Syntax Instruction

Syntax: A Generative Introduction.
2nd Edition
Instructor’s Handbook
Syntax
A Generative Introduction
2nd Edition
Instructor’s Handbook
Andrew Carnie
David P. Medeiros
© 2007 by Andrew Carnie & David P. Medeiros
BLACKWELL PUBLISHING
350 Main Street, Malden, MA 02148-5020, USA
9600 Garsington Road, Oxford OX4 2DQ, UK
550 Swanston Street, Carlton, Victoria 3053, Australia
The right of Andrew Carnie and David P. Medeiros to be identified as the Authors of this Work has been asserted
in accordance with the UK Copyright, Designs, and Patents Act 1988.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in
any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the
UK Copyright, Designs, and Patents Act 1988, without the prior permission of the publisher.
First edition published 2002 by Blackwell Publishers Ltd

Second edition published 2007 by Blackwell Publishing Ltd
The publisher’s policy is to use permanent paper from mills that operate
a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary
chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met
acceptable environmental accreditation standards.
For further information on Blackwell Publishing, visit our website:

www.blackwellpublishing.com
THE USE OF THIS DOCUMENT IS FOR COURSE INSTRUCTORS ONLY. PLEASE
DO NOT PHOTOCOPY OR DISTRIBUTE THESE ANSWER KEYS TO YOUR
STUDENTS.
Contents
Preface and Acknowledgments

1 Generative Grammar
2 Parts of Speech
3 Constituency, Trees, and Rules
4 Structural Relations
5 Binding Theory
6 X-bar Theory
7 Extending X-bar Theory to Functional Categories
8 Constraining X-bar Theory: The Lexicon
9 Head-to-Head Movement
10 DP Movement
11 Wh-movement
12 A Unified Theory of Movement
13 Expanded VPs
14 Raising, Control, and Empty Categories
15 Advanced Topics in Binding Theory
16 Lexical-Functional Grammar
17 Head-Driven Phrase Structure Grammar
Please do not copy or distribute this answer key to your students.

Welcome
TO THE INSTRUCTOR
Thank you for adopting Syntax, 2nd Edition as the textbook for your class. I hope you and your students
will find it useful.
The book is aimed at an introductory level of formal analysis. It is not an introduction to
Minimalism, but presents a view that is Minimalist-informed. For example, I don’t approach phrase
structure from the perspective of Bare Phrase Structure or Antisymmetry. Instead, the student starts with
old-style Phrase Structure Rules, then is led through X-bar theory, and then topics in movement. A new
section in this edition is devoted to more advanced topics that start students towards minimalism. I
didn’t write this text in Minimalism mainly because I feel that MP, although a great theory, is a little too
abstract at this time for beginning students.
The book is designed to take students through many of the fundamental concepts of syntax
(constituency, phrase structure rules, X-bar theory, syntax/lexicon interactions, binding, structural
relations, various kinds of movement and conditions on movement). I certainly don’t claim to have
covered everything. I chose topics that I think are accessible to the beginner and form a coherent whole.

2 Syntax: Instructor’s Handbook
However, you may well want to supplement the text with readings on the topics I don’t cover. You may
also want to look carefully at the problem sets and gray textboxes for each chapter. If I wasn’t able to
include discussion of a topic then there is frequently a problem set that takes it up.
The first three parts of the textbook are designed to be used over a typical 14-week semester,
although I have found that sometimes I want to spend more time on some of the earlier chapters. For
example, I often spend two and a half weeks on chapter 5. I, and my colleagues, have used this textbook
successfully a number of times in our one-semester undergraduate Introduction to Syntax class. This class
(which normally has around 50 students in it) has a prerequisite of a formal introduction to linguistics,
which focuses on the core areas of phonology, syntax, morphology, phonetics, and semantics. Some
people may prefer to break this textbook into two semesters’ worth of material, supplementing some
areas I gloss over more quickly with additional reading. I have also given this textbook to first-year
graduate students who came to our program with limited background in syntax. They have all found it a
successful way to gear them up to the level of our core syntax classes. The last two parts of the book are
more appropriate for this audience. Other people have told me that they have used the book in
introductory graduate classes, although this is not the primary audience for the book. I deliberately made
the tone of the book “informal,” which may be less appropriate to graduate students.
The textbook is designed to be used in the order of presentation, where each chapter builds on
the next. I understand that you may, of course, want to do the chapters in a different order, or skip
chapters entirely (e.g. some people may prefer to skip the discussion of phrase structure rules and go
directly to X-bar theory). In this manual, I’ve mentioned all the major ideas covered in each chapter, so
that if you skip a chapter or do them out of order you’ll know what to cover extramurally.
The problem sets form an important part of the book. They often ask the student to challenge the
presentation of the material in the textbook and think critically about the material. The problem sets come
of a number of types. Some are simply technology practice to cement the knowledge they have learned
(e.g., the tree drawing exercises), others ask the student to apply their new skills to foreign language
problem sets, and still others ask the student to challenge the black and white presentation of the text. In
this second edition, the problem sets have been thoroughly revised and updated. There are many more
practice exercises. The problem sets have also been divided into two major groups: General Problem sets
that I assign to the regular students in my undergraduate class and Challenge Sets which I give to my
honors students and graduate students. In addition, I have attempted to assign a difficulty level and skill
set to each problem set.
Electronic copies of all the problem sets in Microsoft Word and PDF formats are available on the
textbook homepage on Blackwell’s website: You will need a password to access them. Instructions on
how to obtain one are found on the web page. You may also need to download fonts to use the Word
versions. You are welcome to download these problem sets and modify them for your own use. However,
please do not distribute problem sets without a citation to the book on every page, even if you
significantly modify them. I, and my contributing problem set authors, have spent a long time composing
these problem sets, and would like credit for them. On that same website you can download the Keynote
presentation slides I use in my own classes. You are welcome to use these if you like, again with credit.
Please do not copy or distribute this answer key to your students

Introduction 3
On a related note: in order to retain the usefulness of the problem sets in this textbook for use in
classes by other instructors, and for your future classes, please do not distribute this instructor’s
handbook or these answer keys in any form (electronic, print, or otherwise) to your students. Please do
not photocopy this handbook. I thank you in advance for your assistance in this matter, it will ensure that
future instructors will be able to use the problem sets in my textbook with confidence in the originality of
their students’ answers.
Unsurprisingly, this text is going to reflect my own biases. While I tried to follow the “canon” of
thought on particular matters where possible, on occasion, I quite deliberately strayed from the generally
accepted path. When I did this, I did it for one of two reasons: (1) Pedagogical reasons (2) I really don’t
believe the “standard view.” An example of the former (pedagogical reasons) is the fact that until
Chapter 9, I generate all auxiliary verbs under T (not as V). I felt it was just easier for the students to put
them in the right place when you don’t give them the option. I have thought very carefully about all the
cases where I diverge from common practice, and have decided on reflection to leave them all in. Perhaps
these areas will stimulate discussion in your classes.
For those instructors who have used the first edition, the following is a list of the major differences
between the two versions: This list is not comprehensive, many more minor differences can be found.
• The exercise sections of the chapters are now organized differently and are greatly expanded.
Exercises are presented in the order that the material appears in the chapter. I have attempted to
categorize each exercise for level and type.
• There are two types of problem sets: General and Challenge. These two types roughly correspond
to the exercises that I assign to my regular students and my honors students respectively. Challenge
Problem Sets often challenge the straightforward presentation of the material in the main body of the
text.
• The former chapter 2 on structure and parts of speech has been split into two chapters. The new
chapter 2 contains new information on subcategorization that some instructors requested to better
inform students about the role of part of speech in phrase structure processes. Also Adjectives are
now distinguished from Adverbs.
• The phrase structure rules in the new chapter 3 have been completely revised. In particular, I’m
using non-X-bar versions of TP and CP here, and have added embedded clauses to all the relevant
rules.
• The definitions of precedence, exhaustive domination and c-command have all been significantly
revised in the chapter on structural relations. A limited version of government is given for those
instructors who wish to teach it to their students.
• The chapters on X-bar theory have many more trees and examples.
• DPs are used consistently from chapter 7 forward.
• I have added categories to the theta grids in the chapter on the Lexicon in order to tie them to the
subcategories introduced in chapter 2.
• A new section on stacked VPs and affix-hopping has been added to the chapter on head movement.
• VP-internal subjects are used consistently from chapter 9 forward.

• The treatment of passives in chapter 10 is completely different from the previous edition. I have
moved towards a Baker, Johnson and Roberts style approach where the -en morphology is directly
assigned the internal theta role and accusative case by the verb in the syntax rather than in the
lexicon.
• The treatment of locality conditions in the chapter on wh-movement is entirely new. I’ve dropped
subjacency in favor of an MLC based approach. The chapter now includes an inventory of the major
island types; but theoretical coverage is only given to wh-islands. (Although the chapter also
contains a brief discussion of the Head-Movement Constraint and Super-raising in the context of the
MLC).
• Chapter 12 now contains a more accurate discussion of wh-in situ and develops the ideas of feature
checking, covert movement, and SPELLOUT.
• There is a brand new chapter on split VPs in a brand new section on “advanced topics,” including
sections on object shift, ditransitives, a Lasnik style analysis of Pseudogapping and a Hornstein style
analysis of ACD.
• The chapter on Raising and Control has been moved to the new part of the book on advanced topics,
and uses a split VP (vP-AgrOP-VP) structure to avoid ternary branching.
• There is a new chapter on advanced topics in binding theory. This looks at issues on level of
representation, chains and the copy theory of movement. It also takes a relativized view of binding
domain consistent with Chomsky (1986).
• I’ve taken some of the more controversial “comparing theories” language out of the chapters on LFG
and HPSG.
I hope that instructors and students will find these revisions helpful I have attempted where
possible to take into account all the many comments and suggestions I received from people using the
first edition. Although of course in order to maintain consistency, I was unable to do them all.
This instructor’s manual is your personal guide to the major ideas, idiosyncrasies of presentation,
and answer keys. Each chapter has a brief outline of the major topics covered in the corresponding
textbook chapter, along with a discussion of anything unexpected or non-canonical. What follows are the
definitions (also found in the textbook itself) used in the chapter, and answers to most problem sets.
Please feel free to send me – or, if you prefer to remain anonymous, Blackwell – any comments,
corrections, or high praise you might have on the textbook. I’ll be saving those comments for revisions in
future editions and will post regular errata on my website.
A special thanks goes out to Dave Medeiros, who has worked as my TA for the past two
semesters, many of the answers here were painstakingly worked out by him, and he has worked
tirelessly to help me finish this handbook.
Andrew Carnie
Tucson

Generative Grammar
1. TOPICS COVERED IN THIS C HAPTER
This chapter covers some of the basic philosophical issues of syntactic theory. Items marked with an
asterisk (*) have a mention below in Idiosyncrasies in this Chapter section below.
a) Syntax as a science
b) Syntax as a part of cognitive science
c) I-Language (Language) vs. E-Language*
d) The scientific method*
The distribution of person, number agreement in anaphors is used as an example.
e) Modeling syntactic hypotheses using rules
f) Prescriptive vs. Descriptive rules
g) Sources of data: judgments, corpora
h) Distinguishing learning from acquisition
i) Innateness and arguments for it
j) Parameters as an explanation for language variation
k) Choosing among theories: Levels of adequacy (descriptive, observational and explanatory)

2. IF YOU SKIP THIS C HAPTER
Some instructors have expressed to me the desire to do this chapter last instead of first. This is a matter of
personal taste. If you leave it until last, then you will want to review number and agreement and the
notion of anaphor before you do chapter 5, and will probably want to discuss evaluation metrics (levels
of adequacy) before you do chapters 12. Parameter setting should be reviewed before chapter 6. The
things that I would review in class before going on to chapter 2 include prescriptive vs. descriptive rules,
and basic syntactic methodology and perhaps innateness.
3. IDIOSYNCRASIES IN THIS C HAPTER
1 My description of the scientific method will undoubtedly cause some people to raise their eyebrows.
Of course scientific investigation does not necessarily begin with data. (In fact, some people claim
that it never does). Often a hypothesis precedes any data gathering. However, from the perspective of
the student, I think it is important to perceive the data as being the primary driving force behind
linguistic science. So I made this simplification for pedagogic reasons.
2 I distinguish I-language from E-language, but use the terms Language (capital L) and language (lower
case l). The definitions are not precisely identical to I-language and E-language. Language (upper
case) is the capacity and ability to use a particular language (lower case).
GENERAL PROBLEM SETS
1. PRESCRIPTIVE RULES
Common answers include: maintaining a standard across dialects for effective

communication, marking of social and educational status, and so-called
“clarity” in writing.
2. JUDGMENTS
a) Prescriptive (who instead of whom)

b) Prescriptive (me instead of I)
c) Descriptive, syntactic (my red can’t be a subject NP; count nouns like
refrigerator can't appear without either plural morphology or a
determiner)(Some students say semantic; I generally accept this answer
too.)

Chapter 1: Generative Grammar 7
d) Descriptive, syntactic (that-trace effect)

e) Prescriptive (Hopefully is supposed to mean “in a hopeful manner” not “I
hope that…”)
f) Prescriptive (split infinitive)
g) Descriptive, syntactic (wrong word order)
h) Descriptive, semantic (a sister must be a sibling)
3. LEARNING VS. ACQUISITION
Typical answers:
Learned: Reading, writing, mathematics, modern dance, the rules of
basketball, driving.
Acquired: walking, facial recognition, sexuality, taste in food.
Other answers are, of course, possible.
4. UNIVERSALS
Common answers include: shared world view and perception of the way the world
works; a common ancestor language; convergent evolution under similar
conditions of use and acquisition; the idea that syntactic structures reflect
some natural or logical order in the exterior world.
5. INNATENESS
It is really hard to come up with a good answer to this question, which is

the point of the question. It is hard to come up with an argument against
innateness, which suggests the hypothesis is likely correct! The best common
answer I’ve seen suggests performing an unethical experiment where one
exposes a child to a language that violates UG.
6. LEVELS OF ADEQUACY
a) descriptive
b) observational
c) explanatory

CHALLENGE PROBLEM SETS
1. ANAPHORA
Part 1: The antecedent must precede the anaphor.
Part 2: This is the trickiest question in the chapter. The antecedents here are
quantifiers. Everyone allows any gender or number to appear on the
anaphor. Nobody, although it is technically not plural (as shown by the
verb agreement) requires a plural or masculine singular anaphor.
2. YOURSELF
Part 1: Only yourself (and yourselves) are fully grammatical.
Part 2: On the face of it, this seems to contradict the rule. Students may note
that this structure permits an overt subject pronoun (in some registers
and dialects), which must be 2nd person (Don't you hit yourself!). Our
account, of course, will be that this understood silent subject is
syntactically present.
3. IS LANGUAGE REALLY INFINITE?
Q 1: Premise 2 is contradicted by this assumption; if the set is finite,

there is a longest sentence (or set of sentences).
Q 2: Assuming that E is infinite begs the question: we start out assuming

the truth of the conclusion we wish to establish.
Q 3: This is a difficult question. Personally, I do not agree with Pullum

and Sholz that there is anything formally wrong with this proof. It
follows standard practice in establishing the cardinality of infinite
sets (the proof is entirely parallel to, say, Euclid's proof that the
set of prime numbers is infinite).
Q 4: Rough proof using subsets

Assume that the set of sentences S is a subset of possible sentences of
English (as determined by a native speaker), S contains the sentence Calvin
likes tuna. S also contains the embedding construction John said that S which

Chapter 1: Generative Grammar 9
can be applied recursively to any other sentence in S. S is an infinite set,

S is a subset of possible English sentences, therefore the set of possible
sentences of English is infinite.
Rough proof using contradiction.

