In this part of Section II, Kant summarizes his argument, showing in particular that each of the three variations of the categorical imperative is a variation, not an amending, of the general formulation.1 This summary serves both as a review of the “ethical interlude,” and also as a transition from the idea of the categorical imperative to that of autonomy. As such, little commentary will be required.
A. The General Formulation
We end where we began, with the idea of an unconditionally good will. The absolutely good will is incapable of doing evil; its maxims, when made into universal law, are never contradictory. So the supreme principle or law is this: always act on a maxim which you can at the same time will to be a universal law. This rule alone guarantees consistency in the will and so is a categorical imperative, the supreme law for the absolutely good will.
B. The First Variation: The Law of Nature
In view of the analogy between the validity of the will (as the law for possible actions) and Nature in its formal aspect (as the existence of all things in a unified system governed by universal laws), we can state the categorical imperative as follows: always act according to a maxim which you can at the same time consider to be a universal law of nature.
C. The Second Variation: Humanity as an End
Rational nature’s distinguishing characteristic is that it acts for an end. An end is the content of every act of will. The idea of the absolutely good will excludes any particular end which must be brought into being. As a thing good without qualification, the good will cannot be tied to any particular goal, for to do so would give the will only a relative value. The end proper to a good will cannot be brought into existence but must be self-existent. Such an idea can be expressed only negatively: we must never act against this end, but in every act of will treat it not merely as a means but also as an end, as the originator of all conceivable ends, the possessor of a possible absolutely good will. Only by conceiving the unconditional end to be the absolutely good will itself do we avoid the contradiction of subordinating the good will to some other entity. The law can be stated in two ways: always act according to a maxim which holds every rational nature (whether yourself or another) as an end in itself; or, always act according to a maxim which is universally valid for every rational being.
To say that in using means to any end I must regulate my maxim so that it will be universally valid law for every agent is the same as saying that rational nature, as the originator of ends, must be respected as the source of all maxims and never taken merely as a means. As such, rational nature must be treated as the ultimate standard for using any means: it must always be treated as an end also.
Kant’s approach here differs from his discussion in Chapter 7, but his meaning is clear enough. Instead of arguing that rational nature has inherent value, he would have us think of rational nature as the orginator of all maxims and not merely as their object. That is, since rational nature is always itself the legislator through its own maxims, it cannot be considered solely as the object of a maxim, i.e., merely a means for another’s purposes. What is an agent by its nature cannot be considered a mere object of action.
In his statement of the second variation Kant includes the idea of universality, the sole characteristic of all law. A comparison of the original statement of the second variation (in Chapter 7, Section 3, with the two statements expressed above will indicate the shift in emphasis from the concept of human nature as an end in itself (as the source of a possibly good will) to rational nature as the origin of all practical law. The shift in emphasis carries us further toward the concept of autonomy.
D. The Third Variation: The Autonomous Will in a Kingdom of Ends
It follows without question that every rational being must be able to consider himself an end in himself; and with regard to any law to which he is subject he must be able to see himself as making universal law. For the way he makes universal law through his maxims shows him to be an end in himself. It further follows that his dignity (prerogative) compared to mere things means that he must make his maxims accord with his own nature and that of other rational beings as legislative beings, i.e., as persons. In this way the world of rational being (mundus intelligibilis) may be seen as a kingdom of ends, since each person, as a member, makes his own universal law. Thus: every rational reing must act as if by his maxims he were always a lawmaking member in a kingdom of ends.
The formal principle of these maxims is: always act as if your maxims were also the universal law for all rational beings.
Here, as in the second variation above and implicitly in the first, Kant rephrases the general formulation after giving the variation in order to show the close relationship between the general formulation and its variations. When we consider a maxim as universal law, we see its analogy with the universality of the laws of nature (first variation). When we consider a rational nature, one inherently an agent, universalizing its own maxims, we conceive every agent as an end unto himself (second variation). And from the idea of every person legislating his own moral law, in common with every other rational being, we arrive at the community of persons as ends in themselves (third variation).
E. The Kingdom of Ends and the Kingdom of Nature
Consider the analogy between the kingdom of ends and the system of Nature. The former is based on maxims, or self-imposed policies, while the latter is organized by laws which describe causes from outside. By analogy, however, we can call the natural system a “kingdom” of Nature, in so far as Nature appears to be ordered toward rational beings as her end—even though Nature seems to work like a machine. If the maxims which the categorical imperative ordains as law for all rational beings were universally obeyed (as laws of Nature are), then the kingdom of ends would be a fact.
