552LANGUAGE, VOLUME 70, NUMBER 3 (1994) from the preceding chapters. It reports on the methodology followed in the construction of WordNet, a lexical database of English made up of 54,000 entries organized into some 48,000 sets of synonyms. Working on the premise that 'the mental lexicon is organized by semantic relations' (201), WordNet attempts to model this organization. Miller & Fellbaum describe some of the relevant semantic relations that hold between lexical items and between classes oflexical items, and argue that correspondences between meaning and syntactic category are far from arbitrary. In sum, L&CS provides an excellent introduction to contemporary work in lexical semantics. For those already engaged in this area of linguistic research, it offers valuable new insights and suggests many directions for future research. It would make an ideal reference text for a cognitive science course with language as an important focus. Each article is very clearly written (and has been carefully proofread) and would be accessible to students with only a basic knowledge of linguistic metalanguage. The volume contains a language index, a name index, and a subject index which facilitate the reader's task. REFERENCES Bowerman, Melissa. 1990. Mapping thematic roles onto syntactic functions: Are children helped by innate "linking rules"? Linguistics 28.1253-89. Gordon, Peter. 1985. Evaluating the semantic categories hypothesis: The case of the count/mass distinction. Cognition 20.209-42. Hale, Kenneth L. 1986. Notes on world view and semantic categories: Some Warlpiri examples. Features and projections, ed. by Pieter Muysken and Henk van Riemsdijk, 233-54. Dordrecht: Foris. Hopper, Paul J. and Sandra A. Thompson. 1980. Transitivity in grammar and discourse. Lg. 56.251-99. Vendler, Zeno. 1967. Linguistics in philosophy. Ithaca, NY: Cornell University Press. Department of English[Received 8 February 1994.] The University of Queensland Brisbane, QId 4072 Australia Linguistic semantics. By William Frawley. Hillsdale, NJ: Lawrence Erlbaum, 1992. Pp. xvii, 533. Paper $39.95. Reviewed by Victor Raskin, Purdue University By 'linguistic semantics', Frawley means 'the study of literal, decontextualized, grammatical meaning' (1). Other scholars have referred to similar enterprises as 'grammatical semantics', 'syntactical semantics', or 'semantics of syntax'. I will use the first of these terms (GS) for no principled reason. F makes it very clear, actually even before (xiii) the very first sentence of the text partially quoted above, that GS is what the book is about. I think it would have been even better if this had been reflected in the title as well. This review is going to be 'external' in the sense that it is written by a semanticist who is not comfortable with dividing linguistic semantics into GS and the rest of it, let alone with defining that rest of it out of linguistics and into philosophy or some such place. But I believe I understand enough about where GSers REVIEWS553 are coming from to suggest that they may also be uncomfortable with F's slippage away from GS proper, both in his theory and in his examples throughout the book. The book is a very substantial enterprise. Its 500-plus pages are divided into ten chapters, each with its own summary. Many sections, subsections, and sometimes groups of subsections are summarized as well. I found myself reading those summaries first and then delving into the sections, so I wonder if these summaries would not have worked better as road-mapping prefaces. The first two chapters stand apart both in their relative brevity and in their content, which is general theory. The remaining eight chapters deal each with one facet of meaning, often corresponding to a lexical category or a grammatical feature. Ch. 1, 'Semantics and linguistic semantics: Toward grammatical meaning' (1-16), briefly defines the nature of F's enterprise as the study of overtly grammatically encoded meaning. As an example of what is included in and what is excluded from that meaning, the causative and inchoative meanings of kill as 'cause to become dead' are included, while 'death' is not. In Ch. 2, 'Five approaches to meaning' (17-61), F discusses meaning as reference, meaning as logical form, meaning as context and use, meaning as culture, and meaning as conceptual structure. Loosely paraphrased, these can be presented in a more...