We propose an abstract mathematical model of space and time within which to study agents, multi-agent systems and their environments. The model is unusual in three ways: an attempt is made to reduce the structure and behaviour of agents and their environment to the properties of the "matter" of which they are composed, a "block time" perspective is taken rather than a "past/present/future" perspective, and the emphasis is placed on discovering agents within the model, rather than on designing

more »

... ents into it. The model is developed in a little semi-formal detail, some relevant experimental computational results are reported, and questions prompted by the model are discussed. • homogeneous in space i.e. the rules are the same for all S-locations • Markovian in the sense that the LHS and RHS of the rules refer only to successive STaM states. The example rule just given is Markovian. Different rules sets for the same STaM may differ in these properties.