But even when a rational being conscientiously follows a maxim, he cannot reasonably expect all other rational beings to accept it. Neither can he expect that the orderly purpose of a kingdom of Nature will harmonize with his own efforts as a member of the kingdom of ends (which he himself creates) to promote his own happiness. But the law still binds categorically: always act according to the maxims of a universal lawmaking member in a possible kingdom of ends.
Kant presents the kingdom of ends as analogous to a system of nature. If the universe were composed solely of rational beings, the kingdom of nature and morality would coincide. But men do not always act purely from reason and the universe is not a moral realm. The principle difference, which is obvious, is that the laws of nature are always obeyed, so to speak, while the laws of men and morality are often violated. The Utopian felicity of a true kingdom of ends would come about if all men in every action obeyed the moral law, and if nature also harmonized with its purposes.
Herein lies a paradox, that respect for a mere idea—the idea of the dignity of man’s rational nature, excluding any advantages or goals to be achieved—can serve as an inflexible command to the will. In fact, this very independence of a maxim from all influences of self-interest constitutes the sublimity of the maxim and makes the rational agent worthy of membership in the kingdom of ends. If this were not so, man would simply be subject to the natural law of his own desires. Even if the kingdom of Nature and the kingdom of ends were joined into one system so that the kingdom of ends were no longer a mere Ideal but an actuality, this kingdom would surely become a strong impetus to choice, but it would gain no greater inherent value. The sole absolute lawgiver [God] would still have to judge a rational being’s worth on the basis of his ability to command himself to perform disinterested actions solely in view of the dignity of rational nature in man. Things do not lose their nature simply by changing their external relations; a man must be judged by his rational nature, not by how he may be related to other things, ineluding the one who judges him—even though his judge be the Supreme Being.
The point of Kant’s paradox is subtle, and in fact he is making two points. First, it is our daily experience that much of what we intend to accomplish does not happen for one reason or another, even when we try to do our duty. Even were the universe so fortuitous that all our moral intentions turned out successfully, our purposes would still not give our action moral value. Even though the good to be accomplished would have a strong influence on our choices, moral worth would still be based on the moral maxim to act as a self-legislating rational being. What is more (Kant’s second point), the Supreme Being who judges us all will determine the value of our lives—our worthiness to be happy—according to how we sought to achieve a good will, not by how much good we accomplished in our actions. For the actual success of our moral intentions is accidental to the moral value of the will from which such intentions spring. The paradox lies in the moral principle that while our motive must be always to perform our duties, the actual performance of our duties is less important than our motive, at least in the matter of a man’s moral worth as a rational being.
F. General Conclusion: the Definition of Morality
Morality can thus be defined as the relation of actions to the autonomy of the will, that is, to a will which by its maxims can make universal law. Whatever agrees with autonomy of the will is permitted; what conflicts with it is forbidden. A holy will is one whose maxims necessarily agree with the autonomous law; such a will is absolutely good. But a will less than absolutely good is subject to autonomous law (i.e., moral necessity) by what we call obligation. A holy will, consequently, cannot have any duties, since duty is the relationship between a will which is not absolutely compelled to obey the law and the objectively necessary action which is its duty.
We have finally arrived at the definition of morality. Necessity for any action must be grounded in reason; for moral reason alone can legislate that universal law which imposes practical necessity. Autonomy is the only condition by which reason determines its own universal law, which law in turn makes an action objectively necessary, i.e., obligatory. Consequently, we may define morality as the relationship of the self-legislating will to an action perceived as objectively necessary through its own self-legislation.
In the above paragraph, Kant shows that the moral law as expressed by the categorical imperative does not prescribe specific positive duties. Any action is morally permissible which is consistent with an autonomous will, that is, which is not in conflict with it. Thus I may indeed eat ice cream, dance, cheer at a football game, and ride a bicycle—unless any of these actions conflicts with some duty.
