Kant’s final task is to prove a priori that the point of view from the world of thought is valid for a being with will; that is, that a rational being necessarily takes that point of view when he wills. Without an a priori justification of autonomy, Kant’s carefully wrought system of morality will collapse into heteronomy; and the necessity of moral obligation based on autonomy will turn out to be nothing more than causal necessity of desires and inclinations. We cannot rely on the merely logical possibility of autonomy. We must provide an a priori defense of the categorical imperative by proving that the point of view from the world of pure thought is necessary, at least in a practical sense. But how can we prove that every rational being with a will must conceive of himself as a member of the world of thought and so subject to the law of autonomy? Does a man, simply because he is a rational being with a will, necessarily conceive himself subject to the law of freedom, duty-bound by the categorical imperative? This is the crucial question which yet remains to be answered.
how is a categorical imperative possible?
Because he is a thinking being, every rational being counts himself as a member of the world of thought; and because he sees himself as an efficient cause in that world of thought, he can think of his causality as the exercise of will. On the other hand, he is also aware that he is a member of the world of sense impressions, in which his actions seem to be merely events in a causal sequence. If we ask how the will, as a causal agent in the world of thought, can produce these actions, we cannot find any answer since we have no knowledge of that world; rather, all causal explanations must rely on appearances in the world of sense impressions (e.g., desires and inclinations). If I were simply a member of the world of thought, all my actions would conform completely to the law of autonomy (the supreme principle of morality); or if I were simply a member of the world of sense impressions, all my actions would conform completely to the natural causality of desires and inclinations—in short, to the law of heteronomy (the natural principle of happiness ).
But the world of thought contains that reality which under girds reality and law in the world of sense impressions. As a thinking being, I can think of my will only as a member of the world of thought, and so I must recognize that the law which governs the world of thought, which is the law of autonomy, governs my will directly. At the same time that I know myself as a member of the world of sense impressions (and subject to laws of natural causality), I must also think of myself as subject to that law of reason which, in the world of thought, contains the idea of freedom. Thus the law of the world of thought becomes an imperative for me, and actions which conform to this law become my duties.
In this argument Kant seeks to prove that a rational being necessarily conceives himself subject to the categorical imperative. Let us examine a more detailed statement of his argument.
K1. The world of thought is the objective reality which supports the sensible reality of the world of sense impressions.
K2. Since laws of appearance govern the world of sense impressions, laws of thought must govern the world of thought.
K3. Every time I exercise my will I must think myself to be a member of the world of thought.
K4. Thus, every time I exercise my will, I must think myself subject to the law governing the world of thought, the law of autonomy.
K5. Therefore, every time I exercise my will, I must think myself subject to the law of autonomy, expressed in the categorical imperative.
K6. Therefore, the categorical imperative is valid as a law governing my will. Q.E.D.
Thus Kant completes the argument of the Foundation. Not only is the categorical imperative possible, he has argued, but it governs all human activity by imposing moral obligation. The good will, the only absolutely good thing, is within the grasp of every man. If the argument is satisfactory, Kant has proved the validity of the moral command of reason by showing that a free will is subject to a priori law. We cannot then avoid a careful, critical analysis of this crucial argument.
To begin with, we can determine that the conclusion K6 follows from K5, for if I must think of myself as subject to the categorical imperative, then it does govern my will and so is valid for me as law. Furthermore, we can see that K5 follows from K4, for if the law of autonomy is the law governing the world of thought and if I must think myself a member of that world, then I must think myself subject to that law. In addition, K4 follows from K3, for if I must think myself a member of the world of thought whenever I will, then I must think myself subject to the laws of the world of thought. If K1 is true, then K2 must be true (or so Kant believes), for law must have some foundation for its validity and necessity other than itself.1 He believes that if there is law in the world of appearances—and there is—then it should be “supported” by a law in the world of objective reality. Now, if I am a member of this objective world, I am subject to its laws and whenever I will I must think myself a member. In exercising my will I necessarily think myself a member of the world of thought, and this thought relates the concept of law in the world of pure thought to the concept of my being subject to that law. K3 is the statement of Kants “third idea,” which carries the argument from K2 to K4, and so to the conclusion, K6.
