ideas—the possibility of perception under supposed circumstances -may be startling to common sense, but there is nothing in it to baffle scientific acceptance. There is, however, a very simple objection, which neither the subjective idealism of Berkeley nor Hume's modification of it can successfully meet, and which did much to block the further development of English empiricism during a full century. How can things be identified with perceptions, when the same thing can be perceived in so many ways and from so many different points of view? It is not as if the various impressions thus received were fused into a single image-as color and texture unite in the perception of a rose-leaf. They may be in the highest degree incompatible and mutually exclusive, as well as extremely different from each other. Yet they remain perceptions of the same thing. Neither Hume nor Berkeley can offer any explanation-except the denial of the fact. So long as the perception and the thing remain identical, a one-to-many relation between them is a manifest absurdity. The device of getting rid of epistemological dualism by equating one side of the division. with a portion of the other side will not suffice. In conclusion, we wish to remark that, while Berkeley and Hume denounce the representative theory, they in effect fall back upon its very crudest form for the conception of the relation between idea and ideatum. They regard it as necessarily a resemblance, and contemplate no other possibility. This is why, for example, the possibility of an idea of a spirit is rejected by Berkeley. The passive idea and the active spirit are so utterly unlike, that no possible bond of resemblance between them can subsist. This attitude on their part may be thought the more remarkable, as their investigations into general ideas had familiarized them with an altogether different type of representation, —that of ideas by words; and Berkeley had particularly noted the analogy between the signification of words and the visual perception of distance. The strain of eye-convergence means nearness, not because it resembles it in any way, but because it has been constantly associated with it. But representation such as this means simply the ability to suggest, arising from previous association. The thing must first be represented by its ideacopy, before the suggestion of that copy by an associated idea is possible. CHAPTER I THE CRITICAL PHILOSOPHY After the successful carrying-out of a revolution, our wonder may be less excited by the greatness than by the limitedness of the changes that have been effected. Looking beneath the altered surface of things, we find a scarcely modified substratum, which bears witness to an unbroken historical continuity. Very notably is this the case with the more important revolutions in philosophical thought. It is notorious, for example, that the founders of modern philosophy, with all their contempt of scholasticism, were never able to free themselves from its most characteristic concepts-nay, never awoke to their bondage to them. Very similar is the relation which the critical philosophy bears to its forerunners, rationalism and empiricism. The 'Copernican hypothesis' of Kant, despite its magnificent daring, meant no such absolute shift of the center of vision as its author supposed. On the contrary, nothing is more apparent to the reflective student than the far-reaching identity of the fundamental logical conceptions of Kantian and pre-Kantian thought. Indeed, it may safely be asserted, that there is not a single one of the doctrines which we have pointed out as characteristic of the old dogmatism, that is not to be found, either openly expressed or implicitly accepted, in the writings of Kant. And yet is it none the less true, that in the critical philosophy a transformation of the traditional logic is involved. So far as it is possible to regard this transformation as due to a single revolutionary idea, it may be described as having its source in a new conception of the nature of truth and validity. As conceived by rationalism, the warrant for the truth of any proposition could be exhibited only by deducing it from some more general proposition, whose truth in turn must be attested by |