In view of all this, we can easily see how we think of duty as a subjection to law and still ascribe a kind of sublimity and dignity to the person who fulfills all his duties. His sublimity does not come from his being subject to the law, but from his being the law maker, who at the same time subjects himself to the law. We have seen also that neither fear of nor attraction to the law can give an action any moral value—only respect for the law can give that. Our own will, conceived ideally as acting under no other condition but making universal law by its maxims, is the proper object of respect. Thus, the dignity of a human being consists simply in his ability to make universal law to which he is also subject
We do not find our dignity simply in being subject to the law, but in being subject through our own legislation. Kant here restates the essential conditions for the dignity of a human being. The rational human being has dignity because he is autonomous. The autonomous will, however, has two characteristics: it is the lawmaker and at the same time it is subject to self-legislation. The moral law is self-imposed but it binds under moral necessity nevertheless. Indeed, only a self-imposed necessity could qualify as moral obligation.
autonomy of the will is
the supreme principle of morality
Autonomy of the will is that which makes the will a law to itself, without regard for any objects of desire. Thus the principie of autonomy: always choose according to maxims which you can think of as becoming universal laws by your choice. This practical rule is an imperative: the will of every rational being is necessarily subject to it. But we cannot prove this by a simple analysis of the concepts found in the rule, for the rule is a synthetic proposition. Proof would require that we go beyond knowledge of objects of choice to a critical investigation of the subject itself, pure practical reason; for if there be any synthetic proposition which commands us under necessity, we must be capable of knowing it purely a priori. Such a proof does not belong in this section. But we can show by a simple analysis of concepts that the principle of autonomy is the sole principle of morality. Such an analysis shows that the principle of morality must be a categorical imperative, and that such an imperative commands no more nor less than autonomy.
There is but one supreme principle, the autonomy of the rational will, but autonomy may be considered under two aspects:
1. From the viewpoint of the will as subject to its own law, the supreme principle is expressed as a categorical command of morality.
2. From the viewpoint of the will as making its own law, the supreme principle is expressed in the concept of autonomy. But in each expression, the supreme principle is grounded in the rational will, a will which makes its own law and which is subject to its own law.
Consequently we have two statements of the supreme principle, the first being the general formulation of the categorical imperative: Act only on that maxim which you can at the same time will to become a universal law. This emphasizes the universalizing of the maxim. The second statement of the supreme principle, given in the above paragraph, emphasizes the autonomy of the will: Always choose according to maxims which you can think of as becoming universal laws by your choice.
These two aspects also appear in the double meaning of “principle.” In one sense, a principle is a guiding rule, a standard, such as we find in a “principle of logic.” In this sense, the categorical imperative is seen to be the supreme guiding principle of morality, not in the sense of imposing particular duties, but as a standard for determining the moral status of particular maxims. In the second sense, a principle is a foundation, a ground; in this sense autonomy is the supreme principle or ultimate foundation of morality. Some moral philosophers stress the former sense, emphasizing ethical guidance. But Kant wanted primarily to establish the ultimate foundation of morality and so he emphasizes the latter sense.
Again Kant raises the question of the synthetic nature of the categorical imperative. By analyzing the concepts of will and law we cannot prove that the will is subject to the moral law, since subjection of will to law is not part of the meaning of will. But we can prove by such an analysis of concepts that the categorical imperative is the essence of moral law. Neither of these assertions is new. The first half of Section II (Chapters 5 and 6) is an analytical investigation showing that morality requires a categorical imperative. We know what this imperative must be if we assume that the will has the original capacity to act from duty. But analysis can take us only so far; we have not proved that the will has this capacity. Only by examining the will in itself (that is, by a critique of pure practical reason) can we determine that the will is subject to the law of duty and not merely to causal laws of inclination. Such an investigation must be made independently of experience: it must be purely a priori. The problem (not to be examined until Section III) still remains: to prove by a critical examination of practical reason that the will is subject to the law of duty, or, as Kant expresses the problem, to prove how a categorical imperative is possible.
heteronomy of the will is the source
of all spurious principles of morality
Whenever the will seeks its law somewhere other than in the fitness of its maxims to be self-made universal law, that is, whenever the will seeks law in the objects of choice rather than in the will itself, then the result will always be heteronomy. Whenever the will fails to give itself law, seeking instead the source of its rule in a desire for some objects or in some rational ideal, the derived imperative can only be hypothetical: “I ought to do this because I want that.” But the moral (categorical) imperative says: “I ought to do this or that even though I want nothing from doing it.” For instance, the hypothetical imperative might say, “If I want to keep my good name, I ought not to lie;” while the categorical imperative would say, “I ought not to lie, although I would never even be suspected of it.”
