Open Access

Françoise Gadet

• This paper looks at some questions which were considered quite differently before corpora became the ordinary way to describe languages, focusing on the following points: a) Our ultimate objective is to document how wide-reaching the appellation “French” can be, given the extent of variation found in the corpora: is it possible to document the whole variational span of “the French language”, in what Chaudenson (2003: 182) would call “the limits of intra-linguistic variability of French”? b) Are there grammatical phenomena which could be looked at differently and analysed using corpora? c) Is it possible to generalise in an explanatory perspective, and to determine something of the principles which lie behind the difference between standards and vernaculars? d) The discussion of ordinary and non-standard data, mostly spoken (only occasionally written) will lead me to consider if it is possible to qualify vernacular varieties as such, in what Chambers called in 2000 “universal sources of the vernacular” and in 2003 “vernacular roots”; and what I shall choose to call here “vernacular re- source” (see the concluding remarks in section 3). The linguistic variation data which will be looked at in sections 1 and 2 are primarily diatopic, and occasionally diastratic: they mostly come from geographically “periphercal” French (mostly North American) and socially “marginal” French, which means here ways of speaking which have been subjected to / been the target of (albeit sometimes in a limited way) normative pressures, as is the case for ordinary spoken French, “popular French”, youth language, child language, and different types of urban or rural vernaculars.