Assume there is a sentence S which is the longest sentence in the
English language. Native speakers know you can embed any sentence under "John
said that", therefore "John said that S" is a grammatical sentence and is
longer than S. Therefore there can be no S which is the longest sentence in
the English language.
4. ARE INFINITE SYSTEMS REALLY UNLEARNABLE?
There is an important typo in part 3 of this question, replace "infinite

hypothesis" with "alternative hypotheses".
This question requires a fair amount of creativity on the part of the

student, so I can't give you a definitive answer here. (I doubt most
linguists would say there is a definitive answer anyway). In short none of
the equivalents are necessarily incompatible with UG (in fact they all
require something innate even if it is just a good statistical learning
mechanism); they mostly vary in the degree to which innateness is involved in
grammar.
5. LEARNING PARAMETERS: PRO DROP
Q 1: Positive evidence for Pro-drop can only come in the form of hearing
sentences without subjects.
Q 2: The default setting must be non-Pro-drop.
Q 3: This is a difficult question; on the face of it, this is exactly the

sort of trigger that should lead children to analyze English as pro-
drop. One way around this problem is to note that imperatives are
special in several ways, including prosody and (lack of) agreement or
perhaps they are innately specified to avoid listening to imperatives
when determining the pro-drop parameter.

Parts of Speech
This is a new chapter in the second edition. This chapter focuses on parts of speech and how they are
determined (on the basis of distribution, rather than semantically). Among the lexical categories,
subcategories are distinguished on the basis of features.
If you skip this chapter you will want to review the basic ideas of subcategorization before doing chapters
3, 6, or 8.
3. IDIOSYNCRASIES OF THIS C HAPTER
Unlike the first edition I do distinguish Adjectives from Adverbs. Enough of you were annoyed with me
about the A category in the first edition that I've caved to pressure. This makes the rules in chapter 3
considerably more complex than the ones in the first edition. Honestly though, I still don't believe in the
distinction.

Chapter 2: Parts of Speech 11
1. PART OF SPEECH 1
Students are asked to give only the Adjs, Advs, Ns,Ps, and Vs. They are not
asked for Determiners or Ts.
a) The old rusty pot-belly stove has been replaced.

D Adj Adj N(or Adj) N T T(orV) V
b) The red-haired assistant put the vital documents

D Adj N V D Adj N
through the new efficient shredder.

P D Adj Adj N
c) The large evil leathery alligator complained to

D Adj Adj Adj N V P
his aging keeper about his extremely

D Adj N P D Adv
unattractive description.
Adj N
d) I’ve just eaten the last piece of chocolate cake.

N T A V D A N P A N
2. NOOTKA
1) N
2) V
3) V
4) N
5) First position in sentence is a verb, verbs take –ma ending Second
position in a sentence is a noun, nouns take -÷i suffix.
6) The same word appears in different parts of speech in the same
language.

3. GENDER NEUTRAL PRONOUNS
Pronouns are a closed class. Evidently, co-opting a featurally mismatched

form is preferred to innovation.
4. FUNCTIONAL CATEGORIES
The categories appear below the words. The categories D, T, Neg, Conj, C,
and P are all closed class functional categories, so not marked as such below
to save space. Likewise, only the closed lexical categories (here, pronouns)
are marked for open/closed class below.
The propriety of introducing the university slang will be

D N P V/N D N/Adj N T T
readily admitted; it is not less curious than that of the Old

Adv V N-c T Neg Adv Adj Conj N-c P D Adj
Bailey, and is less generally understood. When the number and

N Conj T Adv Adv Adj/V C D N Conj
accuracy of our additions are compared with the price of the

N P D N T V P D N P D
volume, we have no doubt that its editors will meet with the
N N-c T D N C D N T V P D
encouragement that is due to learning, modesty, and virtue.

N C T Adj P N N Conj N
5. PART OF SPEECH 2
You will be surprised at the variety of responses Students will give. A

number of alternatives are reasonable in some cases; these are indicated
where they are known to me.
wabe N; after the

were T; closed class
mome Adj or N after the. If Adj, before a plural N; if N, before an
agreeing verb (depending on whether the -s ending on raths is a [plural]
or [3rd Sg Pres].
outgrabe V; out- prefix
[also possible are Adj or even P; cf. "...and the mole rats ugly", "...and

the mole reads outside."]

Jubjub Adj; after the, before bird
frumious Adj; after the, -ious ending
bandersnatch N; after frumious
vorpal Adj; after his, before sword, possibly -al ending
manxone Adj; after the, before foe
tumtum Adj; after the, before tree
And Conj; closed class
in P; closed class
thought N; after in
uffish Adj; after in, before thought, –ish ending
he This is tricky, I accept D, Pronoun, or Noun
Jabberwock N; after the, before PP
whiffling could be either a V (gerund) (after came, precedes PP) or
an Adverb (after verb came)
the D; closed class
tulgey Adj; between the and wood, -y ending
burbled V; -ed ending
The "modifies" criteria should be avoided here since it isn't clear what
modifies what, given that many of these words are meaningless.
6. SUBCATEGORIES OF NOUNS
Noun Plural Count Proper Pronoun Anaphor

Cats + +
Milk
New York ? +
They + + +
People + +
Language +
Printer +
Himself + + +
Wind
Lightbulb +
New York is tricky, it can't appear with a determiner, but it is of course a

count noun.

7. SUBCATEGORIES OF VERBS
There will be some variability in the answers here.
spray V[NP__NP PP](ditransitive type 2)

sleep V[NP__] (intransitive)
escape V[NP__](intransitive),
V[NP__NP](transitive type 1)
throw V[NP__NP](transitive type 1),
V[NP__NP {NP/PP}](ditransitive type 3)
wipe V[NP__NP](transitive type 1)
say V[NP__NP/CP](transitive type 2)
think V[NP__](intransitive), I think, therefore I am.
V[NP__NP/CP](transitive type 2)
grudge (Some of your students may be unfamiliar with this verb; try
giving them begrudge which is ditransitive V[NP__ NP NP])
thank V[NP__NP](transitive type 1)
?? I thank my lucky stars that it's not true.
pour ?V[NP__] (intransitive ?the water poured onto the ground;
V[NP__NP](transitive type 1) I poured the wine,
V[NP__NP {NP/PP}] I poured her a glass of lemonade, I poured a glass of
lemonade for her.
send ?this sends a message(idiom??);
V[NP__NP {NP/PP}](ditransitive type 3; I sent him the message/the
message to him
promise V[NP__] (intransitive). I promise.
V[NP__NP/CP](transitive type 2)I promised him; I promised that we'd
go.
V[NP__NP {NP/PP}]I promised him an apple, ?I promised an apple to John
Promise is also an object control verb, which they might analyze
as V[NP__NP {NP/CP}]
kiss V[NP__]They kissed;
V[NP__NP]He kissed her.
arrive V[NP__](intransitive),
1. -IAN AND -ISH
This question relies on creativity on the part of students. One answer is

that the words can simply be of either category (Adj or N). Another possible

answer is that the examples without an overt noun head have a silent noun,
something like a null version of "one".
2. NOMINAL PRENOMINAL MODIFIERS
Part 1. Based on distributional criteria, these should be Adjs.

Part 2. These items do not behave at all like other Adjs, not allowing
comparatives or adverbial modification. Note further that there are
adjectival forms of these words (leathery, watery).
3. INTENSIFIERS
Part 1. This is a closed class.

Part 2. Intensifiers cannot be preceded by anything other than another
intensifier within any AdjP or AdvP.
4. COMPLEMENTARY DISTRIBUTION
This of course suggests that complementarity distribution doesn't work for

part of speech categories, which is one of the reasons I think that many
researchers don't use part of speech categories anymore. But, of course, they
are in practical use all the time. This question is meant to make your
students critically evaluate the material in the main body of the text.
5. SUBCATEGORIES OF ADVERBS
Part 1.
1. ___ 2. T ___ 3. end of 4.
Subj V S Obj__PP
ditrans.
luckily + - - -
earnestly - + + +
intently - - + ?
hopefully + ? +? +?
probably + + - -
certainly + + - ?
frequently + + + +?
patiently ? + + +
always - + -? -
completely - + + +
almost - + - -
again ? - + ?

evidently + + - -
frankly + ? + -
demandingly - - + +
yesterday + - + -?
necessarily + + -? ?
Parts 2 & 3. These are HARD questions. We have at least sentence scope
temporal adverbs like yesterday; Speech act adverbs like frankly and
evidently, temporal adverbs like often and always; manner adverbs like
completely and patiently. There probably are some more subtle orderings
within each class. The judgments are fairly subtle I think.
6. SUBCATEGORIES OF ADJECTIVES
two > big > thick,deep > desperate > young > scaly > blue

Constituency, Trees, and Rules
This chapter focuses on the basic notions of constituency (and constituency tests) and phrase structure
rules. The rules are roughly Aspects-style phrase structure rules.
It is possible to skip directly to chapter 5 at this point and do X-bar theory. However, if you do this, you
will need to discuss the mechanics of how rules are written, and trees are generated by rules. Average
students often comment to me that they are glad they did this chapter before doing X-bar, although they
express extreme frustration at having to learn two notations. I try to explain to them that they have to
learn Newtonian physics before they can learn Quantum Mechanics!
You should be warned that the trees in chapters 4 and 5 use the phrase structure rules in this
chapter, not the X-bar ones. So if you skip this chapter, you’ll need to explain that.
3. IDIOSYNCRASIES OF THIS C HAPTER
1 I don’t use AUX, instead auxiliaries are generated in T. I use the categories TP and CP (and not S or S'
as in the first edition).
2 Be careful to note the final form of rules, some rules are revised throughout the chapter.

3 I use + to indicate multiple possible occurrences of an item in a PSR instead of *, since many student
confuse this * with the one meaning ungrammaticality.
4. There is now a brief discussion on headedness; the fact that TP and CP don't always have heads may
require some discussion.
5 I discuss the possibility that particles (blow up, etc.) are actually prepositions without an object.
6 Some people dislike my use of the term “modifies” in my principle of modification for tree
association, feel free to replace with “is closely semantically associated with.”
7 I have mentioned in a grey box relative clauses, there is no easy way to deal with these right now
because they have gaps. I suggest you avoid using them until you cover Wh-movement. Although to
be honest, I don't really treat them at all in this book.
8 Here for convenience are the final rules:

a) CP → (C) TP
b) TP → {NP/CP} (T) VP
c) VP → (AdvP+) V (NP)({NP/CP}) (AdvP+) (PP+) (AdvP+)
d) NP → (D) (AdjP+) N (PP+) (CP)
e) PP → P (NP)
f) AdjP → (AdvP) Adj
g) AdvP → (AdvP) Adv
1. TREES 1: NPS, ADJPS AND A DVPS
a) AdjP
AdvP Adj
smelly
Adv
very

Chapter 3: Constituency, Trees, and Rules 19
b) AdvP
AdvP Adv
quickly
Adv
too
c) AdvP
AdvP Adv
quickly
AdvP Adv
too
Adv
much
d) AdvP
AdvP Adv
quickly
AdvP Adv
too
AdvP Adv
much
Adv
very
e) NP
D AdjP N
the shoelace
Adj
old
f) NP
D AdjP AdjP AdjP N

the noodle
Adj Adj Adj
soggy limp spaghetti

g) NP
D AdjP N
these children
AdvP Adj
finicky
Adv
very
2. TREES: ENGLISH PPS
a) NP
D N PP
the desk
P NP
with
D AdjP N
the drawer
Adj
wobbly
b) PP
P NP
in
D AdjP AdjP N
my boots
Adj Adj
black rubber
c) NP
D N PP
that notebook
P NP
with
D N PP
the scribbles
P NP
in
D N
the margin

d) NP
D N PP
the pen
P NP
at
D N PP
the back
P NP
of
D N PP
the drawer
P NP
in
D N PP
the desk
P NP
near *
NP NP
D AdjP N D AdjP AdjP N

the painting the painting
AdvP Adj Adj Adj
yellow bright yellow
Adv
bright
3. SWEDISH NPS
1) AdjP  (AdvP) Adj

2) Determiners are not obligatory
3) AdjPs are not obligatory in NPs
4) PP  P NP
5) PPs are not obligatory in NPs
6) NP  (D) (AdjP+) N (PP)
7)g) NP
D AdjP N
ett paraply
AdvP Adj
fint
Adv
mycket

h) NP
D AdjP AdjP N
ett paraply
Adj Adj
gammalt fint
i) NP
D AdjP N PP
ett paraply
Adj P NP
rött med
D AdjP N
ett handtag
Adj
gult
8)f) [NP [D ett] [AdjP [Adj gult]] [N äpple]]

i) [NP [D ett] [AdjP [Adj rött]] [ N paraply] [PP [P med] [NP [D ett] [ AdjP [Adj
gult]] [N handtag]]]]
4. ENGLISH
a) TP
NP VP
D N V PP
the kangaroo hopped
P NP
over
D N
the truck
a)[TP[NP[Dthe][Nkangaroo]][VP[Vhopped][PP[Pover][NP[Dthe][Ntruck]]]]]

b) TP
NP T VP
haven't
N V NP PP
I seen
D N P
this sentence before
b)[TP[NP[NI]][Thaven't][VP[Vseen][NP[Dthe][Nsentence]][PP[Pbefore]]]]]
c) TP
NP T VP
will
N AdvP V PP
Susan sing
Adv P NP
never at
N
weddings
c)[TP[NP[NSusan]][Twill][VP[AdvP[Advnever]][Vsing][PP[Pat][NP[Nweddings]]]]]
d) TP
NP VP
D N AdvP V NP
the officer inspected
Adv D N
carefully the license
d)[TP[NP[Dthe][Nofficer]][VP[AdvP[Advcarefully]][Vinspected][NP[Dthe][Nlicense]]]]
e) TP
NP VP
D N AdvP V NP
every cat knows
Adv D N PP
always the location
P NP
of
D AdjP AdjP N
her toy
Adj Adj
favorite catnip

e)[TP[NP[Devery][Ncat]][VP[AdvP[Advalways]][Vknows][NP[Dthe][Nlocation][PP[Pof][NP[Dher][AdjP[Adj
favorite]][AdjP[Adjcatnip]][Ntoy]]]]]]
f) TP
NP VP
D N V NP PP
the cat put
D AdjP N P NP
her toy on
Adj D AdjP N
catnip the mat
Adj
plastic
f)[TP[NP[Dthe][Ncat]][VP[Vput][NP[Dthe][AdjP[Adjcatnip]][Ntoy]][PP[Pon][NP[Dthe][AdjP[Adjplastic]
][Nmat]]]]]
g) TP
NP VP
D AdjP N V PP PP
the child walked
AdvP Adj P NP P NP
young from to
Adv N D N
very school the store
[TP[NP[DThe][AdjP[AdvP[Advvery]][Adjyoung]][Nchild]][VP[Vwalked]
[PP[Pfrom][NP[Nschool]]][PP[Pto][NP[Dthe][Nstore]]]]]
h) TP
NP VP
N V NP PP
John paid
D N P NP
a dollar for
D N PP
a head
P NP
of
N
lettuce
h) [TP[NP[NJohn]][VP[Vpaid][NP[Da][Ndollar]][PP[Pfor]
[NP[Da][Nhead][PP[Pof][NP[Nlettuce]]]]]]].

