Case selection for the direct object in Russian negative clauses By Arto Mustajoki and Hannes Heino (review)

BOOK NOTICES 613 cal framework is functional typology, although the authors are very cautious with typological generalizations. Even in their conclusion (Ch. 18.11; 'Some remarkable characteristics of Samoan syntax'), they refrain from correlating obviously related characteristics in a functional perspective (e.g. the correlation of verbal 'lability ' with ergativity and predicate prominence, or with the syntactically-based aspect system.) M&H's approach is data-oriented, with preference for modern Samoan of the tautala lelei register ('the good language'). Although both authors have done extensive fieldwork, published data was preferred, so that other researchers can go back to the indicated sources and check contexts and usages. The examples (they are legion!) are scrupulously analyzed, glossed, and translated. When the English translation risks disfiguring the characteristic structure of the original, a literal translation is added. The orthography of quoted examples is left unchanged. This is understandable from a philological point of view, but it has the inconvenience that very often the reader has to check vowel length and diacritics in the glossary. The first part of the book (1-65) begins with background information about the language, its history and varieties, Samoan linguistics, and the genesis of this particular grammar (Ch. 1). This is followed by a descriptive analysis of the language's phonology and orthography (Ch. 2) and a preliminary view of its syntactic organization (Ch. 3). This overview reveals the organizational principles of the book, inspired by Georg von der Gabelentz's ideas on grammatography, e.g. the dichotomies of 'form to function' vs. 'function to form', or of the internal structure of a grammatical unit vs. its external relation to other units on a higher hierarchical level. The second part of the book (67-774) is a detailed description of Samoan morphosyntax. Starting from the discussion of word classes (Ch. 4) and morphology (Ch. 5), the authors treat phrasal units (Chs. 6-8), clauses (Chs. 9-14), complex sentences (Chs. 15-16), and coordination (Ch. 17). In a sort of summary, the last chapter takes up the question of case marking and grammatical relations. The book also contains an indexed word list, a subject index, and a detailed bibliography. The grammar addresses typologists and linguists of any theoretical background who are unfamiliar with Samoan. It is also intended to help in the development of teaching materials and school books in Samoa. Because it challenges some of the basic concepts of universal grammar, however, its main benefit will lie in the advancement oflanguage typology and functional linguistics. [Martin Haase, University of Osnabrück.] Case selection for the direct object in Russian negative clauses. By Arto Mustajoki and Hannes Heino. (Slavica Helsingiensia, 9.) Helsinki: University of Helsinki, 1991. Pp. viii, 249. The Finnish Slavist Arto Mustajoki is engaged in a lengthy research project to apply quantitative methods to the thorny problem of Accusative vs. Genitive case marking for the direct object in Russian negated clauses. This volume is the second interim report on his findings , presenting 'the results of a basic analysis of the material in the corpus' (Preface). The first report, published in Russian in 1985 (BN, Lg. 63:189, 1987), gave a comprehensive review of the extensive literature on the subject—summarizing and weighing the various morphological , syntactic, semantic, and pragmatic factors that have previously been used in attempting to explain the choice of case for negated direct objects in Russian—and reported on a preliminary questionnaire experiment. For the second stage of the project, M (together with several graduate students) compiled a relatively large computerized corpus of varied current Russian texts, from which he extracted 4,037 negated clauses with direct objects. He then excluded certain marginal cases (indeterminate case in the corresponding affirmative sentence, idioms, etc.), as well as clauses in which a single strong factor eliminated the variation in case (e.g. , human objects were invariably Accusative, the predicate ¡met' 'have' invariably took Genitive objects when negated, etc.), leaving 2,722 examples in which the choice of case rests upon nondiscrete factors—these are the crucial data, and the reason why Slavists spend so much time on this issue. The bulk of the book consists of a systematic analysis of the effects of 33 'variables' that influence...