which friends does Mary hope that John will like t ?
b. What did the Party Chairman send t to John?
c. Who did the Party Chairman send a book to t ?
d. To whom did the Party Chairman send a book t ?

These examples show us that the direct objects in (32a,b) are constituents, that
the NP following the P to in (32c) is a constituent, and that the PP to whom is a
constituent.
Non-constituents, i.e. strings of words that do not form an independent, unique
phrase, bolded in (28), cannot be fronted:
(28) a. *What to did the Party Chairman send t John?
b. *What to John did the Party Chairman send t ?

Strictly speaking the ungrammaticality of (28) does not give us information about
constituency; only the successful cases of wh-movement do this. Wh-movement
does not apply to VP: the various wh-phrases correspond to different grammatical
categories: who, what are NPs (respectively animate and inanimate), which is a
D, why an AdvP or PP (‘for what reason’), when a temporal NP or PP, where a
PP, how an AP or AdvP and how (many) a measure expression. But there is no
wh-word which questions a VP. Hence, we cannot use wh-movement as a diagnos-
tic for a VP constituent. It does not follow from this that there is no VP constituent.
A further permutation we can apply to sentences in order to isolate constitu-
ents is fronting. This operation ‘highlights’ phrasal constituents by placing them
at the beginning of the sentence. The fronted phrase often functions as a topic

IanRoberts_9781316519493_C003.indd 61 05-09-2022 19:49:23

 62 3 phrase-structure rules and constituency tests

which ‘is commented on’ by the rest of the sentence. So, from the neutral sen-
tence (29a) we can derive (29b), where the direct object the new car is fronted:
(29) a. Mary hopes that John will like the new car.
b. The new car, Mary hopes that John will like t.

Again, we see a trace in the position where the direct object would normally be
in (29b). Fronting can apply to CP and VP, as in (30):
(30) a. That John will like her friends, Mary hopes t.
b. (Mary hoped that John would like her friends) … and [VP like her friends ]
he did t.

For (30b) to sound natural, it helps to give some context, as shown here.
Sentence (30a) may sound a little stilted, but it certainly seems acceptable. The
examples in (30) should be contrasted with (31) (where again non-constituents
are bolded):
(31) a. *A present to, the Party Chairman sent t John.
b. *A present to John, the Party Chairman sent t.

We still have no evidence that the sequences a present to or a present to

John are constituents, but now we have some evidence that VP and CP are
constituents.
A further permutation-based constituency test is clefting. Clefting is the
­permutation of the order of elements illustrated in (32):
(32) a. The Party Chairman sent a book to John.    →
b. It was to John that the Party Chairman sent a book t.
c. It was John that the Party Chairman sent a book to t.

The sentence in (32a) ‘neutral’. In (32b), the sequence to John has been clefted;
in (32c) just the NP has been clefted, ‘stranding’ the preposition to. The general
schema for clefting is given in (33):
(33) S → It was XP that S.

(33) is not a PS-rule. It represents a different kind of rule, a transformational

rule, that we will introduce properly in Chapter 5. The S on the left of the arrow
represents the neutral sentence, e.g. (32a); what is on the right of the arrow is
the clefted version of that sentence, e.g. (32b). The S on the right of the arrow
contains a gap (similar to the gaps in relative clauses which we observed in the
fish sentences in Chapter 1). Here it is marked as ‘t’ (standing for trace); this
indicates where the XP would normally be in the non-clefted version of the
sentence. We can think of ‘t’ as a silent copy of the XP. It stands for a single
constituent, so in (32b) we do not have two traces following book, one for P and
one for NP, but just one, standing for PP.
For our purposes here, the most important thing about clefting as schematised
in (33) is that XP has to be a constituent. In (32b), the clefted XP is the PP to
John. The sentences in (34) are two further cases of clefting:

IanRoberts_9781316519493_C003.indd 62 05-09-2022 19:49:23

 3.4 Constituency Tests 63

(34) a. It was the Party Chairman that t sent a book to John.

b. It was a book that the Party Chairman sent t to John.

In (34a), the NP the Party Chairman is clefted, while in (34b) it is the NP the
present. So the clefting test has isolated three constituents for us: [PP to John ],
[NP the Party Chairman ] and [NP a book ].
If we try to cleft non-constituents, bolded in (35), the result is ungrammatical:
(35) a. *It was the Party Chairman sent that t a book to John.
b. *It was a book to that the Party Chairman sent t John.

In (35a) the sequence the Party Chairman sent is clefted, and in (35b) a book
to. In both cases the result is ungrammatical. Since there could be independent
reasons why clefting some constituents is not good, the failure of a constituency
test does not really tell us anything. For example, clefting VP yields a rather odd
result (although probably not as bad as (35)):
(36) ??It was send a book to John that the Party Chairman did.

Despite the oddity of (36), we do not immediately conclude that VP is not a

constituent in (32a) and, as we have seen, other tests can successfully pick out
VP for us.
So, the success of the test tells us that the fronted sequence is a constituent,
but the failure of the test, strictly speaking, doesn’t tell us anything. Unlike the
VP, however, no constituency test will isolate the strings the Party Chairman
sent or a book to as in (35). These strings are equivalent to the putative VP in
(25c) and the possible constituents discussed in relation to (26).
The fourth type of constituency test involves substitution of pro-forms. Pronouns
(I/me, he/him, she/her, etc.) are noun-like elements that have little semantic content
of their own but stand in for something else. This ‘something else’, known as the
antecedent, determines what the pronoun refers to. So, in an example like (37),
John can be naturally interpreted as the antecedent of he (although it doesn’t have
to be; he could refer to any contextually salient male individual):
(37) John hopes that he will win.

So (37) is interpreted to mean ‘John hopes that he, John, will win.’ Since John is
an NP, we should really call pronouns NPs too, since they stand for whole NPs
rather than just nouns. We can see this from the two cases of pronoun substitu-
tion in (38b) and (38c):
(38) a. [NP The man who wears glasses ] hopes that he will win.
b. * The he who wears glasses hopes that he will win.
c. He hopes that he will win.

Treating he as a noun in (38b) leads to ungrammaticality, while treating he as

occupying the same position as the whole relative clause gives the grammatical
(38c). The ungrammaticality of (38b) doesn’t necessarily imply that the position
of he is not a constituent; in fact, the noun man clearly is a consitituent. It shows
us that pronouns are NPs (and not Ns).

IanRoberts_9781316519493_C003.indd 63 05-09-2022 19:49:23

 64 3 phrase-structure rules and constituency tests

Other categories have pro-forms too. Most VPs can be replaced by (do) so,
as in (39):
(39) John has promised Mary a book, and Bill has done so too.

Here we see the VP promised Mary a book in the first conjunct, i.e. the verb
along with both the direct and the indirect object, are replaced in the second
conjunct by done so. As with pronoun (i.e. pro-NP) substitution in (38), do so
stands for the whole VP:
(40) a. * … , and Bill has done so Mary too.
b. *... , and Bill has done so a book too.
c. *... , and Bill has done so Mary a book too.

In (40a), just the verb and the direct object (promise a present) have been sub-
stituted; in (40b), just the verb and the indirect object (promise Mary), and (40c)
just the verb promise. Again, the fact that these examples are all ungrammatical
doesn’t necessarily imply that the substituted categories are not constituents (in
fact, the verb on its own clearly is a consitituent); what it shows is that do so
substitutes an entire VP (and not, for example, V).
In the light of the conclusion that do so substitutes for a whole VP, consider
the following examples:
(41) a. John put his car in the garage on Tuesday, and Peter did so too.
b. *John put his car in the garage on Tuesday, and Peter did so on the driveway
on Wednesday.
c. *John put his car in the garage on Tuesday, and Peter did so his bike on
Wednesday.
d. John put his car in the garage on Tuesday, and Peter did so on Wednesday.

In (41a), which is grammatical, do so in the second conjunct appears to be replacing

the sequence put his car in the garage on Tuesday, i.e. the verb put, the direct object
his car, the locative PP in the garage and the temporal PP on Tuesday. We there-
fore conclude that all of these elements are in the VP. This conclusion seems to be
confirmed by the ungrammaticality of both (41b) and (41c): in (41b), the locative
PP on the drive is not substituted and ungrammaticality results, while in (41c) the
object NP his bike is not substituted, again leading to ungrammaticality.
But (41d) behaves differently. Here the temporal PP on Wednesday is not
substituted by do so in the second conjunct, but the sentence is grammatical.
The difference between the temporal PPs on Tuesday/on Wednesday in these
examples and the direct object and the locative PPs is that the former are not
obligatory with the verb put. The direct object and the locative are, though, as
the ungrammaticality of sentences like (42b–d), where one or other (or both) of
them is left out, shows in contrast with the grammatical (42a):
(42) a. John put the car in the garage.
b. *John put the car.
c. *John put in the garage.
d. *John put.

IanRoberts_9781316519493_C003.indd 64 05-09-2022 19:49:23

 3.4 Constituency Tests 65

The grammaticality of (42a) also shows us that the temporal PPs are not required
for grammaticality. Such PPs are optional modifiers, or adjuncts, giving ‘extra’
information about the time the event took place, but their presence is not required
in order for the sentence to be intuitively ‘complete’. On the other hand, (42b–d)
feel ‘incomplete’: required information about what is being put (or) where is not
given. This is because the direct object and the locative PP are arguments of the
verb put; they must appear in the VP when put is V. More technically, put cat-
egorially selects for a direct-object NP and a locative PP. But it does not select
for an adjunct temporal PP.
What the ungrammaticality of (41b,c) tells us then is that selected arguments
of V must form part of the VP with the verb, and hence must undergo do so
replacement. Then (41d) can be taken to indicate that adjuncts are outside the
VP, and so do not correspond to do so replacement. But what about (41a)? The
fact that the substituted VP is interpreted as put the car in the garage on Tuesday
indicates that the adjunct PP is part of the VP. So (41d) appears to be telling us
that the adjunct PP is outside VP, and (41a) appears to be telling us it is inside
VP. We conclude for now that adjuncts (of this kind, at least) are optional con-
stituents of VP, while selected arguments are obligatory constituents of VP. This
conclusion accounts for the pattern seen in (41) as well as the examples in (42)
and is consistent with the idea that do so corresponds to VP. We will see a more
sophisticated treatment of adjuncts in the next chapter.
The do so test and the fronting test allow us to find VPs. We can now apply
these tests to our two structures for John ate the cake in (25), repeated here as (43):

(43) a. [S [NP John ] [VP ate the cake ]]

b. [S [VP John ate ] [NP the cake ]]

Fronting gives the following results:

(44) (I thought John would eat the cake and)

a. .. eat the cake he did t!
b. *… John eat did t the cake!

With a slight tweak to the tense of the verb to create a natural context for
VP-fronting, (44a) is perfectly grammatical. Example (44b), on the other hand,
is strongly ungrammatical (bordering on what is sometimes called ‘word salad’,
a sequence of words so unintelligible that it is almost impossible to impose any
kind of structure or meaning on it).
Do so substitution, seen in (45), gives rise to a similar contrast:
(45) a. John ate the cake, and Mary did so too.
b. *John ate the cake, and did so the cake too.

Example (45b) is not word salad, but it is not acceptable. We can also notice that
Mary has completely disappeared from this example. This is a reflection of a
very general fact about substitution and ellipsis operations: substituted and elided
material is subject to a recoverability condition, i.e. the operations of substitu-
tion and ellipsis (the latter of which we will look at in depth in Volume II) must

IanRoberts_9781316519493_C003.indd 65 05-09-2022 19:49:23

 66 3 phrase-structure rules and constituency tests

be such that the semantic representation can find, i.e. recover, what has been sub-
stituted or elided. In (45b), Mary is not recoverable. The contrast in (45) clearly
supports the structure in (43a) over that in (43b). This is a good result because, as
we noted above, it seems that the constituents line up with the functions of sub-
ject and predicate (although our PS-rules cannot and do not state this).
As a final point on substitution, consider the following examples:
(46) a. %The Party Chairman sent it John.
b. The Party Chairman sent it.

Example (46a) is ungrammatical for most English speakers, although there are
varieties in the north-west of England, around Liverpool and Manchester, where
it is accepted (the ‘%’ is used to indicate a form which is acceptable in one vari-
ety of a language but not another; of course, this kind of variation simply reflects
the fact that the sociocultural E-language concept ‘English’ does not correspond
exactly to aggregates of I-languages – people from the relevant area of England
have slightly different I-languages from those from elsewhere, a fact reflected
in their differing grammaticality judgements of examples like (46a). However,
in these varieties it is interpreted as the direct object (a/the book); it cannot be
interpreted as substituting for a present to. Similarly, (46b) is grammatical, but it
can only be the direct object. An indirect object (i.e. to John) cannot be recovered
here. Again, then, a book to and a book to John fail the constituency test.
Ellipsis is the next kind of constituency test. Ellipsis elides, or deletes, mate-
rial, subject to the recoverability requirement we have already seen in relation to
substitution. VP-ellipsis is quite natural in English, as in:
(47) a. John can speak Mandarin and Mary can speak Mandarin too.
b. John will leave tonight and Mary will leave tonight too.
c. John has passed the exam and Mary has passed the exam too.
d. John is writing a book and Mary is writing a book too.

(48) John ate a cake and Mary did eat a cake too.

This example illustrates an aspect of the English auxiliary system that will play
a major role in our analysis of clauses from the next chapter on. The struck-out
sequence in the second conjunct here consists of the verb and its direct object,
and so we can quite reasonably analyse this as VP. This would be consistent with
our other results for constituency tests for VP: fronting and do so substitution (see
(44a) and (45a)). Example (48) differs from (47) in that there is no auxiliary in the
first conjunct and the auxiliary do appears in the second one. There is a difference
between the two conjuncts here: in the first one, the verb bears the tense marking:
we have ate, not eat. In the second conjunct, the past-tense marking is carried by
the auxiliary do, in the form of did, and so the struck-out verb form lacks tense
marking (hence eat, not ate). We see from (47) and (48) that, generally speaking,
auxiliaries do not have to be deleted under VP-ellipsis. This suggests that finite
auxiliaries are not part of VP. VP-fronting confirms this:

IanRoberts_9781316519493_C003.indd 66 05-09-2022 19:49:23

 3.4 Constituency Tests 67

(49) a. (I expected John can speak Mandarin) and speak Mandarin he can t!
b. (I expected John will leave tonight) and leave tonight he will t!
c. (I expected John to pass the exam) and pass the exam he has t!
d. (I expected John was writing a book) and writing a book he is t!

This raises the question of where auxiliaries are situated. Our basic PS-rule (6i)
has nothing to say about this and will need to be revised. We defer this question
to Chapter 4.
The exact counterpart of (47), but with did also apparently deleted is gram-
matical too:
(50) John ate the cake and Mary did eat the cake/ate the cake too.

(The struck-out forms indicate that it’s unclear where the tense marking is inside
the elided sequence: on do or on ate.) The sequence in (50) is ambiguous in
that it can also mean ‘John ate both the cake and Mary too’; this latter reading
involves coordination inside the direct object of ate and no ellipsis. The reading
that the strike-out in (50) indicates is the one where Mary is the subject, not the
object, of ate. Alongside (50), we have ellipsis patterns comparable to those in
(48), but where the auxiliary is also deleted:
(51) a. John can speak Mandarin and Mary can speak Mandarin too.
b. John will leave tonight and Mary will leave tonight too.
c. John has passed the exam and Mary has passed the exam too.
d. John is writing a book and Mary is writing a book too.

These examples, along with (50), indicate that the auxiliary can form a constit-
uent with VP. It may be then that there is a mystery constituent in the clause,
indicated in (52):
(52) [S [NP John ] [?? [Aux has ] [VP passed the exam ]]]

The VP is elided in (47) and (48), and the mystery category ?? in (52). In (23a)
we saw that infinitive to may occupy a mystery category in a similar position;
there we noted that PS-rule (6i) needs modification, a point we will return to in
the next chapter.
Consider once more our alternative structures for the simple sentence John
ate the cake in (43a,b):
(43) a. [S [NP John ] [VP ate the cake ]]
b. [S [VP John ate ] [NP the cake ]]

VP-ellipsis clearly favours (43a) here, just like do so substitution and VP-fronting,
as we have seen. Applying VP-ellipsis to (43b) yields:
(53) John ate the cake and Mary eat did the cake.

(53) is not exactly acceptable, but it is difficult to work out what the second
conjunct means. Moreover, Mary is unrecoverable in (53). Hence we must take

IanRoberts_9781316519493_C003.indd 67 05-09-2022 19:49:23

 68 3 phrase-structure rules and constituency tests

this to be an ill-formed application of ellipsis. Again, then, we see that (43a) is

the correct structure for John ate the cake and (43b) is not.
The second kind of ‘VP-ellipsis’, which appears to apply to our mystery cat-
egory ?? in that it deletes the auxiliary as well as the VP, gives (54) from (43b):
(54) John ate the cake and Mary eat did the cake.

Aside from being rather odd and redundant, here again Mary is unrecoverable.
Once again, (43b) seems to give the wrong results, while (43a) makes correct pre-
dictions. We conclude that (43a) is the correct structure. Thus the traditional func-
tional division of the clause into subject and predicate has a structural correlate.
Coordination can also function as a constituency test. By and large, only con-
stituents of the same type can be coordinated. Thus we can observe the follow-
ing(examples from Radford 2016:134):1
(55) a. The chairman has resigned from the board and the company.
b. The chairman has resigned from the board and from the company.
c. The chairman has resigned from the board and gone abroad.
d. The chairman has resigned from the board and is living in Ruritania.
e. *The chairman has resigned from the board and company has replaced him.
f. The chairman has resigned from the board and the company has replaced him.

In (55a), we see coordination of the NPs the board and the company; in (55b)
we have coordination of the PPs from the board and from the company; in (55c),
we have VP-coordination of resigned from the board and gone abroad. Example
(55d) confirms our suggestion above of a mystery constituent combining the aux-
iliary and the VP, since two occurrences of this constituent appear to be coor-
dinated: has resigned from the board and is living in Ruritania. Example (55e)
is unacceptable, and so we do not have evidence that the sequences chairman
resigned from the board and company has replaced him are constituents; in fact
each seems to consist of the N of the subject NP (chairman, company) and our
mystery category. This confirms the suggested structure in (52) for S where the
mystery category is a constituent of S and not of the subject. The structure for
The chairman has resigned from the board, on analogy with (52), would be (56):
(56) [S [NP [D The ] [N chairman]] [?? [Aux has ] [VP resigned from the board ]]]

Here we clearly see that the sequence chairman has resigned from the board is
not a constituent. This is consistent with the fact that the coordination test yields
an ungrammatical result when we attempt to combine strings of this type as in
(55e). Finally, (55f) shows us that S is a constituent.
The final kind of constituency test relates to fragments. These are most naturally
found in answers to questions. Generally speaking, only phrasal constituents form
grammatical fragments. So, for example, the following answers are ungrammati-
cal (here again, I borrow, and slightly adapt, examples from Radford 2016:137–8):

1
A. Radford (2016), Analysing English Sentences, Cambridge: Cambridge University Press.

IanRoberts_9781316519493_C003.indd 68 05-09-2022 19:49:23

 3.4 Constituency Tests 69

(57) A: What has the chairman resigned from?

*B: The.

(58) A: What has the chairman resigned from?

*B: Board.

The respondent B in (57) could reasonably be accused of being obtusely uncom-

municative in responding by indicating some unspecified but unique, existing
thing or things known to the speaker (recall our discussion of the meaning of
the definite article in Chapter 1). But in (58), B provides a real answer as far as
the content is concerned. However, fragments are unable to be non-phrasal cat-
egories such as single nouns, and, as we have seen, singular count nouns cannot
constitute an NP on their own in English, so the fragment response in (58) is
ungrammatical. The principle that fragment answers must be phrases will also
account for the ungrammaticality of the response in (57), independently of its
pragmatic failings.
The dialogue in (59), on the other hand, is straightforwardly grammatical:
(59) A: What has the chairman resigned from?
B: The board/from the board.

The grammaticality of the response confirms that NP and PP are constituents.

Similarly, (60) confirms once more that VP, minus the auxiliary, is a constituent:
(60) A: What has the chairman done?
B: Resigned from the board.

The plot thickens around our mystery category given the ungrammaticality of
the response in (61), however:
(61) A: What has the chairman done?
B: *Has resigned from the board.

Recall that the failure of a constituency test does not show us that a given
sequence is not a constituent; there may be other reasons certain constituents
cannot pass a given constituency test. We have seen evidence in (50), (51) and
(55e) that the mystery category, the sequence Aux + VP, is a constituent. Some
independent factor must then be responsible for the failure of the fragment test
in (61). If fragments can only be phrasal and the mystery category is not phrasal,
that could be the solution; again, we return to this in the next chapter.
The various constituency tests isolate NP, PP, S, VP and the mystery cate-
gory, as we have seen, although not all of them isolate every category. What
about APs? APs come in two different syntactic and semantic contexts, attribu-
tive (modifying a noun, as in sweet cookies) and predicative (constituting, usu-
ally with the help of be, a predicate: the cookies are sweet). Here the simple
adjective sweet forms an AP on its own (just as John and he form NPs on their
own). Let us see how the different APs fare under our constituency tests, begin-
ning with predicative APs:

IanRoberts_9781316519493_C003.indd 69 05-09-2022 19:49:23

 70 3 phrase-structure rules and constituency tests

(62) Permutation:
a. ??It’s sweet that the cookies are t. (clefting)
b. How sweet are the cookies t?     (wh-movement)
c. ?(I expected those cookies to be sweet), and sweet, they really are!
(fronting)

(63) Pro-form:
John likes his cookies sweet → John likes his cookies (just) so.

(64) Ellipsis:
The cookies are sweet and the cake is sweet too.

(65) Coordination:
The cookies are sweet and extremely fattening.

(66) Fragment:
A: How are the cookies?
B: Sweet/very sweet/*very.

From (62–66) we can conclude that sweet is an AP constituent in a predicative sen-

tence like The cookies are sweet. The permutation tests are little tricky: (62a) is mar-
ginally grammatical (some speakers many find that examples like this sound vaguely
‘Irish’; this is because many varieties of English spoken in Ireland do allow this kind
of cleft – again we see that E-English does not correspond to a homogeneous set
of I-languages). Example (62c) is perhaps a little forced but seems grammatical (it
appears that AP-fronting needs a special discourse context similar to VP-fronting:
compare (35), (44a) and (49)). The AP pro-form is so (note again the similarity with
VP), but seems more natural where there is a substantive verb introducing the pred-
icate rather than auxiliary be, as (63) shows. The other tests work straightforwardly.
Now compare attributive APs:
(67) Permutation:
a. *It’s sweet that John ate [ the t cookies ]. (clefting)
b. *How sweet did John eat [ the t cookies ]? (wh-movement)
c. *Sweet, John ate [ the t cookies ].     (fronting)

(68) Pro-form:
John likes his sweet cookies → *John likes his so cookies.

(69) Ellipsis:
*John ate the sweet cookies and the sweet cake.

(70) Coordination:
John ate the sweet and extremely fattening cookies.

(71) Fragment:
A: What kind of cookies did John eat?
B: Sweet ones/*sweet.

Here the only test that works is coordination. In (67), we can interpret the AP
as predicative, with the examples deriving from John ate the cookies sweet, but

IanRoberts_9781316519493_C003.indd 70 05-09-2022 19:49:23

 3.4 Constituency Tests 71

this, as we have seen, is a different syntactic context. For this reason, I brack-
eted the NP to indicate that sweet here corresponds to the prenominal position
inside NP. Understood this way, these examples are clearly ungrammatical.
Similarly, pro-form so and ellipsis are impossible in the NP-internal position.
In (69), the sequence John ate the sweet cookies and the cake is of course gram-
matical, but sweet cannot be recovered as a modifier for cake and so the ellipsis
indicated is ungrammatical. In (71), sweet seems to need a nominal pro-form
ones for grammaticality. Nevertheless, the fact that coordination yields a gram-
matical result supports the idea that adnominal APs are constituents.
Finally, let’s compare the two possible structures for (26b), repeated here:

(26b) They gave [NP Mary ] [AP sweet ] cookies.

(72) a. They gave [ Mary sweet ] [ cookies ].

b. They gave [ Mary ] [ sweet cookies ].

Here are the results of the tests applied to Mary sweet in (72a):

(73) Permutation:
a. *It’s Mary sweet that they gave t cookies. (clefting)
b. *What/how/who did they give t cookies? (wh-movement)
c. *Mary sweet, they gave t cookies. (fronting)

(74) Pro-form:
John gave Mary sweet cookies → *John gave her so cookies.

(75) Ellipsis:
*John gave Mary sweet cookies and Bill gave her sweet cake.

(76) Coordination:
?John gave Mary sweet and Bill extremely fattening cookies.

(77) Fragment:
A: Who did John give what kind of cookies?
B: *Mary sweet.

The only test which does not yield an ungrammatical result is coordination.
However, the (slightly marginal) structure is a special kind of coordinate
structure known as ‘right node raising’, involving permuting cookies to the
right and deleting a copy of this element. So the structure of (76) is actually
something like:

(78) ?John gave Mary sweet cookies and Bill extremely fattening cookies.

What this shows is that cookies is a constituent (a noun), but nothing about
the sequence Mary sweet. So no tests reveal Mary sweet to be a constituent as
in (72a). Combined with the absence of any semantic or functional motiva-
tion for grouping these words together, in that sweet does not modify Mary,
we may conclude that the structure in (72a) is mistaken.
Now let us apply our tests to sweet cookies in (72b):

IanRoberts_9781316519493_C003.indd 71 05-09-2022 19:49:23

 72 3 phrase-structure rules and constituency tests

(79) Permutation:
a. It’s sweet cookies that they gave Mary t. (clefting)
b. What kind of cookies did they give Mary t? (wh-movement)
c. Sweet cookies, they gave Mary t. (fronting)

(80) Pro-form:
John gave Mary sweet cookies → *John gave her them.

(81) Ellipsis:
*John gave Mary sweet cookies and Bill gave Mary sweet cookies too.

(82) Coordination:
John gave Mary sweet cookies and Bill gave Mary extremely fattening cakes.

(83) Fragment:
A: What did John give Mary?
B: Sweet cookies.

Here all the tests work straightforwardly except ellipsis. English does not
allow ellipsis of argument NPs. We can see this from the following contrast:
(84) a. John brought his friend but Bill didn’t bring his friend.
b. *John brought his friend but Bill didn’t bring his friend.

VP, but not NP, elides in English. The opposite is true in Japanese:
(85) a. *Kai-ga piza-o tabeta, Lina-mo piza-o tabe sita. [Japanese]
Kai-nom pizza-acc eat.pst, Lina-also pizza. acc eat do.pst
b. Taroo-wa zibun-no tomodati-o turetekita, demo Hanako-wa zibun-no
tomodati-o
Taroo.top self.gen friend.acc bring.pst, but Hanako.top self.gen
friend.acc
turetekona katta.
bring not.pst
‘Taroo brought his friend, but Hanako did not bring her friend.’

Here we see ungrammatical VP-ellipsis in (85a), alongside grammatical

NP-ellipsis on (85b). For reasons that are not at all well understood, languages
vary as to which categories they allow to elide; we will return to this important
observation in Volumes II and III. Obviously, the ellipsis possibilities a given
language has will affect the results of ellipsis as a constituency test.
Since all the other tests work for the structure in (72b), and none of them work
for (72a), we take (72b) to be the correct structure for the sequence in question.
This is consistent with the semantic, functional observation that sweet modifies
cookies, but not Mary.
Finally, constituency tests can be used to demonstrate that a string of words
is structurally ambiguous, i.e. that the same string of words corresponds to two
distinct structural descriptions with two distinct meanings. An example is the
following sentence (attributed to Groucho Marx):

IanRoberts_9781316519493_C003.indd 72 05-09-2022 19:49:23

 3.4 Constituency Tests 73

(86) I shot an elephant in my pyjamas.

This sentence has two meanings, depending on who was wearing the pyja-
mas. The natural reading is that I was wearing my pyjamas (perhaps because
the elephant broke into my house one night). The other reading is that the
elephant was wearing the pyjamas (this was the one Groucho intended as he
went on ‘I’ve no idea why he was wearing them’).
The two interpretations of the sentence correspond to the following structural
descriptions:
(87) a. I shot [NP an elephant [PP in my pyjamas ]].
b. I shot [NP an elephant] [PP in my pyjamas ].

In (87a), the PP in my pyjamas is inside the object NP and so modifies ele-

phant; this is Groucho’s intended interpretation: the elephant was wearing
his pyjamas. In (87b), that PP is outside the object NP and is interpreted as
modifying the subject, so this is the perhaps more ‘normal’ reading where I
was wearing my pyjamas.
Let us now apply the constituency tests to the sequence an elephant in my
pyjamas ((i)-examples below) and an elephant ((ii)-examples below). I have
written ‘Groucho’ or ‘normal’ after each unambiguous example, ‘Groucho’
meaning the elephant was wearing the pyjamas (structure (87a)) and ‘normal’
meaning I was wearing the pyjamas (structure (87b)). If the example is ambig-
uous, this is indicated; if the example is ungrammatical, no indication of inter-
pretation is given.
(88) Permutation:
a. i. It was an elephant in my pyjamas I shot. (Groucho)
ii. It was an elephant I shot in my pyjamas. (normal/ambiguous)
b. i. What did I shoot? (Groucho/vague)
ii. What did I shoot in my pyjamas? (normal)
c. i. An elephant in my pyjamas, I shot. (Groucho)
ii. An elephant, I shot in my pyjamas. (normal)

(89) Pro-form:
i. I shot it. (Groucho/vague)
ii. I shot it in my pyjamas. (normal)

(90) (VP-)ellipsis:
i. I shot an elephant in my pyjamas and John did shoot an elephant in
my pyjamas too. (Groucho)
ii. I shot an elephant in my pyjamas and John did shoot an elephant in
my pyjamas too. (John)

(91) Coordination:
i. I shot an elephant in my pyjamas and a tiger in my favourite teeshirt.
(ambiguous)
ii. I shot an elephant and a tiger in my pyjamas. (ambiguous)

IanRoberts_9781316519493_C003.indd 73 05-09-2022 19:49:23

 74 3 phrase-structure rules and constituency tests

(92) Fragment:
A: What did you shoot?
B: i. An elephant in my pyjamas. (Groucho)
ii. An elephant. (normal/vague)

The permutation tests, especially clefting in (88a) and fronting in (88c), clearly
distinguish the two structures in (87) and each reading is consistent: where an
elephant in my pyjamas is a constituent, we have Groucho’s interpretation, where
it is not, we have the ‘normal’ one. Example (88b,i) is slightly equivocal, in that
in my pyjamas may be completely unspecified, hence the designation ‘vague’.
However, if forced to choose between the two readings in (87) for (88b,i), the
Groucho one is clearly preferred. Similar considerations apply to (89i): in my
pyjamas may be left unspecified. But (89ii) allows only the ‘normal’ reading.
VP-ellipsis gives an interesting result in (90ii), as the interpretation is that John
was wearing my pyjamas. This is consistent with the structure in (87b), where
pyjamas modifies the subject. The fragment test in (92) gives a completely
unambiguous result in (92i), the Groucho reading, but, again, the possibility of a
vague reading in (92ii). Leaving aside the coordination result for a moment, all
of these tests, with one or two provisos, disambiguate the structural ambiguity
of (86) and reveal that the ambiguity is attributable to the different structural
descriptions, associated in turn with different semantic interpretations, in (87).
The coordination examples in (91) are both ambiguous, however. This is due
to the ambiguity of coordination more generally. The bracketings given in (93)
illustrate the ambiguities:
(93) i. I shot [ an elephant [ in my pyjamas ]] and [ a tiger [ in my favourite teesh-
irt]]. (Groucho)
I shot [ an elephant ] [ in my pyjamas ] and [ a tiger ] [ in my favourite teesh-
irt]. (normal)
ii. I shot [[ an elephant and a tiger ] [ in my pyjamas ]]. (Groucho)
I shot [ an elephant and a tiger ] [ in my pyjamas ]. (normal)

The first example in (93i) coordinates two direct objects each with the internal
structure seen in (87a), and so unambiguously yields the Groucho interpretation.
The second example coordinates a constituent consisting of the sequence NP
PP, where PP is not a constituent of NP and so does not modify NP. This constit-
uent is probably a VP, where the verb in the second conjunct has been ‘gapped’
(a form of ellipsis affecting just V, not VP, as in I spoke to Mary about John
and to Bill about Pete). In the first example of (92ii), both the elephant and the
tiger were wearing my rather capacious pyjamas; clearly a case of the Groucho
reading which is rather more absurd than even he intended. The second exam-
ple in (92ii) is a straightforward case of the normal reading. So the ambiguities
in (91) result from ambiguities regarding exactly which constituents are being
coordinated, disambiguated by the bracketings in (93). Finally, the bracketing in
(94) gives a further example of the ambiguities of coordination:
(94) I shot [ an elephant ] and [ a tiger [ in my pyjamas ]].

IanRoberts_9781316519493_C003.indd 74 05-09-2022 19:49:23

 3.5 Summary and Conclusion 75

Example (94) is clearly interpreted to mean that the tiger, but not the ele-
phant, was wearing my pyjamas. This reading arises from the fact that in my
pyjamas modifies tiger here, since it is a constituent of the same NP. Once
again, we see that constituency tests can probe for structural ambiguity and
reveal that ambiguity to be a consequence of the different structural descrip-
tions, each unambiguously associated with different semantic interpretations.
In this section we have introduced, explained and applied the principal kinds
of constituency tests. Two important points should always be borne in mind in
connection with these tests. First, only a positive result is a real result: if a test
for constituent X fails, it does not show that X is not a constituent; in fact, it does
not show anything. Second, one aspect of syntactic variation across languages
involves the range of permutation operations, pro-forms, ellipses, coordinations
and even fragments a given language allows. So what may work as a good test
in one language (e.g. VP-ellipsis in English) may not work in another language,
as we have just seen in relation to VP-ellipsis in Japanese.

3.5 Summary and Conclusion

In this chapter we have seen how PS-rules generate the well-
formed structural descriptions in a grammar. This formal, technical notion of
well-formedness (and its contrary, ill-formedness, i.e. putative structural descrip-
tions not generated by PS-rules) aligns fairly closely with the intuitions concern-
ing acceptability of native speakers of the I-language the grammar is a theory
of. The PS-rules are therefore an important part of our account of native-speaker
competence. Some basic PS-rules were introduced and illustrated in Section 3.2.
In Section 3.3, we showed how even quite small sets of PS-rules can interact
so as to generate structural descriptions of unlimited length and depth. This is the
recursive nature of systems of PS-rules, a feature of enormous theoretical impor-
tance, since it accounts for the unlimited expressive power of language. Recursive
PS-rule systems directly capture the fact that natural language makes infinite use of
finite means. Since this is an aspect of human cognition (seen not only in language
but also in numbers), PS-rules formally capture an important aspect of cognition.
The structural descriptions generated by PS-rules consist of particular arrange-
ments of categories and constituents that we can represent as tree diagrams or
labelled bracketings (we have seen numerous examples of both in this chapter).
The particular choices of structural descriptions can be supported by constitu-
ency tests of various kinds, which we introduced and illustrated in Section 3.4.
The constituency tests are, to a degree which remains to be fully determined, in
part language particular; the ones we have seen work quite well for English, but
they may not give such clear results when applied to other languages. For this
reason, and also because they can only verify a postulated structural description,
failure of a constituency test to isolate constituent X does not prove that X is not
a constituent. Constituency tests may also reveal the presence of categories we
did not expect to find, such as the mystery category in (52) and (55d).

IanRoberts_9781316519493_C003.indd 75 05-09-2022 19:49:23

 76 3 phrase-structure rules and constituency tests

As we have pointed out several times, PS-rules simultaneously specify three

kinds of formal information: relations of linear precedence among constituents,
relations of hierarchical dominance/constituency among constituents and the
category labels of the constituents. Being purely formal devices, however, they
do not on their own specify semantic or functional relations among constitu-
ents, such as ‘subject’, ‘predicate’, ‘modifier’, etc. For this, we need to add
something to our theory, a matter we come back to in Section 5.2. But we also
need to limit the formal power of the rules themselves. This will be the goal of
the next chapter.

Exercises
1. Phrase-Structure rules
Take the following set of PS-rules:
i. S → NP VP
ii. VP → V (NP) (PP)
iii. NP → (D) N (PP)
iv. PP → P NP
Give five structurally distinct, grammatical English sentences that
these rules generate, showing their structure as a tree diagram.
2. Give the tree diagrams for the following sentences:
a. Goona hoped that Vurdy would eat the curry.
b. Harry thought that Ron said that Hermione knew the answer.
c. Pep wanted to know how much money Paul would want.
d. Andy wondered whether Lou would write a song.
e. Ron wondered whether to talk to Hermione.
3. Using the constituency tests discussed in this chapter (permu-
tation, pro-forms, ellipsis, coordination, fragments), justify the
constituent structure you have assigned to one of the sentences in
(1) and (2).
4. The following sentences are structurally ambiguous. Explain the
ambiguity, assign distinct structural representations (i.e. draw a tree
or give labelled brackets) corresponding to each reading of the sen-
tence, and show how constituency tests support the structures you
propose in each case.
a. Mary ate the cake on the shelf.
b. Olly dreamt that his team won last night.
c. The chickens are ready to eat.

IanRoberts_9781316519493_C003.indd 76 05-09-2022 19:49:23

 Further Reading 77

For Further Discussion

At the beginning of his 1957 book Syntactic Structures (The Hague: Mouton),
Noam Chomsky writes that ‘[m]ost of our linguistic experience, both as speak-
ers and hearers, is with new sentences; once we have mastered a language, the
class of sentences with which we can operate fluently and without difficulty or
hesitation is so vast that for all practical purposes (and, obviously, for all the-
oretical purposes), we may regard it as infinite’ (p. 7). How do PS-rules of the
kind we have seen in this chapter connect to this observation?

Further Reading
All of the readings given here introduce the notions of category, constituency and con-
stituency tests, with some variations in emphasis and details. Sportiche et al. (2014) are
particularly thorough and Freidin goes into interesting detail on coordination.
Carnie, A. 2010. Constituent Structure. Oxford: Oxford University Press, Chapter 2.
Carnie, A. 2013. Syntax: A Generative Introduction. Oxford: Blackwell, Chapters 3 and 4.
Freidin, R. 2012. Syntax: Basic Concepts and Applications. Cambridge: Cambridge
University Press, Chapter 3.
Larson, R. 2010. Grammar as Science. Cambridge, MA: MIT Press, Unit 7.
Sportiche, D., H. Koopman & E. Stabler. (2014). An Introduction to Syntactic Analysis
and Theory. Oxford: Wiley, Chapter 3.
Tallerman, M. 1998. Understanding Syntax, London: Routledge, Chapter 5.

IanRoberts_9781316519493_C003.indd 77 05-09-2022 19:49:23

 4 X′-theory

4.1 Introduction
In the previous chapter we saw how PS-rules provide structural
descriptions of sentences, in the form of labelled bracketings or tree diagrams,
in terms of the key notions of categories and constituents. PS-rules specify the
grammatical structures (the well-formed structural descriptions) of a given gram-
mar, corresponding as far as possible to the I-language grammar of competent
native speakers. PS-rules, or systems of PS-rules, can be recursive, in that they
apply to their own output; they thereby specify an infinite set of well-formed
structural descriptions. We saw several illustrations of this is in the previous
chapter and in Chapter 1 with the fish examples. Constituency tests of various
kinds show, on the basis of native-speaker well-formedness judgements, which
structural descriptions are correct. Thus these tests can support the postulation
of a given set of particular PS-rules. In this chapter we will revise and simplify
the system of PS-rules.

4.2 Possible and Impossible PS-Rules

In (1), we see some PS-rules of English. We have already seen some
of these, e.g. (1a) in the previous chapter:
(1) a. NP → (D) N (PP)
b. AP → (Mod) A (PP)
c. PP → (Mod) P (NP)
d. VP → (AdvP) V (NP)

In (2a–d), we see the respective structural descriptions rules that (1a–d) gener-
ate, presented as labelled bracketings. Each one corresponds to the full expan-
sion of each rule, i.e. with optional material realised:
(2) a. [NP [D the ] [N picture ] [PP [P of ] [NP [N John ]]]]
b. [AP [Mod very ] [A angry ] [PP [P with ] [NP [N John ]]]]
c. [PP [Mod just ] [P beyond ] [NP [D the ] [N frontier ]]]
d. [VP [AdvP really ] [V enjoy ] [NP [D the ] [N movie ]]]

78

IanRoberts_9781316519493_C004.indd 78 05-09-2022 18:15:35

 4.2 Possible and Impossible PS-Rules 79

The examples in (2) show us that the PS-rules in (1) generate a series of well-
formed structural descriptions and as such are PS-rules in good standing. In each
case there is a lexical category (N, V, A or P) on the right of the arrow corre-
sponding to the phrasal category on the left of the arrow.
Now compare the rules in (3):

(3) i. NP → V AP
ii. PP → N V
iii. AP → V NP

None of these rules has a lexical category on the right of the arrow correspond-
ing to the phrasal category on the left. These rules are certainly not rules of
English: there is no NP constituent which consists of a verb and an AP, no PP
which contains the sequence N V and no AP consisting of V and NP. The strings
found to the right of the arrows do exist: look nice is a case of V AP; Clover slept
is the string N V and any transitive verb followed by its object is a case of V NP
order (eat a mouse, like cake, visit London, etc.). But look nice is not an NP, as
we can see if we apply the basic distributional test for NPs given in Chapter 2.
That test consisted in seeing whether a putative NP could appear in subject posi-
tion, i.e. in the dash position in (4):

(4) ___ can be a pain in the neck.

(5) shows us that the result of inserting look nice in that position is
ungrammatical:

(5) *Look nice can be a pain in the neck.

Similarly, we saw in Chapter 2 that N V sequences like Clover slept are

really NP VP sequences, which make up an S. They cannot form a PP. This
is shown by the fact that PP-modifiers like straight and right can’t precede
them:

(6) *straight/right Clover slept

Furthermore, V NP sequences like eat a mouse, etc., cannot appear in the

AP-frame we identified in Chapter 2:

(7) *The students are very eat a mouse/like cake/visit London, …

We can safely conclude that, while the PS-rules in (2) certainly do gener-
ate well-formed structural descriptions for English, those in (3) do not.
Languages differ somewhat from one another regarding their PS-rules, but
we might reasonably ask whether any language would counter structures of
the kind generated by the rules in (3). Since each PS-rule specifies a relation
of immediate dominance/constituency, the rules in (3) generate trees of the
kind shown in (8):

IanRoberts_9781316519493_C004.indd 79 05-09-2022 18:15:35

 80 4 X′- t h e o r y

(8) i. NP

V AP

ii. PP

N V

iii. AP

V NP

Structures of this kind have never been postulated, and it seems highly unlikely
that they ever will be. If it is true that these structures are part of the grammar of
no natural language, then our theory of UG must rule them out. That means that
the PS-rules in (3) should be ruled out of our theory. The theory of PS-rules, as
we presented it in the previous chapter, does not do this. Hence we must develop
a way of doing this.

4.3 Introducing X′-theory

What is the difference between the ‘good’, attested, PS-rules in (1)
and the ‘crazy’, unattested (as far as we know) PS-rules in (3)? Let us look again
at both sets, side by side:
(1) a. NP → (D) N (PP)
b. AP → (Mod) A (PP)
c. PP → (Mod) P (NP)
d. VP → (AdvP) V (NP)

(3) i. NP → V AP
ii. PP → N V
iii. AP → V NP

The main difference between the rules in (1) and the rules in (3), as we observed
in the previous section, is that those in (1) have ‘different versions’ of the same
category on each side of the arrow, while that is not true of those in (3). In fact,
the rules in (1) have phrasal categories on the left of the arrow and word-level
versions of the same categories on the right of it. Not only that, but everything
else in the expansion of the respective categories is optional. So the rules in (1)
could be written in a more schematic form as:
(9) a. NP → … N …
b. AP →… A …
c. PP →… P…
d. VP →… V …

We have seen examples of the minimal expansions of each of these categories:

IanRoberts_9781316519493_C004.indd 80 05-09-2022 18:15:35

 4.3 Introducing X’-theory 81

(10) a. [NP [N Clover ]]

b. [AP [A sweet ]]
c. [PP [P on ]]
d. [VP [V slept ]]

So one generalisation about the well-formed PS-rules in (1) is that they all consist
of a phrasal category to the left of the arrow and at least a single-word version of
that category to the right. Call the single-word version of the phrasal category the
head of the phrase, or just the head. So, the generalisation is that whatever else
appears to the right of the arrow, the head of the phrase that appears to the left must
appear there. To put it another way, every phrase must have a head. The technical
term for ‘having a head’ is endocentric (as opposed to exocentric, not having a
head). So every phrase generated by the well-formed PS-rules has a head. This is
the endocentricity requirement: virtually all phrasal categories are headed.
Now we can see what is wrong with the PS-rules in (3): they all generate exo-
centric categories. None of the rules in (3) meets the endocentricity requirement,
while all of those in (1) do. We can generalise the endocentricity requirement by
generalising over the partial rules in (9) as follows:
(11) XP → … X …

Here ‘X’ is a categorial variable ranging over the syntactic categories (just as in
algebra x is a variable ranging over numbers). Since all the good PS-rules in (1)
conform to (11) and none of the crazy ones in (3) do, let us make (11) a general
requirement on PS-rules. So all PS-rules must conform to the template in (11).
This significantly limits the number and the form of PS-rules. So we have now
successfully eliminated the PS-rules in (3).
Looking again at the well-formed PS-rules in (1), we can make a further gen-
eralisation. In addition to requiring a head of the same category as the category
on the left of the arrow, all the rules allow optional material both to the left and
to the right of the head. In the case of AP, PP and VP, the material to the left of
the head is always modificational and, as such, optional. In the case of NP, it is
D(eterminer), which we will take for now to also be a kind of modifier although
we know that Ds are sometimes obligatory (e.g. singular count nouns cannot
stand alone in subject or object position: recall *Boy fish from Section 1.4).
Turning to the material following heads in (1), we see that N and A are followed
by PP, while V and P are followed by NP. In (2), we exemplified these as follows:
(12) a. The picture of John
b. very angry with John
c. just beyond the frontier
d. really enjoy the movie

In (12c) and (12d) the NPs are clearly objects of P and V respectively. We can
maintain the same for N and A in (12a,b), noting that these objects are obligato-
rily prepositional, i.e. they must be introduced by a Preposition (of, with).

