Sociolinguistic Method and Linguistic Theory

What we may call ‘the sociolinguistic method’ is neither new to sociolinguistics, nor universally adhered to by sociolinguists, nor-strictly speaking-a method. It is basically a working hypothesis with a distinctive (within linguistics) methodological and concetpual apparatus, built up over the last ten to fifteen years in response to the particular needs of research guided by this hypothesis. The essential point is that the primary data for the study of linguistic structure, function, and change is the spoken language, more specifically spontaneous unreflecting speech in its natural context. This principle is best exemplified by the research of William Labov and associates (LABOV, 1963, 1966, 1972a, 1972b, LABOV et al., 1968, 1972, 1977, 1978), though the present paper is based on the work of the Montreal French Project (e.g. SANKOFF, 1978; THIBAULT, 1979; G. SANKOFF, 1980). We shall trace some of the consequences of the natural speech hypothesis, not only on the methodological level, but also extending into linguistic theory. On the theoretical level we focus on two debates: one about the existence of the ‘syntactic variable’ and the second on the ‘wave’ theory.