Bringing speakers back in? Epistemological reflections on speaker-oriented explanations of language change

Abstract

From the late 1980s historical linguists have repeatedly criticised autonomous, system-internal historical linguistics and emphasised the need to incorporate human agency and intentionality into our explanations of language change, a concern most clearly articulated in sociolinguistic approaches to language history. The article evaluates the ways in which speaker agency has been conceptualised in these discussions and critically reviews the explanatory status of traditional ‘belief–desire' models of action in the light of evolutionary, neuropsychological and sociological contributions to the question of human agency. Considering the generally accepted definition of language as a collective structure, the article discusses further whether explanations should invoke individuals (‘methodological individualism') or collectives (‘methodological holism') as the agents of linguistic change. It is suggested that Giddens's theory of structuration, which attempts to overcome the dichotomy of individualism vs. holism and to disassociate agency from individual intentionality, is a useful starting point for future theorising. While speaker-oriented approaches to language change are an important part of the explanatory inventory of historical linguistics, not all linguistic changes are equally amenable to explanations in terms of speaker agency. It is argued that there is a cluster of areas (grammaticalisation and reanalysis, language contact, native theories of language as well as standardisation) which should feature prominently in the field of socio-historical linguistics.