The moral law, therefore, must completely ignore all objects so that they have no influence whatsoever on the will; in this way practical reason (the will) will not support some outside interest but rather will exhibit its ultimate authority as the supreme lawmaker. For example, I ought to seek the happiness of others, not because their becoming happy is any concern of mine (either from some desire or from some rational satisfaction), but simply because a maxim which conflicts with the happiness of others cannot at the same time be willed as a universal law.
Heteronomy is the opposite of autonomy, describing any principle which is not based on the autonomy of the will. Kant’s position is unambiguous: whenever the will does not make its own law, then the will determines its actions from a heteronomous principle; if the will does not act from duty, then it must act from some inclination or self-interest; if the imperative directing an agent’s action is not categorical, then clearly it must be hypothetical; and most importantly, unless the guiding principle of the will be self-imposed, it cannot produce a good will.
Kant overstates his case. It would be virtually impossible to ignore all objects to the extent that they have no influence whatsoever on the will. It is quite enough, even for Kant’s own purpose, that the autonomous will allow no object to exceed (or even equal) the influence of the moral law. If objects had no influence at all, it would be impossible for me to perform my imperfect duties. How can I choose between giving a needy person bubble gum or a decent job if I am forbidden to evaluate the alternative results?
This particular passage, as well as others, leads some commentators to say that Kant would make it our duty to act from duty. But this is impossible, since it would lead to an infinite series of duties for each duty. I would have a duty2 to keep my promise, and I would also have the duty2 to keep my promise from duty!. But since the second-order duty is no less than the duty to keep my promise, I would also have the duty3 to keep-my-promise-from-duty1 from duty2. There would be no end to it. We can only conclude that we do not have any duty to act from duty. “Duty” refers to actions, not to motives.5
The matter in dispute concerns the necessary condition for moral credit. In order to receive moral credit for doing my duty, I must act without relying on inclinations of self-interest. I must perceive my action as my duty, as an act complying with universai law, and perform the action for this reason. Kant does suggest that we should try to act from duty, even if we do not have a duty to act from duty. It we reconsider his argument in Section I, that only the moral motive can guarantee moral credit, moral goodness demands that we strengthen the moral motive in every way possible. The more we act primarily from a sense of duty, the more likely we are to do our duty in situations which test the very core of moral fiber. On the other hand, the more we rely on inclination or self-interest to spur us to do our duty, the less likely we are to do our duty when self-interest conflicts with duty.
The will which rests on self-interest or desire is heteronomous. In fact, the will which rests on any ground but its own selflegislation is heteronomous. Such grounds Kant considers “spurious,” invalid foundations for a genuine moral law. To climax his argument thus far, he now will examine four major attempts to base morality on a principle other than autonomy.
catalog of all possible principles of
morality derived from heteronomy
Here, just as they do in all cases, philosophers who have not made a critical examination of human reason will try all possible wrong ways before they finally find the one correct way. From this point of view, all principles are either empirical or rational. Empirical principles are based on the idea of happiness and are related either to a physical sensation or to a moral feeling. Rational principles are based on some rule commanding perfection, being derived either from a rational ideal of perfection as a goal to seek or from the concept of the Will of God as an independent, perfecting norm to guide human choice.
Principles of morality must be grounded in either reason or experience. The principles based on experience may be divided into two subcategories:
a. Hedonistic, based upon some kind of pleasure principle; or,
b. Intuitive, allegedly justified by a moral “sense.” Rational principles (other than autonomy) are either:
c. Idealistic, derived from an abstract ideal of human perfection; or,
d. Theological, based upon some conception of God’s will. Kant’s criticism of any empirical principle is easily anticipated: none will justify the objective necessity of obligation—and without obligation there is no true moral law. Rational principles are empty and so have no practical value, even though from them we may derive a conception of necessity.
Although Kant explicitly thinks of Epicurus as the typical exponent of the hedonistic principle,6 readers will probably be more familiar with the modern equivalent, expressed in the utilitarian philosophy of John Stuart Mill (1806-1873). Mill is the best known of the moral philosophers who base morality on man’s natural quest for pleasure and his search for a life of happiness. Mill defines happiness in the following statement: “Actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness. By 'happiness' is intended pleasure, and the absence of pain; by unhappiness, pain, and the privation of pleasure.”7 We determine the right action by calculating which alternative will produce the "greatest happiness (= pleasure) for the greatest number.” Mill came to his conclusions because be believed some kind of pleasure principle to be the spring of all human activity. Kant strongly rejects the principle of hedonism.