IanRoberts_9781316519493_C004.indd 81 05-09-2022 18:15:35

 82 4 X′- t h e o r y

So, the well-formed PS-rules obey the endocentricity requirement in (11),

have modifiers to the left of the head and complements to the right. In other
words all the good rules conform to the template in (13):
(13) XP: (Modifier) X (complement)

The material preceding X is called the Specifier of X; this term generalises

over truly optional modifiers like right, very, etc., and determiners: the notion
is that what is common to all these pre-head elements is that they specify some-
thing about the head. So the Determiner ‘specifies’ something about the Noun
(through quantification, for example); the modifier ‘specifies’ something about
the A, P, etc.
The material following X is called the complement of X. Complements may
be obligatory; this depends on lexical properties of X. This is easiest to see with
verbs, as verbs have the richest range of complements. As we briefly saw with
put in the previous chapter, verbs categorially select, or c-select, for their com-
plements. Here are some further examples of verbal c-selection requirements:
(14) a. watch(V) ___ NP    So: VP → V NP
b. rely(V) ___ PP     So: VP → V PP
c. put(V) ___ NP PP    So: VP → V NP PP
d. say(V) ___ S′      So: VP → V S′

(Strictly speaking, verbs and other categories c-select the head of their comple-
ment, but the endocentricity requirement has the effect that c-selection introduces
phrasal categories corresponding to the selected head.) On the left in (14a–d),
we see examples of the lexical entries of verbs. The lexicon is the mental dic-
tionary (stored in the long-term memory of every competent native speaker of
an I-language) containing idiosyncratic information about all the words in the
I-language (i.e. all the words a given speaker knows). We will say more about
selection and the lexicon in Chapter 8 and in Volume II; for now, the simplified
lexical entries given in (14) suffice.
The lexical entry on the left of (14a) says that watch is a verb that c-selects an
NP; the ‘__NP′ notation means the verb must appear in the position immediately
before NP. On the right in (14a), we have the PS-rule which expands VP so as
to create the structure required by the c-selection property of watch. This rule
generates NP as the complement of V. Clearly, watch here represents the very
large class of basic transitive verbs, verbs which take just a direct object. Having
a lexical entry like this (with further information about the phonological and
semantic properties of the word) stored in long-term memory is a part of what it
means for a native speaker to ‘know’ the word.
The lexical entry in (14b) states that rely c-selects a PP, which in fact must
be headed by on: rely on/*with/*for/*up John. The PS-rule generating the corre-
sponding expansion of VP, needed to accommodate this requirement, generates a
PP in the complement of V. The lexical entry in (14c) is the one for put, which we
briefly saw in Chapter 3. Here VP needs both an NP and a PP position, since put

IanRoberts_9781316519493_C004.indd 82 05-09-2022 18:15:35

 4.3 Introducing X’-theory 83

c-selects both; see the discussion around (42) in Chapter 3. Finally, in (14d) we
have the lexical entry for say, which specifies that this verb c-selects S′: say [S′
that the world is round ]. The corresponding expansion of VP is given on the right.
Putting all the PS-rules on the right of (14) together, we come to something fairly
close to the expansion of VP given in (24iii) of Chapter 3.
In (15), we see the first version of the full ‘Xʹ-schema’ for PS-rules (pro-
nounced ‘X-bar’):
(15) a. XP → Specifier X′
b. X′ → X complement

(Recall that X is a categorial variable whose possible values are one of N, V, A,

P.) The rules in (15) are slightly more complicated than (13), in that they intro-
duce an extra level of structure; this is the X′-level which gives this version of
the theory of phrase structure its name.
The tree structure generated by the rules in (15) is (16):
(16) VP

specifier X′

X complement

In (16) we can readily see how the X′-level creates different levels of syntactic
‘closeness’ for the complement and the Specifier in relation to the head. This
relative ‘closeness’ is structurally represented. The complement is closer to the
head, since it is obligatory and its category depends on the lexical c-selection
properties of the head, as illustrated for various verbs in (14). The Specifier is
usually (but not always, as we have seen) optional and modifies or ‘determines’
the head in various ways. The structure in (16) represents the greater relative
closeness, both syntactic and semantic, of the complement to the head.
The terms ‘Specifier’ and ‘complement’, although more abstract than ‘sub-
ject’, ‘object’, etc., are equally relational. We have repeatedly stated that
PS-rules do not (directly) give relational information. These relational terms
should be replaced with categories. With the apparent exception of S′, all
the complements in (14) are phrasal; in X′-theory terms, they are all XPs.
To distinguish them categorially from the head X that c-selects them, let us
call them YP (where Y is just another, distinct categorial variable). We will
return to the status of S′ below. The Specifier elements, with the exception
of D, are all phrasal modifiers. For uniformity, let us also then treat D as
phrasal, DP. Then Specifiers are always phrasal. To distinguish them from
the X′-category they specify and from the YP complement to X, call them
ZP (Z is yet another distinct categorial variable). So the correct version of the
X′-schema for PS-rules is (17):
(17) a. XP → ZP X′
b. X′ → X YP

IanRoberts_9781316519493_C004.indd 83 05-09-2022 18:15:36

 84 4 X′- t h e o r y

The tree diagram representing the structural description (or the template for
structural descriptions, where X, Y and Z are all categorial variables) generated
by (17) is given in (18):
(18) XP

ZP X′

X YP

Y and Z are further categorial variables, substitutable with different values from
X. So YP and ZP are each expanded as in (17) and so have the internal structure
seen in (18). Clearly, then, the X′-system is fully recursive.
We have posited a new kind of constituent in (17) and (18): the interme-
diate level category X′. So far, we have only given a functional justification
for X′, saying that it allows us to distinguish the relative closeness of comple-
ments to ‘their’ heads, as opposed to Specifiers. But we must be able to prove
its presence using constituency tests of the kind introduced in Section 3.4.
The coordination test clearly isolates X′, in that X′-level constituents can be
coordinated:
(19) a. the [ pictures of John ] and [ books about Mary ]
b. just [ beyond the frontier ] and [ down the road ]
c. very [ angry with John ] and [ worried about Mary ]
d. really [ enjoy the movie ] and [ dislike the book ]

In (19a), the modifies the head of both conjuncts: books about Mary can be
interpreted as definite despite the fact that the string books about Mary can stand
alone as a plural indefinite NP (with a silent determiner). Similarly, in (19b–d)
Mod can apply to both conjuncts in each case.
The permutation tests do not give good results, however:
(20) a. *It’s pictures of John that we saw the t.     (clefting)
b. *Which pictures of John did you see the t?    (wh-movement)
c. *Pictures of John, we saw the t.        (fronting)

The pro-form test does give a good result, using one as the pro-form, as the fol-
lowing examples illustrate:
(21) a. John read a book of poems and Mary read one too.
b. *John read a book of poems and Mary read one of stories.
c. John read a book with a red cover and Mary read one too.
d. John read a book with a red cover and Mary read one with a blue cover.

In (21a), one appears to stand for the whole NP the book of poems. The unac-
ceptability of (21b) shows that one must stand for the complement of book, the
PP of stories. In (21c), one once again appears to stand for the whole NP the
book with a red cover. But in (21d) we see that with a blue cover is not obligato-
rily part of the constituent one stands for. Here we see a pattern similar to the one

IanRoberts_9781316519493_C004.indd 84 05-09-2022 18:15:36

 4.3 Introducing X’-theory 85

we saw with the examples in (41) in Chapter 3 the with-phrase, which is not a
complement of book but an optional modifier, does not have to be substituted by
the pro-form one. On the other hand, the of-phrases in (21a,b) are complements
of book (they are not obligatory, but few complements to nouns are fully oblig-
atory) and, as the unacceptability of (21c) shows, must be part of the constituent
one stands for. Since the occupies the Specifier of N′ (abbreviated SpecN′), and
the complement must be part of what one stands for along with the head, we
conclude that one is an N′-proform. We will come back to the question of the
with-modifiers in Section 4.5 below; clearly they cannot be in the complement
of N. So the structure of the NP one with a blue cover in (21d) is as in (22):
(22) NP

N′

N PP
one with a blue cover

With certain Specifiers, N′-ellipsis is also possible:

(23) a. John likes these shirts but I like those shirts.
b. John read these books of poetry and I read those books of poetry.
c. *John read these books of poetry and I read those books of poetry.
d. John read these books with red covers and I read those books with blue
covers.
e. John read two books and I read three books.
f. *John read every book but I read no book.

Other determiners do not allow N′-ellipsis, either requiring the pro-form one, or
not allowing ellipsis at all. We do not need to go into the details of this here; it
suffices to note, once again, that the conditions under which ellipsis is possible
are hard to define.
Finally, N′-fragments are not allowed:
(24) A: What did you read?
B: *Book of poems.

We conclude that the coordination, pro-form and ellipsis tests reveal that N′is a
constituent.
Concerning the other X′ categories, neither P′ nor A′ can be permuted, like N′:
(25) a. *On the nose, she hit him right t. (Radford 2016:126)1
b. *Proud of him, she certainly seems to be very t.    (Radford 2016:127)

P′ can be substituted by the pro-form there:

(26) She hit him right there.

1
See Chapter 2, Further Reading.

IanRoberts_9781316519493_C004.indd 85 05-09-2022 18:15:37

 86 4 X′- t h e o r y

A′ can, at least marginally, be substituted by the pro-form so (taking the compar-

ative marker more to be in SpecA′) in predicative position but not in attributive
position:
(27) a. John is nice, but Mary is more so.
b. *John is a nice man, but Bill is a more so man.

PPs and APs can elide in predicative position (we did not look at these exam-
ples of PP-ellipsis in the discussion of ellipsis in Chapter 3), but P′ and A′
cannot:
(28) a. John is intelligent/in London, and Bill is intelligent/in London too.
b. *John is very intelligent/right in London, and Bill is very intelligent/right in
London too.

So coordination and pro-forms provide some evidence for P′ and A′. I will leave
the question of evidence for V′ aside, since we will entertain in later chapters
a very different view of what is in SpecV′ from that suggested in (1d) and (2d)
above.
X′-theory presents a category-neutral view of hierarchical relations and linear
precedence; the central idea is that these are the same for all categories. As such,
idiosyncratic categorial information is removed from the PS-rules and supplied
independently by categorial features: thus, the way in which we give a value to
the categorial variable is by assigning a feature to it. However, the endocentric-
ity of all X′-structures has one important consequence: if we specify the head
X of a category as, say, N, then X′ must be N′ and XP must be NP. This is the
notion of projection of categorial features. This idea leads to the following ter-
minological conventions for X′-structures:
(29) XP

ZP X′

X YP

In a representation like this one, XP is called the maximal projection (of X), X′
is called the intermediate projection (of X), X is the head (of both X′ and XP)
and, as we have seen, ZP is the Specifier of X′ (abbreviated SpecX′) and YP is
the complement of the head X. The dominance and constituency relations are
clear: XP dominates everything in XP and immediately dominates ZP and X′,
while X′ immediately dominates X and YP.
As we have seen, the nature of the complement depends on the c-selection
properties of the head. We saw examples of this with various verbs in (14).
Certain heads (common nouns, intransitive verbs, simple adjectives and some
prepositions) may not c-select a complement at all:
(30) a. Clover [VP slept ].
b. [NP the cat ]

IanRoberts_9781316519493_C004.indd 86 05-09-2022 18:15:37

 4.4 X′-theory and Functional Categories 87

c. [AP fat ]
d. John walked [PP out ].

Here, the relevant lexical entries specify that X c-selects nothing. For example,
compare the lexical entry for sleep with those in (14):
(31) sleep(V)__

Given (31), we must allow for a unary-branching expansion of V′. The same
holds for N′, A′ and P′ in (30). Nonetheless, as a default we will assume that the
hierarchical structure is complete. A proper name like Goona, for example, has
the full non-branching structure shown in (31):
(32) NP

N
Goona

Of course, no information is lost if the full specification of this structure is left

out, but X′-theory is assumed to operate in a consistent way. We will return to
this issue in Chapter 8 of Volume II.
In this section we have seen the basics of X′-theory. This approach reduces
the PS-rules to the rule schemata in (17), giving rise to the structural descrip-
tion (or template for structural descriptions) in (18). In this way, crazy rules
like those in (3) are eliminated, since they cannot be reduced to (17). In (17),
there is a further intermediate level of structure, the X′-level. We saw evi-
dence from constituency tests for N′, A′ and P′ in favour of this level, leaving
V′ aside for now. Finally, we introduced the relevant category-neutral ter-
minology along with the idea of projection of categorial features. In contrast
to more traditional PS-rules of the kind introduced in Chapter 3, the rules in
(17) do not give category information: they do not provide constant categorial
labels for nodes other than the categorial variables X, Y... Like the traditional
PS-rules, however, the X′ rules do provide information about hierarchical
relations of dominance and constituency and linear relations of precedence.

4.4 X′-theory and Functional Categories: The Structure

of the Clause
In the previous section, we saw the basics of X′-theory in relation
to the lexical categories N, V, A and P. Here we will extend the approach to
functional categories, concluding that all categories are generated according to
the schema in (17), i.e. resulting in structures like (18) (with the possibility of
YP and ZP being absent).

IanRoberts_9781316519493_C004.indd 87 05-09-2022 18:15:38

 88 4 X′- t h e o r y

Let us begin with auxiliaries. How do they fit the X′-schema? One possibility
would be to treat the combination of the main verb and auxiliary as a complex
verb form inside VP, along the lines shown in (33):
(33) [V′ [V Aux V ] NP ]

Constituency tests show us that this is wrong. We saw in Chapter 3 that

VP-fronting, substitution of the VP pro-form do so, VP-ellipsis and coordination
all show that VP is a constituent separate from the auxiliary:
(34) a. (We expected John to talk to Mary),
… and [VP talk to Mary ], he will/can/should/would t.
b. You said John should [VP talk to Mary] and he has already [VP done so].
c. I have [VP talked to Mary ] and John has talked to Mary too.
d. The chairman has [VP resigned from the board ] and [VP gone abroad ].

This evidence rules out a structure like (33). We also saw in Chapter 3 that the
exact counterpart of (34c), but with has also apparently deleted, is grammatical
too:
(35) I have talked to Mary and John has talked to Mary too.

The sequence in (35) is ambiguous, in that it can also mean ‘I have talked both
to Mary and to John’; this latter reading involves coordination inside the indirect
object and no ellipsis. The reading that the strike-out in (35) indicates is the one
where John is the subject, not the indirect object, of talked. We also saw that
alongside (34d), we also have coordination patterns where the auxiliary is coor-
dinated along with the verb and its complements:
(36) The chairman has resigned from the board and is living in Ruritania.

So we see that constituency tests tell us that the sequence containing the verb
and its complements is separate from the auxiliary, and that the auxiliary can
combine with the verb and its complements to form a constituent. In Chapter 3,
we referred to the latter as the mystery constituent, treating the former as VP:
(37) [S [NP John ] [?? [Aux has ] [VP passed the exam ]]]

(37) represents one possibility. Another possibility is that Aux is in SpecV′ and
what is labelled VP in (37) is actually V′. In other words, (38) could be the cor-
rect representation:
(38) [S [NP John ] [VP [Aux has ] [V′ passed the exam ]]]

There are four reasons to favour (37) over (38). First, the verb+complement
sequence (VP in (37), V′ in (38)) acts more like an XP than an X′ in relation
to constituency tests. Second, complex sequences of auxiliaries can appear
which do not act like a single constituent, but in which each auxiliary acts
like a complement of the preceding one. Third, the auxiliaries carry sentential

IanRoberts_9781316519493_C004.indd 88 05-09-2022 18:15:38

 4.4 X′-theory and Functional Categories 89

information regarding negation and interrogatives; SpecV′ does not seem the
right place for carriers of such information. Fourth, we will see that when we
bring the apparently exocentric category S into line with the X′-template there
is a natural head position that we can place auxiliaries in. Let us look at these
reasons one by one.
First, as we saw above several constituency tests for X′ fail: permutation,
ellipsis (except N′-ellipsis under demonstratives and numerals) and the fragment
test. But the putative V′ in (38) passes them all (permutation is restricted to
fronting, for reasons given in Chapter 3):
(39) a. (I expected John to pass the exam), and [ passed the exam ] he has!
b. John passed the exam and Mary did pass the exam too.
c. A: What did John do?
B: Pass the exam.

The other XPs pass the permutation, ellipsis (with the exception of NP) and
fragment tests and none of the X′s do. This makes the putative V′ in (38) look
very different to the other categories. On the other hand, if we treat this sequence
as VP, as in (37), it falls into line with the other XPs. Since X′-theory treats all
lexical categories as participating in the same structural configurations, (37) thus
appears to be a more desirable option than (38).
The second argument relates to sequences of auxiliaries. English allows up to
four auxiliaries in a row, rigidly ordered as Modal > Perfect have > Progressive
be > Passive be, as in:
(40) (When you arrived), John must have been being interviewed.

If auxiliaries are Specifiers of V′, then (40) must have the structure in (41):
(41) [S [NP John ] [VP [Aux? must have been being ] [V′ interviewed ]]]

Constituency tests, particularly ellipsis, show us, however, that the sequence
‘Aux?’ in (41) comprises several separate constituents, along with V′ (some
speakers, especially Americans, find (42d) ungrammatical, but I don’t; this is
further evidence that E-English does not correspond to a homogeneous set of
I-languages and does not affect the argument being made here):
(42) John must have been being interviewed right then, and …
b. Bill must have been being interviewed too.
c. Bill must have been being interviewed too.
d. Bill must have been being interviewed too.
e. Bill must have been being interviewed too.

Any sequence of auxiliaries and the main verb can be elided, but not a passive
main verb on its own (recall that this doesn’t show that the passive main verb is
not a constituent). So we must have a structure like (43) for the auxiliary+main
verb sequence where each bracketed constituent can be elided:

IanRoberts_9781316519493_C004.indd 89 05-09-2022 18:15:38

 90 4 X′- t h e o r y

(43) [ must [ have [ been [ being interviewed ]]]]

We could label (43) as in (44), in line with (38):

(44) [VP must [V′ have [V′ been [V′ being interviewed ]]]]

But now V′ looks very different from other X′ categories, in that it can have three
Specifiers (or probably four, since being could be a fourth SpecV′ although the
ellipsis test doesn’t show this). Furthermore, X′-ellipsis is hardly found with the
other lexical categories. Instead, then, we could treat each V′ in (44) as an AuxP of
some kind. This is consistent with (37), but not with (38), so again (37) is favoured
The third argument concerns negation and question-formation (in main clauses).
The rule for negation of finite clauses in English is ‘insert not after the first auxil-
iary’. We see this rule in action in (45), where not is contracted to n’t and attached
to the auxiliary (in some cases causing a phonological change to the auxiliary):
(45) a. John can’t speak Mandarin.
b. John won’t leave tonight.
c. John hasn’t passed the exam.
d. John isn’t writing a book.

If there is no auxiliary in the positive sentence, the ‘dummy’ auxiliary do must

appear in the negative:
(46) John doesn’t speak Mandarin.

Negation is semantically a sentence-level operation, denying the proposition

asserted by the sentence. The structure in (38) obliges us to place not in SpecV′,
which seems an odd place for an element whose meaning affects the whole
clause. At the very minimum, this complicates the relation between syntax and
semantics. We will see below that such a complication is unnecessary.
As we saw in Chapter 2, auxiliaries undergo ‘subject-aux inversion’ in main-
clause questions:
(47) a. Cambridge will flood as a consequence of global warming.
b. Will Cambridge flood as a consequence of global warming?

Subject-auxiliary inversion clearly affects the meaning of the whole sentence,

changing it from a declarative to an interrogative. As with negation, this seems
an odd thing to attribute to a category that normally occupies SpecV′ and leads to
an unnecessarily complex relation between syntax and semantics. Furthermore,
we have to find a way to account for the permutation relating SpecV′ and the
initial position in the sentence in order to account for the purely formal relation
between (47a) and (47b), while we will see in Chapter 5 that there is a simpler
alternative in terms of (37). Once again, (37) is favoured over (38).
Our fourth reason to prefer (37) over (38) is that it helps us bring S into line
with X′-theory. Let us look again at the structure in (37):

IanRoberts_9781316519493_C004.indd 90 05-09-2022 18:15:38

 4.4 X′-theory and Functional Categories 91

(37) [S [NP John ] [?? [Aux has ] [VP passed the exam ]]]

When we first discussed the mystery category ‘??’ here in the previous c­ hapter,
we pointed out that it entails a complication of the basic PS-rule for S:
(48) S → NP VP

We could simply add Aux to (48), as follows:

(49) S → NP Aux VP

However, rule (49) does not fit the X′-schema, and we have evidence from coor-
dination and ellipsis that Aux and VP form a constituent distinct from NP here.
At the very minimum, then, we need the two rules in (50):
(50) i. S → NP ??
ii. ?? → Aux VP

The rules in (50) will generate the structure in (37). So, the question we must
now answer is: what category is ‘??’?
X′-theory now suggests a simple answer: Aux is a head, ‘??’ is the interme-
diate projection of Aux, Aux′, and S is really the maximal projection of Aux,
AuxP. In this way, Aux and its projections conform to the general X′-theory
template; S ceases to be an exocentric exception to X′-theory; we know what
‘??’ is (Aux′); the structure is consistent with the results of the various constitu-
ency tests we have been applying and consistent with the fact that negation and
question-formation are sentence-level operations applying to Aux, since Aux is
the head of the sentence.
The tree diagram (51) gives the full structural description of a simple sentence
as AuxP, generated by the rules in (50):
(51) AuxP

NP Aux′
he
Aux VP
should
V PP
talk
P NP
to
D N
this man

Here we see that the subject NP is in SpecAux′ and VP is the complement of

Aux. Everything is fully in line with the X′-template.
What we have seen so far answers the question about the nature of the mystery
category and brings S into line with X′-theory where an auxiliary is present. But
what about sentences without an auxiliary, i.e. simple examples like the following?

IanRoberts_9781316519493_C004.indd 91 05-09-2022 18:15:38

 92 4 X′- t h e o r y

(52) a. John left.

b. Mary smokes.
c. Fish fish fish.

Although there is no auxiliary in examples like (52), there is still a tense: (52a)
is marked for past tense, (52b) is, at least indirectly, marked for present tense
by the third-person singular agreement and, as we argued in Chapter 1, the verb
fish in (52c) has a silent present-tense marker. Auxiliaries often mark tense too,
notably English future will and shall, while have marks perfect, which is at least
partly a tense. So let us rechristen AuxP as TenseP, or TP. This is arguably a
less parochially English label than AuxP, as very many languages have tense
marking but auxiliaries are arguably somewhat rarer. Very few languages have
an auxiliary system as rich as English, allowing, as we saw in (42), up to four
auxiliaries in a row.
So we replace (51) with (53):
(53) TP

NP T′
he
T VP
should
V PP
talk
P NP
to
D N
this man

Here the subject is in SpecT′, the head of the clause is T and VP is the comple-
ment of T in T′.
We can make some sense of the English phenomenon of ‘do-support’ in terms
of the structure in (52). The basic rule is that if no auxiliary is independently
present in the main clause, do appears where VP is elided, where the sentence is
negated or where subject-auxiliary inversion takes place:
(54) a. Sarah finished her paper, and Mary did finish her paper too.
b. Sarah didn’t finish her paper.
c. Did Sarah finish her paper?

The appearance of do where the verb is elided implies that tense and agreement
must always be expressed. If there is no verb or auxiliary independently availa-
ble, do is inserted. Similarly, we can think that the presence of negation in (53b)
and the intervening subject in (53c) mean that the operation combining T with
V, which clearly takes place in examples like (51a,b) (and covertly in (51c)),
is somehow prevented from applying, and so the meaningless auxiliary do is
inserted, in order to mark tense and agreement. The auxiliary do should be dis-
tinguished from the homophonous main verb do, as in do the tango: the latter do
also requires do-support in the relevant contexts, as (54) shows:

IanRoberts_9781316519493_C004.indd 92 05-09-2022 18:15:39

 4.4 X′-theory and Functional Categories 93

(55) a. Mary did the tango.

b. Mary didn’t do the tango.
c. *Mary didn’t the tango.
d. Did Mary do the tango?
e. *Did Mary the tango?

Generalising from AuxP to TenseP also gives us a way to analyse infinitival

to, a question we left open in Section 3.3. There we saw that there were two pos-
sible analyses for the infinitival complement sentence in Goona arranged for [
Vurdy to arrive on time ], repeated here:
(56) a. [S [NP Vurdy] [? to ] [VP arrive on time ]]
b. [S [NP Vurdy ] [VP to arrive on time ]]

Two of the four arguments we gave for treating VP in finite clauses as a con-
stituent separate from Aux/T apply to the infinitives with to. Constituency tests
such as fronting, ellipsis and fragments give very similar results to those seen in
(39) for finite clauses, for example (although (55a) is not quite as natural as the
other examples):
(57) a. ?(Goona arranged for Vurdy to arrive on time) and [ arrive on time ] he
managed to!
b. Goona wanted to arrive on time and Vurdy wanted to arrive on time too.
c. A: What did Goona want Vurdy to do?
B: Arrive on time.

Furthermore, infinitival to can be followed by a sequence of auxiliaries which

shows the
same order as in finite clauses. The only difference with finite clauses is that
modals cannot appear:
(58) We arranged for Bill to (*must) have been being interviewed, when you arrived.

Since modals are intrinsically finite in Modern English, they are in complemen-
tary distribution with to, which is intrinsically non-finite. Modals and to are
always followed by an uninflected verb or auxiliary; similarly, do is in examples
like (54a,b). The arguments related to negation and inversion given above do not
apply to to: the contracted negation n’t is restricted to finite clauses and inver-
sion can only apply in main clauses, while infinitives are always embedded. But
we can conclude from (57) and (58) that the mystery category in (56a) is T and
that to is the marker of non-finite T.
So we can bring S fully into line with X′-theory as TP. The next question con-
cerns S′. Can this apparently exocentric category be reduced to the X′-template?
Here we will see how this can be done, but we will leave one major question
open until the next chapter. Let us look again at the representation of a sentence
containing an embedded complement clause introduced by the complementizer
that (here S has been replaced by TP, Aux by T, and T′s have been added, along
with specification [Pres] in the main-clause T):

IanRoberts_9781316519493_C004.indd 93 05-09-2022 18:15:39

 94 4 X′- t h e o r y

(59) TP

NP T′

T VP
[pres]
V S′
N
hopes
Mary
Comp TP

that NP T′

N T VP
John will
V
arrive

In Chapter 3, we introduced the following rule expanding S′:

(60) S′ → Comp S

If we treat complementisers as heads, then we restate (60) as:

(61) CP → C(omp) TP

Here, CP stands for ‘Complementiser Phrase’, i.e. the maximal projection of

C(omp). So we render S′ endocentric by treating complementisers as heads and
by treating TP as the structural complement of C. This seems justified in that
different elements occupying C select for different kinds of TP: that requires a
finite TP, for requires a non-finite TP and whether allows either (John wondered
whether to leave/whether he should leave).
The step from (60) to (61) does not actually accord with the X′-template,
however. Strictly, to bring CP in line with the template we should have the two
rules in (62):
(62) i. CP → ZP C′
ii. C′ → C TP

This raises the question of what corresponds to ZP in (62i), i.e. what can func-
tion as SpecC′. I will leave this question open until Chapter 5.
Another question is whether we need CP in main clauses. Up to now we
have treated S′/CP as the category for subordinate clauses and S/TP as that
for main clauses. This is justified to the extent that the material following that
in a finite subordinate clause corresponds to a stand-alone main clause. But
this does not, in itself, preclude the possibility of a silent C introducing main
clauses. Furthermore the rules in (62) would clearly generate this structure.
Some languages, e.g. some varieties of Welsh and Gascon, seem to have overt
Cs introducing declarative main clauses, and many languages have interroga-
tive particles similar to English if and whether introducing main-clause ques-
tions, supporting the idea that main clauses are CPs, i.e. introduced by the

IanRoberts_9781316519493_C004.indd 94 05-09-2022 18:15:39

 4.5 Adjuncts in X’-theory 95

rules in (62). In Chapter 7, when we look at the syntax of languages other than
English, we will see that German supports the presence of a main-clause CP.

4.5 Adjuncts in X’-theory

We saw in Chapter 3 that the pro-form test enables us to see whether
a phrase is an argument or an adjunct. Here are our examples from Chapter 3
(see (41) of Chapter 3):
(63) a. John put his car in the garage on Tuesday, and Peter did so too.
b. *John put his car in the garage on Tuesday, and Peter did so on the drive on
Wednesday.
c. *John put his car in the garage on Tuesday, and Peter did so his bike on
Wednesday.
d. John put his car in the garage on Tuesday, and Peter did so on Wednesday.

We saw that (63d) can be taken to indicate that adjuncts are outside the VP and
so are not substituted under do so replacement, but the fact that the substituted
VP in (63a) is interpreted as put his car in the garage on Tuesday indicates that
the adjunct PP is part of the VP. We concluded that adjuncts are optional constit-
uents of VP, while selected arguments are obligatory constituents of VP.
In terms of X′-theory, selected arguments are complements of X inside X′, so
the NP his car and the PP in the garage in (63) are in V′. This, combined with
the two-layer structure of XPs given by the X′-template, allows us to make a
structural distinction between arguments and adjuncts. We can treat adjuncts as
categories adjoined to V′ in a structure like (64):
(64) VP

spec V′

V′ PP

V NP PP P NP
put the car in the on Tuesday
garage

The sequence V – object NP – locative PP forms the V′. It consists of put, a V

head of V′ and VP, and the c-selected arguments of put; NP and (locative) PP.
The optional modifier, the adjunct temporal PP on Tuesday, is adjoined to V′ as
sister of one occurrence of V′ and daughter of another occurrence of V′. If we
now construe do so substitution as applying to V′, it can freely substitute either
the upper or the lower V′ in (64), giving the two possibilities in (63a) and (63d),
and giving rise to the interpretation of the elided larger V′ in (63a) as containing
the PP on Tuesday.
Similar considerations apply in the case of the examples in (21), repeated
here:

IanRoberts_9781316519493_C004.indd 95 05-09-2022 18:15:40

 96 4 X′- t h e o r y

(21) a. John read a book of poems and Mary read one too.
b. *John read a book of poems and Mary read one of stories.
c. John read a book with a red cover and Mary read one too.
d. John read a book with a red cover and Mary read the one with a blue cover.

Since selected arguments are complements of X inside X′, the PPs of poems
and of stories in (21a,b) are in N′, as complements of the noun book. Since one
substitutes for N′, it must substitute the complement to N, hence the ungram-
maticality of (21b). On the other hand, the two with-PPs in (21c,d) are optional
N′-adjuncts in a structure like (65):
(65) NP

D N′
the
N′ PP

N PP P NP
book of stories with the blue/red cover

In cases of N′-substitution the adjunct PPs are optionally substituted, while the
complement PPs are obligatorily substituted.
Putting this approach to adjuncts together with the rest of the X′-template, we
have the following set of category-neutral PS-rules:
(66) a. XP → ZP X′
b. X′ → X′ WP
c. X′ → X YP

These rules define the following relations:

(67) a. Complement = sister of X, daughter of X′
b. Adjunct = sister of X′, daughter of X′
c. Specifier = sister of X′, daughter of XP

We can generalise the notion of adjunction in (67b) to ‘sister of Xn’ and ‘daugh-
ter of Xn’, where Xn may be any bar-level. This allows adjunction to maximal
projections (XPs) and heads (Xs). We will largely leave these possibilities aside
here and return to the nature of adjunction in Volume II.
The universalist assumption is that, by default, all phrasal categories in all lan-
guages will have the internal structure generated by the rules in (64) and so can
be described in terms of (65), along, of course, with the central notions of head
(X) and maximal projection, XP. This represents a very strong claim about UG
and about the possible structures we expect to find in the languages of the world.

4.6 Conclusion
In this chapter we have seen how we can posit a single X′-template
which subsumes the PS-rules of English which we saw in Chapter 3 and explains

IanRoberts_9781316519493_C004.indd 96 05-09-2022 18:15:40

 4.6 Conclusion 97

why only rules of that kind and not ‘crazy’ rules like those in (3) are possible.
The template, complete with the possibility of X′-adjunction, is given in (68):
(68) XP

ZP X′

X′ WP

X YP

Here ZP is the Specifier (SpecX′), WP is an adjunct adjoined to X′ and YP is the

selected complement of X.
X′-theory also captures the endocentricity of phrases, always consisting of a
head, and possibly taking other phrases as its complement and Specifier; exo-
centric categories are predicted not to exist. Furthermore, we saw evidence from
constituency tests for the intermediate levels of phrase structure at X′ (i.e. N′,
P′, A′, etc.). In this connection, the apparently exocentric categories S and S′
were replaced with TP and CP, examples of how functional categories obey
the X′-template. A natural further assumption would be that the other ‘satellite’
functional categories we saw in Chapter 1 (Mood, Agr, Num and D) also con-
form to X′-theory. This point will be further developed in Volume II.
We have left one important question open, though: we have not said what can
fill SpecC′, i.e. what ZP in (61i) can be. This question will be answered when we
look at wh-movement, the topic of the next chapter.

Exercises
1. Give the tree diagrams for the following NPs, and comment on how
they instantiate, or fail to instantiate, the X′-schema:
a. him
b. the composer of Yesterday
c. two pints of beer
d. my cat
e. his professor of phonetics
f. John’s cat
2. Give the tree diagrams for the following APs and PPs, and comment
on how they instantiate, or fail to instantiate, the X′-schema:
a. in Paris
b. just over the fence
c. out from under the bed
d. just there
e. very worried about the result
f. really amused that Boris hates Michel

IanRoberts_9781316519493_C004.indd 97 05-09-2022 18:15:41

 98 4 X′- t h e o r y

3. Give the tree diagrams for the following clauses, and comment on
how they instantiate, or fail to instantiate, the X′-schema:
a. Andy visited.
b. Andy has insulted Lou.
c. We had guessed that Andy would insult Lou.
d. We had thought that Andy would admit that she had
insulted Lou.
e. I have very much enjoyed this lecture.
f. Lou asked if he should ignore Andy.
g. Lou asked whether to ignore Andy.
h. Lou could have ignored Andy.
i. Lou decided on the plane.

For Further Discussion

Discuss the ways in which X′-theory improves on the format of Phrase-­Structure
rules we saw in the previous chapter. Are there are any disadvantages of X′-the-
ory in relation to the earlier format?

Further Reading
Carnie, A. 2013. Syntax: A Generative Introduction. Oxford: Blackwell, Chapters 6–8.
Freidin, R. 2012. Syntax: Basic Concepts and Applications. Cambridge: Cambridge
University Press, Chapter 5.
Fromkin, V., R. Rodman & N. Hyamset. 2000. Linguistics: An Introduction to
Linguistic Theory. Oxford: Blackwell, Chapter 3.
Haegeman, L. 1995. Introduction to Government and Binding Theory. Oxford:
Blackwell, Chapter 3.
Haegeman, L. & J. Guéron. 1999. English Grammar: A Generative Perspective.
Oxford: Blackwell, Sections 1.2.4 and 1.2.5.
Larson, R. 2010. Grammar as Science. Cambridge, MA: MIT Press, Units 14–18.
Sportiche, D., H. Koopman & E. Stabler. 2014. An Introduction to Syntactic Analysis
and Theory. Oxford: Wiley, Chapters 4–6.

IanRoberts_9781316519493_C004.indd 98 05-09-2022 18:15:41

 5 Movement

5.1 Introduction and Recap

In the previous chapter we saw how X′-theory can give us a sim-
plified and more abstract format for PS-rules and, correspondingly, for the struc-
tures they generate. The basic template for categories that X′-theory specifies is
given in (1):
(1) XP

specifier X′

X′ adjunct

X complement

Here, as we saw, the complement is the sister of X and daughter of X′, the
adjunct is the sister of X′ and the daughter of X′ (we mentioned that XP- and
head-adjunction may also be possible) and the Specifier is sister of X′ and
daughter of XP. Specifier, adjunct and complement are all relational notions,
defined in terms of two positions in the tree. The truly primitive notions are the
structural ones X, X′ and XP. These relations are in turn defined in terms of the
categorical value of X (remember that X is a variable ranging over categories)
and the notion of projection from X to X′ to XP.
More generally, what we have seen up to now is that PS-rules generate well-
formed structural descriptions (whether formulated in terms of X′-theory or not,
although in the last chapter we saw good reasons to prefer X′-theory over the
more ‘traditional’ kind of PS-rules). The structural descriptions are represented
as tree diagrams or labelled bracketings, those generated by X′-theory being
more abstract, in particular in that they generate a category-neutral template
which is in principle the same for all lexical and functional categories. In either
format, the PS-rules can create recursive, and therefore potentially infinite,
structures.
All of this gives us a good basis for the theory of constituent structure, but
there are two questions which we need to address, one quite specific and one
much more general. The specific question is one we have held over from our
discussion in the previous chapter of how X′-theory applies to functional cate-
gories: what goes in the Specifier of CP?

99

IanRoberts_9781316519493_C005.indd 99 05-09-2022 19:51:38

 100 5 movement

The general question relates to the kinds of rules PS-rules are, whether
restricted to X′-theory or not. An obvious way to think about PS-rules is as
‘building rules’: as we apply them, we build up the constituent structure. As we
saw in Chapter 3, we can relate each PS-rule to a specific ‘treelet’, a little piece
of a tree diagram; look again at (13) of Chapter 3 and the preceding discussion.
The same can be said of X′-theory. We can relate the X′ PS-rules given in (63)
of Chapter 4 and repeated here as (2) to the tree in (3):
(2) a. XP → ZP X′
b. X′ → X′ WP
c. X′ → X YP

(3) XP
rule (2a)
ZP X′
rule (2b)
X′ WP
rule (2c)
X YP

PS-rules generate structures: to put it in less technical terms they ‘build

trees’.
The very general question we can ask at this point, having seen how PS-rules
and X′-theory work, is whether we need any other kind of rule in our theory. Here
we will propose that there is a further kind of syntactic rule of a somewhat differ-
ent nature from PS-rules. These are known as transformations. Transformations
map phrase markers into other phrase markers. In other words, they convert one
tree into another tree. There are several kinds of transformations, including dele-
tion rules, movement rules and others. The goal of this chapter is to introduce
movement rules, showing how they interact with and complement PS-rules. We
will concentrate on one particular kind of movement rule: wh-movement.

5.2 Types of Movement Rules

Movement rules differ from PS-rules in that they do not create struc-
tures; instead they manipulate them in various ways. The principal way in which
movement rules manipulate trees involves transforming a tree in which a given
constituent occupies one position into one in which it occupies another, where
the positions can be seen and defined in relation to a tree diagram or labelled
bracketing. We see this in the different positions of the auxiliary in (4):
(4) a. Cambridge will flood as a consequence of global warming.
b. Will Cambridge flood as a consequence of global warming?

The auxiliary will is in one position in the declarative (4a) and in another in the
interrogative (4b). Informally, we say that will ‘moves’ from one place in the
structure to another.

IanRoberts_9781316519493_C005.indd 100 05-09-2022 19:51:39

 5.2 Types of Movement Rules 101

We have already seen several examples of movement rules in our earlier

discussions, only earlier we did not draw attention to the technical nature of
these processes. There are in fact three main kinds of movement rules: head-­
movement, NP-movement and wh-movement. Here we briefly describe all
three before moving on to a more detailed discussion of wh-movement.
Head-movement is seen in English subject-auxiliary inversion in questions.
In Chapter 2, we observed how subject-auxiliary inversion forms direct yes/
no questions from declaratives in English in examples like (4). We observed
that the declarative sentence (4a) has the order Subject – Auxiliary – Verb …
while the interrogative (3b) shows the ‘inverted’ order of subject and auxiliary:
Auxiliary – Subject – Verb ... The rule that relates the two sentences takes the
T-position occupied by will in (4a) and moves it to the C-position. Since the sub-
ject, Cambridge, occupies the SpecT′ position and linearly intervenes between
C and T, this has the effect of inverting the order of will (in T, moved to C) and
Cambridge. The rule relates the tree in (5a) to that in (5b) (for simplicity I have
left out the adjunct as a consequence of global warming, which is probably a PP
adjoined to T′; I have also left out V′ as there is very little in VP aside from the
main verb flood):
(5) a. C′

C TP

NP T′
Cambridge
T VP
will
T
flood

b. C′

C TP

T NP T′
will Cambridge
T VP
(will )
T
flood

We see from (5) that subject-auxiliary inversion in English is a case of move-

ment of T to C. Thus the structure in (5a) is mapped into that in (5b). No other
changes are made to the trees here, as can easily be observed. What is left behind
in T is a copy, which we will also refer to as a trace, of the auxiliary in its
original position; this copy is not pronounced, as the strikeout is intended to
indicate. Movement, then, can be usefully thought of as a kind of copy/paste
operation familiar from word-processing applications: a given element is copied

IanRoberts_9781316519493_C005.indd 101 05-09-2022 19:51:39

 102 5 movement

to a new position and pasted in there. From the fact that the original copy of will
in the T position is not heard, we can add that there is a general copy-­deletion
rule applying to moved elements in their original position (we will look at this
operation in more detail in Section 8.6). Since T is a head, this is a case of
head-movement. Head-movement does what it says: it takes a head, copies it to
a new position and deletes the content of the original head (but not the actual
position). The basic tree is unchanged by the movement operation, aside from
the substitution of T into a position dominated by C. So English subject-auxil-
iary inversion is a case of T-to-C movement, itself a case of the general opera-
tion of head-movement. There is much more to say about head-movement, and
we will come back to it in Chapter 7 and at greater length in Volume II. But for
now we have the basics, and an illustration of how movement operates.
The second kind of movement, NP-movement, is found in passives in English.
We saw an example of a passive in the discussion of ellipsis and English auxil-
iaries in Chapter 4:
(6) (When you arrived), John must have been being interviewed.

If we take away the complex auxiliary sequence, we can construct the following
active-passive pair:
(7) a. Oprah interviewed John. (active)
b. John was interviewed (by Oprah).   (passive)

These two sentences are very close to synonymous; in particular, the who’s-
doing-what-to-whom relations (known as argument structure) are the same:
both sentences describe a situation in which Oprah is the interviewer and John
is the interviewee. Furthermore, interview is a transitive verb; in terms of the
c-selection frames given in (14) in Chapter 4 it acts like watch, so it has a lexical
entry like (8):
(8) interview (V) ___ NP

This c-selection requirement is clearly satisfied in (7a), where John is the NP.
We also saw in Chapter 4 that this c-selection frame is associated with the
PS-rule which generates NP as the complement of V. In turn, this position cor-
responds to the direct-object function for NP, so in (7a) John is the direct object
of interview. The position of complement of V and corresponding direct-object
function is typically associated, in English and in many other languages, with
the semantic, or thematic, role of Patient or Undergoer of the action described
by the verb. In other words, John is the interviewee, as we already observed.
In (7a), Oprah is the subject; NPs in this position are typically associated, in
English and in many other languages, with the thematic role of Agent: Oprah is
the interviewer.
Now, if we compare the passive sentence in (7b) with the active one in (7a),
we can make two observations. First, John has the subject function in (7b): we
can see this from the word order (it precedes both the auxiliary and the main

IanRoberts_9781316519493_C005.indd 102 05-09-2022 19:51:39

 5.2 Types of Movement Rules 103

verb) and from subject-verb agreement, in that if we replace John with a plural
NP (the boys) then the auxiliary changes to its plural-agreement form: The boys
were interviewed (by Oprah). There are other ways to see that John is the subject
of (7b), but these will do.
We have already made the second observation: the who’s-doing-what-to-
whom/argument structure relations are the same as in (7a).
So active–passive pairs change grammatical function: the NP which is the
direct object of the active is the subject of the passive (while the NP which is
the subject of the active goes into the by-phrase of the passive and can even
disappear altogether: (7b) is grammatical if we leave out by Oprah). In our brief
discussion of subjects and predicates in Chapter 3, we observed that phrase
structure does not represent grammatical functions: it merely deals with cate-
gories, hierarchies and linear order. However, we can define the functions of
subject and direct object in X′-theoretical terms as follows:

(9) a. The subject is the NP in the Specifier of TP.

b. The direct object is the NP in the complement of V.

Strictly speaking, this reduces traditional grammatical relations to X′-theoretic

relations. In turn, as we saw at the beginning of this chapter, these can be reduced
to structural positions.
It then follows that saying that John is the object in the active (7a) and the
subject in the passive (7b) implies that the same NP, with the same thematic role
of interviewee, is in the complement of V in (7a) and in the Specifier of TP in
(7b). We can account for this by saying that this NP moves from one position to
the other; this is one aspect of the relation between active and passive sentences.
The relationship can be captured by the trees in (10):
(10) a. TP

NP T′

T VP

V NP
interviewed John

b. TP

NP T′
John
T VP
was
V NP
interviewed (John )

(The trees in (10) leave aside the active subject NP Oprah; accounting for the
relationship between the subject in the active sentence and the by-phrase in the

IanRoberts_9781316519493_C005.indd 103 05-09-2022 19:51:40

 104 5 movement

passive sentence is a tricky business which would take us too far afield here.)
Looking at the representation of the direct-object NP John in (10), we can see
that it is moved, by the same copy/paste + deletion operation as that applied
to T in (5), from the VP-internal direct-object position to the Spec,T′ subject
position. As with the head-movement operation in (5), a silent copy, or trace,
is left behind in the original position (indicated by strikeout). Because what is
moved here is an NP we call this NP-movement. In (10), again as in (5), the
tree is unchanged; just the position of one NP is changed. The object NP here is
substituted into the subject NP position in SpecT′.
As with head-movement, there is much more to say about NP-movement, and
we will come back to it at great length in Volume II. But, again, for now we have
the basics, along with a further illustration of how movement operates.
For the rest of this chapter, though, we will focus on the third type of move-
ment, wh-movement.

5.3 Wh-Movement: The Basics

Wh-movement appears in wh-questions, both direct and indirect,
and in a range of other constructions, including relative clauses (which we men-
tioned in our discussion of the fish sentences in Chapter 1). Wh-questions, some-
times called partial or constituent questions, ask about a given constituent of the
clause. (11) is a simple example of a direct wh-question:
(11) What did you catch (what)?

Here the wh-phrase what, an NP, moves from the direct-object position (indi-
cated here by the copy in parentheses) to the initial position of the sentence to
form a direct question. The question is interpreted as ‘asking about’ the nature of
the direct object, so a natural answer could be ‘fish’, for example. Wh-questions
contrast with yes/no-questions of the kind seen in (4b), as the latter simply ask
whether a given situation (e.g. Cambridge flooding as a consequence of global
warming) holds or not.
Wh-questions are so-called because, in English, the words and phrases that
introduce them, i.e. which stand in initial position of the interrogative clause,
nearly all begin with the letters ‘wh’: when, who, which, what, where, why, how,
etc. Wh-movement is the most important and interesting type of movement for
various reasons, some of which we will see directly.
In order to get a reasonably full picture of wh-movement, there are several
questions to be addressed. These include the following (note in passing that
these are all themselves wh-questions):
(12) a. Which XPs can move?
b. Where do they move from?
c. Where do they move to?

IanRoberts_9781316519493_C005.indd 104 05-09-2022 19:51:40

 5.3 Wh-Movement: The Basics 105

d. Why do they move?

e. How far do they move?

Looking first at question (12a), we can see from the following examples
that quite a range of categories can move (although, as (12a) implies, these
are all phrasal categories, XPs, so wh-movement is clearly distinct from
head-movement):
(13) a. Which man will David see t ?
b. Who should John talk to t ?
c. To whom can John talk t ?
d. How disappointed is Fred t ?
e. What does Don believe t ?

In all of these examples and from now on, I abbreviate the copy in the starting
position as ‘t’ for trace; this is purely for convenience; strictly speaking, there
is a whole phrasal copy in the relevant positions. In (13a), as in (11), we see
wh-movement of the object NP. The difference with (11) is that here we have the
wh-determiner which followed by the Noun man, while in (11) we have the pro-
nominal form what which substitutes the whole NP. In (13b), the indirect object
NP is moved; here we have the animate wh-pronoun who. The NP who is moved
from inside the indirect-object PP, headed by to. Since to is unaffected by the
movement, this kind of example is known as Preposition stranding. Preposition
stranding is very natural in English, as (13b) shows. It is, however, quite rare in
the languages of the world; the only other languages which clearly allow it are the
Scandinavian languages. In (13c), we see movement of the whole indirect object
PP. This is known as pied-piping (the terminology alludes to the Pied Piper of
Hamelin, who having led the town’s rats to a watery death and not been paid, led
the town’s children to a similar fate in revenge). Here the Preposition, which is
not itself questioned in (13c), is ‘pied-piped’ along with the questioned NP. In
languages which do not allow Preposition stranding (which, as just observed, is
most languages), pied-piping of the Preposition is the only way to question the
indirect object. In English, on the other hand, pied-piping of this kind, although
grammatical, sounds rather ‘stilted’, at least in the spoken language. The sense of
artificiality is added to by the use of the archaic accusative form of the wh-word:
whom. This form is all but moribund in most varieties of colloquial English and
is associated with formal, literary registers.1

1
However, it is arguably more natural in (13c) than the corresponding example with who:
(i) ??To who can John talk t?

Compare also the Preposition-stranding example (13b) with whom:
(ii) ?Whom did you talk to t?

