Does Agent-Based Modeling Survive in Sociology? a Theoretical First Step Toward "Sociological" Micro-Macro Links

Monday, 11 July 2016: 09:15
Location: Hörsaal 27 (Main Building)
Oral Presentation
Yoshimichi SATO, Sociology, Tohoku University, Japan
Agent-based modeling, a sophisticated technique in rational choice theory, has been gaining popularity in sociology. It is a strong tool with which to study self-organization in society such as the evolution of cooperation and the emergence of trust between strangers. Thus it has found a niche in sociology. However, it is another story whether it will evolve and eventually prevail in sociology. A major obstacle to its evolution in sociology is sociological concepts that are difficult to be translated into concepts used in agent-based modeling. Take social structure for example, which is one of the main topics in this session. Conventionally, social structure has been defined in two ways in sociology. First, it is equivalent to social network structure. Agent-based modeling is a powerful tool with which to study the evolution of social networks and, therefore, has contributed to the progress of the study of social networks. A second definition of social structure is that it is relations of roles to which actors and resources are allocated. Then it is challenging for agent-based modelers to use the concept of “role” in their models. This is because a role is more complex than an actor or an agent. A role is defined as a bundle of expectations by incumbents of other roles. Thus it is necessary for a modeler to define the relationship between a role and expectations in order to define social structure and study its evolution. I will present a theoretical framework to do that. Linking such concepts as role and social structure conventional in sociology with agent-based models would help agent-based modeling prevail in sociology.