BOOK NOTICES 223 ters, but no indication of what principles underUe the distinction between permissible and impermissible clusters. One ofhis few analytical statements—that vowel and consonant length is non-phonemic (29, 67)—would have benefited from more detailed justification; if it is valid, it seems strange that pt and ptt are listed as distinct permitted consonant clusters. [Bernard Comrie, USC] The grammatical hierarchy ofMalayan Cantonese. By Siew-Yue Killingley . Newcastle upon Tyne, England : Grevatt & Grevatt, 1982. Pp. xxiv, 231. £5.70. This work, a revision of K's University of London doctoral thesis, is an extensive catalog of Cantonese sentence, clause, phrase, and word types, presented in a tagmemic model. It contains a great deal of data, presented in a format which makes it relatively accessible; however , most readers wiU probably have to work with the book for an hour or two to famiUarize themselves with the organization and some terminological conventions before they wiU be able to use it easily as a reference tool. Among the strengths of the book are the extensive exemplification, the exhaustive inventory of clause types and syntactic categories, and the unusually careful attention given to intonation and sentence-final particles. Certain weaknesses are inherent in the nature of the work; in particular, the admirable breadth of coverage precludes discussion in depth ofmany of them. Thus, whüe the treatment of interrogative constructions wiU satisfy many readers whose interest lies in that area, the material on sentential complementation (which is, of course, a much larger topic) is less comprehensive . The book wiU be extremely useful to anyone planning to undertake work on Cantonese syntax; typologists and students of Southeast Asian areal phenomena will find much of interest here, but cannot be sure of finding all the detail which their work may require. The physical format of the volume is quite unsatisfactory—a problem acknowledged by K: it is a lithographed typescript, reduced so as to fit two pages onto an 8 x 11" sheet, bound with fragüe and short-Uved tape. There is an exhaustive subject index, but no morpheme index. [Scott DeLancey, University of Oregon.] La lengua tlapaneca de Malinaltepec. By Jorge A. Suárez. México: Universidad Nacional Autónoma de México, 1983. Pp. xxiii, 641. Suárez, who unfortunately died in 1985, has here given us the first detailed monograph description of Tlapanec, an Otomanguean language of southwestern Mexico with about 30,000 speakers. A general introduction is followed by individual chapters on phonology (phonemes, including tones, and their aUophones, phonotactics, general morphophonemic alternations; 5-62), grammatical categories (63-97), morphology (of nouns, verbs, and other grammatical categories; 98-222), and syntax (of phrases and simplex and complex sentences; 223-364). There is also a sample text with morpheme glosses and translation. The second major part of the monograph is an extensive Tlapanec-Spanish vocabulary (377582 ), accompanied by a Spanish-Tlapanec index. FinaUy, there is a bibUography of previous work on Tlapanec. The entire description is comprehensive and on a high level. Tlapanec nominal and verbal morphology is extremely complex, not least in its tonal alternations, and those more interested in general theory may remain dissatisfied with S's essentiaUy word-and-paradigm analysis. But for a hitherto virtually undescribed language, it is essential that the data necessary for an analysis be presented, and S has certainly provided a wealth of data for anyone interested in re-analysing Tlapanec phonology and morphology. After struggling through so many descriptive grammars, trying to make sense out of vague statements about tense and aspect, I found it a pleasure to see S's clear demonstration that Tlapanec operates in terms of a relative tense opposition (75). The discussion of syntax is particularly extensive, and includes some topics that seem tojustify further monographic study. For instance, there are different forms depending on whether a 3rd person inflection refers to an NP mentioned expUcitly in the clause (S caUs this remission; 69, 89, 322-5). Since Tlapanec is a PRO-drop lanaguage, for 'he left' one can either include the subject pronoun (with remission ) or omit it (without remission). S emphasizes that remission is essentiaUy a syntactic phenomenon—although in appropriate cases it can...