Sense-impressions: A New Model

Abstract

It seems to me that there are a large number of philosophers who believe that to admit that perception essentially involves the occurrence of sense-impressions in the sense of visual, tactual, aural, olfactory or gustatory sensations is to be committed to either a representative realist or phenomenalist theory of perception. From this point paths may diverge: some take the view that the reasons for admitting sense-impressions are sound and, therefore, that either representative realism or phenomenalism needs to be defended; others say that both representative realism and phenomenalism are incoherent, so that sense-impressions just cannot be admitted. In this paper I wish to argue that the initial belief is false; one can allow an essential role in perception to sense-impressions without being committed to either theory of perception. I shall argue that the belief that one cannot is due to the use in almost all philosophical theories in which sense-impressions are postulated of an inappropriate model of the role of these impressions. I shall suggest a more appropriate model. One philosopher who takes the first of the paths I have mentioned is J. L. Mackie,1 one who takes the second is D. M. Armstrong.2 Mackie argues that John Locke's distinction between the primary and secondary qualities of material objects (which quite clearly entails the admission into perception of an essential element of sensation-the ideas of the secondary as well as of the primary qualities) requires some sort of representative theory of perception (ibid p. 8). Objections to the representative theory must therefore be seen as fatal to the Lockean distinction (ibid, p. 28). Since Mackie finds at least one of Locke's arguments for the distinction convincing, he sets out to circumvent the objection to the representative theory. Armstrong, on the other hand, argues