As all obligation is entirely dependent on a condition, so also is all duty. Both conceptions are very closely related, indeed almost identical. The only difference between them might be said to be that obligation in general may rest on mere force, whereas duty involves the sense of obligation deliberately undertaken, such as we see between master and servant, principal and subordinate, rulers and the ruled. And since no one undertakes a duty gratis, every duty implies also a right. The slave has no duties, because he has no rights; but he is subject to an obligation which rests on sheer force. In the following Part I shall explain the only meaning which the conception "Duty" has in Ethics.
If we put Ethics in an imperative form, making it a Doctrine of Duties, and regard the moral worth or worthlessness of human conduct as the fulfilment or violation of duties, we must remember that this view of Duty, and of Obligation in general, is undeniably derived solely from theological Morals, and primarily from the Decalogue, and consequently that it rests essentially and inseparably on the assumption of man's dependence on another will which gives him commands and announces reward or punishment. But the more the assumption of such a will is in Theology positive and precise, the less should it be quietly and unsuspectingly introduced into philosophical Morals. Hence we have no right to assume beforehand that for the latter the imperative Form, the ordaining of commands, laws, and duties is an essential and a matter of course; and it is a very poor shift to substitute the word "absolute" or "categorical" for the external condition which is indissolubly attached to such conceptions by their very nature: for this gives rise, as explained above, to a contradictio in adjecto.
Kant, then, without more ado or any close examination, borrowed this imperative Form of Ethics from theological Morals. The hypotheses of the latter (in other words, Theology) really lie at the root of his system, and as these alone in point of fact lend it any meaning or sense, so they cannot be separated from, indeed are implicitly contained in, it. After this, when he had expounded his position the task of developing in turn a Theology out of his Morals—the famous Moraltheologie—was easy enough. For the conceptions which are implicitly involved in his Imperative, and which lie hidden at the base of his Morals, only required to be brought forward and expressed explicitly as postulates of Practical Reason. And so it was that, to the world's great edification, a Theology appeared depending simply on Ethics, indeed actually derived therefrom. But this came about because the ethical system itself rests on concealed theological hypotheses. I mean no derisive comparison, but in its form the process is analogous to that whereby a conjurer prepares a surprise for us, when he lets us find something where he had previously employed his art to place it. Described in the abstract, Kant's procedure is this: what ought to have been his first principle, or hypothesis (viz., Theology) he made the conclusion, and what ought to have been deduced as the conclusion (viz., the Categorical Command) he took as his hypothesis.[4] But after he had thus turned the thing upside down, nobody, not even he himself, recognised it as being what it really was, namely the old well-known system of theological Morals. How this trick was accomplished we shall consider in the sixth and seventh chapters of the present Part.
Ethics was of course frequently put in the imperative form, and treated as a doctrine of duties also in pre-Kantian philosophy; but it was always then based upon the will of a God whose existence had been otherwise proved, and so there was no inconsequence. As soon, however, as the attempt was made, as Kant attempted, to give a foundation to Ethics independent of this will, and establish it without metaphysical hypotheses, there was no longer any justification for taking as its basis the words "thou shalt," and "it is thy duty" (that is, the imperative form), without first deducing the truth thereof from some other source.
[1] Du sollt (sic) nicht lügen.
[2] Sollt is the old form for "sollst." Cf. Eng., shalt: Icel; skalt—(Translator.)
[3] A contradiction in the adjective. This occurs when the epithet applied to a noun contradicts its essential meaning.—(Translator.)
[4] Like the converse of a geometrical proposition, this Kantian inversion is not necessarily true; its validity, in fact, depends on the conclusion being implicitly contained in the hypothesis. —(Translator.)