with relations-first have sensations and then build them up into things and project them out into space. We apprehend all at once the thing-sensational qualities and relations and all. And the thing is apprehended as something permanent existing in space, of which changing accidents are predicable.

But, asks the Realist, do we not also know and believe that the thing existed-and existed as we know it-before we perceived it, and goes on existing as perceived, after we have ceased to perceive it, and would equally be there if we had never perceived it? And is not this exactly what Idealism denies ? Now, if we are speaking of commonplace, unreflective thought, it may be admitted at once that the Realist is right. We do all of us begin by thus thinking of objects in space as independent things. The question is whether on reflection this common-sense view of the Universe can justify itself- -as the truth and the whole truth. But observe-if it is to justify itself, we shall have to say that the secondary qualities are quite as independent of our perception as the primary. That is what our commonplace attitude of mind undoubtedly requires. And yet here even Mr. Prichard-though somewhat reluctantly begins to hang back. Very rarely indeed has any serious thinker maintained that things are colored or hot apart from all relation to our consciousness-that the thing in respect of its secondary qualities is altogether apart from perception just what it is for our perception. Once give up this, and the Realist has admitted that is is not true that knowledge of a thing means in all cases knowing what

4 It is observable that the champion of another new Realism, the Honorable Bertrand Russell, takes exactly the opposite view. "Berkeley," he says, "retains the merit of having shown that the existence of matter is capable of being denied without absurdity, and that if there are any things that exist independently they cannot be immediate objects of our sensations." (The Problems of Philosophy, p. 19). So again: "What we directly see and feel is merely an appearance' which we believe to be a sign of some 'reality' 'behind.''' (Ibid., pp. 23, 24). "The one thing we know about it [the object in question] is that it is not what it seems. (Ibid., p. 24).

it is apart from my knowledge of it. Once admit that the color of things would have no existence in a world in which color never had been and never would be known, and it ceases to be true that knowledge is knowing what things are apart from knowledge. Mr. Prichard only avoids this admission by boldly denying that the secondary qualities belong to the thing at all. They are something that belong solely to our mental experience. "It must therefore be concluded that color is not a quality of bodies." But surely here he contradicts common-sense and accurate psychology, and gives up all the truth that really is contained in his presentative theory of perception. To say that what we immediately perceive is a colorless piece of paper occupying the space outside us, and that on looking at it we experience a sensation of white which is only in our minds, is surely preposterous. It is the paper that is colored, not my sensations. The color is unthinkable apart from the surface. If the surface which we see is out of the mind, the color must be out of the mind. Yet the color by Mr. Prichard's admission could not exist apart from the mind. And if this is so in respect of some of the qualities of the thing, it may be so with regard to others. And the Idealist contends that a little fuller reflection-a little more of that same reflection which has already differentiated the realistic philosopher from the man in the street in his most unreflective moments-will lead him to see that it must be sothat the existence of primary qualities in a mindless world is as unthinkable as the existence of secondary qualities. There are passages in which Mr. Prichard appears to admit that color is a quality of the thing unlike "hot" or "cold" which are merely mental experiences of ours. I do not think he ever succeeds in reconciling these admissions with the theory."

The only way of evading this line of argument is to admit that, though the object does not exist in all ways as 5 Kant, Theory of Knowledge, p. 87.

6 Cf. pp. 87-91.

we perceive it, yet we know that it does exist. But directly you admit that the object as perceived does not exist apart from perception, you are back again at the old Lockian position; the thing-in-itself is not what we immediately perceive but something behind what we perceive; and that mode of thought properly followed out lands you in the Kantian unknown and unknowable thing-in-itself-the difficulty and superfluity of which are a commonplace of philosophical criticism, and are fully recognized by the new Realists. For the Realist there is no way out of the dilem"If the thing which I assert to exist is not what 1 perceive, I have no reason for asserting its existence: if I do perceive it, it must exist in itself just as I perceive it." And yet, as we have seen, this last position cannot seriously be maintained.

ma:

