In this paper, I consider the use of corpora in sociolinguistic research and, more broadly, the relationships between corpus linguistics and sociolinguistics. I consider the distinction between "conventional" and "unconventional" corpora (Beal et al. 2007a, b) and assess why conventional corpora have not had more traction in sociolinguistics. I then discuss the potential utility of corpora for sociolinguistic study in terms of the recent trajectory of sociolinguistic research interests (Eckert under review), acknowledging that, while many sociolinguists are increasingly using more advanced corpus-based techniques, many are, at the same time, moving away from corpus-like studies. I suggest two primary areas where corpus developers, both sociolinguistic and non-, could focus to develop more useful corpora: Corpora containing a wider range of non-standard (spoken) varieties and more flexible annotation and treatment of spoken language data.
Sociolinguistics; conventional and unconventional corpora; spoken language corpora; data management; annotation methods
