ABSTRACT
This chapter examines the empirical basis for the claims and argues that they are questionable because of serious methodological shortcomings in the research. It offers some tentative evidence based on speech perception studies and re-analyses of selected bilingual case studies that young bilingual children are psycholinguistically able to differentiate two languages from the earliest stages of bilingual development and that they can use their two languages in functionally differentiated ways, thereby providing evidence of differentiated underlying language systems. The majority of empirical investigations of bilingual development have found mixing. Phonological, lexical, phrasal, morphological, syntactic, semantic and pragmatic mixing have all been reported. A number of other more specific explanations of bilingual mixing have been suggested. By far the most frequent of these is that bilingual children mix because they lack appropriate lexical items in one language but have them in the other language and, effectively, they borrow from one language for use in the other.