Strictly speaking, whom should also appear in direct-object function in (13a):
(iii) ?Whom will David see t?

IanRoberts_9781316519493_C005.indd 105 05-09-2022 19:51:40

 106 5 movement

In (13d) we see wh-movement of a predicate AP, questioned by how. In (13e)

we have, potentially, questioning of a CP by what; the answer here could be a
CP such as That the world is flat. So we see, in answer to question (12a), that
wh-movement, with the right wh-word, can apply to NPs, PPs, APs and CPs,
whereas TP and VP do not undergo wh-movement.
Let us turn now to question (12b), where do wh-phrases move from? In exam-
ples like (11) and (13) the starting position of wh-movement seems reasonably
clear from the interpretation of the question, but can we demonstrate this syn-
tactically? A related question is whether wh-movement is always obligatory: do
wh-phrases always move? There are two constructions in English which give
us answers to both questions, as they show the wh-phrases in their unmoved
(in-situ) position. The first of these are so-called echo questions. These are
questions which – either because an interlocutor simply fails to properly hear a
previous sentence, or because they are astonished by what has been said – show
the wh-phrase in its original position, as in:
(14) a. Bill bought WHAT?!
b. You talked to WHO?!

Here the capitals and other punctuation are intended to indicate that the
wh-phrase is heavily accented. Most important for our present purposes, though,
these examples clearly show the wh-phrases in their unmoved positions.
The second construction is multiple questions. These are questions featuring
more than one wh-phrase, and hence questioning more than one constituent.
In English, only one wh-phrase can move per clause, and so any others must
remain in situ. Some examples are:
(15) a. Which prizes did we award t to which students this year?
b. Which girls did you say t danced with which boys at the party?

In these examples one wh-phrase moves to the beginning of the sentence, while
the other one stays in situ: we can see that in both of these examples the in-situ
wh-phrase is inside a PP. If we make the moved wh-phrase into a non-wh-phrase,
then the other one must move, and we have the usual choice of Preposition
stranding or pied-piping:
(16) a. Which students did we award the Schlumpfenberger Prize to t this year?
b. To which students did we award the Schlumpfenberger Prize t this year?
c. Which boys did you say Mary danced with t at the party?
d. With which boys did you say Mary danced t at the party?

Native-speaker judgements vary on these marginal examples. My own judgement is that (i)
sounds very strange, as it involves a clash of colloquial (preposition-stranding) and formal reg-
isters, while both (ii) and (iii) sound rather archaic and/or very formal. Native-speaker read-
ers should judge for themselves (don’t take the question marks in front of these examples too
seriously). Here again, we see that the cultural E-language concept ‘English’ masks a range of
I-language diversity.

IanRoberts_9781316519493_C005.indd 106 05-09-2022 19:51:40

 5.3 Wh-Movement: The Basics 107

Clearly, the simplest assumption about in-situ wh-phrases in both echo ques-
tions and multiple questions is that they simply do not move; in that case, these
examples confirm our semantically based intuitions about the original position
of wh-phrases in wh-movement examples in general.
These points are further confirmed by evidence that the position and func-
tion of the wh-phrase are occupied in wh-movement examples, even if the
wh-phrase has apparently moved away from that position. We can see this in
two ways. First, if we insert a wh-phrase in the beginning of a sentence with
no corresponding functional or positional ‘gap’ in the sentence, the result is
ungrammatical:
(17) a. *Who does Grandma like Fido?
b. *Which student did we award the prize to Bloggs?
c. *How many people did you say Mary had been to Chicago?

Clearly, the examples in (17) contrast minimally with those in (18):

(18) a. Who does Grandma like t ?
b. Which student did we award the prize to t ?
c. How many people did you say t had been to Chicago?

The contrasts between (17) and (18) can be straightforwardly accounted for if
we take seriously the copy/paste + deletion idea about movement. In each of
the examples in (18), the wh-phrase is moved from the position indicated by the
trace; hence it fills the position occupied by the trace and, derivatively therefore,
has the function associated with the trace’s position: direct object in (a), indirect
object in (b) and subject in (c). The examples in (17) are ungrammatical because
either the wh-phrase or the constituent in the respective trace positions has no
syntactic home; there is no position or function available for it in the structure.
They are ungrammatical for the same reason as (19):
(19) a. *Grandma likes Fido Bill.
b. *We awarded the prize to Bloggs Smith.
c. *You said Mary more people than me had been to Chicago.

A second way to make the same point is to take, for example, an intransitive
Verb of a kind that really cannot take a direct object however hard we try, e.g.
arrive:
(20) a. The train/Mary arrived.
b. *John arrived the train/Mary.

This example cannot allow what is sometimes called the causative reading of
otherwise intransitive verbs, in that it cannot mean ‘John made the train/Mary
arrive.’ Compare break in (21):
(21) a. The vase broke.
b. John broke the vase.

IanRoberts_9781316519493_C005.indd 107 05-09-2022 19:51:40

 108 5 movement

If we now add a wh-phrase to (20a), ungrammaticality results as we see in (22a).

Compare (22b), based on (21a):
(22) a. *What did John arrive t ?
b. What did John break t ?

Again, in (22a) the wh-phrase, or more precisely its copy/trace, has no syntactic
‘home’ in the structure, no position and therefore no function. Therefore the
result is ungrammatical (and semantically nonsensical).
So the answer to question (12b) is that wh-phrases move from independently
needed structural positions associated with particular grammatical functions:
direct object, indirect object, subject and predicate positions for example. Hence,
going back to our original example (11), we can give the pre-wh-­movement
structure in (23):
(23) C′

C TP

T NP T′
did you
T VP
(did )
V NP
catch what

Here we see what in the direct-object position, which we can motivate from
the semantic interpretation, echo questions, multiple questions and examples
equivalent to (19) and (20b). Since (11) is a direct question, we also have sub-
ject-auxiliary inversion (i.e. head-movement of T to C, as we saw in the previous
section) here, this time moving T containing the ‘dummy’ auxiliary do to C.
Now we can turn to question (12c): where do wh-phrases move to? In the light
of the structure in (11) and the general X′-template introduced in Chapter 4, the
answer to this question is now quite clear. Let us look just at the top part of the
structure in (23) and compare it with the X′-template in (1):
(24) C′

C TP

(25) XP

ZP X′

X′ WP

X YP

IanRoberts_9781316519493_C005.indd 108 05-09-2022 19:51:42

 5.3 Wh-Movement: The Basics 109

Mapping the general template in (25) onto the structure in (24), we can see
that C′ contains the head C (the target of head-movement of T, in fact) and the
complement TP. There is no adjunct WP adjoined to T′ here; recall that adjuncts
are always optional. What is missing in (24) is ZP Specifier of C′. Given the
linear order of the wh-phrases and the inverted auxiliary in all our examples of
wh-movement so far, the obvious conclusion is that the wh-phrase is in SpecC′.
So the structure for (11) after wh-movement has applied is (26):
(26) CP

NP C′
what
C TP

T NP T′
did you
T VP
(did )
V NP
catch (what )

Wh-movement copies the wh-phrase what, pastes it into SpecC′ (without

disturbing the X′ template), and the copy in the starting position is left
silent.
Turning next to question (12d): why do wh-phrases move? This is a much
more difficult question, and I won’t try to give a full answer here. However, we
can make an interesting and perhaps rather telling observation if we look again
at our basic clausal template of CP, TP and VP:
(27) CP

ZP C′

C TP

NP T′

T VP

V NP

Here we see the three phrasal categories CP, TP and VP. Loosely speaking, each
one seems to play a characteristic syntactic-semantic role in building up a full-
fledged clause.
The three clausal domains are, first, the VP, which specifies basic argument
relations, i.e. who does what to whom. In (27), the subject NP is outside VP; this
may be because subjects have a special status in relation to the VP because of the
importance of the subject-predicate relation, or it may be that subjects originate

IanRoberts_9781316519493_C005.indd 109 05-09-2022 19:51:43

 110 5 movement

inside VP and are moved to SpecT′ by a variant of the NP-movement operation

introduced in Section 5.2. For now we will leave this point open.
The second clausal domain is TP: this is the domain of the functional ele-
ments, foremost among them Tense, but also Agreement, Mood and other ‘satel-
lite’ functional categories associated with the clause which we saw in Chapter 1.
In this connection, it is worth briefly reconsidering complex auxiliary sequences
in English such as (6):
(6) (When you arrived), John must have been being interviewed.

In Chapter 4, we considered and rejected a structure like (28) for the auxiliary
sequence in (6):

(28) [VP must [V′ have [V′ been [V′ being interviewed ]]]]

We suggested instead that each V′ of (28) should be treated as ‘an AuxP of

some kind’. Since the first auxiliary is always the finite one (in a finite clause),
we can treat this one as Tense itself. The others encode aspectual (perfect have,
progressive be) and voice (passive be) information. So, we could tentatively
replace (28) with (29):
(29) [TP must [PerfP have [ProgP been [VoiceP being [VP interviewed ]]]]]

Each auxiliary occupies the head of its functional category and takes the next
one to its right as its complement, with Voice taking VP as its complement. The
subject occupies SpecT′; for now we leave open the question of what, if any-
thing, occupies the other Specifier positions (remember that Specifier positions
don’t have to be filled, as in simple NPs like John). Heads expressing Mood,
Tense, Aspect and Voice are often realised by inflections in other languages
(and indeed there are inflections in the English auxiliary sequence, e.g. the pro-
gressive -ing). So what we might loosely call the ‘extended TP′ is the domain
where notions such as Tense, Aspect and Mood are realised. Depending on the
language, these may be realised by auxiliaries, inflections, both together, or not
at all. English relies mainly, but not exclusively, on auxiliaries.
The third domain is the one relevant for wh-movement. We could take CP,
as the highest category in the clause, to be the one that relates the otherwise
complete clause (TP complete with subject and satellites, the VP saying what
the subject is doing to whom) to the ‘outside’. What’s outside the clause will
vary depending on whether the clause is a main clause or not. If the clause is a
main clause, there is no further, higher, syntactic structure. The clause is a root
clause (the CP node is the root of the tree). But root clauses have a function in a
wider discourse context; in particular, it is relevant for the discourse whether the
clause is a declarative, stating something, an interrogative, asking something,
an imperative, ordering someone to do something, etc. In other words, the root
clause must express discourse-relevant information regarding clause type. There
is much more to say about this matter, but for now it suffices to observe that
declaratives and interrogatives are two of the most basic clause types.

IanRoberts_9781316519493_C005.indd 110 05-09-2022 19:51:43

 5.3 Wh-Movement: The Basics 111

We have already observed that root interrogative Cs, expressing direct

questions, have special syntactic properties in English. In yes/no questions,
we observe subject-auxiliary inversion, i.e. T-to-C movement, the example of
head-movement introduced in Section 5.2. There we saw that a yes/no question
like Will Cambridge flood? has the structure in (5b), repeated here (we should
also add the CP node to (5b) for completeness, but for now we’ll continue to
assume that there’s nothing in SpecC′ in yes/no questions):
5b. C′

C TP

T NP T′
will Cambridge
T VP
(will )
T
flood

With this word order, i.e. with subject-auxiliary inversion, the sentence is inter-
preted as a yes/no question. So we could add a special clause-type feature to C,
to indicate this. Let us call the feature Q (for Question) and write CQ to designate
an interrogative C, a particular subcategory of C. So now we can say that CQ
attracts T, i.e. it has the syntactic effect of triggering subject-auxiliary inversion
and the discourse-semantic effect of saying ‘I’m asking you a yes/no question’
(however, exactly that is to be analysed by the semantic component of the gram-
mar, but we leave this aside since it is not a matter for syntax).
Wh-questions are obviously a type of question, so we can further type C with the
feature Wh; CQ, Wh is another subcategory of C. We can say this type of C has the
syntactic effect of triggering wh-movement of some XP into SpecC′. Since in main
clauses wh-questions also involve subject-auxiliary inversion, we could say that Q
is also present on C. So the full structure of a simple wh-question like (11) is (30):
(30) CQ, WhP

NP CQ, Wh′
what
CQ, Wh TP

T NP T′
did you
T VP
(did )
V NP
catch (what )

Following the X′-theoretic notion of projection, we can take it that the Q and
wh-features of this kind of C are marked on C′ and CP. Hence, in the discourse,
CP is a wh-question, with its particular discourse semantics (which, again, we
leave aside here as extra-syntactic). The syntactic property of CQ, Wh in a main

IanRoberts_9781316519493_C005.indd 111 05-09-2022 19:51:44

 112 5 movement

clause is that it triggers both subject-auxiliary inversion and wh-movement. In

this way, we can answer question (12d) above: the clause-typing wh-feature of
C triggers wh-movement to SpecC′.
Finally, we come to question (12e): how far can wh-phrases move? This is a
very important and difficult question that will, in various ways, occupy a great
deal of Volume II. I will postpone addressing this question for now, although
we’ll make a start on it in the next two sections, after looking at some further
aspects of wh-movement.

5.4 Subject Questions, Indirect Questions and Long-

Distance Wh-Movement
We saw in the previous section that wh-movement can affect NPs in
various functions, including direct and indirect objects. The relevant examples
from (13) are repeated here:
(13) a. Which man will David see t ?
b. Who should John talk to t ?
c. To whom can John talk t ?

It is possible to apply wh-movement to subjects. However, complications arise

in this case. In (31b), we see an example of a wh-question formed on the subject:
(31) a. Mary saw John.
b. Who saw John?

Here it looks as though the wh-phrase is in situ, since the example appears to
correspond point-for-point with the declarative in (31a). However, it is clear that
this is not an echo question; that would presumably be (32):
(32) WHO saw John?!

Furthermore, (31b) clearly has the discourse semantics of a wh-question; it just

seems to be syntactically different. We could therefore assign a structure like
(33) to (31b) (I have left the Q-feature off C for reasons that will become appar-
ent immediately below):
(33) CWhP

NP CWh′
who
CWh TP

NP T′
(who )
T VP

V NP
saw John

IanRoberts_9781316519493_C005.indd 112 05-09-2022 19:51:45

 5.4 Subject Questions, Indirect Questions and Other Matters 113

Here, the subject wh-phrase has moved from SpecT′ to SpecC′ in the standard
copy/paste + deletion manner, and this movement is presumably triggered by
the wh-feature on T, as in the examples we saw in the previous section. This
example of wh-movement does not change the linear order of the words in the
sentence, though; it is string-vacuous movement. That is why (31a) and (31b)
appear to line up (while, if (33) is the correct analysis of (31b), they actually
don’t). There is nothing in the notion of movement rule that states that move-
ment operations, copy/paste + deletion, have to change the linear order of ele-
ments; the fact is that they usually do, and all the examples we have seen so far
have involved such a change. But, as long as we don’t require this, then there is
no bar to positing the structure in (33) for (31b).
The problem, of course, is the absence of subject-auxiliary inversion, sig-
nalled in the absence of other auxiliaries by do-support, in (31b). In fact, if we
apply subject-auxiliary inversion to (31b), the result is strongly ungrammatical,
as long as do is left unstressed:
(34) *Who did see John?

If do is stressed, the sentence becomes grammatical, as (35a) shows. But stressed

do, strongly asserting the statement or question being made, is always possible
in finite clauses, as (35b) shows:
(35) a. Who DID see John?!
b. Mary DID see John!

Since, as (35b) shows, stressed do can appear in T, we can assume that the
stressed do in (35a) is also in T. In that case, what seems to give rise to the
ungrammaticality of (34) is subject-auxiliary inversion.
We can connect (34) to another strange feature of subject wh-movement.
As we saw with the multiple wh-question in (15b) in the previous section,
movement of wh-subjects from a subordinate clause to a main clause is
grammatical:
(36) a. Who did you say [ t likes John ] ?
b. Which student do you think [ t will win the prize ] ?

Here the brackets show the subordinate clause, ambiguously indicating either
CP or TP. In these examples the complementiser that is absent; this is consistent
with the general optionality of this complementiser introducing the clausal com-
plements to verbs like say and think in English. However, if we take the option
of inserting that in (36), the examples become ungrammatical:
(37) a. *Who did you say [CP that [TP t likes John ]] ?
b. *Which student do you think [CP that [TP t will win the prize ]] ?

We only see this contrast with subject wh-movement. If we wh-move the object
in examples like this, that is fully optional, just as in subordinate clauses where
no wh-movement takes place:

IanRoberts_9781316519493_C005.indd 113 05-09-2022 19:51:45

 114 5 movement

(38) a. Who did you say [CP (that) [TP John likes t ]] ?
b. Which prize do you think [CP (that) [TP Bloggs will win t ]] ?

We could generalise over (34) and (37) as follows:

(39) C-trace generalisation:

A filled C is not allowed when adjacent to the trace of a wh-moved subject.

The relevant parts of (34) and (37) which violate the constraint in (39) are
shown in (40):
(40) a. * … [C do ] [TP t … (34)
b. * … [C that ] [TP t … (37)

The generalisation in (39) is our first example of a condition limiting the

application of wh-movement. We will see many more, starting in the next
section.
So let’s say that some constraint like (39) applies to subject wh-movement
in English. This is admittedly a very odd constraint: it looks very parochial and
particular, not at all the kind of generalisation we’d like to have as part of the
theory of Universal Grammar. But we can see that there is a motivation for it:
in main clauses, we can maintain the analysis in (33) for string-vacuous subject
wh-movement and, in embedded clauses, we can understand why the otherwise
optional complementiser is banned in examples like (37). We can also observe
further cases, assuming, as we mentioned in Chapter 2 that if and for are com-
plementisers. The complementiser if shows up in indirect questions, as we will
see in more detail directly, and is not optional. Moreover, wh-movement out of
indirect questions is never very good. Nonetheless, we can see a contrast in the
following examples:
(41) a. *Who did you ask [CP if/whether [TP t likes John ]] ?
b. ?Who did you ask [CP if/whether [TP John likes t ]] ?

Similarly with for: many speakers of English allow optional for with certain
infinitive-taking verbs, e.g. (in my English at least) prefer:
(42) I would prefer (for) John to win the prize.

We see the same disappearing optionality with subject wh-movement, but not
with object wh-movement, as we saw with that in (36) and (37):
(43) a. Who would you prefer [ t to win the prize ] ?
b. *Who would you prefer [CP for [TP t to win the prize ]] ?
c. Which prize would you prefer [CP (for) [TP John to win t ]] ?

Strange though it is, (39) seems to account for a range of facts.

Optional complementisers like that and for can, under the right circum-
stances, be omitted (by definition). But we said in the previous section that
the Q-feature associated with C appears in both yes/no and wh-questions and

IanRoberts_9781316519493_C005.indd 114 05-09-2022 19:51:45

 5.4 Subject Questions, Indirect Questions and Other Matters 115

types the clause for discourse semantics. Its syntactic effect is to trigger sub-
ject-auxiliary inversion, T-to-C movement. But subject-auxiliary inversion
is impossible in (34), where we have subject wh-movement. What does this
mean for the Q-feature? Clearly, since (31b) is interpreted as a question, we
want to say that C still has a Q-feature, but (39) prevents it from trigger-
ing subject-auxiliary inversion where there is an adjacent subject trace, i.e.
where there is subject wh-movement. We will see directly that there are fur-
ther examples where C has a Q-feature but does not trigger subject-auxiliary
inversion.
This brings us to the topic of wh-movement in indirect questions. As in direct
questions, we see a range of possibilities for wh-movement:
(44) I wonder …
a. which man John will see t ?
b. who John should talk to t ?
c. to whom John can talk t ?
d. how angry Alex is t ?
e. what John believes t ?

In each of these examples, we have wh-movement to the SpecC′ of the selected

CP complement of wonder, a verb which selects for an indirect question (more
precisely CQ; see below). These examples exactly parallel those in (13): in (44a)
we have wh-movement of an object NP, in (44b) wh-movement of an indirect
object NP with Preposition stranding, in (44c) wh-movement of an indirect
object PP with pied-piping of the Preposition, in (44d) wh-movement of an AP
and in (44e) wh-movement, potentially, of a CP. The embedded questions here
are all exactly like their direct-question counterparts in (13), with the single
exception that there is no subject-auxiliary inversion.
Subject-auxiliary inversion is unacceptable in examples like (44) (for most
English speakers):
(45) *I wonder …
a. … which man will John see t ?
b. … who should John talk to t ?
c. … to whom can John talk t ?

In fact, the embedded C must be empty in all these examples: no complementiser

(that, if or whether) can appear:
(46) *I wonder …
a.   … which man that John will see t ?
b.   … who if John should talk to t ?
c.   … to whom whether John can talk t ?

We mentioned in the previous section that CP types the clause and we saw there
how this works for yes/no questions and wh-questions in main (root) clauses. In
subordinate clauses, CP bears clause-typing features that are c-selected by the main

IanRoberts_9781316519493_C005.indd 115 05-09-2022 19:51:45

 116 5 movement

verb of the superordinate clause. So, for example, the verb wonder selects CQ, since
its complement must be an interrogative of some kind (yes/no or wh-), while think
does not allow the head of its CP complement to be CQ:
(47) a. I wonder if/*that the world is flat.
b. I think *if/that the world is flat.

Unlike the CQ of main clauses, however, the CQ in embedded clauses cannot

trigger subject-auxiliary inversion (i.e. T-to-C movement), although it can be
realised by if or whether. Where C also has the wh-feature, it cannot be real-
ised by any complementiser, as (46) shows; this is parallel to what we observed
above with subject wh-movement in main clauses.
In indirect wh-subject questions, the situation is parallel to main clauses. As
in all other indirect questions, C must be empty. No subject-auxiliary inversion
or complementisers are possible. So all indirect wh-questions have a silent CQ, wh:
(48) a. I wonder who t saw John ?
b. *I wonder who did t see John ?
c. *I wonder who that/if t saw John ?

These restrictions on how CQ and CQ, Wh are realised in main and embedded
clauses are quite complicated and idiosyncratic. In other languages, things can
be different. For example, certain non-standard varieties of Dutch allow the
equivalent of that in indirect wh-questions. After a verb like ‘wonder’, then, we
find clauses like the following:

(49) …wie dat het Gedaan heeft.

…who that it done has
‘… who did it’     [non-standard Dutch]

Here wie (‘who’) occupies SpecC′ and dat (‘that’) is in C. So this example is
exactly equivalent to the ungrammatical English sentence in (46a). Even in ear-
lier stages of English, we can find something similar. The first line of Chaucer’s
General Prologue to the Canterbury Tales, written in the Middle English of the
late fourteenth century, begins with the wh-word whan (‘when’) and is immedi-
ately followed by that and then the subject:
(50) Whan that Aprill with his shoures soote …
When that April with its showers sweet …

Although not a wh-question (but an adjunct clause functioning as a time adver-

bial), we see the wh-phrase directly followed by that, which is not allowed in
Modern English. So the restrictions on the realisation of CQ and CQ, Wh are complex
and idiosyncratic in Modern English. The one thing which is constant, however, is
that where we have CQ, Wh, a wh-phrase always moves to SpecC′.
Let us now begin to address question (12e) from the previous section: how far can
a wh-phrase move? We have already seen some examples of wh-movement from
embedded clauses to the root SpecC′, primarily involving movement of the subject

IanRoberts_9781316519493_C005.indd 116 05-09-2022 19:51:45

 5.4 Subject Questions, Indirect Questions and Other Matters 117

of the embedded clause (see (35), for example). In fact, we can find examples of all
the other kinds of wh-movement from an embedded clause to a main clause:
(51) a. Which man did you say [ (that) John will see t ] ?
b. Who did you say [ (that) John will talk to t ] ?
c. To whom did you say [ (that) John will talk t ] ?
d. How angry did you say [ (that) Alex is t ] ?
e. What did you say [ (that) John believes t ] ?

In (51a), object NP wh-moves from the subordinate clause; in (51b), indirect

object NP moves from the subordinate clause, stranding the Preposition; in
(51c), the indirect object PP moved from subordinate clause, pied-piping the
Preposition; in (51d), the predicate AP moves from the subordinate clause and in
(51e) the wh-CP moves from the subordinate clause. In all of these cases, where
the verb of the main clause is say, the complementiser that is optional.
We saw in Chapter 3 that PS-rules can generate indefinitely long sequences
of subordinate clauses, each one embedded in the one superordinate to it. Hence
examples such as (20) from Chapter 3, repeated here, are possible:
(52) Mary hopes that John expects that Pete thinks that Dave said that …

Since wh-movement can move material from a subordinate clause to the root
SpecC′, we expect this to be possible from any subordinate clause, however
deeply embedded. At first sight, this certainly appears to be true, as the follow-
ing examples show:
(53) a. Who did you say [ (that) Mary believes [ (that) John saw t ]] ?
b. Who did you say [ (that) Mary believes [ (that) Fred knows [ (that) John saw t ]]] ?
c. Who did you say [ (that) Mary believes [ (that) Fred knows [ (that) I asserted
[ (that) John saw t ]]]] ?

There doesn’t appear to be any limit to how deeply embedded the subordinate
clause from which wh-movement takes place can be. So we see that wh-move-
ment can produce constructions with an unbounded dependency. The depend-
ency between the trace/copy of the wh-phrase in the starting position and the
SpecC′ which the wh-phrase moves to appears to be unbounded: wh-movement
can be as ‘long-distance’ as we like. The same appears to be true for wh-move-
ment in indirect questions. If we change the main-clause verb to wonder in (53),
then we have wh-movement to the first embedded SpecC′, the one c-selected by
wonder. Here, too, wh-movement appears to be unbounded:
(54) a. I wonder [ who Mary believes [ (that) John saw t ]] ?
b. I wonder [ who Mary believes [ (that) Fred knows [ (that) John saw t ]]] ?
c. I wonder [ who Mary believes [ (that) Fred knows [ (that) I asserted [ (that)
John saw t ]]] ?

The fact that wh-movement is unbounded does not imply it is unconstrained:

we have already seen a constraint on wh-movement of subjects, the C-trace

IanRoberts_9781316519493_C005.indd 117 05-09-2022 19:51:45

 118 5 movement

generalisation of (38). In the next section, we will see that there is a whole class
of important restrictions on wh-movement.
In this section we have seen some more complex aspects of wh-movement: the
special properties of subject wh-movement, including the strange C-trace gener-
alisation in (38); the behaviour of wh-movement and wh-Cs in embedded clauses
and, most importantly, the apparently unbounded nature of wh-movement. In the
next two sections we will focus on some consequences of this last observation.

5.5 The Nature of Movement Rules

The unbounded nature of wh-movement is important. It gives us a
reason for distinguishing PS-rules from movement rules. Wh-movement is an
operation which places a wh-phrase in a higher SpecC′ by copy/paste, leaving a
copy/trace in its original position. Seemingly, the target SpecC′ can be arbitrar-
ily structurally distant from the original position of the wh-phrase. On the other
hand, as we have seen, PS-rules (including those generating the X′-template) are
local: they define little bits of the tree at a time. For example, a rule like (55a)
generates the ‘treelet’ in (55b):
(55) a. V′ → V CP

b. V′

V CP

Could we generate wh-movement dependencies just using PS-rules? We could

certainly introduce the wh-feature into the rule in (55a), and distinguish CWhP
from a declarative CP. Specifying verbs such as wonder as c-selecting CWh
(which may be a special case of CQ, as our earlier discussion implied) will allow
wonder, but not think, for example, to be V in (55a). We can also specify root
C as optionally CWh. We can then add further PS-rules like (56):
(56) a.   CWhP → WhP C′Wh
b.  C′Wh → CWh TP

Rule (56a) will guarantee that exactly one wh-phrase appears in the Specifier
of CWh and rule (56b) introduces TP as the complement to CWh, seemingly in
the standard way (see (56i) of Chapter 4). But now we run into a problem. The
rules in (56) tell us what the properties of a CP which attracts a wh-phrase are,
but we need to also specify that the TP must contain a trace/copy of the WhP
in (56a). If we do not do this, we cannot rule out ungrammatical examples like
(17), repeated here:
(17) a. *Who does Grandma like Fido?
b. *Which student did we award the prize to Bloggs?
c. *How many people did you say Mary had been to Chicago?

IanRoberts_9781316519493_C005.indd 118 05-09-2022 19:51:46

 5.5 The Nature of Movement Rules 119

We could specify TP in (56b) as Wh, meaning that it dominates a trace/copy

of a wh-phrase. Then we could say that the examples in (17) do not have a
TwhP, so TP doesn’t dominate a trace/copy of a wh-phrase and ungrammaticality
results. We would have to extend this to every node dominating a trace/copy
of a wh-phrase in an example with long-distance movement. So in (57) every
XP-node (except the subject NP-nodes dominating you, Mary and John) would
have a wh-feature:
(57) [CP-wh Who did [TP-wh you [VP-wh say [CP-wh that [TP-wh Mary [VP-wh believes [CP-wh that
[TP-wh John [VP-wh saw t ]]] ?

But the wh-features everywhere except on the root C and CP are not the same
as the ‘true’ clause-typing wh-features we have seen until now. We can see
this particularly clearly in the embedded CPs. These are not indirect ques-
tions, but they have to be marked ‘wh’ to show that they contain a trace/copy.
So we need to find a way to distinguish the ‘trace/copy-marking’ wh-feature
from the clause-typing one. The clause-typing feature clearly has a lexical
motivation (some verbs c-select indirect questions and some don’t) and a dis-
course-semantic motivation, as we saw above. The ‘trace/copy-marking’ fea-
ture is there simply to say that there is a trace/copy somewhere ‘lower down’
in the structure. If traces/copies are not independently existing lexical or func-
tional categories, then indicating them in a PS-rule is no more than a way to
encode the movement dependency. Moreover, this approach goes against the
general locality of PS-rules, in that it states there is a trace/copy somewhere
lower in the structure, but this trace/copy could be arbitrarily deeply embed-
ded, given the unbounded nature of wh-movement. So marking movement
dependencies in PS-rules makes the PS-rules significantly more complex by
effectively introducing whole classes of new categories to mark traces/cop-
ies, and effectively makes PS-rules non-local. Furthermore, the complexity
increases greatly if we try to introduce a similar method for marking head- or
NP-movement dependencies.
So, rather than complicating the PS-rules very significantly as sketched above,
although this can be and has been done, we will stick to the idea that we there are
two relatively simple rule types. On the one hand there are PS-rules of the kind
introduced in Chapters 3 and 4 (including the X′-template), which build struc-
ture, i.e they generate phrase markers in a strictly local, incremental fashion. On
the other hand, there are movement, or transformational rules, which map phrase
markers into other phrase markers, preserving the X′-template. Movement con-
sists of the copy/paste + deletion operations, as we have seen. The former rules
are highly local, the latter, at least wh-movement, appear to be unbounded. The
former rules do not create traces/copies; the latter do.
To summarise what we have seen concerning movement so far, we have seen
three types of movement: head-movement, NP-movement and wh-movement.
Head-movement is illustrated by subject-auxiliary inversion, as in (5b), repeated
here with CP and the Q-feature added:

IanRoberts_9781316519493_C005.indd 119 05-09-2022 19:51:46

 120 5 movement

(5b) CQP

CQ′

CQ TP

T NP T′
will Cambridge
T VP
(will )
V
flood

By definition, head-movement moves a head. The target is also a head. The move-
ment operation copies the head T and pastes it into the ‘target head’. The copy left
behind remains in the structure but is not phonologically realised (from a phono-
logical perspective, it is deleted; from a structural perspective it is not, however).
The second type of movement is NP-movement. The example we saw was
movement of the object NP from the complement position of V to the subject
position, SpecT′, in passives. We saw the derived structure of a passive in (10b),
repeated here:

(10b) TP

NP T′
John
T VP
was
V NP
interviewed (John )

NP-movement moves NPs, as the name implies. Passives are a typical, but not
the only, case. Here the object is copied and pasted into the subject position,
and the copy in object position is phonologically deleted. NP-movement always
targets the subject position, SpecT′.
Third, there is wh-movement, which moves a wh-phrase to the Specifier of
CWh. Wh-phrases can be NPs, APs, PPs or CPs, but not VPs or TPs (or other
clausal functional categories such PerfP, ProgP or VoiceP; see (28)). A basic
case of object wh-movement is (30), repeated here:

(30) CQ, WhP

NP CQ, Wh′
what
CQ, Wh TP

T NP T′
did you
T VP
(did )
V NP
catch (what )

IanRoberts_9781316519493_C005.indd 120 05-09-2022 19:51:48

 5.5 The Nature of Movement Rules 121

Here the wh-phrase is copied and pasted into SpecCQ, wh′, and the copy is pho-
nologically silenced. Wh-movement always targets a SpecCwh′. We have been
taking Cwh to imply CQ, since wh-questions are a kind of question; we will return
to this point below.
All three cases of movement have certain things in common: they copy/
paste a constituent into a distinct position created by the X′-rules of phrase
structure. Technically, what we have informally been calling copy/paste is
known as substitution. So all the movement rules we have seen so far are
substitution transformations. (There are also adjunction transformations,
which create adjoined positions of the kind generated by PS-rules like (2b),
but we will leave these aside for now.) All of the movement rules leave a copy
in the original position and this copy is deleted phonologically; in other words,
it is present in the syntax (we have mostly abbreviated it as ‘t’ for trace) but
lacking phonological content, i.e. silent. We can also note that each movement
rule places the new copy (i.e. the ‘moved’ category) in a position which is, in
an obvious sense, higher up in the tree than the original position: C is higher
than the original position of T in (5b), the subject position SpecT′ is higher
than the object position in (10b) and SpecC′ is higher than the object position
in (30). So movement appears to always move constituents ‘upward’ in the
tree; this is an important generalisation which we will make more precise in
later chapters.
The three operations differ both in what is moved and in the target of the
movement. Head-movement moves heads, and in fact moves the head to another,
higher head position; again, this can be easily seen in (5b). NP-movement moves
NPs into subject position; in fact it moves NPs from one grammatical-function
position (object) to another (subject). This is why operations like passive are
often described in purely grammatical-function terms (‘passives turn objects into
subjects’). As usual, we take the grammatical-function terminology as a useful
shorthand for the correct, phrase-structural description. Finally, wh-movement
moves a wh-XP into SpecCWh′.
In the above description, we can note an interesting match between the start-
ing position and the target position of movement: head-movement moves heads
from one head position to another; NP-movement moves NPs from one gram-
matical-function position to another and wh-movement moves wh-XPs from
one wh-position to another (recall that in-situ wh-phrases in echo questions and
multiple questions show us that any NP position can host a wh-phrase).
Putting the last two observations together, we can give a general characterisa-
tion of movement rules as follows:
(58) Movement substitutes an element of type X into a higher position of type X,
where X ranges over heads, wh-phrases and grammatical-function NPs.

Although somewhat rough and ready, and certainly causing us to wonder why
heads, grammatical-function NPs and wh-phrases might fall together in this
way, (58) contains the germ of a very important set of generalisations, which we
investigate in full in Volume II.

IanRoberts_9781316519493_C005.indd 121 05-09-2022 19:51:48

 122 5 movement

Finally, we have observed that wh-movement is unbounded. The other types

of movement do not seem to be unbounded. Moving a head from an embedded
clause into a main clause gives a strongly ungrammatical result:

(59) *Will you can say [ that Cambridge (will) flood ] ?

Compare the fully grammatical (60), with ‘local’ T-to-C movement in the main
clause:
(60) Can you (can) say [ that Cambridge will flood ] ?

Similarly, NP-movement from a finite embedded clause into a main clause is

very bad, as the contrast between (61a–c) and (61d) shows:
(61) a. It was believed that Oprah interviewed John.
b. It was believed that John was interviewed (by Oprah).
c. John was believed to have been interviewed (by Oprah).
d. *John is believed [ that Oprah interviewed (John) ].

Sentence (61c) shows that movement from a non-finite complement clause is

acceptable.
Compare (62a), where the verb in the embedded clause is passivized and John
moves to the ‘local’ subject position and (62b), where, again, John is moved out
of a finite clause:
(62) a. We believe [ that John was interviewed (John) ].
b. *John was believed [ that (John) was interviewed ].

Wh-movement seems to differ markedly from the other types of movement in

this respect. This brings us back to question (12e), which we have still not fully
answered: how far can wh-movement go? In the next section, we will introduce
an important class of limitations on long-distance wh-movement.

5.6 Limiting Wh-Movement

We have seen that wh-movement can apply to most categories, but
not to VP or TP (or other clausal functional categories such as PerfP etc., see
(28)). We have also seen that wh-movement can move wh-phrases across multi-
ple clause boundaries, as in (53), repeated here:
(53) a. Who did you say [ (that) Mary believes [ (that) John saw t ]] ?
b. Who did you say [ (that) Mary believes [ (that) Fred knows [ (that) John saw
t ]]] ?
c. Who did you say [ (that) Mary believes [ (that) Fred knows [ (that) I asserted
[ (that) John saw t ]]] ?

So a natural question now arises: is movement of wh-XPs unrestricted in gen-

eral? In fact, we already know that the answer to this question is negative, in
that subject wh-phrases have to obey the C-trace generalisation in (38). But we

IanRoberts_9781316519493_C005.indd 122 05-09-2022 19:51:48

 5.6 Limiting Wh-Movement 123

could still ask about non-subject wh-phrases. Here we will see that the answer is
definitely negative, and that the limits to wh-movement are quite significant and
very interesting from a theoretical perspective.
Let us begin by looking at NPs which contain a possessor NP, such as my
guitar or John’s cat. The structure of these NPs is shown in (63) (here the sub-
script numerals 1 and 2 are there just to keep the two NPs distinct; they have no
theoretical significance):
(63) NP1

NP2 N′1

N′2 N1
cat
N2 guitar
John’s
my

Here we see that the possessor NP, NP2, occupies the Specifier of the possessee
NP, NP1. Everything is completely in conformity with X′-theory. Since the pos-
sessor is an NP, it can be a wh-phrase, giving whose cat, for example, with the
same structure as (63):
(64) NP1

NP2 N′1

N′2 N1
cat
N2
whose

NP2 looks like a wh-phrase in good standing. But if we try to apply wh-move-
ment to it we run into ungrammaticality:
(65) * [NP2 Whose ] did you feed [NP1 t cat ] ?

NP2 is unable to move alone. Instead, the whole of NP1 must be pied-piped:
(66) [NP1 Whose cat ] did you feed t ?

In (66), we have a straightforward case of object wh-movement, so of course

there is no difficulty here. But the ungrammaticality of (65) shows us that
wh-movement cannot apply directly to the possessor NP2.
The contrast between (65) and (66) reflects an important limitation on
wh-movement, known as the Left Branch Constraint (LBC):
(67) Wh-movement can’t apply to the left branch of an XP, or to part of a left branch.

Since, in (64), NP2 whose is on the left branch (i.e. the Specifier) of NP1, the LBC
prevents it from moving alone. This accounts for the ungrammaticality of (65).
Pied-piping NP1 as in (66) is unproblematic, since the entire object is not on a

IanRoberts_9781316519493_C005.indd 123 05-09-2022 19:51:49

 124 5 movement

left branch; being the complement of V, it is on the right branch inside V′. So
the LBC doesn’t affect object movement. (In order to allow for movement of
wh-subjects as in (31b) we have to exempt TP from (67).)
The LBC extends beyond possessors to APs. Consider an example like (68),
which we have seen several times:
(68) [AP How angry ] is Alex t ?

Here we have the wh-AP how angry, which we can analyse as containing the
Specifier how and the A′ (and A) angry. The LBC predicts that Specifiers, being
on a left branch cannot undergo wh-movement alone. This prediction is correct
for how angry, as (69) shows:
(69) *How is Alex [AP t angry ] ?

Here again, the whole AP must pied-pipe, as in (68).

The LBC is one of several constraints on wh-movement called island con-
straints. The terminology is based on the idea that islands are difficult to move
off, so the syntactic contexts which block wh-movement are islands for wh-move-
ment. Hence the left branch, or Specifier, of NP and AP are islands. There are
several other island constraints. Here I will briefly illustrate two; in Volume II we
will look at island constraints and how to account for them in detail.
One further island constraint concerns wh-islands. This constraint bans
wh-movement out of indirect questions, or any context introduced by a separate
wh-word. So, we start from an indirect question introduced by ask, e.g. (70a).
Then, we replace the direct object John with another wh-phrase, e.g. which mov-
ies. Then we wh-move which movies to the main-clause SpecC′. This gives rise
to the ungrammatical (70b):
(70) a. Mary asked Bill who t saw John?
b. *Which movies did Mary ask Bill who (who) saw (which movies)?

There is a clear contrast in grammaticality between (70a) and (70b); this is

because wh-movement of which movies in (70b) violates the wh-island con-
straint. Example (70b) can also be compared with (71), where there is compara-
ble long-distance wh-movement, but in this case there is no wh-island:
(71) Which movies did Mary say John saw (which movies)?

Another island constraint is the Complex NP Constraint (CNPC). This con-

straint bans wh-movement from complex NPs meaning, at least to a first approx-
imation, an NP containing a full CP. Hence wh-movement is banned in examples
like the following:
(72) a. *Which hobbit did you read [NP a story [CP that a wizard met t ]] ?
b. *Which hobbit did you read [NP a story [CP which was about t ]] ?

In (72a), there is a CP inside the NP in the complement of story (cf. I read a

story that a hobbit found a ring). In (72b), the CP inside the NP is a relative

IanRoberts_9781316519493_C005.indd 124 05-09-2022 19:51:49

 5.6 Limiting Wh-Movement 125

clause; there is a gap in the subject position corresponding to story here (this
indicates that relative clauses involve movement, a matter we haven’t gone into
here, with string-vacuous movement of which from SpecTP to SpecCP here; see
the discussion of subject wh-questions in Section 5.4). Compare (72b) with the
‘reduced-relative’ NP in (73), where there is no CP inside the NP:
(73) Which hobbit did you read [NP a story about t ] ?

In addition to the LBC, the CNPC and the wh-island constraint, there are several
other island constraints. But this is not the place for a thorough discussion, still
less an analysis, of the phenomenon. Here we can simply observe that, despite
appearing to be unbounded, the operation of wh-movement is not unlimited: it is
subject to island constraints. Subject wh-movement is also subject to the C-trace
generalisation in (38).
So we see that wh-movement combines two contrasting properties: it appears
to be unbounded and yet is subject to island constraints. Here we have con-
centrated on wh-movement in interrogatives, although it is also found in other
constructions, notably relative clauses, as in (74):
(74) a. the hobbit which you read a story about t
b. the hobbit [which Mary believes [ (that) Fred knows [ (that) I asserted [(that)
John saw t at the bottom of the garden ]]]]
c. *the hobbit which you read [NP a story [CP that a wizard met t ]]

Example (74a) illustrates a basic relative clause, with the wh-relative pronoun
which and an associated trace/copy as marked. We can take it that which is in
the SpecC′ of the CP which modifies hobbit. Example (74b) shows that rela-
tive-clause formation is unbounded, and (74c) that it is subject to the CNPC.
So which-movement, forming the relative clause here, is clearly a further case
of wh-movement. Presumably this case of wh-movement is not associated with
CQ, as it does not form an interrogative; we may want to define Crel for relative
clauses, or perhaps generalize the notion of Cwh beyond interrogatives. These are
among many issues we have not explored here.
English also allows relative clauses where there is no apparent wh-relative
pronoun. These can be marked with the complementizer that, which, as usual,
is optional. Where that is optional it appears that there is no overt marker of the
relative clause:
(75) a. the hobbit (that) you read a story about t
b. the hobbit [ (that) Mary believes [ (that) Fred knows [ (that) I asserted [(that)
John saw t at the bottom of the garden ]]]]
c. *the hobbit (that) you read [NP a story [CP that a wizard met t ]]

The evidence from the presence of a trace/copy, the unbounded dependency

in (75b) and the CNPC violation in (75c) shows us that there is wh-movement
here despite the absence of an overt wh-relative pronoun. One possibility is that
the wh-relative marker is deleted here. We cannot go into these questions fully
here, however.

IanRoberts_9781316519493_C005.indd 125 05-09-2022 19:51:49

 126 5 movement

5.7 Conclusion
In this chapter we have seen that movement rules are necessary,
alongside the PS-rules which generate the X′-template. We saw in Section 5.6
that although it is technically possible to complicate PS-rules in such a way that
they can generate the structures that result from movement, the complications
involved change the nature of PS-rules, in particular in that they lose their fun-
damentally local character.
We saw that there are three types of movement: head-movement, NP-movement
and wh-movement. These three types of movement have a number of properties
in common: they all involve substitution of a category into a position generated
by the PS-rules (hence they manipulate, rather than generate, structure), they all
copy the ‘moved’ element, leaving the copy in the starting position to be deleted
by the phonology, and they all involve a ‘like-for-like’ substitution, captured by
the rough generalisation in (58), repeated here:
(58) Movement substitutes an element of type X into a higher position of type X,
where X ranges over heads, wh-phrases and grammatical-function NPs.

Wh-movement seems to differ from the other types of movement in that it

involves an (apparent) unbounded dependency, although we saw that it is none-
theless subject to the LBC and other island constraints, as well as the C-trace
generalisation in (38).
The following structures illustrate each kind of movement that we have intro-
duced in this chapter:
(76) a. Head-movement:
CQP

CQ′

CQ TP

T NP T′
will Cambridge
T VP
(will )
V
flood
b. NP-movement:
TP

NP T′
John
T VP
was
V NP
interviewed (John )

IanRoberts_9781316519493_C005.indd 126 05-09-2022 19:51:50

 Exercises 127

c.  Wh-movement:
CQ, WhP

NP CQ, Wh′
what
CQ, Wh TP

T NP T′
did you
T VP
(did )
V NP
catch (what )

The general characterisation of movement given here combines with the

PS-rules generating the X'-template, given in (2) and repeated here, to give us
the mechanisms which generate and manipulate structure:
(2) a.   XP → ZP X′
b.  X′ → X′ WP
c.  X′ → X YP

We now have quite a good picture of the basics of the theory of syntax. However,
many questions have been left open here. One of the most important ones con-
cerns the ‘upward’ nature of movement, also captured by our generalisation in
(58). In the next chapter, we will look at a different set of rules which account for
certain kinds of meaning dependencies among elements; this will give us a way
of making the ‘upwards’ aspect of movement both more precise and integrating
it into the theory properly.

Exercises
1. Provide the trees for the following sentences. Indicate the gap (t) left
by wh-movement in each case:
a. Which car did Mary buy?
b. What did Mary buy?
c. With which knife did Mary cut the bread?
d. Which knife did Mary [V′ [V′ cut the bread ] [PP with t ]]?
e. Which books did Mary say [CP (that) she will buy t ]?
f. Who did Mary say [CP (that) [TP t sold her the books ]?
g. Which knife did Mary say that she cut the bread with?
h. How did Mary say that she will solve this problem?
i. I remembered [ who Ed said [ t had visited Wanda ]].
j. Ed asked [CP who [TP t knew [CP who [TP Greg had visited
t ]]]].
k. I will ask [CP which books [TP PRO to buy t ]].

