We are the funny organisms that make and follow rules. To understand us, one must understand what is it to institute and follow a rule, to perform correctly or in error. This question is more important than it might at first seem for linguistic meaning is (partly) constituted by rules that govern uses of expressions. For example, the fact that 'squid' is correctly applied to squid and incorrectly applied to cuttlefish is part of what makes 'squid' mean what it does. Philosophers sometimes argue further that all representations are rule-governed so that a theory of representation must explain the possibility of misrepresentation. The rule-following problem makes an explanation of correctness surprisingly difficult. Philosophers have also not been clear on just what is the problem with rules. I provide a statement of the features of rules and the problems they drive. I then show that some well-known accounts of rule following---Kripke's, Wright's, Blackburn's---are inadequate. My formulation also shows that some favored theories of content---causal, evolutionary, interpretivist---are inadequate. Positively, I argue (with Brandom) that an essentially social and irreducibly normative organization---a practice---solves the rule-following problem. Therefore, only participants in social practices can follow rules and thence traffic in linguistic meaning. Still, an account of the correctness of a linguistic performance is not yet an explanation of its representational content. I argue that linguistic content is constituted by correctness facts but also by the representational purport of expressions. I explain an expression's purport to represent naturalistically in terms of syndromes of (non-representational) psychological and behavioral symptoms. A theory of linguistic representation, then, must join the non-natural social facts that constitute correctness with the natural individualistic facts that constitute representational purport. Such a hybrid theory reveals that thought and language are not as intimately connected as philosophers assume. Whereas linguistic content constitutively partakes of the social organization that sets humans apart from other creatures, thought content involves only refinements of the psycho-behavioral syndromes that we share with most vertebrates. It follows that linguistic content and thought content cannot be explained one in terms of the other.