The Syntax of Chinese
The past quarter of a century has seen a surge in Chinese syntactic research that
has produced a sizeable literature on the analysis of almost every construction
in Mandarin Chinese. This guide to Chinese syntax analyzes the majority of
constructions in Chinese that have featured in theoretical linguistics in the past
twenty-five years, using the authors’ own analyses as well as existing or potential
alternative treatments. A broad variety of topics are covered, including categories,
argument structure, passives, and anaphora. The discussion of each topic sums up
the key research results and provides new points of departure for further research.
This book will be invaluable both to students wanting to know more about the
grammar of Chinese, and to graduate students and theoretical linguists interested
in the universal principles that underlie human languages.
james huang is a Professor in the Department of Linguistics at Harvard
University.
audrey li is a Professor in the Department of Linguistics and Department of
East Asian Languages and Cultures at the University of Southern California.
yafei li is a Professor in the Department of Linguistics at the University of
Wisconsin–Madison.
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Frontmatter
More information
cambridge syntax guides
General editors:
P. Austin, B. Comrie, J. Bresnan, D. Lightfoot, I. Roberts, N. V. Smith
Responding to the increasing interest in comparative syntax, the goal of the
Cambridge Syntax Guides is to make available to all linguists major findings,
both descriptive and theoretical, which have emerged from the study of particular

languages. The series is not committed to working in any particular framework,
but rather seeks to make language-specific research available to theoreticians and
practitioners of all persuasions. Written by leading figures in the field, these guides
will each include an overview of the grammatical structures of the language con-
cerned. For the descriptivist, the books will provide an accessible introduction to
the methods and results of the theoretical literature; for the theoretician, they will
show how constructions that have achieved theoretical notoriety fit into the struc-
ture of the language as a whole; for everyone, they will promote cross-theoretical
and cross-linguistic comparison with respect to a well-defined body of data.
Other books available in this series
O. Fischer et al.: The Syntax of Early English
K. Zagona: The Syntax of Spanish
K. Kiss: The Syntax of Hungarian
S. Mchombo: The Syntax of Chichewa
H. Thrainsson: The Syntax of Icelandic
P. Rowlett: The Syntax of French
R. D. Borsley et al.: The Syntax of Welsh
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Frontmatter
More information
The Syntax of Chinese
C T. JAMES HUANG
Harvard University
Y H. AUDREY LI
University of Southern California
YAFEI LI
University of Wisconsin–Madison

© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Frontmatter
More information
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, S˜ao Paulo, Delhi
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521599580
C
Cambridge University Press 2008
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 2008
Printed in the United Kingdom at the University Press, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication Data
Huang, Cheng-Teh James.
The syntax of Chinese / C T. James Huang, Y H. Audrey Li, Yafei Li.
p. cm. – (Cambridge syntax guides)
Includes bibliographical references and index.
ISBN 978-0-521-59058-7 – ISBN 978-0-521-59958-0 (pbk.)
1. Chinese language – Syntax. I. Li, Yen-hui Audrey, 1954– II. Li, Yafei. III. Title.
IV. Series.

PL1241.H855 2008
495.1 – dc22 2008025651
ISBN 978-0-521-59058-7 hardback
ISBN 978-0-521-59958-0 paperback
Cambridge University Press has no responsibility for the persistence or
accuracy of URLs for external or third-party internet websites referred to
in this publication, and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Frontmatter
More information

Table of Contents



Introduction

Chapter 1 - Categories
1.1 Lexical categories
1.1.1 Verbs and nouns - basic distinctions
1.1.2 Localizers
1.1.3 Adjectives
1.1.4 Prepositions
1.2 Functional categories
1.2.1 [Fn], n >
0

1.2.2 [F] and the modifier-introducing de

Chapter 2 - Argument structure
2.1 Arguments and theta-roles
2.1.1 Basic properties of theta-roles
2.1.2 Chinese resultative compounds: a case study
2.1.3 Compounds vs. phrases
2.2 On the nature of theta-roles
2.2.1 Theta-roles produced by the syntax
2.2.2 What’s in a verb?
2.2.3 Squeezing a lexical foot into a functional shoe
2.3 Sketching an alternative theory of theta-roles
2.3.1 How a lexical entry contributes to the argument structure
2.3.2 The theory
2.3.3 Facts explained
2.4 In place of a conclusion

Chapter 3 - The verb phrase
3.1 Adjuncts and complements
3.2 Postverbal constituents
3.2.1 Double objects and the structure of VP
3.2.2 V-de
3.2.3 Frequency/Duration phrases (FP/DrP)
3.3 Preverbal constituents
3.3.1 Aspectual phrase
3.3.2 Modals
3.4 Summary

