Researcher: Corinna Elsenbroich
Different social sciences understand and measure norms in different ways. How can these understandings be integrated to produce effective theories?
A distinctive element of sociality extending across the fields of economics, psychology, sociology, anthropology, criminology and political science is the fact that social behaviour is shaped by normative judgements. From the perspective of moral philosophy, it is widely recognised that the normative domain is independent in a distinctive way from the domain of facts.
Although social science has attempted to recognise this distinction in a variety of ways, existing theory building tools have not really provided an effective framework for trying to reconcile the cognitive, emotional and behavioural dimensions of normative action. This failure can be recognised in a number of approaches to norms across the social sciences.
In economics, for example, it has been argued that professed norms actually reduce to more sophisticated forms of self-interest or that the term norm should simply be used to refer to behavioural regularities with no "normative" content. By contrast, anthropologists and sociologists have often been criticised for treating social actors as "cultural dopes" completely determined by normative (or other) structures and incapable of exercising choice. Neither extreme position sits well with our everyday experience of norms as significant but nonetheless resistable "social facts".
The problem for social science is thus to develop an integrated model of cognition, emotion and behaviour which adequately captures what we mean by normative judgements (as well as their creation and maintenance) but is neither naively reductive nor implausibly determinist.
This model needs to display the same kinds of normative behaviours that actors display in real social contexts with a normative dimension (moral dilemmas, labelling effects, social reproduction of normative systems despite violations, selective disapproval and so on). In particular, the methodological challenge for this kind of simulation will be to integrate cognitive data (acquired by self-report as well as by experiment) and behavioural data.
The same synthetic approach will be used in this strand as in the repeated interaction strand but to meet a different methodological challenge for simulation. There is already a wide range of approaches and research across the social sciences (e.g. in social psychology: Biddle et al. 1987, Christensen et al. 2004, Sripada and Stich 2006, Smith and Terry 2003, Kallgren et al. 2000, Jetten et al. 2002, Cialdini and Goldstein 2004; in sociology: Opp 2002, Morris 1956, Lapinski and Rimal 2005, von Wright 1963, Gouldner 1960, Healy 1998; in philosophy: Bicchieri 1990, Coyle, 2002 and in social science simulation studies: von Scheve et al. 2006, Castelfranchi et al. 1998, Delgado 2001, Saam and Harrer 1999) which can be used as a starting point.