64 SHAD WORTH H. HODGSON. and the ontological which referred it to some transcendent, or it might be some immanent and absolute, source at once of world, and agents, and consciousness. But the distinc- tion between psychology and philosophy falls very differ- ently, on my method. It falls, not between two theories of genesis of experience or consciousness, but between the analysis and the genesis of conscious experience, however that genesis is conceived. Psychology is with me one of the positive sciences, and philosophy is the subjective counter- part of them all. I think it will now be clear to you, as I said above I hoped I should make it, that to take experience simply as given, avoiding the assumption that sensations are given as a chaos, does not involve the assumption of an orderly ex- ternal world. Whether there is an external world, and whether it is orderly, we have to learn from analysis of experience as given. We are not to assume either that it is given orderly or that it is given chaotic. To assume order is to assume too much in the world, to assume chaos is to assume too much in the percipient. In either case a genesis of experience is assumed, and mixed up with pure analysis of it ; and to that extent we are deserting experience and taking our stand upon fancies. If we jind experience orderly, if we Jind some order in the stream of conscious- ness when we analyse it, this surely is not a result to marvel or to grieve at, though it may have something in it humiliating to human conceit, as dispensing with the necessity of our assistance in creation, or, as Kant too modestly puts it, in " constituting objects ". The fact is, that this extreme disorder and isolation in sensations, which Kant's theory logically requires, and which I have called a chaos, is an utter illusion. It is not only without proof as a fact, it is impossible as an imagina- tion. It cannot even be thought. It is not merely a fiction, it is an unthinkable fiction, illusory even as a falsity. Yet it is on this illusion, with its necessary complement, the assumption of a real but unknowable agency in the Subject, that Kant's distinction between the phenomenal and the transcendent, or noumenal, rests, as well as all post-Kantian ontologies which build on that latter distinction. It is the inside of a soap-bubble. But perhaps you will say to me, that without this assump- tion of chaos in sensations Hume is not answered, and therefore that, if it cannot be assumed, Hume cannot be answered. If Kant is not allowed to assume order and nexus in sensations on the ground that this would be 110 reply to