I. Empirical principles never provide the basis for moral law, for moral law must hold for all rational beings without distinction. But whenever man’s human nature, or the particular situation which affects human nature, is taken as the foundation for moral law, such a “law” lacks universality, and the unconditional practical necessity of law is missing. The worst of these empirical principles is that of personal happiness, not merely because it is false, but because :
(a) experience belies the presumption that good conduct and happiness always go together;
(b) the happiness principle is no help to morality, since making a man happy and making him good are two quite different things, just as making him prudent and enlightened in his own self-interest is quite different from making him virtuous; but especially
(c) the happiness principle sets morality on a basis of selfinterest, which undermines and destroys its sublimity; for such a basis eliminates any real distinction between virtue and vice—the only difference being how well we can calculate what is to our advantage.
Though Kant gives three arguments against the hedonistic principle, his primary objection is that we cannot conclude from the fact that all men do seek pleasure that they ought to seek it. Principles derived from experiences of human nature, even if universally true, cannot provide the unconditional necessity required by the moral law. For a moral law must command independently of inclination, desire, or self-interest; often, indeed, it commands us to act in direct opposition to the pleasure principle.
Kant, moreover, flatly denies that pleasure is the spring of all human action: it is just not true that every human action is done only to gain this or that pleasure.9 But even if it were true, our common moral knowledge would lead us to reject hedonism as a basis for morality.
a. A person’s happiness is often disproportionate to his moral character, particularly if happiness be defined as a state of well-being or pleasure.
b. We do not make a man good by making him happy; often just the opposite is the case, especially if happiness is equated with pleasure.
c. Pleasure is notoriously a bad motive to suggest as a moral standard, for it instills into a man a selfish outlook which is contrary to the moral attitude. Expediency is not morality nor is prudence always virtue; sometimes it is simply cautious greed.
Although Kant mentions Francis Hutcheson (1694-1747) by name, the moral sense theory is popularly ascribed to David Hume (1711-1776), who believed the moral sense to be a kind of passional reaction to elements of a moral situation. Essentially, the moral sense theory is a kind of intuitionist ethics,10 asserting that there is in human experience a special moral faculty analogous to the sense of sight. Hume11 refers to it as a sentiment of approbation; Hutcheson specially calls it a moral sense.12 For both, the moral feeling is an experience of the moral tone of a situation. A moral agent somehow “sees” or “feels” the moral character of an action. And just as with any organic sense, the moral sense can be trained to a high sensitivity, enabling a man to make acutely discriminating moral judgments. Nevertheless the moral sense judgments spring from experience, and are merely a posteriori.
Kant’s basic opposition is the same: experience cannot be a foundation for necessity of any kind.
II. By contrast there is the moral sense, an alleged special feeling.* This is a theory without depth, appealing to thoughtless men who trust in their feelings, even in matters dealing with universal law. But feelings can differ almost infinitely from one person to another, and thus simply cannot furnish a uniform standard of good and evil. Furthermore, the theory ignores the fact that one man cannot validly judge others by his own feelings. Even so, the moral feeling is closer to morality and dignity, for it honors Virtue with direct and immediate satisfaction and esteem; unlike the happiness principle, it does not tell her to her face that her worth lies in her favors rather than her beauty.
* I classify the principle of moral feeling under the happiness principle because every empirical principle promises some benefit from agreeable results, either in an immediate satisfaction (e.g., pleasure), or in long-term gains (e.g., happiness). We must also, with Hutcheson, classify the principle of Sympathy (feeling for the happiness of others) under his assumed moral sense.
Feelings of whatever kind, says Kant, are subjective. Why should my feelings, or anyone’s feelings for that matter, serve as the standard for moral judgment? The moral feeling may be a universal capacity in man, but experience shows that men’s reactions to moral situations are frequently incompatible. Further, moral obligation, we know, must often hold counter to feeling. Nevertheless, Kant feels some affinity for this theory since it does emphasize the unique status of moral judgments and concepts and does not make morality a high-sounding name for expediency. Virtue is still admired for her own—albeit empirical—beauty.
Of the empirical principles of morality, one is based on selfinterest while the other is not. The two rational principles are similarly paired. The concept of perfection is not grounded in self-interest, but the theological ideal is. Kant discusses the ideal of perfection first.
