The Issue
The problem social scientists still revolve around is how autonomous systems, like living beings, perform positive behaviors toward one another and comply with existing norms, especially since self-regarding agents are much better-off than other- regarding agents at within-group competition. Since Durkheim, the key to solving the puzzle is found in the theory of internalization of norms (Mead, 1963; Parsons, 1967; Grusec & Kuczynski, 1997; Gintis, 2003).
Norm internalization is one of the red herrings running across all of the social- behavioral disciplines, and there are not many. Not only sociologists but also developmental, social and cognitive psychologists have perceived its crucial role in socialization. Drawing on the early work by Vygotzky (published in the US as late as 1978) and Piaget (1978), psychologists showed that a parental attitude oriented to elicit norm internalization predicts children’s later wellbeing and even their inclination to other-regarding behavior (Ryan & Deci, 2000).
Nonetheless, our scientific definition and understanding of the process of norm internalization is still fragmentary and insufficient. The main purpose of this paper is to argue for the necessity of a rich cognitive modelling of norm internalization in order to (a) provide a unifying view of the phenomenon, accounting for the features it shares with related phenomena (e.g., robust conformity as in automatic behavior) and the specific properties that keep it distinct from them (autonomy); (b) model the process of internalization, i.e. its proximate causes (as compared to the distal, evolutionary, ones; see Gintis, 2003, 2004); (c) characterize it as a progressive process, occurring at various levels of depth and giving rise to more or less robust compliance; and finally (d) allow for flexible conformity, enabling agents to retrieve full control (Bargh et al., 2001) over norms which have been converted into automatic behavioral responses (Epstein, 2007). Thanks to such a model, it will be possible to adapt existing agent architectures (such as EMIL-A, cf. Andrighetto et al., 2007) and simulation platforms (EMIL-S, see Troitzsch, 2008) to test hypotheses concerning (a) individual and social effects of internalization, (b) factors favoring or hindering internalization, and (c) the evolution of internalization in future societies.
Throughout the paper, the process of norm internalization is meant as a mental process that takes (social) norms as inputs and gives new goals of the internalizing agent (from now on, the internalizer) as outputs. Emotions, playing a significant but not necessary role in this process, will not be investigated at this stage.