On the grammar of code-switching

This paper explores an Optimality-Theoretic approach (Prince and Smolensky, 1994) to account for the observed cross-linguistic patterns of code-switching. The proposal is premised on the simple assumption that code-switching strives for well-formedness. I assume that when the embedding constituents are mixed into the matrix language, the syntax operates to optimize well-formedness. I show that optimization of well-formedness in code-switching follows from the (violable) ranked constraints. I will argue that the patterns of code-switching emerge from the interactions among the ranked constraints.