Chapter 4 - Passives
4.1 The Mandarin long passive

4.1.1 Two competing traditions
4.1.2 The analysis: A’-movement and predication


4.1.3 Further evidence for the NOP analysis
4.2 The Mandarin short passive
4.2.1 Against the agent-deletion hypothesis
4.2.2 Analysis of the short passive
4.3 The analysis of indirect passives
4.3.1 Direct vs. indirect passives
4.3.2 The inclusive indirect passive
4.3.3 The adversative passive
4.4 Summary

Chapter 5 - The ba construction
5.1 Ba and bei constructions
5.2 What is ba?
5.2.1 The categorial status of ba
5.2.2 The analysis of ba
5.3 Ba not a theta-role assigner
5.3.1 Ba and the subject
5.3.2 Ba and the post-ba NP
5.4 Structures
5.4.1 A preliminary analysis
5.4.2 Revision
5.5 “Affected”
5.6 Alternatives
5.7 Summary

Chapter 6 - Topic and relative constructions

6.1 Topic structures
6.1.1 Movement or not?
6.1.2 Island conditions
6.2 Relative structures
6.2.1 Distribution and interpretation
6.2.2 Movement
6.2.3 Base generation
6.2.4 Relative operator
6.2.5 NP adjunction
6.3 Gapless structures

Chapter 7 - Questions
7.1 Yes-no questions
7.2 Disjunctive questions
7.3 A-not-A questions
7.3.1 Three types of A-not-A questions
7.3.2 A-not-A questions: a modular approach
7.3.3 Explaining the differences
7.3.4 VP-neg questions
7.3.5 Summary


7.4 Wh-questions
7.4.1 A movement approach to wh-in-situ
7.4.2 LF movement: some problems and alternatives
7.4.3 LF subjacency and pied-piping
7.4.4 Non-movement and unselective binding
7.5 Summary

Chapter 8 - Nominal expressions

8.1 The issues
8.2 Projecting a DP - referential and quantity expressions
8.2.1 Number expressions as indefinite and quantity expressions
8.2.2 Quantity vs. indefiniteness
8.2.3 Number Phrase and Determiner Phrase
8.2.4 Comparison with indefinite wh-elements
8.2.5 Comparison with you expressions
8.2.6 Prohibition against an indefinite subject/topic
8.2.7 Summary
8.3 Order and constituency within a DP
8.3.1 Demonstratives
8.3.2 Pronouns
8.3.3 Proper names
8.3.4 Common nouns
8.3.5 Not appositives or adverbials
8.3.6 Summary
8.4 Extension and revision: plurality
8.4.1 Some puzzles about -men
8.4.2 Plural feature as head of NumP
8.4.3 Proper name + pronoun + demonstrative
8.5 Summary and some empirical complications
8.5.1 Non-quantity indefinite nominals in subject position
8.5.2 Non-root clauses, generic NPs

Chapter 9 - Anaphora
9.1 Binding theory in Chinese
9.1.1 Reflexives and Principle A
9.1.2 Pronouns and Principle B
9.1.3 Principles C and D
9.2 The bare reflexive ziji

9.2.1 Two approaches to the long distance ziji
9.2.2 Logophoricity and anaphoricity
9.2.3 Logophoricity: syntax and semantics
9.3 Bound anaphora and donkey anaphora
9.3.1 Pronouns in co-reference or as bound variables
9.3.2 Variable binding: scope, accessibility, and disjointness
9.3.3 Indefinites and donkey anaphora
9.4 Summary and conclusion

Abbreviations
A, AP adjective, adjectival phrase
AC Adjunct Condition
ACC accusative case
ACD Antecedent Contained Deletion
ADV Adverb
AE Anaphoric Ellipsis
ART Article
Asp, AspP aspect, aspectual phrase
BA marker of the ba construction (see Chapter 5)
BEI passive marker bei (see Chapter 4)
BPA Binding Principle A
C, CP complementizer, complementizer phrase
CED Condition on Extraction Domain
CFC complete functional complex
CL classifier
CNPC Complex NP Constraint
CR Conjunction Reduction
D, DP determiner, determiner phrase
DAT dative case
DC Directionality Constraint

DE pre-nominal modification marker or postverbal resultative
marker de
DECL declarative
DEM demonstrative
DrP Duration Phrase
DRT Discourse Representation Theory
ECP Empty Category Principle
Fn functional element of degree n
FEC free empty category
FI Full Interpretation
FP Frequency Phrase
ix
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Frontmatter
More information
x Abbreviations
GB Government and Binding
GC Governing Category
GCR Generalized Control Rule
GUO experiential aspect marker guo
H head
HMC Head Movement Constraint
IHRC internally headed relative clause
I, IP inflection, inflectional phrase
L, LP localizer, localizer phrase
LBC Left Branch Condition
LD long-distance

