According to a venerable tradition in philosophy and linguistics, expressions have meaning through being subject to conventions or rules. This claim has become a central topic of contemporary philosophy of language and mind in the wake of Wittgenstein and Kripke, largely because the normativity of meaning is regarded as a serious challenge to naturalism. One reaction to this challenge is to deny that the normativity of meaning is genuine. While there are ‘semantic principles’ specifying conditions for the correct application of (...) expressions, these are either not genuinely normative or they are not in fact constitutive of meaning. But this dilemma can be defused if one distinguishes different dimensions of semantic normativity and pays attention to different types of mistakes that can afflict linguistic behavior. In particular, one needs to keep apart norms of truth and norms of meaning, regulative, and constitutive rules, and to appreciate the normative dimensions of the notion of correctness. The final section discusses the role of communally shared rules or conventions for language, in particular the age-old problem of circularity: how could such conventions be essential to language, given that the latter appears prerequisite for establishing and communicating conventions in the first place. (shrink)