its sole form and function, yet one so rich in results that all our empirical knowledge rests upon it. If, as has often been said, the refutation of an error is only complete when the way it originated has been psychologically demonstrated, I believe I have achieved this, with regard to Kant's doctrine of the categories and their schemata, in what I have said above.
After Kant had thus introduced such great errors into the first simple outlines of a theory of the faculty of perception, he adopted a variety of very complicated assumptions. To these belongs first of all the synthetic unity of apperception: a very strange thing, very strangely explained. "The I think must be able to accompany all my ideas." Must – be able: this is a problematic-apodictic enunciation; in plain English, a proposition which takes with one hand what it gives with the other. And what is the meaning of this carefully balanced proposition? That all knowledge of ideas is thinking? That is not the case: and it would be dreadful; there would then be nothing but abstract conceptions, or at any rate a pure perception free from reflection and will, such as that of the beautiful, the deepest comprehension of the true nature of things, i.e., of their Platonic Ideas. And besides, the brutes would then either think also, or else they would not even have ideas. Or is the proposition perhaps intended to mean: no object without a subject? That would be very badly expressed by it, and would come too late. If we collect Kant's utterances on the subject, we shall find that what he understands by the synthetic unity of apperception is, as it were, the extensionless centre of the sphere of all our ideas, whose radii converge to it. It is what I call the subject of knowing, the correlative of all ideas, and it is also that which I have fully described and explained in the 22d chapter of the Supplements, as the focus in which the rays of the activity