IanRoberts_9781316519493_C005.indd 127 05-09-2022 19:51:51

 128 5 movement

2. Provide the trees for the following sentences, both the grammatical
and the ungrammatical ones. Indicate the gap (t) left by wh-move-
ment in each case:
a. I fed [NP [NP my [N′ neighbour’s ]] [N′ cat ]].
b. Which neighbour’s cat did you feed t ?
c. I wonder [CP [NP which neighbour’s cat ] [TP he fed t ]].
d. [CP [NP Which neighbour’s cat ] [C did ] [TP you say [CP
(that) [TP you fed t ]]]] ?
e. *[CP [NP Which neighbour’s ] did you feed [NP t cat ] ?
f. *Which did you say you fed [NP [NP t neighbour’s ] cat ] ?
What constraint do (e) and (f) violate?
3. The following examples feature another island constraint:
a. *Who did [ pictures of t ] annoy John?
b. *Who did they say that [TP [TP PRO to talk to t ] [T is ] [AP
forbidden ]]?
c. *Who was [TP [CP that John spoke to t ] twas a shock?
Can you formulate this constraint (hint: try to generalize the Left Branch
Constraint).

For Further Discussion

4. Here is a very interesting construction, usually known as
‘tough-movement’ or the ‘easy-to-please construction’:
a. It is [AP tough [CP for [ me/John/NASA to build space-
ships ]].
b. Spaceships are [AP tough [CP for [ NASA to build t ]].
c. It is [AP easy [ PRO to cook pasta ]].
d. Pasta is [AP easy [ PRO to cook t ]].
In (b) and (d), it seems natural to suggest that spaceships/pasta have moved to
the main-clause subject position. Is this a further case of WH-movement? How
could we determine whether it is or not?
other

Further Reading
Adger, D. 2003. Core Syntax. Oxford: Oxford University Press, Chapters 9 and 10.
Carnie, A. 2011. Modern Syntax: A Coursebook. Cambridge: Cambridge University
Press, Chapter 4.

IanRoberts_9781316519493_C005.indd 128 05-09-2022 19:51:51

 Further Reading 129

Carnie, A. 2013. Syntax: A Generative Introduction. Oxford: Blackwell, Chapters

9–11.
Larson, R. 2010. Grammar as Science. Cambridge, MA: MIT Press, Units 24–27.
Freidin, R. 2012. Syntax: Basic Concepts and Applications. Cambridge: Cambridge
University Press, Chapter 8.
Radford, A. 2016. Analysing English Sentences. Cambridge: Cambridge University
Press, Chapter 6.
Roberts, I. 1997. Comparative Syntax London: Arnold, Chapter 4.
Sportiche, D., H. Koopman & E. Stabler. (2014). An Introduction to Syntactic Analysis
and Theory. Oxford: Wiley, Chapters 10 and 14.
Tallerman, M. 1998. Understanding Syntax. London: Routledge, Chapter 8.

IanRoberts_9781316519493_C005.indd 129 05-09-2022 19:51:51

 6 Binding

6.1 Introduction and Recap

In the previous chapter we made the distinction between the rules which
generate structure, PS-rules, including those which generate the X′-tem-
plate, and movement rules, which map phrase markers into phrase markers.
In this section I will introduce a third type of rule, interpretative rules,
which specify the relationships between constituents in a hierarchical struc-
ture generated by PS-rules. These rules do not generate structure, so they are
clearly distinct from PS-rules. Furthermore, they do not appear to involve
movement. As the name suggests, though, they are relevant for the seman-
tic interpretation of relations between constituents, including, as we will
see, the relations between moved elements and their copies. The particular
kind of dependency we will be concerned with involves binding of various
types; the exact nature of binding will be elucidated as we proceed. We will
see that the account of binding relations we arrive at provides important
support for the hierarchical organisation of constituents of the kind we have
been developing up to now.

6.2 Pronouns and Anaphors

We introduced pronouns in the context of substitution operations as
tests for constituency in Chapter 3. There we observed that pronouns (e.g. I/
me, he/him, she/her, we/us, you, they/them) are noun-like elements that have
little semantic content of their own. Their semantic content is limited to Person,
Number, Gender and Case. Concerning Person, I/me and we/us are first per-
son, in that they refer to the speaker (or author) of a sentence and, in the plural,
some unspecified other(s); you is second person, referring to the addressee(s),
and the other pronouns are third person, in that they refer to neither speaker
nor addressee. Number refers to the singular vs plural opposition; note that
the Engish second-person pronoun you is ambiguous in Number. Gender, in
English, relates to ‘natural’ rather than ‘grammatical’ gender: animate males
are masculine, normally referred to with he/him and animate females are fem-
inine, referred to as she/her; all inanimate objects are referred to as it. We saw
in Chapter 1 that some languages, such as Italian, have grammatical gender,

130

IanRoberts_9781316519493_C006.indd 130 05-09-2022 17:32:06

 6.2 Pronouns and Anaphors 131

categorising all Nouns as masculine or feminine on what appear to be arbitrary

grounds in many cases. Case refers to the nominative vs accusative distinction
between I and me, he and him, etc. We saw in Chapter 2 that nominative is the
typical case for subjects and accusative for objects. There are also the genitive
forms of pronouns: my, your, his, her, its, our and their. The features of pro-
nouns are features which are found elsewhere in the grammatical system: in
verbal agreement and in the functional categories associated with nominals (e.g.
Number, as we mentioned in Chapter 1).
The featural make-up of pronouns makes them semantically rather ‘light’.
They generally do not pick out referents on their own, unlike typical NPs such
as John, the boy or the man who lives next door. Instead, they stand in for some-
thing else. This ‘something else’ is the antecedent, and this determines what the
pronoun refers to. So, in an example like (1), John can be naturally interpreted
as the antecedent of he:
(1) John hopes that he will win.

In that case, (1) is interpreted to mean ‘John hopes that he, John, will win.’
Alternatively, he in (1) could refer to any contextually salient male individual
distinct from John. In the latter case, we say that the pronoun has a non-linguistic
antecedent, while in the interpretation where he stands for John, the pronoun has
a linguistic antecedent.
As we will see, binding theory makes an important distinction between pro-
nouns and anaphors. This distinction is not made in traditional grammar, where
anaphors are described as a particular type of pronoun: reflexive pronouns such
as herself, himself, myself, yourself, yourselves, etc., and reciprocal pronouns
such as each other. On the other hand, binding theory groups reflexives and
reciprocals together as a class of anaphors distinct from normal, simple pro-
nouns such as I/me, he/him, she/her, we/us, you, they/them. The reasons for mak-
ing this distinction will become clear below.
Both pronouns and anaphors are distinct from non-pronominal NPs, notably
R(eferring)-expressions, NPs with the intrinsic semantics not dependent on the
interpretation of some other NP in the linguistic expression in which they occur,
e.g. John, Mary, the idiot, the pet shop owner, etc.
The goal of binding theory is to account for the distributions and interpretations
of each class of NP. A simple way to do this is to adopt the coindexing convention:
whenever two NPs have the same index, one is dependent on the other for its inter-
pretation. The indices are subscript letters of the alphabet starting from ‘i’: i, j, k …
This is linked to the idea that every distinct individual in our mental representation of
the world has an integer which identifies it: John23, the man next door, the pet shop
boy666, etc. However, this last point has little bearing on what follows. Adopting the
coindexing convention, we can repeat (1) as (2):
(2) Johni hopes that hei will win.

IanRoberts_9781316519493_C006.indd 131 05-09-2022 17:32:06

 132 6 binding

Unlike (1), (2) is unambiguous: the coindexing of John and he means that John
is the antecedent of he, i.e. (2) only has the first of the two interpretations of (1)
discussed above.
So binding theory should be able to explain why the following sentences, with
the interpretations indicated by the indices, are ungrammatical:
(3) a. *Herself thinks that it’s going to rain.
b. *A good friend of Maryi’s has declared herselfi bankrupt.
c. *Johni likes himi.
d. *Hei likes Johni.

Furthermore, it should explain why the following examples are grammatical,

again with the interpretations indicated by the indices:
(4) a. Johni said that the person who had upset himi should be fired.
b. Maryi has declared herselfi bankrupt.
c. Hisi mother thinks Johni is a genius.

Here we will see how binding theory deals with these examples in terms of the
three classes of NP. We will look first at anaphors, then at pronouns and finally
at R-expressions.

6.3 Anaphors
So let’s begin with anaphors, concentrating mainly on reflexives.
Consider first (5):
(5) Goona betrayed herself.

Here, for (5) to be grammatical, we must interpret Goona as the antecedent

of herself. We should therefore write (5) as (6), following the coindexing
convention:
(6) Goonai betrayed herselfi.

Compare now the ungrammatical (7):

(7) *Herself left.

Comparing the grammatical (6), where Goona must be the antecedent of herself,
with the ungrammatical (7), where herself has no antecedent, we can conclude
that herself needs an antecedent.
Now consider (8):
(8) *Herselfi betrayed Goonai.

Here, given the coindexing, herself has an antecedent, namely Goona. The sen-
tence is nonetheless clearly ungrammatical. How can we account for this?
The structure of both (6) and (8) is (9):

IanRoberts_9781316519493_C006.indd 132 05-09-2022 17:32:06

 6.3 Anaphors 133

(9) TP

NP T′

N′ T VP

N′ V′
Goonai
Herselfi V NP
betrayed herselfi
Goonai

We will now review several hypotheses concerning what might be the best anal-
ysis of the required structural relation between the antecedent of the reflexive
and the reflexive. Here is the first hypothesis:
(10) Hypothesis One: The antecedent of an anaphor must be a subject.

Hypothesis One will correctly distinguish the grammatical (6) from the ungram-
matical (8). In (6), the antecedent of the reflexive, Goona, is a subject and in (8)
it is not.
However, in English non-subjects can be the antecedents of reflexives, as in:
(11) I asked Johni about himselfi.

This example is grammatical, but John is not a subject, although it can still be
the antecedent for himself. So this disconfirms Hypothesis One. Compare also:
(12) *I asked himself about John.

Here I is the subject, but it still cannot be the antecedent for himself. This is
because the antecedence relation requires agreement in number, gender and per-
son. First-person I therefore cannot bind third-person himself (contrast Ii asked
myselfi about John where myself and I agree for first person and so I can be the
antecedent of myself). We will see further examples of the agreement condition
on antecedence below. Hence subjecthood, on its own, is neither necessary nor
sufficient for being a well-formed antecedent for a reflexive.
To a good approximation, the structure of (11) and (12) is as in (13):

(13) TP

NP T′

N′ T VP
(past)
N V NP PP
/ asked Johni
*/ asked himself P′

P NP
about himselfi
about John

IanRoberts_9781316519493_C006.indd 133 05-09-2022 17:32:07

 134 6 binding

These examples lead to the postulation of Hypothesis Two, as follows:

(14) Hypothesis Two: The antecedent of an anaphor must precede it.

Hypothesis Two accounts for the grammaticality of (6) and (11), where the
antecedent precedes the anaphor, and for the ungrammaticality of (8) and (12),
where anaphor precedes the antecedent. In fact, precedence may be a necessary
condition for a well-formed antecedent antecedent–anaphor relation, but it is
not a sufficient one, as there are examples where the antecedent precedes the
anaphor which are nonetheless ungrammatical, such as the following:
(15) a. *Johni’s friends talked about himselfi.
b. *Friends of Johni’s talked about himselfi

Here we must pay careful attention to the indices. The coindexation in (15a)
indicates that John is the intended antecedent of himself, and this is impossible.
We can assign a grammatical interpretation to the string in (15a) if we inter-
pret John’s friends as the antecedent of the plural anaphor themselves; here we
see the requirement that the antecedent and anaphor must agree for number in
action. But this would correspond to the coindexing in (16):
(16) [ John’s friends ]i talked about themselvesi.

Here the index i is associated with the whole NP John’s friends (incidentally,
this shows that indices are associated with NPs, not with Nouns). Similar rea-
soning applies to (15b): with the indexing given, John cannot be the antecedent
of himself. John’s friends can be, if himself is changed to themselves as just men-
tioned, but in that case the index is associated to that whole NP, and not to John.
Keeping these points in mind, we can get a very important clue to what is
going on if we look at the structure of (15a) (here, as in our discussion of pos-
sessive NPs in Section 5.3, the numerical subscripts on the NPs serve merely to
keep them distinct; they are not to be taken as indices):
(17) TP

NP1 T′

NP2 N1′ T VP

N2′ N1 V PP
friends talked
N2 P′
John’s
P NP3
about
N3′

N3
themselves

IanRoberts_9781316519493_C006.indd 134 05-09-2022 17:32:07

 6.3 Anaphors 135

Here, if NP1, John’s friends, is the antecedent of themselves, the relation is struc-
turally well-formed, as we pointed out above. On the other, NP2, John’s, can-
not be the antecedent. What seems to make the difference here is the relation
between the antecedent and the anaphor in terms of hierarchical structure. We
could therefore tentatively formulate Hypothesis Three, as follows:
(18) Hypothesis Three (to be revised): An antecedent must be structurally higher
than the anaphor which it binds.

Since NP1 is evidently structurally higher (higher up in the tree) in (17) than
the anaphor NP3, it is able to be its antecedent. NP2, John’s, on the other hand,
is seemingly not in the requisite structural relation with the anaphor for this to
be possible. Clearly, we need to make this intuitive notion of ‘structural height’
more precise.
The formal indicator of relative structural height is the relation of
­constituent-command, or c-command. This structural relation plays a central
role in binding theory and, as we will see in Volume II, many other dependen-
cies. The definition of c-command is given in (19):
(19) A node X c-commands another node Y if and only if:
i. X does not dominate Y;
ii. Y does not dominate X;
iii. The first branching node dominating X dominates Y.

Let us revisit some of the structures we looked at earlier and apply this definition
to the various antecedent–anaphor pairs. First, (9):
(9) TP

NP1 T′

N1′ T VP
[Past]
N1 V′
Goonai
V NP2
betrayed herselfi

Taking the subject NP to be X and the object NP to be Y in terms of the defini-

tion of c-command in (19), we can apply the definition. First, it is easy to see that
the subject NP doesn’t dominate the object NP and NP2 doesn’t dominate NP1
(remember the definition of dominance in Chapter 2, (26)). So the first two con-
ditions in (19) are satisfied. Coming to the third, the first branching node domi-
nating the subject NP is TP, and TP, as the root node here, dominates everything.
Therefore TP dominates the object NP and so the third clause of the definition
in (19) is met: the subject NP c-commands the object NP. If c-command is the
condition for antecedence, then the subject NP Goona can be the antecedent of
the object NP herself.

IanRoberts_9781316519493_C006.indd 135 05-09-2022 17:32:08

 136 6 binding

Now let’s ask if the object NP c-commands the subject NP. As before the first
two clauses of the definition in (19) are met, in that neither NP dominates the
other. Turning to (19iii), though, we see that the first branching node dominating
the object NP is V′, and this node does not dominate the subject NP. Therefore
the object NP does not c-command the subject NP in (9), hence in the ungram-
matical (8), Goona cannot be the antecedent of herself.
Next, the structure in (13) (annotated with numerical subscripts):
(13) TP

NP1 T′

N1′ T VP
[Past]
N1 V NP2 PP
/ asked Johni
*/ asked himself P′

P NP3
about himselfi
about John

Here the question is whether NP2 c-commands NP3. Since it is easy to see nei-
ther NP dominates the other, the question again is whether the first branching
node dominating NP2, VP, dominates NP3. The answer is clearly yes, and so NP2
c-commands NP3 and is able to be the antecedent for NP3. Conversely, since the
first branching node dominating NP3 is P′ and P′ does not dominate NP2, NP3
cannot be the antecedent for NP2; this accounts for the ungrammaticality of (12).
Finally, let us consider (17) in the light of c-command as defined in (19):
(17) TP

NP1 T′

NP2 N1′ T VP
[Past]
N2′ N1 V PP
friends talked
N2 P′
John’s
P NP3
about
N3′

N3
themselves

Here, since the first branching node dominating NP1 is TP, which as the root node
dominates everything, NP1 c-commands NP3. This is why John’s friends can be
the antecedent of themselves in (16). On the other hand, the first branching node
dominating NP2 is NP1, which does not dominate NP3. Therefore John’s cannot
be the antecedent of themselves.

IanRoberts_9781316519493_C006.indd 136 05-09-2022 17:32:09

 6.3 Anaphors 137

So we can restate Hypothesis Three as follows:

(20) Hypothesis Three (revised): An anaphor must be c-commanded by its
antecedent.

Formulated in this way, with c-command defined as in (19), Hypothesis Three

correctly captures all the grammatical and ungrammatical sentences we have
looked at so far.
However, c-command, although of central importance, is just one com-
ponent of binding theory. There is a further aspect, connected to how far
away the anaphor is from its antecedent. We can see this from the following
example:
(21) *Johni said [CP that himselfi left ].

The structure of (21) is (22):

(22) TP

NP1 T′

N1′ T VP
[Past]
N1 V CP
*John said
C TP
that
NP2 T′

himself T VP
[Past]
V
left

Here we can easily see that NP1 c-commands NP2, since, once again, the
first branching node dominating NP1 is TP, the root node, which dominates
everything. With only Hypothesis Three, then, John should be able to be the
antecedent for himself.
We can correctly rule out (22) and retain all our earlier results if we restate
Hypothesis Three as follows:
(23) Hypothesis Four: An anaphor must be c-commanded by a clause-mate
antecedent.

We can define ‘clause-mate’ as ‘dominated by the same CPs’. To put it more

formally (and precisely):
(24) X and Y are clause-mates if and only if there is no CP which dominates Y
and does not dominate X.

By this definition, we see that John and himself are not clause-mates in (22) since
the lower CP dominates himself without dominating John. In (22), then, himself is

IanRoberts_9781316519493_C006.indd 137 05-09-2022 17:32:09

 138 6 binding

not c-commanded by a clause-mate antecedent and so Hypothesis Four rules out

the antecedent-anaphor relation here. All the earlier examples we considered were
monoclausal and the clause-mate condition was satisfied every time.
But compare now the following examples:
(25) a. *Johni thought [CP that himselfi was a genius ].
b. Johni believes [XP himselfi to be a genius ].

We can account for the ungrammaticality of (25a) exactly as for (21): John and
himself are not clause-mates by the definition in (24) and therefore John cannot
be the antecedent of himself, given Hypothesis Four. But where does this leave
(25b)? At a minimum, we have to say that, although XP looks like a clause in
that it contains a subject and a predicate (at least from a semantic perspective), it
does not qualify as a clause for the definition of clause-mates in (24). This would
be the case if XP is not a CP. So instead, we take it to be TP, where infinitival
to is in T. This is supported by the fact that verbs like believe cannot take CWh
complements, as we noted in Chapter 5. So we have contrasts of the following
kind with verbs like wonder:
(26) a. John wondered [CP whether to leave ].
b. *John believed [ whether the world to be flat ].

If there is no CP in the infinitival complement to believe, then there is no possi-

bility of CWh.
So let us assign the structure in (27) to (25b):
(27) TP

NP T′

N′ T VP
[Past]
N V TP
John believes
NP T′
himself
T VP
to
V NP
be
D N′
a
N
genius

Since there are either no CPs dominating John and himself here, or there is
just a root CP which is not shown here, there is no CP which dominates him-
self but does not dominate John. Therefore, John and himself are clause-mates.
Therefore John can be the antecedent of himself. The contrast in (25) is now
accounted for.

IanRoberts_9781316519493_C006.indd 138 05-09-2022 17:32:10

 6.3 Anaphors 139

But now what can we make of the ungrammaticality of (28)?

(28) *Johni believes [TP Mary to like himselfi ].

The structure of (28) is (29):

(29) TP

NP T′

N′ T VP

N V′
John
V TP
believes
NP T′
Mary
T VP
to
V NP
like himself

Unless we posit a CP in some strange place inside the embedded TP, John and
himself will be clause-mates by the definition in (24). It also seems clear that
John c-commands himself, since John is the main-clause subject in main-clause
SpecT′, and so the first branching node dominating John is the root TP, which
dominates everything. It seems we need a further condition in addition to clause-
mates and c-command.
Hypothesis Five can account for (28) while leaving all the earlier results
intact:
(30) Hypothesis Five: An anaphor must be c-commanded by a clause-mate
antecedent such that no other NP intervenes between the antecedent and
anaphor.

We can also define intervention in terms of c-command (a more general version

of this notion of intervention, still defined in terms of c-command, will play a
major role in Volume II):
(31) Any NP which c-commands the anaphor and is c-commanded by the ante-
cedent is an intervener.

Applying Hypothesis Five to (28), we see that the subject of the infinitival TP, Mary,
c-commands the anaphor himself and is c-commanded by the antecedent John. Mary
therefore intervenes between the antecedent and the anaphor, and so the relation
cannot hold. In contrast, there is no such intervener in (25b/27) and so here the ante-
cedent–anaphor relation is well-formed, as it was under Hypothesis Four.
However, the notion of intervener in (31) is too strong, in that it rules out cer-
tain well-formed antecedent–anaphor relations, as in (32) for example:

IanRoberts_9781316519493_C006.indd 139 05-09-2022 17:32:10

 140 6 binding

(32) Johni told Mary about himselfi.

The structure of (32) is (33):

(33) TP

NP T′

N′ T VP
[Past]
N V NP PP
John told Mary
P NP
about himself

Here, Mary c-commands the anaphor and is c-commanded by the antecedent,

just as in (28). So the antecedent–anaphor relation should fail to hold here too.
But in this case it is allowed.
To allow for (33) while retaining our account of (28), we can reformulate
Hypothesis Five as follows:

(34) Hypothesis Five (revised): An anaphor must be c-commanded by a clause-

mate antecedent such that no subject intervenes between the antecedent and
anaphor.

The definition of intervention in (31) remains unchanged, but now Hypothesis

Five only applies to subjects intervening between the antecedent and the anaphor.
This means that the antecedent–anaphor relation is still blocked in (28), while
(33) is allowed, since although Mary is still defined as a intervener by (31), it is
an indirect object and not a subject.
We have now almost arrived at the final characterisation of the structural
domain for well-formed antecedent–anaphor relations. There are two require-
ments: (i) the anaphor and antecedent must be clause-mates and (ii) no subject
can intervene between the anaphor and antecedent. Furthermore, the anteced-
ent must c-command the anaphor. We can replace the notion of clause-mate,
defined as in (24), with the simpler notion of ‘domain containing a finite T′.
Here the key point is that while certain infinitives may be defective in that
they are TP rather than CP, as in the complement of believe discussed above,
finite clauses are never defective; they are always full TPs with an associ-
ated CP (this is true for finite root clauses although we have frequently not
included the root CP in the structures we have given). Hence two NPs will
fail to be clause-mates if they are not contained in the same minimal finite
TP, although they may be clause-mates when contained in distinct non-finite
minimal TPs.
We are now in a position to present the sixth and final hypothesis on anteced-
ent–anaphor relations:

(36) Hypothesis Six: An anaphor must be c-commanded by an antecedent which

is contained in the same binding domain (BD).

IanRoberts_9781316519493_C006.indd 140 05-09-2022 17:32:11

 6.3 Anaphors 141

We define binding domain (BD) as follows:

(37) The BD for an anaphor A is the smallest XP containing A and either:
(a) a subject (distinct from A) or
(b) a finite T.

The crucial cases illustrating how this definition of BD works are the biclausal
ones in (25) and (28), so let us go through those once more:
(25) a. *Johni thought [CP that himselfi was a genius ].
b. Johni believes [XP himselfi to be a genius ].

(28) *Johni believes [TP Mary to like himselfi ].

The BD for himself in (25a) is the embedded TP as this is the smallest XP con-
taining a finite T (note that this TP does not contain a subject distinct from the
anaphor). The antecedent John is outside this BD and hence Hypothesis Six is not
satisfied and the relation does not hold. The sentence is therefore ungrammatical.
The BD for himself in (25b) is the root TP, as this is the smallest category con-
taining both a subject distinct from the anaphor and a finite T. The embedded TP
is non-finite and the subject is the anaphor itself, and therefore not distinct from
the anaphor. Since that BD contains the antecedent John, the relation is well-
formed and the sentence is grammatical. This account has the advantage over the
clause-mate-based definition relying on the notion of clause-mate in (24), since
we are no longer required to consider that XP is not CP.
The BD for himself in (28) is the embedded TP. Although non-finite, this is
the smallest XP containing a subject, Mary, distinct from the anaphor. The ante-
cedent John is outside this BD, and hence Hypothesis Six is not satisfied and the
relation does not hold. The sentence is therefore ungrammatical.
Hypothesis Six therefore accounts for all the examples we have seen in this
section. We can now add a general definition of binding, which will be relevant
in the rest of this chapter, and as a consequence further simplify Hypothesis Six.
The definition of binding is (38):
(38) X binds Y if and only if X c-commands Y and X is coindexed with Y.

As long as antecedents are coindexed with their anaphors, all the cases of well-
formed antecedent–anaphor relations we have seen involve the antecedent bind-
ing the anaphor by this definition. We can now restate Hypothesis Six in simpler
form as follows:
(39) An anaphor must be bound in its binding domain.

The statement in (39) captures all the facts we have seen in this section, both the
well-formed and ill-formed antecedent–anaphor relations.
In the next sections, we will see that the definition of binding generalises to
pronouns and R-expressions. We will also see that the notion of binding domain
is relevant for capturing important aspects of the distribution and interpretation
of pronouns, which is our next topic.

IanRoberts_9781316519493_C006.indd 141 05-09-2022 17:32:11

 142 6 binding

6.4 Pronouns
Remember that the definition of pronouns we are working with here
excludes reflexives (himself, etc.) and reciprocals (e.g. each other); those are
anaphors and the subject of the preceding section.
The first observation we can make about pronouns is that they do not require
a linguistic antecedent (something we already noted in our discussion of (1)).
Hence both examples in (40) grammatical:
(40) a. He grew some vegetables.
b. Vurdy said he grew some vegetables.

In (40a) the pronoun he clearly has no linguistic antecedent. Nonetheless, the

sentence is grammatical. In this respect pronouns contrast with anaphors; com-
pare the ungrammaticality of (7). The pronoun here refers to some contextually
salient male individual.
The following examples illustrate important aspects of the distribution of
pronouns:
(41) a. Vurdyi’s mother loves himi.
b. Goonai believes Vurdy to like heri.

In (41a), we see that the antecedent of the pronoun doesn’t have to c-command
the pronoun; again this contrasts with the behaviour of anaphors seen in (15). In
(41b), we see that the antecedent of the pronoun can be outside the pronoun’s
binding domain, unlike what we saw with anaphors in examples like (28). Here
we are extending the idea of binding domain to pronouns. Binding domains
were defined in (37) explicitly for anaphors; here we generalise that definition
so that it applies to pronouns:
(42) The BD for NPi is the smallest XP containing NPi and either:
a. a subject (distinct from NPi) or
b. a finite T.

This definition applies to any (indexed) NP; therefore it can still apply to anaphors
in exactly the way we saw at the end of the previous section, but now it can also
apply to pronouns (and in principle to other NPs). To repeat, then, (41b) shows
that a pronoun, unlike an anaphor, can be bound from outside its BD.
In fact, while anaphors must be in a local domain with their antecedents, as
we saw, pronouns must not be too close to their antecedents, as the following
examples show:
(43) a. Vurdyi hates him*i/j.
b. Vurdyi believes [ him*i/j to be the best ].
c. Vurdyi told Goona about him*i/j.

Here we have starred certain subscripts in order to indicate the ungrammati-

cal interpretations of the pronouns. So, in (43a) him cannot take John as its

IanRoberts_9781316519493_C006.indd 142 05-09-2022 17:32:11

 6.5 R-expressions 143

antecedent, although it is able to have some other non-linguistic antecedent.

Similarly, in (43b) and in (43c); in each case the sentence as such is grammati-
cal, but it does not allow the interpretation where the antecedent of John is him.
Once again, there is a minimal contrast with anaphors: compare (6) with (43a),
(25b) with (43b) and (32) with (43c).
The condition on binding of pronouns is in fact a negative one. There is no
specification as to where pronouns appear, but rather a constraint which states
where they cannot be interpreted as bound to an antecedent. This constraint can
be stated as follows:
(44) A pronoun cannot be bound in its binding domain.

The definition of binding domain is the generalised one in (42). The definition of
binding is (38), repeated here:
(38) X binds Y if and only if X c-commands Y and X is co-indexed with Y.

So no pronoun can have a coindexed, c-commanding antecedent in its binding

domain. Otherwise, the distribution and interpretation of pronouns is quite free:
they can have non-c-commanding local antecedents (as in (41a)), antecedents
outside their BD (as in (40b) and (41b)) and they can fail to have linguistic ante-
cedents (as in (40a)).
We can add one further definition and then restate (44):
(45) If X is not bound, then X is free.

Now we restate (44) as a positive condition:

(46) A pronoun must be free in its binding domain.

It is obvious that (44) and (46) are equivalent.

We can now move to briefly consider R-expressions.

6.5 R-expressions
In Section 6.2 we defined R-expressions as NPs with the intrinsic
semantic capacity to pick out a referent, a specific individual from the domain
of discourse (Goona, Vurdy, John, Mary, the Swedish Chef, the pet shop boy,
etc.). Exactly how NPs refer, and what the relation of reference is, are important
and difficult questions in semantic theory. For our purposes, it is sufficient to
observe that R-expressions have an intrinsic referential capacity that pronouns
and anaphors lack; at the very minimum, there is an obvious sense in which an
NP like John or that man does not require a relation with an antecedent of some
kind in order for its reference to be determined.
In terms of the notion of freedom introduced at the end of the previous sec-
tion (‘free’ simply means ‘not bound’, with binding defined as in (38)), we might

IanRoberts_9781316519493_C006.indd 143 05-09-2022 17:32:11

 144 6 binding

therefore expect that R-expressions are able to be free. This is, perhaps unsurpris-
ingly, true. But the condition on R-expressions is stronger: R-expressions must
always be free. What this means is that R-expressions can never have a c-com-
manding, coindexed antecedent. There is no locality condition associated with this
constraint; the notion of BD plays no role. If an R-expression is bound by a pronoun,
even at a distance, ungrammaticality results, as examples like the following show:
(47) a. *Hei thought [CP that [the idiot]i must be right ]].
b. *Hei said that John thinks that Mary believes that Nigeli is invited.

In (47a), the BD for the the idiot is the embedded TP, as this is the minimal XP
containing that NP and a finite T, but binding by he in the main clause is nonethe-
less not allowed. If we replace the idiot with a coindexed pronoun, the example
becomes grammatical; if we replace it with an anaphor it remains ungrammat-
ical. These outcomes are predicted by (46) and (39) respectively. Here we can
see the utility of the generalised definition of BD in (42). In (47b) Nigel is three
clauses away from he; nonetheless he cannot bind Nigel. This clearly illustrates
that R-expressions are subject to an absolute freedom requirement.
If R-expressions must be free in the technical sense of not bound, this means
that they can have antecedents; the requirement is that such antecedents must
not c-command the R-expression. So we find well-formed examples such as the
following:
(48) a. [ Hisi mother ] loves Johni.
b. [ The fact that hei failed the exam ] bothered Vurdyi.
c. [ The hobbit hei befriended ] trusted Gandalfi.

In all of these cases, the pronoun is embedded inside a subject constituent which
it cannot c-command out of (it is a good exercise to draw the tree diagrams for
these examples and work out the c-command relations). Hence, although there
is coindexation, and therefore coreference, the R-expressions John, Vurdy and
Gandalf are not bound by the pronouns in the sense defined in (38). Therefore
the relation is allowed and the sentences are grammatical with the interpretations
indicated by the coindexation.

6.6 The Binding Principles

We can now bring together what we have concluded regarding the
conditions on anaphors, pronouns and R-expressions. These are unified as the
three binding principles given in (49):
(49) Principle A: an anaphor must be bound in its binding domain.
Principle B: a pronoun must be free in its binding domain.
Principle C: an R-expression must be free.

IanRoberts_9781316519493_C006.indd 144 05-09-2022 17:32:11

 6.6 The Binding Principles 145

The definitions of binding, free and binding domain are repeated below:
(38) X binds Y if and only if X c-commands Y and X is co-indexed with Y.

(45) If X is not bound, then X is free.

(42) The binding domain for NPi is the smallest XP containing NPi and either:
a. a subject (distinct from NPi) or
b. a finite T.

The binding principles account for a wide range of facts concerning the dis-
tribution and interpretation of NPs. We can get a glimpse of this by reconsid-
ering the grammatical and ungrammatical examples we introduced in (3) and
(4) above:
(3) a. *Herself thinks that it’s going to rain.
b. *A good friend of Maryi’s has declared herselfi bankrupt.
c. *Johni likes himi.
d. *Hei likes Johni.

(4) a. Johni said that the person who had upset himi should be fired.
b. Maryi has declared herselfi bankrupt.
c. Hisi mother thinks Johni is a genius.

Let us go through these examples one by one.

Example (3a) is ungrammatical because the anaphor herself has no (lin-
guistic) antecedent, but Principle A in (49) requires anaphors to be bound,
i.e. to have a c-commanding, coindexed antecedent. Example (3b) is ungram-
matical because the anaphor herself is not c-commanded by its antecedent
Mary (the first branching node dominating Mary is the P′ headed by of;
clearly this node does not dominate herself). Example (3c) is ungrammatical
because the pronoun him is bound in its binding domain (the root TP) by
John, but Principle B in (49) requires pronouns to be free, i.e. not bound,
in their binding domain. Finally, example (3d) is ungrammatical because
the R-expression John is bound by he, but Principle C in (49) requires
R-expressions to be free.
Turning now to the grammatical examples in (4), (4a) is grammatical because
the pronoun him is bound by John, an antecedent outside the pronoun’s binding
domain (which is the TP of the relative clause (who) had upset him, the smallest
XP containing a finite T and a subject distinct from John, the latter presumably
the trace/copy of the moved relative pronoun who). This example is therefore in
conformity with Principle B in (49). Example (4b) is grammatical because the
anaphor herself is bound by the antecedent Mary in the same binding domain,
the root TP (the smallest XP containing a finite T and a subject distinct from
the anaphor). This example therefore obeys Principle A of (49). Finally, (4c)
is grammatical because, although the R-expression John is coindexed with his,

IanRoberts_9781316519493_C006.indd 145 05-09-2022 17:32:11

 146 6 binding

there is no c-command relation between the two and so John is not bound by his,
in conformity with Principle C of (49).
So we see that the binding principles give us a compact account of a wide
range of distributional and interpretational properties of different NP-types.
The binding principles are interpretative rules regulating the dependen-
cies among NPs in structures generated by PS-rules in conformity with
X′-theory, and appear to be quite distinct from movement rules (although
they are sensitive to the operations of movement rules, as we will see in the
next section).

6.7 Variables, Principle C and Movement

In our discussion of movement rules in Chapter 5, we arrived at the
following general characterisation of movement:
(50) Movement substitutes an element of type X into a higher position of type
X, where X ranges over heads, wh-phrases and grammatical-function NPs.
(See Chapter 5, (58).)

The reference to a ‘higher position’ here reflects the fact that we observed that
movement always appears to be upwards. This can be seen in the following
structures illustrating each kind of movement, again repeated from Chapter 5:

(51) a. Head-movement:
CQP

CQ′

CQ TP

T NP T′
will Cambridge
T VP
(will )
V
flood

b. NP-movement:
TP

NP T′
John
T VP
was
V NP
interviewed (John )

IanRoberts_9781316519493_C006.indd 146 05-09-2022 17:32:12

 6.7 Variables, Principle C and Movement 147

c. Wh-movement:
CQ, WhP

NP CQ, Wh′
what
CQ, Wh TP

T NP T′
did you
T VP
(did )
V NP
catch (what )

We can now observe that the correct notion of ‘upwards’ is defined by c-com-
mand: movement targets a c-commanding position. In each of (51a–c), the first
branching node dominating the target position of movement dominates the
trace/copy in the starting position. This is clear in (51b,c) where the moved
category occupies the Specifier of the root node, so that the first branching node
dominating it is the root node, which dominates everything. But it is also the
case in (51a), where the first branching node dominating CQ, the target of T-to-C
movement, is CQ′, which dominates T.
So we can replace (50) with (52):

(52) Movement substitutes an element of type X into a c-commanding position

of type X, where X ranges over heads, wh-phrases and grammatical-function
NPs.

If the moved category c-commands its trace, since c-command is part of the
definition of binding, we can ask whether a moved category binds its trace/
copy. In other words, is the moved category coindexed with its trace/copy? For
NP-movement and wh-movement of NPs the answer to this question is clearly
affirmative: if all NPs bear referential indices, then those indices will be copied
as part of the copy/paste operation of movement, so coindexation automatically
results. We may be able to extend the idea to head-movement and wh-movement
of categories other than NP (although admittedly the notion of ‘referential index’
lacks a clear semantic grounding in these cases), but let us concentrate here on
some interesting consequences of the conclusion that wh-NPs and moved NPs
in passives bind their trace/copies.
There are good semantic reasons to think that wh-NPs bind their traces/
copies quite independently of the fact that we can now see that the configura-
tion created by wh-movement creates a binding relation between them given
the definition of binding (c-command + coindexing) in (38). To see this, we
need to look in a little more detail at the semantic properties of certain kinds
of NPs.
We can start by comparing these two sentences:

IanRoberts_9781316519493_C006.indd 147 05-09-2022 17:32:12

 148 6 binding

(53) a. Goona likes pizza.

b. Every student likes pizza.

The proper-name subject in (53a) is a standard kind of R-expression: it picks

out a particular individual in the universe of discourse. The predication relation
between this R-expression subject can be construed as placing the referent of the
R-expression in the set denoted by the predicate (VP): the set of pizza-likers. So
(53a) is true if and only if Goona is a member of the set of pizza-likers, and false
otherwise. This shows how we can combine an R-expression subject with a simple
predicate and get the meaning, or the truth conditions, for the whole sentence.
Now look at (53b). The subject NP every student doesn’t refer to a single
individual (for binding-theory purposes it is nonetheless an R-expression subject
to Principle C of (49); this is where the syntactic notion of R-expression and the
semantic notion of reference come apart, as we hinted in Section 6.5). Because
of the quantifier every, this NP is a quantified expression. To cut a long, com-
plex and very interesting semantic story short, such NPs refer to a set of entities
(students) and say that every member of that set is also in the set denoted by
the predicate, the set of pizza-likers. (This is a subset relation: there can be piz-
za-likers who are not students but, if (53b) is true, there are no students who are
not also pizza-likers.)
We can give a quasi-logical representation for (53b) as follows (if this were
a proper semantic textbook, we’d have to use a much more mathematically pre-
cise formula here, but this is good enough for us):

(54) For every x, if x is a student, then x likes pizza (or ‘if x is in the set of stu-
dents, then x is in the set of pizza-likers’).

In (54), x is a variable, ranging over individuals in the domain of discourse. To get

the full meaning of (53b), we have to ‘plug in’ individuals for x (technically, we
assign values to the variable) and then see if that individual, say individual 23 =
John23, individual 5 = the man next door5, individual 666 = the pet shop boy666,
etc., is a member of the set of students and a member of the set of pizza-likers.
As long as we don’t find any individuals which are in the set of students and not
in the set of pizza-likers, then the sentence is true.
The above paragraph describes how we can give a precise semantics for NPs
which are interpreted as quantified expressions. It is absolutely crucial for this
account that the two occurrences of the variable x in ‘if x is a student, then x
likes pizza’ (54) always have the same value. The semantics breaks down if we
start asking whether if John is student, the man next door likes pizza. In logic
and semantics, these variables are said to be bound by the quantifier, and so as
we compute the semantics of the quantifier, plugging in values for the variables,
the two variables always have the same value. This notion of binding is similar
to what we have seen in syntactic binding theory: when a pronoun or anaphor is
bound by an antecedent, it has the same referential value as that antecedent, i.e.
it denotes the same (set of) individuals.

IanRoberts_9781316519493_C006.indd 148 05-09-2022 17:32:12

 6.7 Variables, Principle C and Movement 149

Now we can come back to wh-movement. Consider a simple wh-question in

the light of what we’ve just seen regarding quantified expressions:
(55) Which coursei does Goona like ti ?

The crucial observation is that which is also quantificational rather than refer-
ential: the answer to the question could be ‘none of them’, showing that which
course doesn’t have to pick anything out in the domain of discourse. Like every
student, which course applies to a set of entities (courses) and asks for the iden-
tity of a member of that set such that Goona likes it. In quasi-logic:
(56) For which x, x a course, does Goona like x ?

So we see an important semantic similarity between wh-elements and quantifiers:

both instantiate operator-variable structures. The operator (or quantifier)
applies to a set of entities and specifies the subpart of that set we are dealing with
(every member, some member, no member, or, in the case of wh-operators, find
a member). The variable tells us where in the structure you would ‘plug in’ the
specified domain in relation to the predicate so as to compute the relation to the
predicate and see if the sentence is true or false. So, in relation to (56), we plug
in different courses in the variable in object position as values of x, and ask if
Goona is a member of the set of x-likers. As with the account of the semantics
of every, it is crucial that the two x′s following ‘which x’ in (56) always have the
same value. So, these variables are semantically bound by the wh-word.
Leaving aside the syntactic counterpart of ‘x a course’ in (56), it is clear that
the syntactic counterpart of the rightmost x in (56) is the trace/copy of wh-move-
ment. So here we can conclude that syntactic binding, c-command and coindex-
ation correspond exactly to the semantic binding involved in interpreting the
wh-quantification. So now we really have a good answer for why wh-movement
is always upwards: the wh-expression must be placed in a position from which
it c-commands the trace/copy it is coindexed with in order for that trace/copy to
be semantically interpreted as a variable.
So the relation between the wh-phrase and its copy/trace in (55) is a bind-
ing relation in good standing. The obvious question to ask now is which of
the binding principles in (49) is the wh-copy/trace subject to? We can imme-
diately conclude that it can’t be Principle A, since the binding domain for
the trace/copy is the finite TP [tp Goona (does) like ti ]and the wh-phrase is
clearly outside this constituent (note in passing the importance of the copy of
the finite T left by T-to-C movement for this conclusion). But is it Principle
B or Principle C?
To answer this question, we need to introduce a pronoun into our example:
(57) Whoi did Goona say she likes ti ?

As we have seen, pronouns don’t have to have linguistic antecedents, so she here
could refer to some contextually given female individual distinct from Goona.

IanRoberts_9781316519493_C006.indd 149 05-09-2022 17:32:12

 150 6 binding

Equally, she can be coindexed with and therefore bound by Goona (since Goona
c-commands she):
(58) Whoi did Goonaj say shej likes ti ?

Principle B is respected here, since she is in a distinct binding domain from

Goona (the deepest embedded TP is a BD by definition (42)). But now, if instead
we assign index i to she, ungrammaticality results:
(59) *Whoi did Goonaj say shei likes ti ?

The question in (59) cannot have the interpretation that she is the antecedent of
who or its trace/copy. So the trace/copy cannot be an anaphor subject to Principle
A, since it would be bound by shei here, but that interpretation is unavailable.
This leaves no alternative to Principle C.
Let us now look at the interpretation of (59). Since she is now coindexed with
who, it is bound by who. Pronouns bound by quantifiers are semantically inter-
preted as variables (more on this below), so (59) has the interpretation in (60)
(taking who to mean ‘for which x, x a person’):
(60) For which x, x a person, Goona said x likes x

The interpretation in (60) would correspond to a grammatical sentence like (61):

(61) Whoi did Goonaj say ti likes herselfi ?

It is clear, though, that (59) does not have that meaning. So what is the cause
of the ungrammaticality of (59)? The pronoun she does not violate Principle
B, as who is clearly outside its binding domain (the embedded TP). Therefore
it must be the trace/copy. The trace/copy’s binding domain is the embed-
ded TP (the minimal XP containing the trace/copy, a subject and a finite T),
which also contains she. So we could conclude that the trace/copy is subject
to Principle B, which would then be violated here (and not, for example, in
(58)).
However, if we swap the positions of Goona and she in (59), and maintain the
indexing, ungrammaticality results again:
(62) *Whoi did shei say Goonaj likes ti ?

The interpretation of (62) is (63a), which corresponds to the grammatical sen-

tence (63b):
(63) a. For which x, x a person, x said Goona likes x
b. Whoi ti said Goona likes heri ?

The ungrammaticality of (62) cannot be attributed to Principle B applying to

she, since the binding domain for she is the main-clause TP (the minimal XP
that includes she and a finite T) and she is free in this domain. Neither can it
be attributed to Principle B applying to the trace/copy, since Goona, which is
in the trace/copy’s binding domain, has a distinct index and she is now outside

IanRoberts_9781316519493_C006.indd 150 05-09-2022 17:32:13

 6.7 Variables, Principle C and Movement 151

the trace/copy’s binding domain. However, we could attribute it to Principle C

applying to the trace/copy, since the trace/copy is bound by she.
Saying that traces/copies, at least of wh-movement, are subject to Principle
C has one advantage and, at first sight, one major problem. The advantage is
the connection with the semantics of variables described above. Variables are
semantically substituted by expressions with constant values (John23, etc.), as
we saw, and so we can reasonably treat them as R-expressions. On the other
hand, as we have also seen, variables are c-commanded by and coindexed with
their wh-phrases, therefore they are bound by them. On the face of it then, they
should always violate Principle C.
The solution lies in distinguishing two kinds of NP positions. In fact, we have
already implicitly made the right distinction in our discussion of NP-movement
in Section 5.6. There we characterised NP-movement in passives as moving
NPs from one grammatical-function position (object) to another (subject). All
our examples of binding have involved coindexation and c-command relations
between grammatical-function positions, mainly subject positions and object
positions, although in (13) and (33) indirect-object positions were involved. We
can also count the complement of P as a grammatical-function position (either
indirect object or oblique) and the Specifier of a possessive NP (NP2 immediately
dominated by NP1 has the possessor function). Thus, the core cases of binding
theory all involve grammatical-function positions. The target of wh-movement
SpecCWh′, on the other hand, is not a grammatical-function position: as we saw
in Chapter 5, XPs bearing any grammatical function (subject, direct object, indi-
rect object, predicate) or none (e.g. temporal when, locative where, etc.) can
appear there.
So we can distinguish grammatical-function, or A-positions, from non-gram-
matical-function or A′-positions. Now we can say that the binding conditions
hold of binding relations from A-positions (A-binding) only. The A-positions
are subject, object, indirect object, object of preposition and possessor. All other
positions are A′-positions. In that case, binding of a trace/copy by a wh-phrase
in SpecC′ does not violate Principle C, as Principle C holds only of A-binding.
On the other hand, the trace/copy in (62) is A-bound by she, so we can maintain
that the traces/copies of wh-movement, as variables, are R-expressions subject
to Principle C, and Principle C is violated in (62).
Examples like (62) where a trace/copy of wh-movement is A-bound in vio-
lation of Principle C are known as strong crossover (the idea being that the
wh-phrase ‘crosses over’ the pronoun as it moves). There is also a phenomenon
of weak crossover, illustrated by examples like (64):
(64) ?*Whoi does heri mother love ti ?

