Norms in Social Simulation: Balancing Between Realism and Scalability

Agent-based modelling (ABM) has been used to study the dynamics of complex systems, including human societies. However, the design of such models often fails to capture one of the key features of human behavior: norms. While norms and normative behavior are extensively studied in artificial intelligence (AI), especially in the context of multi-agent systems (MAS), their approaches are often very complex and formalized, going against the prevailing discourse of ABM, which advocates keeping the models as simple as possible and pruning any unnecessary complexity. Nevertheless, norms are relevant and integral to many social contexts, and capturing their effect and dynamics often requires agents that, while not as complex as those developed for AI, are capable of sophisticated cognition. We present a normative architecture that attempts to capture the ways norms affect cognition and behavior, while at the same time being lightweight enough to be suitable for ABM use in simulations.