Everything in the Realm of Ends has either a Price or a Dignity. Skill, Diligence, &c., bearing on human likings and needs, have a Market-price; Qualities like Wit, Fancy, &c., appealing to Taste or Emotional Satisfaction, have an Affection-price. But Morality, the only way of being End-in-self, and legislating member in the Realm of Ends, has an intrinsic Worth or Dignity, calculable in nothing else. Its worth is not in results, but in dispositions of Will; its actions need neither recommendation from a subjective disposition or taste, nor prompting from immediate tendency or feeling. Being laid on the Will by Reason, they make the Will, in the execution, the object of an immediate Respect, testifying to a Dignity beyond all price. The grounds of these lofty claims in moral goodness and virtue are the participation by a rational being in the universal legislation, fitness to be a member in a possible Realm of Ends, subjection only to self-imposed laws. Nothing having value but as the law confers it, an unconditional, incomparable worth attaches to the giving of the law, and Respect is the only word that expresses a rational being's appreciation of that. Autonomy is thus the foundation of the dignity of human and of all rational nature.
The three different expressions that have been given to the one general principle of morality imply each the others, and differ merely in their mode of presenting one idea of the Reason to the mind. Universal application of the Maxim of Conduct, as if it were a law of nature, is the formula of the Will as absolutely good; universal prohibition against the use of rational beings ever as means only, has reference to the fact that a good will in a rational being is an altogether independent and ultimate End, an End-in-self in all; universal legislation of each for all recognizes the prerogative or special dignity of rational beings, that they necessarily take their maxims from the point of view of all, and must regard themselves, being Ends-in-self, as members in a Realm of Ends (analogous to the Realm, or Kingdom of Nature), which, though merely an ideal and possible conception, none the less really imposes an imperative upon action. Morality, he concludes, is the relation of actions to the Autonomy of the Will, i.e., to possible universal legislation through its maxims. Actions that can co-exist with this autonomy are allowed; all others are not. A will, whose maxims necessarily accord with the laws of Autonomy, is holy, or absolutely good; the dependence of a will not thus absolutely good is Obligation. The objective necessity of an action from obligation is Duty. Subjection to law is not the only element in duty; the fact of the law being self-imposed gives Dignity.
The Autonomy of the will is its being a law to itself, without respect to the objects of volition; the principle of autonomy is to choose only in such a way as that the maxims of choice are conceived at the same time as a universal law. This rule cannot be proved analytically to be an Imperative, absolutely binding on every will; as a synthetic proposition it requires, besides a knowledge of the objects, a critique of the subject, i.e., pure practical Reason, before, in its apodeictic character, it can be proved completely à priori. Still the mere analysis of moral conceptions has sufficed to prove it the sole principle of morals, because this principle is seen to be a categorical Imperative, and a categorical Imperative enjoins neither more nor less than this Autonomy. If, then, Autonomy of Will is the supreme principle, Heteronomy is the source of all ungenuine principles, of Morality. Heteronomy is whenever the Will does not give itself laws, but some object, in relation to the Will, gives them. There is then never more than a hypothetical Imperative: I am to do something because I wish something else.
There follows a division and criticism of the various possible principles of morality that can be set up on the assumption of Heteronomy, and that have been put forward by human Reason in default of the required Critique of its pure use. Such, are either Empirical or Rational. The Empirical, embodying the principle of Happiness, are founded on (1) physical or (2) moral feeling; the Rational, embodying the principle of perfection, on (1) the rational conception of it as a possible result, or (2) the conception of an independent perfection (the Will of God), as the determining cause of the will. The Empirical principles are altogether to be rejected, because they can give no universal law for all rational beings; of the Rational principles, the first, though setting up an empty and indefinite conception, has the merit of at least making an appeal from sense to pure reason. But the fatal objection to all four is their implying Heteronomy; no imperative founded on them can utter moral, i.e., categorical commands.