Here again, the pronoun her, since it is bound by who, is interpreted as a variable
bound by it. So (64) has the interpretation:
(65) For which x, x a person, x′s mother loves x

IanRoberts_9781316519493_C006.indd 151 05-09-2022 17:32:13

 152 6 binding

Although the interpretation is coherent, (64) is somewhat unacceptable; compare

Who is loved by her mother, which is acceptable with the interpretation in (65).
The phenomenon is called ‘weak crossover’ precisely because the unaccepta-
bility is less strong than in strong-crossover cases such as (62).
What is wrong with (64)? Here there is no Principle C violation as her does
not c-command the trace/copy. Neither is there a Principle B violation as her
is not A-bound by anything here (it is A′-bound by who, giving rise to the
bound-variable interpretation, but we have now excluded A′-binding from
core binding theory). We will leave this question open, merely noting the
phenomenon.
We have now seen some cases where pronouns are interpreted as variables
bound by wh-phrases. Pronouns can also have bound-variable interpretations
when bound by other quantified expressions, e.g. every student in (66):
(66) Every studenti thinks theyi are the best.

The natural interpretation of (66) is (67):

(67) For every x, if x is a student, then x thinks x is the best.

The expression in (67) can be interpreted by the procedure of plugging in values

for variables described above.
There are two things to note about bound-variable pronouns. First, the
interpretation depends on the quantified NP c-commanding the pronoun. So the
interpretation is unavailable in cases like the following:
(68) The man who talked to every studenti thinks theyi are the best.

Here the only interpretation available seems to be that where the man thinks that
the set of students as a whole are the best; the bound-variable interpretation (‘the
man who talked to every x, x a student, thinks x is the best’) is not available.
This is because every student does not c-command they here.
Second, the bound pronoun obeys Principle B, as we would expect, since it
is A-bound. Hence (69a) is ungrammatical on any interpretation, while (69b)
shows that anaphors can also have bound-variable interpretations provided
Principle A is satisfied:
(69) a. *Every boyi admires himi.
b. Every boyi admires himselfi.

As a corollary of obeying Principle B, a pronoun can have a bound-variable

interpretation at a considerable syntactic distance from its quantified antecedent:
(70) Every politiciani betrays the people who vote for himi.

Here, him is inside a relative clause, yet it is c-commanded by every politician,

Principle B is not violated and the bound-variable interpretation is available.
As a final point, if traces/copies are bound by their moved antecedents, the
traces/copies of NP-movement must be A-bound. So, again, we can ask which

IanRoberts_9781316519493_C006.indd 152 05-09-2022 17:32:13

 6.8 Conclusion 153

of the binding principles applies to them. Here the answer seems clear from a
simple passive sentence like (51b), repeated here:
(51b) Johni was interviewed ti.

The trace/copy is clearly in the same binding domain as John, and the sentence
is grammatical. Therefore the trace/copy must be subject to Principle A. If the
trace/copy were subject to either Principle B or Principle C, the sentence would
be ungrammatical. This conclusion is supported by examples like those in (71):
(71) a. *John was believed that (John) was bankrupt.
b. John was believed (John) to be bankrupt.
c. *John was believed Mary to like (John).

Here the object-to-subject movement relation looks similar to the a­ ntecedent-­

reflexive relation seen in examples (21), (25) and (28) in Section 6.3.
In this section we have made some connections between binding theory and
movement. We have seen good reasons, both syntactic and semantic, to ana-
lyse the trace/copy left by wh-movement as a variable. Making the distinction
between A- and A′-binding, we can regard these variables as R-expressions
subject to Principle C. This gives us an account of the strong-crossover phe-
nomenon. We also noted that pronouns can function as variables bound by
quantified antecedents and that the trace/copy of NP-movement is subject to
Principle A. The question of the status of traces/copies of head-movement
remains open. Since heads do not denote (sets of) individuals – only NPs to
do this – then the question may simply not arise; we will return to this point in
Section 8.6.

6.8 Conclusion
This chapter completes our introduction to the central mechanisms
of the theory of syntax. Now we have seen how structures are generated by
PS-rules following the X′-template, how they can be modified by movement
rules and how the rules of binding theory regulate distributional and interpreta-
tive dependencies of various kinds. Obviously, very many questions have been
left open, or barely addressed at all. Nonetheless, the core ideas and mechanisms
are presented here.
One very major question needs to be addressed: how far are the mechanisms
and phenomena they are intended to account for idiosyncratic to English? We
have concentrated almost exclusively on English up to now, largely for exposi-
tory convenience. But we said at the outset that we are trying to construct a gen-
eral theory of syntactic structure, one that applies in principle to all languages.
It is therefore necessary, now that the basic notions are in place, to try to apply
the approach to some data from other languages. Unsurprisingly, it will emerge
that other languages, to varying extents, are organised differently from English.

IanRoberts_9781316519493_C006.indd 153 05-09-2022 17:32:13

 154 6 binding

So a very important question emerges: how can we best account for syntactic
variation across languages? This will be the topic of the next chapter.

Exercises
1. Give the trees for the following sentences:
a. Sarah likes herself.
b. Mark said that Sarah should promote herself.
c. *Sarah said that Mark should promote herself.
d. *Sarah thinks that herself/himself is the best.
e. Sarah considers Mark to like himself/*herself.
f. Sarah’s brother likes himself/*herself.
Say why the ungrammatical examples are ungrammatical. Replace
the reflexives with the corresponding non-reflexive pronouns, give
your judgements as to the grammaticality and the possible inter-
pretations of the resulting examples, and explain them as far as
you can.
2. In the following examples, take each other to be an NP with the
structure [NP each [N′ [N other ]]. There are two generalisations gov-
erning the distribution of this NP, one from binding theory and
another concerning its intrinsic nature. Try to state both (hint: for the
second one, try to make friend plural in (e) and see what happens):
a. The hippies love each other.
b. *Each other love the hippies.
c. I asked the hippies about each other.
d. *I asked each other about the hippies.
e. *The hippies’ friend talked to each other.
f. *The hippies said that each other had left.
g. The hippies believe each other to be geniuses.
h. *John believes Mary to like each other.

For Discussion

1. If you had to decompose the structure of NPs like himself, what

would their internal structure look like?
Think of (a) X′-theory and (b) the 1st- and 2nd-person reflexives
myself, yourself, ourselves, etc., and non-standard forms like hisself,
theirselves.

IanRoberts_9781316519493_C006.indd 154 05-09-2022 17:32:13

 Further Reading 155

What would that structure suggest about (a) the morpheme -self and (b)
pronouns? Can you think of any support for treating self as an autonomous
morpheme?

Further Reading
Carnie, A. 2013. Syntax: A Generative introduction. Oxford: Blackwell, Chapters 4, 5, 15.
Haegeman, L. & J. Guéron. 1999. English Grammar. Oxford: Blackwell, Section 4.1.
Isac, D. & C. Reiss. 2008. I-Language. Oxford: Oxford University Press, Chapter 8.
Roberts, I. 1997. Comparative Syntax. London: Arnold, Chapter 3.
Sportiche, D., H. Koopman & E. Stabler. (2014). An Introduction to Syntactic Analysis
and Theory. Oxford: Wiley, Chapter 7.

IanRoberts_9781316519493_C006.indd 155 05-09-2022 17:32:13

 7 Syntax beyond English

7.1 Introduction and Recap

In the preceding chapters we have gradually put together the building
blocks of the theory of syntax. We have seen that sentences can be broken down into
their constituent structure, and how to represent this and test for it (Chapter 2). Then
we saw how PS-rules can generate the well-formed structures (Chapter 3). Next,
we replaced the rather general PS-rule format with the simpler and more abstract
X′-theory (Chapter 4). Chapter 5 introduced and motivated movement rules, rules
which modify structure rather than directly generating it. Finally, in Chapter 6, we
introduced binding theory, which regulates the distribution and interpretation of
certain syntactic dependencies including at least some cases of movement.
The topic of the current chapter is to investigate how far and in what ways
the syntactic tools we have developed up to now, very much on the basis just
of English data, can apply to other languages. Inevitably, our discussion will be
limited; it is obvious that an exhaustive treatment of comparative syntax cannot
even be seriously begun in this introductory context. But nonetheless, looking
beyond English, even just a little bit, can be very revealing and can give us a
sense of the task at hand. Certain useful ideas and important generalisations will
emerge from our discussion, as we shall see.

7.2 Approaching Universal Grammar

As we said in the Introduction, Universal Grammar (UG) is the theory
of the first factor which makes up the language faculty: our innate genetic endow-
ment for language. In terms of this idea, it is natural to see specific I-languages as
variations on a single theme. Therefore, all other things being equal, we want our
syntactic representations of comparable (i.e. semantically matching, or at least
highly similar) sentences in different languages to be as uniform as possible. Of
course, things may very well not be equal quite a lot of the time, but we can per-
haps apply a kind of comparative Occam’s Razor when we start to do compara-
tive work: do not multiply cross-linguistic differences beyond what is necessary.
In these terms, since the notion of constituent structure is well established
on the basis of English, there is no reason to start out from the assumption that
other languages are organised syntactically along radically different lines. It is

156

IanRoberts_9781316519493_C007.indd 156 9/6/2022 1:55:22 PM

 7.3 Verb Positions in French and English 157

a perfectly reasonable assumption that all languages have constituent structure;

this assumption is a consequence of the fact that the rule systems of all lan-
guages are fundamentally the same. Let us then make the hypothesis that all
languages use PS-rules to generate constituent structure, and that X′-theory dic-
tates a common cross-categorial structure. This, of course, does not imply that
any and every structure that we assign to English sentences will be found in
other languages, only that the building blocks of X′-theory are universal. We
thus expect notions of head, complement and phrasal projection to be universal,
along with, again all other things being equal, Specifiers and adjuncts.
So much, for now, for the category-neutral template. What about the catego-
ries themselves? We mentioned in Chapter 2 that the noun–verb distinction may
well be universal, while adjectives and prepositions probably are not. We can
further add that the inventories of functional categories and features undoubtedly
vary across languages: English lacks grammatical gender features, while Italian
and other Romance languages have it, for example. Many languages (almost all
the Slavonic languages, Latin, Chinese, Japanese and many others) lack articles,
Chinese may well lack Tense, it is uncertain whether Japanese really has com-
plementisers (but see below), and so on. Conversely, other languages have gram-
matical categories, probably structurally realised as functional projections, that
English lacks, such nominal classifiers, evidential markers and many others.
However, a reasonable hypothesis concerning all this variation might be to
say that there is a universal ‘pool’ of categories, and individual languages make
a selection among these. There may be constraints on what the selection can be;
for example, all languages have negation (and there may be functional explana-
tions for those constraints). Certain categories, perhaps Nouns and Verbs, may
fall outside of this pool, being absolutely required by UG. Remarkably little can
be said with any confidence on these points, but the idea of a universal pool of
some kind seems somewhat plausible.
More broadly, we could think of a universal pool of possible ways in which
X′-theory, movement rules and interpretative rules like binding theory may oper-
ate. There may be a universal core to these aspects of syntax, with associated
parameters of variation, specifying various options for how core processes
can apply. This approach to syntactic variation has been fairly widely adopted
since the 1980s, and we will see some examples of how it works in what follows.
Whether this parameter-based approach will be able to account for all syntac-
tic variation remains an open question. But let’s begin with a simple example,
involving an interesting set of differences between English and French.

7.3 Verb Positions in French and English

English and French are both Indo-European languages, and are
structurally fairly similar. In particular, both languages show a basic subject-­
object-verb (SVO) word order, which we can see in (1) ((1b) is a word-for-word
translation of (1a), and is fully grammatical):

IanRoberts_9781316519493_C007.indd 157 9/6/2022 1:55:22 PM

 158 7 syntax beyond english

(1) a. Harry and Hermione follow Filch.

b. Harry et Hermione suivent Rusard.

The SVO order is ‘basic’ in the sense that although it can be changed by move-
ment rules (e.g. wh-movement could place the object in the first position,
SpecC′), it is the neutral order for expressing simple declarative sentences with-
out any particular emphasis on any constituent.
The structure of (1a) is given in (2):
(2) TP

NP T′
Harry et
Hermione T VP
[Pres]
V NP
follow Filch

Our comparative Occam’s Razor would lead us to assign exactly the same struc-
ture to the French sentence (1b). But consider the next pair of sentences:
(3) a. Harry and Hermione often follow Filch.
b. Harry et Hermione suivent souvent Rusard.

Here we see a small but important difference in word order between the two
languages: while the adverb often precedes the Verb follow in English, the order
of their French counterparts is the opposite: suivent (‘follow’) precedes souvent
(‘often’). Moreover, in both languages the opposite order, i.e. the one found in
the other language, is ungrammatical:
(4) a. *Harry and Hermione follow often Filch.
b. *Harry et Hermione souvent suivent Rusard.

For the English example (3a), it seems reasonable to treat often as an AdvP
adjoined to V′, since it provides extra information about the action of following
Filch. So we can give (3a) the structure in (5):
(5) TP

NP T′
Harry and
Hermione T VP
[Pres]
V′

AdvP V′
often
V NP
follow Filch

IanRoberts_9781316519493_C007.indd 158 9/6/2022 1:55:23 PM

 7.3 Verb Positions in French and English 159

What should we say about the French sentence (3b)? We either have to say that
there’s something different about the Verb as compared to English or some-
thing different about the Adverb (we could say both are different, but that
would clearly violate Occam’s Razor). The adverb souvent means the same
thing as English often and modifies the action described by V′ in the same
way; there do not seem to be very good grounds for attributing the difference
between the two languages to this element. On the other hand, the Verb suiv-
ent does differ from English follow in that it is overtly marked for present
tense while follow does not show any overt marking for Tense (this does not
mean that there isn’t an abstract Tense affix; see Section 8.6). French verbs
are also marked to varying degrees for person and number: this Verb has the
third-person plural ending -ent. Forms of this Verb in the third-person plural
in other tenses include suivaient (imperfect, ‘used to follow’), suivirent (past,
‘followed’), suivront (future, ‘will follow’) and suivraient (conditional, ‘would
follow’). As we saw when we compared the fish sentences in English and
Italian in Chapter 2, English is rather poor in overt morphological marking
of Tense, Mood and Agreement, while Italian is quite rich. We can see now
that French is more like Italian in this respect (this isn’t surprising, as French
and Italian are both quite closely related Romance languages). Since we’ve
been assuming all along that Tense information is present on T, we could pro-
pose that the French Verb moves to T. Keeping souvent in exactly the same
V′-adjoined position as in English, this gives us the structure (6) for the French
sentence (3b):
(6) TP

NP T′
Harry et
Hermione T VP
[Pres]

V
suivent V′

AdvP V′
souvent
V NP
(suivent) Rusard

Substituting suivent into T[Pres] may force the two elements to combine as a
single head, hence the correlation between V-to-T movement and morphologi-
cal tense marking.
So we observe a difference between English and French, which we can call
the V-to-T Parameter. This is a parameter defining a specific head-movement
option, whether a given language has V-to-T movement or not. In English, this
parameter is set to negative, so there is no V-to-T head movement. In French, it
is set to positive: there is V-to-T head movement. The difference is not always

IanRoberts_9781316519493_C007.indd 159 9/6/2022 1:55:24 PM

 160 7 syntax beyond english

apparent, as the examples in (1) show, but it emerges where adverbs like often
are adjoined to V′. It also shows up in negative sentences and various other
environments, but our illustration here will suffice for now. The parametric dif-
ference may well be connected to the richer verbal tense inflection of French as
compared to English.

7.4 Word Order and X′-theory

We mentioned in the previous section that English and French are
both SVO languages. Taking these three elements as basic to simple transitive
clauses, a reasonable question is what other orders of these elements are found.
A strongly universalist view might expect to find no variation; then the uni-
versality of SVO would need to be explained. However, it seems clear that the
strong universalist view doesn’t hold up in this case, since all six possible per-
mutations of the three elements are attested in the world’s languages. A strongly
‘egalitarian’ view might expect that all six orders are found roughly equally
distributed across the world’s languages, i.e. about a sixth of all languages show
each order, or roughly 16–17 per cent of languages per order. That is not true
either. Certain orders are frequent, others very rare, although, as already men-
tioned, all the possibilities are found.
According to the most recent data from the World Atlas of Language Structures
(WALS; wals.info), an online database of 192 phonological, morphological, syn-
tactic and lexical features of 2,662 languages, the basic word orders of Subject,
Verb and Object in 1,376 languages pattern as follows:
(7) SOV John cake ate. 564 (e.g. Japanese)
SVO John ate cake. 488 (e.g. English)
VSO Ate John cake. 95 (e.g. Welsh)
VOS Ate cake John. 25 (e.g. Malagasy, Madagascar)
OVS Cake ate John. 11 (e.g. Mangarrayi, Australia)
OSV Cake John ate. 4 (e.g. Warao, Venezuela)
(World Atlas of Language Structures, Feature/Map 81A; https://wals.info/
feature/81A#2/18.0/153.1)

A further 189 languages are listed as ‘no dominant order’, i.e. either languages
for which a single, ‘basic’ order cannot be discerned on the basis of available
data, or for which more than one order is ‘basic’ depending on various factors.
Leaving aside these languages, we see that 564 out of the remaining 1,187
languages (just over 47.5%) are SOV and 488 (just over 41.1%) are SVO. So
together these two orders make up almost 90% (88.6%) of the world’s lan-
guages. Among the other orders, the 95 VSO languages make up 8% of the total
and just over 70% of the 135 non-SOV, non-SVO languages. The other orders
are very rare, at 2.1% (VOS), 0.92% (OVS) and 0.37% (OSV) of the total.

IanRoberts_9781316519493_C007.indd 160 9/6/2022 1:55:24 PM

 7.4 Word Order and X′-theory 161

Map 7.1 Geographical distribution of ‘basic’ S, V, O orders (WALS Map

81A; https://wals.info/feature/81A#2/18.0/153.1)

The geographical distribution of the orders is shown on Map 7.1.

Here we can see clear areal patterns for SOV (the Indian subcontinent, most
of Asia except the South-East, a belt across Central Africa and a preponderance
in North America) and SVO (Europe, South-East Asia, Sub-Saharan Africa and
Indonesia/Papua New Guinea), with pockets of VSO languages in the British
Isles (the Celtic languages), the extreme North-West of Africa, the Philippines,
the Pacific North-West of North America and Mesoamerica. It is not easy to see
areal patterns for the other orders, since they are so rare, although the majority
of O-initial languages are found in Amazonia. Various explanations have been
proposed for these patterns, but we will not go into those here. What is clear is
that two patterns, SOV and SVO, are much more common than all the others, so
the variation is not random.
Let us compare an SOV and an SVO language, taking Japanese as our SOV
language and French as our SVO one. The simple sentences in (8) illustrate the
different orders:
(8) a. Thierry écrivit la lettre
Thierry wrote the letter
S V O
b. Taroo-ga tegami-o kaita
Taroo-nom letter-acc wrote
S O V
Clearly, the basic difference is whether the Verb appears to the left or to the
right of the object. We could say, then, that French has the order V – NP in V′,
Japanese has the opposite order:

IanRoberts_9781316519493_C007.indd 161 9/6/2022 1:55:24 PM

 162 7 syntax beyond english

(9) a. French: V′

V NP

b. Japanese: V′

NP V

In terms of the category-neutral X′-template, we could generalise from (9)

to (10):
(10) a. French: X′

X YP

b. Japanese: X′

YP X

If French generalised the head-initial pattern in (6a) across categories, then we

would expect to find that P precedes its complement, i.e. Prepositions. This is
what we indeed find in French:
(11) sous l’océan
under the sea

On the other hand, (10b) predicts that in Japanese the complement of P should
precede it, i.e. we expect to find Postpositions in Japanese, and we do:
(12) Nihon kara
Japan from
‘from Japan’

We could expect the same with functional categories, so in French T is expected

to precede VP. We showed this order in the structure in (6), but, assuming that
auxiliaries appear in T (without moving there from V), then we expect to find
the order Aux – V in French, which we do:
(13) Thierry a écrit la lettre.
Thierry has written the letter

Again, in Japanese we expect to see VP-T order. It is a little difficult to discern

auxiliaries in Japanese, but the following example may suffice:
(14) John-ga Mary-to renaisite iru.
John-nom Mary-with in-love   is
‘John is in love with Mary.’

Furthermore, we expect C to precede TP in French. Again, this is clearly the

case:

IanRoberts_9781316519493_C007.indd 162 9/6/2022 1:55:25 PM

 7.4 Word Order and X′-theory 163

(15) J’ Ai entendu [CP que [TP Muriel travaille au musée ]]

I have heard   that  Muriel works at.the museum

In Japanese C should follow TP, i.e. complementisers are expected to follow the
embedded clause:
(16) Bill-ga [CP [TP Mary-ga [VP John-ni [NP sono hon-o ] watasita ] to ] itta
Bill-nom Mary-nom John-dat that book-acc hand-pst that said
‘Bill said that Mary handed that book to John’

So we can define the Head Parameter in X′-theoretic terms as follows:

(17) Does X precede or follow its complement in X′?

(It is a useful convention to formulate parameters as yes/no questions since

this brings out their binary nature and suggests how the acquirer might inter-
rogate the data it is exposed to). We see that French is head-initial, taking
the ‘precede’ option in (17). It therefore has VO order (8a), Prepositions
(11), Aux-VP order (13) and C-TP order (15). Japanese, on the other hand, is
head-final, taking the ‘follow’ option in (16). It therefore has OV order (8b),
Postpositions (12), VP-Aux order (14) and TP-C order (16). These are the har-
monic word-order patterns predicted by the X′-parameter in (17). Moreover,
since subjects occupy a Specifier position, they are not affected by (17). If
the universal, non-parametrised position of the subject is SpecT′, then (17)
predicts SOV and SVO orders. These assumptions appear to be correct at first
sight for 88.6 per cent of the languages assigned dominant basic word-order
patterns in WALS, as we saw in (7). We can assume that the four minority pat-
terns involve various marked options, which we do not need to go into here.
WALS does not give data on Aux-VP or C-TP orders. It does, however, give
data on Pre- and Postpositions in 1,089 languages, shown in Map 7.2.
As with SOV and SVO order in Map 7.1, we see clear areal clusters: post-
positions are predominant in the Indian subcontinent and in Asia as a whole,
except for South-East Asia, as well in the Americas, West Africa and most
of Australia, while prepositions dominate in Europe, Sub-Saharan Africa,
South-East Asia, Indonesia/Papua New Guinea, Mesoamerica and the Pacific
North-West.
Table 7.1 gives the data from WALS (Maps 83A Order of Object and Verb,
and 85A Order of Adposition and Noun Phrase), showing the correlation of OV
with Postpositions and VO with Prepositions.
Here the total number of languages is 982, and so the figures almost cor-
respond to percentages to one decimal place. We see the predicted harmonic
patterns, OV and Postpositions and VO and Prepositions, in 928 of the 982 lan-
guages, 94.5 per cent of the languages.
So we see that the X′-based Head Parameter of (17) makes quite good pre-
dictions both for the verb-object/adposition-order case and for the general

IanRoberts_9781316519493_C007.indd 163 9/6/2022 1:55:25 PM

 164 7 syntax beyond english

Map 7.2 Distribution of postpositional and prepositional languages (WALS

Map 85A; https://wals.info/feature/85A#2/16.3/153.1, inpositions, no
­dominant order and no-adpositions omitted)

Table 7.1 Correlations between Order of Object and

Verb and Order of Adposition and Noun Phrase (based
on WALS Maps/Features 81A and 85A.

OV & postp. 472

OV & prep. 14
VO & postp 42
VO & prep. 456
TOTAL 982

preponderance of SOV and SVO orders, although the disharmonic and minority patterns
must of course be accounted for. The parameter in (17) also makes predictions about
the order of Nouns and their complements and of Adjectives and their complements,
which we have not gone into at all here. The parameter-based approach to accounting for
the distributions of some of the major word-order patterns in the world’s, based on the
X′-parameter in (17), certainly shows some initial promise. We will return to this matter
in detail in Volume III, armed with a much more sophisticated approach to word-order
variation.
We have concentrated here on harmonic head-initial or head-final languages, as
defined by (17), taking French and Japanese as our respective case studies. Clearly,
though, there are quite a number of languages with no dominant order, in WALS terms,
or with disharmonic order. In the next section we will look at a well-known case of this
kind: German.

IanRoberts_9781316519493_C007.indd 164 9/6/2022 1:55:25 PM

 7.5 A Case Study: German 165

7.5 A Case Study: German

German is classified by WALS as ‘No Dominant Order’ under Map/
Feature 81A (Order of Subject, Object and Verb), and as displaying both SOV
and SVO order under Map/Feature 81B (Languages with Two Dominant Orders
of Subject, Object and Verb). The sentences in (18) illustrate these orders:
(18) a. Hermann schreibt den Brief.
Hermann writes the letter
S V O

b. Ich weiß daß Hermann den Brief schreibt.

I know that Hermann the letter writes
        S O V

The structure of the SOV embedded clause in (18b) is (19):

(19) TP

NP T′
Hermann
T VP
[Pres]

V′

NP V
den Brief schreibt

If this is the correct structure, then the question of how (18a) is related to it
immediately arises. In fact, as pointed in the commentary text to WALS Feature
81B, German is verb second rather than SVO in main clauses, while it is clearly
SOV in embedded clauses. We now look at the difference between SVO and V2.
The V2 nature of German main clauses can be seen from the following
examples:
(20) a. Am Tisch schreibt Hermann den Brief.
on table writes Hermann the letter
‘Hermann writes the letter on the table.’

 b. Was schreibt Hermann?

what writes Hermann
‘What does Hermann write/what is Hermann writing?’

c. Einen Brief schreibt er.

a    letter   writes he
‘He is writing a letter.’

d. Gestern hat Hermann den Brief geschrieben.

yesterday has   Hermann the letter written
‘Hermann wrote the letter yesterday.’

IanRoberts_9781316519493_C007.indd 165 9/6/2022 1:55:25 PM

 166 7 syntax beyond english

e. Hermann  hat den Brief geschrieben.

Hermann has the letter written
‘Hermann has written the letter.’

Alongside the subject-verb order in (18a), what we see here is PP-verb in (20a),
wh-phrase verb in (20b), object-verb order in (20c), adverb-auxiliary order in
(20d) and subject-auxiliary order in (20e). In each case, exactly one XP precedes
the finite verb, i.e. the verb occupies the linear second position. Examples (20d)
and (20e), where we have a combination of finite auxiliary and a verbal parti-
ciple, shows that it is the finite element, not strictly speaking the lexical verb,
which occupies second position. Here, the main verb is final, following its
object, and so we have OV order.
Furthermore, the V2 constraint is very strict. Orders where two XPs precede
the finite verb in main clauses are, with certain well-defined exceptions which
we need not go into here, ungrammatical. So the word-for-word counterpart of a
simple English sentence like (21a) is ungrammatical in German, as (21b) shows.
Instead, either the subject or the adverb has to occupy a postverbal position, as
seen in (21c) and (21d):
(21) a. Yesterday John wrote a letter.
 b. *Gestern Johann schrieb einen Brief.
c. Gestern schrieb Johann einen Brief.
d. Johann schrieb gestern einen Brief.

So we have two questions to answer about German. First, how do we account

for the V2 order in main clauses? Second, how do we account for OV order in
subordinate clauses (and when the main verb is non-finite in a main clause, as
in (20d) and (20e))?
In order to understand V2, we need to know what the initial XP-position is.
Here the wh-question in (20b) gives us an important clue. We saw in Chapter 5
that wh-phrases raise in English to SpecCWh,Q′, as in (22), repeated from (30) in
Chapter 5:
(22) CQ WhP

NP CQ Wh′
what
CQ Wh TP

T NP T′
did you
T VP
(did)
V NP
catch (what)

Let us assume that wh-movement behaves in the same way in German. Then
the first position in (20b) is SpecC′ and the Verb schreibt is in C. We can think

IanRoberts_9781316519493_C007.indd 166 9/6/2022 1:55:26 PM

 7.5 A Case Study: German 167

that XP-movement to SpecC′ in German and V-to-C movement don’t depend on

CWh, but are operations that take place in any root CP. So we conclude that the
initial XPs in (20) are all in SpecC′. By this logic, the same goes for the subject
in (18a).
If the XP is in SpecC′, then we can get the ‘second-position effect’ directly
from the fact of X′-theory that there is just one Specifier per head (cf. PS-rule
(2a) of Chapter 5), as long as V is in C. Moreover, we already have the ana-
lytical tools to guarantee that V gets to C in German. We saw in Section 7.3
that there is a parameter regulating whether V raises to T. This parameter is
positive in French, but negative in English, as we saw. Suppose it is positive
in German; this is actually much more difficult to demonstrate for German
than it is for French owing to the V2 nature of main clauses and the OV
nature of embedded clauses in German, but let us make the assumption in
any case. In Section 5.2, we saw that English has a rule raising T to C, giv-
ing subject-auxiliary inversion in questions, for example. In a language with
the positive value of the V-to-T parameter, T-to-C movement could place a
main verb which has moved to T in C. In fact, we see this in French, in direct
questions like (23):
(23) Suivent-ils souvent Rusard?
Follow they often Filch?
‘Do they follow Filch often?’

So we can say that French is just like English is having CQ cause T-to-C
movement. The difference between French and English is that main verbs
like suivent in (23) can raise to T, and from there move on to C (where C has
the right features). English, lacking V-to-T movement, has to use the dummy
auxiliary do here.
Coming back to German, we can say that German resembles French in having
both V-to-T and T-to-C, but it differs from French (and English) in that T-to-C
does not depend on CQ; it seems that any root C-type can trigger T-to-C. In
other words, this property generalizes to all (root) Cs in German, as does the
triggering of XP-movement to SpecC′, as we saw. In this way, we arrive at a full
account of V2.
Turning now to OV order, we can account for this in terms of the OV-VO
parameter in (9) in the previous section, repeated here:

(9) a. French: V′

V NP

b. Japanese: V′

NP V

IanRoberts_9781316519493_C007.indd 167 9/6/2022 1:55:26 PM

 168 7 syntax beyond english

If we say that German, like Japanese, chooses (9b), then V will follow the
object.
All of this will give us the structure in (24b) for the sentence in (24a):
(24) a. Hermann hat den Brief geschrieben.
Hermann has the letter written
‘Hermann has written the letter.’
b. CP

NP C′
Hermann
C TP

T NP T′
hat (Hermann)
T VP
(hat)
V′

NP V
den Brief geschrieben

In (24b) we see three important aspects of this kind of V2 sentence, with an

initial subject and a periphrastic tense (consisting of the auxiliary hat and
the participle geschrieben). First, the subject Hermann raises from SpecT′
to SpecC′; this is comparable to what we saw in Section 5.4 with subject
wh-movement in English, except that, thanks to the raising of T to C, the move-
ment is not string-vacuous here (compare (30) and (33) in Chapter 5). Second,
we see T-to-C movement, exactly as in English and French, except that the
movement is not sensitive to an interrogative feature of C, but applies in all
root clause types. Third, we see head-final order in V′: the object den Brief
precedes the main verb geschrieben. This is a good example of how mecha-
nisms with an independent existence in other languages (T-to-C movement as
in English and French, OV order in V′ as in Japanese, XP-movement to SpecC′
as in English) combine to give the specifically German sentence-type we see
here. Aside from extending T-to-C movement and XP-movement to SpecC′ to
non-interrogatives, we have introduced no new mechanisms in our analysis of
German V2; we have merely combined independently needed mechanisms in
a new way.
We can account for the root nature of V2 in terms of the idea that a com-
plementiser is always present in finite, non-root CPs. The complementiser
occupies C and thereby prevents T-to-C movement. The fact that where there
is no T-to-C movement there is no XP-movement to SpecC′ suggests that
the latter operation somehow depends on the former, although this is poorly
understood.

IanRoberts_9781316519493_C007.indd 168 9/6/2022 1:55:26 PM

 7.5 A Case Study: German 169

To complete the picture, (25a) and (25b) represent the case of V2 where no
auxiliary is present and the main verb raises to T and C:
(25) a. Hermann schreibt den Brief.
Hermann writes the letter
b. CP

NP C′
Hermann
C TP

T NP T′
(Hermann)
V T VP
schreibt
V
(schreibt)
V′

NP V
den Brief (schreibt)

Everything is as in (24), except that the main verb schreibt moves first to T (as
in French, see Section 7.3) and then T-to-C movement raises it to C.
We see in (25b) that German has the order OV in V′. As we mentioned above,
this corresponds to the Japanese choice in (9b). However, German clearly does
not take the general category-neutral option of (10b), repeated here, unlike
Japanese:
(10b) X′

YP X

For example, we can see from our V2 examples and from the structures
in (24b) and (25b) that C precedes TP. This is confirmed by the position of
complementisers like daß (‘that’), which precedes TP, as we saw in (18b),
repeated here:
(18b) Ich weiß [CP daß Hermann den Brief schreibt ].
I know that Hermann the letter writes

In fact, it may be correct that T follows VP, contrary to what is shown in (24b)
and (25b). If auxiliaries such as hat are in T, then the fact they follow the main
verb in non-V2 clauses indicates this. This is illustrated in (26a), with the
structure in (26b):
(26) a. … daß Hermann den Brief geschrieben hat.
… that Hermann the letter written has
‘… that Hermann has written the letter’.

IanRoberts_9781316519493_C007.indd 169 9/6/2022 1:55:26 PM

 170 7 syntax beyond english

b. CP

C′

C TP
daß
NP T′
Hermann
VP T
hat
V′

NP V
den Brief geschrieben

If we flip the order of VP and T inside T′ in (24b) and (25b), it makes no differ-
ence to the eventual surface word order, since T is in both cases occupied by the
trace/copy of the auxiliary or the verb.
So German is head-final in VP and TP, but head-initial in CP. Furthermore,
German is head-final in A′:
(27) sich seiner Sache sicher
oneself.dat one’s subject.gen certain
‘to be sure about something’

Strikingly, German has mixed order in adpositional phrases. It is head-initial

with most Ps, but there are several postpositions:
(28) a. nach Berlin
to Berlin

b. den Fluss entlang

the river along
‘along the river’

So German is a good example of a mixed word order, disharmonic lan-

guage. It is certainly not the only one; Latin has quite similar overall word
patterns to German (although Classical Latin was not V2), and Mandarin
Chinese is disharmonic but in quite a different way from German or Latin.
This shows that we cannot deduce all word-order properties in all languages
from the Head Parameter in (17); in some languages we need to look at
subcases of (17). However, in very many languages, as we saw with our
examples from French and Japanese in the previous section, (17) makes the
right predictions.
To sum up, German has the following positive parameters:
(29) a. Head-initial: C, P (partially); Head-final: V, T, A, P (partially)
   b. V-to-T movement
c. T-to-C movement in all main clauses
   d. XP to SpecC′ movement in all main clauses

IanRoberts_9781316519493_C007.indd 170 9/6/2022 1:55:27 PM

 7.6 The Wh-Movement Parameter 171

Compare (29) with the English parameters in (30):

(30) a. Head-initial
   b. No V-to-T movement
c. T-to-C movement in main-clause interrogatives
   d. Wh-XP to SpecC′ movement in wh-interrogatives

French is like English except it has V-to-T movement; in this respect it is like
German, as we have seen. In the next section, we will look at an example of a
language lacking wh-movement.
Here we have gone into some detail on German syntax, although of course
there is much more to say and we have simplified a couple of points. The pur-
pose of the discussion was to illustrate how various parameters combine to
give us the particular properties of German, such as V2 and OV order non-
V2 clauses. We also saw that German is disharmonic in relation to the Head
Parameter, indicating that there is more to say about word-order parameters than
(17), although (17) does a lot of work in a lot of languages. In general, we see
how the parametric approach can be applied; in principle, it could be applied in
this way to any language.

7.6 The Wh-Movement Parameter

There are quite a few languages which apparently lack wh-movement;
WALS (Map/Feature 93; https://wals.info/chapter/93) states that 615 of 902
languages surveyed lack obligatorily initial wh-phrases, which we can at least
tentatively take to mean that wh-movement to SpecC′ (the initial position in
direct questions) is not required. Mandarin Chinese is a well-known example of
a language lacking wh-movement, as the following examples illustrate:
(31) a. Zhangsan xiang-zhidao [CP Lisi mai-le shenme ].
Zhangsan wonder Lisi buy-Perf(P) what
‘Zhangsan wonders what Lisi bought.’

b. Ni weishenme bu qu ne?
You why not go wh
‘Why aren’t you going?’

As the English translation shows, and as we saw in Chapter 5, in English

wh-movement to SpecC′ is required in both direct and indirect interrogative
clauses.
Of course, we can easily add a further movement parameter to our inventory.
We have seen two head-movement parameters: V-to-T movement and T-to-C
movement, and we have seen variation with respect to both in that English lacks
V-to-T movement and T-to-C movement takes place in a wider range of contexts in
German than in English or French. So we can easily add a wh-movement parame-
ter to describe the difference between Chinese and English that we observe in (31).

IanRoberts_9781316519493_C007.indd 171 9/6/2022 1:55:27 PM

 172 7 syntax beyond english

However, if Chinese lacks wh-movement, two questions arise. First, how, if

at all, does Chinese mark wh-interrogatives? Second, if, as we saw in Section
6.7, wh-movement takes place in order to bind the trace/copy as a variable so
that the wh-quantifier can be semantically interpreted, how does this happen in
a language without wh-movement? Here we will answer the first of these ques-
tions, but we will defer answering the second to the next chapter.
There is evidence that Chinese marks wh-interrogatives as a clause type, in
other words that Chinese has CWh in the terms introduced in Chapter 5. First, in
main clauses, a special sentence-final particle can appear to mark the clause as
a wh-question; this is ne in (31b). We could treat ne as occupying a final Cwh, as
shown in (32):
(32) CP

C′

TP C
ne
NP T′
ni
T VP
[Pres]
Adv VP
weishenme
Neg VP
bu
V′

qu

This analysis implies that Chinese C follows TP in C′, like its Japanese counter-
part, as we saw in Section 7.4; we mentioned at the end of the previous section
that Chinese has disharmonic word order, though, so this does not imply that
Chinese is fully head-final the way Japanese is. This particle, which is optional,
contrasts with the yes/no particle ma:
(33) a. Ni shuo Zhongwen.
You speak Chinese.
‘You speak Chinese.’

 b. Ni shuo Zhongwen ma?

You speak Chinese   Q
‘Do you speak Chinese?’

So we could say that, for root clauses, ne optionally realises CWh and ma realises
CQ. This is evidence that C is present, and that it can function, as in English, to
type a clause as an interrogative of one kind or another.
Second, there is evidence from how clauses containing an unmoved (or in-situ)
wh-phrase are interpreted which shows that CWh is present. We saw in Chapter 5

IanRoberts_9781316519493_C007.indd 172 9/6/2022 1:55:27 PM

 7.6 The Wh-Movement Parameter 173

that in English verbs like wonder require an indirect-question complement, i.e.

they obligatorily c-select CQ or CWh. Furthermore, verbs like think do not allow
such a clausal complement, instead obligatorily selecting declaratives. There is
also a third class of verbs, exemplified by remember, which optionally take an
interrogative complement. In English, these differences can be clearly seen from
the possibilities of placing a wh-phrase which originates inside the complement
clause in the Specifier of the CP selected by the verb:
(34) a. We wonder what you bought.
?*What did you wonder (that) I bought?

 b. *I think what you bought.

What did you think (that) I bought?

c. I remember what you bought.

What did you remember (that) I bought?

The possible positions for the wh-phrase reflect the c-selection properties of
each verb class, i.e. whether the embedded C is Cwh, Q. Furthermore, since the
presence of a wh-phrase implies an accessible SpecCwh′, where the selected C is
not CWh, as obligatorily in the complement of think and optionally in the com-
plement of remember, the root C is Cwh. Hence the scope of the interrogative
(whether the main or embedded clause counts as a question, i.e. whether the
main or the embedded clause has Cwh) is indicated by wh-movement.
The striking fact is that the corresponding sentences in Chinese have exactly
the same range of possible and impossible interpretations, despite the absence
of wh-movement:
(35) a. Zhangsan xiang-zhidao [CP Lisi mai-le shenme ].
Zhangsan wonder    Lisi buy-perf what
‘Zhangsan wonders what Lisi bought.’

  b. Zhangsan yiwei [CP Lisi mai-le shenme ] ?

Zhangsan think Lisi buy- perf what
‘What does Zhangsan think that Lisi bought?’

c. Zhangsan jide   [CP Lisi mai-le   shenme ] ?

Zhangsan remember Lisi buy- perf what
‘Zhangsan remembers what Lisi bought.’ Or
‘What did Zhangsan remember that Lisi bought?’

Clearly, the wh-phrase shenme is in the same position in all three cases (the
direct object position of the subordinate clause), but the interpretations vary as
a function of the main verb in exactly the same way as in English. This shows
us that Chinese has Cwh. The difference between Chinese and English does
not lie in the interpretation or clause-typing of wh-questions, or in the classes
of verbs that c-select Cwh optionally or obligatorily, but much more narrowly
in whether a Cwh triggers movement to its Specifier. The difference, then, is

IanRoberts_9781316519493_C007.indd 173 9/6/2022 1:55:27 PM

 174 7 syntax beyond english

purely syntactic, having to do with whether a movement operation takes place

in a given context or not (in this respect the wh-parameter is hardly different
in principle from the V-to-T parameter we saw in Section 7.3). Of course, to
all appearances, the lack of wh-movement means that there is no variable for
the wh-quantifier to bind in examples like (33), raising the question of how
wh-questions are interpreted in languages like Chinese. As already mentioned,
we will revisit this question in the next chapter.
In this section, we have introduced a further parameter, the wh-movement
parameter. Although very simple in itself, this parameter has interesting impli-
cations for clause-typing and for the interpretation of wh-expressions. Chinese
is a standard example of a language without wh-movement, although it is by no
means the only one. In fact, the data from WALS cited at the beginning of this
section suggest that this may be the majority pattern.

7.7 Conclusion
In this chapter, we have taken a first look at how the theory of syntax
developed in the earlier chapters can be made into a cross-linguistically general
theory. This is of course the overall goal of our enterprise, as we want ultimately
to construct an account of the human language faculty (or at least the central
syntactic part of it). Our theory must therefore have universal scope. Although
we may still be some way off that ambitious goal, we can see the general shape
of the approach.
The central idea is that there are universal mechanisms and principles:
X′-theory, the binding principles and movement as an upward copy/paste oper-
ation. These universal features of the language faculty are subject to parametric
variation. Here we saw two types of parameters: one concerning the presence
or absence of particular movement rules (V-to-T, wh-movement) and the Head
Parameter (which may hold in maximally general form as in (17) or, for dishar-
monic languages like German, in a more attenuated form). What the full range
of possible parameters might be is a matter for ongoing research which we will
come back to in the later volumes, particularly Volume III. We can ask one sim-
ple question here though: if there are head-movement parameters (e.g. V-to-T)
and wh-movement parameters (see Section 7.6), then are there NP-movement
parameters? The only case of NP-movement we have seen here is the passive
(see Section 5.2). So even our highly limited look at parameters leads us to ask
whether there are languages without passives, and the answer is positive: WALS
(Map 107A; https://wals.info/feature/107A#2/18.0/148.9) cites 211 languages
lacking passives as opposed to 162 in which passives are found. Maybe, then,
the answer to our question is positive: there is an NP-movement parameter of
some kind.
What we have also seen here, particularly in our discussion of German in
Section 7.5, is that different languages can combine the same (or almost the

IanRoberts_9781316519493_C007.indd 174 9/6/2022 1:55:27 PM

 7.7 Conclusion 175

Table 7.2 Summary of languages and parameters discussed in this chapter

English French German Chinese Japanese

V-to-T No Yes Yes No No?

Head Parameter Initial Initial Mixed Mixed Final
Wh-movement Yes Yes Yes No No

same) parameters differently and quite distinct grammatical systems emerge.

This is a very important and attractive aspect of the parametric approach, which
can obviously be taken significantly further than we have done here; again, this
approach to cross-linguistic variation is the object of much ongoing research and
is yielding some very interesting results.
As a final point on cross-linguistic variation (for now), let us look at the three
parameters we have introduced and exemplified here in relation to the five languages
we have taken examples from. The parameters are V-to-T movement, the Head
Parameter and wh-movement; the languages are English, French, German, Chinese
and Japanese. Table 7.2 gives the value of each parameter for each language:
We have seen the evidence for almost all of these parameter settings in the
foregoing. Of course ‘mixed’ is not really a setting of the Head Parameter, rather
an indication that the language is disharmonic in that some heads precede and
some follow their complements; we saw some of the evidence for the mixed
nature of German head-complement orders in Section 7.5, but a full descrip-
tion of Chinese would take us too far afield here (it should also be noted that
Japanese has some head-initial order, and so Table 7.2 slightly oversimplifies
things). The absence of V-to-T in Chinese is indicated by the fact that almost
all adverbs, as well as clausal negation, precede the verb. We can see this for
negation in (31b), repeated here:
(31b) Ni weishenme bu qu ne?
You why    not go wh
‘Why aren't you going?’

The negator bu is probably located in between T and VP in Chinese, as is English

not and French pas, and the verb always follows it, as qu (‘go’) does here.
Compare this with French, where the finite verb precedes pas (we give a declar-
ative example to avoid the complication of T-to-C movement in interrogatives):
(36) Vous n’allez pas à Paris.
You neg-go not to Paris
‘You aren’t going to Paris.’

The status of V-to-T movement in Japanese is debated, and so I will leave it

aside here.
The only other cell of Table 7.2 we have not illustrated concerns the lack of
wh-movement in Japanese. This is a well-known property of Japanese, illus-
trated by examples like (37):

IanRoberts_9781316519493_C007.indd 175 9/6/2022 1:55:27 PM

 176 7 syntax beyond english

(37) Boku-wa [CP [TP John-ga  nani-o katta] ka]  shiritai.

I-top     John-nom what-acc bought Q    want-to-know
‘I want to know what John bought.’

In Table 7.2, we see that each language has its own set of parameter settings; this
reinforces the idea that each language takes essentially the same set of building
blocks for a grammar but puts them together in its own distinct way. This is the
key to understanding the unity in diversity that we find when we try to reconcile
cross-linguistic diversity with the idea that there is a Universal Grammar. If we can
multiply Table 7.2 by many orders of magnitude, adding thousands of languages
and hundreds of parameters, then we may be able to get an inductive idea of the
true diversity of the world’s languages, whilst at the same time sticking to a clear
universalist perspective. That is one the goals of parametric comparative syntax.
We have been concentrating on the details of the theory of syntax in the last few
chapters. Now it is time to return to a wider perspective and try to see how everything
we’ve seen (constituent structure, X′-theory, movement, binding, parameters) fits
together and fits with the overall architecture of the language faculty, including
phonology and semantics. That is the goal of the next and final chapter.

Exercises

1. We mentioned in the text that finite verbs precede the negator pas
in French. What do we conclude from the following infinitives in
French?
(i) a. *Ne sembler pas content est une condition pour écrire
des romans.
‘To seem not happy is a condition for writing novels.’
b. *Ne posséder pas de voiture en banlieue rend la vie
difficile.
‘To possess not a car in the suburbs makes life difficult.’
c. Ne pas sembler content est une condition pour écrire des
romans.
‘To not seem happy is a condition for writing novels.’
d. Ne pas posséder de voiture en banlieue rend la vie
difficile.
‘To not possess not a car in the suburbs makes life
difficult.’
Now consider these examples:
(ii) a. N’être pas content est une condition pour écrire des
romans.
‘To not be happy is a condition for writing novels.’

IanRoberts_9781316519493_C007.indd 176 9/6/2022 1:55:27 PM

 7.7 Conclusion 177

b. N’avoir pas de voiture en banlieue rend la vie difficile.