Unfortunately, analysis of this argument reveals three weakness es. The first is that the fact of law in the world of sense impressions does not demand a basis in law. True, every law demands some ground or other, but only an existent reality of some kind; the reality need not itself be a law nor even be governed by law. As a consequence, we cannot deduce K2 from K1. A second and more serious difficulty is that we cannot even assert K1 with any assurance: we do not know at all what “supports” the world of appearance. We can think of the objective world as a foundation for the world of sensible appearances, but the thought does not permit us to assert that such a world exists and even less than it “supports” the world of appearance. Without K1 and K2, however, we have no grounds for thinking ourselves subject to any laws of the world of thought. Even if we prove a priori that every time I exercise my will I must think myself a member of the world of thought (which K3 asserts), we have not thereby proved that we must think ourselves governed by any kind of law. K1 and K2 were necessary to establish the validity of the law in the world of thought but, since they fail to do so, the validity of such law and our subjection to it are still unproved. The third fault is that the argument simply assumes the truth of K3. But as we saw clearly at the close of Chapter 9, K3 is the crucial premiss of the proof for freedom of the will. It also expresses Kant’s “third idea,” without which we cannot establish a necessary connection between exercising our wills and acting as moral agents subject to the law of duty. Kant’s argument, which set out to prove this connection necessary, actually assumes it as a premiss: thus his argument begs the question and is a failure.
Since the remainder of Section III shows clearly that Kant was aware of his failure to prove the validity of the moral law, we can legitimately ask why he bothered to present an argument at all. The first and most probable reason is that he wished to point out as accurately as possible the focal problem and the impossibility of solving it. Indeed, the remainder of the Foundation is a kind of apologetic explanation of this impossibility. Secondly, since Kant does insist that K3 is a necessary presupposition, so far as it connects willing with moral obligation, he may have wished to suggest indirectly that this synthetic union of the two ideas cannot itself be proved, but rather is an ultimate synthetic principle of a system which can be deduced analytically from it.
But the most troublesome aspect of his proof, which Kant apparently does not appreciate, is that by begging the question in his argument, he does not even justify our taking the synthetic connection as a necessary presupposition. The validity of the moral law rests squarely on the supposed connection between the fact of willing and law in the world of thought. We can still question whether morality is a sham, a mere stage setting creating the illusion of a good will without its substance. As a defense, let us ask another question: could Kant have given an argument which, although it would not prove the objective reality of the moral law, at least would prove that the law of autonomy is valid as a necessary practical law, which for practical reason would be all the proof we require? I believe he could have; and in light of comments he makes later in this section, I will outline briefly how such an argument might be developed.
The categorical imperative is possible because the idea of freedom makes me a member of the world of thought. If this world were the only world I belonged to, all my actions would invariably conform to the autonomy of the will. But I also perceive that I am at the same time a member of the world of sense impressions—and so my actions in the world of sense impressions ought to conform to the law of autonomy. This categorical “ought” presents us with a synthetic a priori practical proposition: the idea of my will as influenced by sensuous desires is joined to the idea of that same will as a pure, self-directing member of the world of thought. In the idea of the autonomous will we find the ultimate law of reason which governs the will as it appears in the world of sensuous influences.*
* This is similar to the way in which concepts of the understanding, which in themselves are only the forms of law in general, are joined to our immediate impressions of the world of sense and, in this way, make up the synthetic a priori propositions which structure all our knowledge of the universe.
We find the key in the sentence: “This categorical 'ought' presents us with a synthetic a priori practical proposition.” This suggests that a person who understands the practical implication of the moral “ought” does in fact recognize a ground for this “ought,” a ground which cannot be derived from the world of sensuous desires. A world governed solely by laws of causal necessity cannot produce a meaningful “ought.” If my whole realm of thinking were derived only from the world of appearances, any understanding of the moral “ought” as indicating a possible alternative to natural causal necessity would be utterly impossible. Thus by the mere fact that I can understand the meaning of “ought” I must necessarily take the point of view of myself as a member of the world of thought, and thus as a free agent whom the “ought” commands. Should I simply choose this point of view as an imaginative option, I would not thereby recognize an “ought” of moral law; but since I do understand and recognize the “ought” as a moral imperative, I must necessarily see myself subject to the laws of freedom.
The cardinal issue in this argument is not any recognition of a law necessary for a world of thought, nor the mere awareness of acting as a free agent. Rather, it is the assertion of free will subject to law as a necessary precondition for an understanding of the moral term “ought.” Let us put the argument into a rigorous form.
L1. If all understanding came solely from the world of sense impressions, no rational being could understand the implications of the moral “ought.”
L2. Every rational being who understands the implications of the moral “ought” necessarily takes a point of view different from that of the world of sense impressions.
