Social motivations for codeswitching: Evidence from Africa By Carol Myers- Scotton (review)

REVIEWS841 types of cognitive but nonlinguistic factors are involved. For example, if our theory of grammar hypothesizes that linguistic knowledge is categorical, then gradient phenomena are extralinguistic by definition. The interface between the grammar and other systems which contribute to the production of an utterance (the output of an organism, as opposed to the output of a grammar) is in dire need of study. Extragrammatical aspects of cognition are clearly worthy of study but are, strictly speaking, not part of linguistic theory. If external evidence from acquisition and disorders can eventually be integrated into a coherent and principled model of the grammar, then it is likely that the evidence is relevant to linguistic theory. The interesting papers in this volume represent a real contribution to the attempt to demonstrate this relevance. REFERENCES Archangeli, Diana. 1985. Yokuts harmony: Evidence for coplanar representation in nonlinear phonology. Linguistic Inquiry 16.335-72. Berwick, Robert. 1986. The acquisition of syntactic knowledge. Cambridge, MA: MIT Press. Dell, François. 1981. On the learnability of optional phonological rules. Linguistic Inquiry 12.31-7. Hale, Mark, and Charles Reiss. 1997a. How to parse (and how not to) in OT phonology. Proceedings of NELS 27, ed. by Kiyomi Kusumoto. Amherst, MA: UMass Graduate Linguistics Association. -----, -----. 1997b. Formal and empirical arguments concerning phonological acquisition. Linguistic Inquiry, to appear. Inkelas, Sharon. 1994. The consequences of optimization for underspecification. Berkeley: University of California, ms. Ito, Junko. 1988. Syllable theory in prosodie phonology. New York: Garland Press. Reiss, Charles. 1994. A theory ofassimilation with special reference to Old Icelandic phonology. Cambridge, MA: Harvard University dissertation. Wexler, Kenneth, and M. Rita Manzini. 1987. Parameters and learnability in binding theory. Parameter setting, ed. by Thomas Roeper and Edwin Williams, 41-76. Dordrecht: Reidel. Yip, Moira. 1996. Lexicon optimization in languages without alternations. Irvine, CA: University of California , MS. Department of Classics, Modern Languages and Linguistics Concordia University 1455 de Maisonneuve West Montréal, Québec H3G INI Canada [reiss@alcor.concordia.ca] Social motivations for codeswitching: Evidence from Africa. By Carol MyersScotton . (Oxford studies in language contact.) Oxford: Oxford University Press, 1995. Pp. xii, 177. Reviewed by James A. Walker, University of Ottawa Why do bilingual speakers conduct a conversation in two languages rather than one? To answer this question, Myers-Scotton calls on her many years of research experience and an array of examples of codeswitching (CS) data from Africa. As she says in the introduction (1-8), the goals of her overview of CS are descriptive, in that she presents examples from different language pairs, and theoretical, in that she outlines a framework for explaining the psychosocial motivations for CS. Ch. 2, 'The African setting' (9-44), contextualizes the social motivations for CS in a multilingual society (43) by providing an overview of the linguistic situations in two African nations, Zimbabwe and Kenya. As in other societies, language use is associated with different domains: an indigenous mother-tongue with more intimate settings, an (indigenous) lingua franca with multi-ethnic neighborhood and work interaction, and an official European language (here, English ) with official business. The choice of a European language as official is problematic in 842LANGUAGE, VOLUME 73, NUMBER 4 (1997) that it retains associations of colonialism, elitism, and access to education and privilege (28). However, choosing a neutral indigenous official language in each country is made impossible by linguistic diversity and interethnic rivalry, and the indigenous lingua franca is often associated with one ethnic group or is considered a market language. Ch. 3, 'The rise of codeswitching as a research topic' (45-74), discusses the history of the sociopsychological approach to CS, beginning with Blom and Gumperz's (1972) introduction of the notion of situational and metaphorical CS (46). Gumperz's later work introduced the concept of conversational (as opposed to situational) CS and linked CS to monolingual stylistic choices (53-4). Although critical ofthese distinctions, M-S acknowledges Gumperz's importance to the field, especially his idea of language use as a function of interactional dynamics and of language choice as a social strategy. Reviewing subsequent approaches to the social motivations for CS, M-S favors the explanation of CS by abstract constructs as organizing devices within some larger framework, such...