LDR long-distance reflexive
LE perfective marker or sentence-final particle
LF Logical Form
LRS lexical relational structure
LSS lexico-semantic structure
Lv light verb
MDP Minimal Distance Principle
Mod Modifier
MP Minimalist Program
N, NP noun, noun phrase
NOM nominative case
NOP null operator
Num, NumP numeral, number phrase
OP operator
P, PP preposition/postposition, prep/postpositional phrase
PASS passive morpheme
PAST past tense
P&P Principles and Parameters
PF Phonetic Form
PL plural
PLA Principle of Lexical Association
PLI Principle of Lexical Integrity
POV Point-of-View Phrase
PRES present tense
PRO/pro empty pronominal element
PROG progressive
QNP quantificational NP
Q question particle
QR Quantifier Raising
© Cambridge University Press www.cambridge.org

Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Frontmatter
More information
Abbreviations xi
Qu question operator
QVE quantificational variability effect
RNR right-node raising
SC Subject Condition, also for Superiority Condition
SFP sentence-final particle
SourceP Source Phrase
Spec specifier
SUO pronominal element suo marking object relativization or
passivization
t trace of moved element
T, TP tense, tense phrase
TOP topic
UG Universal Grammar
UTAH Uniformity of Theta Assignment Hypothesis
V, V P verb, verb phrase
X
0
syntactic head of type X
XP full syntactic phrase of type X
X’ intermediate syntactic phrase of type X
ZHE durative aspect marker zhe
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese

C T. James Huang, Y H. Audrey Li and Yafei Li
Frontmatter
More information
Introduction
Over the last quarter-century, there has been a surge of research on Chi-
nese syntax. A cursory look at the programs of Chinese linguistics conferences
held since 1985 shows that at least a full continuous session has been devoted
to Chinese syntax throughout each day of every conference. Those who were
involved in organizing such conferences can also recall the large number of syntax
abstracts, routinely accounting for fifty to sixty percent of all abstracts received
for review. It is also during this past quarter-century that a significant number of
theoretically oriented works on Chinese syntax began to appear in major refereed
academic journals published in the West. Several monographic, theoretical treat-
ments of Chinese syntax have also appeared that distinguished themselves from
earlier general descriptions or reference grammars. In the field of theoretical lin-
guistics, more works than before make crucial reference to Chinese syntax. It is
clear that research on Chinese syntax that is informed by modern linguistic theories
has been productive. In turn, it is also clear that the study of Chinese syntax has
played an ever-increasing role in linguists’ construction of modern “mainstream”
syntactic theories.
Most of these “modern syntactic theories” are in one form or another theo-
ries falling under the formal paradigm of generative grammar. Of these formal
treatments, much research has been carried out in the Principles-and-Parameters
(P&P) approach initiated by Chomsky and his colleagues and students around
1980, plus and minus two or three years, in its various incarnations including
the so-called Government-and-Binding (GB) framework, the Barriers framework,
and recent attempts at theoretical economy aimed at the ideals of the Minimalist
Program (MP). The P&P approach marked the beginning of an era that distin-
guished itself from the first quarter-century of generative grammar (since 1957) in
enabling the construction of a theory of grammar that is at once general enough

to capture common properties of human language and flexible enough to account
for language variations. It provided a way to make good sense of the innateness
hypothesis (or “biolinguistic approach”) that characterized Chomsky’s approach
since it was introduced twenty-five years before, a hypothesis that takes the inter-
nalized grammar of any language to be a combined product of nature and nurture.
1
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
2 The Syntax of Chinese
It also allowed for the productive description of languages of various typological
types and, most importantly, for the study of a variety of languages to directly
contribute to the construction of general linguistic theory. The construction of the
GB theory as we know it today, for example, was itself in part informed by some
analyses of Chinese syntax.
The volume of research products that have appeared in this period, quite unlike
the situation ever before, far exceeds the amount anyone can easily recall or enu-
merate when pondering over one particular topic or another. Various grammatical
constructions have been given multiple different treatments. Some grammatical
constructions that seemed irrelevant to generative grammar in its early periods are
now actively analyzed, while objections to certain formal analyses have now lost
ground. Yet we continue to hear objections and questions from scholars unfamiliar
with the paradigm – either those who were educated in the pre-GB model with
many assumptions that are no longer held by current generative researchers or
those who are less informed about formal approaches. Part of this situation, we
believe, arises from misunderstanding or lack of accessible information. The fact
is that, for almost every topic of Chinese syntax, there now exists a sizable amount

