In lieu of an abstract, here is a brief excerpt of the content:

BOOK NOTICES 397 data on N-S languages. Its chief value is in expanding upon the much shorter sketches ofMasalit and Maban by Archibald Tucker & Margaret Bryan in 1966 (in Linguistic analyses: The nonBantu languages ofnortheastern Africa, Oxford U. Press), particularly because of its careful attention to the fascinatingly rich systems of inflection and derivation (e.g. of person, tense, and mood). Tojudge by this work alone, Masalit is in most respects no more typologically striking than many other N-S languages that have been described. Still, there are clues here suggesting that further investigation will yield valuable insights into phenomena like vowel harmony, relationships between transitivity and definiteness, and the death and morphological reanalysis of semantically-based class systems in language. A glaring problem in this description of Masalit is the tendency not to mark tone on examples , although E clearly indicates the lexical and morphological importance of tone. A glossary would have made the sketch much more useful, as would more data exemplifying and supporting E's analysis of morphophonemic alternations and the meaning and use of derived categories, especially in verbs. (For instance, a word meaning 'know'—E's paradigm example—should have been supplemented with verbs of nonstative meaning, both transitive and intransitive.) Another disappointment is the seven pages of notes on syntax, which are telegraphic and scattered . This would be less unfortunate if the author had glossed the texts in the following section or included a glossary sufficient to allow interested readers to interpret these few short selections for themselves. As they stand, the texts (few of which are tone-marked) will probably be useful more for ethnographic or sociolinguistic purposes than for linguistic analysis. While the section on the Miisiirii language is too brief to serve anyone but the comparativist of basic N-S lexicon and morphology, it is valuable as the only published data on this variety. The closing discussion of language, place, and ethnic group names in the region is really just an extended footnote, an interesting if inconclusive foray into the territory of the anthropologist , historian, and etymologist which neither enhances nor detracts from the main body of the work. Despite these shortcomings, what this book attempts to do it does well. It will be gratefully received by Nilo-Saharanists and possibly (because of the quality of the better developed sections of the sketch of Masalit) by morphologists. Too much field work like this remains out of sight; it should be published, if only to serve as a solid base on which more enlightening descriptions may build. [Keith Denning, Eastern Michigan University.) Tense and text: A study of French past tenses. By Dulcie M. Engel. (Croom Helm Romance Linguistics Series.) London & New York: Routledge , 1990. Pp. xii, 147. Cloth $65.00. The two punctual past tenses of modern French, the passé simple (PS) and the passé composé (PC), are regarded by prescriptive grammarians as register variants which should not be mixed in the same text. Typically, PS is said to be limited to use in written, generally impersonal, historical or fictional narratives, while the more colloquial PC is appropriate for narratives of personal experience. However, both tenses occur routinely in the same text in French newspaper prose, where tense choice is generally complex. Engel analyzes a corpus of newspaper and newsmagazine texts with regard to fourteen factors mentioned in the literature on PC/PS choice. Ch. 1 (1-13) reviews the indicative tense system of French and describes the corpus of newspaper and newsmagazine texts on which the study is based. Ch. 2 (14-32) surveys the literature on the PS/PC contrast in the French tense system, reviewing especially studies ofthe contrast between the two tenses in newspaper language (23-27). Ch. 3 (33-55) describes the qualitative and quantitative methodologies used and the characteristics of the newspapers and the articles which show mixing of PC and PS. Upmarket papers like Le Figaro and Le Monde were only slightly more likely to use PS than were downmarket papers like France-Soir (46). In contrast to the figures cited in Waugh & Monville -Burston 1986 (Lg. 62.846-77), national papers in this study were more likely to use PS than the regional...

pdf

Share