A Study on the Cognition of Nature in Connectionism

Abstract

The following two points were thought to be necessary for research on the natural recognition to be useful for science education: clarifying the cause or reason why the characteristics of natural recognition were formed and showing the ideas necessary to arrange each research project. The object of this study was to propose a theory of Connectionism, that is, parallel distributed processing, in order to resolve these points. The following points could be explained by the theory of Connectionism: (1) the mechanism of production of framework such as prototype and theory-ladeness (2) the mechanism of flexibility of schema to conditions (3) the reason why mental models and metaphors were useful for recognition (4) the reason why logical thinking was hard to develop in a short time (5) the reason why naïve concepts were held to and difficult to change (6) the ideas that related research on situated cognition to research on internal activities Furthermore, it could be shown that Connectionism proposed a single idea to arrange each research project through the neural network models.