Book I, Chapter I: The principles of pure practical reason

Book I. V:19 A 35 The Analytic of Pure Practical Reason.

Chapter I. On the Principles of Pure Practical Reason.

§1 Definition.

Practical principles are propositions that contain a general determination of the will under which several practical rules fall. They are subjective, or maxims, when the subject regards the condition as valid only for the subject's own will; they are objective, or practical laws, when the condition is cognized as objective, that is, as valid for the will of every rational being.

Remark.

If we assume that pure reason can contain within itself a ground sufficient for practice, that is, for determining the will, then there are practical A 36 laws; otherwise, all practical principles will be mere maxims. In the will of a rational being affected by feeling, maxims may conflict with practical laws that the being itself recognizes. Someone may, for example, make it a maxim never to suffer an insult without taking revenge, and yet see that this is no practical law but only his own maxim, whereas as a rule for the will of every rational being it could not agree with itself within one and the same maxim. In cognition of nature, the principles of what happens—for example, the principle of the equality of action and reaction in the communication of motion—are also laws of nature; for the use V:20 of reason there is theoretical and determined by the character of the object. In practical cognition, that is, cognition concerned solely with determining grounds of the will, principles we adopt are not on that account laws to which we are inevitably subject, because in practical matters reason is concerned with the subject, namely the faculty of desire, whose particular character allows the rule to vary in many ways.—The practical rule is always a product of reason, because it prescribes an action as a means to an effect intended as an end. But for a being in whom reason is not the sole determining ground of the will, this rule is an imperative: a rule expressed through an ought, which conveys the objective necessitation of the action and signifies that, if reason determined the will completely, the action would unfailingly take place according to this rule. Imperatives therefore hold objectively and differ entirely from maxims, A 37 which are subjective principles. Imperatives, however, determine either the conditions of the causality of a rational being as an efficient cause, solely with regard to the effect and its adequacy to produce that effect, or else the will alone, whether or not it is adequate to the effect. The former would be hypothetical imperatives, containing mere precepts of skill; the latter, by contrast, would be categorical and alone would be practical laws. Maxims, then, are indeed principles, but not imperatives. Imperatives themselves, however, when they are conditional—that is, when they determine the will not simply as will, but only with regard to a desired effect, and are thus hypothetical imperatives—are practical precepts, but not laws. Laws must sufficiently determine the will as will even before I ask whether I possess the capacity required for a desired effect, or what I must do to bring it about. They must therefore be categorical; otherwise they are not laws, because they lack the necessity that, if it is to be practical, must be independent of conditions arising from feeling and thus attaching to the will only contingently. Tell someone, for example, that in youth he must work and save so as not to suffer want in old age: this is a correct and also important practical precept for the will. But it is easy to see that the will is here referred to something else that it is assumed to desire. This desire must be left to the agent himself: he may foresee other resources besides the wealth he has acquired, may have no hope of growing old at all, or may think that in case of need he will one day be able to make do with little. Reason, from which alone every rule A 38 that is to contain necessity can spring, does indeed invest this precept of its own with necessity—for otherwise it would not be an imperative. But that necessity is only subjectively conditioned, and we cannot presuppose it to the same degree in every subject. For reason to legislate, however, it must need to presuppose only itself, V:21 since a rule is objectively and universally valid only when it holds without contingent, subjective conditions that distinguish one rational being from another. Now tell someone that he ought never to make a lying promise: this is a rule that concerns his will alone, whether or not the purposes he may have can be achieved through it. Mere willing is what this rule is to determine entirely a priori. If it now turns out that this rule is practically correct, then it is a law, because it is a categorical imperative. Practical laws therefore refer solely to the will, regardless of what its causality accomplishes; we can abstract from this causality, as belonging to the world of sense, in order to have the laws in their purity.

§2 Theorem I.

All practical principles that presuppose an object (matter) of the faculty of desire as the determining ground of the will are without exception empirical and cannot yield practical laws.

By the matter of the faculty of desire I mean an object whose actuality is desired. Now if the desire for this object precedes the practical rule and is the condition A 39 for adopting it as a principle, I say (first) that this principle is then always empirical. For the determining ground of choice is then the representation of an object and that relation of the representation to the subject by which the faculty of desire is determined to make the object actual. But such a relation to the subject is called pleasure in the actuality of an object. This pleasure would therefore have to be presupposed as a condition of the possibility of determining choice. Yet of no representation of any object whatsoever can it be cognized a priori whether it will be connected with pleasure or displeasure, or will be indifferent. In such a case, therefore, the determining ground of choice must always be empirical, and so too must the material practical principle that presupposed it as a condition.

Now (second), a principle founded only on the subjective condition of receptivity to pleasure or displeasure—which can always be cognized only empirically and cannot hold in the same way for all rational beings—may indeed serve the subject who possesses it as a maxim, but cannot serve even that subject as a law, because it lacks objective V:22 necessity, which must be cognized a priori. Such a principle can therefore never yield a practical A 40 law.

§3 Theorem II.

All material practical principles, as such, are without exception of one and the same kind and fall under the general principle of self-love or one's own happiness.

Pleasure arising from the representation of the existence of a thing, insofar as it is to be a determining ground for desiring that thing, rests on the subject's receptivity, because it depends on the existence of an object. It therefore belongs to sense (feeling), not to the understanding, which expresses a relation of the representation to an object according to concepts, but not to the subject according to feelings. Pleasure is thus practical only insofar as the sensation of the agreeableness that the subject expects from the actuality of the object determines the faculty of desire. Now a rational being's consciousness of the agreeableness of life uninterruptedly accompanying its entire existence is happiness; and the principle of making this the supreme determining ground of choice is the principle of self-love. All material principles that place the determining ground of choice in the pleasure or displeasure to be felt from the actuality A 41 of some object are therefore entirely of the same kind in this respect: they all belong to the principle of self-love or one's own happiness.

Corollary.

All material practical rules place the determining ground of the will in the lower faculty of desire. If there were no merely formal laws of the will that sufficiently determined it, then no higher faculty of desire could be admitted either.

Remark I.

It is astonishing that otherwise acute thinkers can believe they find a distinction between the lower and higher faculty of desire in V:23 whether the representations connected with the feeling of pleasure originate in the senses or in the understanding. For when we ask about the determining grounds of desire and locate them in the agreeableness expected from something, it does not matter at all where the representation of this pleasing object comes from, but only how much it pleases. If a representation—even one whose seat and origin lie in the understanding—can determine choice only by presupposing a feeling of pleasure in the subject, then its being a determining ground of choice depends entirely on the character of inner sense: namely, on its capacity to be agreeably affected by that representation. However different in kind the representations of objects may be, even if they are representations of the understanding A 42 or of reason itself, as opposed to representations of the senses, the feeling of pleasure through which alone they actually constitute the determining ground of the will—the agreeableness, the enjoyment expected from them, which impels activity toward producing the object—is nonetheless of one and the same kind. This is so not only because it can always be cognized only empirically, but also because it affects one and the same vital force manifesting itself in the faculty of desire and, in this respect, can differ from any other determining ground only in degree. How else could we compare two determining grounds wholly different in their manner of representation in respect of magnitude, so as to prefer whichever affects the faculty of desire most? The very same person can return unread an instructive book that comes into his hands only once, so as not to miss a hunt; leave midway through a fine speech, so as not to be late for a meal; abandon the entertainment of rational conversation, which he otherwise greatly values, to sit at the gaming table; and even turn away a poor man whom he otherwise takes pleasure in helping, because he now has no more money in his pocket than he needs to pay for admission to the theater. If the determination of his will rests on the feeling of agreeableness or disagreeableness that he expects from any cause, it is entirely indifferent to him what kind of representation affects him. To decide his choice, he cares only how intense this agreeableness is, how long it lasts, how easily it is acquired, and how often it is repeated. Someone who needs gold to spend is entirely indifferent whether its material, A 43 the gold, was mined from the mountains or washed from sand, provided it is accepted everywhere at the same value. In just the same way, no one concerned solely with the agreeableness of life asks whether representations of the understanding or of the senses provide it, but only how much enjoyment, and how great an enjoyment, they will provide for the longest time. Only those who would gladly deny pure reason the capacity to determine the will without presupposing any feeling V:24 can stray so far from their own definition as afterward to declare wholly different in kind what they themselves previously brought under one and the same principle. We find, for example, that we can take enjoyment even in the mere exercise of our powers, in consciousness of our strength of soul in overcoming obstacles to our resolve, in cultivating our mental talents, and so on. We rightly call these finer joys and delights, because they are more within our control than others, do not wear out, but instead strengthen our capacity to feel still greater enjoyment of them, and cultivate us even as they delight us. But to present them on that account as a way of determining the will other than through sense alone—when, as the first condition of this satisfaction, they already presuppose a feeling implanted in us that makes these enjoyments possible—is just like the ignorant would-be dabblers in metaphysics who imagine matter so fine, so exceedingly fine that the thought might make them dizzy, and then believe they have conceived in this way a spiritual yet extended being. If, with Epicurus, we make virtue determine the A 44 will solely through the enjoyment it promises, we cannot afterward reproach him for regarding this enjoyment as entirely of the same kind as the pleasures of the coarsest senses. For we have no grounds for charging him with having attributed the representations that arouse this feeling in us solely to the bodily senses. As far as one can guess, he sought the source of many of them just as much in the use of the higher cognitive faculty. But that did not prevent him, and could not prevent him, from regarding, in accordance with the principle just mentioned, the enjoyment itself that these possibly intellectual representations afford us—and through which alone they can become determining grounds of the will—as entirely of the same kind. To be consistent is the greatest obligation of a philosopher, yet the one most rarely fulfilled. The ancient Greek schools give us more examples of it than we find in our syncretistic age, when a certain coalition system of contradictory principles is contrived with great dishonesty and shallowness, because it appeals more readily to a public content to know a little of everything and nothing as a whole, while remaining adept in its old ways. However much understanding and reason the principle of one's own happiness may employ, it would still contain no determining grounds for the will other than those suited to the lower faculty of desire. Either, then, there is no higher faculty of desire at all, or pure reason must be practical by itself alone: it must be able to determine A 45 the will through the mere form of the practical rule, without presupposing any feeling, and thus without representations of the agreeable or disagreeable as the matter of the faculty of desire, which is always an empirical condition of principles. Only then, insofar as reason by itself determines the will V:25 and is not in the service of inclinations, is it a true higher faculty of desire, to which the faculty determinable through feeling is subordinate. It is really, indeed specifically, distinct from that faculty, so that even the slightest admixture of the latter's impulses diminishes its strength and superiority, just as the least empirical element introduced as a condition into a mathematical demonstration lowers and destroys its dignity and force. In a practical law, reason determines the will immediately, not through an intervening feeling of pleasure or displeasure, not even pleasure in this law itself; and only its capacity to be practical as pure reason makes it possible for reason to be legislative.

Remark II.

To be happy is necessarily the desire of every rational but finite being, and thus an unavoidable determining ground of its faculty of desire. For satisfaction with its entire existence is not an original possession, a blessedness that would presuppose consciousness of its independent self-sufficiency. Rather, it is a problem imposed on it by its finite nature itself, because it has needs. These needs concern the matter of its faculty of desire, that is, something related to a subjectively underlying feeling of pleasure or displeasure, by which what it needs for satisfaction with its condition is determined. But precisely because this material determining ground can be cognized by the subject only empirically, it is impossible to regard this task as a law, since a law, as objective, would have to contain in every case and for every rational being the very same A 46 determining ground of the will. For although the concept of happiness everywhere underlies the practical relation of objects to the faculty of desire, it is only the general title for subjective determining grounds and determines nothing specifically. Yet this practical task is concerned solely with such specific determination and cannot be solved without it. What each person is to place his happiness in depends on his particular feeling of pleasure and displeasure, and even in one and the same subject it depends on differences of need as this feeling changes. A subjectively necessary law, as a law of nature, is therefore objectively a highly contingent practical principle, which can and must differ greatly in different subjects and hence can never yield a law. For in the desire for happiness, what matters is not the form of conformity to law, but solely the matter: whether and how much enjoyment I can expect from following the law. Principles of self-love can indeed contain general rules of skill for finding means to ends, V:26 but in that case they are merely theoretical principles⁷—for example, how someone who would like to eat bread should devise a mill. A 47 But practical precepts founded on them can never be universal, since the determining ground of the faculty of desire rests on the feeling of pleasure and displeasure, which can never be assumed to be universally directed toward the same objects.

But suppose that finite rational beings all thought alike both about what they should accept as objects of their feelings of enjoyment or pain, and even about the means they must use to attain the former and avert the latter. They could still by no means present the principle of self-love as a practical law, for this unanimity would itself be merely contingent. The determining ground would still be only subjectively valid and merely empirical, and would lack the necessity thought in every law: objective necessity from grounds a priori. Unless, that is, one were to present this necessity not as practical at all, but as merely physical, meaning that our inclination compelled the action from us just as inevitably as we are compelled to yawn when we see others yawning. One could sooner maintain that there are no practical laws at all, but only recommendations for the benefit of our desires, than elevate merely subjective principles to the rank of practical laws. Such laws must possess objective, not merely subjective, necessity throughout, and must be cognized through reason a priori, not through experience, however empirically universal experience may be. Even the rules of concordant appearances are called laws of nature—for example, mechanical laws—only if we either actually cognize them a priori or at least assume, as with A 48 chemical laws, that they would be cognized a priori from objective grounds if our insight went deeper. But with merely subjective practical principles, it is expressly made a condition that they must be grounded not in objective but in subjective conditions of choice; hence they may always be represented only as mere maxims, never as practical laws. This last remark seems at first glance mere quibbling over words; yet it gives verbal definition to the most important distinction, one that can come into consideration only in practical investigations.

§4 V:27 Theorem III.

If a rational being is to think of its maxims as universal practical laws, it can think of them only as principles that contain the determining ground of the will not in their matter but solely in their form.

The matter of a practical principle is the object of the will. This object either is or is not the determining ground of the will. If it is, then the rule of the will would be subject to an empirical condition—the relation of the determining representation to the feeling of pleasure and displeasure—and consequently would not be a practical law. Now if we separate all matter from a law, that is, every object of the will as a determining ground, nothing remains but the mere form of a universal legislation. A 49 A rational being, therefore, either cannot think of its subjectively practical principles, that is, maxims, as universal laws at the same time, or must assume that their mere form, by which they are suited to universal legislation, makes them practical laws by itself alone.

Remark.

The most ordinary understanding can distinguish without instruction which form in a maxim is suited to universal legislation and which is not. Suppose, for example, I have made it my maxim to increase my wealth by every safe means. Now I hold a deposit whose owner has died without leaving any written acknowledgment of it. This is naturally a case for my maxim. I now want to know only whether that maxim can also hold as a universal practical law. I therefore apply it to the present case and ask whether it could indeed assume the form of a law, and hence whether through my maxim I could at the same time give a law such as this: everyone may deny having received a deposit whose entrustment to him no one can prove. I immediately see that such a principle, as a law, would destroy itself, since it would bring it about that there were no deposits at all. A practical law that I recognize as such must qualify for universal legislation; this is an identical proposition and thus clear in itself. If I now say that my will is subject to a practical law, I cannot cite my inclination—for example, in the present case, my greed—as a determining ground of it suited to a universal practical law. V:28 For this inclination, far from A 50 being fit for universal legislation, must instead consume itself in the form of a universal law.