L3. Anyone who takes the point of view other than that of the world of sense impressions with its law of causality necessarily thinks himself an agent who can act independently of the law of causality: he thinks of himself as a free agent.
L4. Since “ought” signifies the will’s awareness of its relationship to an objectively necessary action, which only a law can determine, a rational agent must think of himself subject to whatever law determines the action to be objectively necessary: that is, he must think himself subject to the law of freedom expressed in the categorical imperative.
L5. Therefore, any rational human being who understands the implications of the moral “ought” necessarily thinks himself subject to the categorical imperative.
L6. Therefore, the categorical imperative is a valid moral law for anyone who understands what “ought” means.
The advantage of this argument lies in its initial premiss, unspoken in the argument: that we do in fact understand what “ought” means. In order to understand this at all we must think of ourselves in some special way, a way quite different from our membership in the world of causal necessity. But once we think of ourselves as independent of this sensuous world, the categorical imperative becomes a valid command of reason, since some law must relate the will to an objectively necessary action, and the only law which remains is that principle which commands the will to act as a free agent—the law of autonomy. The law is therefore valid, not because we have proved that human beings have free will, but because we have proved that we must think ourselves free agents subject to the law of free agency. More proof than this we cannot ask, could not find—and do not need. The moral law is valid for all practical purposes; but then it is valid, since it need be valid only for practical purposes.2
Of course, we are still creatures of desire, subject as well to the laws of natural inclination, which exert at times an almost overwhelming influence on our wills. Yet no matter how compelling the pull of desire and inclination, so long as we exercise our reason when we act, we are subject to the law of morality. Moreover, we are always aware of this, Kant insists; no matter how debauched we may be, we still respect the moral law.
The practical use of ordinary human reason supports this argument. No one, not even the most hardened criminal—assuming that he uses his reason in other ways—will fail to wish that he could be a person honest in his purpose, faithful to good maxims, sympathetic and generally benevolent, even when it takes great sacrifice, disadvantage, and discomfort. His desires and inclinations prevent him from acting in that way, but he would like to be free of those impulses which weigh so heavily upon him.
Man shows here that he takes himself in thought to a realm in which his will is free, a world totally different from the sensible world of desire. He is aware that he could not gratify his desires in the world of freedom, nor would such a world satisfy his genuine or imaginary needs: this anticipation would destroy the supreme position of that very idea of freedom which makes him wish for virtue. He could only expect to become a person of greater inner worth. The idea of freedom forces him to transfer himself to membership in the world of thought, a world in which he is free from the natural causality of the world of sense impressions. In this world of freedom he becomes aware of a good will which by his own admission stands judge over the bad will he has as a member of the sensuous world. Even while he transgresses the moral law he recognizes its authority. The moral law is his own will as a member of the world of thought; by thinking of himself at the same time as a member of the world of sense, he thinks this law as an “ought.”
In more modern terms, so long as a man knows the difference between right and wrong he somehow wishes that he were virtuous, even when his life and habits show not the least evidence of it. He still respects the law and the good will as having an unqualified value, even though he knows well that living by the moral law in pursuit of the good will could not gain him the satisfaction and pleasures which his life of immorality occasionally offers. As long as a man must believe that he acts from his own choosing—whether or not this be so—he must respect the law expressed in the moral “ought.”
The essential proof, for better or worse, has been given. The remainder of Section III examines two related questions, the one regarding the conflict between the two points of view and the other explaining why an absolute theoretical proof of freedom is not possible. The first question asks : how can two entirely different sets of laws, both governing the will, be equally valid for the same act of will?
the extreme limit of practical philosophy
Every man thinks of himself as a free agent. This way of thinking leads us to judge that some actions ought to have been done, even though they were not done. The idea of freedom cannot be derived from experience; experience may reveal demands which conflict head-on with the a priori commands arising necessarily from the idea of freedom, yet the commands of freedom still hold. On the other hand, it is equally necessary that everything that happens should be completely determined by natural causal law; thus the causal law is also a priori, since it involves the idea of necessity. Experience confirms the idea of Nature as a system, because experience (as knowledge of objects given through sensations and interrelated by universal laws) necessarily rests on the validity of such a priori laws. Freedom, then, is only an Idea of Reason which cannot substantiate its own objective reality. Nature, on the other hand, is a concept of the understanding, which can prove itself objectively valid by its application to and confirmation by events in experience.
