Uncommonly Taught Languages

This study focuses on the morphosyntactic aspects of Cantonese-English code-mixing as commonly spoken by bilinguals, most of whom are Cantonese. A corpus of Cantonese-English code-switching collected from informal conversations is analyzed in terms of structural properties, followed by a critique of the major constraints or principles that have been proposed in the literature. It is suggested that none of these constraints or principles are descriptively adequate in view of the data collected, and three alternative constraints--the category equivalence constraint, the bound morpheme constraint, and the specifier constraint are proposed. The possible implications of the constraints on bilingual processing are also discussed. (Author/JL) * Reproductions supplied by EDRS are the best thRt can be made * * from the original document.