The study of linguistic variation as the creative construction of meaning almost inevitably leads to the questioning of some nuclear principles and methods of traditional variationist sociolinguistics – and the label ‘traditional’ should not be taken to mean for its views to have been significantly modified over time. In fact, the theoretical nucleus of the paradigm has remained virtually unaltered to this day, at least as regards the status granted to meaning as a component of variation. Meaning equivalence or identity between linguistic variants is still usually held to be a prerequisite for scientific analyses to be feasible; such identity is argued for even in cases when it proves scarcely …