The Lillooet language: Phonology, morphology, syntax by Jan van Eijk (review)

680 LANGUAGE, VOLUME 74, NUMBER 3 (1998) using constituent structure to integrate words into coherent meaning and that these children are not using the last-word strategy. Ch. 5 (99-122) presents experiments designed to test comprehension of word order. Each videotaped sequence shows an agent, an action, and a patient of the action, differing only in which character (Big Bird or Cookie Monster) plays which role. Audio prompts consist of reversible sentences such as, 'Cookie Monster is feeding Big Bird!' The control experiment again uses an audio prompt with no actual matching video screen, such as, "Find Big Bird!" H & M conclude that the experiments show comprehension of word order cues by children who are not yet producing multiword speech. Ch. 6 (123-58) presents experiments in which the children's comprehension of syntactic frames is investigated . Children are tested with verbs that have both transitive and intransitive interpretations, and with intransitive sentences using the grammatical marker 'with' to block transitive misinterpretations, e.g. "Cookie Monster is turning Big Bird" versus "Cookie Monster is turning with Big Bird." Conjoined subjects such as "Cookie Monster and Big Bird bending'' appear to cause comprehension difficulties in one experiment. In a subsequent experiment the audio prompt is revised to include the plural marker "are", and comprehension improves. In Ch. 7 (159-202) the authors return to discussion of the theoretical debates introduced in the opening chapters and develop their viewpoint that many of the supposed oppositions are really differences in emphasis. They argue that first language acquisition is not a matter, for example, of 'syntax versus semantics ' or 'innateness versus socialization', but rather a coalition of many factors working in concert, with prosody, semantics, and syntax each taking a turn as the dominating factor in the growing child's linguistic analysis. Because the book is clearly written and well organized , there is no confusion between what the authors present as facts versus possibilities. H & M themselves describe their model as "speculative", but it is speculation which is well-founded in both thenown research and the research of others and which suggests creative directions for future investigations. [J. Daniel May, Dominican College ofSan Rafael.] The Lillooet language: Phonology, morphology , syntax. By Jan van Eijk. (First Nations languages, 1.) Vancouver : University of British Columbia Press, 1997. Pp. xxx, 279. Cloth $75.00, paper $24.95. Jan van Eijk's long-awaited grammar of St'át'imcets (also known as Lillooet), a Salish language spoken in British Columbia, Canada, is based on over ten years of fieldwork and is invariably insightful and well-written. In its earlier form as the author's 1985 dissertation, The Lillooet language laid the groundwork for other scholars to continue researching anddocumenting St'át'imcets, anditspublication now will surely help to extend and renew efforts to maintain this severely endangered language. Following the 'Introduction' (xxi-xxviii), which briefly acquaints readers with the Salish language family and traditional St'át'imc culture, the book is divided into three main parts. Part 1 (1-37) surveys the phonology of St'át'imcets, which has a rich consonantal inventory; its 44 contrastive consonants include velars, uvulars, pharyngeals, and glottalized counterparts of most consonants. The vowel inventory , however, is much smaller, having as few as three vowels, depending on the analysis. Attention is also given to the shape of roots, the majority of which are CVC. Data here and throughout the book are given in standard Americanist phonetic transcription ; some works on St'át'imcets have cited forms using the practical orthography, a helpful key to which may be found near the back of the book (251-52). St'át'imcets is a polysynthetic language, entailing that arguments are obligatorily inflected on the predicate . Overt noun phrases are optional, and actually dispreferred in discourse, and so a well-formed sentence often may constitute a single (though heavily affixed) word. It is not surprising, then, that Part 2 ('Morphology', 39-221) forms the body ofthe book; it includes sections on the small set of prefixes and on the larger set of suffixes, which encode aspect, transitivity, and reflexivity as well as further lexical information. The proclitics encompass articles, conjunctions , and prepositions, while enclitics include adverbials and interrogative...