Accounting for constitutive rules in game theory

Game theory and rules are deeply intertwined for at least two reasons: first, in many cases rules are necessary to break the indeterminacy that surrounds most of the games; second, in the past 30 years game theory has been increasingly used as a major tool to build a theory of social rules. Interestingly, though the concept of rules is now part of most game theorists' tool box, none of them has explicitly entertained the important distinction between regulative rules and constitutive rules. This distinction, which finds its roots in Ludwig Wittgenstein's profound discussion of ‘language games', is at the core of modern philosophy of social sciences. This article asks whether game theory can account for constitutive rules. I distinguish between three game-theoretic accounts of rules: the rules-as-behavioral-regularities account, the rules-as-normative-expectations account, and the rules-as-correlated-devices account. I show that only the latter two are able to make sense of constitutive rules by giving up any pure individualistic understanding of game theory.