to these labours than that anyone should make the unexpected discovery that there neither is, nor can be, any a priori knowledge at all. But there is no danger of this. This would be the same thing as if one sought to prove by reason that there is no reason. For we only say that we know something by reason, when we are conscious that we could have known it, even if it had not been given to us in experience; hence rational knowledge and knowledge a priori are one and the same. It is a clear contradiction to try to extract necessity from a principle of experience (ex pumice aquam), and to try by this to give a judgement true universality (without which there is no rational inference, not even inference from analogy, which is at least a presumed universality and objective necessity). To substitute subjective necessity, that is, custom, for objective, which exists only in a priori judgements, is to deny to reason the power of judging about the object, i.e., of knowing it, and what belongs to it. It implies, for example, that we must not say of something which often or always follows a certain antecedent state that we can conclude from this to that (for this would imply objective necessity and the notion of an a priori connexion), but only that we may expect similar cases (just as animals do), that is that we reject the notion of cause altogether as false and a mere delusion. As to attempting to remedy this want of objective and consequently universal validity by saying that we can see no ground for attributing any other sort of knowledge to other rational beings, if this reasoning were valid, our ignorance would do more for the enlargement of our knowledge than all our meditation. For, then, on this very ground that we have no knowledge of any other rational beings besides man, we should have a right to suppose them to be of the same nature as we know ourselves to be: that is, we should really know them. I omit to mention that universal assent does not prove the objective validity of a judgement (i.e., its validity as a cognition), and although this universal assent should accidentally happen, it could furnish no proof of agreement with the object; on the contrary, it is the objective validity which alone constitutes the basis of a necessary universal consent.
Hume would be quite satisfied with this system of universal empiricism, for, as is well known, he desired nothing more than that, instead of ascribing any objective meaning to the necessity in the concept of cause, a merely subjective one should be assumed, viz., custom, in order to deny that reason could judge about God, freedom, and immortality; and if once his principles were granted, he was certainly well able to deduce his conclusions therefrom, with all logical coherence. But even Hume did not make his empiricism so universal as to include mathematics. He holds the principles of mathematics to be analytical; and if this were correct, they would certainly be apodeictic also: but we could not infer from this that reason has the faculty of forming apodeictic judgements in philosophy also — that is to say, those which are synthetical judgements, like the judgement of causality. But if we adopt a universal empiricism, then mathematics will be included.
Now if this science is in contradiction with a reason that admits only empirical principles, as it inevitably is in the antinomy in which mathematics prove the infinite divisibility of space, which empiricism cannot admit; then the greatest possible evidence of demonstration is in manifest contradiction with the alleged conclusions from experience, and we are driven to ask, like Cheselden’s blind patient, “Which deceives me, sight or touch?” (for empiricism is based on a necessity felt, rationalism on a necessity seen). And thus universal empiricism reveals itself as absolute scepticism. It is erroneous to attribute this in such an unqualified sense to Hume,7 since he left at least one certain touchstone (which can only be found in a priori principles), although experience consists not only of feelings, but also of judgements.
However, as in this philosophical and critical age such empiricism can scarcely be serious, and it is probably put forward only as an intellectual exercise and for the purpose of putting in a clearer light, by contrast, the necessity of rational a priori principles, we can only be grateful to those who employ themselves in this otherwise uninstructive labour.
1 Lest any one should imagine that he finds an inconsistency here when I call freedom the condition of the moral law, and hereafter maintain in the treatise itself that the moral law is the condition under which we can first become conscious of freedom, I will merely remark that freedom is the ratio essendi of the moral law, while the moral law is the ratio cognoscendi of freedom. For had not the moral law been previously distinctly thought in our reason, we should never consider ourselves justified in assuming such a thing as freedom, although it be not contradictory. But were there no freedom it would be impossible to trace the moral law in ourselves at all.
2 The union of causality as freedom with causality as rational mechanism, the former established by the moral law, the latter by the law of nature in the same subject, namely, man, is impossible, unless we conceive him with reference to the former as a being in himself, and with reference to the latter as a phenomenon — the former in pure consciousness, the latter in empirical consciousness. Otherwise reason inevitably contradicts itself.
3 A reviewer who wanted to find some fault with this work has hit the truth better, perhaps, than he thought, when he says that no new principle of morality is set forth in it, but only a new formula. But who would think of introducing a new principle of all morality and making himself as it were the first discoverer of it, just as if all the world before him were ignorant what duty was or had been in thorough-going error? But whoever knows of what importance to a mathematician a formula is, which defines accurately what is to be done to work a problem, will not think that a formula is insignificant and useless which does the same for all duty in general.
4 [See Kant’s “Das mag in der Theoric ricktig seyn,” etc. Werke, vol. vii, p. 182.]
5 It might also have been objected to me that I have not first defined the notion of the faculty of desire, or of the feeling of pleasure, although this reproach would be unfair, because this definition might reasonably be presupposed as given in psychology. However, the definition there given might be such as to found the determination of the faculty of desire on the feeling of pleasure (as is commonly done), and thus the supreme principle of practical philosophy would be necessarily made empirical, which, however, remains to be proved and in this critique is altogether refuted. It will, therefore, give this definition here in such a manner as it ought to be given, in order to leave this contested point open at the beginning, as it should be. LIFE is the faculty a being has of acting according to laws of the faculty of desire. The faculty of DESIRE is the being’s faculty of becoming by means of its ideas the cause of the actual existence of the objects of these ideas. PLEASURE is the idea of the agreement of the object, or the action with the subjective conditions of life, i.e., with the faculty of causality of an idea in respect of the actuality of its object (or with the determination of the forces of the subject to action which produces it). I have no further need for the purposes of this critique of notions borrowed from psychology; the critique itself supplies the rest. It is easily seen that the question whether the faculty of desire is always based on pleasure, or whether under certain conditions pleasure only follows the determination of desire, is by this definition left undecided, for it is composed only of terms belonging to the pure understanding, i.e., of categories which contain nothing empirical. Such precaution is very desirable in all philosophy and yet is often neglected; namely, not to prejudge questions by adventuring definitions before the notion has been completely analysed, which is often very late. It may be observed through the whole course of the critical philosophy (of the theoretical as well as the practical reason) that frequent opportunity offers of supplying defects in the old dogmatic method of philosophy, and of correcting errors which are not observed until we make such rational use of these notions viewing them as a whole.
6 I am more afraid in the present treatise of occasional misconception in respect of some expressions which I