i) TP
NP VP
N V AdvP
teenagers drive
AdvP Adv
quickly
Adv
rather
i) [TP[NP[NTeenagers]][VP[Vdrive][AdvP[AdvP[Advrather]][Advquickly]]]]
j) TP
NP T VP
can
D AdjP N PP V NP AdvP
a magician fool
Adj P NP D N Adv
clever with the audience easily
D AdjP N
the equipment
Adj
right
j) [TP[NP[Da][AdjP[Adjclever]][Nmagician][PP[Pwith][NP[Dthe]
[AdjP[Adjright]][Nequipment]]]][Tcan][VP[Vfool][NP[Dthe]
[Naudience]][AdvP[Adveasily]]]].
k) TP
NP T VP
might
D N V NP PP
the police plant
D N P NP
the drugs in
D N
the apartment
k) [TP[NP[DThe][Npolice]][Tmight][VP[Vplant][NP[Dthe][Ndrugs]]
[PP[Pin][NP[Dthe][Napartment]]]]].

l) TP
NP T VP
should
D AdjP N V AdvP AdvP
those hopefuls practice
Adj Adv Adv
Olympic diligently daily
l) [TP[NP[DThose][AdjP[AdjOlympic]][Nhopefuls]][Tshould]
[VP[Vpractice][AdvP[Advdiligently][AdvP[Advdaily]]]].
m) TP
NP VP
D AdjP N PP AdvP V NP PP
the research warns
Adj P NP Adv N P NP
latest on always people about
N D N
dieting the dangers
PP
P NP
of
AdjP N
chol.
AdvP Adj
much
Adv
too
m)[TP[NP[DThe][AdjP[Adjlatest]][Nresearch][PP[Pon][NP[Ndieting]]]]
[VP[AdvP[Advalways]][Vwarns][NP[Npeople]][PP[Pabout][NP[Dthe][Ndangers][PP[Pof][NP[AdjP[AdvP[Advto
o]][Adjmuch]][Ncholesterol]]]]]]]
n) TP
NP T VP
was
D AdjP N V AdvP PP
that faucet dripping
Adj Adv P NP
annoying constantly for
N
months
n) [TP[NP[DThat][AdjP[Adjannoying]][Nfaucet]][Twas][VP[Vdripping]
[AdvP[Advconstantly]][PP[Pfor]
[NP[Nmonths]]]]]

o) TP
NP VP
N V CP
Marian wonders
C TP
if
NP T VP
will
D N PP AdvP V
the package arrive
P NP Adv
from ever
N
Boston
o)
[TP[NP[NMarian]][VP[Vwonders][CP[Cif][TP[NP[Dthe][Npackage][PP[Pfrom][NP[NBoston]]][Twill][VP[A
dvP[Advever][Varrive]]]]]]
p) TP
NP VP
N V CP
I said
C TP
that
NP T VP
should
N V NP
Bonny do
D N PP
some dances
P NP
from
D AdjP N
the East
Adj
Middle
p)[TP[NP[NI]][VP[Vsaid][CP[Cthat][TP[NP[NBonny]][Tshould]][VP[Vdo]][NP[Dsome][Ndances][PP[Pfrom
][NP[Dthe][AdjP[AdjMiddle]][NEast]]]]]]]]]

q) TP
CP VP
C TP AdvP V NP
that bothers
NP VP Adv N
really Alina
N V PP
Dan smokes
P NP
in
D N
the office
q)[TP[CP[Cthat][TP[NP[NDan]][VP[Vsmokes][PP[Pin][NP[Dthe][Noffice]]]]][VP[AdvP[Advreally]][Vboth
ers][NP[NAlina]]]]
r) TP
NP VP
D N CP V NP
the belief emboldened
C TP D N AdjP
that the professor
NP VP AdvP Adj
cocky
AdjP N V NP AdvP Adv
theory reveals too
Adj D AdjP N PP AdvP Adv
syntactic the structure much
Adj P NP Adv
inner of already
N
sentences
r)[TP[NP[Dthe][Nbelief][CP[Cthat][TP[NP[AdjP[Adjsyntactic][Ntheory]][VP[Vreveals][NP[Dthe][AdjP[A
djinner]][N structure][PP[P of][NP [Nsentences]]]]]]]][VP [Vemboldened][NP[D the][AdjP[AdvP [AdvP[AdvP
[Advalready]][Advmuch]][Advtoo]][Adjcocky]][Nprofessor]]]]
5. BAMBARA
1) Yes
2) No
3) NP  N
4) PP  NP P
5) VP  (NP) V (PP)
6) TP  NP (T) VP

7) a) TP
NP VP
N V
A kasi-ra
b) TP
NP T VP
ye
N NP V
Den min
N
ji
c) TP
NP VP
N V PP
N son-na
NP P
ma
N
a
8) b) [TP[NP[NDen]][Tye][VP[NP[Nji]][Vmin]]]
c) [TP[NP[NN]][VP[Vson-na][PP[NP[Na]][Pma]]]]
6. HIXKARYANA
1) Yes
2) NP  N (AdjP) (D)
3) VP  (NP) V
4) TP  VP NP
5) Verb, it precedes the subject NP
6) a) TP
VP NP
NP V N
yonyohoryeno biyekomo
N
kuraha

c) TP
VP NP
NP V N
yonoye kamara
N AdjP D
toto komo
Adj
heno
7) a)[TP[VP[NP[NKuraha]][Vyonyhoryeno]][NP[Nbiyekomo]]].
c)[TP[VP[NP[NToto][AdjP[Adjheno]][Dkomo]][Vyonoye]][NP[Nkamara ]]].
7. DUTCH
1) NP  (D) (AdjP) N (PP)

2) PP  P NP
3) VP  (PP) (NP) V (Note: VP  (NP) (PP) V is also consistent with the data.)
4) TP  NP T VP
5) a) TP
NP T VP
is
D N PP PP V
de man gegaan
P NP P NP
in naar
D N N
de regenjas Amsterdam
b) TP
NP T VP
heeft
D N NP V
de man gekocht
D AdjP N PP
een auto
Adj P NP
gele met
D N
een aanhanger

8. AMBIGUITY
(Students often give the trees for their paraphrases, not the original sentence,
although the question is worded so they don’t. You might draw their attention to
this.)
a) i) John said quickly that Mary went to the store.
TP
NP VP
N V CP AdvP
John said
TP Adv
quickly
NP VP
N V PP
Mary went
P NP
to
D N
the store
ii) John said that Mary went quickly to the store.
TP
NP VP
N V CP
John said
TP
NP VP
N V PP AdvP
Mary went
P NP Adv
to quickly
D N
the store
b) i) I discovered a poem written in Old English.

TP
NP VP
N V NP
I discovered
D AdjP N
an poem
AdvP Adj
English
Adv
Old
(Some students will also treat Old English as a compound and generate it under a
single AdjP node. This seems like a reasonable alternative to me.)
ii) I discovered an old poem written in Modern English (or written in England,
or from England).
TP
NP VP
N V NP
I discovered
D AdjP AdjP N
an poem
Adj Adj
old English
c) TP
NP VP
D N V PP PP
two sisters reunited
P NP P NP
after in
D N AdjP N
18 years counter
Adj
checkout
A miraculous reunion

TP
NP VP
D N V PP
two sisters reunited
P NP
after
D N PP
18 years
P NP
in
AdjP N
counter
Adj
checkout
A very long checkout line
d) TP
NP VP
AdjP N V NP
cow injures
Adj N PP
enraged farmer
P NP
with
N
axe
axe-wielding farmer
TP
NP VP
AdjP N V NP PP
cow injures
Adj N P NP
enraged farmer with
N
axe
axe-wielding cow

e) TP
NP T VP
are
N V PP
hospitals sued
P NP
by
D AdjP N
seven doctors
Adj
foot
seven podiatrists
TP
NP T VP
are
N V PP
hospitals sued
P NP
by
AdjP N
doctors
AdvP Adj
foot
Adv
seven
tall doctors
f) TP
NP T VP
will
N V NP PP
dealers hear
AdjP N P NP
talk after
Adj N
car noon
A talk about cars

TP
NP T VP
will
N V CP PP
dealers hear
TP P NP
after
NP VP N
noon
N V
car talk
a talking car
(note there's another ambiguity in the second example: [after noon] could attach to
either VP)
9. STRUCTURE
a) *It was a minivan to Petunia that Susan gave.

*Susan gave it. (where ellipsis hasn’t occurred)
The following sentence will give bad results:

Susan gave the minivan to Petunia and the car to Phil.
b) The constituency here depends upon whether the letter is from Stacy, or
whether Clyde merely got it from Stacy (but it was from Louise). So both
answers are possible.
10. ENGLISH PREPOSITIONS
*He blew out the candle and up the building.
But
He turned off and blew up the lightbulb.

1. QUANTIFIERS
The simplistic answer is revise the PSR with a + after the determiner… but
then we need an alternative explanation why not all Ds can do this
(presumably in terms of subcategorial features). This question can be a
discussion that might lead up to DPs in chapter 7 or QPs (in the context of
floating quantifiers) in chapter 9
2. NOMINAL ADVERBS
Part 1: Every is a Q/D so that marks these as Nouns. In some cases we can also
get adjectives Every cloudy Sunday.
Part 2: not really, without actually allowing OTHER NPs in this position.
This question can act as a lead in for case theory in chapter 10
3. POSSESSIVE NPS
Part 1: Can't have an NP before the N head.

Part 2: Allow a choice function between the D and an NP's : NP  ({NP's/D})
(AdjP+) N (PP+}
4. CONSTITUENCY TESTS
This is a very difficult question. [Barbie and Ken kissing] is a small clause. Various
tests will show this. But [Barbie and Ken] is displaced by movement.

Structural Relations
0. INTRODUCTION
This chapter is pretty straightforward. However, I would like to draw a couple of things to your
attention.
1) There is now a discussion of government (new to this edition), but it may safely be skipped or
referred to much later.
2) The definition of “mother” is limited to immediate domination.
3) The discussion of “indirect object” is very sketchy. If you care about this you might want to expand
on the issue.
4) The definitions of “precedence” and “exhaustive domination” have been completely revised from
previous editions.

1. TREES
a) TP
NP1 VP
D AdjP1 N1 PP1 V NP2

The man loves
Adj1 P1 NP4* N2 PP2
big from bagels
N4 P2 NP3*
New York with
AP2 N3
cheese
A2
cream
*Some students may treat New York as an N with an adjective modifier, and similarly
some students may treat cream cheese as a compound.
It's probably also possible to attach [with cream cheese] to the VP rather than the
NP.
b) TP
NP VP
N V NP PP
Susan rode
D AdjP N P NP
a train from
AdvP Adj N
blue NY
Adv
bright
Riding is from NY
OR

Chapter 4: Structural Relations 39
TP
NP VP
N V NP
Susan rode
D AdjP N PP
a train
AdvP Adj P NP
blue from
Adv N
bright NY
Train is from NY
You could also have NY as below, though it’s probably a compound.

NP
AdjP N
York
Adj
New
It may also be possible to have bright form an AdjP of its own which modifies the NP
directly.
c) TP
NP VP
D AdjP N V NP PP PP
The platypus kicked
Adj D N PP P NP P NP
plucky a can from to
P NP N N
of NY Tuc.
N
soup
The kicking happened from NY to Tucson

TP
NP VP
D AdjP N V NP PP
The platypus kicked
Adj D N PP PP P NP
plucky a can to
P NP P NP N
of from Tuc.
N N
soup NY
The can is from NY

It might also be possible to have [from NY] modify soup.
d) TP
NP VP
N V CP
John said
TP
NP VP
N V NP PP
Martha sang
D N P NP
the aria with
N
gusto
Martha sang with gusto
d) TP
NP VP
N V CP
John said
TP PP
NP VP
N V NP
Martha sang
D N P NP
the aria with
N
gusto
John said so with gusto

e) TP
NP VP
N V CP
Martha said
TP
NP VP
N V NP
John sang
D N PP
the aria
P NP
from
N
La Bohème
f) TP
NP VP
D N PP PP V
the book stinks
P NP P NP
of with
N D AdjP N
poems the cover
AdvP Adj
red
Adv
bright
g) TP
NP VP
N V CP
Louis hinted
TP
NP VP
N V NP AdvP
Mary stole
D N Adv
the purse deftly
The stealing was deft.

