Um conceito de informação para a teoria representacional da mente

Abstract

In this thesis, we approach the definition of computation as information processing to check if this definition, although broadly accepted, is an appropriate notion of computation, at least to the cognitive sciences. More precisely, we seek a definition of information as applied to the human mind, i.e., an analysis of what we call mental information. In order to achieve these goals, we will support one of the Computational Theories of Mind, the Representational Theory of Mind, and will offer a concept of information to it from the idea that the bearers of information in mind are mental representations. Moreover, we will make explicit the semantic character of mental information (through a justification of its particular character as information) from its relation to mental representations and will show how, in our view, it is processed. We will also approach some alternative concepts of information and computation, showing in which points we agree and in which points we disagree with them. At last, we will give an answer to one of the main problems put to Computation Theories of Mind: that all computational atributions are trivial (i.e., dependent on an observer or external agent). As we put forward our answer, we will briefly deal with some other problems put to artificial intelligence and cognitive science, such as the Chinese Room argument and the Homunculus problem, and will briefly discuss defenses of semantic requirements to computation and their relation to our defense of a representational requirement to mind.