It is therefore strange that sensible thinkers could have conceived of presenting the desire for happiness as a universal practical law simply because this desire, and thus also the maxim by which each person makes happiness the determining ground of his will, is universal. For whereas a universal law of nature otherwise brings everything into accord, here, if we wished to give the maxim the universality of a law, precisely the extreme opposite of accord would follow: the worst conflict and the complete destruction of the maxim itself and its purpose. For everyone's will does not then have one and the same object; instead, each has his own—his own well-being—which may indeed happen to agree with the purposes of others, likewise directed toward themselves, but falls far short of sufficing for a law, because the exceptions one is entitled to make on occasion are endless and cannot be comprehended determinately in a universal rule. This produces a harmony like the one depicted in a certain satirical poem about the spiritual accord of a married couple ruining each other: O marvelous harmony: what he wants, so does she, etc.; or like the declaration attributed to King Francis I in regard to Emperor Charles V: what my brother Charles wants to have—Milan—I want to have too. Empirical determining grounds are unfit for any universal external legislation, but equally unfit for internal legislation. For each person takes his own subject as the basis of inclination, while another takes a different subject, A 51 and even within each subject, now one inclination and now another has the predominant influence. To discover a law that would govern them all under this condition, namely with agreement on every side, is utterly impossible.

§5 Problem I.

Assuming that the mere legislative form of maxims alone is the sufficient determining ground of a will, to find the character of the will that is determinable by this alone.

Since the mere form of a law can be represented only by reason, and is therefore no object of the senses and hence does not belong among appearances, the representation of this form as a determining ground of the will differs from all determining grounds of events in nature according to the law of causality, V:29 because in their case the determining grounds must themselves be appearances. But if no determining ground of the will other than that universal legislative form alone can serve it as a law, then such a will must be thought entirely independent of the natural law of appearances in their relation to one another, namely the law of causality. Such independence, however, is called freedom in the strictest, that is, transcendental sense. Thus a will A 52 for which the mere legislative form of a maxim alone can serve as a law is a free will.

§6 Problem II.

Assuming that a will is free, to find the law that alone is fit to determine it necessarily.

Since the matter of a practical law, that is, an object of the maxim, can never be given otherwise than empirically, while a free will, as independent of empirical conditions—that is, conditions belonging to the world of sense—must nevertheless be determinable, a free will must find a determining ground in the law independently of its matter. But apart from the matter, nothing more is contained in the law than its legislative form. The legislative form, insofar as it is contained in the maxim, is therefore the only thing that can constitute a determining ground of the will.

Remark.

Freedom and unconditional practical law thus refer reciprocally to one another. I am not asking here whether they are in fact distinct, or whether, instead, an unconditional law is merely the self-consciousness of a pure practical reason, and this in turn entirely identical with the positive concept of freedom. I ask rather where our cognition of the unconditionally practical begins: with freedom or with the practical law. It cannot begin with freedom, A 53 for we can neither become immediately conscious of it, since its first concept is negative, nor infer it from experience, since experience lets us cognize only the law of appearances, and thus the mechanism of nature, the exact opposite of freedom. It is therefore the moral law, of which we become immediately conscious as soon as we frame maxims of the will, that presents itself to us first. And since reason V:30 presents it as a determining ground that no sensible conditions can outweigh, indeed as entirely independent of them, it leads directly to the concept of freedom. But how is consciousness of that moral law itself possible? We can become conscious of pure practical laws just as we are conscious of pure theoretical principles: by attending to the necessity with which reason prescribes them to us, and to the separation of all empirical conditions to which it directs us. The concept of a pure will arises from the former, just as consciousness of a pure understanding arises from the latter. That this is the true order of subordination among our concepts, and that morality first discloses to us the concept of freedom, so that practical reason first confronts speculative reason with its most insoluble problem in this concept, placing it in the greatest perplexity, is already evident from the following. Nothing in appearances can be explained from the concept of freedom; here the mechanism of nature must always provide the guiding thread. Moreover, the antinomy of pure reason, when it seeks to ascend to the unconditioned in the series of causes, becomes entangled in incomprehensibilities with either concept alike; yet the latter, mechanism, at least has A 54 usefulness in explaining appearances. Thus no one would ever have ventured to introduce freedom into science had not the moral law, and with it practical reason, intervened and forced this concept upon us. Experience, too, confirms this order of concepts within us. Suppose someone claims that his lustful inclination is entirely irresistible whenever the desired object and the opportunity present themselves. Ask whether he would not master his inclination if a gallows were erected in front of the house where he finds this opportunity, in order to hang him immediately after he had enjoyed his lust. We need not guess long what he would answer. But ask him whether, if his prince demanded on pain of the same immediate death that he give false testimony against an honest man whom the prince would like to destroy under plausible pretexts, he would consider it possible to overcome his love of life, however great it might be. He may perhaps not venture to assure us whether he would do it or not; but he must admit without hesitation that it would be possible for him. He judges, therefore, that he can do something because he is conscious that he ought to do it, and cognizes within himself the freedom that, without the moral law, would otherwise have remained unknown to him.

§7 Fundamental Law of Pure Practical Reason.

Act so that the maxim of your will could always hold at the same time as the principle of a universal legislation.

V:31

Remark. A 55

Pure geometry has postulates as practical propositions, but these contain nothing more than the presupposition that one can do something if one were required to do it; and these are its only propositions that concern existence. They are thus practical rules under a problematic condition of the will. Here, however, the rule says that one ought to proceed in a certain way unconditionally. The practical rule is therefore unconditional, and hence represented a priori as a categorical practical proposition, by which the will is objectively determined absolutely and immediately—through the practical rule itself, which is therefore a law here. For pure reason, practical in itself, is here immediately legislative. The will is thought as independent of empirical conditions, and thus, as pure will, as determined through the mere form of the law; and this determining ground is regarded as the supreme condition of all maxims. The matter is strange enough and has no parallel in the whole remainder of practical cognition. For the thought a priori of a possible universal legislation, which is thus merely problematic, is unconditionally commanded as a law, without borrowing anything from experience or from any external will. Nor is it a precept according to which an action ought to take place through which a desired effect becomes possible—for then the rule would always be physically conditioned. It is instead a rule that determines only the will a priori with regard to the form of its maxims. Here, then, it is at least not impossible to think of a law serving solely the subjective form of principles as a determining ground through the objective form of a law in general. Consciousness of this fundamental law may be called a fact of reason, A 56 because we cannot reason it out from antecedent data of reason, such as consciousness of freedom (for this is not given to us beforehand), but because it forces itself upon us of itself as a synthetic proposition a priori, founded on no intuition, either pure or empirical. It would indeed be analytic if freedom of the will were presupposed; but for freedom as a positive concept an intellectual intuition would be required, which we may by no means assume here. Yet to regard this law as given without misinterpretation, we must carefully note that it is no empirical fact, but the sole fact of pure reason, which thereby announces itself as originally legislative: thus I will, thus I command.

Corollary.

Pure reason is practical by itself alone and gives (to the human being) a universal law, which we call the moral law.

V:32

Remark.

The fact just mentioned is undeniable. We need only analyze the judgment people make about the conformity of their actions to law. We will always find that, whatever inclination may interject, their reason, incorruptible and constrained by itself, nevertheless always measures the maxim of the will in an action against the pure will, that is, against itself, insofar as it regards itself as practical a priori. Now this principle of morality, precisely because of the universality of the legislation that it makes the supreme formal determining ground of the will regardless of all the will's subjective differences, is declared by reason to be a law for all rational A 57 beings, insofar as they have a will at all, that is, a capacity to determine their causality through the representation of rules. It holds, therefore, insofar as they are capable of actions according to principles, and hence also according to practical principles a priori, for these alone possess the necessity that reason requires of a principle. It is thus not restricted to human beings, but extends to all finite beings that have reason and will, and indeed includes even the infinite being as supreme intelligence. In the former case, however, the law takes the form of an imperative. For although we can presuppose in a finite being, as rational, a pure will, we cannot presuppose in it, as a being affected by needs and sensible motives, a holy will: one incapable of any maxim conflicting with the moral law. For such beings the moral law is therefore an imperative that commands categorically, because the law is unconditional. The relation of such a will to this law is dependence, under the name of obligation, which signifies a necessitation, though through reason alone and its objective law, to an action that is called duty for this reason: choice affected by feeling—although not determined by it, and thus still always free—carries with it a wish arising from subjective causes, which can therefore often oppose the pure objective determining ground. It consequently requires, as moral necessitation, a resistance from practical reason that may be called an internal but intellectual constraint. In the supremely self-sufficient intelligence, choice is rightly represented as incapable of any maxim that could not at the same time be objectively a law; and the concept of holiness, which belongs to it for that reason, places it A 58 not above all practical laws, but above all practically restrictive laws, and thus above obligation and duty. This holiness of the will is nonetheless a practical idea that must necessarily serve as the archetype, an infinite approach to which is all that is possible for finite rational beings. The pure moral law, itself therefore called holy, constantly and accurately holds this idea before their eyes. To be assured of this infinite progress V:33 of one's maxims and of their immutability in continued advance—that is, virtue—is the highest achievement of finite practical reason. Yet virtue itself, at least as a naturally acquired capacity, can never be complete, since assurance in such a case never becomes apodictic certainty and, as persuasion, is very dangerous.

§8 Theorem IV.

The autonomy of the will is the sole principle of all moral laws and of the duties that accord with them. All heteronomy of choice, by contrast, not only grounds no obligation at all, but is instead opposed to the principle of obligation and to the morality of the will. The sole principle of morality consists in independence from all matter of the law, namely a desired object, and yet at the same time in the determination of choice through the mere universal legislative form of which a maxim must be capable. That independence is freedom in the negative sense, while this legislation of its own A 59 by pure reason, practical as such, is freedom in the positive sense. Thus the moral law expresses nothing other than the autonomy of pure practical reason, that is, of freedom; and this is itself the formal condition of all maxims under which alone they can agree with the supreme practical law. If, therefore, the matter of willing—which can be nothing other than the object of a desire connected with the law—enters the practical law as a condition of its possibility, heteronomy of choice results: dependence on the natural law of following some impulse or inclination. The will then does not give itself the law, but only the precept for rationally following laws of feeling. The maxim, however, which can never in this way contain the universally legislative form, not only establishes no obligation in this manner, but is itself opposed to the principle of a pure practical reason, and thus also to a moral disposition, even if the action arising from it should conform to law.

V:34

Remark I.

A practical precept carrying a material, and thus empirical, condition must therefore never be counted as a practical law. For the A 60 law of the pure will, which is free, places it in a sphere entirely different from the empirical; and the necessity it expresses, since it is not to be natural necessity, can consist solely in formal conditions of the possibility of a law in general. All matter of practical rules always rests on subjective conditions, which confer on it no universality for rational beings other than a conditional one: if I desire this or that, what must I then do to make it actual? These rules all revolve around the principle of one's own happiness. Now it is certainly undeniable that all willing must also have an object, and thus a matter; but this does not make that matter the determining ground and condition of the maxim. For if it is, the maxim cannot be presented in a universally legislative form, since expectation of the object's existence would then be the determining cause of choice, and willing would have to be grounded in the faculty of desire's dependence on the existence of some thing. Such dependence can always be sought only in empirical conditions and can therefore never provide the ground for a necessary and universal rule. Thus the happiness of other beings can be the object of a rational being's will. But if it were the determining ground of the maxim, we would have to presuppose that we find in the well-being of others not only a natural enjoyment but also a need, as is indeed the case with the sympathetic temperament in human beings. But I cannot presuppose this need in every rational being, and not at all in God. The matter of the maxim may therefore remain, but it must not be the maxim's condition; otherwise A 61 the maxim would be unfit to be a law. Thus the mere form of a law, which limits the matter, must at the same time be a ground for adding this matter to the will, but not for presupposing it. Let the matter, for example, be my own happiness. If I attribute this to everyone—as I may indeed do in the case of finite beings—it can become an objective practical law only if I include the happiness of others within it. The law of promoting others' happiness therefore arises not from the presupposition that this is an object of everyone's choice, but solely from the fact that the form of universality—which reason requires as a condition for giving a maxim of self-love the objective validity of a law—becomes the determining ground of the will. Thus the object, the happiness of others, was not the determining ground of the pure will. It was the mere lawful form alone, by which I limited my maxim grounded on inclination in order to give it the universality V:35 of a law and so make it accord with pure practical reason. Only from this limitation, and not from the addition of an external incentive, could the concept of the obligation to extend the maxim of my self-love to the happiness of others arise.

Remark II.

The exact opposite of the principle of morality is to make one's own happiness the determining ground of the will. As I have shown above, this includes everything that locates the determining ground which is to serve as a law anywhere other than in the legislative form of the maxim. This opposition is not merely logical, like the A 62 opposition involved in empirically conditioned rules that one nevertheless wished to elevate into necessary principles of cognition. It is practical, and would utterly destroy morality if the voice of reason in relation to the will were not so clear, so impossible to drown out, so audible even to the most ordinary person. As things stand, it can maintain itself only in the mind-confusing speculations of the schools, which are bold enough to turn a deaf ear to that heavenly voice in order to uphold a theory that costs them no mental exertion.

Suppose a companion whose company you otherwise enjoy thought to justify to you his having given false testimony by first invoking what he claimed was the sacred duty of securing his own happiness. He then listed all the advantages he had gained by it and pointed out the prudence he had exercised to secure himself against any discovery, even on your part: he had revealed the secret to you solely so that he could deny it at any time. But suppose he then claimed in all seriousness to have fulfilled a genuine human duty. You would either laugh in his face or recoil from it in disgust, although, if someone had directed his principles solely toward his own advantages, you would have not the slightest objection to these measures. Or suppose someone recommended a man as a steward to whom you could blindly entrust all your affairs, and, to inspire your confidence, praised him as a shrewd person who understood his own advantage masterfully and as tirelessly active, letting no opportunity for it pass unused. Finally, so that A 63 no misgivings about vulgar selfishness should stand in the way, suppose he praised the man's refined way of life: he sought his enjoyment not in hoarding money or in coarse extravagance, but in extending his knowledge, in well-chosen, instructive company, even in helping those in need. Yet as for the means—which borrow their worth or worthlessness only from the end—he had no scruples, and other people's money and property were, V:36 for these purposes, as good to him as his own whenever he knew he could use them without detection or hindrance. You would believe either that the person recommending him was making a fool of you or that he had lost his mind.—So clearly and sharply are the boundaries between morality and self-love drawn that even the most ordinary eye cannot fail to distinguish whether something belongs to the one or the other. The following few remarks may indeed seem superfluous in the case of so evident a truth, but they at least serve to lend somewhat greater clarity to the judgment of ordinary human reason.

The principle of happiness can indeed yield maxims, but never ones fit to be laws of the will, even if one made universal happiness one's object. For cognition of happiness rests entirely on data of experience, and every judgment about it depends greatly on each person's opinion, which is itself also highly changeable. There can therefore be general, but never universal, rules: rules that hold most often on average, but not rules that must hold always and necessarily. Hence no practical laws can be founded on them. Precisely because an object of choice is here laid at the foundation of its rule and must therefore precede it, the rule can be referred A 64 to and founded on nothing other than what is recommended, and thus on experience; and there differences of judgment must be endless. This principle therefore does not prescribe the very same practical rules to all rational beings, although they stand under a common title, namely happiness. The moral law, however, is thought as objectively necessary only because it is to hold for everyone who has reason and will.

