Socially meaningful syntactic variation in sign-based grammar

This study begins from two observations. The first is that linguistic variation is socially meaningful (see Labov 1963, Ochs 1992, California Style Collective 1993, Eckert 2000, Irvine 2001, Podesva et al. 2001, Bucholtz and Hall 2005, Campbell-Kibler 2006, inter alia). The second is that speakers use their knowledge of language (i.e., linguistic competence) to engage in various linguistic activities (i.e., linguistic performance) and therefore in modeling linguistic competence we are well-served to strive for models which are ‘performance-plausible’ (Bresnan 1978, Pollard and Sag 1994). Among other things, this means that, if sociolinguistic meaning-making is bound up with syntactic constraints, then our model of (syntactic) linguistic competence must interface smoothly with models of knowledge (and use) of social meaning, and that systems which look optimal under an artificial modularization might fare less well when viewed against this broader scope of data. As we extend models of syntactic competence to encompass sociolinguistic variation, we must account for the following types of data: