A survey on the nature, reasons for compliance and emergence of social norms

The aim of the paper is to offer a critical review of some of the most important contributions on the subject of social norms. The analysis will hinge upon a fundamental distinction between self-regarding and other-regarding reasons to action, which are supposed to represent two basic types of motivations making up the individual system of rational deliberation. Such a dichotomy will make it possible to divide the account of rules of behaviour in three categories, depending on the weight accorded to one rather than the other reason to action. The narrower concept is given by what will be called strictly-conceived convention, which coincides with Lewis's classical account of a convention. In this case, it is the self-regarding motive that actually provides an underpinning of such a regularity of behaviour, making it possible to build a system of convergent mutual expectations. The next category, that of broadly-conceived conventions are based on Sugden's earliest works on the subject; in this case, the self-regarding motive is still the crucial one, but conventions are not necessarily mutually beneficial, thus coming down to a standard Nash equilibrium, or evolutionary stable strategy, notion. Finally, with the final category of norms a decisive shift out of the self-interested justification is accomplished. In fact, these regularities are grounded on some forms of other-regarding behaviour, as Sudgen's model of normative expectations sets out clearly. The cognitive structure needed in order to generate such type of expectations leading to norms is then critically examined, reaching the conclusion that the account of norms provided in the received theory does not prove thoroughly satisfactory. Particularly, the notion of "empirical" expectation, as opposed to that of "causal" expectation, is deemed as the relevant one in order to build a system of mutual expectation in the case of norms; however, this concept does not help to explain how a norm comes out as an equilibrium of the social interaction, thus making the whole explanation arguably circular. We finally argue for the importance to ground the concept of norm on the dynamic evolution of expectations.