The maxim of self-love, or prudence, merely advises; the law of morality commands. Yet there is a great difference between what we are advised to do and what we are obligated to do.

What is to be done according to the principle of the autonomy of choice is quite easy for the most ordinary understanding to discern without hesitation. What is to be done on the presupposition of its heteronomy is difficult and requires knowledge of the world. That is, what duty is presents itself to everyone of its own accord; but what brings true and lasting advantage, if this advantage is to extend over one's whole existence, is always shrouded in impenetrable darkness. Much prudence is required to adapt the practical rule directed toward it to the purposes of life, even tolerably, through skillful exceptions. Nevertheless, the moral law commands everyone to follow it, and indeed with the greatest exactness. Judging what is to be done according to it must therefore not be so difficult that even the most ordinary and unpracticed understanding could not manage it without worldly prudence.

To satisfy the categorical command of morality is within everyone's power V:37 at all times; to satisfy the empirically conditioned precept of happiness is only rarely A 65 possible, and far from possible for everyone even with regard to a single purpose. The reason is that in the former case only the maxim matters, which must be genuine and pure; in the latter, one's powers and physical capacity to make a desired object actual matter as well. A command that everyone should seek to make himself happy would be foolish, for one never commands someone to do what he already inevitably wills of himself. One would have only to command the measures, or rather provide them, since he cannot do everything he wills. But to command morality under the name of duty is entirely reasonable; for, first, not everyone is willing to obey its precept when it conflicts with inclinations; and as for the measures by which he can follow this law, these need not be taught here, for what he wills in this respect he can also do.

Someone who has lost at gambling may be annoyed with himself and his imprudence. But if he is conscious of having cheated at gambling, even though he won by doing so, he must despise himself as soon as he compares himself with the moral law. This law must therefore be something other than the principle of one's own happiness. For having to say to oneself, “I am a worthless person, although I have filled my purse,” must involve a different standard of judgment from applauding oneself and saying, “I am a prudent man, for I have enriched my coffers.”

Finally, there is something else in the idea of our practical reason that accompanies the transgression of a moral law: its deserving punishment. Now the concept of punishment as such cannot be connected with the A 66 attainment of happiness. For although the one who punishes may at the same time have the benevolent intention of directing this punishment toward that end as well, it must first be justified in itself as punishment, that is, as a mere ill. Thus the person punished, if matters rested there and he could see no favor hidden behind this severity, would himself have to admit that justice had been done to him and that his lot was perfectly suited to his conduct. In every punishment as such there must first be justice, and this constitutes the essence of the concept. Benevolence may indeed be joined to it, but after his conduct the one deserving punishment has not the slightest ground to count on it. Punishment is therefore a physical ill that, even if it were not connected with moral evil as a natural consequence, would nevertheless have to be connected with it as a consequence according to principles of moral legislation. Now if every crime is punishable in itself, even without regard to its physical consequences for the agent—that is, if it forfeits happiness, at least in part—then it would plainly be absurd to say that the crime consisted precisely in his having brought punishment upon himself by impairing his own V:38 happiness, which according to the principle of self-love would have to be the proper concept of every crime. In this way punishment would be the ground for calling something a crime, and justice would instead have to consist in refraining from all punishment and even preventing natural punishment. For then there would no longer be anything evil in the action, since the ills that otherwise followed it, and on account of which alone the action was called evil, would now have been averted. To regard all punishment and reward, moreover, merely as A 67 machinery in the hands of a higher power, serving solely to set rational beings in motion toward their final end, happiness, is all too visibly a mechanism of their will abolishing all freedom for us to need to dwell on it.

More subtle still, though equally untrue, is the claim of those who assume a certain special moral sense that, rather than reason, determines the moral law. According to them, consciousness of virtue is immediately connected with satisfaction and enjoyment, consciousness of vice with inner unrest and pain; and so they still stake everything on the desire for one's own happiness. Without bringing in what has been said above, I will only point out the deception involved here. To represent the vicious person as tormented by inner unrest through consciousness of his transgressions, they must already represent him in advance, in the chief basis of his character, as morally good at least to some degree. Likewise, they must already represent the person delighted by consciousness of dutiful actions as virtuous. The concept of morality and duty must therefore have preceded all consideration of this satisfaction and cannot be derived from it at all. We must first esteem the importance of what we call duty, the authority of the moral law, and the immediate worth that following it gives a person in his own eyes, in order to feel that satisfaction in consciousness of our conformity to duty and that bitter reproach when we can accuse ourselves of transgressing it. We cannot, therefore, feel this satisfaction A 68 or inner unrest before cognition of obligation and make it the ground of obligation. One must already be at least halfway an honest man even to form a representation of those sentiments. I by no means deny, moreover, that just as the human will, by virtue of freedom, can be immediately determined by the moral law, so frequent practice in accordance with this determining ground can finally produce, subjectively, a feeling of satisfaction with oneself. Rather, it is itself a duty to establish and cultivate this feeling, which alone properly deserves to be called moral feeling. But the concept of duty cannot be derived from it; otherwise we would have to think of a feeling of a law as such, and make an object of sensation V:39 what can be thought only through reason. Unless this is to become a flat contradiction, it would entirely abolish every concept of duty and replace it with a merely mechanical play of finer inclinations that sometimes conflict with the coarser ones.

If we now compare our formal supreme principle of pure practical reason, as an autonomy of the will, with all previous material principles of morality, we can present all the others as such in a table that actually exhausts every possible alternative apart from a single formal principle. We can thus demonstrate at a glance that it is futile to look for any principle other than the one now presented.—All possible determining grounds of the will are either merely subjective and thus empirical, or else objective and rational; but both are either external or internal.

Practical material determining groundsV:40 in the principle of morality are A 69Subjectiveexternal• Education (according to Montaigne)• Civil constitution (according to Mandeville)internal• Physical feeling (according to Epicurus)• Moral feeling (according to Hutcheson)Objectiveinternal• Perfection (according to Wolff and the Stoics)external• The will of God (according to Crusius and other theological moralists)

V:41 Those on the left are all empirical and are plainly A 70 entirely unfit to serve as the universal principle of morality. Those on the right, however, are founded on reason, for perfection as a quality of things and supreme perfection represented in substance, that is, God, can both be thought only through concepts of reason. But the first concept, namely perfection, can be taken in a theoretical sense, where it means nothing but the completeness of each thing in its kind (transcendental perfection), or of a thing merely as a thing in general (metaphysical perfection); and neither can be at issue here. The concept of perfection in a practical sense, however, is a thing's fitness or adequacy for all sorts of ends. This perfection as a quality of the human being, and thus internal, is nothing other than talent and what strengthens or supplements it, skill. Supreme perfection in substance, that is, God, and thus external perfection, considered in a practical respect, is the adequacy of this being for all ends in general. Now ends must be given to us beforehand, in relation to which alone the concept of perfection—whether internal in ourselves or external in God—can become a determining ground of the will. But an end, as an object that must precede the determination of the will through a practical rule and contain the ground of the possibility of such a determination, and thus as the matter of the will taken as its determining ground, is always empirical. It can therefore serve the Epicurean principle of the doctrine of happiness, but never the pure rational principle of the doctrine of morals and duty. For talents and their cultivation can become motives of the will only because they contribute to the advantages of life; and the A 71 will of God, if conformity with it is made the object of the will without an antecedent practical principle independent of its idea, can become a motive only through the happiness we expect from it. It follows, first, that all the principles set out here are material; second, that they comprise all possible material principles; and finally, from this, the conclusion follows that, since material principles are entirely unfit to serve as the supreme moral law, as has been proved, the formal practical principle of pure reason is the only possible one suited to categorical imperatives, that is, practical laws that make actions duties, and to the principle of morality in general, both in judgment and in its application to determining the human will. According to this principle, the mere form of a universal legislation possible through our maxims must constitute the supreme and immediate determining ground of the will.

I. A 72 V:42 On the Deduction of the Principles of Pure Practical Reason.

This Analytic demonstrates that pure reason can be practical, that is, can determine the will by itself, independently of everything empirical. It does so through a fact in which pure reason proves itself actually practical in us: autonomy in the principle of morality, by which reason determines the will to action.—It shows at the same time that this fact is inseparably connected with consciousness of freedom of the will, indeed is identical with it. Through this consciousness, the will of a rational being—which, as belonging to the world of sense, cognizes itself as necessarily subject, like other efficient causes, to the laws of causality—is nevertheless at the same time conscious in practical matters of its existence as determinable on another side, namely as a being in itself, within an intelligible order of things. This is not in accordance with a special intuition of itself, but with certain dynamic laws that can determine its causality in the world of sense. For it has been sufficiently proved elsewhere that freedom, if attributed to us, places us in an intelligible order of things.

If we now compare this with the analytic part of the Critique of Pure A 73 Speculative Reason, a remarkable contrast between the two emerges. There, not principles but pure sensible intuition—space and time—was the first datum that made cognition a priori possible, and only for objects of the senses.—Synthetic principles from mere concepts without intuition were impossible. Rather, they could arise only in relation to that intuition, which was sensible, and hence only in relation to objects of possible experience, since concepts of the understanding, combined with this intuition, alone make possible the cognition we call experience.—Beyond objects of experience, and thus concerning things as noumena, speculative reason was quite rightly denied every positive element of cognition.—Nevertheless, it accomplished this much: it secured the concept of noumena, that is, the possibility, indeed the necessity, of thinking such things. It defended, for example, against all V:43 objections the assumption of freedom, considered negatively, as entirely compatible with those principles and limitations of pure theoretical reason, yet without providing cognition of anything determinate and expansive concerning such objects; rather, it completely cut off every prospect of that.

The moral law, by contrast, though it gives no prospect, A 74 nevertheless places in our hands a fact absolutely inexplicable from all the data of the world of sense and the entire compass of our theoretical use of reason. This fact points to a pure world of understanding, indeed even determines it positively and lets us cognize something of it, namely a law.

This law is to give the world of sense, as a sensible nature, the form, as regards rational beings, of a world of understanding, that is, of a supersensible nature, without impairing the mechanism of the former. Now nature in the most general sense is the existence of things under laws. The sensible nature of rational beings in general is their existence under empirically conditioned laws, and thus, for reason, heteronomy. The supersensible nature of these same beings, by contrast, is their existence according to laws independent of every empirical condition and thus belonging to the autonomy of pure reason. And since the laws according to which the existence of things depends on cognition are practical, supersensible nature, insofar as we can form a concept of it, is nothing other than a nature under the autonomy of pure practical reason. The law of this autonomy, however, is the moral law, which is therefore the fundamental law of a supersensible nature and a pure world of understanding, whose counterpart is to exist in the world of sense, yet A 75 without impairing its laws. We could call the former the original model (natura archetypa), which we cognize only in reason, and the latter the reproduction (natura ectypa), because it contains the possible effect of the idea of the former as a determining ground of the will. For the moral law does indeed transport us, in idea, into a nature in which pure reason, if accompanied by a physical capacity adequate to it, would produce the highest good; and it determines our will to impart this form to the world of sense as a whole of rational beings.

The most ordinary attention to ourselves confirms that this idea actually underlies our determinations of the will as a model, a kind of preliminary design.

V:44 When the maxim according to which I intend to give testimony is examined by practical reason, I always consider what it would be like if it held as a universal law of nature. Clearly, in this form it would necessitate everyone to truthfulness. For it cannot be consistent with the universality of a law of nature to allow statements to count as evidence and yet as intentionally untrue. Likewise, the maxim I adopt concerning the free disposal of my A 76 life is immediately determined when I ask myself what it would have to be like for a nature to maintain itself according to a law of that maxim. Clearly, in such a nature no one could end his life at will, for such a constitution would not be an enduring order of nature; and so in all other cases. In actual nature, however, as an object of experience, the free will is not of itself determined to maxims that could by themselves establish a nature according to universal laws, or that would of themselves fit into a nature ordered according to such laws. Rather, it is private inclinations that constitute a natural whole according to laws of feeling—physical laws—but not a nature that would be possible solely through our will according to pure practical laws. Nevertheless, through reason we are conscious of a law to which all our maxims are subject, as though an order of nature must at the same time arise through our will. This must therefore be the idea of a nature not empirically given and yet possible through freedom, and thus supersensible. To this nature we give objective reality, at least in a practical respect, because we regard it as the object of our will as pure rational beings.

The difference, then, between the laws of a nature to which the A 77 will is subject and those of a nature subject to a will, with regard to what bears on the will's free actions, rests on this: in the former, objects must be causes of the representations that determine the will; in the latter, the will is to be the cause of the objects, so that its causality has its determining ground solely in the pure faculty of reason, which may therefore also be called a pure practical reason.

Thus the two problems—how pure reason can, on the one hand, cognize objects a priori, and how, on the other, it can immediately be a determining ground V:45 of the will, that is, of the causality of a rational being with regard to the actuality of objects, solely through the thought of the universal validity of its own maxims as law—are very different.

The first, belonging to the Critique of Pure Speculative Reason, requires a prior explanation of how intuitions are possible a priori, since without them no object at all can be given to us and hence none can be synthetically cognized. Its solution is that these intuitions are all merely sensible, and therefore allow no speculative cognition extending beyond the reach of possible experience. Hence all the principles of that pure speculative reason accomplish nothing A 78 more than making experience possible, whether of given objects or of those that may be given to infinity but are never given completely.

The second, belonging to the Critique of Practical Reason, requires no explanation of how the objects of the faculty of desire are possible. For that, as a problem of theoretical cognition of nature, is left to the Critique of Speculative Reason. It asks only how reason can determine the maxim of the will: whether this occurs only through empirical representations as determining grounds, or whether pure reason too would be practical and a law of a possible order of nature that is not empirically cognizable at all. The possibility of such a supersensible nature, whose concept could at the same time be the ground of its actuality through our free will, requires no intuition a priori of an intelligible world—an intuition that in this case, as supersensible, would also have to be impossible for us. For what matters is only the determining ground of willing in its maxims: whether it is empirical or a concept of pure reason, concerning the conformity of those maxims to law in general, and how it can be the latter. Whether or not the causality of the will suffices to make the objects actual is left to the theoretical principles of reason to judge, as an investigation A 79 of the possibility of the objects of willing. Their intuition therefore constitutes no element at all in the practical problem. Here only the determination of the will and the determining ground of its maxim as a free will matter, not the outcome. For provided the will conforms to law for pure reason, whatever the state of its capacity for execution, V:46 whether or not such a nature actually arises from these maxims of the legislation of a possible nature, this is of no concern to the Critique, which investigates whether and how pure reason can be practical, that is, can determine the will immediately.

In this undertaking it can therefore, without reproach, and indeed must, begin with pure practical laws and their actuality. But in place of intuition it grounds them in the concept of their existence in the intelligible world, namely freedom. For this concept means nothing else; and those laws are possible only in relation to freedom of the will, but necessary on the presupposition of it—or conversely, freedom is necessary because those laws are necessary as practical postulates. How this consciousness of moral laws, or, what is the same, of freedom, is possible cannot be further explained; only their admissibility can quite readily be A 80 defended in the theoretical Critique.

