Published online by Cambridge University Press: 30 January 2013
This article outlines the diglossic approach to intra-speaker grammatical variation (Ferguson 1959), wherein speaker-hearers acquire two grammars which are socio-stylistically distinct – one H(igh), the other L(ow) – but linguistically related (to the extent that users regard them as the same language), and then engage one or other of them (but do not mix them) in their active productions. It then sets out how a case could be made for such a model to capture variation in contemporary France, in place of the variationist model which envisages a single, flexible grammar, e.g., the bipolarity, strength and non-random nature of the sociolinguistic H–L distinction, the differing pattern of acquisition of H and L forms, the tendency for L forms to encroach on H terrain (rather than vice versa), and the internal coherence of each of the H and L varieties. Finally, the article sketches the politico-moral dimension to the debate, extending beyond scientific objectivity, and relating to the treatment of non-standard linguistic behaviour in context of the socio-cultural status of the standard.
Nous remercions Fabienne Martin et Andreas Dufter pour leurs remarques précieuses sur une version antérieure de ce texte.