The Language of Word Meaning (review)

This volume collects a range of articles promoting and critiquing generative approaches to the lexicon, particularly James Pustejovsky’s generative lexicon framework (GL). The contributions range from the philosophical (dealing with the nature and role of creativity in the lexicon) to the very practical (reports on the development of particular computational resources). After a preface by James Pustejovsky (JP) and the editors’ introduction, the volume is divided into four sections, each with its own introduction by the editors (who are not shy about their own theoretical allegiances).

The first section, ‘Linguistic creativity and the lexicon’, starts with philosopher James McGilvray’s review of Chomsky’s relation to the topic, focusing on the role of modularity in allowing for lexical creativity and overlap between Chomsky’s and JP’s approaches. The next two chapters reprint Fodor and Lepore’s (F&L) 1998 critique of The generative lexicon and JP’s reply (both from Linguistic Inquiry). Time has not altered the fact that the response does not fully address the issues in the critique, but Yorick Wilks enters the discussion with an argument that F&L’s attack rested on false assumptions and data (‘The “Fodor”–fodor fallacy bites back’). JP addresses the issue of semantic relations (one of F& L’s concerns) and qualia structure in ‘Type construction and the logic of concepts’, which leads the section on ‘The syntax of word meanings’.

Jacques Jayez, in ‘Underspecification, context selection, and generativity’, agrees that semantic representations are underspecified but disagrees with JP’s attribution of generativity to semantic representation, and instead argues that a two-way relationship between lexicon and context is best treated through a construction grammar approach. Bouillon and Busa apply GL to the French verb attendre. Patrick Saint-Dizier develops the theory (particularly the telic role) to account for adjectival modification, and a range of selective binding and metonymic verb phenomena. Salvador Climent looks at the role of constitutive information in the representation of Spanish partitive constructions, then Laurence Danlos develops the GL approach to direct causation, raising questions about JP’s approach to unaccusatives.

Papers in the next section, ‘Interfacing the lexicon’, deal most directly with the issue of creativity. Julius Moravcsik provides an account of metaphor built upon JP’s approach to polysemy. The article is strangely silent on the extensive cognitivist literature on the topic. Nicholas Asher and Alex Lascarides present a rule-based account of metaphor that is compatible with a constraint-based grammar, illustrating their arguments with metaphorical uses of French verbs. Jerry Hobbs approaches metonymy in the ‘interpretation as abduction’ framework. The last contribution to this section is Adam Kilgarriff’s test of GL predictions against corpus evidence. While Kilgarriff’s experiment is an interesting one and his conclusion (‘GL is a theory for some lexical phenomena, not all’, p. 327) is undoubtedly correct, it can be difficult to come to one’s own conclusions because Kilgarriff’s verdicts on the data are often given without much illustration. Since some of these ‘unaccountable’ examples are metaphorical, and a GL approach to metaphor is included in the volume, a bit of head-scratching is inevitable.

The final section of the book, ‘Building resources’, includes two articles on the SIMPLE project, which implements various aspects of the GL framework (Busa, Nicoletta Calzolari, and Alessandro Lenci; Nilda Ruimy, Elisabetta Gola, and Monica Monachini), and one on ‘condensed meaning’ in EuroWordNet (Piek Vossen). This book is a major contribution to the development of the GL framework, but also very important for the critiques of GL that it includes.