Abstract
Among the rules which govern the "language-games" discussed by Wittgenstein there are some which seem to have particular functions which can be more effectively brought to light by considering the logical and pragmatic effects of their breakage. Indeed, if we extend progressively the analysis of possible breakages of such rules from particular language-games to broader and broader areas of language, we arrive at a point where (as happened in the Tractatus) it seems possible to draw a limit between what, in general terms, is endowed with sense, and what is devoid of it. This possibility, offered by a "rule-breaking" approach, also opens a promising perspective from which to look afresh at some classical problems connected with skepticism.