But are we then to admit, the Realist may reply, that the thing comes into existence when I perceive it, and ceases to exist when I cease to perceive it? No. It is quite true that, whether instinctively or after the fullest reflection, we do recognize that there must be a sense in which things have existed not merely before I personally have perceived them, but before any human mind has perceived them. For what I know about things compels me to think so. It is by no means true that we know this immediately and a priori. It is only by a long course of experience and inference from experience and comparison with other people's experiences that we learn to distinguish purely subjective experiences from what we call objective realities. What gradually leads us to this distinction is the impossibility of harmoniously co-ordinating our experiences without it. To the savage the world of his dreams is as real as his waking world. It is a region into which his soul occasionally wanders while it leaves the body; or the spirit which appears to him is a spirit which has really come back to its old home from a distant region. And so when seen in the light of the whole mass of human experience, my present perception of a piece of granite does imply that it existed

as a molten mass centuries before any human being apprehended it as granite, and when there were (so far as we know) no human spirits to know it in its molten state. It is just this impossibility of reconciling the permanence of material things, which is forced upon us by the inferences of science, with the fleetingness and fragmentariness of human knowledge which leads most Idealists to postulate the existence of a divine Mind in and for which all things eternally exist. And even for such a Mind we must not suppose that all distinction between the self which knows and the thing which is known is lost-that the space-occupying object is dissolved, as it were, into a mass of purely subjective and spaceless experiences. A thing will always be different from a sensation or a mind. But because the object is distinguishable from the subject, it does not follow that it can exist without it. The convex is not the concave, the color is not the surface, and yet you cannot have the convex without the concave or the color without the surface which is colored. Equally unintelligible from the Idealistic point of view is the object without a subject.

(2) A second criticism on Idealism which is now widely circulated in Oxford is that Idealism uses the word "perception" in an ambiguous sense. It confuses the act of perceiving with the object perceived. The plausibility of the system, it is suggested, depends upon the device of getting the disputant to admit that the act of perception cannot exist without a perceiving mind, and then triumphantly claiming to have proved that the object perceived cannot exist apart from such a mind. Here again there is nothing new in the much-vaunted discoveries of my Oxford friends. It is the very pith and marrow of Reid's criticism upon Berkeley.

And once again I submit that the contention is good as against Berkeley-not good as against modern forms of Idealism. Undoubtedly the thing is not simply a bundle of

7 An old objection of Reid's-also urged by Mr. Bertrand Russell (The Problems of Philosophy, p. 65).

sensations. The sensations of sight and touch considered as acts are in my mind: in so far as they are localized, they are localized somewhere in my organism. I am conscious of seeing with my eyes and of feeling with my fingers. The thing I regard as occupying space, and a space outside my body. The sensation is momentary and ever-changing; the thing I think of as permanently persisting while my sensations change. The sensations considered as sensations are in my mind: the qualities which they reveal in the thing are outside in the thing, and I regard the qualities as persisting while the sensations in me come and go. All this is true, but I contend that the truth of it is fully recognized by Kant, and by those who have learned of Kant. The thing for the post-Kantian Idealist is not the momentary sensation or even a bundle of sensations. It is not the thing as it is for sense but the thing as it is for thought which he regards as being the reality of the world or rather part of the reality. We must not, indeed, go to the opposite extreme, and talk as if the reality of things was simply what they are for abstract thinking quite apart from and independent of perception. We must not, with Green, say that the world "is a system of relations"-relations without anything to relate. Of such an Idealism Professor L. T. Hobhouse is quite justified in saying that it is difficult to discover from Professor Green's writings what position he attributes to sensation in the formation of our knowledge except that it is a contemptible one. Such an attitude has, I hope, forever been made impossible by Mr. Bradley. It is in actual perception that we come into the closest contact with the reality of things. All our thought about things-the thought that we express in predicating a universal predicate of a subject an attempt to represent the thing as it is for perception, an attempt which (as Mr. Bradley has taught us) is never wholly successful. The moment we begin to form universal conceptions, we have left out all the wealth of detail that we actually perceive. But it is not the momentary perception which constitutes the reality of the thing, or even the suc

is