We have seen in Chapter 9 that the understanding produces knowledge of the natural universe by ordering sensations to conform to fundamental a priori principles. These principles have already been proved valid by an appeal to reason (in the Transcendental Deduction of the first Critique) and confirmed by experience. Ideas of reason, on the other hand, have no such justification and cannot be proved valid as principles of knowledge. How can it be then that an unjustified Idea of practical reason can take precedence over a thoroughly justified concept of the understanding?
From this there arises a dialectic of reason: the freedom we attribute to the will seems to contradict Nature’s causal necessity. For scientific knowledge, reason relies on the idea of natural causal necessity, since only in this way can it explain facts of experience; but for the sake of its own agency reason must rely solely on the idea of freedom, since only with this idea can practical reason (will) be considered a cause of action.
Neither the cleverest philosophy nor commonsense can argue freedom away. But if the idea of freedom should contradict either itself or Nature, the already established a priori validity of Nature’s causal necessity would take precedence and the idea of freedom would have to be abandoned. Since we cannot abandon the idea of freedom any more than that of natural necessity, we must demand that moral philosophy somehow avoid the apparent contradiction between freedom and natural necessity in human action.
This contradiction would be forced upon us if, when we choose to act, we think ourselves subject to natural causality in the same way and in the same relationship as we consider ourselves free agents. The inescapable task of theoretical philosophy must be to explain how this contradiction is illusory: that the illusion arises when we fail to take a different point of view and affirm a different relationship between the agent and his action when we think of the agent as free, and when we think of him as a part of Nature and subject to her laws.
Furthermore, philosophy must explain not only how we can regard a single agent from both points of view at the same time, but why we must do so. For if we could not explain why we must necessarily take both points of view, the mere fact that both could be combined without contradiction would not give us sufficient leave to trouble theoretical reason with such an embarrassing difficulty. Theoretical philosophy must do its duty, however, in order to clear the way for practical philosophy. The philosopher has no choice but to erase the apparent contradiction. If he should leave the theory alone, it would be a bonum vacans, an abandoned dwelling, which a fatalist might rightly occupy, driving morality from its domain as an unwarranted intruder.
But this would not bring us to the border of practical philosophy. In solving the above controversy, practical philosophy can play no part; it only insists that theoretical philosophy pull free from the tangles of theoretical questions, so that practical reason itself may find peace and security from outside attacks which might contest the ground on which it wants to build.
The dual point of view saves us from contradiction: the two sets of laws, natural causality and freedom, do not conflict so long as we continue to take both points of view. The two worlds have no point of contact whereat the law of one world could suddenly become applicable in the other. But practical reason, since it operates from both points of view, cannot itself explain or justify them. That is the task of metaphysical (theoretical) reasoning. In the following paragraphs, Kant reviews the operation of will in this dualistic situation of agency.
Even common reason claims title to freedom of the will, based on its awareness and assumption of reason’s independence from purely subjective causal influences, all of which belong to sensation and combine to form the world of appearance. When a man thinks of himself as an intelligent being with a will (a causal agent ), he puts himself into a system of causal relationships entirely different from that other system of sensible appearances to which he also belongs, one in which his outward actions are determined by natural necessity. There is no contradiction at all in saying that an object as it appears in the sensible world is subject to laws governing appearances, and that the same object as it exists in itself may be independent of those laws.
Consequently man does and must think of himself as existing simultaneously in both systems: on the one hand, he is conscious of himself as a being influenced by sensations, so he is subject to the laws governing appearances; on the other hand, he is conscious of himself as an intelligent being, a member of a world of thought which is independent of the laws governing appearances. On these grounds man claims to have a will which can ignore desires and inclinations, and by doing so can-indeed must-think of its actions as results of its own agency. He is the cause of his actions because he is an intelligent being; through his actions he produces effects according to the principles governing the world of thought, which principles he knows only by pure reason, independently of experience.
Furthermore, since he conceives his own true nature as an intelligent being only by considering himself in the world of thought-for he knows his own human nature through experience-these principles rule him directly and categorically. Not even the allurements, influences, and temptations of the entire natural world of sense impressions can weaken in the least the laws governing his will as an intelligent being. He does not even hold himself responsible for these influences and desires for they do not arise from his true nature as will; he is responsible only for letting them influence his maxims to the extent that he violates the rational laws governing the will.
If knowledge of the world of thought is impossible, can practical reason find there the foundation for its law? Would this not constitute an illegitimate inference for reason to make, one which would end in an antinomy, an unavoidable contradiction, of practical reason? No, Kant would reply, for our idea of the world of pure thought is a negative idea, containing in itself only the characteristic of being independent of causal necessity. It gains positive value when we think ourselves as members of this world.