The exposition of the supreme principle of practical reason has now been accomplished. That is, we have shown, first, what it contains: that it subsists on its own entirely a priori and independently of empirical principles; and then, how it differs from all other practical principles. With the deduction—the justification of its objective and universal validity and the insight into the possibility of such a synthetic proposition a priori—we cannot hope to fare as well as we did with the principles of the pure theoretical understanding. For these referred to objects of possible experience, namely appearances, and it could be proved that only by bringing these appearances under the categories in accordance with those laws can the appearances be cognized as objects of experience; consequently all possible experience must conform to these laws. But I cannot follow such a course with the deduction of the moral law. For it concerns not cognition of the character of objects that may be given to reason in some way from elsewhere, but cognition insofar as it can become the ground of the existence of the objects themselves, and through which reason has causality in a rational being A 81: that is, pure reason, which can be regarded as a faculty immediately determining the will.

Now all human insight reaches its limit as soon as we arrive at fundamental forces V:47 or fundamental faculties. For their possibility cannot be comprehended through anything, yet neither may they be arbitrarily invented and assumed. Hence, in the theoretical use of reason, only experience can entitle us to assume them. But here, with regard to the pure practical faculty of reason, even this substitute—offering empirical proofs instead of a deduction from sources of cognition a priori—is denied us. For anything that needs to draw the ground of proof of its actuality from experience must, as to the grounds of its possibility, depend on principles of experience. Yet pure and nevertheless practical reason cannot, by its very concept, be regarded as such a thing. Moreover, the moral law is given, as it were, as a fact of pure reason of which we are conscious a priori and which is apodictically certain, even supposing that no example of its exact observance could be found in experience. Thus the objective reality of the moral law cannot be proved through any deduction, through any exertion of theoretical reason, whether speculative or supported by experience; nor, even if we were A 82 willing to forgo apodictic certainty, can it be confirmed through experience and so proved a posteriori. Nevertheless, it stands firm by itself.

But something different and wholly paradoxical takes the place of this vainly sought deduction of the moral principle: conversely, the principle itself serves as the principle for deducing an inscrutable faculty that no experience can prove, but that speculative reason had to assume as at least possible in order to find, among its cosmological ideas, the unconditioned in respect of its causality and thus avoid contradicting itself. This is the faculty of freedom, whose possibility and indeed actuality in beings that recognize this law as binding on them is proved by the moral law, which itself needs no justifying grounds. The moral law is in fact a law of causality through freedom, and thus of the possibility of a supersensible nature, just as the metaphysical law of events in the world of sense was a law of the causality of sensible nature. The former therefore determines what speculative philosophy had to leave undetermined: the law for a causality whose concept in speculative philosophy was merely negative. It thereby first confers objective reality on this concept.

V:48 This sort of credential of the moral law, in which the law itself is established as a A 83 principle for deducing freedom as a causality of pure reason, is entirely sufficient in place of any justification a priori to meet a need of theoretical reason, since theoretical reason was compelled at least to assume the possibility of freedom. For the moral law proves its reality to the satisfaction even of the Critique of Speculative Reason by adding a positive determination to a causality thought merely negatively, whose possibility speculative reason could not comprehend and yet had to assume. This positive determination is the concept of a reason immediately determining the will through the condition of a universal lawful form of its maxims. In this way the moral law is able for the first time to give objective, though only practical, reality to reason, which with its ideas always went beyond bounds whenever it wished to proceed speculatively. It transforms reason's transcendent use into an immanent one: being, in the field of experience, causes that are themselves effective through ideas.

The determination of the causality of beings in the world of sense, as such a world, could never be unconditional; yet for every series of conditions there must necessarily be something unconditioned, and thus also a causality determining itself entirely by itself. Hence the A 84 idea of freedom as a faculty of absolute spontaneity was not a need but, as regards its possibility, an analytic principle of pure speculative reason. Yet it is absolutely impossible to give an example conforming to it in any experience, because among the causes of things as appearances no determination of causality can be found that is absolutely unconditioned. We could therefore only defend the thought of a freely acting cause, when applying it to a being in the world of sense insofar as that being is also considered, from another side, as a noumenon. We did so by showing that it is not contradictory to regard all its actions as physically conditioned insofar as they are appearances, yet at the same time to regard their causality as physically unconditioned insofar as the acting being is a being of understanding. In this way we made the concept of freedom a regulative principle of reason. Through it I do not at all cognize what the object to which such causality is attributed is, but I nevertheless remove the obstacle: on the one hand, in explaining events in the world, and hence also the actions of rational V:49 beings, I do justice to the mechanism of natural necessity by proceeding backward from the conditioned to its condition to infinity; on the other hand, I keep open for speculative reason the place empty A 85 for it, namely the intelligible, in which to locate the unconditioned. But I could not give reality to this thought, that is, transform it into cognition of a being acting in this way, even merely as to its possibility. Pure practical reason now fills this empty place with a determinate law of causality in an intelligible world through freedom, namely the moral law. Speculative reason thereby gains nothing in insight, but it does gain in the securing of its problematic concept of freedom, to which objective and, though only practical, nonetheless indubitable reality is here imparted. Practical reason does not even enlarge the concept of causality—whose application, and hence also its meaning, properly takes place only in relation to appearances, in order to connect them into experiences, as the Critique of Pure Reason proves—so as to extend its use beyond those limits. For if it aimed at this, it would have to seek to show how the logical relation of ground and consequence could be used synthetically with a kind of intuition other than sensible intuition, that is, how a causa noumenon is possible. This it cannot accomplish at all; but as practical reason it also takes no account of this question. It merely places the determining ground of the causality of the human being as a sensible being A 86—a causality that is given—in pure reason, which is therefore called practical. Thus it uses the concept of cause itself, from whose application to objects for the sake of theoretical cognition it can here abstract entirely—since this concept is always found in the understanding a priori, independently of all intuition—not to cognize objects, but to determine causality with regard to them in general, and thus for no purpose other than a practical one. It can therefore locate the determining ground of the will in the intelligible order of things, while readily admitting that it does not at all understand what determination the concept of cause might have for the cognition of these things. It must indeed cognize in a determinate way causality with regard to actions of the will in the world of sense, for otherwise practical reason could actually produce no deed. But the concept it forms of its own V:50 causality as a noumenon need not be determined theoretically for the sake of cognizing its supersensible existence, nor need practical reason be able to give it meaning in that respect. For it acquires meaning in any case, though only for practical use, namely through the moral law. Considered even theoretically, it always remains a pure concept of the understanding given a priori, which can be applied A 87 to objects whether they are given sensibly or nonsensibly. In the latter case, however, it has no determinate theoretical meaning or application, but is merely a formal, though still essential, thought of the understanding concerning an object in general. The meaning reason confers on it through the moral law is solely practical: the idea of the law of a causality, that of the will, itself has causality, or is its determining ground.

II. On the Entitlement of Pure Reason, in Its Practical Use, to an Extension Not Possible for It by Itself in Its Speculative Use.

In the moral principle we have established a law of causality that places the determining ground of causality beyond all conditions of the world of sense. We have not merely thought the will as determinable insofar as it belongs to an intelligible world, and thus the subject of this will, the human being, as belonging to a pure world of understanding, though in this respect unknown to us—as could be done according to the Critique of Pure A 88 Speculative Reason. We have also determined it with regard to its causality through a law that cannot be counted among any laws of nature in the world of sense, and have thus extended our cognition beyond the limits of that world, a presumption the Critique of Pure Reason declared null in all speculation. How, then, is the practical use of pure reason here to be reconciled with the theoretical use of that same reason as regards defining the limits of its faculty?

David Hume, of whom it may be said that he truly began all the challenges to the rights of pure reason that made a complete investigation of it necessary, reasoned as follows. The concept of V:51 cause is a concept containing the necessity of the connection of the existence of different things, precisely insofar as they are different, so that if A is posited, I cognize that something entirely different from it, B, must necessarily exist as well. But necessity can be attributed to a connection only insofar as it is cognized a priori; for experience would allow us to cognize only that a connection exists, not that it necessarily exists in this way. Now it is impossible, he says, to cognize a priori and as necessary the connection between one thing and another, or between one determination and another entirely different from it, A 89 when they are not given in perception. Thus the very concept of a cause is false and deceptive; or, to speak of it most leniently, it is a deception excusable to this extent: habit, a subjective necessity arising from frequently perceiving certain things or their determinations as associated in their existence alongside or after one another, is imperceptibly taken for an objective necessity of positing such a connection in the objects themselves. The concept of a cause is thus surreptitiously obtained, not legitimately acquired; indeed, it can never be acquired or authenticated, because it demands a connection that is in itself null, chimerical, and indefensible before any reason, to which no object can ever correspond.—In this way empiricism was first introduced as the sole source of principles with regard to all cognition concerning the existence of things—mathematics thus still remained excepted—but with it also the most severe skepticism, even with regard to the whole of natural science as philosophy. For according to such principles we can never infer a consequence from given determinations of things as to their existence, since this would require the concept of a cause, which contains the necessity of such a connection. We can only, according to the A 90 rule of imagination, expect cases similar to those that have occurred before; but this expectation is never certain, however often it may have been fulfilled. Indeed, of no event could we say that something must have preceded it upon which it necessarily followed, that is, that it must have a cause. Thus even if we knew ever so many cases in which something of this sort preceded it, so that a rule could be drawn from them, we could not on that account assume that it always and necessarily happens in this way. We would therefore also have to allow blind chance its due—a chance with which all use of reason V:52 ceases. This firmly establishes and renders irrefutable skepticism concerning inferences ascending from effects to causes.

Mathematics had so far escaped unscathed because Hume held that its propositions were all analytic, that is, that they proceeded from one determination to another on account of identity, and hence according to the principle of contradiction. But this is false, since they are instead all synthetic. Even though geometry, for example, deals not with the existence of things but only with their determination a priori in a possible intuition, it nevertheless passes, just as through causal concepts, from one determination A to an entirely different one B as nonetheless necessarily connected A 91 with it. But in the end that science, so highly praised for its apodictic certainty, must also succumb to empiricism in principles, for the very same reason that Hume substituted habit for objective necessity in the concept of cause. Despite all its pride, it must submit to lowering its bold claims that command assent a priori, and must await approval of the universal validity of its propositions from the favor of observers. As witnesses, they would surely not refuse to admit that they had always perceived what the geometer presents as principles to be so, and consequently would permit the expectation that it would continue to be so, although it was not strictly necessary. In this way Hume's empiricism in principles also leads inevitably to skepticism even with regard to mathematics, and consequently in all scientific theoretical use of reason, since this belongs either to philosophy or to mathematics. Whether the ordinary use of reason would fare any better amid so dreadful an overthrow as we see befalling the leading branches of cognition, or would instead become still more irretrievably entangled in this same destruction of all knowledge, and hence whether a universal skepticism must follow from the same principles—though this would of course A 92 affect only the learned—I leave everyone to judge for himself.

As for my treatment in the Critique of Pure Reason, which was occasioned by Hume's skeptical doctrine but went much further, embracing the entire field of pure theoretical reason in its synthetic use, and thus also what is called metaphysics V:53 in general, I proceeded as follows with regard to the Scottish philosopher's doubts concerning the concept of causality. Hume was entirely right to declare the concept of cause deceptive and a false illusion if he took the objects of experience—as indeed is almost universally done—for things in themselves. For with things in themselves and their determinations as such, we cannot discern how, because something A is posited, something else B must necessarily also be posited; and thus he could not admit any such cognition a priori of things in themselves. Still less could that acute thinker allow this concept an empirical origin, because such an origin directly contradicts the necessity of connection that constitutes the essence of the concept of causality. The concept was therefore outlawed, and habit in observing the course of perceptions took its place.

From my investigations, however, it emerged that the objects A 93 with which we deal in experience are by no means things in themselves but mere appearances. Although with things in themselves we cannot at all discern—indeed it is impossible to discern—how, if A is posited, it should be contradictory not to posit B, which is entirely different from A, this being the necessity of the connection between A as cause and B as effect, we can nevertheless quite well think that, as appearances in one experience, they must necessarily be connected in a certain way, for example in regard to temporal relations, and cannot be separated without contradicting that connection by means of which this experience is possible, in which they are objects and alone cognizable to us. And so it was indeed found to be. I could therefore not only prove the concept of cause in its objective reality with regard to objects of experience, but also deduce it as a concept a priori on account of the necessity of connection it carries with it: that is, demonstrate its possibility from pure understanding without empirical sources. Having thus removed the empiricism of its origin, I could eradicate its inevitable consequence, skepticism, first with regard to natural science and then also, because what followed did so entirely from the same grounds, with regard to mathematics—both sciences being referred A 94 to objects of possible experience—and thereby eradicate total V:54 doubt about everything theoretical reason claims to discern.

But what becomes of the application of this category of causality—and likewise of all the others, for without them no cognition of what exists can be achieved—to things that are not objects of possible experience but lie beyond its limits? For I was able to deduce the objective reality of these concepts only with regard to objects of possible experience. But precisely this, that I have rescued them even if only in this case, that I have shown that through them objects can at least be thought, though not determined a priori, is what gives them a place in the pure understanding, from which they are referred to objects in general, sensible or nonsensible. If anything is still lacking, it is the condition for the application of these categories, and specifically that of causality, to objects: namely intuition. Where intuition is not given, it makes application for the sake of theoretical cognition of the object as a noumenon impossible; such application, if anyone ventures it, is therefore completely forbidden, as it was in the Critique of Pure Reason. Meanwhile, the objective reality of the concept always remains, and it can also A 95 be used of noumena, but without our being able to determine this concept theoretically in the least and thereby produce cognition. For that this concept contains nothing impossible even in relation to an object was proved by securing its seat in the pure understanding throughout its application to objects of the senses. Even if, when subsequently referred to things in themselves, which cannot be objects of experience, it is incapable of determination for the representation of a determinate object for the sake of theoretical cognition, it could nevertheless still be capable of a determination for its application for some other purpose, perhaps a practical one. This would not be so if, as Hume held, this concept of causality contained something utterly impossible to think.

To discover this condition for applying the concept in question to noumena, we need only look back at why we are not satisfied with applying it to objects of experience, but would also like to use it of things in themselves. For it soon becomes apparent that it is not a theoretical but a practical purpose that makes this necessary for us. V:55 For speculation, even if we succeeded, we would still A 96 make no genuine gain in cognition of nature, or generally with regard to any objects that might be given to us. At most we would take a great step from the sensibly conditioned—where remaining and diligently traversing the chain of causes already gives us enough to do—to the supersensible, in order to complete and delimit our cognition on the side of its grounds. Yet an infinite gulf between that boundary and what we know would always remain unfilled, and we would have listened more to a vain passion for questions than to a thoroughgoing desire for knowledge.

