When Penelope Eckert and Sally McConnell-Ginet call for grounding language and gender research in ‘local communities of practice’ (1992), their challenge is to move away from global generalizations and stereotyped conclusions about language and about gender. Actual instances of language use must be considered as part of a complex of interlinking cultural, social, political, psychological and linguistic systems. Earlier analytic categories, such as the unified, dichotomous notion of ‘gender’ and theoretical frameworks like the ‘difference’ and ‘dominance’ models (Thorne and Henley 1975, Thorne et al. 1983), have proved to be rigid and unyielding; such formulations are unable to account for the vast array of phenomena that occur when people, in all their diverse and fuzzy categorizations and situated in particular political climates, speak to each other. We need to develop a methodological and theoretical approach to issues of language and gender that allows for the integration of varied, complex, even disparate information.