This note focuses on the interaction between logic and grammar and discusses whether logical rules are irrelevant for syntactic formation. Classical evidence for the independence of syntax from logic comes from the acceptability of analytic propositions. Based on this piece of evidence, various authors in traditional discussions pronounced that language is illogical. This note focuses instead on two ways to oppose such pronouncements. In Gramscian linguistics, one can maintain that language is not illogical by broadening the notion of language one (...) adopts. In an alternative account we submit, one can oppose those pronouncements by restricting the notion of analyticity which is assumed to be relevant for syntactic formation. (shrink)