Syntax: A functional-typological introduction, I By Talmy Givón (review)

160LANGUAGE, VOLUME 62, NUMBER 1 (1986) word. This would permit the generalization that a modifier precedes its head when both are part of a word. Nevertheless, a lot of ingenuity is reflected in this book, and any linguist can profit by reading it. In fact, I think a viable theory of linguistic structure must eventually incorporate a number of features that distinguish word grammar from the theoretical approaches currently in favor. Word grammar consists of a preface, six chapters ('Introduction', 'Morphology and lexical relations', 'Syntax', 'Semantics', 'Co-ordinate structures', and 'Spoken and written utterances'), a list ofreferences, and name and subject indexes. Hudson's prose is generally quite clear, though it is sometimes difficult to grasp his complex formulas and to follow the intricate details of his consideration of alternate analyses. Physically, the book is reasonably attractive. Mechanically, there are far too many errors, some with the potential for hindering comprehension. Certain diagrams appear to be faulty, with missing arrowheads (47, 109), incorrect subscripts (64), inconsistencies in the direction of dependency (99), or arrows that are misdrawn (230). Something has gone awry with the printing of one formula (230), and 'big' is omitted from the gloss of one example (82). Numerous minor errors of spelling etc. occur in the references , and a few in the text. Extra lines of text are inserted in a few places; extra spaces are sometimes put between paragraphs; and formatting mistakes can be found. Finally, I noted a fair number of cross-references to page 000 which I have been unable to locate. REFERENCES Hudson, Richard A. 1976. Arguments for a non-transformational grammar. Chicago: University of Chicago Press. Jackendoff, Ray. 1983. Semantics and cognition. (Current studies in linguistics, 8.) Cambridge, MA.: MIT Press. Langacker, Ronald W. 1986. Foundations of cognitive grammar, I: Theoretical prerequisites . Stanford: University Press. [Received 12 July 1985.] Syntax: A functional-typological introduction, I. By Talmy Givón. Amsterdam & Philadelphia: Benjamins, 1984. Pp. xx, 464. Reviewed by Jeffrey Heath, Harvard University This is half of a two-volume work designed to be used as a textbook for a typological survey, or a typologically oriented syntax course (p. 26). In many ways, though, it is written as a compendium of Givonian thinking; the preface hints that it represents the fleshing out of G's polemic On understanding grammar (Academic Press, 1979), the examples are principally from G's own field materials (many unpublished), and few opportunities are missed to cite G's earlier publications in footnotes. There is nothing wrong with this kind of compendium from such a prolific and wide-ranging author, but it is somewhat difficult to use such a work as a classroom text unless the instructor is sympathetic REVIEWS161 to and well-versed in the relevant corpus—in which case the book can be recommended. Chap. 1 , 'Background' (1-28), gives a capsule review ofWestern intellectual history, culminating in prototype theory; G explains that typological correlations are usually probabilistic rather than absolute, because of overlapping or conflicting factors. Chap. 2, 'Methodological preliminaries' (29-45), uses biological analogies, identifies major formal /functional divisions of communication, proceeds to discuss 'functional domains' such as referential tracing, and philosophizes about types of functional explanation. Chap. 3, 'Word classes' (47-84), motivates noun/adjective/verb divisions through time stability, describes nominal hierarchies, surveys basic grammatical categories, and identifies classes of adverbs. Chap. 4, 'Simple sentences' (85-134), analyses predicators in terms of case (and complement) frames; 'dative' is used for conscious non-agentive participants, such as the subject of 'know' or the object of 'frighten', as well as indirect objects. There is a section on transitivity scales, and discourse patterns are used to elucidate cognate objects, 'bleached' verbs, and object incorporation. Factive, implicative, and manipulative verbs are distinguished with attention to the logic of negation. Chap. 5, 'Case-marking typology' (135-85), presents case systems as the result of conflicting semantic and pragmatic factors. 'Subject' is maximally topical, DO (direct object) has secondary topicality, and others are generally non-topical. Nominative/accusative systems thus emphasize subject topicality; split-intransitive ('active') systems favor semantics; ergative systems are depicted as hybrids which put most emphasis on transitivity. Split ergativity and antipassives are analysed as functions...