But besides the relation in which the understanding stands to objects in theoretical cognition, it also has a relation to the faculty of desire, which is therefore called the will, and the pure will insofar as the pure understanding—which in such a case is called reason—is practical through the mere representation of a law. The objective reality of a pure will, or, what is the same, of a pure practical reason, is given a priori in the moral law as though through a fact; for this is what we may call a determination of the will that is unavoidable although it does not rest on empirical principles. But the concept of a will already contains the concept of causality; consequently the concept of a pure will contains the concept of a causality with freedom, that is, one not A 97 determinable according to laws of nature and hence incapable of any empirical intuition as proof of its reality. Yet in the pure practical law a priori this concept fully justifies its objective reality, though, as is easy to see, not for the theoretical but solely for the practical use of reason. Now the concept of a being that has free will is the concept of a causa noumenon. We are already assured that this concept does not contradict itself, because the concept of a cause, as originating wholly from the pure understanding and at the same time secured through the deduction in its objective reality with regard to objects in general, is in its origin independent of all sensible conditions. Hence it is not in itself restricted to phenomena, except where a theoretically determinate use is intended, and can indeed be applied to things as pure beings of understanding. But since no intuition, which can always be only sensible, can be laid at the foundation of this application, a causa noumenon is, with regard to the theoretical use of reason, V:56 a possible and thinkable but nevertheless empty concept. Now I do not seek through it to know theoretically the character of a being insofar as it has a pure will. It is enough for me A 98 merely to designate it through this concept as such a being, and thus merely to connect the concept of causality with that of freedom and, inseparably from it, with the moral law as its determining ground. I am certainly entitled to do this by virtue of the pure, nonempirical origin of the concept of cause, insofar as I regard myself as authorized to make no use of it other than in relation to the moral law that determines its reality, that is, only a practical use.

Had I, with Hume, deprived the concept of causality of objective reality in theoretical use, not only with regard to things in themselves, the supersensible, but also with regard to objects of the senses, it would have lost all meaning and been declared entirely unusable as a theoretically impossible concept. And since no use can be made of nothing, the practical use of a theoretically null concept would have been utterly absurd. But as things stand, the concept of an empirically unconditioned causality is indeed theoretically empty, without an intuition suited to it, yet still possible and related to an indeterminate object; and in place of that object, meaning is nevertheless given to it in the moral law, and hence in a practical respect. Thus although I have no intuition that would determine its objective theoretical reality, it nonetheless has actual A 99 application, which can be exhibited in the concrete in dispositions or maxims: that is, practical reality that can be indicated. This suffices to justify it even with regard to noumena.

But once this objective reality of a pure concept of the understanding has been introduced into the field of the supersensible, it now also gives all the remaining categories objective reality, though only a practically applicable reality, and always only insofar as they stand in a necessary connection with the determining ground of the pure will, the moral law. Yet this has not the slightest influence in extending theoretical cognition of these objects, as insight into their nature through pure reason. As we will also find in what follows, the categories always refer only to beings as intelligences, and in these only to the relation of reason to the V:57 will, hence always only to the practical; beyond this they claim no cognition of these beings. Whatever further properties belonging to the theoretical manner of representing such supersensible things might be brought in along with them, all these must then be counted not as knowledge, but only as an entitlement—and in a practical respect even a necessity—to assume and presuppose them. This holds even where supersensible A 100 beings, such as God, are assumed according to an analogy, that is, according to the pure relation of reason that we employ practically with regard to sensible beings. Thus, by applying pure theoretical reason to the supersensible, but only for a practical purpose, we give it not the slightest encouragement to lose itself in extravagance beyond all bounds.

Of the Analytic of Practical Reason

7 Propositions called practical in mathematics or natural science should properly be called technical. For these disciplines are not at all concerned with determining the will; they indicate only the manifold of possible action sufficient to produce a certain effect, and are therefore just as theoretical as all propositions stating the connection of a cause with an effect. Whoever desires the latter must also consent to be the former.

Book I. V:19 A 35 The Analytic of Pure Practical Reason.

Chapter I. On the Principles of Pure Practical Reason.

§1 Definition.

Practical principles are statements containing a general determination of the will that includes several practical rules. They are subjective, or maxims, when the subject regards the condition as valid only for their own will. They are objective, or practical laws, when the condition is recognized as objectively valid, that is, valid for the will of every rational being.

Remark.

If we assume that pure reason can contain a ground that is practically sufficient, that is, sufficient for determining the will, then there are practical A 36 laws. Otherwise, all practical principles will be mere maxims. In a rational being whose will is affected by feelings and desires, maxims can conflict with practical laws that the being itself recognizes. For example, someone can adopt the maxim never to let an insult go unavenged, while also seeing that this is not a practical law but only their own maxim. As a rule for the will of every rational being, it could not be consistent with itself in one and the same maxim. In cognition of nature, the principles of what happens—for example, the principle that action and reaction are equal in the transfer of motion—are also laws of nature. That is because the use V:20 of reason there is theoretical and determined by the character of the object. In practical cognition, which deals only with determining grounds of the will, principles we adopt are not, just for that reason, laws to which we are unavoidably subject. That is because in practical matters reason deals with the subject, namely the faculty of desire. The rule can vary in many ways according to that faculty's particular character.—A practical rule is always a product of reason, because it prescribes an action as a means to an effect intended as an end. But for a being in whom reason is not the sole determining ground of the will, this rule is an imperative. That is, it is a rule marked by an “ought,” which expresses the objective necessitation to act. It means that if reason completely determined the will, the action would inevitably follow this rule. Imperatives therefore hold objectively and are entirely different from maxims, A 37 which are subjective principles. Imperatives determine either the conditions of a rational being's causality as an effective cause, solely with regard to the effect and what is sufficient to produce it, or simply the will, whether or not it is sufficient to produce the effect. The first would be hypothetical imperatives and contain only instructions for skill. The second, by contrast, would be categorical and alone would be practical laws. So maxims are principles, but not imperatives. Imperatives themselves, when conditional, determine the will not simply as will but only in relation to a desired effect. These hypothetical imperatives are practical instructions, but not laws. Laws must determine the will sufficiently as will, before I even ask whether I have the ability needed for a desired effect, or what I must do to produce it. They must therefore be categorical. Otherwise they are not laws, because they lack necessity. If necessity is to be practical, it must be independent of conditions arising from feelings and desires, which attach to the will only contingently. Tell someone, for example, that they must work and save when young so as not to suffer want in old age. This is a correct and also important practical instruction for the will. But we can easily see that the will is being directed toward something else, which we assume the person desires. We must leave that desire to the person acting: perhaps they foresee other resources besides wealth earned by their own efforts; perhaps they do not hope to grow old at all; or perhaps they think they can get by poorly if they should ever be in need. Reason is the only source of any rule A 38 that is to contain necessity. It does put necessity into this instruction too, since otherwise it would not be an imperative. But this necessity is only subjectively conditioned, and we cannot assume it to the same degree in all subjects. For reason to give laws, however, it must need to presuppose only itself. V:21 This is because a rule is objectively and universally valid only when it holds without the contingent, subjective conditions that distinguish one rational being from another. Now tell someone never to make a lying promise. This rule concerns only their will, whether or not it can achieve whatever aims they may have. Willing alone is what this rule is to determine entirely a priori. If this rule proves practically correct, then it is a law, because it is a categorical imperative. Thus practical laws concern only the will, regardless of what its causality accomplishes. We can leave that causality aside, since it belongs to the sensible world, in order to have these laws in their purity.

§2 Theorem I.

All practical principles that presuppose an object (matter) of the faculty of desire as the determining ground of the will are empirical, without exception, and cannot provide practical laws.

By the matter of the faculty of desire I mean an object whose reality is desired. Now suppose the desire for this object comes before the practical rule and is the condition A 39 for adopting that rule as a principle. Then I say, first, that this principle is always empirical. For in that case the determining ground of choice is the representation of an object, together with that representation's relation to the subject through which the faculty of desire is determined to make the object real. Such a relation to the subject is called pleasure in the reality of an object. So this pleasure would have to be presupposed as a condition that makes the determination of choice possible. But no representation of any object, whatever it may be, allows us to know a priori whether it will involve pleasure, displeasure, or indifference. In such a case, then, the determining ground of choice must always be empirical. So too must the material practical principle that presupposes it as a condition.

Now, second, suppose a principle rests only on the subjective condition of being receptive to pleasure or displeasure. This condition can always be known only empirically and cannot hold in the same way for all rational beings. Such a principle can certainly serve as a maxim for the subject who has that receptivity. But it cannot serve as a law, even for that subject, because it lacks objective V:22 necessity, which must be known a priori. So a principle of this kind can never provide a practical A 40 law.

§3 Theorem II.

All material practical principles, as such, are of one and the same kind. They fall under the general principle of self-love or personal happiness.

Pleasure in the representation of a thing's existence, insofar as it is to be a determining ground for desiring that thing, rests on the subject's receptivity, because it depends on the existence of an object. It therefore belongs to sense (feeling), not to the understanding. Understanding expresses a relation of the representation to an object in terms of concepts, not to the subject in terms of feelings. So this pleasure is practical only insofar as the feeling of enjoyment that the subject expects from the object's reality determines the faculty of desire. Now a rational being's awareness of enjoyment in life, continuing uninterrupted throughout its whole existence, is happiness. The principle of making this the highest determining ground of choice is the principle of self-love. Thus all material principles that locate the determining ground of choice in the pleasure or displeasure expected from some object's reality A 41 are entirely of the same kind in this respect: they all belong to the principle of self-love or personal happiness.

Corollary.

All material practical rules locate the determining ground of the will in the lower faculty of desire. If there were no merely formal laws of this faculty that sufficiently determined the will, then we could not acknowledge any higher faculty of desire either.

Remark I.

It is surprising that otherwise perceptive thinkers believe they can distinguish the lower from the higher faculty of desire V:23 by asking whether the representations connected with the feeling of pleasure originate in the senses or in the understanding. When we ask about the determining grounds of desire and locate them in the enjoyment expected from something, it makes no difference where the representation of this pleasing object comes from. All that matters is how much it pleases. A representation may have its seat and origin in the understanding. But if it can determine choice only by presupposing a feeling of pleasure in the subject, then its being a determining ground of choice depends entirely on the character of inner sense: namely, on its capacity to be pleasantly affected by that representation. Representations of objects may be of very different kinds. They may be representations of the understanding, A 42 or even of reason, as opposed to representations of the senses. Still, the feeling of pleasure through which alone they actually determine the will is of the same kind. This is the enjoyment, the pleasure expected from them, that drives the activity of producing the object. These feelings are alike not only because they can always be known only empirically, but also because they affect one and the same vital force, which expresses itself in the faculty of desire. In this respect, each can differ from any other determining ground only in degree. Otherwise, how could we compare two determining grounds that are completely different in their kind of representation, in terms of their magnitude, so as to prefer the one that affects the faculty of desire most? The very same person can return unread an instructive book that comes into their hands only once, so as not to miss a hunt. They can leave in the middle of a fine speech so as not to be late for a meal. They can leave an entertaining rational conversation, which they otherwise value highly, to sit down at the gaming table. They can even turn away a poor person whom they would otherwise enjoy helping, because they have only enough money in their pocket to pay for admission to a play. If the determination of their will rests on the feeling of enjoyment or unpleasantness they expect from some cause, it makes no difference to them what kind of representation affects them. To make their choice, they care only about how intense and lasting the enjoyment is, how easily it can be obtained, and how often it can be repeated. Someone who needs gold to spend is completely indifferent to whether its material, A 43 the gold, was dug out of a mountain or washed out of sand, provided it is everywhere accepted at the same value. In just the same way, someone concerned only with enjoying life does not ask whether representations come from understanding or sense, but only how much pleasure, and how great a pleasure, they will provide for the longest time. Only those who would like to deny pure reason the ability to determine the will without presupposing any feeling V:24 can stray so far from their own definition. They first bring things under one and the same principle, only to declare afterward that they are entirely different in kind. For example, we find that people can take pleasure even in the mere exercise of their powers: in awareness of their mental strength when overcoming obstacles to their purposes, in developing their intellectual talents, and so on. We rightly call these finer pleasures and enjoyments. More than other pleasures, they are under our control. They do not wear out; instead, they strengthen our capacity to enjoy them further, and they develop us at the same time as they please us. But this does not justify presenting them as a way of determining the will different from determination merely through sense. These pleasures themselves depend for their very possibility on a feeling built into us, which is the first condition of this satisfaction. To call this a different kind of determination is just like ignorant people who want to dabble in metaphysics: they imagine matter so fine, so excessively fine that the very thought could make them dizzy, and then believe they have conceived a spiritual yet extended being. If, like Epicurus, we rely on virtue's promised pleasure alone to determine A 44 the will, we cannot then criticize him for treating that pleasure as entirely the same kind as the pleasures of the crudest senses. For we have no reason to charge him with attributing the representations that arouse this feeling in us only to the bodily senses. As far as we can guess, he sought the source of many of them just as much in the use of the higher cognitive faculty. But that did not and could not prevent him, on the principle just mentioned, from regarding pleasure itself as entirely the same kind throughout. This includes the pleasure those possibly intellectual representations give us, through which alone they can be determining grounds of the will. To be consistent is a philosopher's greatest obligation, yet it is the one most rarely met. The ancient Greek schools give us more examples of it than we find in our syncretistic age. In this age a certain coalition system of conflicting principles, full of dishonesty and shallowness, is contrived because it recommends itself better to a public content to know a little of everything and nothing as a whole, while remaining firmly seated in familiar positions. However much understanding and reason the principle of personal happiness may employ, it would still contain no determining grounds of the will except those suited to the lower faculty of desire. So either there is no higher faculty of desire at all, or pure reason must be practical on its own. That is, it must be able to determine A 45 the will through the mere form of a practical rule, without presupposing any feeling. It must therefore do so without representations of the pleasant or unpleasant as the matter of the faculty of desire, which is always an empirical condition of principles. Only then is reason a true higher faculty of desire, and only insofar as it determines V:25 the will by itself, rather than serving the inclinations. The faculty that can be determined by feelings and desires is subordinate to it. Reason is genuinely, indeed specifically, different from that faculty. Even the slightest mixture of the latter's impulses diminishes reason's strength and superiority, just as the slightest empirical element introduced as a condition into a mathematical proof lowers and destroys its standing and force. In a practical law, reason determines the will directly, not through an intervening feeling of pleasure or displeasure, not even a feeling about this law itself. Only its ability to be practical as pure reason makes it possible for reason to give laws.

Remark II.

To be happy is necessarily the wish of every rational but finite being, and so an unavoidable determining ground of its faculty of desire. For satisfaction with its whole existence is not something it possesses from the start. It is not a state of bliss that would presuppose awareness of its independent self-sufficiency. Instead, its finite nature itself imposes this problem on it, because it has needs. These needs concern the matter of its faculty of desire: something related to an underlying subjective feeling of pleasure or displeasure, which determines what it needs to be satisfied with its condition. But precisely because the subject can know this material determining ground only empirically, it is impossible to regard this task as a law, since a law, as objective, would have to contain the very same A 46 determining ground of the will in all cases and for all rational beings. For the concept of happiness does indeed underlie the practical relation of objects to the faculty of desire everywhere. But it is only a general heading for subjective determining grounds. It determines nothing specifically, although such specific determination is all that matters in this practical problem, and without it the problem cannot be solved at all. What each person is to count as their happiness depends on their own particular feeling of pleasure and displeasure. Even in one and the same subject, it depends on differences in need as this feeling changes. A subjectively necessary law, as a law of nature, is therefore objectively a highly contingent practical principle. It can and must differ greatly between subjects, and so can never provide a law. For the desire for happiness is concerned not with the form of lawfulness but only with the matter: whether, and how much, pleasure I can expect from following the law. Principles of self-love can certainly contain general rules of skill for finding means to ends. V:26 But in that case they are merely theoretical principles⁷—for example, how someone who wants to eat bread should devise a mill. A 47 Practical instructions based on them, however, can never be universal. For the determining ground of the faculty of desire rests on the feeling of pleasure and displeasure, which can never be assumed to be universally directed at the same objects.