TP
NP VP
N V CP
Louis hinted
TP
NP VP
N V NP AdvP
Mary stole
D N Adv
the purse deftly
The hinting was deft.
h) TP
NP VP
D AdjP N V NP PP
the students hated
AdvP Adj AdjP N P NP
tired trees with
Adv Adj D N
extremely syntactic a passion
There may also be a weird reading where the trees are passionate.
i) TP
NP T VP
have
D N V CP
many soldiers claimed
TP
NP VP
AdjP N V NP AdvP
water quenches
Adj N Adv
bottled thirst best
There may also be a really weird reading where best modifies claimed.

j) TP
NP VP
N V CP
networking helps
TP
NP VP
N V NP
you grow
D N
your business.
2. DOMINATION
a) NP1, TP1
b) NP2, TP2, CP, VP1, TP1
c) NP1, TP1
d) NP2, TP2, CP, VP1, TP1
e) VP1, TP1
f) VP2, TP2, CP, VP1, TP1
g) AdvP, VP2, TP2, CP, VP1, TP1
h) CP, VP1, TP1
i) none
j) CP, VP1, TP1
k) TP1
l) TP2, CP, VP1, TP1
m) TP1
n) TP2, CP, VP1, TP1
o) VP1, TP1
p) VP2, TP2, CP, VP1, TP1
q) TP1
r) D1, N1, V1, C, D2, N2, V2, Adv
s) TP1, NP1, VP1, CP, TP2, NP2, VP2, AdvP
t) V2, AdvP, Adv
u) C, TP2, NP2, D2, N2, VP2, V2, AdvP, Adv
v) D1, N1
3. EXHAUSTIVE DOMINATION
a) No.
b) Yes, NP1.
c) Yes, VP2.
d) Yes, TP2.
e) No.
f) No.

g) Yes, CP.
h) V1, C, D2, N2, V2, Adv
i) No, no node exhaustively dominates C, D2, N2.
j) Yes, those nodes are exhaustively dominated by TP1.
k) Yes.
l) Yes.
m) No.
n) Yes.
o) Yes.
p) Yes.
q) Yes (should read Adv, AdvP)
4. IMMEDIATE DOMINATION
a) NP1
b) NP2
c) NP1
d) NP2
e) VP1
f) VP2
g) AdvP
h) CP
i) none
j) CP
k) TP1
l) TP2
m) TP1
n) TP2
o) VP1
p) VP2
5. PRECEDENCE
a) None
b) NP1, D1, N1, V1, C
c) D1
d) NP1, D1, N1, V1, C, D2
e) NP1, D1, N1
f) NP1, D1, N1, V1, C, NP2, D2, N2
g) NP1, D1, N1, V1, C, NP2, D2, N2, V2
h) NP1, D1, N1, V1
i) none
j) NP1, D1, N1, V1, C
k) None
l) NP1, D1, N1, V1, C
m) NP1, D1, N1
n) NP1, D1, N1, V1, C, NP2, D2, N2
o) NP1, D1, N1, V1
p) NP1, D1, N1, V1, C, NP2, D2, N2, V2

6. IMMEDIATE PRECEDENCE
a) None
b) C
c) D1
d) D2
e) NP1, N1
f) NP2, N2
g) V2
h) V1
i) none
j) C
k) None
l) C
m) NP1, N1
n) NP2, N2
o) V1
p) V2
7. C-COMMAND
a) N1
b) N2
c) D1
d) D2
e) CP, C, TP2, NP2, D2, N2, VP2, V2, AdvP, Adv
f) AdvP, Adv
g) None
h) TP2, NP2, D2, N2, VP2, V2, AdvP, Adv
i) None
j) C
k) VP1, V1, CP, C, TP2, NP2, D2, N2, VP2, V2, AdvP, Adv
l) VP2, V2, AdvP, Adv
m) NP1, D1, N1
n) NP2, D2, N2
o) V1
p) V2
q) NP1, V1, C
r) VP1
s) NP1, V1, TP2
8. SYMMETRIC AND A SYMMETRIC C-COMMAND
a) N1
b) N2
c) D1
d) D2

e) CP
f) AdvP
g) None
h) TP2
i) None
j) C
k) VP1
l) VP2
m) NP1
n) NP2
o) V1
p) V2
q) None
r) None
s) None
t) None
u) C, TP2, NP2, D2, N2, VP2, V2, AdvP, Adv
v) Adv
w) None
x) NP2, D2, N2, VP2, V2, AdvP, Adv
y) None
z) None
aa) V1, CP, C, TP2, NP2, D2, N2, VP2, V2, AdvP, Adv
bb) V2, AdvP, Adv
cc) D1, N1
dd) D2, N2
ee) None
ff) None
gg) NP1, V1, C, NP2
hh) VP1
ii) NP1, V1
jj) TP2
9. GOVERNMENT
a) No. Although NP1 c-commands VP2, there are nodes which NP1 c-commands (V1,
C, NP2) which also c-command VP2.
b) No. NP1 c-commands C, but NP1 c-commands V1, which also c-commands VP2.
c) D1
d) No. V1 c-commands V2, but V1 c-commands C, a head which also c-commands V2.
e) V2
f) No. NP1 c-commands AdvP, but NP1 c-commands NP2, which also c-commands
AdvP.
g) Yes. VP2 c-commands N2, and there is no phrase G which VP2 c-commands which
also c-commands N2.
10. GRAMMATICAL RELATIONS I
a) NP1

b) NP2
c) CP
d) No, it has no CP or NP daughters which might be
candidates for direct object.
11. GRAMMATICAL RELATIONS II
a) NP1
b) NP3
c) NP5
d) NP6
e) CP
f) NP4
g) NP2
h) PP1
i) Oblique; it is not a subject, direct or indirect object or object of a
preposition.
12. GRAMMATICAL RELATIONS III
a) TP
NP VP
N AdvP V AdvP PP
it rains
Adv Adv P NP
never violently in
AdjP N
California
Adj
southern
Subject: [NPit]
Object of P: [NPsouthern California]
verb: [Vrains]

b) WARNING TELL YOUR STUDENTS TO DELETE THE SOON! THERE IS NO RULE THAT GENERATES
AN ADVERB INITIAL TREE
TP
AdvP NP T VP
should
Adv N V NP NP
soon we give
D AdjP N D N
the dog a bath
Adj
family
Subject: [NPwe]
Direct Object: [NPa bath]
Indirect Object: [NPthe family dog]
verb: [Vgive]
c) TP
NP VP
D AdjP N AdvP V NP PP
the contestant made
AdjP Adj Adv D AdjP N P NP
show bravely a guess about
Adj Adj D N
quiz wild the answer
Subject: [NPthe quiz show contestant]
Direct Object: [NPa wild guess]
Object of P: [NPthe answer]
verb: [Vmade]
13. STRUCTURAL RELATIONS I
a) NP3, PP, VP, TP

b) NP3
c) No
d) D1, AdjP, Adj
e) T, VP, V, NP2, N2, PP, P, NP3, D3, N3
f) VP
g) VP, V, NP2, N2, PP, P, NP3, D3, N3
h) {NP1,T,VP}, {D1,AdjP,N1}, {V,NP2,PP}, {P,NP3}, {D3,N3}
i) VP
j) Symmetrically c-command
k) NP2, N2, PP, P, NP3, D3, N3
l) NP1 [the big bully]
m) NP2 THIS QUESTION SHOULD BE CORRECTED TO SAY "DIRECT OBJECT"
n) NP3
o) Yes
p) PP

q) V, N2, PP, P, D3, N3

r) TP
s) D1, Adj, N1, T, V, N2, P, D3, N3
t) D3
14. STRUCTURAL RELATIONS II
The exact answers to these questions may vary depending upon whether the student
treated New in New York and cream in cream cheese as adjectives or as part of compound
nouns, and whether they attached [with cream cheese] to the VP or NP.
a) VP, V, NP2, N2, PP2, P2, NP3, (AdjP2), (Adj2), N3

b) V, NP2, N2, PP2, P2, NP3, (AdjP2), (Adj2), N3
c) D, AdjP1, Adj1, N1, PP1, P1, NP4, N4
d) D, AdjP1, N1, PP1
e) VP, V, NP2, N2, PP2, P2, NP3, (AdjP2), (Adj2), N3
f) NP1, D, AdjP1, Adj1, N1, PP1, P1, NP4, N4
g) D, AdjP1, Adj1, N1, PP1, P1, NP4, N4
h) V, NP2, N2, PP2, P2, NP3, (AdjP2), (Adj2), N3
i) V, NP2
j) None
k) NP1, D, AdjP1, Adj1, N1, PP1, P1, NP4, N4
15. TZOTZIL
1) NP  (D) N
2) PP  P NP
3) VP  V (AdvP) (NP) (PP)
(There is not enough evidence to determine the order of NP and PP. Some students
may treat “away” as P, in which case there is no AdvP in the rule)
4) TP  VP NP
(Some students may treat the subject as optional if they note that the church in
(c) is semantically an object.)
5) [NPli Maruche]
6) A trick question, could be either depending upon how you treat passives also
depending upon their trees in (10) below.
7) yes
8) yes
9) Yes
10) b) TP
VP NP
V PP D N
‘ibat li Maruche
P NP
xchi’uk
N
smalal

c) TP OR TP
VP NP VP
V D N V NP
pas ti ‘eklixa’une pas
D N
ti ‘eklixa’une
16. HIAKI
1) NP  (D) N (AdjP)
2) No, prepositions are morphologically marked.
3) VP  {AdvP/NP} V
(Don’t put any order on these complements because there isn’t any evidence for
the order in the data.)
4) TP  NP VP
5) b) TP
NP VP
D N NP2 V
Hunáa'a yá'uraa nokriak
D2 N2
hunáka'a hámutta
c) TP
NP1 VP
N1 NP2 V
Taáwe hiba-tu'ure
N2 AdjP
tótoi'asó'olam
Adj
káamomólim
d) TP
NP VP
N AdvP V
Tá’abwikasu yépsak.
Adv
’áma
6) [NPHunáa'a yá'uraa] ‘that chief’

7) Technically no, since the students are told to assume 'áma is an adverb and object
is defined as an NP daughter of VP.

8) [N yá'uraa]
9) VP, NP2, N2, D2, V
10) [V yépsak] ‘arrived’
11) AdjP
12) AdjP, NP2, VP, TP
13) N2,
14) NP1, N1, N2
15) No
16) Yes
1. DISCONTINUOUS CONSTITUENTS
No, the principle of modification would force us to cross lines to capture the
meanings.
2. NEGATIVE POLARITY ITEMS
Both c-command and precedence are consistent with the data in this question.
3. IRISH
1) None or AdjP Adj

2) NP  (D) N (AdjP) (must be consistent with the answer to (1) allowing an AP if
they posit one)
3) No
4) Object = Second NP daughter of TP
5) TP  V NP (NP) (this question will make students, rightfully, very uneasy – the solution to
the problem is discussed in chapter 8).
6) c) TP
V NP NP
Phóg
D N D N
an fear an mhuc
d) TP
V NP NP
Chonaic
N D N Adj
mé an mhuc mhór

(mhór may be dominated by an AdjP, depending on the answer to question 1)
e) TP
V NP
Rince
D N
an bheán

Binding Theory
0. INTRODUCTION
This chapter presents a fairly straightforward presentation of the "classics" of binding theory. I
oversimplify the description of what binding domain is. I stipulate that the binding domain is equivalent
to the clause. So there is no discussion of accessible subject or governing category. A more sophisticated
version of binding theory is given in chapter 15.
1. NP TYPES
their pronoun
each cat R-expression
folk dancing R-expression
oneself anaphor
each other anaphor
she pronoun
her pronoun
themselves anaphor

2. C-COMMAND AND BINDING
a) TP
NP T VP
didn't
D N PP V NPi
the book bother
P NPi N
about him
D N
the president
(i) No binding.
(ii) Although the NPs corefer, neither c-commands the other.
b) TP
NPk T VP
didn't
D N PP V NPi
the book bother
P NP N
about him
D N
the president
(i) No binding.
(ii) The NPS do not corefer.
c) TP
NPk VP
D N PP V NPk
the book sold
P NP N
about itself
D N
the president
(i) Yes, there is a binding relationship.
(ii) [NPthe book about the president] binds [NPitself].
d) TP
NP VP
NPi AdjP N PP V NP
lack dismayed
N Adj P NP Di N
Andy's constant of his father
N

Chapter 5: Binding Theory 55
effort
(i) No, there is no binding relationship.
(ii) The NPs corefer, but neither c-commands the other.
e) TP
NP VP
NPi AdjP N PP V NP
lack dismayed
N Adj P NP Dn N
Andy's constant of his father
N
effort
(i) No, there is no binding relationship.
(ii) The NPs do not corefer.
3. BINDING DOMAIN
a) TP
NP VP
D N V NP CP
the students told
N C TP
themselves that
NP T VP
wouldn't
D N V AdjP
the exam be
AdvP Adj
hard
Adv
too
WARNING TELL YOUR STUDENTS TO TREAT BE AS CATEGORY V

b) TP
NP VP
D N V NP CP
the students told
D N C TP
their professor that
NP T AdjP
weren't
N Adj PP
they worried
P NP

about
AdJP N
theory
Adj
binding
WERE HERE IS T, BUT IN THE NEXT SENTENCE IT'S A V!!!!!
c) TP
NP VP
N V CP
Michael said
TP
NP VP
D AdjP N
the judgments were wrong
Adj
binding
4. BINDING PRINCIPLES
a) [NPhim] is c-commanded and bound within its domain (the entire sentence),
violating Condition B.
b) [NPMichael] is c-commanded and bound within its domain (the entire
sentence), violating Condition C.
Many students will answer B here, but of course the pronoun isn't bound,
it's just coindexed. Answering B means they've probably not got the fact that
coindexing and binding are not the same thing.
c) [NPhimself] is not c-commanded (thus not bound) within its domain (the
entire sentence), violating Condition A.
d) [NPhim] is c-commanded and bound within its domain (the entire sentence),
violating Condition B.
e) [NPherself] is not c-commanded (thus not bound) within its domain (the
embedded sentence [TPJohn should marry herself]), violating Condition A.
f) [NPher] is c-commanded and bound within its domain (the embedded sentence
[TPSusan should kiss her]), violating Condition B.

Chapter 5: Binding Theory 57
1. WH-QUESTIONS
The anaphor in the wh-phrase isn’t c-commanded by its antecedent in its

surface position. We return to this question in the chapter on wh-movement.
2. BINDING DOMAIN
The pronoun is bound within its domain, so this should violate Condition B.
3. PERSIAN
Xodesh can either be bound locally or at a distance, like Japanese zibunzisin

Xod must be locally bound (and behaves like an anaphor)
4. JAPANESE
Question 1: It appears to be an anaphor in (a), as it must be locally bound.
Question 2: It has the properties of both, since in (b) it can be bound by

something outside the clause. (The solution, not asked for, is that
zibunzisin must be bound by the first NP that precedes it.)
Question 3: Condition C (Mary bound by zibunzisinga).
5. COUNTEREXAMPLES?
a) Me is c-commanded and coindexed with I, should be a principle B violation.

b) Should be a condition A violation, because himself is bound by an NP
outside its binding domain (clause).
c) The obligatorily bound anaphor is bound by an NP two clauses up.
d) In (i) the pronoun ye can’t be locally bound.
In (ii) the pronoun ye can’t be free.
So in (i) it’s behaving like a pronoun, and in (ii) like an anaphor.

6. C-COMMAND OR PRECEDENCE?
If precedence defines binding, then these sentences would constitute

condition C violations. However, the sentences are grammatical. By contrast,
under c-command, the pronoun does not c-command the R-expression, so these do
not constitute condition C violations (correctly).
Some students will try to use condition B instead of condition C. But
condition B does not distinguish the two hypotheses -- under both precedence
and c-command the pronoun is free, so both hypotheses are consistent.
Condition C is the crucial case.

X-bar Theory
0. INTRODUCTION
This chapter introduces the basic notions of X-bar theory as applied to NPs, AdjPs, AdvPs, PPs, and VPs.
CP, TP (which I use instead of IP), and DP are saved until the next chapter. There is significant
discussion on motivating more general principles instead of language specific phrase structure rules. I
also discuss at length the distinction between complements and adjuncts, and the tests to distinguish
them. The notion of phrase structure parameters is also introduced.
As in Radford (1988) and the original work in X-bar theory (but not consistent with common
current practice), adjuncts in this chapter are not Chomsky-adjoined (ie attached above XP). Instead they
are daughters of and sisters to the single bar level.
[XP spec [X' adjunct [X' X complement]]]
I’m using a version of X-bar theory where only the single bar level iterates, XP is what I use for Xmax or X''.
There is relatively little discussion of specifiers (I’m going to reserve specifiers for subjects in later
chapters). I put determiners in the specifier of NP in this chapter but this is the only specifier they see. In
chapter 7, I adopt a DP hypothesis getting rid of this exception. Unlike Jackendoff’s original form, I don’t
put very in the spec of VP, because I find this confuses the students later when they do the VP-internal
subject hypothesis. So students may need some additional discussion reassuring that specifiers are well
motivated.

