The Dynamic Evolution of Cooperation amongst spatially-distributed agents

The aim of this paper is to present a model of how systems can organise into multi-celled communities based around the biological concept of mutualism (Boucher, 1985). The resulting model will be shown to be applicable to understanding phenomena in a variety of domains including economics, sociology, and other environments where simple units are subject to local and global forces. The paper questions the relevance of the Prisoner’s Dilemma in understanding the behaviour of these systems, and speculates that swarm-like models may actually comprise an equivalence class. It is suggested that a movement to models more like those traditionally defined in population biology may well provide fruitful insights into the construction of multiple adaptive systems. Target areas: Parallel models, Collective behaviour, Biological models, Cooperation in Multi-agent systems