But suppose finite rational beings all thought exactly alike about what to take as objects of their feelings of pleasure or pain, and even about the means they must use to obtain the former and avoid the latter. Even then they could by no means present the principle of self-love as a practical law. For this agreement would itself still be contingent. The determining ground would still be only subjectively valid and merely empirical. It would lack the necessity conceived in every law: objective necessity based on a priori grounds. Unless, that is, we treated this necessity not as practical at all but as merely physical: our inclination would compel the action just as inevitably as we are made to yawn when we see others yawn. We could more readily maintain that there are no practical laws at all, but only recommendations serving our desires, than raise merely subjective principles to the rank of practical laws. Such laws have objective, not merely subjective, necessity without exception. They must be known by reason a priori, not through experience, however empirically universal that experience may be. Even rules describing uniform appearances are called laws of nature—for example, mechanical laws—only when we either actually know them a priori, or assume, as with A 48 chemical laws, that they would be known a priori from objective grounds if our insight went deeper. But merely subjective practical principles expressly require that they rest on subjective, not objective, conditions of choice. So they may always be presented only as mere maxims, never as practical laws. At first this last remark seems like mere quibbling over words. But it defines in words the most important distinction, one that comes into consideration only in practical investigations.

§4 V:27 Theorem III.

If a rational being is to think of its maxims as universal practical laws, it can think of them only as principles whose determining ground of the will lies not in their matter but solely in their form.

The matter of a practical principle is the object of the will. This either is or is not the will's determining ground. If it is, then the rule of the will would be subject to an empirical condition: the relation of the determining representation to the feeling of pleasure and displeasure. It would therefore not be a practical law. Now if we remove from a law all its matter, that is, every object of the will as a determining ground, nothing remains but the mere form of giving universal law. A 49 So a rational being either cannot think of its subjective practical principles, or maxims, as universal laws at the same time, or must assume that their mere form, by which they are suited to giving universal law, makes them practical laws on its own.

Remark.

Even the most ordinary understanding can distinguish, without instruction, which form in a maxim is suited to giving universal law and which is not. For example, I have made it my maxim to increase my wealth by every safe means. Now I hold a deposit whose owner has died without leaving a written record of it. Naturally, this is a case for my maxim. All I now want to know is whether that maxim can also hold as a universal practical law. So I apply it to the present case and ask whether it could take the form of a law. Could I, through my maxim, also give this law: everyone may deny holding a deposit when no one can prove it was entrusted to them? I immediately see that this principle, as a law, would destroy itself, because it would mean there were no deposits at all. A practical law that I recognize as such must qualify for giving universal law. This is an identical proposition and so is clear in itself. If I now say that my will is subject to a practical law, I cannot cite my inclination—in this case, my greed—as its determining ground suitable for a universal practical law. V:28 For, far from being A 50 suited to giving universal law, this inclination would destroy itself in the form of a universal law.

The desire for happiness is universal, and so is the maxim by which everyone makes happiness the determining ground of their will. Still, it is strange that intelligent people could think of presenting it, on that account, as a universal practical law. For a universal law of nature otherwise brings everything into agreement. But here, if we tried to give the maxim the universality of a law, precisely the opposite of agreement would result: the worst conflict, and the complete destruction of the maxim itself and its aim. For in that case the wills of all do not share one and the same object. Instead, each has their own object—their own well-being. This may happen to be compatible with other people's aims, which are likewise directed toward themselves. But that falls far short of being enough for a law, because the exceptions we are entitled to make as occasions arise are endless and cannot be precisely included in a universal rule. What results is a harmony like the one a certain satirical poem describes in the unity of minds between two spouses ruining each other: Oh, wonderful harmony: what he wants, she wants too, etc. Or it is like the declaration King Francis I is said to have made to Emperor Charles V: what my brother Charles wants—Milan—I want too. Empirical determining grounds are not suited to any universal external lawgiving, but neither are they suited to internal lawgiving. For each person bases inclination on their own subject, while another bases it on a different subject. A 51 Even within each subject, now one inclination and now another has the greatest influence. It is simply impossible to find a law that would govern them all under this condition, namely with agreement on every side.

§5 Problem I.

Assuming that the mere lawgiving form of maxims alone is the sufficient determining ground of a will, find the character of a will that can be determined by this alone.

The mere form of a law can be represented only by reason. It is therefore not an object of the senses, and so does not belong among appearances. Its representation as the determining ground of the will thus differs from all determining grounds of events in nature under the law of causality. V:29 For in those events the determining grounds must themselves be appearances. Now if no determining ground of the will other than this universal lawgiving form alone can serve as a law for it, such a will must be thought of as entirely independent of the natural law governing appearances in relation to one another: the law of causality. This independence is called freedom in the strictest, that is, transcendental, sense. Thus a will A 52 for which the mere lawgiving form of its maxim alone can serve as a law is a free will.

§6 Problem II.

Assuming that a will is free, find the law that alone is suited to determine it necessarily.

The matter of a practical law, that is, an object of a maxim, can be given only empirically. But a free will, though independent of empirical conditions—that is, conditions belonging to the sensible world—must still be capable of determination. So, independently of the matter of the law, a free will must still find a determining ground in the law. Yet apart from its matter, the law contains nothing but its lawgiving form. So the lawgiving form, insofar as it is contained in the maxim, is the only thing that can constitute a determining ground of the will.

Remark.

Freedom and unconditional practical law thus point back to each other. I am not asking here whether they really are different, or whether an unconditional law is simply the self-consciousness of pure practical reason, and this in turn is identical with the positive concept of freedom. Instead, I ask where our cognition of the unconditionally practical begins: with freedom or with practical law. It cannot begin with freedom. A 53 For we cannot be directly conscious of freedom, since its first concept is negative. Nor can we infer it from experience, because experience gives us cognition only of the law of appearances, and thus of nature's mechanism, the exact opposite of freedom. So it is the moral law, of which we become directly conscious as soon as we formulate maxims of the will, that presents itself to us first. Reason V:30 presents this law as a determining ground that no sensible conditions can outweigh, and that is indeed entirely independent of them. In doing so, it leads directly to the concept of freedom. But how is consciousness of this moral law itself possible? We can become conscious of pure practical laws just as we are conscious of pure theoretical principles: by attending to the necessity with which reason prescribes them to us, and to the exclusion of all empirical conditions, to which reason directs us. The concept of a pure will arises from the former, just as consciousness of a pure understanding arises from the latter. This is the true order of dependence among our concepts: morality first reveals the concept of freedom to us. Thus practical reason first confronts speculative reason with its most insoluble problem through this concept, putting it in the greatest difficulty. This is already clear from the following consideration. Nothing in appearances can be explained by the concept of freedom; nature's mechanism must always be our guide here. Moreover, when pure reason tries to ascend to the unconditioned in the series of causes, its antinomy entangles it in things beyond comprehension on either view. Yet the latter view, mechanism, at least A 54 has a use in explaining appearances. So no one would ever have ventured to introduce freedom into science if the moral law, and with it practical reason, had not come along and forced this concept upon us. Experience also confirms this order of concepts in us. Suppose someone claims that their lustful inclination is completely irresistible whenever the desired object and the opportunity are present. Ask whether they would not control their inclination if a gallows stood in front of the house where they found this opportunity, and they were to be hanged on it immediately after satisfying their lust. We need not spend long guessing their answer. But now ask this: suppose their ruler demanded, under threat of the same immediate death penalty, that they give false testimony against an honest man whom the ruler wanted to destroy on plausible pretexts. Would they consider it possible to overcome their love of life, however great it might be? Perhaps they would not dare to assert whether they would do it or not. But they must admit without hesitation that it would be possible for them. They therefore judge that they can do something because they are conscious that they ought to do it. They recognize in themselves the freedom that, without the moral law, would otherwise have remained unknown to them.

§7 Fundamental Law of Pure Practical Reason.

Act so that the maxim of your will could always hold at the same time as a principle for giving universal law.

V:31

Remark. A 55

Pure geometry has postulates as practical propositions. But they contain nothing beyond the presupposition that we can do something if we should be required to do it. These are its only propositions that concern existence. They are therefore practical rules under a problematic condition of the will. Here, however, the rule says that we ought, unconditionally, to act in a certain way. So the practical rule is unconditional. It is therefore represented as a categorical practical proposition a priori, through which the will is objectively determined absolutely and directly—by the practical rule itself, which is therefore a law here. For pure reason, practical in itself, directly gives laws here. The will is thought of as independent of empirical conditions, and thus, as a pure will, as determined by the mere form of the law. This determining ground is regarded as the highest condition of all maxims. This is strange enough, and has no parallel anywhere else in practical cognition. For the a priori thought of a possible universal lawgiving, which is therefore merely problematic, is commanded unconditionally as a law, without borrowing anything from experience or from any external will. Nor is it an instruction saying that an action ought to occur through which a desired effect becomes possible, since then the rule would always be physically conditioned. Instead, it is a rule that determines only the will a priori, with regard to the form of its maxims. Now it is at least not impossible to think of a law that serves only the subjective form of principles, yet acts as a determining ground through the objective form of a law in general. We can call consciousness of this fundamental law a fact of reason. A 56 For we cannot reason it out from previously given data of reason, such as consciousness of freedom, since that is not given to us beforehand. Instead, it forces itself upon us on its own as a synthetic proposition a priori, grounded on no intuition, either pure or empirical. It would indeed be analytic if we presupposed freedom of the will. But freedom as a positive concept would require an intellectual intuition, which we must not assume here at all. Still, to regard this law as given without misunderstanding it, we must note carefully that it is not an empirical fact. It is the sole fact of pure reason, which thereby announces itself as originally giving laws: this is what I will; this is what I command.

Corollary.

Pure reason is practical on its own and gives human beings a universal law, which we call the moral law.

V:32

Remark.

The fact just mentioned cannot be denied. We need only analyze people's judgments about whether their actions conform to law. We will always find that, whatever inclination may say, their reason remains incorruptible and compelled by itself to measure the maxim of the will in an action against the pure will—that is, against itself, insofar as it regards itself as practical a priori. Now reason also declares this principle of morality to be a law for all rational A 57 beings, precisely because of the universality of the lawgiving that it makes the highest formal determining ground of the will, regardless of all subjective differences in it. This applies insofar as they have a will at all: an ability to determine their causality through the representation of rules. Thus it applies insofar as they are capable of actions according to principles, and therefore also according to practical principles a priori, since these alone have the necessity that reason requires of a principle. So it is not restricted to human beings. It extends to all finite beings that have reason and will, and even includes the infinite being as the supreme intelligence. In the former case, however, the law has the form of an imperative. For we can presuppose a pure will in a finite being as rational, but not a holy will in it as a being affected by needs and sensible motives. A holy will would be incapable of any maxim conflicting with the moral law. For finite beings, therefore, the moral law is an imperative that commands categorically, because the law is unconditional. Such a will's relation to this law is one of dependence, called obligation. Obligation means a necessitation to an action, though only through reason and its objective law. This action is called duty, because a power of choice affected by feelings and desires—though not determined by them, and thus always free—carries with it a wish arising from subjective causes. This wish can therefore often oppose the pure objective determining ground. So it requires resistance from practical reason as moral necessitation. This resistance can be called an internal but intellectual constraint. In the all-sufficient intelligence, choice is rightly represented as incapable of any maxim that could not also be an objective law. The concept of holiness that belongs to it on this account places it A 58 above, not all practical laws, but all practically restrictive laws, and thus beyond obligation and duty. Yet this holiness of the will is a practical idea that must necessarily serve as the model. All that finite rational beings can do is approach it without end. The pure moral law, which is therefore itself called holy, constantly and correctly holds this model before them. To be sure of this endless progress V:33 in one's maxims, and of their constancy in continual advancement—that is, virtue—is the highest thing finite practical reason can achieve. Virtue itself, however, at least as a naturally acquired capacity, can never be complete. For assurance in such a case never becomes absolute certainty, and as persuasion it is very dangerous.

§8 Theorem IV.

The autonomy of the will is the sole principle of all moral laws and the duties that conform to them. By contrast, all heteronomy of choice not only fails to ground any obligation, but actually opposes the principle of obligation and the morality of the will. The sole principle of morality consists in independence from all matter of the law—that is, from a desired object—together with determination of choice by the mere universal lawgiving form of which a maxim must be capable. That independence is freedom in the negative sense. This lawgiving of its own A 59 by pure reason, practical as such, is freedom in the positive sense. Thus the moral law expresses nothing other than the autonomy of pure practical reason, that is, freedom. This is itself the formal condition of all maxims, under which alone they can agree with the highest practical law. Now the matter of willing can only be the object of a desire connected with the law. So if it enters into the practical law as a condition of the law's possibility, the result is heteronomy of choice: dependence on the natural law of following some impulse or inclination. The will then gives itself not the law, but only an instruction for rationally following laws of feeling and desire. A maxim formed in this way can never contain the form of universal lawgiving. It therefore not only fails to establish any obligation, but itself opposes the principle of pure practical reason, and thus also a moral disposition, even if the action arising from it should conform to the law.

V:34

Remark I.

So a practical instruction that carries a material, and therefore empirical, condition must never be counted as a practical law. For the A 60 law of the pure will, which is free, places that will in an entirely different sphere from the empirical one. The necessity it expresses is not to be natural necessity. It can therefore consist only in formal conditions that make a law in general possible. All matter of practical rules always rests on subjective conditions. These give it no universality for rational beings except conditional universality: if I desire this or that, what must I then do to make it real? All such rules revolve around the principle of personal happiness. Now it certainly cannot be denied that all willing must also have an object, and thus matter. But that does not mean this matter is the determining ground and condition of the maxim. If it is, the maxim cannot be presented in a universally lawgiving form. For then the expectation of the object's existence would be the determining cause of choice. Willing would have to rest on the dependence of the faculty of desire on the existence of some thing. This dependence can always be sought only in empirical conditions and so can never provide the ground for a necessary and universal rule. Thus the happiness of other beings can be the object of a rational being's will. But if it were the maxim's determining ground, we would have to presuppose that we find not only a natural pleasure but also a need in others' well-being, as the sympathetic temperament in human beings involves. Yet I cannot presuppose this need in every rational being, and not in God at all. So the matter of the maxim can remain, but it must not be the maxim's condition. Otherwise A 61 the maxim would not be suited to be a law. The mere form of a law, which restricts the matter, must therefore also be a ground for adding this matter to the will, but not for presupposing it. Let the matter be, for example, my own happiness. If I attribute this to everyone—as I indeed may in the case of finite beings—it can become an objective practical law only if I include the happiness of others in it. Thus the law of promoting others' happiness does not arise from presupposing that this is an object of everyone's choice. It arises solely because the form of universality becomes the determining ground of the will. Reason requires this form as the condition for giving a maxim of self-love the objective validity of a law. So the object, others' happiness, was not the determining ground of the pure will. It was solely the mere lawful form by which I restricted my maxim based on inclination, to give it the universality V:35 of a law and so make it conform to pure practical reason. Only from this restriction, and not from adding an external incentive, could the concept of an obligation to extend my maxim of self-love to others' happiness then arise.