One area that may require further discussion is the treatment of pre-head complements (as in the
linguistics professor). Some people prefer to treat these as N-N compounds rather than AdjP N
complement head structures. There is no discussion in the text of why pre-head modifiers must not be
phrasal. This will probably need some classroom work. In the first edition this was in the main body of
the text. In this edition I've taken that out and moved it into exercises.
Smarter students will get annoyed with my appeal to parsimony for explaining X-bar structure in
AdjPs, AdvPs, and PPs. You have my sympathy on this point. But frankly my experience is that it is more
of a challenge to get the bulk of students through X-bar tree drawing and understanding the difference
between complements and adjuncts than to worry too much about those students who (rightly) question
its motivations. To answer the clever students' questions I usually just point them to a primary reading
like Jackendoff and let them struggle it out for themselves.
1. COMPLEMENTS VS. ADJUNCTS IN NPS
a) [of flour] is a complement.

b) [with a glass lid] is an adjunct.
c) [of figurines] is a complement, [in the window] is an adjunct.
d) [of Napoleon] is a complement, [on the corner] is an
adjunct.
e) [in the building] and [with a broken pane] are both
adjuncts.
2. ADJECTIVES
Adjectives like hot, red, big, tiny, and ugly are adjuncts.
3. GERMAN NOUN PHRASES
1) Elide or delete an N' node. This is like one-replacement in English.
2) I believe mit Sahne modifies the verb, not the NP cake, but this is up for
debate. Simin Karimi tells me that on the basis of certain
pronominalization facts, she thinks the PP modifies Kuchen. The exact
analysis of the mit Sahne PP is irrelevant to the content of the question,
which is actually about the Subject NP.

Chapter 6: X-bar Theory 61
a) TP
NP VP
D N' V'
Die
N' PP V' PP
AdjP N' P' V NP P'

isst
Adj' N P NP N' P NP
Frau aus mit
Adj N' N N'
schlanke Kuchen
N N
Frankreich Sahne
OR
TP
NP VP
D N' V'
Die
AdjP N' V' PP
Adj' N' PP V NP P'

isst
Adj N P' N' P NP
schlanke Frau mit
P NP N N'
aus Kuchen
N' N
Sahne
N
Frankreich
f) TP
NP VP
D N' V'
Die
AdjP N' V NP
liebte
Adj' N PP D N'
Koenigin die
Adj P' N
junge Prinzessin
P NP

von
N'
N
England
3) von England is a complement, so is a sister to N, not N', so deletion of the bare

noun is impossible. By contrast aus Frankreich is an adjunct, sister to N', so
there is an N' node dominating just the noun head which can be deleted.
4. COMPLEMENTS AND ADJUNCTS IN VPS
a) [her pencils] is a complement, while [never] and [in the correct drawer] are
adjuncts.
b) [to New Mexico], [in the rain], and [last year] are
adjuncts.
There is a strong tendency to put the path PP [to NM] first, but note that it could be
preceded by [from AZ] or [a mile].
5. JAPANESE
1) -o marks direct objects and –ga marks subjects

2) SOV
3) Complements precede heads.
4) Adjuncts precede heads.
5) Specifiers precede heads.
6) TP
NP VP
N' V'
N NP V
Toru-ga mita
D N'
sono
AdjP N'
Adj' N
hon-o
Adj
akai

6. PARAMETERS
NOTE: Irish is a challenge problem set, so you may wish to exclude it from
this question.
Specifier Adjunct Complement

Hiaki (YP) X' Both (Adv/Adj) (WP) X
Irish (YP) X' X' (ZP) No evidence
(subject
between only
comp: V and O)
Bambara No evidence X' (ZP) (WP) X
agreed [to it]
Hixkaryana X' (YP) X" (ZP) (WP) X
Swedish (YP) X' Both(AdjP/PP) X (WP)
Dutch (YP) X' Both(AdjP/PP) Both(VPs/PPs)
Tzotzil (YP) X' X' (ZP), if X (WP)
the PPS are
adjuncts.
7. TREES
a) TP
NP VP
N' V' PP
N P'
Abelard
V' PP P NP
for
V NP P' N'
wrote
D N' P NP N
a in Heloise
N PP N'
volume
P' N
Latin
P NP
of
N'
N
poems

There are six possible trees for this sentence: [in Latin] may attach as an adjunct to
the VP, the NP headed by volume, or the NP headed by poems. [for Heloise] can be an
adjunct to any of these phrases not dominated by the phrase that [in Latin] attaches
to.
b) TP
NP VP
N' V'
N' PP AdvP V'
N P' Adv' V' PP

armadillos
P NP Adv V NP P'
from often destroy
N' N' P NP
with
AdjP N' AdjP N' D N'
their
Adj' N Adj' N N
York pillowcases snouts
Adj Adj
New old
You are on your own with the rest. Notice that “from Italy” in (i) is ambiguous and
could either modify volume or verse, in either case as an adjunct. Likewise in (j),
"from County Kerry" can attach either within the VP (describing a path) or the NP
(describing the historical origin of the jig), as an adjunct in either case.
i) The red volume of obscene verse from Italy shocked the puritan soul of the minister
with the beard quite thoroughly yesterday.
j) The biggest man in the room said that John danced an Irish jig from County Kerry to
County Tipperary on Thursday.
1. INTERMEDIATE STRUCTURE
Example: Adam frequently buys paintings from Natasha,

...but Dan seldom does (so).
...and Dan does (so) from Sheila.

IMPORTANT: There is an error in the question. The relevant rule is V' (AdvP)
V' not V'  V' (AdvP).
2. COMPLEMENT ADJPS?
Part 1. These are complements.
One-replacement:
*the physics professor, not the chemistry one
Coordination of likes:
*the tall and chemistry student
One complement:
*the chemistry student of physics
leather behaves similarly, but not identically:

?the canvas shoe, not the leather one
*the leather and ugly shoe
But note:
the leather book of poems.
Be aware that many students do not share these judgments.
Part 2. The answer to this question relies on creativity on the part of students.
(c) can be taken as evidence against the (b) analysis, while (d) and especially (e)
argue against (a). Note also that the purported N can have phrasal syntax:
the molecular chemistry professor
?the obviously leather shoes
3. AMBIGUOUS ADJPS?
Part 1. Parameters have to be set up specific to a category rather than to X.
Part 2. The ambiguity results from the fact that German can attach as either a
complement or an adjunct. Note that (b) is unambiguous; it can only mean someone from
France who teaches German (and the proposed complement must occur adjacent to the
head). The (c) and (d) examples show that coordination of likes (complement or
adjunct modifiers, respectively) can eliminate the ambiguity.
4. COMPLEMENTS TO A DJ H EADS?
Part 1. no: delightful

clearly yes: full
more thought provoking: familiar with, sensitive to, adjacent to.
Part 2. OOPS they don't really have rules for "as" constructions.

Extending X-bar
Theory to Functional Categories
0. INTRODUCTION
This chapter presents a short survey of clause types (main, embedded, complement, adjunct, specifier,
tensed, untensed). There is no discussion of small clauses, I find students have too much difficulty with
the concept to identify them at this stage in their intellectual development. I identify a clause as a subject
and predicate phrase (I use the term predicate phrase here, instead of predicate due to the fact in the next
chapter we use predicate to refer only to the head of the predicate phrase).
DPs are introduced and the X-bar versions of TPs (note not IP!) and CPs with null and "affix"
heads are all introduced. I make the claim that all clauses have a CP even if there is no overt C (and
similarly with TP and T). To simplify matters I simply generate inflectional affixes in T and then have
them move for morphological reasons. This is revisited in chapter 12, where checking theory is more fully
spelled out. There is also a brief introduction to subject/aux inversion, which we return to in more detail
in chapter 9.

Chapter 7: Extending X-bar Theory to Functional Categories 67
1. SUBJECTS AND PREDICATE PHRASES
a) Subject: The peanut butter

Predicate Phrase: got moldy
b) Subject: The duffer’s swing
Predicate Phrase: blasted the golf ball across the green
c) Main Clause Subject: That Harry loves dancing
Main Predicate Phrase: is evidenced by his shiny tap shoes
Embedded clause subject: Harry
Embedded clause predicate phrase: loves dancing
d) Subject: The Brazilians
Predicate Phrase: pumped the oil across the river
2. CLAUSE TYPES
a) Main clause: Stalin may think that Roosevelt is a fool.

complementizer: Ø
T: may
subject: Stalin
finite
Embedded clause: that Roosevelt is a fool
complementizer: that
T: is
subject: Roosevelt
finite
b) Main clause: Lenin believes the Tsar to be a power-hungry dictator.

complementizer: Ø
T: -s
subject: Lenin
finite
Embedded clause: the Tsar to be a power-hungry dictator
complementizer: Ø
T: to (to be?)
subject: the Tsar
non-finite
c) Main clause: Brezhnev had said for Andropov to leave.

complementizer: Ø
T: had
subject: Brezhnev
finite
Embedded clause: for Andropov to leave
complementizer: for
T: to

subject: Andropov
non-finite
d) Main clause: Yeltsin saw Chernyenko holding the bag.

complementizer: Ø
T: -ed?? or [past]
subject: Yeltsin
finite
Embedded clause – this is tricky, since they have to identify the clause headed by
hold as a clause. Chernyenko holding the bag
complementizer: Ø
T: -ing? (this requires some discussion)
subject: Chernyenko
non-finite
3. ENGLISH THAT
a) Complementizer
b) Determiner, as shown by number agreement (that student/those students)
Also you can say "that that student" but not "that that students"
TP
DP T'
D' T VP
-s
D NP V'
Ø
N' V CP
think
N C'
Robert
C TP
Ø/that
DP T'
D' T VP
should
D NP V DP
that/Ø eat
N' D'
N D NP
student(s) Ø
N'
N
asparagus

Chapter 7: Extending X-bar Theory to Functional Categories 69
4. TREES
a) CP
C'
C TP
Ø
DP T'
D' T VP
-ed
D NP V'
the
N' V' PP
AdjP N' V' PP P'
Adj' N V P' P DP
child walk to
AdvP Adj' P DP D'
from
Adv' Adj D' D NP
young the
Adv D NP N'
very Ø
N' N
store
N
school
You are on your own for the rest:

b) Linguistics students like phonetics tutorials.
c) John paid a dollar for a head of lettuce.
d) Teenagers drive rather quickly.
e) Martha said that Bill loved his Cheerios in the morning.
f) Eloise wants you to study a new language. [assume to = T]
g) For Maurice to quarrel with Joel frightened Maggie.
h) John’s drum will always bother me.
5. TREES II
Again you are on your own. Warning, many of the trees are ambiguous

1. HUNGARIAN
No, the possessor has to appear in spec, NP. In the second construction it appears in
spec, DP.
2. NPI LICENSERS
If the D were buried within the NP, it couldn't c-command the NPI to license it (so we
would predict (a) to be bad).
3. ENGLISH MODALS AND AUXILIARIES
You are only allowed one modal (=T), but multiple Auxs (=V) (e.g., I should have had
been being kissed when Margo walked into the room and interrupted the frivolity).

Constraining X-bar Theory: The
Lexicon
0. INTRODUCTION
The system of X-bar theory developed in the previous chapter seriously over-generates. This chapter
introduces theta roles and thematic relations as a way to constrain this. It builds upon the discussion in
chapter 2, so if you skipped chapter 2, you may want to revisit concepts like transitivity and feature
structures.
Two important points of note: (1) I distinguish thematic relations from theta roles. Thematic relations
are the semantic notions. Theta roles are the syntactic slots associated with a group of theta roles. The
theta criterion holds of theta roles, not thematic relations. (2) I use Haegeman’s box notation for theta
grids.
The EPP and expletives are also introduced in this chapter, as is the argument that expletive
insertion is motivated by the EPP, but that this happens after the theta criterion has been checked.
1. IDENTIFYING THEMATIC R ELATIONS
Part 1
a) Shannon: Agent, Source.
Dan: Recipient (Goal).

email: theme
b) Jerid: Experiencer
[that Sumayya ...]: Proposition
Sumayya: Agent
beef waffles: theme
him: Beneficiary
c) Stacy: Agent, Source.
a baseball: Theme
Yosuke: Goal (Recipient?)
d) Jaime: Agent
jig: Theme
e) Yuko: Agent
the pizza: Theme
a garlic clove: Instrument
f) It: none (expletive)
San Francisco: Location
Part 2

Chapter 8: Constraining X-bar Theory: The Lexicon 73
b) CP
C'
C TP
Ø
DP T'
D' T VP
-s
D NP V'
N' V CP
think
N C'
Jerid
C TP
that
DP T'
D' T VP
-ed
D NP V'
Ø
N' V' PP
N V DP P'
Sumayya cook
D' P DP
for
D NP D'
some
N' D NP
Ø
AdjP N N'
waffles
Adj' N
him
Adj
beef

c) CP
C'
C TP
Ø
DP T"
D' T VP
Ø
D NP V'
Ø
N' V' PP
N V DP P'
Stacy hit
D' P DP
to
D NP D'
a
N' D NP
Ø
N N'
baseball
N
Yosuke
You are on your own for the rest.
2. WARLPIRI
ngka (A): location

kurra (B): goal
ngirli (C): source
ngku (D): agent
wana (E): location/path
3. THETA GRIDS
a) [The stodgy professor]i left with his teaching assistant.

leave
Agent
DP
i

b) [I]i played [a tune]j on my iPod.

play
Agent Theme
DP DP
i j
c) [Molly]i gave [Calvin]j [a kiss]k.

give
Agent Recipient Theme
DP DP DP
i j k
d) [Mercedes]i gave [a test]j [to the students]k in the lecture hall.

give
Agent Theme Recipient
DP DP PP
i j k
e) [Pangur]i ate [a cat treat]j.

eat
Agent Theme
DP DP
i j
f) [Susan]i ate yesterday at the restaurant.

eat
Agent
DP
i
g) [Gwen]i saw [a fire truck]j.

see
Experiencer Theme
DP DP
i j
h) look at. This is tricky, Presumably the goal is specified to be a PP.
i) [Michael]i asked [a question]j.

ask
Agent Theme
DP DP
i j
j) [Adam]i asked [if [Hyacinth]k likes [pineapples]l]j.

ask
Agent Proposition
DP CP
i j

like
Experiencer Theme
DP DP
k l
k) It is sunny [in San Francisico]i.

