In this final chapter let us explore the role of lexical variation in the sociolinguistic history of Paris. We have left this subject till last, not because it is the least important, but because the issues involved are recurrent and are best dealt with together rather than distributed piecemeal across the different periods. To many laypersons, what marks off Parisian speech from other dialects of French is not so much ‘accent’ or ‘grammar’ but its ‘slang vocabulary’. Lexical differences between dialects are highly salient and are readily apparent to all speakers of the varieties concerned, without any linguistic training. Dialectologists and philologists have always been deeply interested in variation in the lexicon, and, in the case of French, they continue to produce exhaustive etymological descriptions of particular items (see von Wartburg 1923–) and impressive inventories of dialect-specific words (see, for example, Rézeau 2001). It is slightly anomalous, therefore, that sociolinguists, in the Labovian tradition at least, should have tended hitherto to keep lexical variation at arm's length.

The reasons for this are comprehensible enough. A variety's lexicon is less tightly structured than its phonetics and grammar: individual lexical items can be modified or exchanged with greater freedom than phonological and grammatical ones, and lexical choices, usually highly conscious, appear to be more random and short-lived. Whereas tokens of particular phonological and morphological variants can occur very frequently in short stretches of discourse, this is not the case with lexical variants, precluding classical Labovian exercises in quantification.