Abstract

The task of this dissertation is to address what I believe is the essential philosophical question in the discussion over pictorial representation: whether pictorial representations, like language, are arbitrarily linked to the world they represent or whether there is some non-arbitrary link. Nelson Goodman in Languages of Art and other works maintains that pictorial representations are arbitrarily linked to the portion of the world they represent, whereas the various forms of the similarity theory maintain that the link is non-arbitrary because pictures are joined to the world by similarity. In chapter I of this dissertation I give an idea of the range of similarity theories and discuss Ernst Gombrich's Art and Illusion. In chapter II Goodman's objections to any version of the similarity theory are explained. After presenting details of Goodman's theory of pictorial representation and portions of his epistemology and ontology that are integral to it , it is shown that Goodman's theory of pictorial representation is not adequate . Although I agree with the orientation of all similarity theories that pictorial representations are not arbitrarily linked to the world and that similarity plays, or may play, some role in linking them, I accept Goodman's arguments that similarity is not a sufficient condition. I do not accept Goodman's argument that similarity is not a necessary condition, although I concede that it is not for reasons of my own . Because of these negative conclusions the task of examining the link between pictorial representations and the portions of the world they represent would not be complete without a proposal for how pictorial representation does work. In the final chapter I propose a new version of the similarity theory which avoids Goodman's criticisms