(be) sunny
Location
PP
i
l) [I]i feel [it is unfortunate [that [television]l is so

vulgar these days]k]j.
feel
Experiencer Proposition
DP CP
i j
(be) unfortunate
Proposition
CP
k
(be) vulgar
Theme
DP
l
m) [That [Angus]j hates [Sushi]k]i is mysterious.

(be) mysterious
Proposition
CP
i
hates
Experiencer Theme
DP DP
j k
4. SINHALA
1) See below
2) See below
3) Experiencer
4) maʈə is used with experiencers
5) A attaches to Agents, B attaches to experiencers

a) Mamə kawi kiənəwa.

agent theme
i j
b) Maʈə kawi kiəwenəwa

experiencer theme
i j
c) Lamea kataawə ahanəwa

agent theme




æhenəwa.
exa.
exə æhenəwa.
experiencer theme
i j
e) Mamə naʈənəwa.
agent
i
f) Maʈə næʈəenəwa.
experiencer
i
g) Hæmə irida mə mamə koləmbə yanəwa

agent goal
i j
h) Hæmə irida mə maʈə koləmbə yæwenəwa

experiencer goal
i j
i) Malli nitərəmə aňɖənəwa.

agent
i
j) Malliʈə nitərəmə æňɖənəwa̤.

experiencer
i
k) Mamə untə baninəwa

agent theme
i j
l) Maʈə untə bænenəwa.

experiencer theme
i j
m) Apiʈə pansələ peenəwa

experiencer Theme
i j
5. THETA CRITERION

a) hates
Experiencer Theme
DP DP
j k
No Theme DP, as required by grid.
b) No theta-role available for "breadbox".
c) Give requires three arguments; only two in sentence.
d) Give requires three arguments; only two in sentence.
e) Place requires Agent DP, not in in sentence.
f) Place requires Theme DP, not in sentence.
g) Place requires location PP, not in sentence.
h) No theta-role for fourth argument.
i) "The rock" can't be an animate Agent
j) Placed requires a PP; no theta-role for third DP.
1. IRISH AND THE THETA CRITERION
a) The theme role seems to be applied twice, once to An fear, the other time to é (a
resumptive pronoun).
b) The subject role is in the subject agreement (pro-drop).
c) The object appears twice, once in the object, the second time in the possessive
clitic (clitic doubling).
2. OBJECT EXPLETIVES
Maybe. It is the double of the relative clause that modifies it. Otherwise this
sentence would be an example of a counterexample to the theta criterion.
3. PASSIVES
Part 1:
a) Johni bit the applej.
agent theme
i j
b) Susani forgave Louisj.

agent theme
experiencer
i j

c) The jockeyi rides the horsej.

agent theme
i j
d) Phillipi gave the medalj to the soldierk.

agent theme recipient
i j k
e) Phillipi gave the soldierj the medalk.

agent recipient theme
i j k
Part 2: Delete external argument (delete agent might also be consistent). Some students
(usually the bright ones) will also say that the second theta role becomes the
external argument. While technically this is not correct, they can’t know this until
they do chapter 9. Other students claim that the theme becomes an experiencer, I’m
less thrilled with this answer, especially in light of how one can experience giving.
But since the definition of experiencer is so wishy-washy, it is hard to argue
against.
4. HIAKI -WA
Part 1: Deletes agents, similar to English –en.

Part 2: Only agents are affected by –wa. Also note that intransitives can be affected
by the suffix.
5. ANTIPASSIVES
Note the question is only about theta grids, but of course there are also case changes
going on here too (the element normally marked with absolutive is deleted, and the
element that otherwise is marked with ergative becomes absolutive).
Part 1: Delete the theme.

Part 2: Similar to antipassive, but doesn’t allow a by-phrase expression of the theme.
Also seems to be limited to certain verbs (not required by question).

Head-to-Head Movement
0. INTRODUCTION
This is the first chapter on transformations. It presents a relatively straightforward view of head-to-head
movement, starting by contrasting French and English adverbial and negative placement, moving on to
VSO order in Irish (the VP-internal subject hypothesis is also introduced here). Then we move on to T 
C movement and the interaction of V  T and T  C.
One thing that is unusual about this chapter, and that you should be aware of, is that in the first
half of the chapter, I continue to treat all auxiliaries as generated under T, and not raised from V  T.
However, I attempt to motivate a stacked VP structure in the second half to account for multiple auxiliary
constructions. This is a radical change from the first edition. I use VPs here instead of vPs, which are
introduced in chapter 13.
I cast transformations in the D-structure  Transformations  S-structure model, leaving aside
questions of the Minimalist model until chapter 12.
1. ITALIAN
V raising.

Chapter 9: Head-to-Head Movement 85
2. HAITIAN CREOLE VERB PLACEMENT
Affix lowering.
3. FLAT VS. HIERARCHICAL STRUCTURE: BERBER
If the NPs c-commanded each other (the flat structure analysis), (a) ought to
be out due to Condition C. In (b), both theories correctly predict
ungrammaticality, but for different reasons: Condition A on the VP analysis,
Condition C on the flat structure analysis.
4. WELSH
The verb shows up in the auxiliary slot left of the subject when there is no
independent T (non-affix) node.

5. VP-INTERNAL SUBJECTS: PRACTICE
a) CP
C'
C TP
Ø
DP T'
D' T NegP
D NP Neg'
Ø
N' Neg VP
not
N V'
Tiffany
V VP
be+ing
t V'
V' PP
V DP P'
take
D' P DP
until
D NP D'
her
N' D NP
Ø
AdjP N N'
class
Adj' AdjP N'
Adj Adj' N
syntax year
Adj
next

b) CP
C'
C TP
Ø
DP T'
D' T
-s VP
D NP
N' t V'
N V' PP
Christine
V DP P'
like
D' P DP
with
D NP D'
Ø
N' D NP
a
AdjP N N'
furniture
Adj' AdjP N'
Adj Adj' N
wood finish
Adj
dark

c) CP
C'
C TP
Ø
DP T'
D' T NegP
D NP Neg'
les
N' Neg VP
pas
N V'
enfants
V VP
(n)ont
t V'
V
travaillé
d) CP
C'
C TP
Ø
DP T'
D' T NegP
(ne)-ent
D NP Neg'
les
N' Neg
pas
N
enfants
VP
t V'
V
travaill-

6. AMERICAN VS. BRITISH ENGLISH VERB HAVE
Part 1: Auxiliary have undergoes T  C but possessive have does not. (never
intervenes)
Part 2: Yes, (suggesting that it has undergone V  T).
7. HEBREW CONSTRUCT STATE (N → D)
Part 1: Overt complementizers are in complementary distribution with construct states.
Part 2: The agreement phenomena show that the construct state N, agrees with the final
adjective. If we take seriously the idea that modifiers are always sisters (or more
accurately are attached to a projection of the element they modify), then at some
point the construct N had to be low and near the adjective.
8. ENGLISH
Adjectives normally precede nouns in English, the adjectives here follow. Suggesting
that at least part of the noun has moved around them. This is supported by the
alternation between some spicy thing and something spicy. (However, of course,
contrast a spicy something!)
9. ENGLISH TREES

a) CP
C'
C TP
T'
T VP
V'
VPerf VP
have en DP V'
D' AdvP V'
D NP Adv' V DP
love
N' Adv D'
always
N D NP
I
N'
N
peanut butter
(some students may treat peanut butter as a phrase)

b) CP
C'
C TP
T'
T NegP
do
Neg'
Neg VP
not
DP
D' V'
D NP V DP
love
N' D'
N D NP
I
N'
N
peanut butter

c) CP
C'
C TP
T'
T VP
-s
DP V'
D' AdvP V'
D NP
Adv' V CP
think
N' Adv C'
often
N C TP
Martha
T'
T VP
-s
DP V'
D' V DP
hate
D NP D'
N' D NP
N N'
Kim
N
phonology

d) CP
C'
C TP
T'
T VP
do
DP V'
D' V DP
like
D NP D'
N' D NP
N N'
you
N
peanut butter
e) CP
C'
C TP
T'
T VP
V'
VPerf VP
have en DP V'
D' AdvP V'
D NP Adv' V DP
hate
N' Adv D'
always
N D NP
you
N'
N
peanut butter

f) CP
C'
C TP
DP T'
D' T AdjP
are
D NP AdvP Adj'
N' always Adj'
N AdvP Adj
you obtuse
Adv'
Adv
so
g) CP
C'
C TP
T'
T VP
will
DP V'
D' V DP
bring
D NP D'
N' D NP
your
N N'
you
N
spouse

h) CP
C'
C TP
T'
T VP
-s
V'
VPass VP
have en V'
V VP
be en V'
V DP
eat
D'
D NP
the
N'
N
food

i) CP
C'
C TP
T'
T VP
-s
V'
VProg VP
be -ing DP V'
D' AdvP V'
D NP Adv' V DP
eat
N' Adv D'
always
N D NP
Mike
N'
N
peanuts
1. FLOATING QUANTIFIERS
In example (b), there is a discontinuous constituent [the men ... all]. This
could be created by movement from the specifier of VP if just the embedded DP
is moved, stranding the embedding DP (including the quantifier) within the
VP.
2. VERB MOVEMENT
German: Yes, main verbs appear higher than negation and undergo T  C.
Persian: Yes the main verb undergoes T  C. The data in (c) and (d) is
trickier, but there it appears as if negation attaches to T, and in (c) it is

attached to the V, suggesting the verb has raised to T. The word order is
surprising, however.
3. GERMANIC VERB SECOND
Part 1: You don’t get V2 when there is an overt complementizer. The

phenomenon is in complementary distribution with overt complementizers,
suggesting that complementizers are involved.
Part 2: V raising.
bonus: depends on their answer, but yes
4. PROPER NAMES AND PRONOUNS
Part 1: Proper names are (usually) in complementary distribution with

determiners. The fact that you can have The Smiths, suggests that under
certain circumstances the N remains downstairs. The alternation (Smith vs.
The Smiths) suggests that movement is involved.
Part 2: D, they are in complementary distribution with determiners (it is an

open question as to whether they undergo N  D movement).
5. ITALIAN N → D
When there is no determiner, the adjective follows the noun, when there is,
it precedes the noun. Assuming that adjectives are fixed in position, the
alternation suggests that the N has raised around the AP. The answer to this
is very similar to the answer to question 7.

DP Movement
0. INTRODUCTION
This chapter presents a fairly standard view of DP movement. I treat both passives and subject-to-subject
raising in this chapter. Subject to object raising and control constructions are left until chapter 14 (I
introduce AgrO in chapter 13 to give a landing site for object movement). I start out with an EPP based
explanation but develop it into Case theory. Movement is explained in terms of feature checking for the
first time.
1. HAITIAN CREOLE
1) DP movement in Haitian Creole leaves a resumptive pronoun. 2) This pronoun is an

“overt trace”, creating a DP–Pronoun chain (thus not violating the theta criterion).

Chapter 10: DP Movement 99
2. ARIZONA TEWA
1) Spec: XP  (YP) X'

Adjunct: X'  (ZP) X'
Complement: X'  (WP) X but DP is X (WP)
2a) CP
C'
C TP
Ø
DP T'
D' T VP
Ø
D NP V'
hȩ’i
N' DP V
mánkhwɛ̧́di.
N D'
sen
D NP
nɛ́’i
N'
N
‘enú

c) CP
C'
C TP
Ø
DP T'
D' T VP
Ø
D NP V'
na:bí
N' DP V
mánsunt’ó
N D'
kwiyó
D NP
hȩ’i
N'
N
p’o
3) Same as the English passive

4) b) CP
C'
C TP
Ø
T'
T VP
Ø
V'
DP V'
D' DP V
mánkhwɛ̧́di
D NP D'
hȩ’i
N' D NP
nɛ́’i
N N'
sen-di
N
‘enú
d) CP
C'
C TP
Ø
T'
T VP
Ø
V'
DP V
nasunt’íi
D'
D NP
he¶’i
N'
N
p’o

3. PERSIAN ACCUSATIVE CASE
Features don’t have to be overtly morphologically expressed. We can have abstract case
features. This parallels the discussion of English nouns in the body of the text. In
fact, it looks like râ realizes both Acc and definite or specific.
4. TURKISH
The DP already has case in the embedded clause, therefore there is no motivation for
raising.
5. IMPERSONALS IN UKRAINIAN, KANNADA, AND IRISH
In these languages, impersonals can still assign accusative case, unlike passives.

6. ENGLISH
a) CP
C'
C TP
T'
T AdjP
is
Adj'
Adj CP
likely
C'
C TP
T'
T VP
to
DP V'
D' V DP
leave
D NP D'
N' D NP
the
N N'
Marie
N
store

b) CP
C'
C TP
T'
T VP
-ed
V'
VPass VP
be en V'
V' PP
V DP P'
hide
D' P DP
in
D NP D'
the
N' D NP
the
N N'
money
N
drawer

c) CP
C'
C TP
T'
T AdjP
is
Adj'
Adj CP
likely
C'
C TP
T'
T VP
to
V'
VPerf VP
have en V'
VPass VP
be en V'
V' PP
V DP P'
kiss
D' P DP
by
D NP D'
N' D NP
the
N N'
Donny
N
puppy

d) CP
C'
C TP
T'
T VP
-s
It V'
V CP
seem
C'
C TP
that
T'
T VP
-s
DP V'
D' V DP
love
D NP D'
N' D NP
N N'
Sonny
N
Cher

e) CP
C'
C TP
T'
T VP
-s
V'
VPass VP
have en V'
V VP
be en V'
V DP
eat
D'
D NP
the
N'
N
rice
7. ENGLISH UNGRAMMATICAL SENTENCES
a) Sonny can’t get case in the embedded subject position, because of non-finite T.
b) Well this is a theta criterion violation, but in terms of case, at least one DP
doesn’t get case, since only the [NOM] DP gets case.
c) Donny already has case in the embedded clause.
8. UNACCUSATIVES AND PASSIVES
You can only delete the external argument once, and only suppress the ACC case feature
once. (Again the 1AEX of relational grammar.)

9. ICELANDIC QUIRKY CASE
These DPs already have case. There is no motivation for the movement. The solution may
lie in distinguishing abstract case from overt morphological case. The quirky case
marked NPs in Icelandic take both a quirky case (dative) and an abstract case (either
accusative or nominative).
1 MIDDLES AND PASSIVES
Part 1: Middle: theme

Active: instrument theme
Causative: agent theme (instrument)
Part 2: This is essentially the Relational Grammar Law 1AEX. Middles are already
passive like, so they don’t allow further passivization.
2. PASSIVES AND DOUBLE OBJECTS
Only the NP adjacent to the verb may passivize, suggesting that this is the element
that gets accusative case.
3. TWO KINDS OF RAISING
d) shows accusative case marking on the embedded subject, since acc case is assigned
to a sister of V, then this is the correct surface position of the NP.
f,g,h) These three pieces of data show that the subject of the embedded clause is
behaving like it is part of the higher clause with respect to binding theory, it
allows disjoint reference in (e). For many people it may be bound by the subject when
it is an anaphor in (f), and in (g) it cannot be coreferent with the main clause
subject.
h) It can get raised to subject position of the main clause in a passive, just like a
regular object.