‘To not have a car in the suburbs makes life difficult.’
What is the generalisation regarding V-to-T movement of
infinitives in French?
2. Here is some Danish data (from Vikner 1995):1
(i) a. Jeg tror  at Johan ikke købte bogen.
I believe that John not bought book-the
‘I believe that John didn’t buy the book.’
   b. Hvorfor købte Johan ikke bogen?
Why bought John not book-the
‘Why didn’t John buy the book?’
  c. Derfor har Peter uden tvivl ikke læst den.
Therefore has Peter without doubt not read it.
“Therefore Peter has without doubt not read it.”
These examples suffice to situate Danish in Table 7.2. Show how
Danish fits in. Now add a further parameter to Table 7.2: verb sec-
ond. State the values of this parameter for the languages in Table
7.2. The following examples, along with (ic), suffice to tell you the
value of this parameter in Danish:
(ii) a. Denne film har børnene set.
This film have children-the seen
‘The children have seen this film.’
b. I går så børnene filmen.
Yesterday saw children-the film-the
‘Yesterday the children saw the film.’
  c. Børnene     så   filmen.
Children-the saw   film-the
‘The children saw the film.’
3. Now consider Icelandic. The examples in (iii) (also from Vikner
1995) correspond to the Danish ones in (ii) and clearly show that
Icelandic is V2:
(iii) a. Þessi mynd hefur börnin séð.
This film have children-the seen
‘The children have seen this film.’
   b. Í gær    sáu börnin    myndina.
Yesterday saw children-the film-the
‘Yesterday the children saw the film.’
   c. Börnin     sáu   myndina.
Children-the saw film-the
‘The children saw the film.’

1
S. Vikner (1995), Verb Movement and Expletive Subjects in the Germanic languages, Oxford: Oxford
University Press.

IanRoberts_9781316519493_C007.indd 177 9/6/2022 1:55:27 PM

 178 7 syntax beyond english

Now consider the Icelandic translation of (ia):

(iv) Ég held  að   Johan keypti ekki bókina.
I believe that John   bought not book-the
‘I believe that John didn’t buy the book.’
You should now be able to add Icelandic to your expanded Table 7.2.

For Further Discussion

At the end of Chapter 1, we raised the following questions about
I-language (see (57) of Chapter 1):
(i)  a. how does a given I-language relate to UG?
b. how does adult I-language result from acquisition?
c. how is I-language related to general cognition and the third factors?
Consider how parameters of the kind illustrated in Table 7.2 may provide
answers to these questions.

Further Reading
Adger, D. 2003. Core Syntax: A Minimalist Approach. Oxford: Oxford University
Press.
Haegeman, L. & J. Guéron. 1999. English Grammar: A Generative Perspective.
Oxford: Blackwell, Chapter 6.
Larson, R. 2010. Grammar as Science. Cambridge, MA: MIT Press, Unit 28.
Roberts, I. 1997. Comparative Syntax. London: Arnold, Chapter 5.
You should also take a look at the website of the World Atlas of Language Structures:
www.wals.info.

IanRoberts_9781316519493_C007.indd 178 9/6/2022 1:55:27 PM

 8 The Architecture of Grammar

8.1 Introduction and Recap

Starting from Chapter 2, where our real investigation of the theory
of syntax began, we have looked at constituent structure (Chapter 2), PS-rules
(Chapter 3), X′-theory (Chapter 4), movement (Chapter 5), binding (Chapter 6)
and parameters (Chapter 7). In this final chapter of this volume, the aim is to see
how all this fits together and, returning to the broader perspective of the discus-
sion in the Introduction, to consider these components of the theory of syntax
in relation to the overall architecture of grammar. Ultimately, this should be a
model of I-language, Universal Grammar.
We will begin by describing a particular overall model of the grammar, which
has five components and a fairly clear specification of how those components
interact. We will then look at each component in turn. To some extent, we will
revisit topics we have already seen in earlier chapters, but often from a slightly
different perspective.

8.2 A Model of Grammar

The organisation of the grammar we will assume has the form in (1):

(1) LEXICON

D-STRUCTURE

S-STRUCTURE

PHONOLOGICAL LOGICAL FORM

FORM

Here I briefly describe each component in turn, before going on to a more

detailed discussion of each one in the sections to follow.
The lexicon is a list of words and morphemes, a kind of mental dictionary
(although of course the entries aren’t alphabetised and definitions may be
given in terms of concepts rather than other words). A native speaker of a

179

IanRoberts_9781316519493_C008.indd 179 05-09-2022 17:44:19

 180 8 the architecture of grammar

given I-language has, stored in long-term memory but readily accessible at a

millisecond’s notice, a collection of lexical items which include idiosyncratic
information, i.e. information that cannot be predicted by any linguistic rule.
Estimates as to the size of the average adult’s vocabulary vary, but they are
invariably in the tens of thousands. An important aspect of the lexicon for syntax
is that it contains functional elements such as auxiliaries, complementisers, etc.
The lexicon is important for what we have seen in the foregoing chapters in that
it interacts with X′-theory in specifying the syntactic categories of lexical items,
with movement theory in specifying the presence and nature of movement-
triggering features such as Wh, Q, etc., and with binding theory in specifying
which elements are anaphors, pronouns and R-expressions (i.e. which lexical
items are subject to which binding principles).
D-structure, or deep structure, is a syntactic representation built by
PS-rules in conformity with X′-theory. No movement rules apply in the
derivation of D-structure, however. This is the level where thematic roles
such as Agent, Patient, etc. (which we briefly saw in Section 5.2) are
structurally represented.
S-structure, or surface structure, is a further syntactic representation derived
from D-structure by movement rules. After S-structure, the derivation divides
into two distinct parts: Phonological Form (PF) and Logical Form (LF). The
importance of this division will become apparent below. The respective roles of
PF and LF are apparent from their names. PF is the level of representation which
converts a syntactic representation (a tree or labelled bracketing) into a rep-
resentation which can be input to phonological and, ultimately, phonetic rules
and representations. LF is the level of representation which converts the syntac-
tic representation into a representation which can be the input to the semantic
representation which, in standard theories of semantics, is capable of bearing
truth conditions, i.e. of being evaluated as true or false. Collectively, we refer
to PF and LF as the interface levels, as they connect syntax to aspects of the
grammar which are external to syntax, i.e. phonology and semantics respec-
tively. The split is required because phonological features cannot be interpreted
at the semantic level and non-phonological features cannot be interpreted by the
phonological level.
We now look at each component in more detail. It should be borne in mind
throughout that, aside from the lexicon, each component represents a stage of
the derivation of a sentence; since the derivation splits after S-structure, the der-
ivation has two outputs: a PF one and an LF one.

8.3 The Lexicon

In order to see the role of the lexicon, consider the following
examples:

IanRoberts_9781316519493_C008.indd 180 05-09-2022 17:44:19

 8.3 The Lexicon 181

(2) a. John dined at 8.

b. *John devoured at 8.
c. *John dined the sandwich at 8.
d. John devoured the sandwich at 8.

What accounts for the grammaticality contrasts here? In (2a,b) we have a V′

containing just V (the adjunct at 8 is adjoined to V′ or T′, which one is immate-
rial here). In (2c,d) V′ contains V along with a complement NP. Either structure
is allowed by the X′-template, and so in this sense the examples are all well-
formed. What we can see from (2), though, is that dine is intransitive, i.e. it does
not allow an NP complement, and devour is transitive, i.e. it requires an NP
complement. Both dine and devour entail eating, each one adding a particular
specification as to the manner of eating. The more generic verb eat is optionally
transitive:
(3) a. John ate at 8.
b. John ate the sandwich at 8.

All of this indicates that it is difficult to predict the precise syntactic require-
ments a verb may have (e.g. NP-complement or not) from its meaning. This
idiosyncratic information is stored in the lexicon, which specifies (among other
things) categorial selection (c-selection) and semantic selection (s-selection)
properties of lexical items. Together, these selection frames specify the argu-
ment structure for each verb, i.e. who’s doing what to whom. For the three verbs
we have been looking at here, then, the c-selection and s-selection properties are
specified as shown in (4):
(4) a. devour (V): __ NP <Agent, Theme>
e.g. John devoured the pie at 5 pm.

b. dine (V): __ <Agent>

e.g. Larry always dines at 5 pm.

c. eat (V): __ (NP) <Agent, (Theme)>

e.g. John ate (the pie) at 5 pm.

These entries give information as to the word’s category (V in each case here),
the c-selection frame, i.e. the syntactic material that must appear in X′ as its
sister, and the s-selection frame, i.e. the thematic roles the word’s arguments
bear. Optional material is in brackets in both the c-selection and the s-selection
specification.
The c-selection frame only specifies the arguments that appear in V′:
the internal arguments. Subjects are not c-selected, since they occupy a
Specifier position, which is always available independently of the specific
properties of any lexical item. General information of this kind, which is not
specific to lexical items, is not specified in the lexicon. Instead, the PS-rules
always generate subject positions. Verbs do differ, however, in the thematic

IanRoberts_9781316519493_C008.indd 181 05-09-2022 17:44:19

 182 8 the architecture of grammar

roles they assign to subjects. Some verbs which describe mental states or
emotions, known as psych-verbs, assign an Experiencer thematic role to
their subjects, e.g. love in (5):
(5) John loves Mary.

If John loves Mary, John doesn’t have to actually do anything (although of

course he might do absolutely anything); the verb simply describes his emo-
tional state in relation to Mary. So the lexical entry for love might look like (6):
(6) love (V): __ NP <Experiencer, Theme>

By convention, the leftmost thematic role in the s-selection frame is assigned

to the subject, which we can also think of as the external argument, i.e. the
argument which occupies a non-c-selected Specifier position outside the core
predicative category consisting of V and its c-selected complements (we will
see in Volume II that this position does not have to be SpecT′ but could be a
lower Specifier position, perhaps in VP). So the subject of the Verbs in (4a–c)
bears the Agent role, while in (6) specifies that the subject of love receives the
Experiencer role.
There are also verbs that assign no subject/external thematic role at all. This
is true of verbs like seem, for example. With these verbs, the subject position
is nonetheless filled by a dummy, or expletive, element it. This element has no
semantic interpretation, bears no thematic role and appears to be present purely
in order to fill the subject position:
(7) It seems that the world is round.

This shows that the subject position is not c-selected; it appears even where the
verb imposes no lexical-semantic requirement for its presence at all, i.e. it is
not s-selected either. This is a fundamental difference between subjects/external
arguments and internal arguments: internal arguments appear as a function of
c-selection and s-selection properties of lexical items (i.e. heads in X′ terms),
while subject positions always appear and they may, but do not have to, house
external arguments s-selected by a lexical head. If some verbs do not s-select an
external argument, the convention that the leftmost s-selected thematic role is
assigned to the external argument will not suffice. For most verbs, two conven-
tions will suffice:
(8) a. Agents are always lexically external arguments.
b. Themes are never lexically external arguments.

These statements capture redundancies in lexical entries. For psych-verbs, the

convention that the leftmost thematic role is assigned to the external argument
is still valid.
In Chapter 4 we saw the following examples of c-selection frames for various
English verbs (changing the original S′ to CP in the case of say):

IanRoberts_9781316519493_C008.indd 182 05-09-2022 17:44:19

 8.3 The Lexicon 183

(9) a. watch (V) ___ NP

b. rely (V) ___ PP
c. put (V) ___ NP PP
d. say (V) ___ CP

To these we can now add the s-selection frames for each verb as in (9′):
(9′) a. watch (V) ___ NP <Agent, Theme>
b. rely (V) ___ PP <Agent, Theme>
c. put (V) ___ NP PP <Agent, Theme, Location>
d. say (V) ___CP <Agent, Theme>

Here we see further divergences between c-selection and s-selection, in particu-

lar in that the Theme role can be borne by a direct-object NP, as with watch and
put, and by a CP in the case of say.
Other semantic features are associated with thematic roles and so are
also s-selected. Agents and Experiencers are always animate, while Themes
may be animate or inanimate. Hence we can observe contrasts such as the
following:
(10) a. Sincerity delights John.
b. ?*Sincerity relies on Mary.
c. Mary relies on sincerity.

Examples like (10b) may be found in poetry or perhaps other registers in which
aspects of word meanings are ‘stretched’ for literary or aesthetic effect, and so it
may not be strictly correct to treat them as ungrammatical. We can in fact regard
them as syntactically well-formed in that they can be generated by X′-theory
(along with whatever movement rules might be necessary). The anomalous
nature of such examples, at least in most registers, is due to a violation of an
s-selection requirement. On the other hand, c-selection violations, as in (11),
give rise to genuine ungrammaticality, i.e. syntactic ill-formedness:
(11) a. *Sincerity delights that the world is round.
b. *Mary relies sincerity.

The contrast in (12), which we mentioned in Chapter 1, also reflect s-selection:

(12) a. Colourless green ideas sleep furiously.
b. Revolutionary new ideas spread quickly.

As we pointed in Chapter 1, (12a) is grammatical in the sense that it is generated

by the syntactic rules. However, we can see that the s-selection property of sleep
that it requires an animate external argument (whether this is best described as
an Agent is another matter) is violated here, among other things.
Verbs have the richest argument structure of all lexical categories, but the
other categories also impose c-selection and s-selection requirements, as illus-
trated in (13):

IanRoberts_9781316519493_C008.indd 183 05-09-2022 17:44:19

 184 8 the architecture of grammar

(13) a. picture (N) ___ (PP) <Theme>

b. angry (A) ___ (PP) <Experiencer, Theme>
c. beyond (P) ___ NP <Location>

It is debatable whether Prepositions ever have external arguments; clearly the

Location role is borne by the sole NP argument of beyond in (13c). The internal
arguments of Nouns and Adjectives are typically optional. Moreover, some lexical
items may lack argument structure altogether: a simple common Noun like dog
or cat does not appear to have either an external or an internal argument. On the
other hand, it is unclear whether there are any Verbs which entirely lack argument
structure. Meteorological Verbs such as rain and snow may be a case in point, as
they can appear with an apparently expletive subject it and no internal argument:
(14) It’s raining.

This is debatable however; rain can take an internal argument as in (15), although
this might be a fixed, idiomatic expression:
(15) It’s raining cats and dogs.

We will leave the very interesting question of the argument structure of meteor-
ological Verbs aside.
Some reason to think that Nouns do not have to have argument structure while
Verbs do comes from a reconsideration the fish sentences from Chapter 1:
(16) FishN fishV.

Here the Noun fish acts like a typical common Noun and appears to have no
argument structure. The Verb fish, on the other hand, presumably has the argu-
ment structure given in (17):
(17) fish (V) __ (NP) <Agent, (Theme)>

So the Verb fish has an Agent external argument and an optional Theme internal
argument, realised as an NP. The latter option is taken in the three-fish example:
(18) Fish fish fish.

The Noun fish, on the other hand, appears to have no argument structure at all.
Perhaps here we are seeing a fundamental difference between Verbs and Nouns.
As we mentioned above, functional categories are also listed in the lexicon.
We can give the lexical entries for some English auxiliaries as follows:
(19) a. can (T): __ VP
b. have (Perf): __ V-enP
b. be (Prog): __ V-ingP

These lexical entries specify the c-selection properties of each auxiliary, includ-
ing the information in (19b), for example, that perfect have requires the head
of the selected VP to be a perfect participle, here roughly indicated as V-en.
Similarly, progressive be requires the selected VP to be a progressive participle
in -ing. (Of course, there are other kinds of have, e.g. the possessive, and other

IanRoberts_9781316519493_C008.indd 184 05-09-2022 17:44:19

 8.3 The Lexicon 185

kinds of be, e.g. the copula; (19) implies that these are actually distinct lexical
items with their own lexical entries.) Modals such as can select a bare V. It is
fairly straightforward to state the c-selection requirements of these auxiliaries,
but what about s-selection? The auxiliaries appear to have semantic content, but
it is not clear that they are associated with thematic roles, at least not thematic
roles of the Agent, Experiencer, Theme (etc.) type. This is a major open question
for our understanding of auxiliaries and perhaps other functional categories.
We also stated that functional features may have their own lexical entries. For
example [Present] and [Past] may have entries like the following:
(20) a. [Present] (T): __ VP
b. [Past] (T): __ VP

These lexical entries simply specify that these features are realised on T and
c-select for VP. As with the auxiliaries in (19), they clearly have semantic con-
tent, but the question of their s-selection properties is a difficult one.
Up to now we have concentrated on the categorial, c-selection and s-selection
specifications in lexical entries. But, as with conventional dictionaries, lexical
entries must also provide phonological and semantic information. For entries like
(6), (13a) and (19a), we can take the phonological specification to simply be a
phonemic representation of the word, e.g. /lʌv/, /pɪkʧɘ(r)/ and /kæn/. But for (20)
we need to specify that these lexical items are realised as affixes. For [Past], for
example, we could perhaps try to specify something like /+d/, with the +-sign
indicating that it is obligatorily realised as an affix. This is adequate for the major-
ity of regular Verbs, it does not take into account the existence of irregular past-
tense forms like sang, went, etc. It is impossible for a lexical item such as [Past],
a member of the category T, to change its phonological form as a function of the
particular lexical item in V. The fact that sing, go and about 150 other English
Verbs have irregular past tenses can only be listed in the lexicon. Therefore, these
Verbs have to be lexically specified for their past-tense forms, roughly as in (21)
for sing:
(21) sing (V): __ (NP, PP) <Agent, Theme, Goal>. PF: /sɪŋ ~ sæŋ[Past]/.

For regular verbs, we can maintain that the PF form of [Past] is /d/. This default
(or elsewhere) realisation of [Past] is blocked by the lexically pre-specified
form for sing. The idea that a more specific realisation of a formative, such
as the particular realisation of [Past] when combined with sing, blocks the
default/elsewhere realisation, is known as disjunctive ordering, and plays a
pervasive role in morphology and phonology. All of this of course presupposes
that V and T combine when T is [Past]. This must be specified in some way
in the lexical entry, perhaps by means of a simple diacritic like ‘+’, as part
of the PF-representation. Since English lacks V-to-T movement, as we saw
in Section 7.3, it must have T-to-V movement, a rare case of a ‘downward’
movement rule (recall that we concluded in Section 6.7 that it is not clear that
head-movement must result in a binding relation, hence downward head-move-
ment, unlike wh- or NP-movement, may be possible). The downward-movement

IanRoberts_9781316519493_C008.indd 185 05-09-2022 17:44:19

 186 8 the architecture of grammar

option applies to lexical items specified as ‘+’ where upward movement is ruled
out by the negative setting of the V-to-T parameter.
Lexical entries must also give a specification of the semantics of lexical items.
It is difficult to illustrate this properly here without providing an introduction
to semantics. For many verbs, a kind of ‘action schema’ related to the thematic
roles can be given. Very approximate ‘action schemata’ for three common
English Verbs are given in (22):
(22) a. say: ‘the Agent causes a propositional Theme to exist by speaking’
b. eat: ‘the Agent consumes the Theme by oral ingestion’
c. give: ‘the Agent causes the Theme to become located at/possessed by the Goal.

These are of course just approximations to the meanings of these Verbs.

Obviously, there’s a lot more to say, but it would take us too far afield here.
Many functional items have logical or quasi-logical semantics. The quantifier
every, for example, which we saw in Section 6.7, has a semantics which approx-
imates closely to that of the universal quantifier of standard logical systems,
usually written with the symbol ". There is also the question of real-world,
encyclopedic knowledge. For example, the lexical entry for the proper noun
Paris may specify that the word refers to the capital of France. Obviously, there
is much much more to say about all of this.
Here, as sample lexical entries, are the full entries for the Noun fish and the
Verb fish:
(23) a. fish (N): __. PF: /fɪʃ/. LF: λx [fish (x)]
b. fish (V): __ (NP). <Agent, (Theme)>. PF: /fɪʃ/. LF: HUNT(Agent) [FISH
(Theme)]

(The formula in (23a) says ‘the set of individuals such that they are fish’; a
similar, much more complex formula, can be given for the LF in (23b), but this
is not necessary here. The semantics must also somehow specify the intrinsic
content of ‘fish’ as an aquatic animal or whatever, a matter at least partially
bearing on world knowledge which we can safely leave aside here.) What we see
here is that the Verb fish has a more complex lexical entry than the correspond-
ing Noun, and that the Verb’s lexical entry makes reference to the Noun’s: the
occurrence of the term FISH in the LF for the Verb fish must be connected to that
of the Noun in order to avoid unnecessary redundancy in specifying the meaning
of the concept ‘fish’ (presumably λx [fish (x)]) twice over. As we already hinted,
an intriguing question concerns the extent to which the greater complexity of the
lexical entry of the Verb is due to the fact that it is a Verb. Fully exploring this
question would take us too far afield here.
As a last point, here are the lexical entries for the Italian counterparts to the
English Noun and Verb fish, which we saw in Chapter 1:
(24) a. pesce (N; Gender: Masc): __. PF: /’pεʃe/. LF: λx
[fish (x)]
b. pescare (V; Conj: 1): __ (NP). <Agent, (Theme)>. PF: /pεs’kare/. LF:
HUNT(Agent) [FISH (Theme)]

IanRoberts_9781316519493_C008.indd 186 05-09-2022 17:44:19

 8.4 D-structure 187

Here, the phonological information is obviously different compared to English.

The semantic information, however, is identical to English. The syntactic infor-
mation is similar, but two extra pieces of lexical information need to be added
in Italian: for Nouns, the grammatical gender must be specified (in (24a), this is
‘Gender: Masc’), and for Verbs, information about conjugation class (‘Conj: 1’
in (24b)). This information corresponds exactly to what we suggested in Chapter
1 should not be specified by silent functional categories in English (unlike, for
example, Agreement information; see the discussion of examples (14–21) of
Chapter 1). In fact, we already pointed out in Chapter 1 that the syntax of English
and Italian are ‘(near-)uniform and the semantics entirely uniform’; comparison
of the lexical entries in (23) and (24) shows this to be the case.
In this section we have taken a look at aspects of lexical entries. The most
important notions are those of c-selection and s-selection, argument structure,
including thematic role and the distinction between external and internal argu-
ments. We have also looked briefly, and rather inconclusively, at the lexical
entries of functional categories. Here the issues connected to the representation
of irregular forms and default/elsewhere realisations are important, as is the PF
specification of certain formatives as inherently affixal. We also took a very
sketchy look at the semantic part of lexical entries. Without bringing in a great
deal more semantic theory, it is difficult to give a clearer picture of this aspect of
lexical entries, but we saw that verbs, at least, might be represented by ‘action
schemata’ of the kind in (22). Finally, we gave the lexical entries for the Noun
and the Verb fish in (23); these suggest some intriguing connections between the
semantics and selectional properties of Nouns and Verbs.
We will leave our discussion of lexical entries here, with many matters left
unresolved. What is important now is to see how this static, dictionary-like list
is connected to the generative rules of syntax.

8.4 D-structure
As we said in Section 8.2, D-structure is a syntactic representation
built by PS-rules in conformity with X′-theory. No movement rules apply in the
derivation of D-structure, however. This is the level where the c- and s-selection
requirements of lexical items are directly represented structurally. But how do
the Lexicon and Syntax ‘communicate’? How is the information regarding c-
and s-selection of the kind we saw in the previous section transmitted from the
lexical entries to the syntactic structures?
This is achieved by the Projection Principle, which can be stated as follows:
(25) All and only featural information that is stored in the Lexicon must be
reflected in the Syntax, at all levels of representation.

Looking at (23a), for example, we can take the effect of the Projection
Principle to mean that the Noun fish can only undergo lexical insertion into
an N-node. Since its c-selection and s-selection frames are null, this Noun

IanRoberts_9781316519493_C008.indd 187 05-09-2022 17:44:20

 188 8 the architecture of grammar

cannot have a complement. This accounts for the ungrammaticality of NPs

such as (26):
(26) a. *a fish of physics (with large fins)
b. *every fish about/on Ancient Greece

On the other hand, Nouns like student are specified for an optional PP com-
plement headed by of, with an associated Theme thematic role, and Nouns like
book are specified for an optional PP complement headed by about or on:
(27) a. a student (of physics)
b. every book (on/about Ancient Greece)

Adjuncts, such as with large fins in (26a), and Specifiers, e.g. the determiners
a and every here, are not c-selected and hence the Projection Principle neither
requires nor forbids their appearance in syntax. This is not to imply that there
are no constraints on adjuncts or Specifiers: a student of physics with large fins
conjures up a possibly anomalous image, but this is a matter of world knowledge
rather than grammar or syntax (and this, as we saw with (10b), may depend on
register in that such an NP would not necessarily be anomalous in a science-fic-
tion context, for example). One important lexical property of Nouns does relate
to what kind of Specifier they are compatible with: whether they are count or
mass. We introduced this distinction in Section 2.2, pointing out that mass nouns
do not have plurals. The Noun fish can function as both a mass and a count
noun, but this is hard to see with plural marking since, as we saw in Chapter 1,
the usual plural of fish (except in the sense of ‘species of fish’) has a null ending
and so looks the same as the singular; see (21) of Chapter 1. In English, mass
nouns can appear in the singular with no determiner, but count nouns cannot,
while count nouns can appear with the indefinite article but mass nouns cannot.
The Noun fish can appear in either context; compare the obligatorily mass Noun
water and the obligatory count Noun ant in (28):
(28) a. Fish/water is good for you.
b. A fish/an ant is on your plate.

So we see that mass or count nature of a Noun, which should also be stated in
its lexical entry (although we have not seen exactly how), interacts with what
can appear in its Specifier. This, too, is arguably a matter of semantics, since the
distinction has to do with countable instances of something as opposed to undif-
ferentiated quantities of something. As we mentioned in Section 2.2, many mass
nouns can be ‘coerced’ into count readings, including water if one is concerned
with distinguishing different types of water (cf. a fine water from the foothills of
the Alps), but fish cannot really be coerced into giving coherent readings where
its c-selection property is violated in examples like (26).
As we said in the previous section, many common Nouns like cat, dog and
fish lack external arguments. This is implicit in the lexical entry in (23a), since
no Agent or other potential external argument is specified.

IanRoberts_9781316519493_C008.indd 188 05-09-2022 17:44:20

 8.4 D-structure 189

The PF and LF specifications in (23a) only become relevant at the PF and LF

levels. However, we can think of them as ‘carried along’ with the lexical item
through the syntactic parts of the grammar. The symbols which make up these
specifications (e.g. phonemes like /f/, the λ-operator) are illegible to the syntax
and so they are ignored until the derivation reaches the level of representation
which can ‘read’ them. An analogy from cookery might help here. Pasta requires
boiling water in order to cook, but you can put pasta in water before the water
boils and nothing will happen to it. We could say that pasta can’t ‘read’ water
at less than 100°C, but water is water with the physical and chemical proper-
ties it has, below and above 100°C. Similarly, we can insert a lexical item at
D-structure complete with PF and LF symbols, but nothing happens to those
symbols until the derivation reaches the relevant level of representation while
the lexical item remains what it is throughout the syntactic derivation. In fact,
a strong interpretation of the Projection Principle guarantees this, in that it pre-
vents us inserting fish at D-structure and giving it the PF form /kæt/ and the LF
representation λx [dog (x)], possibilities we clearly need to rule out (this entails
treating the PF and LF specifications in the Lexicon as featural).
Turning now to the Verb fish whose lexical entry is given in (23b), we see
that it optionally c-selects an NP complement, corresponding to an optionally
s-selected internal Theme argument, and obligatorily s-selects an external Agent
argument. The Projection Principle therefore allows this Verb to appear in the
contexts in (29) and rules out those in (30):
(29) a. John fishes (in the pond).
b. Fish fish fish.
(30) a. *It fishes that the world is round.
b. *There fished.
c. *John fished a book to Mary.
d. *The economy fishes John.

Again, the occurrence of adjuncts is independent of the Projection Principle,

hence in the pond is optional in all these examples and does not affect their
grammaticality. As before, I will leave aside the question of the content of
SpecV′.
We can see further effects of s-selection in examples like (31). Inanimate nouns
cannot be Agents and abstract nouns cannot be the Themes of Verbs like fish:
(31) a. ??Philosophy fishes fish.
b. ??John fished sincerity.

As usual, metaphorical or poetic interpretations may be possible here, unlike

in (30). It does seem that an inanimate external argument is possible with the
interpretation that it bears the role of Instrument:
(32) The new Ocean-Depleter 1000 ships will fish (our seas) more efficiently
than ever.

IanRoberts_9781316519493_C008.indd 189 05-09-2022 17:44:20

 190 8 the architecture of grammar

This suggests that we may need to modify (23b) as regards the thematical roles
s-selected as external argument. Of course, there are various further possible
additions to be made to (23b), but what we have seen here suffices to illustrate
the role of the Projection Principle.
D-structure, then, is the syntactic level at which lexical items are inserted
into phrase markers generated by X′-theory in conformity with the Projection
Principle. This means that (29b) has the D-structure in (33), where the c- and
s-selection properties of all lexical items (including the functional item [Present],
see (20a)) are satisfied:
(33) TP

NP T′

D N′ T VP
Ø Present
N V′
fish
V NP
fish
D N′
Ø
N
fish

The Projection Principle also guarantees that the external argument [np [d ø ]
[n′ [n fish ]] is interpreted as the Agent and the structurally identical internal
argument is interpreted as the Theme in relation to the action schema in (23b) at
LF. It also ensures that the correct PF representations are associated with each
category, including null representations for D and T. We see that the two oper-
ative constraints at D-structure are X′-theory and the Projection Principle, with
the latter guaranteeing the connection between the lexicon and independently
generated phrase structure.

8.5 S-structure
In terms of the architecture of grammar shown in (1), the next stage
of the derivation is S-structure. As we said above S-structure is derived from
D-structure by movement rules. We introduced movement rules in Chapter 5.
There we saw three types of movement rules: wh-movement, head-movement
and NP-movement. Let us briefly look again at each of these, in the light of
what we have seen in this chapter regarding the lexicon, D-structure and the
Projection Principle.
Let us look first at wh-movement. The structure in (34), corresponding closely
to (30) of Chapter 5, is an example of an S-structure derived by wh-movement:

IanRoberts_9781316519493_C008.indd 190 05-09-2022 17:44:20

 8.5 S-structure 191

(34) CQ, WhP

NP CQ, Wh′
what
CQ, Wh TP

T NP T′
have you
T VP
(have )
V NP
catch (what )

Here we see that the wh-NP what moves from the object position, complement
of V, to SpecCQ,Wh′. The D-structure from which (34) is derived is (35) (here
head-movement of T to C has not taken place either, a point I return to directly):
(35)
CQ, WhP

CQ, Wh′

CQ, Wh TP

NP T′
you
T VP
have
V NP
catch what

As a transitive verb, catch has a lexical entry similar to that of fish in (23b) but
without the specification that the complement NP is optional:
(36) catch(V): __ NP. <Agent, Theme>. PF: /kæʧ/. LF: HUNT(Agent) [
CONCRETE-OBJECT (Theme)]

In (35), what satisfies the c- and s-selection properties that catch imposes on its
internal argument, and the external argument you is the s-selected Agent.
CQ, Wh c-selects finite TP, like any root C. As we saw in Chapter 5, it also has
the property of triggering wh-movement. So, in the derivation from D-structure
to S-structure, the wh-NP what moves to SpecCWh, Q′, as seen in (34). Since the
Projection Principle requires lexical properties to hold at all levels of the gram-
mar (as given in (1)), the c-selection property of catch that it has a complement
NP must be satisfied at S-structure. This is why there must be a trace or copy
of what in that position. We did not take a position on which of these it is in
Chapter 5, but we can now observe that positing a copy seems a simpler option
than positing a trace, at least at S-structure, since the latter involves deleting

IanRoberts_9781316519493_C008.indd 191 05-09-2022 17:44:21

 192 8 the architecture of grammar

the copy of what left by movement and inserting a trace in its place, while the
former may not affect the direct-object position at all. X′-theory generates the
landing site of wh-movement, SpecCQ, Wh′. The S-structure in (34) forms the
input to PF and LF, whose properties in relation to wh-movement we will look
at in the next two sections.
Comparing the S-structure in (34) with the D-structure in (35), we can also
see that head-movement from T to C takes place in the derivation of (34) from
(35). We know from yes/no questions, as already discussed several times in
earlier chapters, that CQ triggers T-to-C movement in main clauses whether the
wh-feature is present on C or not:
(37) a. Will Cambridge flood as a consequence of global warming?
b. Did Cambridge flood after the heavy rain last year?

We can also see that, where there is no other auxiliary present, do appears in T
in main-clause interrogatives and various other environments (e.g. negation and
VP-ellipsis). We can clearly assign will a lexical entry like the one given for can
in (19a), as a T-element c-selecting VP (leaving aside the difficult question of
the s-selection properties of modals, as we mentioned there). Then CQ will trig-
ger T-to-C movement. But in (34) and (35), T is occupied by the feature [Past].
In our discussion of the PF properties of [Past] in Section 8.3, we said that it
must combine with V by the ‘downward’ movement rule T-to-V. The operation
combining T with V is somehow prevented from applying in main-clause inter-
rogatives, and so the meaningless auxiliary did is inserted as the realisation of
[Past]. If T-to-C movement places T in a position too distant from V to allow
the PF T-to-V rule to take place, then this may be why do, in the form of did,
is inserted. This implies that at D-structure and S-structure, T is simply [Past],
with the default PF form /d/ as discussed in Section 8.3. (We will come back to
do-support in more detail in the next section and slightly revise what we have
said here.)
The Projection Principle, X′-theory, wh-movement, head-movement and the
sketch of do-support just given, along with the lexical entry for catch in (36),
together account for the paradigm in (38):
(38) a. What have you caught (what)?
b. You caught a fish.
c. *What have you caught a fish?
d. You caught WHAT?

In (38a), we have wh-movement from the complement position of V, T-to-C

movement triggered by CQ and copy-deletion. In (38b), we have T-to-V move-
ment (and note that catch must be specified for its irregular Past form caught
(/kɔ:t/), which we omitted in (36)). Sentence (38c) is ungrammatical because the
wh-phrase cannot be directly inserted into SpecC′, as there is no head to select
it there, and there is no c-selected position inside TP from which it can have
moved, as a fish occupies the object position c-selected by catch. In the echo
question in (38d), there is neither wh-movement nor T-to-C movement.

IanRoberts_9781316519493_C008.indd 192 05-09-2022 17:44:21

 8.5 S-structure 193

The third type of movement we have seen is passives, as in (39) (repeated

from Chapter 5):
(39) a. Oprah interviewed John. (active)
b. John was interviewed (by Oprah). (passive)

As we saw in Chapter 5, interview has the c-selection frame in (40), to which we

add the standard s-selection specification for a transitive verb:
(40) in=terview (V) ___ NP. <Agent, Theme>.

These selection requirements are clearly satisfied in (39a), where Oprah is the
Agent external argument and John the c-selected NP complement bearing the
Theme thematic role. In the passive (39b) we see the same NP John, with the
same thematic role of Theme, in SpecT′. We account for this by saying that this
NP moves from one position to the other. The relationship can be captured by
the trees in (41) (again repeated from Chapter 5):
(41) a. TP

NP T′

T VP

V NP
interviewed John

b. TP

NP T′
John
T VP
was
V NP
interviewed (John )

Here John moves to the subject position, which we defined as an A(rgument)-

position in Section 6.7. As we saw there, this means that there is an A-binding
relation between John and its copy in (41b). In some sense, then, this copy is
able to ‘count’ as an anaphor for binding theory purposes, if we want to maintain
that movement always entails binding of the copy; we will leave this important
and tricky question aside here.
The Projection Principle forces the copy of John to remain in the c-selected
complement position in (41b). But what of the s-selected external Agent argu-
ment in the passive? One clue comes from the optional by-phrase in (39b); here
Oprah clearly bears the s-selected external Agent thematic role. The by-phrase
is an adjunct and, like all adjuncts, is optional. Since there is no c-selection
requirement for the external argument, there is nothing that actually requires that
argument to be realised in subject position (SpecT′). Hence it may be free to be
realised in a by-phrase. However, the Projection Principle requires s-selection

IanRoberts_9781316519493_C008.indd 193 05-09-2022 17:44:22

 194 8 the architecture of grammar

for the external argument to be satisfied. In fact, there is evidence that this argu-
ment is present even in ‘short’ passives where the by-phrase is absent. A full
discussion of this evidence would take us too far afield here; perhaps the most
important observation is that in (39b) where the by-phrase is absent, the interpre-
tation of the sentence is ‘someone interviewed John’, where ‘someone’ clearly
functions as the s-selected external argument of interview. So in short passives
the external argument is interpreted as an arbitrary implicit argument, a silent
(or implicit) version of ‘someone’, referring to an arbitrary individual or indi-
viduals. The presence of this argument is presumably required by the Projection
Principle. The precise structural realisation of the arbitrary implicit external
argument is much debated; arguably the simplest approach would be to treat
it as a silent version of the by-phrase, as a silent adjunct which, unlike typical
adjuncts, is not optional because of the Projection Principle.
Here we have illustrated how S-structure is derived from D-structure by three
different kinds of movement rules. The Projection Principle requires lexically
specified c- and s-selection requirements to be satisfied at this level, as at all
others. The principal consequence of this is that copies must be present in the
positions vacated by movement; we also saw that the implicit external argument
of passives must be in some sense present.
Let us now look again at the architecture of the grammar in (1):
(1) LEXICON

D-STRUCTURE

S-STRUCTURE

PHONOLOGICAL LOGICAL FORM

FORM

Here we see that the derivation splits into two paths after S-structure, one path
leading to Phonological Form (PF) and the other to Logical Form (LF). As we
will see in more detail in the next section, PF ultimately converts the representa-
tion of the sentence into a phonological and a phonetic form; this is what we
hear. LF, on the other hand, converts the syntactic representation, which is a
purely formal object, into a representation which can be the input to semantic
interpretation (in standard semantic theory, this means a representation which
can be read as a proposition bearing truth conditions, capable of being eval-
uated as true or false); this is what we understand. It is very important to see
that the level of representation we hear (PF) and the level of representation we
understand (LF) are distinct and that they do not directly interact in that neither
is directly derived from the other. The ‘bridge’ between the two is constituted
by the earlier levels of representation D-structure and S-structure, which we can
collectively designate as the narrow syntax, and the Lexicon. The Projection

IanRoberts_9781316519493_C008.indd 194 05-09-2022 17:44:22

 8.6 Phonological Form (PF) 195

Principle plays a vital role too, in that, as we saw, it prevents PF /kæt/ from
corresponding to LF λx [ dog(x) ] (as long as PF and LF specifications in lexical
entries are taken to be featural).
The split in the derivation seen in (1) allows us to account for the ubiquitous
fact that we understand more than we hear; as we have repeatedly emphasised, a
great deal of syntax is covert, hence silent. In terms of the architecture of gram-
mar in (1), aspects of narrow syntax feed into LF representations but are inert
at PF. The opposite is also true: there are syntactic elements which we hear but
which play no role at LF, e.g. the dummy English auxiliary do. Additionally,
phonological and semantic aspects of lexical entries are inert in narrow syntax
but active at their respective interfaces, as we mentioned earlier. We will see
examples of PF/LF mismatches in the next two sections. These examples moti-
vate the split derivation of sentences entailed by the architecture in (1).

8.6 Phonological Form (PF)

As we have said, PF interprets the representation sent to it by the
narrow syntax (D-structure and S-structure) and converts it into phonological
form. In other words, it converts (42a) to something like (42b):
(42) a. TP

NP T′

D N′ T VP
Ø Present
N V′
fish
V NP
fish
D N′
Ø
N
fish
The PF in (42b) is seriously incomplete, in that only the segmental aspects of
the phonological representation are specified; at the very minimum the third
/fɪʃ/ receives greater accentual prominence than the preceding two (this is the
English Nuclear Stress Rule).
Clearly, large parts of the S-structure representation in (42a) have no coun-
terpart in (42b). None of the phrase structure determined by X′-theory appears
to have a PF counterpart, although this may be a little too hasty a conclusion,
as phrase structure information is required in order to determine the Nuclear
Stress, i.e. the greater accentual prominence of the third /fɪʃ/ just noted. Of the
lexical items in (42a), only the occurrences of the noun fish and the verb fish
have any PF realisation, the former twice over. The plural indefinite D and
Present Tense are null, as a matter of lexical specification. Different values of

IanRoberts_9781316519493_C008.indd 195 05-09-2022 17:44:23

 196 8 the architecture of grammar

D, e.g. singular indefinite a, and of T, e.g. Past /d/, do have PF realisations as

shown in (43):
(43) a. TP

NP T′

D N′ T VP
a +d
N V′
fish
V NP
fish
D N′
a
N
fish

Here the singular indefinite D a is straightforwardly realised as /ɘ/ (in unstressed

positions, where it normally occurs). The realisation of Past /d/ is more com-
plex, though. First, it must undergo T-to-V movement, attaching it to fish, giving
fish+d. Then there is a regular phonological assimilation rule which converts a
voiced obstruent to its voiceless counterpart immediately following a voiceless
obstruent in the same word: this rule converts /d/ to /t/ here. There is absolutely
no doubt that the latter operation is a productive phonological rule of English
(a variant of it converts the voiced obstruent marking plural, /z/, to its voiceless
counterpart /s/ in cats /kæts/ but not in dogs /dɔgz/) and as such belongs firmly
in the PF component. But what about the T-to-V operation? This appears to be a
case of head-movement, the opposite of the French and German V-to-T rule we
discussed in Sections 7.1 and 7.5.
In our discussion of the PF properties of [Past] in Sections 8.3 and 8.5, we said
that the T-to-V rule takes place in the derivation from S-structure to PF, while the
T-to-C rule triggered by CQ in root interrogatives, which gives to subject-auxil-
iary inversion, takes place in the derivation from D-structure to S-structure. Let
us now look more closely at this idea, in order to get an idea of some of aspects
of how PF might interact with narrow syntax.
In Section 6.7 we gave the following general definition of movement:
(44) Movement substitutes an element of type X into a c-commanding position
of type X, where X ranges over head, grammatical-function NP and wh.

We suggested there that movement creates binding configurations, since indices

are copied as part of the copy/paste operation, and explored the consequences of
this idea for wh-movement especially; we will return to this idea in the section
on LF to follow. We also observed that the notion of referential index is not so
natural in the context of head-movement.
Since PF is entirely independent of LF, and binding and coindexation are rela-
tions with impact at LF since they affect semantic interpretation, it is possible

IanRoberts_9781316519493_C008.indd 196 05-09-2022 17:44:23

 8.6 Phonological Form (PF) 197

that downward movement is allowed in PF since there is no associated binding

requirement. If so, then T-to-V movement must be a PF operation: if it applied
in narrow syntax, it would be fed to LF where the ensuing ill-formed binding
relation leads to ungrammaticality. We also suggested in Section 8.3 that certain
lexical items are specified as inherently affixal, marked with the diacritic ‘+’;
this is true for Present and Past Tense in English. So we can add the PF prop-
erties of Present and Past to the lexical entry we gave in (20), as follows (still
leaving aside the s-selection and semantic specifications, which are not directly
relevant here in any case):
(45) a. [Present] (T): __ VP. PF: /+ø/
b. [Past] (T): __ VP. PF: /+d/

If T is not raised to C in the narrow syntax, then Present/Past satisfy their affixa-
tion requirement at PF by lowering to V. This gives the following derived struc-
ture for (43a):
(46) TP

NP T′

D N′ T VP
a +d
N V′
fish
V NP

V T D N′
fish +d a
N
fish

Here we see [t +d ], i.e. Past Tense, lowered to V (and in fact adjoined to it).
Various PF rules now apply: the copy of [t +d ] in T is deleted by regular copy
deletion, on which see more below; and the derived [v [v fish ][t +d ]] is realised
(‘spelled out’) as /fɪʃ+d/ and input to the assimilation rule we described above,
converting /d/ to /t/. Phonological realisation of [t +d ], i.e. ‘spell out’ as /d/, is
blocked by disjunctive ordering where the lexical entry of the verb is specified
as having a special past-tense form, as we saw in the case of sing in (21): the
availability of sang (/sæŋ/) as sing+Past overrides singed (/sɪŋd/).
Where CQ attracts T in the narrow syntax, T-to-V lowering cannot take place.
This is presumably because T in C is too distant from V. In that case, the affixal
property of T is satisfied by insertion of do at PF, which combines with +d to
give /dɪd/, again overriding the default form, which would presumably be /
du:d/.
‘Do-support’ is also required in negative contexts. Here it is presumably the
presence of the negation, not/n’t, which prevents T-to-V lowering. It is tempt-
ing to attribute this to linear adjacency, but other adverbs, including negative

IanRoberts_9781316519493_C008.indd 197 05-09-2022 17:44:24

 198 8 the architecture of grammar

adverbs such as never, can intervene linearly between T and V without causing
do-insertion:
(47) a. John did not go on holiday in Bangor.
b. John often went on holiday in Bangor.
c. John never went on holiday in Bangor.

As we saw in our comparison of English and French verb positions in Section 7.3,
often can be analysed as left-adjoined to V′; never is semantically similar to
often in that it modifies the action described by V′, adding intrinsically negative
content. Nonetheless, it does not block T-to-V lowering, as (47c) shows. What
is the generalisation over T-to-C raising and not-insertion which blocks T-to-V
lowering in both cases?
It is clear that, after T-to-C raising, a head, namely the copy of the raised T,
sits in between T in C and V. Similarly, we can treat the clausal negator not as
a head; one piece of evidence for this is that in its contracted form it raises to T,
giving forms such as won’t, can’t, etc., which can then raise further to C giving
sentences like (48):
(48) a. Won’t Vurdy talk to Goona?
b. Can’t you understand this?

(This shows, incidentally, that n’t-raising is a narrow-syntactic operation, not a

PF one.) So, in negative sentences, we have a structure along the lines of (49)
(omitting Neg′ and V′ for convenience and glossing over the c-selection proper-
ties of T this structure appears to introduce):
(49) TP

NP T′
John
T NegP
+d
Neg VP
not
V NP
interview Oprah

Compare (49) with the representation of a subject-auxiliary inversion example

in (50):
(50) CQP

CQ, Wh TP

T NP T′
+d you
T VP
+d
V NP
catch a fish

IanRoberts_9781316519493_C008.indd 198 05-09-2022 17:44:25

 8.6 Phonological Form (PF) 199

In both cases, we see that there is a head in between T and V: Neg in (49) and the
copy of T in (50). So we can conclude that T-to-V lowering is blocked if a head
intervenes between T (containing +d) and V; this prevents T lowering from C to
V here. In Volume II, we will see that this is a subcase of a much more general
constraint on movement in general.
Where T is Present, everything is the same as with Past. But there is one
important difference, which shows us that PF is quite an abstract level of rep-
resentation. The PF specification of Present is as a null affix: /+ø/. We see this
clearly in (42), where the present-tense form of the verb fish is /fɪʃ/. The reason
we posit T-to-V lowering here, although it does not affect the surface phonolog-
ical form since the affix is null, is that do-insertion is obligatory in main-clause
interrogatives and in negative clauses:
(51) a. Do fish fish fish?
b. Fish do not fish fish.