Practical reason does not overstep its limitations by thinking itself in the world of thought, although it would do so if it tried to enter this world through intuition or sensation. As a mere negation of the world of sense impressions, the world of thought does not give reason any laws which govern the will. The only positive element we find there is that the negative characteristic of freedom is joined with a positive one, the will as a causal agent. The will is the ability to act according to principles which comply with the essential condition of rational causality: that is, principles or maxims which have the validity of universal law.
But if practical reason tried to find an object of will (a motive) in the world of thought, it would exceed its limitations and pretend to be acquainted with something which it actually cannot know at all. The idea of a world of thought is simply the point of view outside the world of sense impressions which reason must take in order to think of itself as practical. Such a point of view would be impossible if the influences of sense impressions completely determined man. As is, this point of view is necessary in order for man to be aware of himself as an intelligent being, a rational cause, following the dictates of reason, and acting freely.
This point of view certainly involves the idea of an order and system of laws different from that of a mechanical world of sense impressions; it necessitates the idea of a world of thought populated by rational beings existing as things in themselves. But this point of view gives us nothing more than the formal condition of the world of thought, namely, that maxims of the will be valid as universal law and thus conform to the will’s autonomy, which is the only way will can consistently exercise its freedom. For any law which aims at some object is heteronomous, a law of Nature which applies only in the world of sense impressions.
Practical reason cannot legitimately look to the world of thought for some incentive or goal to aid the will in complying with the moral imperative. Not only is such a quest impossible—the world of thought cannot be an object of such knowledge—but further, even if it were possible, by finding such an incentive or goal, reason would thereby negate the law of morality and nullify the moral motive, which is to act for dutys sake alone. Paradoxical as it may seem, the moral law is valid because practical reason cannot know the world of thought.
To ask how reason can be practical is to ask how a human will, living in a determined universe of appearances which is governed through and through by laws of causal necessity, can be a free agent in its activity. But in order to explain how reason is practical, we require a set of laws which govern the will, by means of which we can give the explanation demanded. But the only law we know in the world of thought is the law of freedom, which presupposes the freedom of will and consequently cannot in turn explain this freedom. A presupposition, however necessary it may be, cannot justify the principles which require it; on the contrary, the presupposition itself must be justified by those principles. Because we must think of ourselves as members of the world of thought, we must presuppose freedom as the foundation for this world; otherwise we could not even conceive of this world in any meaningful sense. Or to put it briefly: any will which thinks of itself as an agent must presuppose that condition which makes agency possible, namely, freedom.
If reason tried to explain how pure reason can he practical, it would certainly exceed its limitations. But this is the same as trying to explain how freedom is possi- hie. In order to explain anything, we must subsume it under some law whose object can be experienced. But freedom is a mere idea; we cannot explain its objective validity by some law of Nature or through some possible experience. No example or analogy will help us to comprehend or even imagine how freedom is possible. The idea of freedom has only one kind of validity: it is a necessary presupposition for a rational being which is aware of its own will as the ability to determine itself to act as an intelligent being according to laws of reason, independently of desires which are subject to Nature’s direction. But if we cannot explain freedom by the laws of Nature, we cannot explain it at all.
But we can still defend freedom by refuting the arguments of those who claim a special insight into the essences of things and boldly assert that freedom is impossible. We can show them that the apparent contradiction which they point to rests on their thinking of man as necessarily an object of appearance, so that the laws of Nature can explain human activity. When they consider man as an intelligent being, they should likewise think of him as a thing in itself—but no, they persist in making him an object of appearance there as well. According to this view, to say that man’s will is a free cause in a world of determined natural causes is an obvious contradiction. But if they are reasonable and are willing to reconsider, they must grant that things in themselves must exist “behind” the appearances as a hidden support; and when they recognize that the laws which govern appearances cannot be extended to govern things in themselves, the apparent contradiction fades away.
The impossibility of proving that the will is free has one beneficial result at least: no one can prove that will is not free. The attempts by determinists to deny genuine freedom have for the most part relied on the law of cause and effect, according to which every event, whether physical or psychical, is what it is because some prior states or events were what they were. Objective freedom, they claim, is simply a self-contradictory idea, that a being subject to causality can act independently of causality. Of course, implicit in this argument is the singularity of the determinist’s point of view; he sees the will simply as an object in the world as we know it, the world of appearances. From this point of view alone the will appears completely determined. It is only when we also think of ourselves as members of another world that freedom becomes possible—yet we know nothing else of this other world. Consequently, there is nothing else known which could possibly conflict with freedom in the world of pure thought.5
If I know the world of thought only as a world in which I think myself a free agent, then nothing in this world can attract me, nothing can influence my will, nothing can appear as an incentive, nothing can interest me—other than the law itself. And yet interest seems to demand some basis other than its own mere existence. Why do I find this interest in moral law? How is it that I find a deep satisfaction in obeying the call of duty, particularly when duty commands me with no promise whatsoever of satisfaction of any kind?
