ABSTRACT
Trom the fact that language is a social institution, it follows that linguistics is a social science, and the only variable to which we can turn to account for linguistic change is social change, of which linguistic variations are only consequences. Antoine Meillet cites half a century later by one of the leading methodologists of contemporary American sociolinguistics, William Labov. The alternative approach, the integrational approach, sees language as manifested in a complex of human abilities and activities that are all integrated in social interaction, often intricately so and in such a manner that it makes little sense to segregate the linguistic from the non-linguistic components. For the segregationalist, to speak of language as social interaction is at best a kind of ellipsis, rather like describing alcohol as a social problem. In America, the attempt to integrate linguistics into the general study of communicative behaviour was pursued most systematically by Kenneth Pike.
