If laws are just made up, then why do we have to follow them? What normative force do they have? It has seemed to most philosophers of law for the last two centuries that law either has moral or prudential normativity, or that it has no normativity at all. As a result, the normativity of law has seemed to be a serious obstacle for theories that attempt to explain law as a social phenomenon, explicable with descriptive resources. But this obstacle can be overcome. The central claim of my dissertation is that law—and other normative practices, such as language and games—are normative in an alternative, non-moral and non-prudential, fashion. As a result, it is possible to explain this normative practices by appeal to descriptive states of affairs.

This possibility has been ignored because philosophers have often conflated two varieties of normativity. One sense in which a practice might be normative is that it has weight in practical deliberation. That is, it generates considerations suited, by their nature, to be included in deliberation about what to do. But a different sense in which a practice might be normative, which I argue still constitutes a form of normativity, is merely that it consist of rules. To get a sense of this variety of normativity, consider the difference between the fact that (a) cereal is eaten with milk and the fact that (b) cereal is eaten with a spoon. The first is a regularity. Eating cereal dry is unusual. The second, by contrast, is a rule (of table manners). Scooping cereal with one’s hand is not just unusual; it is forbidden. This is the hallmark of a rule: it entails that an evaluative or deontic concept, like impermissible or impolite, applies to instances of behavior. Practices like law, language, and games are normative in the sense that they consist of rules, but not in the sense that they necessarily generate reasons with deliberative weight.

The dissertation develops an account of this rule-constituted variety of normativity and applies it law and language. It does this in three parts. The first part outlines the general view. To avoid the complexity and controversy of law and language, the focus here is on a simpler normative practice—table manners—which is taken as a kind of case study. The second part applies the view to law, defending a version of legal positivism that many take to have been decisively refuted. Since law does not necessarily generate reasons, it can be reduced to social facts—facts about what people do and say and think. Moreover, in contrast to prevailing wisdom, the existence of a legal system does not require either officials or ordinary citizens to judge any laws to be morally acceptable.

The third part takes up the issue of the normativity of meaning. In the last half century, there have been many attempts—under the labels of “causal” or “informational” theories of meaning—to reduce linguistic meaning and mental content to descriptive states of affairs. Though many of these accounts fail for idiosyncratic reasons, there is a general question as to whether they are doomed from the start because meaning facts are normative and therefore irreducible. But I argue that meaning is normative only in the sense of being rule-constituted, and that this transforms debates about the nature of meaning and content.