The Nubi language of Kibera: An Arabic creole By Bernd Heine (review)

BOOK NOTICES 991 trast with the often meticulous coverage of the dialect geography tradition, are the lacunae. D's decision (quoted above) to deal only with American studies of social variation is not explained; but certainly the failure to mention L. Milroy's Belfast study (1980), one of the major and influential works of the 1970's, is regrettable. The omission of P. Trudgill's Norwich work (1974) is glaring in the short chapter on structural dialectology . Several important works in American social dialectology are also left unmentioned , including W. Labov's Martha's Vineyard study (1972), and B. Lavandera's muchcited article 'Where does the sociolinguistic variable stop?' (Language in Society 7: 171-82). D. Bickerton's work on implicational scaling is not mentioned; C-J. N. Bailey's important Variation and linguistic theory (1973) receives one sentence of text. The variable rule model, standard in social dialectology studies of the last ten years, is considered beyond the scope of this textbook. D devotes much of Chap. Ill to pointing out errors in the application ofstatistical techniques in studies by C. Feagin (1979) R. Shuy (1968), and W. Wolfram (1969). Similar criticism of the pre-1975 studies, and of variationist methodology in general, may be found in R. Berdan's unpublished but highly influential dissertation, not mentioned by D. Chap. Ill is therefore sketchy at best. As in the two other major chapters, D's approach is basically chronological; but in general, I do not find such an approach suitable for teaching dialectology to students not already familiar with methodology and theory. Other students are easily lost in the plethora of names and dates. The long statistics section in the middle of D's book disrupts the flow of the text, particularly since all the examples are non-linguistic. A text which would concentrate more on the theory and practice of modern dialectology (in its broadest sense), supplemented by an elementary statistics textbook, would better fill the needs ofan undergraduate course. [Ruth King, York University, Downsview, Ontario.] The Nubi language of Kibera: An Arabic creóle. By Bernd Heine. (Language and dialect atlas ofKenya, 3.) Berlin: Reimer, 1982. Pp. 84, with map. Cloth DM 55.00, paper DM 43.00. This is a compact descriptive study of an Arabic creóle spoken in various locations in Kenya and Uganda by people who are known as 'Nubians ' or 'Sudanese', but who are linguistically unrelated to the Nubian groups of the Sudan. H states that the present description covers Nubi as spoken in the Kibera community of Nairobi (p. 1 1); but it should be noted that his main consultant was a multilingual (with English, Swahili , and Luganda), educated in Uganda, where the largest Nubi-speaking community is found. Nubi is used as first language in some communities ; according to H, it represents creolization of a pidgin Arabic used in armies under Egyptian and then British command going back to the last century. Arabic is clearly the main lexifier language, though there is now also a substantial Swahili component (in addition to recent Arabic importations filtered through Swahili). Phonological characteristics include consonantal shifts including deletion of Ar. f, deletion of Ar. h (or shift to h), shift of Ar. q to g, shift of Ar. ? and y to k, and frequent loss of wordfinal consonant, especially Ar. t. H states that former Ar. g became/ 'prior to the pidginization of Nubi' (p. 18); he refers here to original Ar. g, prior to the voicing of Ar. q to g. However, the dialectal Ar. g in question is a reflex of Class. Ar. j, preserved as/ (or z) in other dialects ; so one may suspect that Nubij may be a direct reflex ofolder Ar.jwithout an intervening g stage. Nubi morphology is sharply distinct from that of Class. Ar. or ordinary Ar. dialects. A few Ar. sg/pl. pairs have been preserved, but the productive plural marker is a suffix -a, perhaps based mainly on Class. Ar. fern. pi. -àt, but also involving other possible Ar. forms. A human collective prefix nas-, as in nas-babâ 'group of fathers' , goes back to an Ar. noun meaning 'people '. Most of the prepositions are...