Just as we find it impossible to explain freedom of the will, so too we find it impossible to locate and explain the kind of interest* we take in moral law. But we do take an interest in it, relating this interest to a moral feeling. Some moral philosophers have mistaken this feeling for the standard of our moral judgments; rather it should be seen as a subjective effect of the law on a will whose objective principles are found in reason alone.
* Interest makes reason practical; it is a cause which determines the will. Only rational beings can take an interest in something; irrational beings merely feel sensuous impulses. When the objective validity of the law suffices to determine the will, we say that reason takes a direct, pure interest in the action. But when reason can determine the will only by an appeal to some object of desire or to a particular feeling in the agent, we call this an indirect interest. Since reason by itself cannot discover any such object or feeling outside of experience, indirect interest is merely empirical, not a pure, rational interest. The logical interest which reason takes in unifying the many objects of experience into a meaningful system cannot be a direct interest, but must rest on the presupposition that some purpose will be served by the exercise.
If reason alone can command a rational will (which is influenced by sensuous desires) to act as it ought to act, then reason must be able to instill a feeling of pleasure or satisfaction in the performance of duty. This means that reason must be able to arouse sensuous influences in support of its own principies. But we simply cannot explain a priori how a mere idea which is completely empty of any empirical content can arouse a feeling of pleasure or displeasure. This presents a special instance of the general law of cause and effect and can be known not a priori, but only through our experience of the relationship between two objects in appearance.
How then can we hope to explain why the universality of the maxim as moral law can interest us? We would have to explain how a pure idea of reason, which gives no object to experience, can produce an effect in the world of experience. And this is impossible for human intelligence. One thing we do know: the law is not valid because it interests us; this would be heteronomy, basing practical reason on a sensuous feeling, which will never produce moral law. Rather, the law interests us because it is valid, because it arises from the will of intelligent being, which we take to be our own true nature. In this way, whatever characterizes the will in the world of mere appearances is necessarily subordinated to the true nature of the will in itself.
Could it be that the interest we find in the moral law conveys to us in some unfathomable way a feeling of true rationality, of progress towards the achievement of a good will, which we somehow recognize as the truly unqualified good? Certainly when we obey the call of duty we find a deep, intangible satisfaction which can exceed even the most intense earthy delight—and yet we cannot explain it.
Granted that the idea of freedom is a presupposition of practical reason, is there possibly some line of reasoning which might help to establish the validity of this idea as a presupposition? For example, in what sense can we call the idea of freedom a necessary presupposition? Is its necessity as a presupposition anything more than a merely logical relationship? And would some other presupposition serve as well?
Thus the question, “How is a categorical imperative possible?” can be answered to the extent that we can state the only presupposition which makes it possible: the idea of freedom. We can also see how this presupposition is a necessary and sufficient condition for the practical exercise of reason, convincing us of the validity of the categorical imperative and of the moral law. But human reason cannot possibly prove the truth of this presupposition.
Once we grant this presupposition—that an intelligent being has free will—autonomy necessarily follows as the sole formal condition for determining this will. The possible truth of this presupposition can easily be proved by theoretical philosophy, since it involves nothing contradictory to the laws of natural causality which connect appearances in the world of sense impressions. But what is more, we can prove its practical necessity: freedom is an unconditional principle which must be presupposed by any rational being aware of its own causality as will acting independently of desire; that is, we presuppose the idea of reason as the underlying condition for all our acts of will.
The idea of freedom is undoubtedly a logical possibility; it is not self-contradictory, nor does it contradict any other known truth—so long as we adhere to the two points of view. This is a merely negative appraisal of truth, however; it is a minimum condition which every judgment must meet before any further consideration will be given it. From a practical standpoint, on the other hand, freedom can be proved as a necessary presupposition for every exercise of rational will. This does not mean that a person is forced to acknowledge the truth of free will simply because he wills. Rather, free will must be assumed at least implicitly by a person who wills, because if he thought of himself as a mere effect, he would not think of himself as a causal agent and so he would not bother to will at all. Willing demands a ground in freedom (as its ratio essendi ), even though it does not demand that the fact of freedom be recognized as a necessary truly. We do not assume free will in order to prove that man exercises will; on the contrary, willing is a datum of experience, but it is a datum totally unexplainable by means of causal necessity. Consequently, either our capacity to will is absolutely unexplainable or we must presuppose freedom as that condition which explains our ability to will.
