It is difficult to assess the place of “style” in sociolinguistics. On the one hand, style is everything and everywhere – to the extent that we define styles as context-related varieties, and contextuality as the rationale for sociolinguistics. At this level of generalization, it would seem futile to try to theorize style, since a theory of style would be a theory of everything. On the other hand, style was operationalized as a single quantifiable dimension of sociolinguistic variation in Labovian surveys, and it is still with this focus that sociolinguists tend to address the issue of stylistic variation. From this standpoint, style may not have appeared to merit theorizing; it was (and for many still is) a patterning principle in numerical arrays, an axis on a graph. Sociolinguists have found the consequences of stylistic mapping to be informative, but style itself has needed no more explanatory effort than, at one time, did class or sex or age, as correlates of or as supposed determinants of language variation.

Sociolinguistic critiques of the Labovian conception of style surfaced in the 1970s and 1980s (e.g. Coupland 1980, Cheshire 1982, Bell 1984, Milroy 1987), in many cases voicing dissatisfaction with Labov's claim that stylistic variation was organized according to the degree of attention speakers pay to their own speech. Others wanted to establish social psychological and in particular motivational processes as a basis for understanding style-shifting (Giles 1970, Coupland 1984, 1985).