The Syntax of Language Contact

The success of (socio)linguistic theory depends largely on its ability to demonstrate the systematic nature of language variation and use in language-contact situations. Empirical issues of language variation and use, though central to sociolinguistic theorizing, remain marginal in modern syntactic theory because of its insistence on language invariance as the methodological imperative for a scientific study of language. I have resolved this paradox in my research using the insights of Optimality Theory and have developed accounts that capture the logic of variable language use in two empirical domains: intra-sentential code-switching and language variation.

Intra-sentential code-switching is the alternate use of two languages within a clause. The linguistic interest in this verbal strategy is that mixers have clear, unambiguous intuitions about what is, and also what is not, a possible code-switched utterance. Several syntactic categorical constraints were proposed in the past as holding universally across pairs of languages involved in code-switching. However, subsequent empirical studies falsified the universality claims of the proposed constraints. The specific proposal I have been developing has evolved from the important cross-linguistic observation that speakers of languages involved in code-switching have "preferences" for what constitutes a well-formed code-switch. Replacing categorical constraints with "violable" (soft) constraints as specified in Optimality Theory, I developed an account of code-switching premised on the assumption that a linguistic structure (e.g., code-switched constituent) that violates a particular constraint has its well-formedness "reduced" by a certain amount (1996, 1997, to appear). The framework I developed along these lines integrates successfully the empirical generalizations in terms of different syntactic constraints on well-formedness and allows them all to interact to account for the observed range of code-switching data, e.g., in Spanish-English, Kashmiri-English, Hindi-English, and Swahili-English.

Using the Optimality-theoretic insight that a grammar of a linguistic variety is a set of ranked constraints, I was able to provide an account of the theoretically problematic differences between two varieties of Indian English: vernacular Indian English (VIE) and standard Indian English (SIE). The differences between these two grammars are most conspicuous in the syntax of wh-questions, Focus, non-overt subjects/objects, and expletive (it) subjects. My research reveals that while VIE allows subject-verb inversion in indirect questions and forbids it in direct questions, the mirror opposite holds of SIE. VIE also allows non-overt subjects and objects as well as non-overt expletive subjects; these are disallowed in SIE. Finally, in VIE, but not in SIE, object noun phrases that are focused do not appear in positions where they are (Case) licensed. In my analysis (1995, 1997, 2000), I conclude that these grammatical differences arise due to different rankings of the same constraints, a theoretically desirable account.