That freedom is compatible with the given experience of willing does not prove freedom true, but only that freedom is the sole presupposition which can make willing possible—or else we hold the datum itself to be a hallucination. Common sense, however, forbids our classifying as hallucinatory an experience as universal and commonplace as exercising our capacity to will. Therefore, from the fact that I will, reason demands that I will as a free agent, demands that freedom be presupposed as the objective reality which “supports” my capacity to will—even though reason cannot prove that freedom is a fact of objective reality. If I will, I necessarily will as if my freedom were a proven fact; unless I will under this implicit assumption, I do not will at all. On this basis, then, because I will as a self-conscious, rational being, freedom must be a necessary presupposition of practical reason.
But if we seek to answer the question: (a) how can pure reason be practical without relying on motives drawn from somewhere else; or (b) how can the bare principle of the universai validity of all its maxims as laws (surely the only way pure practical reason gives commands) arouse a purely moral motive and interest without depending on some object of desire; or, to put it simply, (c) how can pure reason be practical?—then human reason fails completely and all the effort and work of looking for answers is futile.
I would again waste my time if I tried to find out how freedom itself is possible as the causality of a will. I would have to abandon the empirical basis for expianation—but it is the only one I have. Yes, I could gambol about in the world of thought, the world of pure ideas, for I do have good ground for these ideas. But I have not the least knowledge of the world, nor can I ever hope to gain any, no matter how hard I exert my natural powers of reason. The world of thought is only an abstract idea which remains after I have eliminated all sensuous influences on the will. This idea permits me to keep the moral motive distinct from the world of appearances, showing that it does not contain absolutely everything, that something more must exist outside it.
Yet of this “more” I have no knowledge at all. Once I exclude all material content from pure reason which frames this ideal, taking away all its knowledge of objects, I have left only its form, the practical law that all its maxims must be universally valid. By conforming to this ideal, reason can be related to a world of pure thought as a possible efficient cause which governs the will. No other motive is allowed except this idea of a world of thought in which reason itself provides the motive. But how to explain it any further is the very problem we cannot solve.
I cannot doubt that I am exercising my will when I will. The exercise of will is a fact of experience. Freedom must be presupposed as the necessary condition to account for it. But I am wasting my time if I try to explain how the mere idea of freedom can produce an event in the world of sense impressions, particularly an event which cannot be explained by means of freedom and yet cannot be explained without it. This is the paradox of free will: the causal laws of the world of appearance cannot explain will, since through causal necessity no explanation of will is possible, and yet causal necessity is the only way we have for explaining events. On the other hand, the experience of willing cannot be explained without the presupposition of freedom; but the idea of freedom cannot further explain how a rational agent can will, since the idea of freedom does not involve a system of causal laws. That the idea of freedom is a necessary presupposition for a rational will may justify the validity of the moral law, but it explains no fact of human experience whatsoever.
We have thus reached the limit of all moral inquiry. Yet it was important that we find this limit: (a) so that reason would not look to the world of appearances for some understandable but empirical supreme motive, one that would destroy morality; and (b) so that reason would not uselessly flap its wings in the vacuum of transcendent ideas we call the world of thought, unable to move except into a world of fantasies. Moreover, the idea of a world of pure thought to which we and all intelligent beings belong—while we belong at the same time to the world of sense impressions—gives us a basis for a rational faith. Although we know nothing about it, this ideal universal kingdom of rational ends in themselves can arouse in us a lively interest in the moral law. We can become members of that kingdom only by living strictly by the maxims of freedom as if they were laws of Nature.
Kant’s idea of a rational faith in an ideal universe of rational ends can lead to a form of Absolute Idealism. The reasoning might run something like this. If we believe that the will of every rational being is free, then each person becomes an autonomous member of the world of thought by his very willing. This world of objective reality is governed by laws of freedom, which are the laws of Reason. Thus all men are ends in themselves in a universe governed by Reason, and when they seek to conform their maxims to the law of autonomous reason, they join their own finite minds and wills harmoniously to Absolute Reason, which is superior to the collective whole of finite beings. Thus through reason and will we can come to partake of the Absolute Reason which governs the objective world of thought.
