Computation is a central concept in computer science. The mathematical theory of computation, with its historical origins that predate the computer, has been most influential in the development of programming languages, tools and methodologies. The study of psychological aspects of computation is by comparison a new concern that has developed largely as a post hoc investigation of contemporary methods of programming; it has yet to make a serious impact upon computer science theory. In this paper, we shall argue that concerns about perception and interpretation that are significant in psychology should be a primary source of insight into the new models of computation that will shape future developments in computer programming. To this end, we reconsider fundamental issues concerning the nature of computation and its relationship to mathematics and cognition. In particular, we shall explain how computation has to be understood with reference to
- physical processes that are perceived to undergo changes of state
- conceptual methods for interpreting state changes in computational terms.
Brief references to our work on agent-oriented programming using definition-based representations of state will be used to explain our thesis.