| AUTHOR: Carnie, Andrew
TITLE: Constituent Structure
SERIES: Oxford Surveys in Syntax and Morphology
PUBLISHER: Oxford University Press
Fredrik Heinat, Department of Swedish, University of Gothenburg
In his book Carnie reviews past and current approaches to phrase structure from
a wide range of (mostly generative) theoretical frameworks. The stated purpose
of the book, and the others in the series, is to provide the reader with
''accessible, critical and up-to-date'' information about central topics in syntax
and morphology (general preface). The book is not meant to be introductory and
the intended reader is someone who has taken at least an introductory course in
syntax. The book contains eleven chapters, divided into three parts:
preliminaries, phrase structure grammars and X-bar Theory, and controversies.
There are also an index and a list of abbreviations and symbols used.
The first chapter briefly describes the topic of the book and its intended
audience. The topic is constituent structure, which is the combination of words
into phrases and clauses. Carnie also lists the various theories that will be
discussed. They are: all Chomskyan grammars (from syntactic structures to
minimalism), relational grammar, lexical-functional grammar, Tree adjoining
grammar, generalized phrase structure grammar, head driven phrase structure
grammar, role and reference grammar, simpler syntax, and to much a lesser extent
dependency grammar, word grammar, categorial grammar, functional grammar,
cognitive grammar and construction grammar.
In the second chapter, Constituent Structure, Carnie starts with pointing out
the problems of viewing constituent structure as simple linear concatenation.
Having established that the hierarchical structure must be taken into account,
he continues to discuss tree structures and their mathematical properties in
chapter three, Basic Properties of Trees. The third chapter is concluded by a
discussion about domination and precedence.
The fourth chapter deals with the second order relations c-command and
government. Carnie gives a brief history of how the syntactic relations, which
are particular to government and binding (GB) and minimalism (MP), developed. He
also describes attempts to reduce these relations to more primitive relations
such as 'unambiguous paths' or 'sisterhood'.
The fifth chapter introduces the second part of the book. The chapter deals with
phrase structure grammars (PSG) and discusses such issues as context free and
context sensitive grammars. It finishes with a discussion about tree structures
and the different meanings 'phrase structure grammar' can have in different
theoretical frameworks; it can be top-down rewrite rules (early generative
grammar), structure creating projection (GB and MP), or it can mean a set of
conditions that filter out tree structures ( Generalized Phrase Structure
Grammar (GPSG) and Lexical Functional Grammar (LFG)).
In chapter six Carnie discusses extensions of PSGs. He touches briefly on topics
such as structure changing and structure building transformations. He also
presents a more detailed view of alternative PSGs in GPSG and LFG than in the
The seventh chapter is dedicated to X-bar theory and its historical development.
Carnie makes use of various substitution tests as arguments for intermediate
structure. He also discusses the use (or non-use)of X-bar structure in relation
to functional projections in various generative frameworks.
Chapter eight introduces the last part of the book, Controversies. This, too, is
a chapter that deals exclusively with GB and MP. Carnie traces the development
from X-bar theory to 'bare phrase structure'. He also touches briefly on topics
such as adjuncts, Kayne's LCA (1994), three dimensional trees and top down
versus bottom up construction of structure.
In chapter nine Carnie discusses the relation between syntactic dependencies
(such as subject and object) and constituent structure, i.e tree structure. He
briefly touches on several approaches, including more semantically oriented
The tenth chapter presents alternatives to strict compositional phrase
structure. Among the different versions of phrase structure trees that Carnie
discusses are line crossing, multi domination and multiplanar structures.
The last chapter deals with functional categories. The focus is on the clausal
structure and its three parts. According to Carnie, the ''fact'' that the clause
consists of three parts (VP, IP and CP in Chomskyan phrase structure, but
obviously encoded differently in other frameworks) is one of the greatest
linguistic discoveries of the last century. The chapter also briefly deals with
the structure of the noun/determiner phrase. Again, the focus is on GB and
I think this is a very good book. Carnie writes in a clear and lucid style.
Still, my opinions about it are somewhat divided. The reason is that the book
does not really do what it sets out to do, but on the other hand it does what it
does very well.
So what are the intentions of the book? As stated in the general preface to the
series, a book in the series ''provides overviews of the major approaches to
subjects and questions at the centre of linguistic theory''(general preface).
Given this aim of the book and the impressive list of theories that Carnie sets
out to survey, I was full of expectations. Unfortunately the book deals with
frameworks other than Chomskyan ones extremely little. Admittedly, Carnie comes
from the generative side and it is difficult to present the views of other
A better approach to meet the aims of the series would perhaps have been to
identify a number of empirical issues related to constituent structure and let
experts in each framework present analyses of these. As it is now many arguments
are presented for various developments in GB and minimalism, but none of the
other frameworks have implemented these changes, and unfortunately Carnie never
presents the arguments against these developments (for example substituting CP
for S, or splitting the VP) and at many times the reader only gets the
references to the literature in other frameworks rather than the actual
arguments. Carnie apologizes for this slanted approach and expects that scholars
in other frameworks will be disappointed in the sparse space allotted their
analyses (p260). This may very well be the case, but what I, and probably other
generative/minimalist scholars with me, miss is a detailed account of the
workings and the argumentation of the other frameworks. What I particularly miss
is a discussion of how frameworks that assume that there is no constituent
structure (in the syntactic sense) would account for the data that Carnie
presents. He mentions Cognitive grammar in about half a page and connectionist
models are only mentioned when he says that they can model constituency by
reference to linear order without making reference to hierarchical structure
(p16). This seems indeed to be a controversy that overshadows all other disputes
discussed in the book, but Carnie doesn't mention the connectionist approach
again, not even with a reference to the literature where arguments against it
can be found (for example Pinker and Prince 1988).
And what does the book really do? In my opinion, the book is an excellent survey
of phrase structure in generative grammar. Carnie traces the arguments and the
data that have been used in the history of Chomskyan grammars in particular, but
also GPSG and HPSG are discussed to a large extent, LFG not so much. The
presentation is clear and easy to follow, and interesting at that. Meticulously
Carnie works his way through the arguments that lie behind the changes in phrase
structure theory that have taken place in the past, but he also discusses the
very latest (improvements?) such as 'bare phrase structure' and 'label free'
syntax. It is also in the discussions about the generative approaches that
Carnie allows himself to be critical (see for example his discussion on
AGR-phrases on pages 245-250) which I think is a good thing.
All in all, I think this book is very well written and interesting and it
definitely deserves a place on every syntactician's bookshelf (after it's been
read, of course).
Kayne, R. 1994. _The antisymmetry of syntax_. Cambridge, Mass: MIT Press
Pinker, S. and A. Prince. 1988. On language and connectionism. In Pinker and
Mehler (eds). _Connections and symbols_. Cambridge, Mass: MIT Press, 73-194.
ABOUT THE REVIEWER
Fredrik Heinat is currently working as postdoc on a project about light verbs.
The approach is generative in broad terms. His interests are, among other
things, argument projection, anaphoric dependencies and linguistic theory.