Simulation & Norms in social science

When it comes to norms in the social sciences, the problem can be called the Paradox of Agency. It is the problem how to take norms, or social structure, seriously whilst allowing for agent autonomy. The current agent-orientated solution is often rational choice theory. The methodology of agent-based simulation is a possibility to approach the paradox differently.  

  • Simulation can take emergence seriously, no longer making it something like a ‘black box’ carried through theory.
  • Simulation can take structure seriously at the same time as autonomy as feedback between structure and agency can be modelled (EMIL).
  • Simulation can implement kinds of reasoning other than rationality to inform choice and guide behaviour.

In criminology the structure-agency dichotomy is particularly strong and has important policy origins and consequences. Theories of crime strongly shape the criminal justice system at the same time as political ideology shapes prevalence of certain theories of crime. Currently crime simulations are either very strongly structural, simulating towns with opportunities for crime etc. or they are rational choice based.

I argue that the analysis of crime needs another kind of reasoning, as cost/benefit analysis does not suffice to explain most crime phenomena. I call this ‘other kind’ moral reasoning. In the talk I will discuss possibilities for what moral reasoning might be in this context and how it can be formalised and implemented in simulation.

Thursday, 7 May, 2009 - 12:30 to 14:00
Corinna Elsenbroich (CRESS)