Computation, behavior, and structure in fixed and growing automata.

The theory of automata is relevant to behavioral science in the sense that it is an absolutely rigorous theory of processes which are themselves absolutely rigorous abstractions of behavior. If we can specify exactly how we wish a system to behave (i.e., what shall happen in response to what), we can in principle construct the system. To what extent, then, can “higher order” systems of this sort be constructed, systems which will construct other systems with prescribed behavior characteristics? In particular, a question which has intrigued automata theoreticians is whether an automaton can be constructed which will reproduce itself and, if so, how complex does it need to be? These questions and related matters are discussed in this article.