teaching us to take these into consideration. We shall therefore have to investigate a priori the possibility of a categorical imperative, as we have not in this case the advantage of its reality being given in experience, so that [the elucidation of] its possibility should be requisite only for its explanation, not for its establishment. In the meantime it may be discerned beforehand that the categorical imperative alone has the purport of a practical law; all the rest may indeed be called principles of the will but not laws, since whatever is only necessary for the attainment of some arbitrary purpose may be considered as in itself contingent, and we can at any time be free from the precept if we give up the purpose; on the contrary, the unconditional command leaves the will no liberty to choose the opposite; consequently it alone carries with it that necessity which we require in a law. Secondly, in the case of this categorical imperative or law of morality, the difficulty (of discerning its possibility) is a very profound one. It is an a priori synthetical practical proposition; * and as there is so much difficulty in discerning the possibility of speculative propositions of this kind, it may readily be supposed that the difficulty will be no less with the practical. * I connect the act with the will without presupposing any condition resulting from any inclination, but a priori, and therefore necessarily (though only objectively, i.e., assuming the idea of a reason possessing full power over all subjective motives). This is accordingly a practical proposition which does not deduce the willing of an action by mere analysis from another already presupposed (for we have not such a perfect will), but connects it immediately with the conception of the will of a rational being, as something not contained in it. In this problem we will first inquire whether the mere conception of a categorical imperative may not perhaps supply us also with the formula of it, containing the proposition which alone can be a categorical imperative; for even if we know the tenor of such an absolute command, yet how it is possible will require further special and laborious study, which we postpone to the last section. When I conceive a hypothetical imperative, in general I do not know beforehand what it will contain until I am given the condition. But when I conceive a categorical imperative, I know at once what it contains. For as the imperative contains besides the law only the necessity that the maxims * shall conform to this law, while the law contains no conditions restricting it, there remains nothing but the general statement that the maxim of the action should conform to a