III. Among the rational principles, we first see the ontological ideal of perfection. But this concept is empty; as a standard for selecting from the limitless universe of possibilities that group of attributes which would be most appropriate for us, this concept is completely useless. Furthermore, when we try to use this ideal to distinguish what is from what ought to be, we usually end up arguing in circles by sneaking into the argument that very morality we are trying to explain.
The expression “ontological ideal of perfection” means roughly the idealized condition of a particular thing. The perfection of man, for instance, might be expressed as a man perfectly rational, perfectly moral, perfectly capable of all human operations. Plato espoused such a theory,13 which holds up to our minds the concept of the Philosopher, the perfect man, whom we should imitäte. But this ideal is vacuous. “Perfect” depends for its meaning on the kind of thing talked about, as well as on some conception of the ideal condition of such a thing. There are “perfect” apples, cars, wills, etc., but there is no meaningful content to the idea of perfection by itself. If I were to tell someone simply, “Be perfeet,” what practical rule could he derive from my command alone? If he has any positive conception of the characteristics of the perfect man, he did not get them from my bare mention of perfection. In some way he had to supply any concrete content from some other source of information. But if that is so, then the idea of perfection by itself is no help as a moral principle. We must still add moral content to it.
Even though it is empty and thus useless, the concept of perfection is preferable to the concept of God’s Will as the foundation of morality,14 since the latter ideal is usually founded on personal concern for salvation, not on respect for moral law.
IV. Even so, the ideal of perfection is better than the theological ideal, which bases morality on the supremely perfect Divine Will. For we cannot know God’s perfect Will directly, but must deduce it from other ideas, of which morality is the foremost. But if we do not deduce God’s Will from other concepts (and this in itself would provide a most circular explanation), then the only idea we could have would be of a will which seeks glory for itself and dominion over others, a will characterized by awesome power and vengeance. Any ethical system based on this kind of will would contradict morality.
The nature of God’s Will cannot be known. What is more, we have no direct insight into God’s Will even as it is revealed in scripture. Kant does not reject the commands of God found in the Old and New Testaments; he simply rejects them as sources of moral law, since they cannot be known a priori. Even if God were to give me a direct command, the moral question still remains: ought I obey God’s command? It is not selfevident that it is my duty to do what God commands, and thus the actuality of an expressed divine command still does not provide us with the a priori grounds for moral judgment. Whenever someone suggests that an action has been commanded by God, we pre-judge the command according to moral principles derived from some other source before we will agree with the original suggestion. For example, we would reject completely the suggestion that God commands us to sacrifice children as a form of worship.
Still, if I had to choose between the ideal of a moral sense and that of perfection in general, I would decide for the latter. While neither of these ideals undermines morality—though they are wholly invalid grounds for it—the ideal of perfection at least keeps the question of morality free from sensibility, bringing it to the control of pure reason. And while this does not solve the problem, it does preserve free from corruption the undefined idea of a will good in itself, until an answer to the question can be found.
I see no point in a drawn out refutation of the rest of the theories. Being such an easy task, doing so here would be superfluous. What is more, such a refutation presumably is well understood by those who must choose the official theory—persons in their charge simply would not put up with any hesitation in judgment. But more important for us here is to recognize that all these principles propose nothing but heteronomy of the will as the ultimate foundation of morality. And so they necessarily fail to achieve their purpose.
Whenever we posit some object of desire as determining the rule which guides the will, we must judge the resultant rule to be heteronomy. Such an imperative is conditional: "If (or since) you want this object, you ought to do such and such,” a rule which cannot be a moral (categorical) command. No matter how the object may determine the will—whether through desire (such as the happiness principle) or through reason as it considers objects of will in general (such as the principle of perfection)—such a will never determines itself by thinking of the action, but rather determines itself under the influence of some impulse which the expected results arouse. In other words, “I ought to do this because I want that.”
But this demands a second law tying me as agent to the first imperative; that is, a law which not only commands that I desire such an object, but also that I will my maxim according to the moral standard. Whenever I consider an object which I can possibly obtain, my will is moved in a perfectly natural and human way; but this influence on my will depends on the natural constitution of my will. I might be moved by sensations, such as desire or taste, or by understanding and even reason. These impulses afford varying satisfactions as they move me to exercise my will. But as a consequence my natural constitution would determine the law, and such a law, since it can be known and proved only by experience, would be contingent. It could never serve as an apodictic practical rule, which the moral law must be. A law of this kind always arises from heteronomy of the will: the will does not make the law for itself, but accepts it from some external impulse arising from the particular tendencies of a responsive human nature.