The weakness in this kind of argument lies in inferring from what is at best an ideal that the ideal is a reality. True, a man may find a certain meaning in existence by thinking that he unites himself to Absolute Mind when he performs his duty; nor would this conflict with the moral motive. But we must keep in mind that this is an interpretation of human existence, not a known fact. The ideal may make life more positively meaningful than concepts like “worthiness of happiness,” but it cannot posit a goal which can take precedence over the moral objective of achieving the good will.6
A theoretical philosophy of Nature points to the absolute necessity of a First Cause of the universe. A practical philosophy of freedom also points to an absolute necessity, but only the necessity of the laws governing the actions of rational beings as such. All reasoning follows the basic rule that knowledge be pursued to the awareness of necessity, otherwise it is not rational knowledge. But a basic limitation on this reasoning prevents our knowing the necessity of what exists, what happens, or what ought to happen, unless we presuppose some condition for what exists, happens, or ought to happen. Reason then looks to higher and higher conditions in its search for satisfaction, continuously seeking the ultimate unconditional necessity in knowledge. It finds itself forced to assume this unconditional necessity without being able to prove it or fully understand it, and is content simply to find some idea which fits the presupposition.
Reason seeks the unconditional condition. In metaphysics the unconditioned is represented by the Uncaused Cause of all existence. In moral philosophy it is the absolutely supreme principle which is not derivable from some more remote principle. Human reasoning, however, is wholly incapable of reaching the unconditional condition, either of theoretical or moral metaphysics: the intellectual instinct which sends human reason on its quest keeps it unceasingly searching. Built into reason is the need for an explanation of everything by means of the conditions necessary for this or that to exist or happen. The ideal goal of this rational instinct is ultimate discovery, that ultimate condition necessary to everything else, but itself unconditioned by any higher necessity. There is only one fault in all this: reason would not recognize the unconditioned condition if it found it. Implanted in Reason is the instinctive restlessness to search for the unconditioned condition—but no means is provided for recognizing such a condition. Consequently, no matter what is presented to reason as a condition, reason must seek the condition for this condition, and the condition for this condition, and so on. Thus when we hear the statement, “The universe was caused by God,” we feel a familiar and irresistible urge to ask, “Yes, but what caused God?” while at the same time we recognize the impatience and frustration, the necessity and absurdity, of asking such an impossible question. If we believe God to be the First Cause, we must accept this belief as a necessary presupposition, itself unexplainable, but which explains finite existence. Freedom, too, is a necessary presupposition, forced upon us by our inability as human beings to explain rational will, and our inability to justify freedom.
We cannot criticize our deduction of the supreme principle of morality for failing to explain how the unconditional practical law is absolutely necessary (as the categorical imperative must be). The fault lies in human reason itself; and yet we cannot blame reason for being unwilling to explain the moral law by an appeal to some conditional interest, for any such law would not be moral—it could not be the supreme law of freedom. And so while we cannot comprehend the unconditional practical necessity of the moral imperative, we can at least explain why we cannot comprehend it—which is all that we can ask fairly of a philosophy which tries by its principles to reach the very limit of human reason.
In the final estimate, Kant’s work has been both a success and a failure. He succeeded admirably in showing the necessary elements, the fundamental principles, and the ultimate ground which can constitute a genuine metaphysics of morals. But he failed in the end to prove that his metaphysics is not an illusion of human reason, one having a certain internal coherence but no genuine effectiveness in human affairs. The failure cannot be imputed to Kant’s inadequacy as a moral philosopher. As he analysed the compexities of human reasoning, he concluded that some questions were beyond man’s capacity to answer. The question of whether or not freedom is real was one of them.
One advantage of writing a commentary is that the writer has the chance to say the last word, and I will briefly exercise this irreverent advantage. Kant’s failure, it seems to me, lies not so much in reason’s inability to find the truth of freedom as in Kant’s asking a question which needs no answer. When I employ a system of logic erected on the principle of contradiction, I cannot legitimately ask whether the principle can be proved valid or not. A man reasons by the chosen principle or he does not reason at all, in any intelligible sense. This is the essence of the supplementary proof for the validity of the categorical imperative that I presented earlier in this chapter. It simply cannot seriously be questioned whether the moral law is valid for any rational being with a will. Merely to be aware of one’s existence is enough to “prove” the moral law valid, for morality is nothing more than a formalized definition of what it means to act as a rational being. A human being is duty-bound because he is a human being.