Kwaio dictionary By R. M. Keesing (review)

498 LANGUAGE, VOLUME 53, NUMBER 2 (1977) need to consult all three works. Such inclusion would have been more useful than the full bibliographical coverage which Bakalla occasionally gives to works which only marginally touch upon Arabic (e.g. Sound pattern of English). In general, however, this is a useful bibliography, bringing to light a number of otherwise obscure publications. [John Hayes, UCLA.] A semantic study of transitivity relations in Chinese. By Shou-hsin Teng. (University ofCalifornia publications in linguistics, 80.) Berkeley & Los Angeles: University of California Press, 1975. Pp. ix, 177. $9.00. This refreshingly original volume is a study of the semantic relations between nouns and verbs in (Mandarin) Chinese sentences. It was submitted as Teng's doctoral dissertation in 1971 ; but because of its independence from any of the more or less rigidly codified grammatical models current at that time, its value has not diminished since it was written. In fact, as emphasis in grammatical studies continues to shift away from mechanisms and toward semantic explanations, the worth of this book will only be enhanced. Teng proposes a synthesis of three previous approaches to the study of semantic relations, those of Fillmore, Halliday, and Chafe. In his first two chapters, he presents and compares these three approaches. The last two-thirds of the book is devoted to the introduction and justification of Teng's own set of transitivity relations in terms of the grammar of Chinese. By 'transitivity relations' he means those which he labels Agent, Patient, Goal, Range, and Causer. These can be distinguished from 'circumstantial relations' such as Benefactive, Instrumental, and Comitative on several grounds: verbs are adequately characterized by transitivity relations alone, and these relations are neither optional nor predictable for a given verb: 'Transitivity relations cluster around (notionally , not syntactically) the verbs and form the inner core of all semantic structures. Circumstantial relations, on the other hand, form the peripheral parts of sentences. They are selected, obligatorily (rare) or optionally, by the inner core' (36). Teng makes a convincing case for the usefulness of this semantically quite reasonable distinction for Chinese grammar: it turns out that transitivity relations do not appear with prepositions in Chinese, while circumstantial relations do. Certain heretofore intractable problems in Chinese grammar seem to yield to Teng's analysis. To take one example, the question ofwhich directobjects canoccur in the 'ba construction ' has been at issue since the inception of Chinese grammar studies. Teng shows that, defined on independent semantic grounds,only his Patient, but not Range or Goal, noun phrases can appear in this construction: Patient: Ta ba nei-feng xin sMo Ie he that letter burn asp 'He burned that letter.' Range: *Ta ba faguo-hua shuo Ie he Frenchspeak asp 'He spoke French.' Goal: *TabaSHU mai le he book buy asp 'He bought the book.' Similarly, certain syntactic oppositions in complement clauses can be explained in terms of whether sentential Goals or Patients are involved. At the same time, as would be expected in any classificatory scheme, there are some indeterminacies, for which a solution seems arbitrary: thus a sentence like Dlren tui Ie 'enemy withdraw asp', i.e. 'The enemy retreated' is analysed as having a subject which is both Agent and Patient. For the specialist in Chinese grammar, this book is a must. It provides a rational discussion of a wide range of grammatical problems which future studies will build on. However, the book also contains much value, as I have indicated, for the general linguist interested in theories of semantic relations. It is informative, clear, generally well-argued; and the data are accessible. It is a particularly good example of the kind of insights that can be derived from an approach which has as its goal the specification of the basic semantic organization of a language. [Sandra A. Thompson, UCLA.] Kwaio dictionary. By R. M. Keesing. (Pacific linguistics, series C, 35.) Canberra: Department ofLinguistics, Research School of Pacific Studies, The Australian National University, 1975. Pp. xxxiv, 296. A$10.50. Keesing's Kwaio dictionary represents contributions of several collaborators, ac- BRIEF NOTICES 499 cumulated as a by-product of three years of anthropological field work. It is the first published information available on Kwaio...