O problema da representação mental como covariação

• Ano de defesa: 2023
• Autor(a) principal: Lima, Wisley Sousa
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Representação mental; Covariação; Representação simulada; Semântica interpretativa
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/74006

From the modern era, covariationist theories of mental representation have become more accepted in comparison to similarity theories because covariance, by asserting that the semantic content of mental representations is determined by the causal links between these representations and the world, solves some typical problems of similarity theories, such as the problem of abstraction. However, covariationist theories present serious difficulties in explaining how representation errors are possible. Moreover, the means to address this problem are limited, being restricted to stipulations about ideal conditions for cognitive functioning or ideal circumstances of perception. Robert Cummins argues that covariationist theories share the characteristic of being describable by a computational metaphor, yet they are unable to satisfactorily define representation within the Computational Theory of Mind. Cummins proposes abandoning covariationism in favor of an interpretive semantics that requires simulated representations to underpin representation. However, an analysis of the debate between Cummins and covariationist theories by John Locke and Jerry Fodor indicates that Cummins' criticisms of covariationism are sometimes inadequate or flawed and do not definitively settle the debate. Furthermore, his interpretive semantics poses problems in connecting the mechanics of simulated representation with the intentionality required by mental representations. These factors lead to the conclusion that both Cummins' critique is not pertinent and his proposed solution, although interesting, fails to conclusively address the debate on mental representation satisfactorily.