So do is inserted here to satisfy the affixal property of Present Tense, even

though no affix is phonologically realised.
There is a further complication involving Present Tense. Where the subject
is third singular, Present is realised as /+z/. In fact, we could say that there is a
separate Agreement head which is realised in this way in this context (as in (18a)
of Chapter 1), or we could combine Tense and Agreement features on T and add
(52) to (45a):
(52) PF: /+ø/; /+z/ __ /[Person:3, Number: Sg]

Again, disjunctive ordering will force /+z/ in the more specified context where
the features [Person: 3] and [Number: Sg] are present, while /+ø/ is the else-
where form. For English, this choice between a separate Agreement (or Agr)
head or (52) may be somewhat moot. In languages with richer tense and agree-
ment inflection (e.g. Italian, as we saw in Chapter 1), distinguishing Tense and
Agr as separate heads may be advantageous. As we also mentioned in Chapter 1,
it may be desirable to ‘line up’ English and Italian verb forms as much as possi-
ble, which would entail postulating a separate Agr head in English.
The +z realisation of Present behaves exactly like Past +d, both in relation to
do-insertion (applying where T-to-V lowering is blocked by an intervening head
in the contexts in (49) and (50)) and in relation to the phonological assimilation
rule, which changes /z/ to /s/ following a voiced obstruent (cf. knots /nɒts/ vs
nods /nɒdz/) and inserts an epenthetic vowel when the root ends in a coronal
fricative or affricate as in fishes (/fɪʃɪz/). Furthermore, a few verbs show stem
allomorphy with +z: do becomes /dʌz/, not /du:z/, and says is pronounced /sεz/
by many speakers (see Section 2.2).
What we see in the contrasting PF realisations of T (and of D in (42) and (43))
is simply differences in the lexical specifications of functional heads: Present
(aside from third singular) happens to be null, just as fish happens to start with
/f/; these are simply matters of lexical arbitrariness (i.e. why we need the lexi-
con in the first place). Copies resulting from movement are different: here we

IanRoberts_9781316519493_C008.indd 199 05-09-2022 17:44:25

 200 8 the architecture of grammar

have a productive rule of copy deletion, as we have seen in our examples of all
three movement types. Since copy deletion removes phonological content, we
take it to be a PF process. We now have three PF processes: T-to-V lowering,
do-support and copy deletion. In addition we have the morphological processes
of lexically conditioned stem allomorphy forming the irregular past tenses of
verbs like sing and catch, and the regular phonological assimilation rules apply-
ing to the affixes +d and +z. It is important to see that these rules must apply in
a particular order, or ungrammaticality will result.
We can easily see that there is an intrinsic ordering of T-to-V lowering and
copy deletion in that the former creates the context for the latter to apply; this
is known as feeding order. Furthermore, lowering clearly creates the context
for the irregular past-tense forms and for the assimilation rules applying to +d
and +z; it therefore feeds and precedes those rules. Finally, if do-insertion is a
‘repair’ rule which applies when lowering fails in order to satisfy the affixation
requirements of +d, +z and +ø, then it must also follow T-to-V lowering. Do-
support seemingly applies where lowering does not apply; these rules are in
bleeding order as lowering removes the context for do-insertion. Do-support
also prevents the formation of irregular past tenses, and as we saw, do itself has
an irregular past form and so must feed, and therefore precede, that rule. It also
clearly precedes the assimilation rules. So it seems clear that we have the follow-
ing partial ordering of the PF operations we have been considering:
(53) T-to-V lowering > do-insertion > irregular past-tense formation > assimilation

It remains to be seen how copy deletion is ordered in relation to (53). We have

already mentioned that it must follow T-to-V lowering. How is it ordered in rela-
tion to do-insertion? The account we gave above of (46) provides the answer.
Let us look again at (46).
(46) TP

NP T′

D N′ T VP
a +d
N V′
fish
V NP

V T D N′
fish +d a
N
fish

Here, if do-support precedes copy deletion, do will be inserted in T, giving rise

to the ungrammatical *A fish did fished a fish; here do-insertion would in fact
bleed copy deletion, since did is not a copy of +d (alternatively, copy deletion
could delete +d giving the equally ungrammatical *A fish do fished a fish).

IanRoberts_9781316519493_C008.indd 200 05-09-2022 17:44:25

 8.6 Phonological Form (PF) 201

Let us now look again at the result of narrow-syntactic T-to-C raising of +d,
repeated from (50):
(50′) CQP

CQ, Wh TP

T NP T′
+d you
T VP
+d
V NP
catch a fish

(50′) differs from (50) in that the copy of +d in T has not been deleted. If T-to-V
lowering precedes copy-deletion, and T-to-C movement, as a syntactic rule,
precedes all the rules in (53), then (50’) is the representation which is the input
to PF. Following (53), we can in fact apply T-to-V lowering here, giving a form
such as [v [v catch ] +d ] in the V position (see again (46)). If copy-deletion
precedes do-support, the copy of +d in the original T position will disappear and
there will be no head in between the moved T in C and V. However, this is not
a problem since T-to-V lowering has already applied; if T-to-V lowering were
allowed to apply iteratively it could then lower T from C to V, bleeding do-in-
sertion, so we see that this is not a possibility. Copy deletion must in any case
precede the other rules in (53), irregular past-tense formation and assimilation,
as it will then delete the occurrence of +d attached to V and bleed these rules,
guaranteeing that where do is inserted in C the verb always appears uninflected.
So we arrive at the following ordering of the PF-rules:
(54) T-to-V lowering > copy deletion > do-insertion > irregular past-tense for-
mation > assimilation

Evidence from child language is relevant here. It has been observed that children
sometimes ‘repeat’ the auxiliary in inversion contexts at early stages of language
acquisition, producing examples like (55):
(55) a. Can its wheels can spin?
b. Is the steam is hot?
c. Was that was Anna?

We could account for this by saying that, at the relevant stage of first-language
acquisition, children have not fully acquired copy deletion, at least in the T-to-C
context. Very strikingly, the same applies with do-support, examples like (56)
being attested:
(56) Did the kitchen light did flash?

We predict that (56) would arise from failure of copy-deletion of the occurrence
of do+d in the original T position in (50’). This form is then input to past-tense

IanRoberts_9781316519493_C008.indd 201 05-09-2022 17:44:26

 202 8 the architecture of grammar

formation, and did results. Failure of deletion of the lowest copy of T, the one
attached to V, would give rise to examples like (57):
(57) Did the light (did) flashed?

These are also attested in the spontaneous speech of children acquiring English.
A further environment for do-support is VP-ellipsis, as in (58), repeated from
Section 3.4:
(58) John ate the cake and Mary did eat the cake too.

Here, as we argued in Section 3.4, the VP eat the cake is elided. One approach
to English VP-ellipsis is to treat it as PF deletion. We interpret the elided VP
as eat the cake because it is present at LF but absent at PF (a good example of
understanding more than we hear). If VP-ellipsis is a PF operation, we can ask
how it is ordered in relation to the operations we have been looking at here. The
answer is clear from the presence of do-support in (58): ellipsis bleeds T-to-V
lowering by deleting V and thereby feeds do-insertion by creating the context
where do is required to meet the affixal requirement of +d in T. The order of
operations is shown in (59):
(59) a. … and [TP Mary [T′ [T +d ] [VP eat the cake ]]] (S-structure)
b. … and [TP Mary [T′ [T +d ] [VP eat the cake ]]] (VP-ellipsis)
c. … and [TP Mary [T′ [T do+d ] [VP eat the cake ]]] (do-insertion)
d. … and [TP Mary [T′ [T did ] [VP eat the cake ]]] (past-tense formation)

Because T-to-V lowering doesn’t apply in (59), copy deletion doesn’t apply.
If VP-ellipsis followed T-to-V lowering, the result would be (60):
(60) John ate the cake and Mary too.

(Here the interpretation where Mary eats the cake is the relevant one, not the one
where John eats Mary; see Section 3.4.) The sentence in (60) could be derived
in the following way:
(61) a. … and [TP Mary [T′ [T +d ] [VP eat the cake ]]] (S-structure)
b. … and [TP Mary [T′ [T +d ] [VP [ eat ] + d] the cake ]]] (by T-to-V lowering)
c. … and [TP Mary [T′ [T +d ] [VP [ eat ] + d] the cake ]]] (by copy deletion)
d. … and [TP Mary [T′ [T +d ] [VP [ eat ] + d] the cake ]]] (by VP-ellipsis)

Example (60) is in fact grammatical, but it is does not involve VP-ellipsis. As we

saw in Section 3.4, (60) is the result of T′-deletion, which can apply with other
auxiliaries as in (62) and so is not connected to do-support:
(62) a. (John can speak Mandarin) and Mary [T′ can speak Mandarin ] too.
b. (John has written a novel) and Mary [T′ has written a novel ] too.

The T′-deletion operation can derive the string in (60) independently of

VP-ellipsis and the operations in in (61), but not vice versa.
We conclude that the ordering among the PF operations we have been consid-
ering is the following:

IanRoberts_9781316519493_C008.indd 202 05-09-2022 17:44:26

 8.7 Logical Form (LF) 203

(63) VP-ellipsis > T-to-V lowering > copy deletion > do-insertion > irregular
past-tense formation > assimilation

It seems natural to suppose that the more purely phonological operation of

assimilation should be the last of the series, while operations more similar to
syntax, e.g. T-to-V lowering, which is a form of head-movement, should occur
earlier. Similarly, the more morphological operation of irregular past-tense for-
mation, i.e. stem allomorphy, perhaps naturally occurs in between. Furthermore,
as we have already observed, T-to-V lowering must precede copy deletion. The
place of VP-ellipsis and do-insertion must be as given here, given the evidence
we have seen, but there is no clear theoretical ground for this.
In this section I have gone into some detail concerning a small number of
PF operations which apply in English, in order to illustrate the workings of this
component of the grammar and its interactions with narrow-syntactic rules, the
lexicon and both phonological and morphological rules. We can see that the
relevant rule systems are highly intricate. But a very important point, which we
have already stressed, is that these rules do not interact with semantics, i.e with
Logical Form. Now it is time to look at that component of the grammar.

8.7 Logical Form (LF)

As we said in Section 8.2, Logical Form (LF) is the level of rep-
resentation which forms the input to the semantic component. It is important to
see that LF itself is not the semantic representation: it is a syntactic representa-
tion (a phrase marker) which can be readily converted into a representation
from which, on standard assumptions, truth conditions can be read and hence
the meaning deduced. As we have seen, the architecture of the grammar in (1)
involves a split into two paths after S-structure, one leading to PF, the other to
LF. Since the PF path, which we looked at in the previous section, exclusively
represents what we actually hear, the derivation from S-structure to LF is some-
times referred to as ‘covert’; we are not directly presented even with LF strings,
instead we must deduce them from various forms of indirect evidence mainly,
since we are concerned with the input to semantics, concerning possible and
impossible interpretations of various kinds.
In Section 6.7, we gave a very brief introduction to the semantic notions of
quantifier and variable. There we saw that an NP like every student doesn’t refer
to a single individual (the way proper names like Goona and Vurdy do); instead,
owing to the nature of the determiner every, this kind of NP is a quantified
expression. Such NPs refer to sets of entities and relate them in various ways to
the set denoted by a predicate, such as likes pizza. Every student in (64) says that
every member of the set of students is also in the set denoted by the predicate,
the set of pizza-likers:
(64) Every student in our class likes pizza.

IanRoberts_9781316519493_C008.indd 203 05-09-2022 17:44:26

 204 8 the architecture of grammar

With a quantifier like every (a universal quantifier), this is a subset relation. We

went on to give the quasi-logical representation for (64) in (65):
(65) For every x, if x is a student in our class, then x likes pizza (or ‘if x is in the
set of students in our class, then x is in the set of pizza-likers’).

Here x is a variable, ranging over individuals in the domain of discourse. To get

the full meaning of (64), we have to ‘plug in’ individuals for x (technically, we
assign values to the variable) and then see if that individual, say individual 23 =
John23, individual 5 = the man next door5, individual 666 = the pet shop boy666,
etc., is a member of the set of students and a member of the set of pizza-likers.
As long as we don’t find any individuals which are in the set of students and not
in the set of pizza-likers, then the sentence is true. What is crucial here is that the
two occurrences of the variable x in ‘if x is a student, then x likes pizza’ always
have the same value. In logic and semantics, these variables are said to be bound
by the quantifier, and so as we compute the semantics of the quantifier, plugging
in values for the variables, the two variables always have the same value.
We next related quantified expressions like (64) to wh-questions like (66):
(66) Which coursei does Goona like ti ?

The crucial observation is that which is also quantificational rather than refer-
ential: the answer to the question could be ‘none of them’, showing that which
course doesn’t have to pick anything out in the domain of discourse. Like every
student, which course applies to a set of entities (courses) and asks for the iden-
tity of a member of that set such that Goona likes it. In quasi-logic:
(67) For which x, x a course, does Goona like x ?

In this we observed an important semantic similarity between wh-elements and

quantifiers: both instantiate operator-variable structures. The operator (or quan-
tifier) applies to a set of entities and specifies the subpart of that set we are
dealing with (in the case of wh-operators, find a member). The variable tells us
where in the structure you would ‘plug in’ the specified constant in relation to
the predicate so as to compute its relation to the predicate and see if the sentence
is true or false. So, in relation to (67), we plug in different courses in the variable
in object position as values of x, and ask if Goona is a member of the set of x-lik-
ers. Again, it is crucial that the two x′s following ‘which x′ in (67) always have
the same value. So, we concluded that these variables are semantically bound
by the wh-word and that the rightmost x in (67) corresponds to the trace/copy
of wh-movement. So we arrived at the important conclusion that syntactic bind-
ing, c-command and coindexation correspond exactly to the semantic binding
involved in interpreting the wh-quantification. This is an important conclusion
for LF, as it means that wh-movement creates a syntactic configuration which
can be interpreted at LF as one kind of quantificational structure, subject to the
general semantic rules for interpreting such structures.
We said above that it is vital for this system of interpreting quantifier-variable
relations that in expressions like (65) and (67) each x should be ‘filled in’ with

IanRoberts_9781316519493_C008.indd 204 05-09-2022 17:44:26

 8.7 Logical Form (LF) 205

the same value. The variables that have this property are said to be in the scope
of the quantifier. So all the x′s in both (65) and (67) are in the scope of every or
the wh-operator. In predicate logic, (65) could be written as follows:
(68) "x [ student(x) → likes-pizza(x) ]

Here the symbol " is the universal quantifier, meaning roughly ‘every’ or ‘all’,
and the arrow indicates the ‘if–then’ implication relation, so (68) restates (65).
The important point is that, by convention, the scope of the quantifier is indi-
cated by the brackets following it; all and only occurrences of x inside that
bracket are in the scope of the quantifier, i.e. are interpreted as bound by it. This
is a convention developed by logicians.
If we now look at (66) and (67) in the light of these logical conventions, we can
say that the interrogative C (CQ, Wh) which triggers wh-movement to its Specifier
thereby does two things. First, as we saw in our discussion of wh-movement in
Chapter 5, it determines the position of the wh-operator by triggering wh-move-
ment, thereby creating a copy in the original position. Second, it determines
the scope of the wh-phrase. We can directly connect these two points by the
following postulate:
(69) A quantified phrase has scope over the portion of phrase structure it
­c-commands at LF which contains a variable that it binds.

Let us look again at our original example of wh-movement from Chapter 5:

(70) CQ, WhP

NP CQ, Wh′
what
CQ, Wh TP

T NP T′
did you
T VP
(did )
V NP
catch (what )

Here CQ, Wh triggers movement of what from the object position to SpecCQ, WhP.
The Projection Principle requires the presence of the copy in object position.
This copy is deleted at PF by the general copy-deletion rule we discussed in
the previous section, but given that PF and LF do not interact, it remains at LF.
In fact, it is interpreted as the variable bound by the wh-operator. The variable
is in the scope of the operator since it is c-commanded by the operator. This is
guaranteed by the general fact that narrow-syntactic movement always targets a
c-commanding position (we saw in the previous section that PF-movement need
not have this property); furthermore, the fact that the variable is bound by the
wh-operator results from the definition of binding as c-command combined with
coindexing (see Chapter 6, (38)). Hence, (70) can be read at LF as ‘for which
thing x, you caught x′ (glossing what as ‘which thing x′), and this can be read by

IanRoberts_9781316519493_C008.indd 205 05-09-2022 17:44:27

 206 8 the architecture of grammar

interpretative rules for quantifier-variables along the lines sketched above and in
a little more detail in Section 6.7.
Since it is CQ, Wh which triggers wh-movement to its Specifier, this element
effectively determines the scope of wh-operators. We saw this in Section 7.6 in
relation to the different matrix verbs in the following examples:
(71) a. We wonder what you bought.
?*What did you wonder (that) I bought?

b. *I think what you bought.

What did you think (that) I bought?

c. I remember what you bought.

What did you remember (that) I bought?

As we observed there, the scope of the interrogative, i.e. the wh-operator, is

determined by the c-selection properties of the main verbs: wonder selects
CQ, Wh and so the operator only moves to the lower SpecC′ and has scope just over
the embedded clause; think cannot c-select CQ, Wh and so in (71b) it is the matrix
CQ, Wh which determines the movement and scope domain of the ­wh-operator,
and remember optionally c-selects CQ, Wh, and so either possibility is available.
C-selection correlates with s-selection here, in that wonder obligatorily ­s-selects
a question, think obligatorily s-selects a statement and remember ­s-selects either.
We see in general, then, that the movement and scope domains of wh-operators
coincide in English.
However, we also saw in Section 7.6 that there are languages in which
wh-phrases do not move. Mandarin Chinese is probably the best-known exam-
ple. As we saw in Section 7.6, the same scope and selection facts as in (71)
hold in Mandarin, although the wh-phrase does not move overtly. Here are the
relevant examples once again:
(72) a. Zhangsan xiang-zhidao [CP Lisi mai-le shenme ].
Zhangsan wonder     Lisi buy-perf what
‘Zhangsan wonders what Lisi bought.’

b. Zhangsan yiwei [CP Lisi mai-le shenme ] ?

Zhangsan think Lisi buy-perf what
‘What does Zhangsan think that Lisi bought?’

c. Zhangsan jide [CP Lisi mai- le shenme ] ?

Zhangsan remember Lisi buy- perf what
‘Zhangsan remembers what Lisi bought.’ Or
‘What did Zhangsan remember that Lisi bought?’

We also saw evidence, from sentential particles, that C can bear the Q and
wh-features. So the question arises of how the wh-operator is interpreted in
Mandarin, a question we left open in Section 7.4.
The covert derivation to LF allows us to give a very simple answer to this ques-
tion: the Mandarin sentences in (72) have different structures at LF, where covert

IanRoberts_9781316519493_C008.indd 206 05-09-2022 17:44:27

 8.7 Logical Form (LF) 207

wh-movement operations parallel to those that take place overtly in languages like
English create the required operator-variable structures. So, a simple wh-question
in Chinese such as (73a) has the S-structure in (73b) and the LF in (73c):
(73) a. Lisi mai-le   shenme?
Lisi buy- perf what
‘What did Lisi buy?’

b. CQ, WhP

CQ, Wh TP

NP T′
Lisi
T VP

V NP
mai-le shenme
c. CQ, WhP

NP CQ, Wh′
shenme
CQ, Wh TP

NP T′
Lisi
T VP

V NP
mai-le (shenme)

(Since the sentential particles marking the clause as Q or wh are final in

Mandarin – see examples (30) and (31b) in Chapter 7 – we should perhaps place
C in final position, but I leave this complication aside here.) Thanks to the oper-
ation of covert wh-movement, at LF, Mandarin wh-questions are interpreted just
like their English counterparts in terms of scope and variable-binding. The LFs
of the two languages are close to uniform, while their S-structures (and PFs)
appear quite different; this is in line with the fact that the interpretation of quan-
tifiers, including wh-quantifiers, and variables is cross-linguistically uniform.
Covert wh-movement can distinguish the interpretations of the sentences in
(72). The indirect question readings in (72a,c) result from wh-movement to the
intermediate SpecC′, and the direct question readings in (72b,c) result from
wh-movement to the root SpecC′. So we can treat the Mandarin c-selection
restrictions seen in (72) in the same way as those in English in (71).
So, in a nutshell: Mandarin has wh-movement. But this movement is cov-
ert, taking place after the derivation has split into the PF and LF paths. Since
it only takes place on the LF path, we hear the Mandarin wh-phrases in situ,
but we understand them in their moved positions. Thus, both languages feature
wh-movement; the only difference lies in the point of the derivation at which

IanRoberts_9781316519493_C008.indd 207 05-09-2022 17:44:27

 208 8 the architecture of grammar

the operation takes place. This difference is regulated by a parameter of the kind
described in Section 7.6. A consequence of this is that if Mandarin has covert
wh-movement, then we might expect it to have covert island phenomena com-
parable to the overt ones of English seen in Section 5.6. I will not go into this
question here, but return to it in Volume II.
Are there any cases of covert movement in English? If we take seriously the
idea that LF represents quantifier-variable relations in terms of binding (coin-
dexation and c-command) of a variable created by movement, then we should
extend our account of wh-questions to sentences containing quantifiers like
every, as in (64) and (65), repeated here:
(64) Every student likes pizza.
(65) For every x, if x is a student, then x likes pizza (or ‘if x is in the set of stu-
dents, then x is in the set of pizza-likers’).

If (65) approximates the LF of (64), then every student raises and its copy is
interpreted as a variable bound by it. So (64) has the LF in (74):
(74) TP

NPi TP

every student NPi T′

every student T VP
−z
V NP
like pizza

Here every student adjoins to TP (this is the standard kind of X′-adjunction

which we saw in Chapter 4, extended from the X′ to the XP-level). As in all
cases of narrow-syntactic movement, it c-commands and is coindexed with its
copy and therefore binds it: the copy is thus interpreted as a variable. Thus (74)
can be interpreted as in (65) or, more formally, as in (68). The operation adjoin-
ing every student to TP here is known as Quantifier Raising, or QR.
This approach allows us to capture quantifier-scope ambiguities of the kind
illustrated by (75):
(75) Every student dreads some lecture.

This sentence has two interpretations. In one, every student has scope over some
lecture, so the interpretation is ‘for every student x, there is some lecture y such
that x dreads y′. The dreaded lecture could be a different one in each case (e.g.
Goona dreads Syntax, Vurdy dreads Phonetics, Pete dreads Semantics, etc.). In
the second reading, some lecture has scope over every student, so the meaning
is ‘there is some lecture x such that for every student y, y dreads x′, i.e. one
lecture so dreadful that all students dread it (clearly this cannot be a Syntax lec-
ture …). The first reading is the wide-scope reading for every student (it’s also

IanRoberts_9781316519493_C008.indd 208 05-09-2022 17:44:28

 8.7 Logical Form (LF) 209

the ‘surface scope’ reading, in that the linearly first quantifier has wider scope).
The second reading is the wide-scope reading for some lecture (it is also known
as the ‘inverse scope’ reading as the scope relations don’t follow surface linear
precedence).
If scope is determined by c-command, then we can disambiguate (75) by
applying QR in different orders where there is more than one quantified NP. Thus
the first reading of (75) (wide scope for every student) is represented in (76a),
while the second reading (wide scope for some lecture) is represented in (76b):
(76) a. TP

NPi TP

every student NPj TP

some lecture NPi T′

every student T VP
−z
V NPj
dread
some lecture

b. TP

NPj TP

some lecture NPi TP

every student NPi T′

every student T VP
−z
V NPj
dread
some lecture
The postulate in (69) that scope reduces to c-command determines that every
student has scope over some lecture in (76a) and some lecture has scope over
every student in (76b). These representations can then be input to logical rep-
resentations of the kind seen in (68) as necessary, in order to derive the semantic
interpretations.
We also see covert wh-movement in certain contexts in English. This occurs
in multiple questions (which we saw briefly in Section 5.3), illustrated in (77):
(77) Who wrote which songs?

Here we assume the subject wh-phrase who has moved from SpecT′ to SpecC′
(recall the discussion of this in Section 5.4). The object wh-phrase which songs
has clearly not moved. In English, the rule for multiple questions is that where

IanRoberts_9781316519493_C008.indd 209 05-09-2022 17:44:28

 210 8 the architecture of grammar

there is more than one wh-phrase in the clause only one can move (this isn’t
true in all languages, e.g. Russian and other Slavic languages have multiple
wh-movement). Nonetheless, which songs must bind a variable in (77) in order
to be properly interpreted as a quantifier. Moreover, it interacts with who: the
natural answer pairs people and songs (e.g. Paul wrote Hey Jude, John wrote I
am the Walrus, George wrote Here Comes the Sun, etc.). This is known as the
pair-list interpretation of multiple questions.
We can account for this by proposing that in multiple questions the wh-phrases
combine (by a rule known as Absorption) to form a complex operator, giving an
LF representation of the kind in (78):
(78) [CP [whoi, which songsj] [TP xi wrote xj ]

The interpretation of (78) is ‘for which x, x a person, and for which y, y a song,
x wrote y′, which naturally elicits a pair-list answer of the kind seen in the pre-
vious paragraph. As with covert wh-movement in Mandarin, the postulation of
covert movement in English leads to the prediction of covert island effects; we
will come back to this in Volume II.
The final LF phenomenon we will look at here is reconstruction. This is
illustrated in (79):
(79) [NP Which joke about himselfi ] does Donaldi dislike --- most ?

This sentence is grammatical. It involves wh-movement of the complex wh-NP

from the object position marked by the dash (for convenience). There is clearly
nothing unusual about the wh-movement here, but the wh-phrase contains the
anaphor himself. Recall from Chapter 6 that anaphors are subject to Principle A
of the binding theory, which requires them to be bound in their binding domain.
Binding involves c-command by and coindexation with the antecedent. But here
the anaphor inside the moved wh-phrase is not c-commanded by an anteced-
ent. Furthermore, it is clear that the antecedent is Donald, as the coindexation
indicates. Assuming that the binding principles apply at S-structure or LF, i.e.
after wh-movement has taken place, this example appears to be a problem for
Principle A of the binding theory.
The idea that movement leaves a copy provides a solution to this problem.
The full representation of (79) at S-structure is (80):
(80) [NP Which joke about himselfi ] does Donaldi dislike [NP which joke about
himselfi ] most ?

The lower copy is deleted at PF, giving (79) (without the trace). But, again
thanks to the independence of PF and LF, we can retain the copy at LF. In that
context, the anaphor himself satisfies Principle A: it is c-commanded by its coin-
dexed antecedent Donald within its binding domain (which is TP, the smallest
XP containing the anaphor and a subject as well as a finite T; see (42) of Chapter
6). So LF looks at the copy for satisfaction of Principle A. In fact, since the
occurrence of the anaphor inside the moved wh-phrase in (80) would otherwise

IanRoberts_9781316519493_C008.indd 210 05-09-2022 17:44:28

 8.8 Conclusion 211

violate Principle A, it may be that copy deletion deletes this copy at LF (but not
the whole wh-phrase, as we need an operator here).
This concludes our very brief introduction to LF. We have seen three examples
of covert movement, i.e. movement that takes place after the derivation splits
into its PF and LF branches at S-structure and whose effects therefore cannot be
directly ‘heard’: wh-movement in Mandarin, QR in English and wh-movement
in multiple questions in English. Finally, we looked at reconstruction in relation
to copies and copy deletion. All of these operations are motivated by the need to
create LF representations which can be directly fed into semantic interpretation.

8.8 Conclusion
To end this chapter, let us look once more at the overall architecture
of the grammar given in (1):

(1) LEXICON

D-STRUCTURE

S-STRUCTURE

PHONOLOGICAL LOGICAL FORM

FORM

The lexicon is a list of lexical items with an associated specification of syntactic

category, c- and s-selection features, a phonological representation and a specifi-
cation of the item’s semantics. Functional items are also listed in essentially
the same way, although functional items may be phonologically null (presum-
ably not a possibility for lexical items) and may have rather simple c-selection
specifications. If all lexical specifications can be reduced to features, then the
Projection Principle requires them to hold at all levels of representation.
D-structure is the level of syntactic representation generated by PS-rules in
line with X′-theory. Lexical and functional items are inserted at this level in con-
formity with the Projection Principle; c- and s-selection properties must there-
fore be respected. This directly determines what appears in X′. To the extent
that they have direct structural correlates, arguments bearing particular thematic
roles may appear in specific positions at this level, e.g. Agents always occupy an
external-argument position.
S-structure is derived from D-structure by movement rules: head-move-
ment, NP-movement and wh-movement. These operations all share the central
property of substituting an element of type X into a c-commanding position
of type X, where X ranges over head position, grammatical-function position
(or A-position, e.g. SpecT′) and wh-position (or A′-position, such as SpecC′.

IanRoberts_9781316519493_C008.indd 211 05-09-2022 17:44:29

 212 8 the architecture of grammar

Movement leaves a copy in the original position. Wh-movement is subject to

island constraints of various kinds. At S-structure the derivation divides and fol-
lows two separate paths, one to PF and one to LF. The derivations of these inter-
face levels are therefore independent of one other and cannot interact directly.
Phonological Form, PF, takes S-structure as its input and gives a phonological
representation as its output. As such a great deal of syntactic information, includ-
ing much (but almost certainly not all) hierarchical structure, is lost. Unlike in
narrow syntax (the derivation from D-structure to S-structure and S-structure to
LF), ‘downward’ movements such as T-to-V lowering in English are allowed.
The operations at PF appear to be ordered in a partially intrinsic fashion, with
more syntactic operations such as head-movement (lowering) taking precedence
over deletion operations such as copy-deletion and ellipsis, as well as repair
operations like do-insertion, which in turn precede morphological stem allomor-
phy rules and phonological assimilation rules.
Logical Form, LF, takes S-structure as its input and derives a representation
which can directly feed the semantic interpretation proper. The operations in the
derivation from S-sructure to LF are covert, in the sense that they do not inter-
act with phonology, given the architecture in (1). We saw examples of covert
wh-movement and Quantifier Raising. We also saw that copy deletion may not
have to apply as it does in PF and may, in fact, affect only parts of copies. Copies
left by wh-movement and Quantifier Raising are interpreted as variables.
This concludes our overview of the architecture of grammar. Of the five lev-
els in (1), two are interface levels (PF and LF) and one must be posited in order
to account, at the very least, for arbitrary idiosyncracies of a given language,
namely the Lexicon. The other two levels, D-structure and S-structure, are
‘internal’ to the grammar. In part they reflect purely syntactic properties; in part
they act as a bridge between the other levels, with D-structure particularly close
to the lexicon and S-structure closer to PF and LF. The core mechanisms at these
levels have been our principle focus in earlier chapters.

Exercises
1. There are at least four kinds of have in English, shown in (a–d):
a. John has written a book. [perfect have]
b. John has to write a book. [modal have (to)]
c. John has many books. [possessive have]
d. John has his students proofread his books. [causative have]
Which of these requires or allows (i) subject-auxiliary inversion, (ii)
negation with n’t, (iii) contraction to ‘ve? What can we conclude
about the position(s) the different kinds of have can occupy? There is
some dialectal variation concerning the possessive and modal forms.

IanRoberts_9781316519493_C008.indd 212 05-09-2022 17:44:29

 Exercises 213

You may also want to try replacing have with (have) got; here again
there is some dialectal variation. Try to formulate and analyse the
data based on your own judgements.
2. Modals have complex and intriguing semantic properties and have
been the object of a great deal of study. One central idea in many
analyses of modal semantics is that they involve quantification over
possible worlds. So, (ia) can be approximately glossed as in (ib), and
(iia) as (iib):
(i) a. I may have lost my keys.
b. In some possible world (consistent with what I know), I
have lost my keys.
(ii) a. I must have lost my keys.
   b. In all possible worlds (consistent with what I know), I
have lost my keys.
‘Possibility’ modals like may involve existential quantification over
possible worlds and ‘necessity’ modals like must involve universal
quantification. If modals have quantificational properties, the ques-
tion of their scope, and possible scope ambiguities, arises. What are
the scopal relations between the modal and negation in the following
examples?
(iii) a. John cannot come along today.
    b. Confronting the enemy cannot be avoided.
   c. You mustn’t drink and drive.
   d. You needn’t drink and drive.
(Can is a possibility modal, need is a necessity modal). If c-com-
mand relations at LF determine scope, what LF representations do
these examples have?
3. There is a scope ambiguity in example (i):
(i) What did everyone buy for Bill?
Explain the ambiguity and give the relevant LF representations.
What does this tell us about the nature of QR?
4. Describe the derivation of the following sentence:
(i) How many students were arrested?
Give the D-structure, S-structure, PF and LF representations and
describe how the later levels are derived from the earlier ones. At
PF, describe how the forms were and arrested are derived. At LF,
bear in mind the implicit agent (‘by someone or other’).

IanRoberts_9781316519493_C008.indd 213 05-09-2022 17:44:29

 214 8 the architecture of grammar

Further Reading
Most of the textbooks cited as Further Reading in earlier chapters discuss the organisa-
tion of the grammar (from slightly varying perspectives) and various more advanced top-
ics not covered here. On some of the specific topics covered here (the lexicon, passives,
do-support and affix lowering), see the following:

Adger, D. 2003. Core Syntax. Oxford: Oxford University Press, Chapter 3.5 (on
thematic roles) and 5.5 (on do-support).
Carnie, A. 2002. Syntax: A Generative Introduction. Oxford: Blackwell, Chapters 8 (on
the lexicon), 9 (on head-movement and do-support) and 10 (on passives
and related matters).
Larson, R. 2010. Grammar as Science. Cambridge, MA: MIT Press, Unit 13 (on the
lexicon).
Freidin, R. 2012. Syntax: Basic Concepts and Applications. Cambridge: Cambridge
University Press, Chapter 7.2 (on do).
Radford, A. 2016. Analysing English Sentences. Cambridge: Cambridge University
Press, Chapter 5.10 (on do-support).
Sportiche, D., H. Koopman & E. Stabler. 2014. An Introduction to Syntactic Analysis
and Theory. Oxford: Wiley, Chapter 6.8 and 6.9 (on the lexicon).
Tallerman, M. 1998. Understanding Syntax. London: Routledge, Chapter 7.1 (on
passives).

IanRoberts_9781316519493_C008.indd 214 05-09-2022 17:44:29

 Conclusion

In the foregoing chapters, I have tried to present the core elements of the the-
ory of syntax. We began with the basic notions of categories and constituents,
and how to find them (Chapter 2). Then we saw how PS-rules can generate
the well-formed structures, creating phrase markers which we can represent
either as tree diagrams or as labelled bracketings. We saw how to justify the
proposed rules and the structures they generate using constituency tests (Chapter
3). Next, we introduced X′-theory, as a simpler, more general and more abstract
format for PS-rules (Chapter 4). In Chapter 5, we introduced movement rules,
concentrating on wh-movement (how it works and where it is unable to apply,
i.e. islands). Chapter 6 introduced the interpretative rules of binding theory,
along with the very important structural relation of c-command and the related
notion of variable binding; here we also made the distinction between A- and
A′-positions. In Chapter 7, we looked a little beyond English and introduced the
important notion of parameter, giving several examples from a small range of
languages along with some data from the World Atlas of Language Structures.
Finally, Chapter 8 introduced the architecture of grammar we adopt, namely the
‘Y-model’, so-called because the derivation looks like an inverted Y, splitting at
S-structure into PF and LF. This led to a discussion of certain PF rules and the
important idea of covert movement in the derivation to LF.
There are many questions that we can raise concerning the model of syn-
tax described here, even while remaining within a broadly Chomskyan concep-
tion of grammar and the language faculty (and of course, if we move beyond
that conception, still more questions arise). One very important question is
whether our assumption that constituency relations (dominance, constituency
and ­c-command) are primary and dependency relations (subject, object, predi-
cate) secondarily defined in terms of constituency is justified. It has often been
claimed that syntactic descriptions based on constituency work well in languages
like English which have fairly fixed word order and rather little inflectional mor-
phology, especially case morphology, while many languages have much freer
word order and appear to mark relations like subject, object, etc. morphologi-
cally (often with case); such languages appear to lend themselves less well to
a constituency-based approach to syntactic analysis and more naturally to a
dependency-based one. Clearly, this question is partly a cross-linguistic one,
and something we were unable to properly investigate in our brief comparative
discussion in Chapter 7. It is of central importance for syntactic theory though.

215

IanRoberts_9781316519493_Conclusion.indd 215 05-09-2022 10:10:37

 216 conclusion

A further set of questions concerns clause structure. In Chapter 5, we proposed

that the tripartite division of the clause into CP, TP and VP, repeated in (1), cor-
responds to different domains, and each domain plays a characteristic role in
building up the clause:
(1) CP

ZP C′

C TP

NP T′

T VP

V NP

CP is the discourse domain, notably featuring clause-typing features like Q and

wh (and associated movement operations). TP is the domain of the ‘inflectional’
functional categories Tense, Agreement, Mood and possibly PerfP and ProgP,
the latter playing a role in the English auxiliary system. VP is the domain of argu-
ment structure: all internal arguments are realised here, while subjects/external
arguments are somewhat distinct from other arguments (they are not c-selected,
many languages including English require a subject position no matter what the
argument structure of the verb is, and predication seems to be a central rela-
tion). We touched on the possibility that external arguments are first inserted
in SpecV′ and raise to SpecT′, but we have not developed this possibility at all.
We mentioned in Chapter 1 and in Chapter 5 that TP may be expanded into a
range of further categories (Mood, Agreement, Aspect, etc.) and in Section 8.6
saw some motivation for a negation phrase, NegP. Can the other domains be
similarly expanded? What are the principles determining this and what are the
limits to such expansions? Can the nature or degree of expansion vary across
languages? A further very major question is whether anything comparable can
be said of other categories, notably nominals.
Another area of grammar we did not develop here is A-movement. Passive
is clearly a central example. A further case of A-movement is seen in the alter-
nation found with verbs like break mentioned in Section 5.3 and repeated here:
(2) a. The vase broke.
b. John broke the vase.

The sentence in (2a) bears a resemblance, both syntactic and semantic, to the
passive The vase was broken. It seems natural to propose that the Theme argu-
ment the vase moves from the object position to the subject position. Finally,
there are raising verbs like seem, which show the alternation in (3):
(3) a. It seems that the world is round.
b. The world seems to be round.

IanRoberts_9781316519493_Conclusion.indd 216 05-09-2022 10:10:37

 Conclusion 217

In (3b), the world arguably raises from the lower subject position to the higher one.
A very general question, which we only addressed rather sketchily, is
what causes movement to happen. We said that a wh-feature on C triggers
­wh-movement, although we did not go into detail as to how or why this happens,
and we didn’t mention any kind of trigger for head-movement or A-movement.
There is much more to say on these points, most of it of a rather technical nature.
Related to this is the question of what underlies the fact that wh-movement is
(mainly) overt in English but covert in Mandarin, that V raises to T in French
and German but not in English, and T raises to all main-clause Cs in German but
only in certain clause types (mainly interrogatives) in English and French. These
last observations suggest that the triggers for movement should be parametrised
in some way.
The Head Parameter clearly has great potential for accounting for typolog-
ical generalisations about cross-linguistic word-order variation, as we saw in
Section 7.4. However, the nature of disharmonic languages like German and
Mandarin needs clarification. Further, while the other parameters we have looked
at here mainly involve the presence or absence of movement (head-movement
and wh-movement, and we mentioned that there are languages which lack pas-
sives and therefore possibly NP-movement), the Head Parameter seems to con-
cern X′-theory rather than movement. Again, the question arises as to the range
of possible parametric variation: which rule systems can be parametrised?
Finally, there are two very large theoretical questions. First, we have seen that
there are two main rule systems: PS-rules (X′-theory) and movement. We have
treated these as quite distinct and we have treated movement as substitution into
positions created by X′-theory. But we also saw two examples where movement
created new adjunction structures: T-to-V lowering in Section 8.6 and Quantifier
Raising in Section 8.7. Can movement also create structure? If so, how differ-
ent are our two rule systems really? As a related question, we saw that lexical
insertion involves substituting a lexical item for a categorially matching head
generated by PS-rules (e.g. substituting fishN for N). But the noun fish has a
feature N, so if we simply place it in a structure it creates its own N-node. If we
could do this, would we need to say that lexical insertion takes place all at once
at D-structure? Nothing in the lexical entries themselves or in the formulation
of the Projection Principle requires this. Since D-sructure is an internal level,
as we observed at the end of Chapter 8, perhaps it would then not be strictly
necessary. This, in turn, would raise the question of whether we need the other
internal level, namely S-structure: perhaps it is enough to simply say that the
derivation splits into its PF and LF paths at some point, nothing more. All of
these questions are interrelated and form the central object of research in post-
1990 generative grammar; collectively they form a large part of the minimalist
programme for linguistic theory. They are addressed in depth in Volume II, but
the groundwork laid here is essential for understanding their scope and import.
In our conclusion to Chapter 1, after the discussion of the fish sentences in
both English and Italian, which was intended as a demonstration simultaneously

IanRoberts_9781316519493_Conclusion.indd 217 05-09-2022 10:10:37

 218 conclusion

of tacit native-speaker competence in understanding strange sentences never

heard before and of how similar and different two languages such as English and
Italian might be, in response to the question ‘what are the properties of individ-
ual I-languages and what are the properties of UG that underlie them?’ I pointed
out that ‘this must involve the operations that are capable of producing the kinds
of structures we saw in relation to the fish-examples’. Well, now we have a
pretty good idea of what those operations are: PS-rules in the form of X′-theory,
movement rules, interpretative rules such as binding theory, parameters distin-
guishing English and Italian, and derivations giving the PFs we observe and the
LFs we understand, manipulating lexical items whose syntactic manifestations
are regulated by the Projection Principle. We have also just seen some of the
­theory-internal questions our expositions have raised. Then, of course, there is
the overarching and crucial question of how children acquire all this and how
much of it is given by the genetic endowment (UG), how much is parametrised,
how parameter settings are acquired and what is attributable to third factors.
There is clearly much to do, but we have made a solid start. As Lao Tzu said
‘The journey of a thousand miles begins with one step.’ We might have made
more than one step here, but we do have quite a journey ahead of us.

IanRoberts_9781316519493_Conclusion.indd 218 05-09-2022 10:10:38

 Glossary

A-position an XP-position which is a grammatical-function position, e.g. SpecT′, the

subject position.
A′-position an XP-position which is not a grammatical-function position, e.g. SpecC′,
the landing site for wh-movement.
A-binding a binding (q.v.) relation in which the antecedent occupies an A-position
(q.v.).
acceptability judgements given by native speakers regarding well-formedness, ambiguity,
etc., of sentences in their language are acceptability judgements rather than grammaticality
judgements. Sentences generated by the rules of the grammar are grammatical; those
judged well-formed by native speakers are acceptable. Grammaticality is a matter of
competence, while performance factors such as memory and processing interfere with
acceptability judgements. A famous example of grammatical but unacceptable sentences
is centre-embedding (q.v.).
accusative one of the case forms found in Latin, Sanskrit and many other languages. It
typically marks the direct object (q.v.) of the verb.
Adjective a lexical category (q.v.) whose typical semantic value is to describe a quality
or property. Adnominal adjectives modify nouns directly, occupying a position inside
NP, as in red bus. Predicative adjectives modify subjects from within the predicative
constituent (AP or VP), as in the bus is red.
adjunct optional, usually modificational, material, which is not c-selected (q.v.), e.g. on
Monday in I ate a pizza on Monday.
adjunction (transformation) the transformational (or movement) operation which
creates a new position attached to the target of movement. Opposed to substitution
(q.v.), which is movement to an already existing position.
Adverb a lexical category (q.v.) whose typical semantic value is to describe the manner
in which an eventuality takes place, or the speaker’s beliefs or attitudes towards that
eventuality. There are numerous subtypes of adverb, including speaker-oriented (e.g.
frankly), subject-oriented (e.g. cleverly) and manner adverbs (e.g. quickly), as well as
adverbs of time, place and reason.
Agent the ‘doer’ of an action described by a verb of the relevant type, e.g. Mary in Mary
ate her dinner. The Agent corresponds to the subject of an active clause.
anaphor according to the binding principles (q.v.), an NP subject to Principle A, e.g.
reflexive and reciprocal pronouns (q.v.).

219

IanRoberts_9781316519493_Glossary.indd 219 05-09-2022 17:47:31

 220 glossary

antecedent (of an anaphoric (q.v.) or pronominal (q.v.) expression), the nominal to

which the anaphoric/pronominal nominal is semantically connected.
arbitrary implicit argument an argument with arbitrary reference is a special kind
indefinite, often with a generic, plural, human interpretation, such as they in They eat
pasta in Italy. The implicit argument of a passive is the ‘understood’ external argument
in cases where there is no by-phrase as in Pasta is eaten in Italy.
argument structure the property of a lexical item which determines the distribution
of its thematic roles (q.v.) so that the predicate and the clause can be fully interpreted
regarding who is doing what to whom.
Basque a language spoken by approximately 750,000 people in north-western Spain
and south-western France. Basque is of great interest to linguistics, since it is probably
the best-known example of a language isolate, a language that has no proven historical
relation with any other known language.
behaviorist psychology an approach to psychology dominant in the English-speaking
world in the first half of the twentieth century and which still has many adherents. The
approach rejects introspection and regards observable behavior, and various methods for
conditioning that behavior, as central to psychological research. On this view, language
is a set of habits that can be acquired through conditioning.
binding principles the three principles which determine the anaphoric relations which
anaphors (q.v.), pronouns and R-expressions (q.v.).
binding the chief way to establish anaphoric relations, defined as c-command (q.v.)
combined with coindexing (q.v.); A-binding (q.v.) is regulated by the binding principles
(q.v.), while operator-variable binding involves binding of a variable by an operator
(q.v.) in an A′-position (q.v.).
bleeding order the order of two rules, A and B, such that A applies before B and removes
the structural environment for the application of B, preventing B from applying. For
example, English T-to-V lowering in PF removes the context for do-insertion.
branch a line connecting two nodes (q.v.) in a tree diagram (q.v.).
C(ategorially)-select part of the lexical entry (q.v.) of a lexical item which specifies the
syntactic category of the item’s complement (if any). For example, transitive verbs have
the c-selection frame __ NP, indicating that they take a single NP complement (q.v.).
C(onstituent)-command the configuration relation defining relative structural
prominence between syntactic positions. The relation plays a fundamental role in many
aspects of syntax, including, notably, binding.
case a class of inflectional endings (typically) which attach to nouns in many languages
to mark the grammatical function of an NP in a clause.
categorial feature a class of features (N, V, etc.) defining lexical categories (q.v.) and
functional categories (q.v.).
category a class of words sharing the same distribution, morphological markings and (roughly)
semantic properties, e.g. noun, verb, etc. Roughly corresponds to the traditional notion of ‘art
of speech’. Standardly divided into lexical (q.v.) and functional categories (q.v.).

