Codeswitching at the interface of language-specific lexicons and the computational system

Abstract

It has been noted by many researchers that young bilingual children pass through a stage of early mixing which extends approximately until the age of 2;6 and ends abruptly. Research on bilingual first language acquisition has clearly excluded the possibility to explain mixed utterances as the result of a fused lexical or grammatical system. However, the actual debate on the reasons for early mixing still continues. Two main approaches have dominated the field of language mixing in adults: One assumes that adult's codeswitching is constrained by grammatical principles, suggesting that a third grammar is responsible for the grammaticality of mixed utterances (e.g., the Equivalence Constraint formulated by Poplack (1980), or the Functional Head Constraint formulated by Belazi, Rubin & Toribio (1994) among others). Since child grammar is supposed to be organized differently from adult grammar, the grammatical “ingredients” of the principles which constrain codeswitching are absent in early child language (Köppe & Meisel, 1995). It therefore follows that early mixing in young bilingual children is not to be considered as constrained by any grammatical principles. The other approach argues that codeswitching in adults is not regulated by external grammatical principles, but that the only constraints which govern codeswitching are those required by the two languages involved (MacSwan, 2000). We will show that this assumption holds for child language as well. Our analysis of mixing opens the perspective that child grammar can be considered to be organized in the same way as adult grammar. Furthermore, we will argue that early mixing is related to developing performance abilities, in the present paper to the readiness on the bilingual child's part to speak the language(s).