The foundations of linguistic theory: Selected writings of Roy Harris Ed. by Nigel Love (review)

BOOK NOTICES 655 In Ch. 1 L discusses the structure of the English lexicon and the varieties of English, and concludes with an informative survey of the most important monolingual English dictionaries . Dealing in Ch. 2 with the various aspects of the linguistic sign, L introduces three sign models—Saussure's, Ogden & Richards's, and Buhler's. After examining the meaning of signs and the different kinds of meaning, he defines and discusses the concepts of morpheme and lexeme. In general, L's conception of the morpheme and its nature derives from classical American structuralism. The main topics of Ch. 3 are word-formation processes, semantic features , and componential analysis. L advocates the integration of feature semantics and prototype semantics as the most urgent task of semantic theory and practice. He also presents his typology of semantic features, consisting of seven types: denotative, connotative, relational , transfer, deictic, inferential, and distinctive . All these types except the inferential function as distinctive features (the seventh, comprehensive type). Ch. 4 is primarily concerned with paradigmatic and syntagmatic relations between lexemes and lexical units. Among other things. L considers the distinction between homonymy and polysemy, the various kinds of paradigmatic sense relations, and linear and hierarchic lexical fields. Ch. 5 contrasts context (linguistic context) and context (extralinguistic setting), both of which contribute to the resolution of polysemy. L welcomes the growing attention contemporary linguists are paying to the lexicon. He gives the metaphor as an example: long excluded from purely linguistic studies, it has now become prominent and its study respectable. According to L, 'the function of words in context and the creative forces of lexical rules and semantic processes must be seen as the fundamental tools which help man to come to grips with the universe around him' (188). Lipka has contributed to the development of modern semantics and lexicology in his numerous writings over the past two decades. This work is a good introduction to and survey of English lexicology. It is supplemented by a listing of over fifty important monolingual English dictionaries, an extensive bibliography (195203 ), and indexes of subjects and names. The book is well printed and misprints have been kept to a minimum. [Zdenek Salzmann, Northern Arizona University.] The foundations of linguistic theory: Selected writings of Roy Harris. Ed. by Nigel Love. London & New York: Routledge, 1990. Pp. xv, 236. Cloth $65.00. This collection of thirteen essays and book excerpts from twenty years of Roy Harris's career as a Socratic gadfly on the back of autonomous linguistics supports Nigel Love's claim in his preface that H should be read more, not only by linguists who engage in autonomous linguistics but also by those who are looking for an alternate model of discourse with which to talk about language as a nonautonomous mode of social interaction. Love's helpful orienting 'Introduction' (1-21) argues that H's work as a whole is an analysis of the thesis that 'linguistics has been led astray by the fact that . . . "man is capable intellectually of decontextuahzing his own verbal behaviour "' (19). H's argument in the work represented in this volume is compelling, if for no other reason than the catholicity of his targets , which range over Ferdinand de Saussure, Leonard Bloomfield, Edward Sapir, Zelig Harris , Charles Hockett, John Lyons. Noam Chomsky , John Searle, and William Labov, among many others. Regardless of the target, H seems to have in mind not a personal attack on any individual, but a direct attack on the isolation of linguistic form from meaning and communication. For example , H argues that post-Bloomfieldian concerns with distributional linguistic analysis not based on meaning usually relied on post-hoc rationalization for an analysis already reached by the use of meaning ('Synonymy and morphological analysis', 22-43). Linguistic arguments which analyze meaning explicitly within an autonomous framework fare no better with H. In 'Truth-conditional semantics and natural languages' (96-111), he attacks the assumption in truth-conditional semantics that there is a 'universal communication situation with universal purposes' (109), such that analytic sentences like Triangles have three sides and synthetic sentences like Triangles are unlucky are analyzed with the same context-neutral truth conditions (106-7). H turns his...