of generative literature within the P&P paradigm. The problem, for those who
for one reason or another have not been able to follow the recent developments,
comes in part from the fact that most research products come in single articles –
from journals, edited volumes, and conference presentations – and there is no
work as yet that attempts to take stock of the major results that have been pro-
duced and describe them in some depth – within one volume – that might serve
the double purpose of informing the readers less familiar with (or less committed
to) formal linguistics and the current status (in our view) of formal Chinese syn-
tax, and of bringing further questions onto the research agenda for other scholars
and students interested in the enterprise of providing rigorous analyses of Chinese
linguistic facts and bringing them to bear on the construction of an optimal theory
of human linguistic competence and its possible variations, as part of a theory of
the “mirror of the mind.”
The desire to take a first step toward filling this gap was a major motivation
that led us to take up the project of writing this book. It is our hope that a volume
consisting of the topics we have chosen will present a more comprehensive outlook
of the syntactic system of the whole language to the reader, and that our discussion
of the various analyses on each topic will help both to sum up some of the important
results and to provide new points of departure for further research. It is also our
intention to use this book to demonstrate, for each topic selected, how a formal
generative analysis may help make sense of certain observed properties of the
language, perhaps in ways better than other imagined approaches, and how it may
be seen as a contribution to linguistic theory.
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
Introduction 3

Before we go on to present the details of what this book is about, however, we
must make clear what it is not. First, it is not meant to be a reference grammar for
the learner of Chinese, though it might be seen as a (somewhat biased) reference
on the formal linguistic analyses of Mandarin Chinese syntax. Although we have
tried to include as many references as we can to the large volume of works avail-
able, we are sure to have inadvertently missed some. Even where references are
included, we do not provide a detailed discussion of all alternative analyses that
are worthy of consideration, other than those closely related to our own analyses.
We have also excluded most references that are explicitly non-formal. Second, it
is not a comprehensive treatment of Chinese syntax. As it turns out, even within
formal approaches, it is impossible to touch on all the important aspects of Chinese
syntax. Rather than briefly summarizing results on a comprehensive list of top-
ics, we have chosen to provide fairly detailed analyses and argumentation of a
selected number of topics, excluding some owing to space limitation and others
where we have nothing new to offer. For each topic our discussion is driven by the
goal of providing one or two specific analyses and explaining the rationale behind
them, with the general theory of grammar in mind. It is often said that Y R.
Chao’s (1968) Grammar of Spoken Chinese is a comprehensive single-volume
masterpiece that represents the best of the American descriptive and structural-
ist tradition. No single-volume formal treatment comparable to Chao’s in scope
has appeared in the last several decades. The rich observations and insights con-
tained in that volume remain unsurpassed to this day. We have not attempted a
comprehensive treatment of Chinese syntax in the generative tradition. Our goals
are both different and limited: the book presents grammatical analyses that cover
most of the constructions of (Mandarin) Chinese that have figured in the field of
theoretical linguistics in the past twenty-five years, focusing on our own analyses
in most cases. It is intended to show how the facts of Chinese may be profitably
understood with the tools of generative linguistics, and in turn how the analyses
may help settle important issues and guide further research in linguistic theory.
It is intended as a contribution to Chinese syntax as a distinct subject of Chinese

studies, and also to generative grammar as a hypothesis about human linguistic
competence.
The rest of this book is organized into four parts comprising nine chapters.
Part I (Chapters 1–3) investigates the building blocks and “canonical” structures
of sentences, including the grammatically relevant properties of words and the
combinatorial algorithm by which phrases are formed. Chapter 1 presents a theory
of parts of speech, which we call categories. Lack of sufficient inflectional and
derivational clues has made the identification of categories difficult for Chinese.
Drawing on the insights gained from other languages, we rely primarily on the
syntactic behaviors of a word to determine its category. It is also shown that
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
4 The Syntax of Chinese
a category is best viewed as a cluster of plus- or minus-valued features, which
enables us not only to distinguish syntactically relevant categories but also to
explain why some of them display identical properties. The analysis of localizers,
one of the perpetually question-begging categories in Chinese, makes use of the
notion of computational cost and opens up a new possibility to understand how
categorial changes happen and a new category comes into existence.
Chapter 2 focuses on the nature of argument structure. Capitalizing on a long-
known and puzzling fact, i.e., that the semantic relations between a verb and its
subject or object are much less restricted in Chinese than in English, and drawing
on recent works by others, we propose a theory of lexico-semantic decomposition
of verbs that minimizes the amount of stipulated mechanisms and components and
thereby maximizes the explanatory power of the theory. In particular, it is argued
that a tiny set of event-typing elements interact with a lexical root to produce the