That the absolutely good Will must be autonomous—i.e., without any kind of motive or interest, lay Commands on itself that are at the same time fit to be laws for all rational beings, appears, then, from a deeper consideration of even the popular conceptions of morality. But now the question can no longer be put off: Is Morality, of which this is the only conception, a reality or a phantom? All the different expressions given to the Categorical Imperatives are synthetic practical propositions a priori; they postulate a possible synthetic use of the pure practical reason. Is there, and how is there, such a possible synthetic use? This is the question (the same as the other) that Kant proceeds to answer in the Third Section, by giving, in default of a complete Critique of the faculty, as much as is necessary for the purpose. But here, since he afterwards undertook the full Critique, it is better to stop the analysis of the earlier work, and summarily draw upon both for the remainder of the argument, and the rather because some important points have to be added that occur only in the later treatise. The foregoing is a sufficient example of his method of treatment.
The synthetic use of the pure practical reason, in the Categorical Imperative, is legitimized; Autonomy of the Will is explained; Duty is shown to be no phantom—through the conception of Freedom of Will, properly understood. Theoretically (speculatively), Freedom is undemonstrable; being eternally met, in one of the (cosmological) Antinomies of the Pure Reason, by the counter-assertion that everything in the universe takes place according to unchanging laws of nature. Even theoretically, however, Freedom is not inconceivable, and morally we become certain of it; for we are conscious of the 'ought' of duty, and with the 'ought' there must go a 'can.' It is not, however, as Phenomenon or Sensible Ens that a man 'can,' is free, has an absolute initiative; all phenomena or Sensible Entia, being in space and time, are subject to the Natural Law of Causality. But man is also Noumenon, Thing-in-self, Intelligible Ens; and as such, being free from conditions of time and space, stands outside of the sequence of Nature. Now, the Noumenon or Ens of the Reason (he assumes) stands higher than, or has a value above, the Phenomenon or Sensible Ens (as much as Reason stands higher than Sense and Inclination); accordingly, while it is only man as Noumenon that 'can,' it is to man as Phenomenon that the 'ought' is properly addressed; it is upon man as Phenomenon that the law of Duty, prescribed, with perfect freedom from motive, by Man as Noumenon, is laid.
Freedom of Will in Man as Rational End or Thing-in-self is thus the great Postulate of the pure Practical Reason; we can be sure of the fact (although it must always remain speculatively undemonstrable), because else there could be no explanation of the Categorical Imperative of Duty. But inasmuch as the Practical Reason, besides enjoining a law of Duty, must provide also a final end of action in the idea of an unconditioned Supreme Good, it contains also two other Postulates: Man being a sentient as well as a rational being, Happiness as well as Perfect Virtue or Moral Perfection must enter into the Summum Bonum (not, one of them to the exclusion of the other, as the Stoics and Epicureans, in different senses, declared). Now, since there is no such necessary conjunction of the two in nature, it must be sought otherwise. It is found in postulating Immortality and God.
Immortality is required to render possible the attainment of moral perfection. Virtue out of respect for law, with a constant tendency to fall away, is all that is attainable in life. The Holiness, or complete accommodation of the will to the Moral Law, implied in the Summum Bonum, can be attained to only in the course of an infinite progression; which means personal Immortality. [As in the former case, the speculative impossibility of proving the immateriality, &c., of the supernatural soul is not here overcome; but Immortality is morally certain, being demanded by the Practical Reason.]
Moral perfection thus provided for, God must be postulated in order to find the ground of the required conjunction of Felicity. Happiness is the condition of the rational being in whose whole existence everything goes according to wish and will; and this is not the condition of man, for in him observance of the moral law is not conjoined with power of disposal over the laws of nature. But, as Practical Reason demands the conjunction, it is to be found only in a being who is the author at once of Nature and of the Moral Law; and this is God. [The same remark once more applies, that here what is obtained is a moral certainty of the existence of the Deity: the negative result of the Critique of the Pure (speculative) Reason abides what it was.]
We may now attempt to summarize this abstruse Ethical theory of Kant.