Guyer, P., (ed.), Critique of the Power of Judgment, Cambridge: Cambridge University Press. 2001.
We proceed quite correctly if, as usual, we divide Philosophy, as containing the principles of the rational cognition of things by means of concepts (not merely, as logic does, principles of the form of thought in general without distinction of Objects), into theoretical and practical. But then the concepts, which furnish their Object to the principles of this rational cognition, must be specifically distinct; otherwise they would not justify a division, which always presupposes a contrast between the principles of the rational cognition belonging to the different parts of a science.
Now there are only two kinds of concepts, and these admit as many distinct principles of the possibility of their objects, viz. natural concepts and the concept of freedom. The former render possible theoretical cognition according to principles a priori; the latter in respect of this theoretical cognition only supplies in itself a negative principle (that of mere contrast), but on the other hand it furnishes fundamental propositions which extend the sphere of the determination of the will and are therefore called practical. Thus Philosophy is correctly divided into two parts, quite distinct in their principles; the theoretical part or Natural Philosophy, and the practical part or Moral Philosophy (for that is the name given to the practical legislation of Reason in accordance with the concept of freedom). But up to the present a gross misuse of these expressions has prevailed, both in the division of the different principles and consequently also of Philosophy itself. For what is practical according to natural concepts has been identified with the practical according to the concept of freedom; and so with the like titles, ‘theoretical’ and ‘practical’ Philosophy, a division has been made, by which in fact nothing has been divided (for both parts might in such case have principles of the same kind).
The will, regarded as the faculty of desire, is (in this view) one of the many natural causes in the world, viz. that cause which acts in accordance with concepts. All that is represented as possible (or necessary) by means of a will is called practically possible (or necessary); as distinguished from the physical possibility or necessity of an effect, whose cause is not determined to causality by concepts (but in lifeless matter by mechanism and in animals by instinct). Here, in respect of the practical, it is left undetermined whether the concept which gives the rule to the causality of the will, is a natural concept or a concept of freedom.
But the last distinction is essential. For if the concept which determines the causality is a natural concept, then the principles are technically practical; whereas, if it is a concept of freedom they are morally practical. And as the division of a rational science depends on the distinction between objects whose cognition needs distinct principles, the former will belong to theoretical Philosophy (doctrine of Nature), but the latter alone will constitute the second part, viz. practical Philosophy (doctrine of Morals).