more “rigid” argument structures found with English verbs, whereas the option
of using bare roots as verbs in the absence of event-typers, aided with world
knowledge, is responsible for the degree of semantic freedom in Chinese.
Chapter 3 covers a broad range of topics on the “canonical” structures of the
sentence, with particular focus on the verb phrase and its components. It examines
the systematic distinctions between adjuncts and complements, looks for the best
structural representations of five different postverbal constituents (the double-
object, two V-de’s, and the frequency and duration expressions), and discusses
how such semantic notions as aspect and modality are handled in the syntax of
Chinese. In the course of presentation, it is proposed that the behavioral disparity
between the resultative V-de and its manner counterpart may be attributed to the
superficially unrelated fact that Chinese has resultative compounds but not ones
with a postverbal manner modifier. Attention is also given to constructions which
appear to display syntax–semantics mismatches. What unifies this large collection
of miscellaneous topics is a single phrase structure pattern whose restriction on
possible syntactic analyses highlights an important characteristic of this model
of linguistic theory: using the least amount of independently motivated tools to
account for the maximal amount of data.
In Part II (Chapters 4 and 5), we take a closer look at argument structure and its
relation with lexical semantics and its effects on syntactic structure, by focusing
on two constructions that have been in the center of debate from the inception of
Modern Chinese syntax as a field. Chapter 4 deals with the passive bei construction,
which takes two forms depending on the presence or absence of an Agent phrase
(the long and short passive respectively). After exhibiting the pros and cons of a
movement-based approach and one based on complementation, it is argued that the
Chinese passive involves both movement and complementation. The long passive
is derived via clausal complementation where the embedded object is brought to
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese

C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
Introduction 5
the periphery of the complement clause (a process of “operator movement”) and
is predicated on the main-clause subject. The short passive, on the other hand,
involves verbal complementation and the object is brought to the periphery of a
verb phrase (a process of “argument movement”) where it is interpreted with the
subject.
Building on the findings of the passive construction, Chapter 5 compares and
contrasts passives and the closely related ba construction. Bei and ba construc-
tions are similar in argument structures. However, they differ in the range of
(un)acceptable cases, which is attributed to the different subcategorization require-
ments of ba and bei, reflected in the syntactic structures with which they are asso-
ciated. Nonetheless, the extant literature on the ba construction has not been as
focused on its syntactic properties as on the special meaning of this construction
and how to account for it. The ba construction has been noted as expressing “dis-
posal” or “affectedness.” We show that the special meaning cannot be due to any
thematic-assigning capabilities of ba. Every ba sentence has a non-ba counterpart,
which points to the irrelevance of ba in contributing to the argument structures. In
the most typical examples, ba seems to be related to the notion of boundedness
or requires a result expression. However, the complexities of the ba construc-
tion require the search for further possibilities and additional mechanisms for an
“affected” interpretation.
While the passive and the ba construction exemplify how modifications in lex-
ical structure affect the syntactic relations between arguments such as subject
and object, other constructions exhibit syntactic properties independent of lexical
semantics. Such constructions involve operations on or beyond clauses, and often
concern the logical relations between clausal peripheral elements and the clauses
as a whole. Part III takes up two types of logical structure: one involving (often)

overt antecedent–gap relations and the other involving, as we shall argue, covert
dependency relations. The first type, dealt with in Chapter 6, is best illustrated by
topic and relative clause structures, in which a clause is used to modify a head
noun phrase. There have been claims that a relative construction is derived from
a topic structure; however, we show that the two constructions are similar but
not identical. They are alike in the set of locality conditions restricting the well-
formedness of these constructions, phrased in terms of constraints on movement
and rules governing the distribution of empty categories. They differ in exactly
which element undergoes movement and where it lands. Variations with respect
to these factors are also manifested within relative constructions in a cluster of
empirical generalizations that can be traced to the absence/presence of a relative
operator.
In Chapter 7, we turn to the syntax of interrogative sentences with particular
attention to wh-questions and a special type of disjunctive question called the
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
6 The Syntax of Chinese
“A-not-A question.” After clarifying the distinctness of this question type from
normal yes-no questions, we propose and defend a modular approach to the A-
not-A questions. We discuss a number of approaches to the syntax and semantics of
wh-questions, which exhibit covert long-distance dependencies with restrictions
that, we argue, follow from an appropriately formulated theory of movement,
binding, and the syntax–semantics interface.
Our presentation of the syntax of Chinese would be inappropriately incomplete
without some in-depth discussion of the syntax of nominal expressions and their
meanings. Part IV of this book is devoted to this area. Chapter 8 takes up the

