Since Labov's initial demonstration of the .. limits" of individual gram- mars. it has been generally accepted (at least by sociolinguists) that pan- dialectal competence, active or passive, is extremely sharply constrained. However, there is an implicit tendency to assume that it is competence in non-standard grammars which is limited. It is significant, for example, and quite characteristic of work in this area, that all the sentences discussed in TrudgiH's 1981 paper are designed to test knowledge of non-standard grammars. In an earlier publication, Trudgill expressed the view more explicitly that non-standard speakers had at least a passive competence in standard English (Trudgill, 1979) and Trudgill's view does not seem at all uncommon.