Remark II.

The direct opposite of the principle of morality is making personal happiness the determining ground of the will. As I showed above, this includes absolutely everything that locates the determining ground that is to serve as a law anywhere other than in the lawgiving form of the maxim. This conflict is not merely logical, like the A 62 conflict involved in trying to raise empirically conditioned rules to necessary principles of cognition. It is practical. It would completely destroy morality if reason's voice regarding the will were not so clear, so impossible to drown out, and so audible even to the most ordinary person. As things stand, this view can survive only in the bewildering speculations of the schools. They are bold enough to turn a deaf ear to that heavenly voice in order to uphold a theory that saves them the trouble of hard thinking.

Suppose a friend whose company you otherwise enjoy tried to justify having given false testimony as follows. First they invoked what they claimed was the sacred duty of securing their own happiness. Then they listed all the advantages they had gained, and pointed to the clever precautions they had taken against any discovery—even discovery by you. They had told you the secret only so that they could deny it at any time. Finally, in complete seriousness, they claimed to have fulfilled a genuine human duty. You would either laugh right in their face or recoil in disgust. Yet if someone has based their principles solely on personal advantage, you would have nothing whatever to object to in these measures. Or suppose someone recommends a man as a manager to whom you can blindly entrust all your affairs. To inspire your confidence, they praise him as a clever man with a masterful grasp of his own advantage, and as a tirelessly active man who lets no opportunity for it go unused. Finally, to make sure A 63 that no worries about his vulgar selfishness get in the way, they praise his refined way of living. He seeks pleasure not in hoarding money or crude extravagance, but in expanding his knowledge, in well-chosen, instructive company, and even in helping the needy. But he has no scruples about the means, which after all derive their worth or worthlessness only from the end. Other people's money and property serve his purposes V:36 just as well as his own, whenever he knows he can use them without being discovered or stopped. You would either think the person recommending him was making a fool of you, or that they had lost their mind.—The boundaries between morality and self-love are so clear and sharp that even the most ordinary eye cannot fail to distinguish which of the two something belongs to. The following few remarks may seem unnecessary when the truth is so obvious. Still, they at least serve to give ordinary human reason's judgment somewhat greater clarity.

The principle of happiness can provide maxims, but never maxims suited to be laws of the will, even if we make everyone's happiness our object. For cognition of happiness rests entirely on data from experience. Every judgment about it depends very heavily on each person's opinion, which is itself also very changeable. So it can provide general rules, but never universal ones: rules that on average apply most often, but not rules that must hold always and necessarily. Thus no practical laws can be based on it. Precisely because an object of choice is made the basis of its rule here, and must therefore precede it, the rule can relate to and be grounded on nothing A 64 but what is recommended, and thus on experience. Differences of judgment must then be endless. So this principle does not prescribe the very same practical rules to all rational beings, even though the rules fall under a common heading, namely happiness. The moral law, however, is thought of as objectively necessary only because it is to hold for everyone who has reason and will.

The maxim of self-love, or prudence, merely advises; the law of morality commands. There is a great difference between what we are advised to do and what we are obligated to do.

Even the most ordinary understanding can see quite easily and without hesitation what to do according to the principle of autonomy of choice. What to do on the assumption of heteronomy is difficult and requires knowledge of the world. That is, what duty requires presents itself to everyone on its own. But what brings genuine, lasting advantage is always shrouded in impenetrable darkness if this advantage is to extend over one's whole existence. It takes great prudence to use skillful exceptions to adapt a practical rule directed at such advantage to life's purposes even tolerably well. Yet the moral law commands everyone to obey, and to obey with the greatest exactness. So judging what to do according to it cannot be so difficult that even the most ordinary, untrained understanding would not know how to manage it without worldly wisdom.

Fulfilling morality's categorical command is within everyone's power V:37 at all times. Fulfilling the empirically conditioned instruction of happiness is only rarely A 65 possible, and is far from possible for everyone even with regard to a single aim. The reason is that the first concerns only the maxim, which must be genuine and pure. The second also concerns the powers and physical ability to make a desired object real. A command that everyone should seek to make themselves happy would be foolish. For we never command anyone to do what they already inevitably want on their own. We would have to command only the measures to take, or rather supply them, since people cannot do everything they want. But commanding morality under the name of duty is entirely reasonable. For, first, not everyone is willing to obey its instruction when it conflicts with inclinations. And as for the measures by which someone can follow this law, these need not be taught here. For what they will in this respect, they can also do.

Someone who has lost at gambling may be annoyed with themselves and their imprudence. But someone conscious of having cheated at gambling, even if they won by doing so, must despise themselves as soon as they measure themselves against the moral law. So this law must surely be something other than the principle of personal happiness. For having to tell oneself, “I am worthless, even though I have filled my purse,” must involve a different standard of judgment from applauding oneself and saying, “I am a clever person, because I have increased my funds.”

Finally, in the idea of our practical reason, something else accompanies the violation of a moral law: its deserving punishment. Now the concept of punishment as such cannot be connected with A 66 sharing in happiness. For the person imposing punishment may also have the kind intention of directing it toward that end. But it must first be justified in itself as punishment, that is, as a mere ill. The person punished must admit that justice has been done to them and that their lot completely fits their conduct, even if matters stopped there and they could look forward to no favor hidden behind this harshness. Every punishment as such must first involve justice, and justice makes up the essence of this concept. Kindness may also be combined with it, but the person deserving punishment has, by their conduct, not the slightest ground for counting on this. So punishment is a physical ill that, even if it were not connected with moral evil as a natural consequence, would still have to be connected with it as a consequence according to principles of moral lawgiving. Now every crime is punishable in itself, even without regard to the physical consequences for the offender. That is, the offender forfeits happiness, at least in part. So it would clearly be absurd to say that their crime consisted precisely in bringing punishment upon themselves by damaging their own V:38 happiness. Yet according to the principle of self-love, this would have to be the proper concept of every crime. On that view, punishment would be the reason for calling something a crime. Justice would instead have to consist in withholding all punishment and even preventing natural punishment. For then there would no longer be anything evil in the action, since the ills that would otherwise follow it, and for whose sake alone it was called evil, would now have been prevented. To regard all punishment and reward merely as A 67 machinery in the hands of a higher power, serving only to set rational beings to work toward their final end, happiness, is even more clearly a mechanism of their will that abolishes all freedom. It is too obviously so for us to need to dwell on it.

More subtle, though just as false, is the claim of those who assume a special moral sense that determines the moral law instead of reason. On their view, consciousness of virtue would be directly connected with satisfaction and pleasure, and consciousness of vice with mental unrest and pain. They thus make everything depend, after all, on the desire for personal happiness. Without bringing in what was said above, I want only to point out the illusion involved. To represent a vicious person as tormented by mental unrest through consciousness of their offenses, they must already represent that person in advance as morally good to at least some degree in the basic foundation of their character. Similarly, they must already represent someone pleased by the consciousness of dutiful actions as virtuous. So the concept of morality and duty must have preceded all consideration of this satisfaction and cannot be derived from it at all. We must already value the importance of what we call duty, the authority of the moral law, and the immediate worth that following it gives a person in their own eyes. Only then can we feel that satisfaction in the consciousness of conforming to duty, and that bitter self-reproach when we can charge ourselves with violating it. So we cannot feel this satisfaction A 68 or mental unrest before cognizing obligation, and make it the basis of obligation. We must already be at least halfway to being an honest person even to form a representation of these feelings. I do not deny that, just as freedom makes the human will directly determinable by the moral law, frequent action in accordance with this determining ground can eventually produce a subjective feeling of satisfaction with oneself. On the contrary, it is itself a duty to establish and cultivate this feeling, which alone properly deserves to be called moral feeling. But the concept of duty cannot be derived from it. Otherwise we would have to think of a feeling of a law as such, and make something an object of sensation V:39 that can be thought only through reason. Unless this is to be a flat contradiction, it would completely abolish every concept of duty, replacing it with a mere mechanical play of finer inclinations, which sometimes clash with the coarser ones.

Now let us compare our highest formal principle of pure practical reason, as autonomy of the will, with all previous material principles of morality. We can set out all the others in a table, in a way that actually exhausts all other possible cases except the single formal one. This will visibly prove that it is useless to look for a principle other than the one just presented.—All possible determining grounds of the will are either merely subjective, and therefore empirical, or objective and rational. Both kinds, in turn, are either external or internal.

Material practical determining groundsV:40 in the principle of morality are A 69Subjectiveexternal• Education (according to Montaigne)• The civil constitution (according to Mandeville)internal• Physical feeling (according to Epicurus)• Moral feeling (according to Hutcheson)Objectiveinternal• Perfection (according to Wolff and the Stoics)external• The will of God (according to Crusius and other theological moralists)

V:41 Those on the left are all empirical and clearly A 70 entirely unsuited to be the universal principle of morality. Those on the right, however, rest on reason. For perfection as a quality of things, and supreme perfection represented in a substance, that is, God, can both be thought only through concepts of reason. But the first concept, that of perfection, can be taken in a theoretical sense. Then it means only the completeness of each thing within its kind—transcendental perfection—or of a thing merely as a thing in general—metaphysical perfection. Neither can be at issue here. In its practical sense, however, the concept of perfection means a thing's fitness or adequacy for all sorts of ends. This perfection as a quality of a human being, and thus internal, is nothing but talent and what strengthens or supplements it, skill. Supreme perfection in a substance, that is, God, and thus external perfection considered practically, is this being's adequacy for all ends in general. Ends must therefore be given to us beforehand. Only in relation to them can the concept of perfection, whether internal in ourselves or external in God, become a determining ground of the will. But an end, as an object that must precede the determination of the will by a practical rule and contain the ground of that rule's possibility, is the will's matter. Taken as the will's determining ground, it is always empirical. So it can serve the Epicurean principle of the doctrine of happiness, but never the pure rational principle of the doctrine of morality and duty. Talents and their development can become motives only because they contribute to life's advantages. Similarly, the A 71 will of God can become a motive only through the happiness we expect from it, if we make agreement with it the object of our will without a prior practical principle independent of the idea of God. It follows, first, that all the principles listed here are material; second, that they cover all possible material principles. Finally, we can conclude that material principles are entirely unsuited to be the highest moral law, as has been proved. So the formal practical principle of pure reason is the only possible principle suited to categorical imperatives, that is, practical laws that make actions duties. According to this principle, the mere form of a universal lawgiving possible through our maxims must constitute the highest and immediate determining ground of the will. It alone is also suited to be the principle of morality in general, both for judging the human will and for applying it to determine that will.

I. A 72 V:42 The Deduction of the Principles of Pure Practical Reason.

This Analytic shows that pure reason can be practical: it can determine the will on its own, independently of everything empirical. It does so through a fact in which pure reason actually proves itself practical in us: autonomy in the principle of morality, through which reason determines the will to act.—At the same time, it shows that this fact is inseparably connected with consciousness of freedom of the will, indeed identical with it. A rational being, as belonging to the sensible world, recognizes itself as necessarily subject to the laws of causality, like other effective causes. Through this consciousness of freedom, however, its will is at the same time, in practical matters, conscious of itself on another side, namely as a being in itself, and of its existence as determinable in an intelligible order of things. This is not in accordance with a special intuition of itself, but with certain dynamic laws that can determine its causality in the sensible world. For it has been sufficiently proved elsewhere that freedom, when attributed to us, places us in an intelligible order of things.

If we now compare this with the analytic part of the Critique of Pure A 73 Speculative Reason, a striking contrast appears. There, the first given element making cognition a priori possible was not principles but pure sensible intuition, space and time. It made such cognition possible only for objects of the senses.—Synthetic principles from mere concepts without intuition were impossible. They could occur only in relation to that intuition, which was sensible, and thus only in relation to objects of possible experience. For concepts of the understanding, joined with this intuition, alone make possible the cognition we call experience.—Beyond objects of experience, and thus concerning things as noumena, speculative reason was rightly denied all positive cognition.—Yet it did secure the concept of noumena: the possibility, and indeed necessity, of thinking such things. For example, it defended against every V:43 objection the assumption of freedom, understood negatively, as entirely compatible with those principles and restrictions of pure theoretical reason. But it did not give us any determinate or expansive cognition of such objects. Instead, it completely cut off every prospect of it.

By contrast, although the moral law gives no prospect, A 74 it does provide a fact that is completely inexplicable from all the data of the sensible world and the entire scope of our theoretical use of reason. This fact points to a pure world of the understanding, indeed positively determines it, and lets us cognize something about it: a law.

This law is to give the sensible world, as sensible nature, the form of a world of the understanding, that is, of a supersensible nature, as regards rational beings, yet without impairing its mechanism. Now nature in the most general sense is the existence of things under laws. The sensible nature of rational beings in general is their existence under empirically conditioned laws, and thus heteronomy for reason. By contrast, the supersensible nature of these same beings is their existence according to laws independent of every empirical condition, and thus belonging to the autonomy of pure reason. The laws according to which the existence of things depends on cognition are practical laws. So supersensible nature, insofar as we can form a concept of it, is nothing other than a nature under the autonomy of pure practical reason. The law of this autonomy is the moral law. It is therefore the fundamental law of a supersensible nature and a pure world of the understanding, whose counterpart is to exist in the sensible world, yet A 75 without impairing that world's laws. We could call the former, which we cognize only in reason, archetypal nature (nature as a model). We could call the latter copied nature (nature as a copy), since it contains the possible effect of the idea of the former as a determining ground of the will. For the moral law does in fact place us, in idea, within a nature in which pure reason, if accompanied by physical power adequate to it, would produce the highest good. It determines our will to give the sensible world this form, as a whole of rational beings.

Even the most ordinary attention to ourselves confirms that this idea really serves as a model, like a preliminary sketch, for our determinations of the will.

V:44 When practical reason tests the maxim by which I intend to give testimony, I always consider what it would be like if it held as a universal law of nature. Clearly, in this form it would necessitate truthfulness in everyone. For treating statements as evidence while also allowing them to be deliberately untrue cannot coexist with the universality of a law of nature. Likewise, the maxim I adopt about freely disposing of my A 76 life is determined at once when I ask what it would have to be like for a nature to maintain itself according to a law based on it. Clearly, in such a nature no one could end their life at will, because such an arrangement would not be an enduring natural order. The same holds in all other cases. But in actual nature, as an object of experience, the free will is not automatically determined to maxims that could by themselves establish a nature under universal laws, or that would automatically fit into a nature arranged according to them. Instead, private inclinations make up a natural whole under laws of feeling and desire—physical laws—but not a nature possible solely through our will according to pure practical laws. Yet through reason we are conscious of a law to which all our maxims are subject, as if a natural order also had to arise through our will. So this must be the idea of a nature not empirically given, yet possible through freedom, and therefore supersensible. We give this nature objective reality, at least in a practical respect, because we regard it as the object of our will as pure rational beings.

The difference, then, between the laws of a nature to which the A 77 will is subject and those of a nature subject to a will, insofar as that will relates to its free actions, rests on the following. In the former, objects must be causes of the representations that determine the will. In the latter, the will is to be the cause of the objects. Its causality therefore has its determining ground solely in the pure faculty of reason, which can for this reason also be called pure practical reason.

So these two problems are very different: how pure reason can on the one hand cognize objects a priori, and how it can on the other hand directly be a determining ground V:45 of the will, that is, of a rational being's causality with regard to the reality of objects, merely through the thought of the universal validity of its own maxims as law.