syntactic structure of nominal expressions – nouns and phrases built around nouns
as their heads. We note that Chinese noun phrases, on their surface, are at once
more complex and more simple than their counterparts in, say, English and other
languages (for example, with respect to the requirement of numeral-classifiers,
the lack of true determiners, and the occurrence of “bare” singular count nouns).
We argue, however, that appearance notwithstanding, Chinese noun phrases (like
those in many other languages) have more structure than meets the eye. We
propose a full determiner phrase that may contain other smaller phrases headed
by a numeral expression, a classifier, and a noun, and show that this allows for
the derivation and explanation of certain facts of (in)definiteness, specificity, and
compositional semantics.
Another important aspect of the semantics of noun phrases concerns their
reference and the referential dependencies they exhibit on each other. This is the
subject of the final chapter. Here our discussion addresses both the syntax and
semantics of coreference and of variable binding. We show that the referential
properties of nominal expressions are tied to their intrinsic properties (whether
they need an antecedent or not), the syntactic position of their antecedents (if
they need one), and the nature of the antecedents themselves (whether they are
referential or quantificational). With respect to definite noun phrase anaphora,
we devote substantial space to a discussion of the Chinese reflexive pronoun
ziji, and show that it is both an anaphor in the sense of classical Binding Theory
and a logophor within contexts of “attitudes de se” that describe the speech, the
mental state, or the perspective of an appropriate protagonist. With respect to
variable binding, we show that the crucial requirement is c-command in a proper
Logical Form representation. We finish Chapter 9 with a discussion of so-called
“donkey anaphora,” something that has a status between definite coreference
and variable binding. We present two types of “donkey sentences,” each with a
set of distinguishing properties, and show that a proper analysis of them helps
settle an important debate between two competing theories that have figured
prominently in recent treatments of indefinite noun phrases and their referential

properties.
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
Introduction 7
There are clearly other interesting topics of Chinese syntax that deserve
coverage in a book with this title, but we have had to leave them out. Several other
constructions that bear on lexical structure and syntactic projection could each
deserve a chapter-length full treatment. For example, the resultative construction
(both the compound and the phrasal versions), touched upon briefly in Chapter
3, has further interesting properties bearing on the structure of events and their
projection in syntax. The syntax of adverbials and that of aspectual markers are
two other areas that have received considerable renewed interest in recent years.
Other topics falling under the area of argument structure and syntactic structure
include the syntax of unaccusatives, the two types of double-object constructions,
and the proper syntactic treatment of various conjunctives. With respect to logical
structure and the syntax–semantics interface, we have left out much work on
quantification and structures bearing on focus and presupposition. And our
discussion of noun phrase anaphora also does not touch upon the distribution and
reference of zero pronouns, a topic of major interest to parametric theory with
implications for the interface between syntax and discourse. In selecting topics
for inclusion in this work, we have used three criteria. The first is our perception
of relative priority in trying to strike a balance between breadth and depth within
a limited space. The second is the availability of the literature: a topic is not
included when it has been extensively discussed in easily accessible monographs
or journals. The third one has to do with the scope of our own research: we have
left out topics on which we have not ourselves carried out sufficient research and

to which we do not have something new to contribute.
A word about the intended readers of this work: we prepared these chapters
originally for university courses that we offer on the linguistic structure of
Chinese, so the most immediate intended readers of this book are those graduate
students and upper-level undergraduates who have some basic knowledge of
linguistic structure. Such students, or any professional linguist of any theoretical
persuasion, should find the book fully accessible, even without prior experience
with the Chinese language. A student of the Chinese language may also find this
work accessible with occasional reference to linguistic terminology available
from syntax textbooks or linguistics glossaries. In writing the book, we have
also had in mind the non-specialists who are curious about Chinese grammar
and generative syntax, and have tried to briefly explain technical notions as they
are first introduced. As such, we hope the book will be useful to teachers and
researchers in such Chinese-related fields as language teaching, natural language
processing, machine translation, language acquisition, philosophy of language,
and other related areas of cognitive science.
As usual, the completion of a book of this size owes itself to the help of numerous
people. It is impossible to enumerate the scholars from whom we have learned
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
8 The Syntax of Chinese
the body of knowledge represented here. We should, however, mention a few
colleagues who have collaborated with us on one topic or another with results that
have been included in this work. In particular, the materials on donkey anaphora
and long-distance reflexives are derived from earlier work conducted with Lisa
Cheng and C S. Luther Liu, respectively. Our discussion of argument structure