Wh-movement
0. INTRODUCTION
Basic wh-movement is addressed in this chapter, as well as some of the easier constraints. Wh-movement
is cast as movement for feature checking reasons.
New to this edition: the book no longer uses subjacency, instead a Minimal Link Condition
account is given of wh-islands is given, it's also extended to DP movement and head-movement. There is
no longer any extensive discussion of complex DP islands, although there is a very brief description of DP
islands and other island types. There is also an improved discussion of wh-in situ.
1. ENGLISH MOVEMENT SENTENCES
For ease of exposition, I’ve triangled all the DPs, although these
should be spelled out.

a) CP WH-FEATURE HERE
C’
C TP CASE & EPP HERE

[+Q, +WH]
T’
T[NOM] VP
[pres]
V’
V VP THETA HERE
is+ing
DP[nom] V’
what V DP[acc]
v bother
you
b) CP WH FEATURE HERE
C’

[+Q, +WH]
T’
T[NOM] VP
[pres]
V’
V VP THETA HERE
has+en
DP[nom] V’
who V DP[acc]
see
my snorkle

Chapter 11: Wh-Movement 111
c) CP WH FEATURES ON HOW
C’
C TP CASE & EPP ON PLOT

[+Q, +WH]
T’
T[NOM] VP
[pres]
V’
V VP
was+en
V’
V’ AP
V’ PP how
THE PLOT IS A THEME! V DP[nom] by the auths

discover
the plot

d) Which animals appear to have lost their collars?
CP WH FEATURES HERE
C’
C TP CASE AND EPP HERE

[+wh]
T’
T VP NO SUBJECT/AUX INVERSION???
pres
V’
V CP
appear
C’
C TP EPP HERE
[-wh]
T’
T VP
to
V’
V VP THETA ROLE HERE

have-en
DP V’
which animals V DP
lose
their collars

e) What did Jean think was likely to have been stolen?
CP WH FEATURES HERE
C’
C TP CASE AND EPP For Jean HERE

[+Q+wh]
T’
T VP DO Insertion
past
DP V’
Jean V CP Wh-movement STOPS here for the MLC

think
C’

[-wh]
T’
T VP
past
nom V’
V AP
is
A’
What does NOT
A CP stop here (no
likely case yet!!)
C’
C TP EPP
T’
T VP
to
V’
V VP
have+en
V’
NO Agent role V VP
Passive be+en
V’
V DP[Nom]
steal
what

f)
CP
C’
C TP CASE & EPP

[-Q, -WH]
T’
T[NOM] VP
[pres]
V’
V VP
has+en
DP[nom] V’
car sales V DP[acc]

surprise
the stockb..
g) CP
C’
C TP CASE & EPP

[+Q, -WH]
T’
T[NOM] VP
[pres]
V’
V VP
has+en
DP[nom] V’
you V DP[acc]
see
my model …

Chapter 11: Wh-movement 115
h) CP
C’
C TP
[+Q, -WH]
T’
T[NOM] VP
can
DP V’
you V DP[acc]
find
the lightbulb store
i) CP
C’
C TP Case & EPP

[+Q, -WH]
T’
T[NOM] VP
[pres]
V’
V VP
was+en
V’
V’ PP
V DP[nom] by the ad...

bite
John

j) CP
C’
C TP Expletive Insertion
[-Q-wh]
T’
T VP
pres
nom V’
V AP
is
A’
A CP
likely
C’
C TP EPP & Case

that
T’
T VP
will
DP V’
Tami V DP
leave
NY

k)
CP
C’
C TP Case and EPP

[-Q-wh]
T’
T VP
pres
nom V’
V AP
is
A’
A CP NOT a potential landing site for case

likely
C’
C TP EPP but not Case

Ø
T’
T VP
to
DP V’
Tami V DP
leave
NY

l)
CP
C’
[-Q-wh]
T’
T VP
-s
nom V’
V CP
seem
C’
C TP EPP & Case

that
T’
T VP
[past]
V’
V VP PASSIVE!!!!
be+en
V’
V DP
mug
Susy

m)
CP
C’
C TP Case and EPP

[-Q-wh]
T’
T VP
-s
nom V’
V CP does not stop here, because no case yet

seem
C’
C TP EPP
Ø
T’
T VP
to
V’
V VP
have+en
V’
V VP PASSIVE!!!!
be+en
V’
V DP
mug
Susy

n) CP
C’
C TP Case & EPP for you

[+Q, +WH]
T’
T[NOM] VP
[past]
DP V’
DID
you V’ PP
V DP at the supermarket
buy
what
Gets case & theta here

o) CP
C’
C TP EPP and case for I

[-Q-wh]
T’
T VP
-ed
DP V’
I V CP Wh-features checked here

ask
C’
C TP EPP & Case for Beth

+wh
T’
T VP
[past]
DP V’
Beth V’ PP
buy
what
gets theta and case here

p) What is it likely for Beth to have bought at the supermarket?

(Treat the PP for Beth as appearing the specifier of the embedded TP.)
CP
C’
[+Q+wh]
T’
T VP
pres
nom V’
V AP
is
A’
A CP STOPS HERE FOR THE MLC

likely
C’
C TP EPP for “For Beth”

[-Q-wh]
T’
T VP
to
V’
V VP
have+en
PP V’
for Beth V’ PP
bought
what
gets theta and case here

q) What is likely to have been bought at the supermarket?

CP
C’
C TP Case and EPP

[+Q+wh]
T’
T VP
pres
nom V’
V AP
is
A’
A CP Does NOT stop here

likely
C’
C TP EPP for “what”

[-Q-wh]
T’
T VP
to
V’
V VP
have+en
V’ PASSIVE!!
V VP
be+en
V’
V’ PP
bought
what
gets theta here

2. BINDING THEORY
Binding conditions hold at D-structure, before the transformations have moved

the wh-phrase.
3. BINDING AND SCRAMBLING
Wh-movement, because the anaphor can still be bound even though it has moved.
4. PICTURE DPS
Should be a violation of the subjacency condition. Who crosses the NP node

[a picture of] and the TP node of the sentence. There is no intermediate CP
to stop off in.
5. LOCALITY
Students are supposed to draw the tree. I’ve just circled the bounding nodes
here:
*Whoj did [TP George try to find out [CP whati [TP tj wanted ti]?
The wh-phrase has to cross both these, creating a subjacency violation. It

can’t stop off in the spec of CP because of the what.
6. IRISH
Complementizer agreement shows that the wh-word stops off in the intermediate
specifier. There would be no aL comp in the embedded CP if the wh-word didn’t
stop there.

1. WHO ATE THE PIZZA?
1. The vacuous movement account leads us to expect (i).

(i) *Who did eat the pizza?
Maintaining the vacuous movement account would require that English
main verbs undergo V-to_T movement just in this case.
2. Hard question. Maybe selection?
2. IRISH
You get the aN–resumptive strategy obligatorily when you have an island (=
subjacency violation). A note on the data: this is a simplification of the
facts. You can actually get the aN-resumptive strategy in any position, except
for the highest subject position, but it is only obligatory in an island. See
McCloskey (1991) for more details.
3. SERBIAN/CROATIAN/BOSNIAN
All the wh-words seem to move. Various alternatives can be proposed, some
more likely than others: Multiple CPs, (or other multiple functional
categories high in the tree). Other students have suggested that the wh-words
all combine into one wh-phrase. This latter answer is obviously less
satisfying.
4. FRENCH N EGATION
The data argues for the (b) analysis, as in the (a) analysis the sentence would
violate the MLC.

A Unified Theory of Movement
0. INTRODUCTION
This chapter introduces some of the ideas that take us towards Minimalism. it is by no means a complete
introduction to the paradigm, but it offers hints in that direction. Some of the main points:
a) motivating all movement via feature checking and Full Interpretation

b) distinguishing LF from PF movement
c) accounting for language-specific variation in terms of timing between overt and covert movement
d) a brief discussion of LF operations (like QR).
e) I use SPELLOUT instead of S-structure (new to this edition)
1. FRENCH
Overt DP movement (subject always appears left of T).

Chapter 12: A Unified Theory of Movement 127
2. IRISH
overt wh-movement,
covert DP-movement,
overt head-movement.
3. PF MOVEMENT
I was looking for sentences that might be synonymous or nearly synonymous, for
example, particle movement. Of course there are always those that will claim that that
all movement types have some effect on meaning (broadly interpreted to include so
called “pragmatic” issues like topic and focus). The answers you’ll get on this
question will depend upon the semantic sophistication of your students.
1. SERBIAN/CROATIAN/BOSNIAN
English only allows one overt wh-movement. Serbo-Croatian requires that all movement
be overt.
2. NEPALI & MONGOLIAN
Wh-movement is covert in both languages.
3. ECHO QUESTIONS IN ENGLISH
The MLC holds that movement is to the closest potential landing site. With wh-islands
there is an intermediate, but filled spec, CP that serves to block further movement.
With NP islands, however, there is no such “closer potential landing site” to block
extraction.

4. SCOPE OF NEGATION
Part 1.
(i) CP
QPj C'
many C TP
mistakes Ø
DPi T'
the editor T NegP

did
Neg'
Neg VP
not
ti V'
V tj
find
(ii) CP
C'
C TP
Ø
DPi T'
the editor T NegP

did
Neg'
Neg VP
not
ti V'
V QP
find
many mistakes
Part 2. Since negation cannot move higher than the moved QP, and the QP cannot lower,
the QP must take wide scope over negation.
Part 3. Actually, what's crucial here is not that the movement is overt per se, but
just that it happens at all. Obligatory covert movement, as in Japanese and Chinese
wh-questions, would produce the same effect. But since it's overt, it must be part of
LF.

Split VPs
0. INTRODUCTION
This chapter is the first of a series of new chapters to this edition. I will be frank with you, these chapters
represent a big jump in the level of the book. I personally don't recommend using them as part of a
typical 14-week introductory class. Instead I think they are best suited for giving enrichment material to
good students, or for use in those rare cases when this book is used for a graduate level text. I taught all
the way through this material the last time I taught LING300 and all but the best students were pretty
lost, so I need to give them a lot of handholding. Supplementing these chapters with primary material is
probably a good idea.
This chapter introduces two new functional categories: AgrO and vP. AgrS is not treated in this
chapter, except in an exercise. AgrO (and AgrIO) are taken to be interleaved with atomic v heads (such as
CAUSE or GOAL). Empirical motivation is given in terms of object shift, scrambling, dative movement, and
the interaction of these with Antecedent Contained Deletion, Pseudogapping and Ellipsis. The treatments
are by no means complete. I mean the chapter to give the student a taste of this material without getting
too much into the details of this complicated debate.

1. NPIS AND DOUBLE OBJECT VERBS
The NPIs must be c-commanded by a negative element to be licensed. The data in (a)
and (b) show that the first object c-commands the second, and that the second does not
c-command the first. This is consistent with the structure in (5).
2. COMPLEX VERBS
Persian and Chichewa have morphologically complex items where English would use a
single verb. The (b) example is particularly interesting, as it has something like
CAUSE+FALL -- exactly what the light verb analysis leads us to expect as the
underlying structure for a meaning like "knock over".
3. PARTICLES
I. In the proposed structure, both "blow" and "blow up" count as V heads. Then
either the verb alone or the V+P complex can undergo head movement to the light verb
position. The first option produces (i); the second, (ii):
(i) [Vblow]i the building ti up.
(ii) [Vblow up]i the building ti.
4. THETA GRIDS
vCAUSE
Agent Predicate
DP AgrOP
i j
√CLEAN
Theme
DP
k

Chapter 13: Split VPs 131
5. TREES
a) CP
C'
C TP
Ø
T'
T vP
-ed
DP v'
Susan v AgrOP
CAUSE
AgrO'
AgrO VP
DP V'
the package V PP
√SEND
to Heidi

b) CP
C'
C TP
Ø
T'
T vP
-ed
DP v'
Carolyn v AgrIOP
CAUSE
AgrIO'
AgrIO vP
DP v'
Heidi v AgrOP
LOCATE
AgrO'
AgrO VP
V'
V DP
√SEND
a package

c) CP
C'
C TP
Ø
T'
T vP
-ed
DP v'
Peter v AgrOP
CAUSE
AgrO'
AgrO VP
DP V'
the letter V PP
√PLACE
in the envelope

d) CP
C'
C TP
Ø
T'
T vP
-ed
DP v'
I V AgrOP
CAUSE
AgrO'
AgrO VP
DP V'
Mike V CP
√ASK
C'
C TP
if
he had seen the yeti

e) CP
C'
C TP
Ø
T'
T vP
-ed
DP v'
I v AgrOP
CAUSE
AgrO'
AgrO VP
DP V'
some flowers V PP
√BUY
for Manuel

f) CP
C'
C TP
Ø
T'
T vP
-ed
DP v'
I v AgrIOP
CAUSE
AgrIO'
AgrIO vP
DP v'
Manuel v AgrOP
LOCATE
AgrO'
AgrO VP
V'
V DP
√BUY
some flowers
6. APPLICATIVES IN BAHASA INDONESIAN
The applicative morpheme may be analyzed as the AgrIOP head giving Case to the extra
DP, attached to the verb by head-movement. Note that it does not appear in (a), where
we postulate less structure (the DP gets Case from the preposition).

1. THAT DOG DOESN'T HUNT
No, this is a sentential idiom. Note in particular that negation is required; "That
dog hunts" cannot have the idiomatic reading. Also, Hunt can normally take an object,
which we expect it to be able to do here (with any suitable object DP); again, this
makes the idiomatic reading unavailable: "That dog doesn't hunt rabbits".
2. AGAIN
This is consistent with the idea that the modifier can be attached to either the vP or
the VP. Analyzing open as CAUSE+OPEN, it can either be caused-by-John again, or made
open again.
In fact, it has been argued that there is a third reading wherein the door was shipped
from the factory open (no one had caused it to be open before), suggesting the actual
structure is CAUSE+BECOME+OPEN.
3. AGRS
Part 1. We analyze "there" as occupying the Specifier of TP for EPP reasons, and the
subjects "a man/four men" as occupying the Specifier of AgrS. With head movement of
the verb to T, this yields the correct surface order. Note further that it is the
lower subjects, not "there", which determine the number agreement on the verb,
consistent with the claim that whatever position they move to is associated with
agreement.
Part 2. There is an Agr particle aL between the subject and the v element.

Raising, Control, and Empty
Categories
0. INTRODUCTION
This chapter introduces the distinctions between raising and control constructions, between subject to
subject raising and subject to object raising, between obligatory and optional control, and between subject
control and object control. There is also a short discussion of little pro and the null subject parameter.
Subject to Object raising is treated as raising to the spec of the AgrO in the higher clause.
Control theory is notoriously problematic. Here I give a less than satisfying answer, but that I
think reflects the less that satisfying status of control theory at this time. I discuss the various syntactic,
thematic/semantic, and pragmatic accounts that have been suggested for controlling PRO. I do not
consider movement analyses of PRO. Some people were critical of my treatment of PRO in terms of Case,
but that seems to be the recent trend at least within Chomsky's work. Case here is taken to be a short
hand for government, so feel free to substitute in government.
You may want to discuss the raising of PRO for EPP reasons.

