A grammar of Kambera By Marian Klamer (review)

BOOK NOTICES 469 clause types as well as relative clauses. Ch. 4 deals with heads and phrases and introduces the distinction between head-marking and dependent-marking. Ch. 5 examines constituent structure and introduces various diagnostics for constituency. Ch. 6 looks at relationships within the clause—that is, ordering, case, agreement, and grammatical relations. Ch. 7 looks at syntactic processes (in essence, valency-changing derivations) and movement rules. Ch. 8 is a short conclusion. At least two chapters in the book could stand to be expanded. The section on relative clauses in Ch. 3 is somewhat superficial and could stand to go into greater depth, especially given the amount of attention that relativization has received in the functionaltypological literature. Also, movement rules are lumped together with valency-changing derivations in Ch. 7, but they really merit their own chapter, in part because they are (arguably) different in kind from valency-changing derivations but mostly because so much attention has been lavished on them in the literature. There are also some problems with the book on a pedagogical level. First, a few aspects of the book's organization are questionable. T discusses word classes, heads and phrases, and constituency in three separate chapters, but the order ofpresentation forces some pedagogical leaps. For example, the structure of sentences and the notion of head words and phrases are both introduced before the basic concept of constituency is laid out. Second, the book sometimes has some problems shifting back and forth between English and other languages. For example, in the chapter on constituency, T introduces three tests for constituency: (1) sentence fragments, (2) echo questions, and (3) clefts. (Later she also introduces ellipsis and coordination.) However, these are a somewhat mixed group. While all languages have sentence fragments of some sort, not all languages have clefts. The notions of constituency that are developed are therefore too English-specific to be widely applied across languages. It would be better to introduce general types ofconstituency tests—isolability , replacement, movement, e.g.—and then illustrate them with specific examples taken from English. Two things distinguish this book from most other introductory syntax textbooks. First, unlike many (if not most) introductory syntax textbooks, Understanding syntax takes crosslinguistic variation quite seriously and provides examples and exercises from numerous languages (with wide areal and genetic sampling). Second, no particular syntactic framework is assumed in the book, and formalism is kept to a minimum. For these two reasons alone, the book will be a valuable resource. [Stuart Robinson, Australian National University.] A grammar of Kambera. By Marian Klamer. (Mouton grammar library 1 8.) Berlin & New York: Mouton de Gruyter, 1998. Pp. 448. Cloth DM 318, $199.00. Marian Klamer presents us with the culmination of nearly nine years of research in this grammar, an augmented version of her PhD dissertation focusing on the phonology and morphosyntax of the Central Malayo-Polynesian language spoken by approximately 150,000 speakers in the eastern region of the island of Sumba in Eastern Indonesia. The language is known by various names, hilu Humba 'Sumbanese language' to the native speakers, Sumbaneesch, Sumba (a)sch, Oost-Sumbaas, and Kamberaas to the Dutch, and Bahasa SumbalKambera to the Indonesians . The grammar outlines the structural properties of Kambera in more or less theory-neutral terms, thus ensuring its accessibility in years to come. The phonology section, however, does appeal to nonlinear models from prosodie morphology that may be less accessible to the uninitiated. The book is partitioned into eight chapters: 'Introduction ', 'Phonology', 'Morphological and syntactic units', 'Word classes', 'Intransitive argument linking ', 'Derivational morphology', 'Complex verbs' (including serial verb constructions), and 'Subordinate clauses'. The volume closes with 39 pages of interlinearized texts, 1 1 pages of references, and an index. K describes Kambera in a nominative/accusative framework, providing reasonable justification that A (the agent of a transitive verb) patterns with S (the subject of an intransitive verb), more than S patterns with O (the object of a transitive verb). In controlled clauses and raising constructions, for instance, the controlled pivot or raised argument must be the S or A argument, not the O. A and S also pattern together in relative clauses and in coordination. However, a few quirks inthe...