Linguistic criticism 2nd edn. By Roger Fowler (review)

448 LANGUAGE, VOLUME 74, NUMBER 2 (1998) Lexiques-grammaires comparés en français: Actes du colloque international de Montréal (3-5 juin 1992). Ed. by Jacques Labelle and Christian Leclère. (Lingvisticae investigationes: Supplementa 17.) Amsterdam & Philadelphia : John Benjamins, 1995. Pp. 217. This is a collection of sixteen papers based on presentations made at an international colloquium on lexicon-grammars in 1992, by researchers from Belgium , France, and Quebec. AU the papers focus on the French language and address at least one of the following topics: (1) characterization of Belgium (B), France (F), or Quebec (Q) variants of the French language; (2) studies in the lexicon-grammar approach ; and (3) natural language processing. Papers are presented below in their order of appearance in the book. Jacques Labelle (Quebec) proposes a contrastive study of F and Q idiomatic expressions, based on the lexicon-grammar descriptive approach. André Goosse (Belgium) demonstrates the problems posed in the characterization and identification of B as opposed to other varieties of French. Jean-René Klein and Béatrice Lamiroy (Belgium) present an inventory ofB idiomatic expressions, based on the lexicongrammar approach. Maurice Gross (France) offers some remarks on the French indefinite article un as determiner of nouns in predicate position. Alain Guillet (France) demonstrates the use of lexicongrammars for automatic part of speech tagging. Gaston Gross (France) studies the syntactico-semantic class of 'human' nouns. Max Silberzteem (France) presents the lexical parser INTEX, which performs word identification. Eric Laporte (France) offers a continuation of the preceding paper that focuses on the problem of solving ambiguities. David Clemenceau (France) presents a formal model of French derivation to be used to improve the performance of parsers based on lexicon-grammars. Jocelyn Desbiens , Lorne Bouchard, and Louisette Emirkanian (Quebec) analyze an algorithm that performs suffixal analysis ofFrench word-forms in linear time. Francine Caviola (Quebec) presents a study of process nouns based on the lexicon-grammar approach for the purpose of making comparisons between Q and F. Jean-Marcel Leard and Pierre Larochelle (Quebec) contrast verbal constructions in Q and F, based on the identification of variations in the argument structure of verbs. Gaétane Dostie (Quebec) offers a semantic analysis ofsentences that are lexical units, e.g. French Sais-tu 'You know', using the meaning-text approach. Pierre Martel and Sylvie Thiboutot (Quebec) study Québécois' perception of their language, based on experiments using dictionary usage notes. Chantal Contant and Éric Brunelle (Quebec) present Exploratexte, software for improving students' knowledge of written French that uses both lexicon-grammar tables and meaningtext syntactic rules. Sophie Rosset and Jacques Labelle (Quebec) present the TRACO system, which automatically performs lexical transfer between Q and F words using two distinct electronic dictionaries . This collection of papers is rather unbalanced. Some contributions tackle interesting problems and offer constructive analyses while others seem more like conference papers that can be valued only as pointers to more developed pieces of work. As usual for books in this series, the presentation is impeccable . It should be noted that, m spite of the book's title, six of the sixteen papers have no direct relation to lexicon-grammars. [Alain Polguère, University of Montreal.] Linguistic criticism. 2nd edn. By Roger Fowler. Oxford & New York: Oxford University Press, 1996. Pp. viii, 262. Paper $13.95. The second edition of a book that originally appeared in 1986, this is an introduction to the critical study of discourse, with emphasis on 'literary' texts. Fowler's approach draws on various methods of analysis, among them Russian formalism (esp. Viktor Shklovsky), French poststructuralism (esp. Roland Barthes), dialogism (esp. Mikhail Bakhtin), modern literary criticism, and functional-textual grammar (esp. M. A. K. Halhday). Reading the book demands considerable concentration, but the reader is rewarded by becoming acquainted with a wide range ofrelated scholarship. Each ofthe twelve chapters is followed by references and suggestions for further reading. The basic purpose of the book is 'to argue and to demonstrate the value to criticism of an analytic method drawn from linguistics' (7). If literature is a creative use of language, then the first question to be answered is, What is created? (Ch. 2). Linguistic codes do not represent the subjects of discourse neutrally...