APPEARANCES AND REALITY. 225 reality at least primarily relates to the spatial qualities of things. The only exception seems to be the case of temporal relations ; we certainly say that a certain event appeared to take a long time, though it really did not. 1 Again these cases bring before us by way of contrast the existence of spheres of perception where the distinction is wholly inapplicable, e.g., those of pain and sensation generally. A pain is necessarily what I feel it to be ; distinction between what it is, and what I feel or perceive it as, is meaning- less. Similarly a noise, in the sense of a sensation as distinct from its physical conditions, is what I hear it as. In fact it is just the absence of the distinction here that gives much of its plausibility to the view that we only know phenomena. If we once allow that our objects are states of ourselves, we seem unable to deny that at least we apprehend tJiem as they are, i.e., that we really know phenomena. We may now proceed to the analysis. The general form of statement to be analysed is, ' a thing looks or appears so and sa, though we know that it is not so in reality '. The following seem the main points : (1) Such a statement is in no sense about ' appearances '. For (a) its subject is not the appearance or look of the thing but the thing. When I assert that the moon looks as large as the sun, I make an assertion about the moon and not about its look. (/8) Its predicate is always 'real,' i.e., it is always of a kind appropriate to real things, as opposed to their look or appearance. Thus the stick looks bent ; the moon looks as large as the sun ; the railway lines look convergent. If we seek for a predicate suit- able to ' appearances,' we naturally think of such terms as ' decep- tive ' and ' untrustworthy '. Our statements then about appearances are expressed in the same terms as the reality from which we distinguish them. It at once follows (a) That statements about appearance imply that we at least know enough about reality to say that real things have certain possible predicates, e.g., bent or convergent. To deny this is to be wholly unable to state how things look. (6) That the issue involved in distinguishing appearance from reality concerns not the general character of the attributes of real things but their relation to a particular subject. The question is not, ' Is convergence or bentness the attribute of any real thing ? ' but, ' Is it an attribute of particular real things, this stick or these lines ? ' It follows that the distinction between reality and appearance relates solely to the details of our knowledge and not to its general or structural character. The attributes of reality cannot belong to ycvr) different from those of appearance. Doubt about its details implies certainty as to its general character. 1 ' Appearing ' and ' looking ' are treated as distinct in meaning from 'seeming' ; they stand for $aivrdat as opposed to doKfw.