IanRoberts_9781316519493_Glossary.indd 220 05-09-2022 17:47:31

 Glossary 221

centre-embedding relative clauses formed on an object which are embedded in subject

position and therefore follow the head noun and precede the verb of the main clause. If,
as is possible with object relatives in English, no relative marker is present, the string
N N V V results with one relative clause, e.g. the mouse the cat caught died. Adding a
further relative of the same kind on the second N gives rise to sentences which are very
hard to process and often judged unacceptable, although they must be grammatical, e.g.
the mouse the cat the dog chased caught died.
classifiers markers of noun classes with an approximate semantic basis, typically along
the lines of animacy, shape and other properties. Very common in the languages of East
Asia, including Mandarin and Japanese, e.g. san-ge xuesheng ‘three Classifier student’,
i.e. ‘three students’.
clause a minimal sentence, containing a subject and a predicate. Not all clauses can stand
alone: infinitives generally cannot (e.g. John to be the best in I believe John to be the
best).
clefting a syntactic operation which emphasises (or focuses) a constituent XP by placing
it in the frame It was XP that S, where S contains a gap corresponding to the neutral
position of XP. A standard constituency test (q.v.) in English.
coindexing a notation for indicating anaphoric relations between NPs (and possibly
heads). Each referential NP has a constant index, written i, j, k …, and quantifiers are
associated with variable indices. Combined with c-command (q.v.), gives rise to binding
(q.v.).
complement a general term for a c-selected (q.v.) internal argument (q.v.) of a head X.
competence the tacit knowledge an adult has of their language. Competence is usually
contrasted with performance, which refers to the actualisation of competence in
behaviour.
Complex NP Constraint (CNPC) an island constraint (q.v.) which bans wh-movement
(q.v.) out of an NP containing a CP which modifies or is the complement of the head
noun.
conjugation morphological classes of verbs; in Latin and the Romance languages, the
three or four morphological classes formed by theme vowels.
constituency tests manupulations of syntactic structure which can only apply to XPs
(often a subclass of XPs), designed to validate proposed constituent structure and
associated Phrase-Structure rules (q.v.). The tests include movement or permutation
operations, substitution of pro-forms (q.v.), ellipsis (q.v.), coordination (q.v.) and
fragments (q.v.).
constituency the fundamental syntactic relation, specified by Phrase-Structure rules
(q.v.) and indicated by dominance (q.v.) relations in tree diagrams (q.v.).
constituent a category (q.v.) A is a constituent of another category B just where a
continuous set of branches (q.v.) can be traced from A to B going consistently ‘upward’
in the tree diagram (q.v.); the inverse of dominance (q.v.).
constituent structure the core set of relations in syntax, generated by Phrase-Structure
rules (q.v.) and represented equivalently by labelled bracketings or tree diagrams (q.v.).

IanRoberts_9781316519493_Glossary.indd 221 05-09-2022 17:47:31

 222 glossary

constructed languages (conlangs) languages which have been deliberately constructed

for one purpose or another rather than developing naturally by historical generation-to-
generation transmission.
coordination the combination of two words or phrases with and (canonically). Can serve
as a constituency test (q.v.) for XPs.
copy the occurrence of a moved category which remains in the original, pre-movement
position after movement has taken place. Copies are usually deleted by the copy-deletion
rule (q.v.).
copy-deletion rule the rule which deletes copies of movement.
dative one of the case forms found in Latin, Sanskrit and many other languages. It
typically marks the indirect object (q.v.) of the verb.
D(eep)-structure the level of syntactic representation where c-selection and s-selection
specifications of lexical entries are directly realized in phrase structure. D-structure is
generated by Phrase-Structure rules (q.v.), but no movement applies prior to this level.
declarative the clause type used for making statements, opposed to interrogative,
imperative and exclamative, among others.
definite article marks an NP as definite (the in English). A definite NP refers to
something which is taken to exist and to be known to both speaker and hearer in the
context. The definite article also implies exhaustive reference to the members of the set
of entities denoted by the noun.
direct object an NP which typically marks the ‘undergoer’ of an action described by the
verb, and forms part of the VP with the verb.
disjunctive ordering the principle of ordering two rules A and B such that the rule with
the more specific structural description precedes the more general one. The more general
rule may be the elsewhere (q.v.) case.
distribution (as a test for categories) a way of identifying and distinguishing categories
according to the syntactic positions they are able, and unable, to appear in, i.e. how a
given category is distributed across the overall set of syntactic positions the grammar
makes available. NPs, for example, are almost the only category able to appear both in
subject position and as the complement of a preposition.
dominance the inverse of constituency (q.v.): a category (q.v.) A dominates another
category B just where a continuous set of branches (q.v.) can be traced from A to B
going consistently ‘downward’ in the tree diagram (q.v.).
do-support the phenomenon peculiar to English which involves the obligatory appearance
of the meaningless dummy auxiliary do in main-clause interrogatives, negative clauses
and VP-ellipsis (q.v.) contexts where no other auxiliary (modal, have or be) is present.
The phenomenon is discussed in detail in Section 8.6.
echo questions wh-questions (q.v.) in which the wh-phrase (q.v.) is not moved, but
appears in situ, typically associated with an interpretation of surprise or disbelief on the
part of the speaker.
E-language external or extensional conceptions of language. Opposed to I-language
(q.v.). Examples of E-language are texts, corpora, or in fact any occurrence of language

IanRoberts_9781316519493_Glossary.indd 222 05-09-2022 17:47:31

 Glossary 223

which is not I-language. The everyday concepts of languages such as English, French,
etc. are E-language concepts, since they are largely social, cultural and historical.
ellipsis elided categories are ‘understood’, i.e. they are present in the semantic
interpretation (in some way) but absent in the phonological representation (i.e. silent).
Can be used as a constituency test (q.v.) for XPs. A prominent type of ellipsis in English
is VP-ellipsis (q.v.).
elsewhere the default, non-specified or minimally specified context for the application of
a rule; for example +d for English Past Tense.
endocentric a phrasal category which has a head (q.v.). According to the X′-schema for
PS-rules (q.v.) all categories are endocentric.
ethnolect a language variety associated with a specific ethnic group. It may be a
distinguishing mark of social identity, both within the group and for outsiders. Probably
the best-studied ethnolect is Afro-American Vernacular English (AAVE).
evidential markers clausal markers which indicate the speaker’s commitment to the
evidential basis for what is asserted by the clause: typical evidentials mark direct vs
indirect knowledge, witnessed or not witnessed by the speaker, etc.
exocentric a phrasal category which has no head (q.v.). According to the Xʹ-schema for
PS-rules (q.v.) no categories are exocentric.
Experiencer a thematic role designating the sentient being affected by an emotion or
psychological state, one of the arguments of a psych-verb (q.v.).
external argument the s-selected argument of a head which appears outside the maximal
projection of that head. External arguments are never c-selected. Agents are always
external arguments, but external arguments are not always Agents.
feeding order the order of two rules, A and B, such that A applies before B and creates
the structural environment for the application of B, making it possible for B to apply. For
example, movement rules feed copy deletion by creating the environment in which copy
deletion applies.
finite finite clauses contain a tense-marked verb or auxiliary. Main clauses are typically
finite; declarative (q.v.) main clauses always are.
first person the type of personal pronoun or person inflection on a verb used to refer to
the speaker or (in the plural) a group including the speaker, I/me/we/us in English.
fragment a piece of a sentence, often occurring the answer to a question, possibly
derived from one or more applications of ellipsis (q.v.). Can be used as a constituency
test (q.v.) for XPs.
fronting a syntactic operation which places a constituent XP at the beginning of the
sentence; the remainder of the sentence contains a gap corresponding to the neutral
position of XP. A standard constituency test (q.v.) in English.
functional categories the subclass of categories (q.v.) that represent grammatical
rather than lexical information. Functional categories tend to have rather few members,
which are frequently phonologically reduced, affixal, or silent, and to be closed classes.
Some seem to lack a semantic value, others have complex or rather abstract semantics.
Examples include auxiliaries, complementisers and determiners.

IanRoberts_9781316519493_Glossary.indd 223 05-09-2022 17:47:31

 224 glossary

gender a grammatical category distinguishing classes of nouns. As mentioned in the

Introduction, in many languages, including Italian and the other Romance languages,
gender is grammatical, not semantic; in other words, it has no connection with biological
or social definitions. In Italian, all nouns have masculine or feminine gender, and this has
no semantic basis at all in many cases: ‘sun’ is masculine, ‘moon’ is feminine; ‘sincerity’
is feminine, ‘feminism’ is masculine; ‘violence’ is feminine, ‘socialism’ is masculine
and so on. Nouns are arbitrarily assigned to two distinct morphological classes.
generate technically a way to enumerate the members of a set, e.g. the members of
the set of grammatical sentences of English. Phrase-Structure rules (q.v.) generate
constituent structure. The term gives its name to the theory of grammar adopted here,
generative grammar (q.v.).
generative grammar the theory of language put forward by Noam Chomsky. Generative
grammars recursively enumerate the set of well-formed expressions in a language, along
with their structural descriptions. The well-formed sentences of the language are akin to
the theorems of a deductive system, and the rule systems that specify them are akin to
rules of inference. The theory of generative grammar aims to provide an account of the
human language faculty (q.v.) by specifying the class of humanly attainable grammars.
generic a use of NPs, typically as subjects, to express general statements, e.g. Fish live
in water. The most natural case is with bare indefinite (q.v.) plurals, as here, but singular
definites can also be generic: The fish is an aquatic animal. Generics are usually taken to
denote natural kinds (q.v.).
Germanic a branch of Indo European (q.v.) spoken natively by a about 515 million
people mainly in north-western Europe. All Germanic languages are derived from Proto
Germanic, one of the daughters of Indo European.
harmonic word-order patterns word-order sequences in which all, or at least the
clear majority, of phrasal categories show a consistent value for the Head Parameter
(q.v.). Japanese is an example of a harmonically head-final (q.v.) language, French of a
harmonically head-initial (q.v.) one. Languages with mixed patterns are disharmonic; a
well-known example is German, as discussed in Section 7.5.
head a lexical item, functional or lexical, of category X which combines with a
complement and forms an intermediate X′ projection (q.v.) of the same category.
Head Parameter the parameter of variation (q.v.) determining whether the head X
precedes or follows its complement in X′. Where the choice is ‘precede’, XP is head-
initial, where the choice is ‘follow’, XP is head-final.
head-final (order/language) the case where the Head Parameter (q.v.) is set to ‘follow’.
Where all or most XPs show this setting the pattern is harmonic (q.v.).
head-initial (order/language) the case where the Head Parameter (q.v.) is set to
‘precede’. Where all or most XPs show this setting the pattern is harmonic (q.v.).
head-movement a movement operation which copies a head into another head position,
as in English subject-auxiliary inversion.
idiolect a language variety characteristic of an individual. It is extremely plausible that
all individual idiolects are different in one respect or another.

IanRoberts_9781316519493_Glossary.indd 224 05-09-2022 17:47:31

 Glossary 225

I-language the internal, intensional, individual conception of language. Generative

grammar (q.v.) aims to provide a theory of I-language by providing a theory of the
language faculty, which makes individual I-languages possible.
immediate constituency a category (q.v.) A is an immediate constituent of another
category B just where a continuous set of branches (q.v.) can be traced from A to B going
‘upward’ in the tree diagram (q.v.), and there is no category C such that C intervenes
between A and B, i.e. there is no node (q.v.) C such that B is a constituent of C and A is
not a constituent of C; the inverse of immediate dominance (q.v.).
immediate dominance a category (q.v.) A is immediately dominates another category B
just where a continuous set of branches (q.v.) can be traced from A to B going ‘downward’
in the tree diagram (q.v.), and there is no category C such that C intervenes between A
and B, i.e. there is no node C such that C dominates B and C does not dominate A; the
inverse of immediate constituency (q.v.).
indefinite indefinite NPs are marked by the indefinite article a(n) in English, where the
head noun is a count noun: a cat is on the mat. Indefinite mass nouns have no singular
article: milk is good for you. Generally, indefinite NPs have none of the properties of
NPs marked with the definite article (q.v.) and so may be used with no commitment to
the existence, the familiarity or the exhaustive reference of the set denoted by the noun.
indicative mood the mood used to express statements that the speaker believes to be
factually correct. In many languages, the indicative contrasts with the subjunctive (used
to express various degrees of doubt, unlikelihood, unreality, etc.), but in Modern English
this contrast barely survives in the verb-inflection system. Modern English more readily
employs modal auxiliaries in these situations: John may/might/could/should be home by
now.
indirect object an NP or PP which typically marks the goal or beneficiary of the action
described by the verb, and forms part of the VP with the verb. Indo European a large
family of languages spoken across western and southern Eurasia. Most, but not all, of
the languages of Europe, Northern India and Iran are Indo European. The Indo-European
family is divided into several branches or sub-families: Albanian, Anatolian, Armenian,
Balto-Slavic, Celtic, Germanic (q.v.), Hellenic, Indo-Iranian, Italic and Tocharian.
infinitival clause in English, clauses marked with the infinitive to, a subset of non-finite
clauses, those which lack a tense-marked verb or auxiliary (see finite)
inflectional endings one way of forming complex words, which involves adding a bound
morpheme at the end of a root to indicate grammatical information such as tense on verbs
or case on nouns. English has a rather impoverished set of inflections compared to many
other Indo-European languages.
interface levels those levels of representation which connect narrow syntax (q.v.) to
non-syntactic parts the grammar. Phonological Form (PF) (q.v.) connects narrow syntax
to phonology and Logical Form (LF) (q.v.) connects it to semantics.
Interlingua a constructed language ‘developed between 1937 and 1951 by the American
International Auxiliary Language Association (IALA). It ranks among the most widely
used IALs and is the most widely used naturalistic IAL [footnote omitted] – in other
words, those IALs whose vocabulary, grammar and other characteristics are derived from

IanRoberts_9781316519493_Glossary.indd 225 05-09-2022 17:47:31

 226 glossary

natural languages, rather than being centrally planned. Interlingua literature maintains
that (written) Interlingua is comprehensible to the hundreds of millions of people who
speak Romance languages [footnote omitted], though it is actively spoken by only a few
hundred’ (https://en.wikipedia.org/wiki/Interlingua, accessed 14/9/21).
internal argument the c-selected (q.v.) argument(s) of a head X which form its
complement(s) (q.v.) inside X′. Agents are never internal arguments.
interpretative rule rules which specify relations between syntactic constituents which
affect semantic interpretation, distinct from both Phrase-Structure (q.v.) and movement
rules (q.v.). Binding of various kinds is a prominent example.
intervention the case where a syntactic relation between A and B may be blocked if there
is a third element ‘in between’ A and B. Very frequently defined in terms of c-command
(q.v.), i.e. C intervenes between A and B where A c-commands C and B, C c-commands
B and C does not c-command A; see (31) of this Chapter.
intrinsic ordering the order of two rules, A and B, such that one can only precede the
other. For example, the feeding order (q.v.) of movement rules and copy deletion is a
case of intrinsic ordering, as there are no copies to delete before movement applies.
island constraints a class of restrictions on wh-movement (q.v.) including the Complex
NP Constraint (q.v.), the Left Branch Constraint (q.v.) and wh-islands (q.v.). Island
constraints do not permit wh-movement to move a wh-phrase (q.v.) from a position
inside the island to a position outside it.
labelled bracketing in syntax, one of the standard ways of displaying constituent
structure, along with tree diagrams (which are equivalent).
language acquisition the process by which people acquire the ability to speak and
understand a language. First-language acquisition takes place in early childhood. From
the perspective of generative grammar (q.v.), first-language acquisition involves the
development of the language faculty (q.v.) on the basis of the three factors of language
design (q.v.).
language faculty whatever cognitive structure underlies our ability to acquire our native
language, to store the knowledge so acquired in the mind/brain and to put it to use in
production and comprehension. In generative grammar (q.v.), Universal Grammar
(q.v.) is the theory of the genetic endowment underlying the human language faculty.
Left Branch Constraint (LBC) an island constraint (q.v.) which prevents a wh-phrase
(q.v.) moving off a left branch, e.g. SpecN′ in English possessives.
lexical category a category (q.v.) whose head is a substantive lexical item, e.g. a noun,
verb or adjective. Lexical categories are open classes, i.e. it is possible to invent new
members and seemingly always have a phonological form, sometimes quite a complex
multisyllable one. Opposed to functional categories (q.v.).
lexical entry the specification in the lexicon of all and only the idiosyncratic properties
of a lexical item. Lexical entries are discussed in more detail in Section 8.3.
lexical insertion the operation which places lexical items in D-structure (q.v.). As
presented in Section 8.4, this is an operation which substitutes a lexical item into a head
position at that level.

IanRoberts_9781316519493_Glossary.indd 226 05-09-2022 17:47:31

 Glossary 227

Lexicon the mental dictionary, in which, in principle, all and only idiosyncratic
information about individual lexical items is listed. This information includes at least a
specification of s-selection, c-selection, phonological and semantic properties.
Linnean taxonomy a system of biological classification based on the system established
by the eighteenth-century Swiss biologist Carl Linnaeus. Linnaeus introduced the system
of two-part Latin names for species such as Homo sapiens for humans and felix catus for
domestic cats.
Logic the study of the laws of valid inference, relevant for semantics (q.v.), since sound
inferences are truth-preserving, and truth provides a way to elucidate meaning in truth-
conditional semantics.
Logical Form (LF) the level of representation which converts the syntactic representation
into a representation which can be the input to the semantic representation.
marked the marked value of an asymmetric opposition is the more complex term of
that opposition. Parameters of variation (q.v.) may have marked and unmarked values,
or perhaps a combination of parameter settings may give rise to a marked system, as
suggested in Section 7.4 for the rarer sentential word orders.
maximal projection the phrasal projection of a head (q.v.) X and the intermediate
projection X′. Can be defined as the projection of X immediately dominated by Y X.
morphology the study of the internal structure of complex words.
movement rule one kind of transformational rule (q.v.) which copies categories into a
c-commanding position.
multiple questions questions featuring more than one wh-phrase (q.v.). In English,
since only one wh-phrase can move per clause, the other(s) must remain in situ.
narrow syntax the collective term for the ‘internal’ syntactic levels of representation,
D-structure (q.v.) and S-structure (q.v.), which do not interface with non-syntactic
parts of the grammar.
natural kind ‘To say that a kind is natural is to say that it corresponds to a grouping that
reflects the structure of the natural world rather than the interests and actions of human
beings. We tend to assume that science is often successful in revealing these kinds; it is a
corollary of scientific realism that when all goes well the classifications and taxonomies
employed by science correspond to the real kinds in nature’, Stanford Encyclopedia of
Philosophy (https://plato.stanford.edu/entries/natural-kinds/#SemNatKinTer, accessed
14/9/21).
Navajo a Southern Athabaskan language spoken by 170,000 people primarily in the
Southwestern United States.
negative concord the phenomenon, found in French and many other languages, whereby
the co-occurrence of two negative morphemes in a sentence does not give rise to a double
negation but rather to single negation, also found in many varieties of non-standard
English.
node a position in a tree diagram (q.v.) from which either at least one branch (q.v.)
emanates ‘downwards’ (a non-terminal node (q.v.)) or a site of lexical insertion (q.v.).

IanRoberts_9781316519493_Glossary.indd 227 05-09-2022 17:47:31

 228 glossary

nominative one of the case forms found in Latin, Sanskrit and many other languages. It
typically marks the subject (q.v.) of a sentence.
non-terminal node a node (q.v.) in a tree diagram (q.v.) from which at least one branch
emanates ‘downwards’ and hence which dominates (q.v.) at least one other node (q.v.).
Noun Phrase (NP) a phrase which contains a noun or pronoun and occupies the same
syntactic positions as a noun or pronoun (subject, direct object, object of preposition,
etc.).
NP-movement the operation which moves an object NP into the subject position in
passives.
number the inflectional category which distinguishes singular from plural, marked on
nouns in many languages, and on verbs in quite a few (in English, only in the present);
some languages, e.g. Mandarin, have no obligatory number marking (either on nouns or
verbs).
operator roughly equivalent to quantifier (q.v.), a category which binds a variable from
an Aʹ-position (q.v.), giving rise to quantificational interpretation. Wh-phrases are a
typical example.
operator-variable structures a syntactic configuration in which an operator (q.v.)
or quantifier (q.v.) binds a variable from an Aʹ-position (q.v.), giving rise to a
quantificational interpretation.
pair-list interpretation the usual interpretation of multiple wh-questions, in which the
answers associated with each wh-phrase are paired, e.g. John drank beer and Paul drank
whisky in answer to Who drank what?
parameters of variation a way to define and analyse syntactic variation among languages.
A given grammatical system (I-language) may be seen as the set of parameter values
characteristic of that system, combined with the invariant aspects of Universal Grammar.
Patient the thematic role (q.v.) of the argument that is affected by the action described
the verb, e.g. her dinner in Mary ate her dinner.
performance putting competence (q.v.) into practice in production and comprehension;
performance involves competence, combined with short-term memory, attention and
other non-linguistic cognitive capacities.
person the inflectional category which disambiguates between first, second and third,
marked on pronouns in very many languages, and on verbs in many languages (in English
only in the third-person singular of the present).
Phonological Form (PF) the level of representation which converts a syntactic
representation (S-structure (q.v.), given the standard architecture seen in (1) of this
chapter) into a representation which can be input to phonological and phonetic rules and
representations.
phonology the branch of linguistics dealing with how speech sounds are organized in
linguistic systems.
phrasal category in syntax (q.v.?), a unit of organization of words, intermediate
between the word level and the sentence level; phrasal categories are non-root, non-
terminal nodes (see Chapter 3).

IanRoberts_9781316519493_Glossary.indd 228 05-09-2022 17:47:31

 Glossary 229

phrasal stress the accent (marker of phonological prominence through pitch, loudness,
length or some combination of these) which falls on the designated most prominent
syllable in a phonological phrase (often close to a syntactic XP), as in Oxford Róad.
phrase marker synonymous with tree diagram (q.v.) or labelled bracketing, the means
of displaying constituent structure (q.v.).
Phrase-Structure (PS) rules rules which generate (q.v.) constituent structure (q.v.).
The recursive nature of these rule systems accounts for the fact that natural languages
make infinite use of finite means.
pied-piping a case of wh-movement (q.v.) where the moving wh-phrase (q.v.) takes
along other parts of a constituent to which it belongs in its pre-movement position.
postposition the category P is postpositional if it has the head-final (q.v.) value of the
Head Parameter (q.v.); if head-initial (q.v.) it is prepositional. Strictly speaking, P
should be taken as the abbreviation for the order-neutral term ‘adposition’.
predicate logic the branch of logic which breaks propositions (roughly, logical-semantic
representations of sentences) down into predicates and arguments; a key aspect of
predicate logic is its ability to represent quantification.
predicate traditionally seen as the part of a sentence which describes a property or action of
the subject, typically corresponding to VPs, PPs or APs, although NPs can be predicates too.
Preposition a lexical category (q.v.) typically denoting a location or path and taking
an NP complement. Unlike other lexical categories, Prepositions are closed-class items.
Preposition stranding (q.v.) of the complement of a preposition without associated
pied-piping (q.v.) of the preposition. Found in English (e.g. in Who did you talk to?) but
cross-linguistically rare.
prescriptive grammar ‘the establishment of rules defining preferred or correct usage
of language … such normative practices often suggest that some usages are incorrect,
inconsistent, illogical, lack communicative effect, or are of low aesthetic value, even in
cases where such usage is more common than the prescribed usage’ (https://en.wikipedia
.org/wiki/Linguistic_prescription, accessed 14/9/21).
present tense a form of the verb which, in English, is just the root in all persons except
the third-person singular, where it is marked by -s. The present tense situates the event or
situation described by the sentence in the present time.
pro-form a reduced form of an XP that can substitute for that XP in the appropriate
context. Pronouns are pro-NPs, do so can function as a pro-VP and so as a pro-AP.
progressive a verbal aspect indicating, roughly, ongoing action. Formed by adding -ing
to the verb and marking tense and agreement inflection on the auxiliary be, as in John
was singing.
projection the occurrences of the category of a head (q.v.) which contain further material in
addition to the head: the X′ projection contains the head and its lexically specified complement,
and the maximal projection (q.v.) XP contains the X′ projection and the Specifier (q.v.).
Projection Principle the principle which conditions lexical insertion (q.v.) and holds
lexical properties constant through the syntactic derivation by requiring these properties
to hold at all levels. A particular formulation is given in (25) of Chapter 8.

IanRoberts_9781316519493_Glossary.indd 229 05-09-2022 17:47:31

 230 glossary

pronoun an underspecified NP, bearing a particular anaphoric relation to its antecedent

(q.v.), if there is one.
proposition a semantic representation of a sentence, which bears a truth value.
psych-verbs a class of verbs denoting psychological, mental or emotional events or
states, typically having an Experiencer (q.v.) and a Theme, or Cause, argument. At first
sight, the two arguments both seem able to appear as either subject or object, depending
on the verb, as in John fears snakes (Experiencer-Theme) vs Snakes frighten John
(Theme-Experiencer).
quantifier a category involved in expressions of generality which are not predicated
of individuals; a quantifier binds a variable in an A-position (q.v.) from an Aʹ-position
(q.v.).
quantifier-scope ambiguities ambiguities involving at least two quantifiers in which
the different interpretations result from the apparent logical scopal relations between the
quantifiers, as in Every student bought some new book, where there could be a different
book for each student or each student could have bought the same book.
R(eferring)-expression NPs with independent semantic content, subject to Binding
Principle C (q.v.).
reciprocal pronoun a class of pronouns subject to Binding Principle A (q.v.), e.g. each
other. Owing to their semantics, reciprocals always require a plural antecedent.
reconstruction the case where movement, typically wh-movement, distorts the
c-command relations required for binding interpretations, e.g. of reflexives inside
complex fronted wh-phrases. Postulating that movement always leaves copies offers a
good account for many such phenomena.
recoverability a condition on elided (deleted) constituents: there must be a way to
‘recover’ their content such that they can be ‘understood’, i.e. accessible to semantic
interpretation.
recursion the property of Phrase-Structure rules (q.v.) that allows them to apply to
their own output and thereby create an unbounded number of well-formed sentences.
reflexive pronoun a class of pronouns subject to Binding Principle A (q.v.), e.g. himself.
In English, all reflexives end in -self.
relative clause a complex modifier of a noun or NP which contains a disguised sentence,
as in [ the person [RelativeClause who I met yesterday ].
root the node (q.v.) in a tree diagram (q.v.) from which all the other nodes emanate. As
such, the root dominates (q.v.) every category in the tree diagram, and every category in
the tree diagram is a constituent (q.v.) of the root.
S(urface)-structure the level of syntactic representation derived from D-structure
(q.v.) by movement rules, after which the derivation divides into the two distinct paths to
Phonological Form (q.v.) and Logical Form (q.v.).
scope the domain over which a quantifier/operator acts, typically by binding variables.
Scope is defined in terms of c-command: a quantifier/operator has scope over the
constituents it c-commands.

IanRoberts_9781316519493_Glossary.indd 230 05-09-2022 17:47:31

 Glossary 231

second-person pronoun the type of personal pronoun used to refer to an addressee, you
in English. English does not distinguish the singular and plural of this person, neither
does it make politeness distinctions indicating the status of the addressee, as many other
languages do.
semantics the study of meaning, particularly sentence meaning.
sign language language(s) transmitted through the visual-gestural rather than the oral-
aural medium, mostly used by deaf communities; sign languages are now known to be
languages in every sense of the word, having all the salient structural properties of oral-
aural languages.
sociolect a variety of language used by a particular socioeconomic class.
Specifier the category (itself a maximal projection (q.v.)) which combines with
the intermediate X′ projection to form the maximal projection (q.v.). Specifiers are
typically, but not only, modifiers in a rather broad sense.
string-vacuous movement movement of a category in such a way that the constituent
structure is changed, but the linear order of terminals is not. The best-known example is
wh-movement (q.v.) of subjects in English, as in Who left?
strong crossover the configuration where a wh-phrase moves ‘over’ a pronoun, which is
unable to take it as its antecedent. Falls under Binding Principle C (q.v.); see Section 6.7.
structural description the precise description of the structure of a syntactic object,
usually a sentence. The standard form of such a description is in terms of constituent
structure (q.v.) generated by Phrase-Structure rules (q.v.) and transformational rules
(q.v.).
Subject an NP which typically marks the Agent (q.v.) of an action described a VP.
substitution (transformation) a movement rule copying a category into a position
which has been created by the Phrase-Structure rules (q.v.). Both NP-movement
(q.v.) and wh-movement (q.v.) substitute into Specifier positions (SpecT′ and SpecC′
respectively).
syntax the study of the structure of sentences.
terminal node a node (q.v.) in a tree diagram (q.v.) which dominates (q.v.) no other
node; the site of lexical insertion.
thematic role lexico-semantic properties of predicates which designate the way in which
their arguments participate in the eventuality they describe. An important component of
argument structure (q.v.).
third-person plural the third-person plural pronoun in English is they/them/their.
Generally used to refer to a group of known individuals who are neither the speaker nor
the addressee(s), although it may also be used as an arbitrary pronoun (they eat pasta in
Italy) or as a singular pronoun to avoid reference to gender. In many languages, verbs
have special forms for the third-person plural distinct from other verb forms; English
does not, however.
three factors in language design the three elements that together constitute the adult
language faculty: (i) the genetic endowment (Universal Grammar), (ii) the primary

IanRoberts_9781316519493_Glossary.indd 231 05-09-2022 17:47:31

 232 glossary

linguistic data which acts as the input to language acquisition, (iii) third factors of a
general cognitive nature.
Topic typically an XP that has undergone fronting (q.v.). Topics generally stand for
‘old’ or known information, something already talked about or given in the context.
trace a cover term for a position from which movement takes place. More precisely,
copies (q.v.) occupy these positions.
transformational rule syntactic rules which map phrase markers (q.v.) into other
phrase markers. Deletion and movement rules (q.v.) are types of transformations.
tree diagram one of the standard ways of displaying constituent structure (q.v.), along
with labelled brackets. Synonymous with phrase marker (q.v.).
truth conditions the standard way of understanding sentence meaning; the central idea is
that to know the meaning of a declarative sentence is to understand what the world would
have to be like for the sentence to be true.
unbounded dependency the property of wh-movement (q.v.) which gives the
appearance of movement across arbitrarily long stretches of material.
Undergoer a further thematic role, almost synonymous with Patient (q.v.).
Universal Grammar (UG) the initial state of the language faculty, assumed to be
genetically endowed. The first fator in language design.
universal quantifier a quantifier, written " in predicate logic (q.v.), corresponding
roughly to every, each or all in English. Its semantics involves set-inclusion, i.e. Every A
is B means that all members of A are included in B but not necessarily vice versa.
verb second the syntactic constraint operative in German (see Section 7.5) and other
Germanic languages, but not English, which requires that the finite verb or auxiliary be
preceded by exactly one XP in declarative main clauses.
Volapük a conlang created by the German Catholic priest Johann Martin Schleyer in
1879–80. It was later displaced by Esperanto.
VP-ellipsis a kind of ellipsis (q.v.) which affects VPs. Quite common in English, and a
good constituency test (q.v.) for VPs.
V-to-T Parameter the parameter which determines whether the lexical verb V raises to
T to combine with tense-agreement marking. Positive in French and German; negative
in English.
Warlpiri an Australian language spoken by about 3,000 people Australia’s Northern
Territory. It is a member of the Pama–Nyungan family.
weak crossover a further configuration where a wh-phrase moves ‘over’ a pronoun,
like strong crossover (q.v.). In this case, though, the crossed-over pronoun does not
c-command (q.v.) the variable bound by the wh-phrase. The effect of unacceptability is
weaker than in the case of strong crossover, hence the name.
wh-islands an island constraint (q.v.) which prevents wh-movement (q.v.) out of a CP
whose Specifier is already filled by a wh-phrase (q.v.).

IanRoberts_9781316519493_Glossary.indd 232 05-09-2022 17:47:32

 Glossary 233

wh-movement the operation which moves a wh-phrase (q.v.) to a SpecC′ position,

leaving a copy (q.v.) in the original position. Can be used as a constituency test (q.v.). A
very important operation, which later chapters look at in detail.
wh-movement parameter the parameter which determines whether wh-phrases move to
SpecC′ in wh-questions. Positive in English; negative in Mandarin.
wh-phrase an XP either consisting just of a wh-word (e.g. who, what, etc.) or having a
wh-word in its Specifier (e.g. which man, how big, etc.).
wh-question an interrogative CP with wh-movement (q.v.) to its Specifier.
word stress the accent (marker of phonological prominence through pitch, loudness,
length or some combination of these) which falls on the designated most prominent
syllable in a phonological word (often close to a syntactic head X), as in Óxford Street.
Xʹ-schema (for PS-rules) the category-neutral template that restricts PS-rules to the
three given in (2) of Chapter 3. This makes the PS-rules highly abstract but also tightly
constrained.

IanRoberts_9781316519493_Glossary.indd 233 05-09-2022 17:47:32

 Index

A-binding, 151–3 coindexing, 131–2

A-movement, 216–17 pronouns, 130–2, 142–3
A’-positions, 151–3 R-expressions, 143–4
acceptability variables, principle C and movement, 146–53
and grammaticality, 35–6, 55 bleeding order, 200
native-speaker competence, 12, 217–18 body language, 6–7
pro-form test, 84–5 bracketing see labelled bracketing
relative clauses, 27–8 branches (tree diagrams), 46
accusative, 105, 131 building rules, 100
accusative objects, 39 building structure, 119
accusative singular, 39 see also Phrase-Structure rules
active–passive pairs, 102–4
Adjective Phrase (AP) case (in English), 130–1
constituency tests, 69–75 case marking, 39
syntactic diagnostics, 40 Catalan, 8
adjectives categorial features, 86
internal arguments, 184–5 categories, 36
lexical categories, 36 identification of, 44
morphological diagnostics, 39 lexical and functional, 36
adjunction transformations, 121 morphological diagnostics for, 38–9
adjuncts, 65, 95–6 phonological diagnostics, 42
adpositions, 36–7 and semantics, 42–3
adverbs, 39 syntactic diagnostics, 39–42
lexical categories, 36 causative reading, 107–8
word order, 159 centre-embedding, 27–30
Agent (argument structure), 102 child language, 201
agreement Chinese
functional categories, 37–8 V-to-T movement, the Head Parameter and
wh-movement, 109–11 wh-movement, 175–6
anaphors, 131–41 wh-movement, 172–4, 206–8
antecedent–anaphor relations, 132–41 classifiers, 157
antecedents, 63–71 clause structure, 216
arbitrary implicit argument, 193–4 clause type, 110–11
argument structure, 102–104, 184–5 clauses, 23–4
artificial languages, 4–7 wh-movement, 109–11
auxiliaries, 39–42 X′-theory and functional categories,
child language, 201 87–95
subject-auxiliary inversion, 90, 101–2 clefting, 61–3
types of movement rules, 100–4 cognitive perspective on language, 9–12
wh-movement, 109–11 coindexing, 131–2, 134
X′-theory, 89–90 competence
native speakers, 12, 217–18
binding, 130 and performance, 30
anaphors, 132–41 complement-clause structure, 57–9
the binding principles, 144–6 Complementiser Phrase (CP), 94–5, 110

234

IanRoberts_9781316519493_Index.indd 234 10-09-2022 16:10:18

 Index 235

complements exocentric phrases, 81–2

Phrase-Structure rules, 82–7 Experiencers, 181–2
wh-movement, 114–15 external arguments, 181–2, 184–5
Complex NP Constraint (CNPC), 124–5
computer-programming languages, 5 feeding order, 200
conjugation, 18–19 fictional languages, 6
constituency, 46–7, 215 finite subordinate clauses, 58–9
constituency tests, 59–75, 87–95 first person, 130–1
constituent-command (c-command), 135–9 fish
intervention, 139–40 lexical entries, 186–7
movement, 147 six- and seven-fish sentences, 25–30
constituent structure, 36 and some relative clauses, 22–5
hierarchical relations, 45–7 tacit knowledge, 15–22
phrasal constituents, 44–5 formal theory, 11
constructed languages, 6 fragments, 61
coordination, constituency tests, 61, 68, 71 constituency tests, 68, 69
copy-deletion rule, 101–2 X′-theory, 89–90
copy, head-movement, 101–2 French
culture, language as part of, 4 V-to-T movement, the Head Parameter and
wh-movement, 175–6
D-structure (deep structure), 180, 187–90, 211–12 verb positions, 157–60
Logical Form (LF), 203–11 fronting, 61–2, 65
Phonological Form (PF), 195–203 functional categories, 36, 157
dative singular, 39
declarative, 39–42, 58–9, 100–1, 110–11 gender (in English), 130–1
declension, 39 gendered nouns, 20–21
definite article, 17–18 generic reading of NP, 21–2
deletion rules, 100 German
dependency relations, 215 dialects, 4–5
determiners, 37–8 syntax in, 165–71
dialects, 8–9 V-to-T movement, the Head Parameter and
disjunctive ordering, 185–6 wh-movement, 175–6
distributional tests, 39–42 grammar, architecture of, 179, 211–2
do-support, 92–3 D-structure, 180, 187–90
dominance, hierarchical relations, 46–7 the lexicon, 179–87
Logical Form (LF), 180, 203–11
E-language, 10–12, 31 a model of grammar, 179–80
echo questions, 106 Phonological Form (PF), 180, 195–203
ellipsis S-structure, 180, 190–5
constituency tests, 61, 66–8, 72 see also Universal Grammar (UG)
X′-theory, 84–6, 89–90 grammaticality
endocentric phrases, 81–2 and acceptability, 35–6
English binding theory, 131–2
comparison with Italian, 17–22, 31–2, 217–18 competence, 27–8
as cultural concept, 4 Phrase-Structure rules, 52–5, 75–6
defining Standard English, 8–9 recursion, 55–9
emergence of Standard English, 8 structural descriptions, 51
‘good’ English vs native speech, 7–8 syntactic categories, 44
history of language, 4–5 wh-movement, 107–108, 113–14
tacit knowledge, 15
theory of syntax, 153–4, 156 harmonic word-order patterns, 163
V-to-T movement, the Head Parameter and head-final, 163, 172
wh-movement, 175–6 head-initial pattern, 162
‘English speakers’ as term, 12 head-movement, 101–2
Esperanto, 6 across five languages, 175–6
ethnolects, 9 S-structure, 190–5
evidential markers, 157 ‘target head’, 119–20

IanRoberts_9781316519493_Index.indd 235 10-09-2022 16:10:18

 236 index

head of the phrase, 81 language faculty, 10

Head Parameter, 163, 217 language instinct, 10
hierarchical relations Left Branch Constraint (LBC), 123–4
constituent structure, 45–7 lexical categories, 36
X′-theory, 86 the lexicon, 179–87, 211
Homo sapiens, 4 lingua franca, 6
linguistics
I-language, 10–11 scientific approach, 1–3
linguistic theory, 31–2 sub-disciplines, 1
and performance, 30 Linnean taxonomy, 4
in relation to syntax, 35 local rules, 118–22
Universal Grammar, 156–7 logic
idiolects, 9 computer-programming languages, 5
immediate constituency, 46–7 predicate logic, 204–5
immediate dominance, 46–7 Logical Form (LF), 180, 203–11
indefinite interpretation of NP, 21–2 S-structure, 194–5, 211–12
indicative mood, 18–19
indices, coindexing, 131–2, 134 marked options (word order), 163
indirect questions, 58–9, 112–18 maximal projection, 86
Indo-European dialects, 5 modals, 92–3
infinitival clauses, 58–9 modifiers, Phrase-Structure rules, 82–7
inflectional endings, 16 mood
interface levels, 180 functional categories, 37–8
intermediate projection, 86 wh-movement, 109–11
internal arguments, 181–2, 184–5 morphological diagnostics, 38–9
interpretative rules, 130 morphology, 1
anaphors, 131–41 movement, causes of, 217
the binding principles, 144–6 movement, characterisation of, 127, 146
pronouns, 130–2, 142–3 movement rules, 100
R-expressions, 131, 143–4 conclusions about, 217
variables, principle C and movement, nature of, 118–22
146–53 S-structure, 190–5
intervention, c-command, 139–40 types of, 100–4, 119–20, 126–7
intransitive verbs, 107–108 variables, principle C and movement, 146–53
intrinsic ordering, 200 see also wh-movement
island constraints, 124–5 multiple questions, 106–7
Italian, comparison with English, 17–22, 31–2, music, as a language, 6–7
217–18
N′-ellipsis, 85–6
Japanese narrow syntax, 194–5
head-final, 172 native-speaker competence, 12, 217–18
V-to-T movement, the Head Parameter and ‘natural kind’, 17–18
wh-movement, 175–6 natural languages
word order, 161–3, 169–70 scientific approach, 9–12
vs artificial, 4–7
labelled bracketing, 16, 19 negation rule, 90
hierarchical relations, 45–7 negative concord, 7
Phrase-Structure rules, 52–5 nodes (tree diagrams), 46
structural descriptions, 51 nominative case marking, 39, 131
language nominative singular, 39
as concept vs individual languages, 3–4 Noun Phrase (NP)
natural vs artificial, 4–7 binding, 147–8
perspectives on, 7–9 centre-embedding, 28–9
theory of syntax, 153–4 fish, 17–18, 26–7
see also syntax beyond English generic vs indefinite interpretation, 17–18
language acquisition, 10 NP-movement, 102–4
language design, 10 phrasal constituents, 45

IanRoberts_9781316519493_Index.indd 236 10-09-2022 16:10:18

 Index 237

Phrase-Structure rules, 55–9 interpretative rules, 130

possible and impossible PS-rules, 79–80 possible and impossible, 78–80
pro-forms, 63–71 recursion, 55–9
R-expressions, 143–4 as structural descriptions, 52–5, 75–6, 99
syntactic diagnostics, 40 X′-theory, 80–7
wh-movement, 61, 109–11, 122–5 X′-theory, adjuncts in, 95–6
X′-theory, 82–4 X′-theory and functional categories, 87–95
nouns pied-piping, 105–6
fish, 15–22 plurals
internal arguments, 184–5 English vs Italian, 20–1
lexical categories, 36 morphological diagnostics, 38–9
lexical entries, 186–7 postpositions, 162–4
morphological diagnostics, 38–9 predicate, 22–3, 40
semantic criteria, 42–3 predicate logic, 204–5
NP-movement, 102–4 preposition stranding, 105–6
existence of parameter, 174 Prepositional Phrase (PP)
kinds of NP positions, 151 pro-forms, 63–71
S-structure, 190–5 syntactic diagnostics, 40
structure of, 120 prepositions
target of, 121 external arguments, 184–5
number (in English), 130–1 lexical categories, 36
Phrase-Structure rules, 60–1
operator-variable structures, 149 word order, 162–4
operators, 149 present tense, English vs Italian, 18–20
Principle A, 151–3
pair-list interpretation, 209–10 Principle B, 149–51
parameters of variation, 157 Principle C, 146–53
parts of speech, 42–3 pro-forms
passives constituency tests, 61, 63–71
languages without, 174 X′-theory, 84–6, 95–6
NP-movement, 101–2, 120, 151, 193 progressive forms, 40
Patient (argument structure), 102 projection of categorial features, 86
performance Projection Principle, 187–90, 193–4, 205–6
and competence, 30 pronouns
relative clauses, 27–8 constituency tests, 63–71
periphrastic tense, 168–9 interpretative rules, 130–2, 142–3
permutation second-person, 16, 17
clefting, 62–3 propositions, X′-theory, 90
tests, 70–5 PS rules see Phrase-Structure rules
X′-theory, 89–90 psych-verbs, 181–3
person, 130–1
see also pronouns Q-feature, 114–15, 119–20
philosophy of mind, 11 quantified expression, 148
phonological diagnostics, 42 quantifier-scope ambiguities, 208–9
Phonological Form (PF), 180, 195–203 quantifiers, 149, 186, 203–11
S-structure, 194–5, 211–12
phonology, 1 R-expressions, 131
phrasal category, 17–18 interpretative rules, 143–4
phrasal constituents, 44–5 proper-name subject, 148
phrasal stress, 42 reciprocal pronouns, 131
phrase markers, 45–7 see also anaphors
Phrase-Structure (PS) rules reconstruction, 210
as building rules, 100 recoverability, 65–6
conclusions about, 217 recursion, 55–9
constituency tests, 59–75 reflexive pronouns, 131
D-structure, 187–90 see also anaphors
distinguishing from movement rules, 118–22 relative clause, 23–4, 27–30

IanRoberts_9781316519493_Index.indd 237 10-09-2022 16:10:18

 238 index

S-structure (surface structure), 180, 190–5, fish and some relative clauses, 22–5
211–12 linguistic theory, 31–2
Logical Form (LF), 203–11 six- and seven-fish sentences, 25–30
Phonological Form (PF), 195–203 teleological perspective, 8, 9
satellite words, 21–2, 37 tense
scientific approach to linguistics, 1–3, 9–12 functional categories, 37–8
second-person pronouns, 16, 17 wh-movement, 109–11
semantics, 1, 42–3 TenseP (TP), 92–5, 110, 122–5
sign language, 5–6 terminal nodes (tree diagrams), 46
sociopolitical perspective, 8 thematic roles, 102, 180
sociolects, 9 third-person plural, 18–19
Specifiers, Phrase-Structure rules, 82–7 to, 92–3
Standard English, 8, 9 Tolkien, J. R. R., 6
stress (phonology), 42 topic, fronting, 61–2
string-vacuous movement, 113 trace, head-movement, 101–2
strong crossover, 151–2 transformations, 100
structural descriptions see also movement rules
constituency tests, 59–75 tree diagrams, 45–7
grammaticality, 51 local PS-rules, 118–22
Phrase-Structure rules, 52–5, 75–6, 99 Phrase-Structure rules, 52–5
possible and impossible PS-rules, 78–80 structural descriptions, 51
presentation of, 51 truth conditions, 148
recursion, 55–9
syntactic representations, 29 unbounded dependency, 117–18
subject-auxiliary inversion, 39–40, 90, 101–2, Undergoer of the action (argument structure), 102
114–15 ungrammatical language see grammaticality
subject-object-verb (SOV) word order Universal Grammar (UG)
French vs English, 157–60 competence, 12
Japanese, 161–3 language acquisition, 10
X′-theory, 160–4 syntax beyond English, 156–7
subject questions, 112–18 universal quantifiers, 186
subject-verb-object (SVO) word order
French vs English, 157–60 V-to-T movement, 171, 175–6
German, 165–71 V-to-T parameter, 159–60, 167
substitution, 121 Verb Phrase (VP)
substitution transformations, 121 fish, 22–3, 26–7
syntactic diagnostics, 39–42 phrasal constituents, 45
syntactic representations, 29 Phrase-Structure rules, 55–9
syntactic theory, goal of, 32 possible and impossible PS-rules, 79–80
syntax syntactic diagnostics, 40
as constituent structure, 35–6 VP-ellipsis, 66–8
I-language, 35 wh-movement, 61, 109–11, 122–5
as sub-discipline, 1 X′-theory, 82–4
syntax beyond English, 156 verb positions, French and English, 157–60
approaching Universal Grammar, 156–7 verb second, 165
German, 165–71 see also subject-verb-object (SVO) word order
verb positions in French and English, 157–60 verbs
wh-movement, 171–4 English vs Italian, 18–19
word order and X′-theory, 160–4 external arguments, 184–5
fish, 15–22
T-to-C movement lexical categories, 36
S-structure, 192 lexical entries, 186–7
subject-auxiliary inversion, 114–15 morphological diagnostics, 38–9
target of, 122 semantic criteria, 42–3
word order, 168–9
tacit knowledge, 15 weak crossover, 151–2
fish, 15–22 Weinreich, Max, 8

IanRoberts_9781316519493_Index.indd 238 10-09-2022 16:10:18

 Index 239

wh-islands, 124 syntax beyond English, 160–4

wh-movement, 61–2, 104–12 verb positions in French and English, 157–60
across five languages, 175–6 word stress, 42
binding, 147–8 World Atlas of Language Structures (WALS), 160,
limitations, 122–5 163–5, 171–4
Logical Form (LF), 205–6
long-distance, 112–18 ‘X′-schema’ for PS-rules, 83
nature of, 118–22 X′-theory, 80–7
S-structure, 190–5 adjuncts in, 95–6
syntax beyond English, 171–4 conclusions about, 217
wh-phrase, 104, 120–1 functional categories, 87–95
wh-questions, 104 syntax beyond English, 160–4
wide scope, 208–9 template for categories, 99
word order Universal Grammar, 156–7

IanRoberts_9781316519493_Index.indd 239 10-09-2022 16:10:18

IanRoberts_9781316519493_Index.indd 240 10-09-2022 16:10:18