ABSTRACT
The linguistic performance of bilinguals has been used to support syntactic theories — for example, Woolford's study of government and binding and code-switching and White's research into the relationship between Universal Grammar and second language acquisition — but there are no theories about the bilingual speaker that aim at a description of the entire language production process. This chapter gives a brief and global description of Levelt’s model, subsequently and considers how such a model should be adapted to make it suitable to describe the bilingual speaker, and finally some alternatives are presented for parts of the model. Levelt's model aims at describing the normal, spontaneous language production of adults. It is a ‘steady-state’ model, and not a language learning model, and it hardly says anything about language perception. The first-language system is flexible enough to add an additional register to those already in existence.
