all a priori knowledge in general, which The Critique of Pure Reason is."[1] He felt, therefore, no incompatibility in describing the transcendental philosophy of the Critique as having for its object the determination of the possibility of a priori knowledge and yet claiming that its chief undertaking was to explain the possibility of experience. We can know a priori only what we ourselves think into things. And this faculty of thinking a priori, this original and invariable spontaneity without which we should know nothing a priori, is at the same time the condition of the possibility of all other appearances, of nature itself as a system of knowable objects.[2] It is thus the ground of a posteriori as well as of a priori synthetic judgments. And in one of the Reflections, Kant says, "The principal inquiry is, How do we come by knowledge in general, and especially a priori knowledge?"[3] But that this statement is not to be strained beyond the natural interpretation here adopted, is clear from the circumstance that it is intercalated between two other reflections, both of which revert to the original formula of a priori synthetic judgments. "It is," runs the one, "the possibility of all a priori knowledge, which is self-contained and borrows nothing from the object, that forms our first and most important inquiry."[4] "The first question," runs the other, "is, how we can have notions which have not been learned from any appearance of things, or principles, which no experience has taught."[5]
This last quotation brings us to a point of view from which the whole question may, with great advantage, be finally surveyed. Kant's problem, it has been here maintained, was correctly stated by Kant himself. How are a priori synthetic judgments possible? How and what may reason know independently of instruction from sense? The scope of such knowledge, it was settled by the Critique, does not go beyond the sensible world. And the conditions of it are the a priori functions of the mind, which may be expressed in notions (cause, substance, etc.). Our problem, therefore, may be said to
