Page:Kant's Prolegomena etc (1883).djvu/138

ment is therefore analytic, is merely the ambiguity of expression. We ought, namely, to cogitate a certain predicate to a given conception, and this necessity adheres even to the conceptions themselves. But the question is not what we ought to, but what we actually do, although obscurely, cogitate in them; this shows us that the predicate of those conceptions is dependent indeed necessarily, though not immediately (but by means of an added intuition), upon its subject.

This division is in view of the Critique of human understanding indispensable, and deserves therefore to be classic in this department; though I am not aware of any other in which it has any important use. And here I also find the cause why dogmatic philosophers who looked for the sources of metaphysical judgments in metaphysics itself (rather than outside of it, in the laws of the pure Reason in general), have always neglected this division, that seems so naturally to offer itself, and like the celebrated Wolff, or the acute Baumgarten, who followed in his steps, have sought the proof of the principle of sufficient reason, which is obviously synthetic, in that of contradiction. On the other hand, I can trace already in "Locke's Essays on the Human Understanding" a notion of this division. For in the third chapter of the fourth book, (Chap. III. § 9 et seq.,) after he has spoken of the connection of different presentations in judgments, and of their sources, one of which he places in identity or contradiction (analytic judgments), and the other in the existence of presentations in a subject (synthetic judgments), he confesses, § 10, that our knowledge (à priori) of the last is very limited, amounting almost to nothing. But there is so little that is definite and reduced to rule in what he says respecting this kind of knowledge, that one cannot wonder that nobody, strange to say, not excepting Hume, was induced thereby to institute investigations into the