An examination of the thesis of models as representations

Abstract

This paper aims to discuss four versions of the thesis of models as representations that are used to deal with the problem of scientific representation: models as structures, analogs, fictions, and mental representations. In particular, the paper focuses on an examination of the problems for the thesis of models as structure and shows that (i) structure cannot be viewed as the essence of models; (ii) isomorphism cannot define a representational relation; and (iii) models involve linguistic descriptions instead of pure abstract mathematical entities. Based on the conception of models as mental representations, the paper suggests a naturalist approach to scientific representation and a reduction of the problem of scientific representation into the problem of mental representation, by which the representational role of models in science may be explained by means of the representational function of mental representations.