Of any principle of morality we should ask the following questions. Does the principle base compliance on the satisfaction of some desire or on some interest of the agent? Does the principle rely on experience of any kind? Does the agent presuppose any other moral standard in using the principle? If the answer to any of these questions is yes, the principle is spurious, for its foundation is contrary to the idea of the self-legislation of the will. Of all these heteronomous principles, Kant prefers the ideal of perfection. But being an empty concept, representing merely the theoretical limit of human excellence, it has no practical value as a basis for moral law.
Kant concludes Section II with a precise statement of the supreme principle of morality.
Therefore, the absolutely good will, which must have a categorical imperative as its principle, will not be determined by any particular object. Its essence contains only the general form of willing—autonomy. Thus: the sole basis for that law which the will of every rational being imposes on itself is the fitness of the maxims of every good will to become universal laws, without recourse to some impulse of self-interest.
The ultimate foundation of morality is the good will, which is an autonomous will; and the supreme law of morality commands us to act as a good will acts, namely, in a way which recognizes the lawmaking of its maxims.
Kant’s search for the supreme principle of morality, the ultimate ground of obligation, is now finished. The outline of his argument is this:
1. Only that which has absolute value in itself can serve as the basis of obligation, since only such an entity ought to be sought universally and unconditionally.
2. The good will is the only thing which has absolute value.
3. Thus the good will (i.e., the moral will) is the foundation of obligation expressed by the moral law.
4. One acts with a good will when he acts solely out of respect for moral law, without regard for self-interest.
5. As the basis of all obligation, the good will must be the source of its own moral law; for were it subject to some external law, it could be subject only through self-interest or natural inclination.
6. The good will imposes its own moral law on itself when it wills that the maxims of its actions could become universal law for itself and every rational being.
7. The supreme moral imperative, then, contains only the form of law: universality. The statement of this law is the categorical imperative.
Kant’s argument then concludes with the statement: “The essence of the absolutely good will contains only the general form of willing—autonomy “ Having begun with the concept of the only absolutely good thing, he has developed it by seeking its necessary foundation, finding it in autonomy. The moral will, the good will, the autonomous will—these are only different ways of conceiving the one ultimate basis of moral obligation. Further, morality imposes valid constraint on the will of a rational being if and only if the will has imposed upon itself the moral necessity of acting.
However, the categorical imperative has not yet been shown to be a genuine moral fact. It is one thing to analyze concepts; it is quite another to prove that there is something to which the concept applies. From an analysis of concepts we know, first, that autonomy is necessary if the will can truly act from duty, and second that if the will is capable of self-legislation, then it is subject to the moral law. But neither experience nor analytical reasoning can give us any information as to the actual validity of the moral law, since neither experience nor analytical reasoning can prove that the will has freedom of choice.
How this synthetic practical a priori proposition is possible, and why it is necessary, are problems which cannot be solved within the system of a metaphysics of morals. We have not solved them, nor have we even claimed to have a solution. We simply showed that by developing the commonsense concept of morality, we found autonomy of the will inextricably connected to it—indeed, we found that autonomy is the very basis of morality. Thus anyone who takes morality to be real and not a mere fantasy must at the same time admit this principle of autonomy. Section II, then, like Section I, has been an analytical investigation.
If we are to prove that morality is no mere fantasy—which is surely proved if the categorical imperative and autonomy of the will are true, i.e., are absolutely necessary a priori principles—we must show that a synthetic use of pure practical reason is possible. However, we cannot do this without first undertaking a critical investigation of the faculty of reason. In Section III, then, we shall outline the main steps of such an investigation, which will suffice for the purpose of this Foundation.
Although we have discovered the ultimate foundation of morality in the autonomy of the will, we have not yet proved that the will of man is actually self-legislative. Section III, then, will be devoted to an examination of this fundamental issue of morality, the freedom of the will. For if the will of man is not free, then moral obligation is truly a mere fantasy. If Kant can prove the real, and not merely theoretical, possibility of human freedom, he will have shown the real authority of the categorical imperative, and to that extent the validity of the moral law.