Free Russell - A Very Short Indroduction by A. C. Grayling

concerns space, and particularly the distinction between the all-embracing public space assumed by science, and the private spaces in which the sense-data of individual perceivers exist. Private space is built out of the various visual, tactual, and other experiences which a perceiver coordinates into a matrix with himself at the centre. But because we do not have acquaintance with the public space of science, its existence and nature is wholly a matter of inference.
    Thus Russell’s first version of a theory of knowledge and perception, as set out in PP . It has a brisk common-sense feel about it on first encounter, but it is far from unproblematic. For example, Russell speaks of ‘primitive’ knowledge and describes it as intuitive; but he does not offer an account of what such knowledge is, beyond saying that it does not require the support of anything more self-evident than itself. But this definition is hardly adequate, and it is obscured further when he adds that there are two kinds of self-evidence, only one of which is basic. Does this distinction make sense? What is ‘self-evidence’ anyway? Nor does he consider the possibility that two propositions might contradict each other despite appearing self-evident when considered separately. If this were to happen, which is one to choose, and on what additional principles of self-evidence?
    Another criticism levelled at Russell’s view is that it makes an important but questionable assumption about the basic nature of senseexperience. This is that sense-data, qua sensory minima such as particular colours, smells, or sounds, are simply given in experience, and are its most primitive elements. But in fact sensory experience is not ‘thin’ and immediate in this way at all. Rather, it is a rich and complex experience of houses, trees, people, cats, and clouds – it is phenomenologically ‘thick’, and sense-data are only arrived at by a sophisticated process of emptying ordinary perceptual experience of everything it normally means to us. Thus we do not see a rectangle and infer that it is a table; we see a table, and when we come to concentrate upon its shape we see that it is a rectangle.
    This criticism is undoubtedly right as far as it goes, but there are ways in which it can be accommodated while still allowing us to describe the purely sensory aspect of experience independently of the usual load of beliefs and theories it carries. Since the whole point is that we are trying to justify possession of those beliefs by showing that perceptual experience entitles us to them, we obviously need an account of our perceptual experience considered purely as such, so that we can evaluate its adequacy to the task. Russell’s aim in talking of sense-data is to do just that. Moreover Russell recognized that sense-data are not the immediately perceptually given ; in writings during the decade after PP he points out repeatedly that specifications of sense-data come last in analysis, not first in experience.
    Another criticism is that Russell assumes that immediate experience is expressible in propositions which, despite the fact that they describe only what is subjectively ‘given’, can be used as a basis for knowledge of the world. But how can what seems to apply only to private experience, and carries no reference to what is beyond that experience, be the basis for a theory of knowledge? It does not help to say that Russell also allows a priori knowledge of logical principles which permit inferences from these propositions, for there would be no motivation to draw them unless, in addition, the subject possessed some general empirical beliefs to serve as the major premisses in such inferences, and some empirical hypotheses which the inferences in effect test or support. But these are not available to an experiencer possessed only, as Russell presents him, with sense-data and the self-evident truths of logic.
    This problem carried weight with Russell himself, and much