Chierchia, Gennaro, and Sally McConnell-Ginet (2000) Meaning and
Grammar: An Introduction to Semantics, 2nd ed. MIT Press, xv+573pp,
paperback ISBN 0-262-03269-4, $35.00 (1st ed., 1990).
Martin Sch�fer, Institute of Linguistics, University of Leipzig.
DESCRIPTION OF THE BOOK'S PURPOSE AND CONTENT
This is a textbook introducing formal semantics to students with little
or no knowledge of logic and some of linguistics. It was announced at
The book consists of nine chapters and one appendix.
Chapter one: The empirical domain of semantics, pp. 1-52. This chapter
sets the stage for the book: it will focus on truth conditional
semantics within a generative framework. Problems for this approach are
Chapter two: Denotation, Truth, and Meaning, pp. 53-112. Here, the
distinction between sense and reference is introduced. The importance
of full sentences for interpretation is pointed out. A fragment of
English F_1 is introduced and some problems discussed.
Chapter three: Quantification and Logical Form, pp. 113-193.
Quantification in English and the idea of quantifier raising is
discussed. A new fragment is introduced, and pronouns are dealt with.
Chapter four: Speaking, Meaning, and Doing, pp. 195-255. Covers
pragmatics classics Grice and Austin. The authors also introduce the
main idea behind dynamic semantics.
Chapter five: Intensionality, pp. 257-328. An elementary intensional
logic is introduced, along with a new fragment.
Chapter six: Contexts: Indexicality, Discourse, and Presupposition, pp.
329-389. Introduces an new stage in language interpretation, namely
the relativization to context, and discusses presuppositions in great
Chapter seven: Lambda abstraction, pp. 392-429. Introduces lambda
abstraction. In discussing its application for cases like
dis/conjunction, relative clauses and VP anaphora, its usefulness is
Chapter eight: Word meaning, pp. 431-500. Discusses decomposition and
alternative approaches. The need for more types (for adj. like
'former') and more primitives (like 'events') is discussed. Via the
discussion of theta-roles it ends in a discussion of fuzzy vs.
supervaluational logic in dealing with semantic imprecision.
Chapter nine: Generalized Quantifiers, pp. 501-528. Introduces
generalized quantifiers and shows their advantages. Shortest chapter.
Appendix: Set-Theoretic Notation and Concepts, pp. 529-540 This is a
basic but good introduction to the set-theoretic notation and concepts
used throughout the book.
Changes from the first edition: according to the authors, chapter
three, five, seven and eight were substantially rewritten. It is now
intended to be modular, meaning that after the third chapter,
everything necessary for the rest is introduced.
Since this is a second addition, and assuming that the first edition's
success already has established this book's overall quality, I will not
discuss the merits of the book but rather show what I still think to be
inadequate. Furthermore, I will look at it purely from the student's
point of view. (For the teacher's viewpoint, cf. (Carlson:1991) on the
first edition). From this vantage point, I see three main points that
diminish the value and the usefulness of this book. They are, in turn,
its coherence, its excessive content at certain points, and the layout.
The modular setup of the book is already commented upon by the authors.
Given that, I will not say much about the order of chapters relative to
each other. The only obvious outlier seems to be the choice to put the
chapter on generalized quantification last. This seems very much
uninspired, especially since its conclusion, which in effect ends the
book, seems hardly appropriate to serve as an conclusion to the whole
book. However, even within chapters, subsections do often not connect
up to each other. Often, all sense of direction is lost. A case in
point is the chapter on lambda abstraction. Here, after introducing the
lambda operator, the authors try to give evidence for the usefulness of
the new operator. In order to do so, they start discussing VP-
disjunction and conjunction. After what reads like a conclusion to this
topic, they start a new subsection titled 'More on TP'. This is quite
startling, as so far TP has played no role in this chapter (It is
discussed 150 pages earlier in the chapter on intensionality.). It
turns out, however, that the section is again on problems in dealing
with conjunction, this time involving tense (The objective still being
to show the usefulness of the lambda operator.). With very few changes
in organization and wording this would have been much easier to follow.
On the whole, one often wishes for a clearer sense of direction, to
avoid the many times where one looks up from what is in itself
stimulating reading to wonder why this section is placed here.
Another question of coherence arises when the authors introduce event
semantics in order to deal with adverbs (new in this edition). Although
it is perhaps useful to show students that there is such a thing as
event semantics even at this stage, it should at least be pointed out
that this approach would force major changes in everything that has
been said before on fragments and interpretation. Without this caveat,
the student is left in the dark as far as the integration of this
approach with what has been described before is concerned.
Oftentimes, the textbook suffers from the authors putting too much
material between its pages. Outstanding examples are the discussion of
presupposition in chapter six and that of the reasons for using
supervaluation instead of fuzzy logic in chapter eight. While these are
interesting topics (cf. Carlson's remarks), they clearly are far too
advanced for a textbook that purports to be a general introduction. The
same can be said for the general tendency to point the reader into the
direction of dynamic semantics. The discussion is always stimulating,
but seems to be misplaced here.
There is only one place where the discussion falls away from the
generally high standards, which is the representation of decomposition
and theories using decomposition, again in chapter eight. The authors
try to show that meaning postulates should be preferred over
decomposition. In order to support their view, they use, besides
theoretic arguments such as 'translucent' words, general considerations
of semantic competence, language acquisition and even experimental
evidence. The representation does not differ from that in the first
edition, except that now, at the end of the discussion (p. 454), they
point to the fact that all these phenomena are not necessarily
inconsistent with decompositional approaches. While this is OK as it
stands, one would have liked to see the whole psycholinguistic
discussion removed from the pages of this book. It seems to question
the whole approach of the book up to that point. In fact, in earlier
chapters, the authors explicitly note that they adapt the stance
propagated by Stanley Peters, that the theory is about what the mind
must compute, not how (p. 192, cf. p. 110). In my view, this is the
correct attitude to have. Selectively using psycholinguistic data
undermines this approach.
The layout of the new edition is probably what differs most from the
first one. It is most of the times irritating and sometimes downright
misleading. Thus, boxes to which the text refers are often placed on
the wrong page (cf. p. 125f.), so that reading involves lots of page-
turning. Another variant is where the boxes are on the same page, but a
paragraph after they are discussed (for no reason, it seems), without
any referral to them (cf. p. 138). In turn, the exact opposite happens
with the next two boxes (p.139f.) . The highlight as far as box
management is concerned is certainly the box on page 419f., which is
itself broken on two separate pages (similarly the box on p. 131f.),
with the numbered elements on the one page being referred to on the
other page. In addition, it seems that the authors are to hesitating in
repeating examples sentences, which would often make reading smoother.
Very misleading are cases like the one on page 87, where what is
positioned like a section heading turns out to be the last line of the
exercise on the opposite page. Even in simple tables, like the one on
p. 66, readability could be much improved.
This is, all in all, a very good textbook. However, it could be even
better if it had a clearer organization and reduced the amount of
material covered. Then, the view on those parts of the book, where the
authors try to show how formal semantics is not an abstract enterprise
but makes clear empiric predictions (Highlights are the discussion of
gerunds vs. infinitives and that of generalized quantifiers) would
become much clearer.
Carlson, Greg (1991). Review of 'Meaning and Grammar (first edition)'.
Language, 67, 4, 805-813.
ABOUT THE REVIEWER
Martin Sch�fer is a linguist and works half-time in a project entitled
'Semantics and World Knowledge' at the University of Leipzig. His
interests include semantics, the syntax-semantics interface, adverbs
and Chinese linguistics.