GOING BACK TO KAXT. 275 opposite principle in the very heart of that thought, then perhaps it will appear that the Ding-an-sich is not such an irrational element in the philosophy of Kant as his critics suppose, but springs from a deep- seated need lying hid in the very principle vith which it apparently conflicts. The exact form in which this conception appears in Kant may have to be given up, but the truth which it expresses, the dependence of thought upon fact which determines rather than is determined by thought, will remain. To demonstrate that this is the case we cannot leave out of sight the work of subsequent speculation. "What we pro- pose to do is rather to show that idealistic speculation although it may not understand it carries Realism in its womb. It is not enough to insist upon the necessity of going back to Kant. All depends upon the way in which we go back to him, and there are different ways of going back. If we glance for a moment at the last century, it is easy to conceive a similar feeling arising with regard to Locke, and Berkeley and Hume. Many who saw the external world disappear in the subjective idealism of Berkeley, and rational truth disappear in the sceptical pheno- menalism of Hume, must have felt strongly that there was a necessity for going back to Locke. Locke must be studied again, they would say ; the simple common sense of the Essay on the Human Understanding can lead to no such results ; and accordingly a going back to Locke did take place with great consequences for the history of philosophy. Kant went back to Locke, and studied again the problem of Locke, but he went back with a full and perfect consciousness of the subtle question Hume had asked, of the hidden problem of which his philosophy was the expression. If a going back to Kant is to take place with fruitful results it must be made in the same way. Just as Kant went back to the problem of Locke and raised the question Is it really true that all our knowledge not merely begins with experience but also arises out of experience '? so it is necessary now to go back to the problem of Kant, and to inquire whether the presupposition under which Kant solved that problem is really true and is really necessary. The problem of Kant was the problem of synthesis and of synthesis a priori. How are synthetical judgments a priori possible ? is the question prefixed to the Critick of Pure Reason. This is the determining point in his philosophy by which its scope is bounded. The necessity of h priori synthesis to know- ledge, and the necessity of his system of Pure Season to a priori synthesis, are the principles upon which he proceeds. Now, if it can be shown that the presupposition, under which alone Kant conceived it possible that a priori synthesis could exist, effectually precludes the possibility of the knowledge in question being either synthetical or a priori, and if the same can be shown to be true of Kant's successors, then the way will have been prepared towards such a modification of the Kantian principle and of their
