The computational model of mind has captured the fancy of cognitive scientists, linguists, and theoreticians in artificial intelligence and philosophers of (behavioral) science. Philosophers of science found the computational model particularly appealing since it comports so nicely with the materialistic and mechanistic tradition of natural science dating from the time of Hobbes, Newton, and Descartes. Cartesian materialism, however, was not thoroughgoing, acknowledging a nonmaterial substance that interacts causally, and thus problematically, with matter. The computer model of mind seemed to its advocates to resolve the problems of causal interactionism. 1 They argued that likening the mind to a computer's software and the body to its hardware made possible an explanatory account of the relation of the mental to the material that was fully compatible with a thoroughgoing materialism.
Right from the start this view had its detractors who vehemently objected to the very idea of human mentality as being anything at all like computation or calculation. They argued that human mentality may well be realized in human cognitive processes and capacities but none of these is computer-like in any important way. It seems that the detractors have persevered not so much from the force of their objections but from the failure of the computer model to solve many of the problems that it set for itself. In Kuhnian fashion, as the paradigm atrophied (or seemed to in certain areas), the objections gained in strength. But worthy paradigms offer strong resistance, and change slowly, and die hard, for besides the areas of vulnerability they have their distinct areas of potency.
Two recent defenses of the model have appeared that notwithstanding their common purpose are very different in almost all other respects. Barbara Von Eckardt's What is Cognitive Science? 3 is a comprehensive, detailed, and technical rational reconstruction of computational cognitive science. Jerry Fodor's The Elm and the Expert 4 is a slim volume comprising the 1993 Jean Nicod Lectures he delivered in Paris, France. The essays exhibit the typically Fodorian flamboyance but beneath the pizzazz is a stout defense of a thesis under siege. We should expect no less, since Fodor's contribution to the computational model is both foundational and seminal.
Von Eckardt's defense of computational cognitive science is predicated on its "rational pursuability" as an "immature" science. The immaturity of cognitive science is relative to historically successful sciences such as physics. To this end she develops a Kuhnian analysis of the unifying characteristics of the cognitive science community vis-a-vis its diverse research activities. She defines her approach in terms of a "framework of shared characteristics" (FSC), proceeding from actual theoretical and research practices that fall under the rubric "cognitive science." She wants to show the rational justification for the substantive and methodological assumptions that human cognitive capacities are computational and representational. These two controversial assumptions are among those comprising the FSCs that bind cognitive scientists into a community. Her highly detailed account traverses the gamut of issues and conflicts that have occupied cognitive scientists and philosophers of cognitive science for the last thirty years. Since the scope of her discussion allows attention to several key points only, I look at those issues that overlap with Fodor's discussion.