It is argued that the traditional distinction between artificial intelligence and cognitive simulation amounts to little more than a difference in style of research - a different ordering in goal priorities and different methodological allegiances. Both enterprises are constrained by empirical considerations and both are directed at understanding classes of tasks that are defined by essentially psychological criteria. Because of the different ordering of priorities, however, they occasionally take somewhat different stands on such issues as the power/generality trade-off and on the relevance of the sort of data collected in experimental psychology laboratories.
Computational systems are more than a tool for checking the consistency and completeness of theoretical ideas. They are ways of empirically exploring the adequacy of methods and of discovering task demands. For psychologists, computational systems should be viewed as functional models quite independent of (and likely not reducible to) neurophysiological systems, and cast at a level of abstraction appropriate for capturing cognitive generalizations. As model objects, however, they do present a serious problem of interpretation and communication since the task of extracting the relevant theoretical principles from a large complex program may be formidable.
Methodologies for validating computer programs as cognitive models are briefly described. These may be classified as intermediate state, relative complexity, and component analysis methods. Compared with the constraints imposed by criteria such as sufficiency, breadth, and extendability, these experimentally based methods are relatively weak and may be most useful after some top-down progress is made in the understanding of methods sufficient for relevant tasks - such as may be forthcoming from artificial intelligence research.