Chapter 14: Raising, Control, and Empty Categories 139
1. THE EXISTENCE OF PRO
Himself is too far away from Robert to meet condition A, so PRO must serve as the
antecedent.
2. RAISING TO OBJECT
The landing site must be in the main clause; if these items remained in the lower
clause, we would expect them to form a continuous string with the embedded clause.

140 Syntax: Instructor's Handbook
(b) CP
C'
C TP
Ø
T'
T vP
-ed
DP v'
D' v AgrOP
D NP AgrO'
Ø
N' AgrO VP
N V'
Mike
AdvP V'
Adv' V CP
expect
Adv C'
incorrectly
C TP
Ø
T'
T vP
to
DP v'
D' v AgrOP
D NP AgrO'
Ø
N' AgrO VP
N V'
Greg
V DP
V P D'
take out
D NP
the
N'
N Trash

Chapter 14: Raising, Control, and Empty Categories 141
3. ICELANDIC PRO AND QUIRKY CASE
The quantifier here agrees in case with the quirky case marking that PRO
would get (note, not the case on the main clause subject that controls PRO).
So this means that PRO must be getting case. This problem is related to the
one in the previous chapter, where passive occurred even though the NP got
quirky case. A similar solution applies: we distinguish morphological from
abstract case.
4. ENGLISH PREDICATES
Warning: this is a very long (and tedious) problem set to do!
is eager Subject control

is believed SSR
seems SSR
is ready SC
persuaded OC
urged OC
requested SC (and maybe Object control)
hoped SC
expect SOR
force OC
tell OC
advise OC
ask OC, SC
assure This is a really weird verb. It doesn’t allow any of them,
but it does allow a wh-trace in the subject position of the
embedded clause. Is assured is a SSR raising verb.
imagine SOR
promise SC
want SC or SOR
is likely SSR
consent SC
imagine SOR
encouraged OC
intended SOR, SC
5. TREES AND DERIVATIONS
I haven’t bothered to do the trees here. I’ve just indicated traces and PROs. You’ll
also note I’m ignoring the existence of the VP-internal subject hypothesis and
movement in embedded clauses for EPP reasons.

a) Jean wants Billi [CP ti to do the Macarena]. SOR, affix lowering

b) Roberti is eager PROi to do his homework. no transformations
c) Jeani seems [CP ti to be in a good mood]. affix lowering, SSR
d) Rosemaryi tried PROi to get a new car. affix lowering
e) Susan begged Billi PROi to let her sing in the concert.
Students may require some discussion of the “let her sing in the concert” part
of this sentence (in terms of SOR)
f) Susani begged PROi to be allowed ti to sing in the concert.
g) Christinai is ready PROi to leave.
h) Fred was believed ti to have wanted PROi to try PROi to dance.
I leave it for discussion as to whether there is a trace only in the specifier
of the first embedded TP, or whether there is also one in the main clause
object position.
i) Susan consented PROi to try PROi to seem ti to have been kissed ti.
Again the existence of the chain of the two traces and the lower PRO needs to
be discussed, since none of these are case positions.
j) Alan told me who wanted to seem to be invincible.
k) What did John want to eat?
6. IRISH pro
Irish doesn’t allow you to have both agreement and the pronoun.
1. IS EASY
Both, there is an arbitrary PRO as the subject of the embedded clause, the object of
the embedded clause raises to main clause subject position.
2. CONTROLLERS
a) To improve myself is a goal for next year.

PRO = me obligatorily (as shown by the anaphor bound by PRO) but there is no me
to control it.
b) To improve yourself would be a good idea.
same answer as (a)
c) To improve himself, Bruce should consider therapy.
PRO = Bruce obligatorily, but Bruce does not c-command it.
d) To improve herself, Jane went to a health spa.
same answer as (c)

Advanced Binding
This chapter offers a Knowledge of Language-style analysis of binding, where the definition of
binding domain is relativized to the type of NP. Anaphors look for the smallest CP/DP containing a
potential antecedent. Pronouns look for the smallest CP/DP not containing a potential antecedent. Facts
about anaphors in the subject position of tensed embedded clause are attributed to the lack of a
nominative anaphor.


1. BINDING DOMAIN FOR ANAPHORS
a) CP The binding domain is the CP.
C'
C TP
Ø
T'
T vP
-s
DP v'
D' v AgrOP
D NP AgrO'
Ø
N' AgrO VP
N V'
John
V DP
love
D'
D NP
Ø
N'
N
himself
I leave b-j for you to figure out on your own

Chapter 15: Advanced Binding 145
2. BINDING DOMAIN FOR PRONOUNS
a) CP The binding domain is the CP.

The pronoun is bound within it's domain,
C' violating Condition B.
C TP
Ø
T'
T vP
-s
DP v'
D' v AgrOP
D NP AgrO'
Ø
N' AgrO VP
N V'
John
V DP
love
D'
D NP
Ø
N'
N
him
I leave b-f for you to figure out on your own
1. PRONOUNS
No. Consider a SOR structure:

(i) Johni expects himselfi [ti to win]
(ii) *Johni expects himi [ti to win]
So while for anaphors ANY link of the chain can satisfy the binding condition,
for Pronouns EVERY chain link must satisfy the condition (be free in its domain).

2. POSSESSIVE PRONOUNS
If the pronominal element is in the specifier of the DP, it can count as a potential
antecedent in the correct position to restrict the anaphor's binding domain to the DP
(rather than the whole CP).
3. WESTERN A MERICAN DIALECTS OF ENGLISH
We can delete the mention of DPs as a potential binding domain:
Binding Condition A': One copy of an anaphor in a chain must be bound within the
smallest CP containing it and a potential antecedent.

Alternative Approaches: Lexical-
Functional Grammar
0. ALTERNATIVE THEORIES
While I myself am a Minimalist, I think it is important for my students to be able to read papers and work
written in other theoretical frameworks.
This chapter and the next take a fairly big turn from the rest of the book. These are quite dense
and quite technical surveys of the machinery involved in LFG and HPSG. People who have read the
manuscript of this book tell me that these are the hardest chapters in the book and I agree. These are not
designed to be user-friendly or in-depth introductions to these frameworks. If you really want to teach
your students to do LFG or HPSG properly (as opposed to giving them a familiarity with the mechanics
and ideas underlying the theories) then these chapters are not for you. Instead, I’d recommend that you
use one of the textbooks written by practitioners of these approaches (such as Bresnan 2001 and Sag,
Wasow and Bender 2003). This chapter and the next one are designed to be quick and dirty introductions
to the technical apparatus of these approaches.
I’ve written these chapters with an eye towards a smart syntax student who has understood the
basics of P&P/Minimalism, but would like to understand the mathematical tools of these other
approaches. In doing this, I haven’t always been faithful to the ideological or philosophical
underpinnings of the approaches. For example, I occasionally say things like “this is the equivalent of
head movement” when what I really mean is “this does roughly the work that head movement does in
P&P, even though it has different assumptions and motivations.” This is likely to annoy those people
who actually work in these frameworks, as it gives a certain P&P feel to them. Again this is for
pedagogical reasons. I reiterate, if you want proper introductions to the influential ideas of LFG or

HPSG, you’d be better off going to the source material. But if you want a brief explanation of what a
metavariable or a SYN-SEM structure is, then these chapters will do the trick.
One other caveat has to do with my discussion of evaluating competing theories. A number of
people have given me quite a bit of flack about the “all theories are roughly equal” or “aesthetics play a
big role” language I use in these chapters. Let me state for the record, that of course I believe it is possible
to evaluate theoretical approaches on empirical grounds. However, I do want to point out that doing so is
extremely difficult. Some theoretical machinery is better suited to certain empirical tasks, but other
machinery might be better at different tasks. This makes empirical comparison, while not impossible,
certainly more difficult than many scholars would have us believe. This practical consideration has the
effect that for the most part people work in the theoretical framework that appeals to them on aesthetic
grounds or works best in their own particular sub-area of interest. This isn’t an ideal situation, but I think
it is wrong to try to conceal this fact from students.
This chapter, on LFG, introduces some of the original motivations for abandoning
transformations. It introduces the basic notions of c-structure, f-structure, a-structure, variables,
metavariables, grammatical functions, AVMs, functional equations, f-descriptions, unification, lexical
rules, and functional control. It treats none of these topics in depth.
1. ENGLISH
a) TPf1
(↑SUBJ) =↓ ↑=↓
NP f2 VP f3
↑=↓ ↑=↓
N' f5 V' f6
↑=↓ ↑=↓ (↑OBJ)=↓

N f7 V f8 NP f9
Susie loves
(↑PRED) = ‘Susie’ (↑PRED) = ‘love <…>’ ↑=↓ ↑=↓
(↑NUM) = sng (↑TENSE) = present D f4 N' f10
(↑SUBJ NUM) = sng the
(↑SUBJ PERS) = 3rd (↑DEF) =+ ↑=↓
N f11
rain

Chapter 16: Alternatives: Lexical Functional Grammar 149
(↑PRED) = ‘rain’
f1, f3, f6, f8, PRED ‘love <SUBJ, OBJ>’
TENSE present
SUBJ f2 f5, f7, NUM sng

PRED 'Susie'
OBJ f4, f9, f10, f11 PRED 'rain'

DEF +
b) This one requires the use of one f-structure embedded inside the other
using the COMP feature. The COMP feature isn't explicitly discussed but is
seen in the f-structures in section 5.4
c) This one requires the use of a focus function, and a line indicating that
the functions occupying the FOCUS function are identical to those of the
object:
FOCUS PRON wh
PRED 'what'
NUM sg
TENSE present perfective
PRED ‘read <(↑subj), (↑obj)>’
SUBJ [“you”]
OBJ [ ]
1. ICELANDIC (AGAIN)
No. Because case is introduced lexically; LFG doesn't use Case for licensing.
PRO is neither cased or caseless in LFG.
2. TRANSFORMATIONS OR NOT?

The answer to this question involves creativity on the part of the student. A
typical answer might include a timing experiment where transformations would
take longer than a licensing process. Of course such an argument is
fallacious due to the fact the experiment would test performance rather than
competence. Other answers may be possible.
3. WANNA-CONTRACTION
The answer to this question involves creativity on the part of the student.
One answer might be that the lexical entry for wanna (which must be a single
lexical item in LFG) contains a constraining equation preventing the Focus
from referring to it's SUBJ function. Other answers may be possible.

Alternative Approaches: Head-Driven
Phrase Structure Grammar
Please read the caveat about the alternatives chapters at the beginning of chapter 16 of this instructor’s handbook.
This chapter covers all the basics of HPSG: Features, SYN-SEM structures, tags, realization principles, gap
(slash) features, lexical rules, compositionality, unification, phrase structure rules, and binding theory. It
does not get into some of the more difficult issues, such as inheritance hierarchies. As in the last chapter,
the metaphors used are sometimes those of P&P not HPSG, for pedagogic reasons, and I make no claims
that this chapter represents HPSG as its practitioners would present it
1. ENGLISH
This is a pretty tedious question, so be warned you may have rebellion on your hands
if you assign it. The details of how to set up lexical entries and do complete SYN-SEM
structures aren't really worked out well in this chapter. I'd advise only doing this
exercise if you do it together with you class.

1. SUBJECT/AUX INVERSION
The answer to this question involves creativity on the part of the student. See Sag,
Wasow and Bender 2003 for a worked out answer consistent with current thinking in
HPSG.
2. ISLAND CONSTRAINTS
The answer to this question involves creativity on the part of the student. See Sag,
Wasow and Bender 2003 for a worked out answer consistent with current thinking in
HPSG.

Syntax Instruction

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Syntax Instruction

Uploaded by

Copyright:

Available Formats

Syntax: A Generative Introduction.

First edition published 2002 by Blackwell Publishers Ltd

For further information on Blackwell Publishing, visit our website:

Preface and Acknowledgments

Please do not copy or distribute this answer key to your students.

Please do not copy or distribute this answer key to your students.

Please do not copy or distribute this answer key to your students

Please do not copy or distribute this answer key to your students

Please do not copy or distribute this answer key to your students

1. TOPICS COVERED IN THIS C HAPTER

Please do not copy or distribute this answer key to your students.

2. IF YOU SKIP THIS C HAPTER

3. IDIOSYNCRASIES IN THIS C HAPTER

GENERAL PROBLEM SETS

Common answers include: maintaining a standard across dialects for effective

a) Prescriptive (who instead of whom)

Please do not copy or distribute this answer key to your students

d) Descriptive, syntactic (that-trace effect)

3. LEARNING VS. ACQUISITION

It is really hard to come up with a good answer to this question, which is

Please do not copy or distribute this answer key to your students

CHALLENGE PROBLEM SETS

Part 1: The antecedent must precede the anaphor.

Part 1: Only yourself (and yourselves) are fully grammatical.

3. IS LANGUAGE REALLY INFINITE?

Q 1: Premise 2 is contradicted by this assumption; if the set is finite,

Q 2: Assuming that E is infinite begs the question: we start out assuming

Q 3: This is a difficult question. Personally, I do not agree with Pullum

Q 4: Rough proof using subsets

Please do not copy or distribute this answer key to your students

can be applied recursively to any other sentence in S. S is an infinite set,

Rough proof using contradiction.

4. ARE INFINITE SYSTEMS REALLY UNLEARNABLE?

There is an important typo in part 3 of this question, replace "infinite

This question requires a fair amount of creativity on the part of the

5. LEARNING PARAMETERS: PRO DROP

Q 2: The default setting must be non-Pro-drop.

Q 3: This is a difficult question; on the face of it, this is exactly the

Please do not copy or distribute this answer key to your students

1. TOPICS COVERED IN THIS C HAPTER

2. IF YOU SKIP THIS C HAPTER

3. IDIOSYNCRASIES OF THIS C HAPTER

Please do not copy or distribute this answer key to your students.

GENERAL PROBLEM SETS

a) The old rusty pot-belly stove has been replaced.

b) The red-haired assistant put the vital documents

through the new efficient shredder.

c) The large evil leathery alligator complained to

his aging keeper about his extremely

d) I’ve just eaten the last piece of chocolate cake.

Please do not copy or distribute this answer key to your students

3. GENDER NEUTRAL PRONOUNS

Pronouns are a closed class. Evidently, co-opting a featurally mismatched

The propriety of introducing the university slang will be

readily admitted; it is not less curious than that of the Old

Bailey, and is less generally understood. When the number and

accuracy of our additions are compared with the price of the

encouragement that is due to learning, modesty, and virtue.

You will be surprised at the variety of responses Students will give. A

wabe N; after the

Please do not copy or distribute this answer key to your students

the mole reads outside."]

Noun Plural Count Proper Pronoun Anaphor