and lexical relations has also benefited from our erstwhile collaboration with Lisa
Cheng and C C. Jane Tang. Some sections on relative constructions and wh-
questions are incorporated from work in collaboration with Joseph Aoun. The
analysis of the V-de constructions draws on our joint work with Jen Ting. And the
discussion of the interactions among different adverb classes is a direct application
of the discoveries we made together with Vivian Lin and Rebecca Shields on the
intervention effects of adverbs in English and Russian. We continue to appreciate
the opportunities we have had to work with them. Parts of this manuscript in
one of its earlier versions have been tried out in classes and read by some of the
students and faculty at Harvard University, the University of Southern California,
and the University of Wisconsin–Madison, as well as the National Taiwan Normal
University, Stanford University, and the University of Venice. We are gratified by
the interest and support shown to us by the instructors and participants and, in some
cases, for their comments and suggestions – especially those of Ressy Ai, Shengli
Feng, Francesca del Gobbo, Miaoling Hsieh, So-One Hwang, Soo-Yeon Jeong,
Julie Jiang, Daphne Liao, Jing Rong, Peter Sells, Yang Shen, Yuan Shen, Fuzhen Si,
Jen Ting, and Yaqing Wu. In our final efforts to bring this work to fruition, we owe
special gratitude to Bridget Samuels for her help in making the whole manuscript
more readable than it otherwise could be. Finally, but not the least, our deep-felt
thanks go to Emily, Qing, and Yu-Chin for all the best of things that life can offer;
something that we have taken all these years but, probably too often, for granted.
JH, AL,&YL
© Cambridge University Press www.cambridge.org
Cambridge University Press
978-0-521-59958-0 - The Syntax of Chinese
C T. James Huang, Y H. Audrey Li and Yafei Li
Excerpt
More information
Chapter 1 Categories


We take it as our starting point that a Chinese sentence is composed of words and that
words have different behaviors in a sentence. For instance, while dayan fei ‘wild.goose fly’
is an acceptable sentence, *fei dayan ‘fly wild.goose’ is not. The most obvious reason for
the contrast is that dayan ‘wild.goose’ is a noun that canonically serves as the subject of the
sentence and fei ‘fly’ is a verb whose canonical function is to be the predicate occurring
after the subject. This means that in order to understand the syntax of Chinese, or the
syntax of any language for that matter, we minimally need to understand how the words in
a language are classified and how these different classes of words are put together to form
sentences. In this book, word classes are referred to as lexical categories, or just categories
for short, following the terminological convention of generative syntax.
While the basic distinction between nouns and verbs is universally recognized in
modern literature on Chinese syntax, scholars differ, sometimes drastically, on other
categories. See Chao (1968), Li and Thompson (1981), Zhu (1982), and Xing and Ma
(1992) for a few examples. The differences in opinion arise partly because linguists with
different theoretical backgrounds may employ different criteria for word classification, and
partly because we still lack sufficient knowledge about certain words and their properties.
Regardless, it is without question that the ultimate task for anyone studying lexical
categories in Chinese is to identify them in such a way that they both allow an accurate
description of the syntactic behaviors of the language, and provide insights into the nature
of word classification.

1
With this goal in mind, we will introduce a theory of lexical categorization in
Mandarin Chinese in this chapter. The theory consists of two intertwined parts. First, a set
of categories is confirmed and examined on the basis of the syntactic behaviors of Chinese
words and morphemes. Second, a decompositional theory that characterizes the intrinsic
relations among these categories is defended. It is important to mention up front, however,
that we do not intend to spread our discussions evenly among all issues related to lexical
categorization, nor do we attempt to provide an exhaustive list of categories in the language.
Rather, the chapter concerns itself primarily with where we believe new insights are

available from recent research. The same approach also applies to the organization of the
whole book.

1.1. Lexical categories

This section focuses on verbs (V), nouns (N), prepositions (P), and adjectives (A).

