Variation in small claims court: Style-shifting versus codeswitching in litigant testimony Codeswitching between different languages is often likened to style-shifting within a single language

Codeswitching between different languages is often likened to style-shifting within a single language (c.f. Gumperz 1982, Myers-Scotton 1993, Ervin-Tripp 2001, Schilling-Estes 2001), yet the two phenomena are generally investigated separately and in studies that fall into different linguistic subfields. This paper compares style-shifting and codeswitching in the same setting, drawing on a data set of forty informal courtroom interactions recorded in New York small claims courts. These interactions include native speakers of English as well as speakers of other languages (Haitian Creole, Polish, Russian, or Spanish) who are assisted by interpreters. Both the interethnic character of the interaction and the institutional context make the data particularly suitable for investigations of style shifting, as individual participants bring different linguistic varieties to the interaction (compare Bell 2001, Schilling-Estes 2004) and as lay court users (litigants) are confronted with a professional register used by attorneys and legal decision makers (see e.g. O'Barr 1982, Jacquemet 1996). All lay court users in the sample are found to engage in style-shifting (exemplified by phonological, morphosyntactic, and lexical variables), yet interpreter-assisted litigants also codeswitch to L2 English in addition to style-shifting within their L1. Across interactional episodes as well as in the speech of individuals, codeswitching and style-shifting occur in similar environments, yet differences are found as well. The paper shows that, unlike codeswitching, many examples of style-shifting can be fruitfully interpreted in the attention to speech paradigm (Labov 1972), as litigants begin their testimony in a careful style that often contains rehearsed elements, but revert to their vernacular when their emotional involvement increases, for example when reacting to their opponent’s conflicting testimony. At the same time, examples of both codeswitching and style-shifting lend themselves to analyses following the audience design model of Bell (1984) and approaches based on speaker design and identity (Coupland 2001, Schilling-Estes 2001). It is shown that the litigants’ use of careful style in L1 or of codeswitching to L2 English can be viewed as accommodation to legal decision makers who generally speak Standard English and use a legal register, but also as speaker-initiated shifts that seek to project a self-image of professionalism and social respectability. The comparison of style-shifting to codeswitching thus relates to questions of speakers' metalinguistic awareness of the "codes" and "styles" in their repertoire, suggesting that switching between languages may be more salient than switching between different varieties of the same language. The paper thus argues for a more central place of codeswitching and language contact studies within variationist sociolinguistics, showing that comparisons of monolingual to bilingual data can yield fruitful results for the field as a whole, where the tendency has been to marginalize bilingual speakers and view language contact as a complicating factor that is best abstracted away from when studying linguistic variation (e.g. Labov 2001: 20).