Mind and Computation (II)

The paper deals with the fundamental problems of cognitive science, starting from the epistemic and ontological limitations which are immanent to the very attempt to describe mental phenomena in terms of objective (scientific) language, up to the problems of the formal representation of common-sense knowledge. A many-level model of the cognitive system has been proposed; in that context, we analyse the Classical and Connectionist approach to the description of the cognitive system, and we argue that (1) these two approaches should be conceived as two different levels of speech about the same phenomena, and that (2) they face essentially the same basic problems. The second part of the paper discusses various positions concerning the range and limits of artificial intelligence; in that context we put forward the Background and Care hypotheses, both of which call into question the very possibility of the existence of machines with (any) real cognitive abilities. The paper concludes that the requirements which are put before Al should be more realistic (than they usually are) if we are to deal with reasonable research projects