1.1.1. Verbs and Nouns – Basic distinctions

It is common wisdom in modern linguistics that N and V are two basic categories. In
Chinese, the two categories can be clearly distinguished on the basis of their modifiability
by the negative morpheme bu. The basic data is given in (1)-(2):

(1) Verbs
a. bu shui ‘not sleep’

2
b. bu tongzhi ‘not inform’
c. bu sai-qiu ‘not play ball’

(2) Nouns
a. *bu shu ‘not tree’
b. *bu xiaoxi ‘not news’
c. *bu qiu-sai ‘not ball game’

To our knowledge, all verbs can be negated by bu, and no noun can. It must be pointed out
that bu can also negate adjectives such as da ‘big’ and lei ‘tired’. As we will see in
subsequent sections, this similarity between verbs and adjectives poses no problem for the
N-V distinction.
Examples exist in modern Chinese that seem to suggest that nouns can be modified

by bu, such as bu-ren-bu-gui ‘not-human-not-ghost’. However, there are reasons for not
regarding such examples as a problem for the bu-test of the noun/verb distinction. First,
they are not formed with a productive process. A change of nouns typically results in
unacceptability:

(3) a. *bu-shu-bu-bao ‘not-book-not-newspaper’
b. *bu-fan-bu-cha ‘not-food-not-tea’


3
Second, the nouns in these examples must be monosyllabic, even when multi-syllabic
counterparts exist, further confirming that bu cannot really modify a noun in modern
Chinese:

(4) a. *bu-huoren-bu-sigui ‘not-live.human-not-dead.ghost’
b. *bu-renlei-bu-guilei ‘not-humankind-not-ghost.kind’

Lastly, even with the nouns that bu can accompany, a single bu-N pair is not permitted,
contrasting sharply with verbs in (1):

(5) a. *bu-ren ‘not-human’
b. *bu-gui ‘not-ghost’

As a result, we regard the few exceptions not as undermining the reliability of the bu-test,
but as idiomatic expressions not subject to the general rules we are pursing.
N and V also differ in many other ways reported in various grammar books (e.g., a
subset of V allows aspectual suffixation, while no word used as N does). For the present
chapter, the data below are of particular interest:

(6) Verbs

a. meiti baodao-le na-ci shigu.
media report-LE that-CL accident
‘The media reported that accident.’

4
b. Zhangsan fanyi-le yi-bu xiaoshuo.
Zhangsan translate-LE one-CL novel
‘Zhangsan translated a novel.’
c. laoshi piping-le zhe ji-ge yanjiusheng.
teacher criticize-
LE these some-CL graduate.student
‘The teacher criticized these graduate students.’

(7) Nouns
a. meiti *(dui) na-ci shigu de baodao
1

media on that-CL accident DE report
‘the media’s report of that accident’
b. Zhangsan *(dui) yi-bu xiaoshuo de fanyi
Zhangsan on one-CL novel DE translation
‘Zhangsan’s translation of a novel’
c. laoshi *(dui) zhe ji-ge yanjiusheng de piping
teacher on these some-
CL graduate.student DE criticism
‘the teacher’s criticism of these graduate students’


1
Parentheses are another notational convention. The expression between a pair of parentheses is optional.

E.g., A(B)C indicates that both AC and ABC are acceptable facts. If an asterisk “*” immediately precedes the
expression inside the parentheses, as in A(*B)C, then AC is acceptable but ABC is not. If instead the asterisk
immediately precedes the left parenthesis, as in A*(B)C, then ABC is acceptable but AC is not. All the
examples in (7) are of this type.


5
The two groups of examples, though both based on baodao, fanyi and fanxiu, exhibit three
differences. Take (6a) and (7a) for example. First, the semantic object occurs to the right of
baodao in (6a) but to the left in (7a); second, a preposition dui is required to introduce the
object only in (6a); third, the morpheme de is required before baodao in (7a). The nature of
these facts will become clearer as we proceed. For now, it is sufficient to note that nouns
depend on prepositions like dui for the grammaticality of their object whereas verbs do not.
This is a very reliable test to separate N from V, with the limitation that it only applies
where the semantic subject of the N/V is present.
2

The fundamental distinction between N and V might be a reflection of
proto-categories,
3
a concept that traces its origin to psychological studies of human
cognition. It is possible that our brain divides the world into two elementary kinds of
entities: things that exist and situations that take place. Proto-N is the linguistic
representation of the former kind and proto-V, that of the latter kind. All specific lexical
categories are then the derivatives of these two proto-categories. Let us represent the
proto-categories as two features, [N] and [V]. Since a word either belongs to proto-N or
does not belong to proto-N, the feature for this proto-category has two values, [+
N]. The
same logic leads to [+
V]. These two binary-valued features yield four possible

combinations: [+N, -V], [+N, +V], [-N, -V], and [-N, +V]. If these feature combinations
indeed correspond to lexical categories in languages, then it is obvious that nouns are [+N,
-V] and verbs are [-N, +V]. That is, a noun conforms to proto-N but not to proto-V,


2
This analysis is adopted from Y. Li (1997a). See Fu (1994) for a different treatment of the data.

3
Cf. Givon (1984) and the references cited there.


6