Multi-Agent Systems, Time Geography, and Microsimulations

Conclusion We have argued that time geography provides a perspective that helps unify the two paradigms of (a) multi-agent systems, as developed within computer science, and (b) microsimulations, as developed within the social sciences. By identifying and defining these two paradigms, and by reasoning about the central concepts of each of them, we have taken a first step in amalgamating them. We have attempted to take a general systems approach in order to avoid myopia and jargon limitations, and hopefully avoid being too narrow in scope (an approach different from, e. g., Gimblett, 2002). Our claim is that developments based on a synthesis of the three paradigms offer a rich potential for substantial advance of systems analysis methodology. It gives a new angle to classical problems like how to achieve consistency with the world outside a defined core system boundary, how to simultaneously represent processes on very different spatial and temporal scales, how to enable agents to concurrently obey internal and external rules, and how to integrate observable and postulated behavior while preserving achievability of endogenous emergence.