Customs as conventional rules
The formation process of customary law has frequently been analysed from the positive perspective of the description of some particular types of societies or practices. International trade is frequently quoted as a significant example (Benson, 1998a, 1998b).
Primitive (Landa, 1983; Benson, 1991) or medieval (Greif, 1989; Milgrom et al., 1990, Greif et al., 1994) societies also provide evidence that help us to understand how customary rules emerge. These various customary orders interestingly reveal how customs are conventional rules: they emerge as the result of a spontaneous and decentralized process which functions like a market process. Customs facilitate coordination in allowing the harmonization and stabilization of mutual expectations about each other’s behaviour. Emerging rules have a crucial role when a player wants to assess other players’ preferences. They reduce the number of required interactions to a relatively low and manageable number of instances. However, such rules are not supposed to emerge in every kind of situation. Not only does the adoption of rules of conduct require a cooperative environment because individuals have to agree to interact or to initiate interactions, but also recognition and compliance with rules are usually unlikely if individuals engage in discrete interactions such as one-shot games.The first condition relates to the motivations of individuals who initiate interactions in a decentralized context. The players must be assumed to display a certain willingness to cooperate and, simultaneously, they must be able to identify those individuals who are characterized by the same feature. To get acquainted with others’ characteristics, sympathy is of decisive importance. Sympathy can be defined as a principle of communication which allows individuals to feel - rather than understand - the way others behave.
It implies not only that individuals have access to a knowledge they have not directly experienced but also that, as long as individuals sympathize with each other, they are capable of imagining themselves as being another person: when I sympathize, ‘I consider what I should suffer if I was really you, and I not only change circumstances with you, but I change persons and characters’ (Smith, 1976, p. 317). In this context, inclination towards others replaces self-interested calculation. Therefore, in a society where individuals feel sympathy with others, social order is the product of a consensus which is grounded on secured expectations. It clearly indicates that interactions do not take place in an environment of conflict that could be represented by a prisoner’s dilemma, but rather take the form of a coordination game. The situation of reference for understanding the emergence of customary rules is not that of the ‘war against all’ that Thomas Hobbes viewed as the manifestation of the state of nature, but that of a Humean, peaceful and already structured society.It is thus a major assumption put forward by the defenders of the spontaneous order tradition, that the ex ante existence of some willingness to cooperate - for instance, under the form of sympathy - is a prerequisite for the emergence of customary rules. Once this first condition is satisfied, individuals can anticipate a secured repetition of interactions. Therefore, conventional customary rules may come into effect under the second condition that interactions be repeated. The process can be summarized as follows: customs emerge from unorganized interactions when a regularity of behaviour among a fixed number of individuals occurs in a recurrent situation. When individuals face these recurrent situations, trial and elimination of errors allow the emergence of patterns of actions which are progressively incorporated into customary rules. Thus, customs endogenously but spontaneously emerge from successful practices and attempts to solve recurrent problems.
From this perspective, Vanberg (1989) rightly insists on the unintended nature of customs: while no one can be identified as their explicit creator, everyone uses them and it is publicly known that everybody does so. In other words, following the definition given by Lewis (1969), customs are common knowledge rules. An organizational benefit is identified here, on the one hand, because of the limitation in the range of circumstances everyone has to pay attention to and, on the other, because of more secured interactions. Indeed, customs will not be challenged by players because, and as long as, everyone follows such rules of action and believes that the others do the same. Thus, successful coordination does not require rules to be explicitly known and discussed but solely to exist. No formal or explicit communication between the individuals about the positive meaning of the rule or about mutual intentions is required in so far as human actions spontaneously produce their own coordination.Such coordination tells us about the impossibility of human beings explicitly knowing the sense customs convey. Customs acknowledge the irremediable ignorance of men, which is emphasized by Friedrich von Hayek but also dates back to the theories of the Scottish founding fathers of political economy, David Hume and Adam Smith. In conflict with Cartesian rationality, the Scottish Enlightenment sensualism proposes a theory of human nature in which men perceive the world through their senses rather than constructing it by means of reason. Knowledge is then viewed as the result of accumulation and association of impressions. Men are thus shaped by what they experience. Of course, since the domain of individual experience is limited, induction plays a major role in the accumulation of knowledge by requiring participation in various repeated interactions. Therefore, the process through which customs emerge rests upon an inductive accumulation of knowledge: repeated interactions and frequent non-formal communication between individuals lead to the emergence of a common tacit knowledge.