The first belongs to the critique of pure speculative reason. It requires a prior explanation of how intuitions are possible a priori. Without intuitions no object at all can be given to us, and therefore none can be synthetically cognized. Its solution is that all these intuitions are only sensible. So they cannot make possible any speculative cognition extending beyond possible experience. Thus all principles of pure speculative reason accomplish nothing A 78 beyond making experience possible, either of given objects or of objects that may be given without end but are never given completely.

The second belongs to the critique of practical reason. It requires no explanation of how objects of the faculty of desire are possible. That task remains with the critique of speculative reason as a problem of theoretical cognition of nature. It asks only how reason can determine the maxim of the will: whether it does so only through empirical representations as determining grounds, or whether pure reason too would be practical and a law of a possible natural order that cannot be cognized empirically at all. The possibility of such a supersensible nature, whose concept could also be the ground of its reality through our free will, requires no a priori intuition of an intelligible world. Such an intuition, being supersensible, would in this case be impossible for us. For the issue is only the determining ground of willing in its maxims: whether it is empirical or a concept of pure reason concerning their lawfulness in general, and how it can be the latter. Whether the will's causality is sufficient to make its objects real is left to the theoretical principles of reason to judge, as an investigation A 79 of the possibility of the objects of willing. Intuition of these objects is therefore no part at all of the practical problem. What matters here is only the determination of the will and the determining ground of its maxim as a free will, not the outcome. For provided the will conforms to the law of pure reason, its ability to carry things out may be whatever V:46 it is. A nature may or may not actually arise from these maxims for giving laws to a possible nature. None of this concerns the critique that investigates whether and how pure reason can be practical, that is, directly determine the will.

In this task, then, the critique can blamelessly—and indeed must—begin with pure practical laws and their reality. But instead of intuition, it takes as their basis the concept of their existence in the intelligible world, namely freedom. For this concept means nothing else. Those laws are possible only in relation to freedom of the will, but are necessary if freedom is presupposed. Or conversely, freedom is necessary because those laws are necessary as practical postulates. How this consciousness of moral laws—or, what amounts to the same thing, of freedom—is possible cannot be explained further. But their admissibility can certainly A 80 be defended in the theoretical critique.

The exposition of the highest principle of practical reason is now complete. First, we have shown what it contains: that it stands on its own, entirely a priori and independently of empirical principles. Then we have shown how it differs from all other practical principles. With its deduction—justifying its objective and universal validity, and gaining insight into the possibility of such a synthetic proposition a priori—we cannot hope to succeed as well as we did with the principles of the pure theoretical understanding. For those referred to objects of possible experience, namely appearances. We could prove that these appearances can be cognized as objects of experience only by being brought under the categories in accordance with those laws. Consequently, all possible experience must conform to these laws. But I cannot follow that course in deducing the moral law. For it does not concern cognition of the character of objects that may somehow be given to reason from elsewhere. Instead, it concerns cognition insofar as it can become the ground of the existence of the objects themselves, and reason through it has causality in a rational being. A 81 That is, it concerns pure reason considered as a faculty that directly determines the will.

But all human insight comes to an end once we reach fundamental forces V:47 or fundamental faculties. Their possibility cannot be comprehended through anything else; yet neither may they be invented and assumed at will. So in the theoretical use of reason, only experience can justify our assuming them. But here, regarding the pure practical faculty of reason, even this substitute is unavailable: we cannot offer empirical proofs instead of a deduction from a priori sources of cognition. For something that needs experience to prove its reality must depend, in the grounds of its possibility, on principles of experience. But pure reason that is nevertheless practical cannot possibly be counted among such things, given its very concept. Moreover, the moral law is given, as it were, as a fact of pure reason. We are conscious of it a priori, and it is absolutely certain, even if experience could supply no example in which it was followed exactly. So the objective reality of the moral law cannot be proved by any deduction or by any effort of theoretical reason, whether speculative or empirically supported. Nor, even if A 82 we gave up absolute certainty, could it be confirmed through experience and so proved a posteriori. Yet it stands firm on its own.

But something different and quite paradoxical takes the place of this vainly sought deduction of the moral principle. Conversely, that principle itself serves as the principle for deducing an inscrutable faculty: freedom. No experience can prove this faculty. Yet speculative reason had to assume it at least as possible, in order to find among its cosmological ideas the unconditioned in respect of its causality, and thus avoid contradicting itself. The moral law, which itself needs no grounds to justify it, proves not merely freedom's possibility but its reality in beings who recognize this law as binding on them. The moral law is in fact a law of causality through freedom, and thus of the possibility of a supersensible nature. Likewise, the metaphysical law of events in the sensible world was a law of the causality of sensible nature. So the moral law determines what speculative philosophy had to leave undetermined: the law for a causality whose concept in speculative philosophy was merely negative. It thereby first gives this concept objective reality.

V:48 The moral law provides this kind of credential by itself being established as a A 83 principle for deducing freedom as a causality of pure reason. Since theoretical reason was compelled at least to assume the possibility of freedom, this credential is completely sufficient to meet its need, in place of any justification a priori. For the moral law proves its reality in a way that also satisfies the critique of speculative reason. It adds a positive determination to a causality that was thought of only negatively, whose possibility speculative reason could not comprehend and yet had to assume. This positive determination is the concept of a reason that directly determines the will through the condition of a universal lawful form of its maxims. In this way it can, for the first time, give objective, though only practical, reality to reason, which always went beyond its bounds with its ideas whenever it tried to proceed speculatively. It transforms reason's transcendent use into an immanent one: being, through ideas, itself an effective cause in the field of experience.

The determination of the causality of beings in the sensible world, as such, could never be unconditional. Yet for every series of conditions there must necessarily be something unconditioned, and thus also a causality that determines itself entirely on its own. So the A 84 idea of freedom as a faculty of absolute spontaneity was not a need but, as regards its possibility, an analytic principle of pure speculative reason. But it is absolutely impossible to give an example of it in any experience. Among the causes of things as appearances, no determination of causality can be found that is absolutely unconditioned. So we could only defend the thought of a freely acting cause, when applying it to a being in the sensible world that is also considered, on another side, as a noumenon. We did so by showing that there is no contradiction in regarding all its actions as physically conditioned insofar as they are appearances, while also regarding their causality as physically unconditioned insofar as the acting being is a being of the understanding. In this way we made the concept of freedom a regulative principle of reason. Through it I do not cognize at all what the object to which this causality is attributed is. But I remove the obstacle: on the one hand, in explaining events in the world, including actions of rational V:49 beings, I do justice to the mechanism of natural necessity, going back endlessly from the conditioned to its condition. On the other hand, I keep open for speculative reason the place that is empty A 85 for it, namely the intelligible, in which to locate the unconditioned. But I could not give reality to this thought: I could not transform it into cognition of a being acting in this way, even merely as regards its possibility. Pure practical reason now fills this empty place with a determinate law of causality in an intelligible world through freedom: the moral law. This adds nothing to speculative reason's insight, but it does add to the security of its problematic concept of freedom. This concept is here given objective reality that, although only practical, is nevertheless beyond doubt. Practical reason does not even expand the concept of causality so as to extend its use beyond those limits. Its application, and therefore its meaning, properly occurs only in relation to appearances, to connect them into experiences, as the Critique of Pure Reason proves. For if practical reason aimed to extend that use, it would have to show how the logical relation of ground and consequence could be used synthetically with a kind of intuition other than sensible intuition. That is, it would have to show how a noumenal cause is possible. This it cannot do at all. But as practical reason, it takes no account of that task either. It simply locates the determining ground of a human being's causality as a sensible being A 86—a causality that is given—in pure reason, which is therefore called practical. Thus it uses the concept of cause itself solely for a practical purpose: not to cognize objects, but to determine causality with regard to them in general. Here it can leave entirely aside the concept's application to objects for the purposes of theoretical cognition, because this concept is always found in the understanding a priori, independently of all intuition. It can therefore locate the will's determining ground in the intelligible order of things, while readily admitting that it does not understand at all what determination the concept of cause might have for cognizing these things. It must, of course, cognize in a determinate way causality with regard to the will's actions in the sensible world. Otherwise practical reason could not actually produce a deed. But it need not be able to determine theoretically the concept it forms of its own V:50 causality as a noumenon, for the purpose of cognizing its supersensible existence, and thus give it meaning in that respect. For it acquires meaning anyway, though only for practical use, through the moral law. Even considered theoretically, it always remains a pure concept of the understanding, given a priori, that can be applied A 87 to objects whether they are given sensibly or not. In the latter case, however, it has no determinate theoretical meaning or application. It is only a formal, yet essential, thought of the understanding concerning an object in general. The meaning reason gives it through the moral law is solely practical: the idea of the law of a causality (the causality of the will) itself has causality, or is its determining ground.

II. The Right of Pure Reason in Its Practical Use to an Extension That Is Not Possible for It on Its Own in Its Speculative Use.

In the moral principle we have established a law of causality that places causality's determining ground beyond all conditions of the sensible world. We have not merely thought of the will as determinable through belonging to an intelligible world, and thus of the subject of this will, the human being, as belonging to a pure world of the understanding, though unknown to us in that respect. That much was possible according to the Critique of Pure A 88 Speculative Reason. We have also determined this subject with regard to its causality by means of a law that cannot be counted among any laws of nature in the sensible world. Thus we have extended our cognition beyond that world's limits, a presumption that the Critique of Pure Reason declared invalid in all speculation. How, then, can the practical use of pure reason here be reconciled with the theoretical use of this same reason as regards defining the limits of its power?

David Hume, who can be said to have really begun all the attacks on the rights of pure reason that made a complete investigation of it necessary, reasoned as follows. The concept of V:51 cause contains the necessity of a connection between the existence of different things, precisely insofar as they are different. Thus, if A is posited, I cognize that something entirely different from it, B, must necessarily also exist. But necessity can be attributed to a connection only insofar as it is cognized a priori. For experience would let us cognize only that a connection exists, not that it necessarily exists in that way. Now, he says, it is impossible to cognize the connection between one thing and another, or between one determination and another entirely different from it, A 89 a priori and as necessary, if they are not given in perception. So the very concept of a cause is false and deceptive. To put it as mildly as possible, it is an illusion that may still be excused to this extent: the habit, a subjective necessity, of frequently perceiving certain things or their determinations as existing together or in succession is imperceptibly taken for an objective necessity to posit such a connection in the objects themselves. Thus the concept of a cause is acquired by stealth, not legitimately, and can never be legitimately acquired or authenticated. For it requires a connection that is in itself empty and imaginary, that cannot stand before any reason, and to which no object could ever correspond.—Thus, for the first time, empiricism was introduced as the sole source of principles concerning all cognition that deals with the existence of things. Mathematics was therefore still exempt. But with it came the most extreme skepticism, even concerning the whole science of nature as philosophy. For according to such principles we can never infer a consequence from given determinations of things in respect of their existence. That would require the concept of a cause, which contains the necessity of such a connection. We can only expect similar cases as before, according to the A 90 rule of the imagination. But this expectation is never certain, however often it may have been fulfilled. Indeed, of no event could we say that something must have preceded it on which it necessarily followed; that is, that it must have a cause. Even if we knew of very many cases where something of this sort came first, so that a rule could be drawn from them, we still could not assume that things always and necessarily happen in this way. We would therefore have to give blind chance its due as well, where all use of reason V:52 ends. This firmly grounds skepticism concerning inferences that ascend from effects to causes and makes it irrefutable.

Mathematics had escaped unscathed so far because Hume thought all its propositions were analytic: they progressed from one determination to another on grounds of identity, and thus according to the principle of contradiction. But this is false, since they are all synthetic. Geometry, for example, deals not with the existence of things but only with their a priori determination in a possible intuition. Yet it still passes, just as causal concepts do, from one determination A to an entirely different one B, as nevertheless A 91 necessarily connected with it. But eventually this science, so highly praised for its absolute certainty, must also succumb to empiricism in principles. The reason is the same one that led Hume to substitute habit for objective necessity in the concept of cause. Despite all its pride, mathematics must accept a lowering of its bold claims that command assent a priori. It must await approval of the universal validity of its propositions from the goodwill of observers. As witnesses, they would not refuse to admit that they too had always perceived matters as the geometer presents them in principles. So, although it was not actually necessary, they would permit us to expect the same in the future. In this way Hume's empiricism in principles inevitably leads to skepticism even about mathematics, and therefore about all scientific theoretical use of reason, since this belongs either to philosophy or to mathematics. Will ordinary reason fare better amid such a terrible overthrow as we see befalling the leaders of cognition? Or will it instead be caught even more irretrievably in the same destruction of all knowledge? Must a universal skepticism not follow from the same principles, though it would admittedly A 92 affect only the learned? I leave everyone to judge for themselves.

Now my treatment in the Critique of Pure Reason was prompted by Hume's doctrine of doubt, but went much further. It covered the entire field of pure theoretical reason in its synthetic use, and thus also what is generally called metaphysics. V:53 Regarding the Scottish philosopher's doubts about the concept of causality, I proceeded as follows. If Hume took objects of experience to be things in themselves, as indeed happens almost everywhere, he was entirely right to declare the concept of cause deceptive and a false illusion. For with things in themselves and their determinations as such, we cannot see why, because something A is posited, something else B must necessarily also be posited. So he could not admit any such cognition a priori of things in themselves. Still less could this perceptive thinker allow an empirical origin for this concept. That directly contradicts the necessity of connection that forms the essence of the concept of causality. So the concept was outlawed, and habit in observing the course of perceptions took its place.

But my investigations showed that the objects A 93 we deal with in experience are by no means things in themselves, but only appearances. With things in themselves we cannot see—and it is indeed impossible to see—how, if A is posited, it should be contradictory not to posit B, which is entirely different from A. This is the necessity of the connection between A as cause and B as effect. Yet we can quite well think that, as appearances in one experience, they must be necessarily connected in certain ways, for example in their temporal relations. They cannot be separated without contradicting the connection through which this experience is possible, in which they are objects and alone can be cognized by us. This did in fact turn out to be the case. So I could not only prove the objective reality of the concept of cause with regard to objects of experience, but also deduce it as an a priori concept because of the necessity of connection it involves. That is, I could show its possibility from pure understanding without empirical sources. Having thus removed empiricism concerning its origin, I could eliminate at the root its inevitable consequence, skepticism: first regarding natural science, then also regarding mathematics, since exactly the same grounds yield exactly the same result. Both sciences refer to A 94 objects of possible experience. In this way I could also eliminate the complete V:54 doubt concerning everything theoretical reason claims to understand.

But what about applying this category of causality to things that are not objects of possible experience, but lie beyond its limits? The same question applies to all the other categories, since without them no cognition of what exists can be achieved. For I could deduce the objective reality of these concepts only with regard to objects of possible experience. But precisely this—that I rescued them even in this case alone, and showed that through them objects can at least be thought, though not determined a priori—gives them a place in the pure understanding. From there they are referred to objects in general, sensible or non-sensible. If anything is still missing, it is the condition for applying these categories, especially causality, to objects: intuition. Where intuition is not given, applying them for theoretical cognition of an object as noumenon is impossible. Any attempt at this is therefore completely forbidden, as it was in the Critique of Pure Reason. Yet the concept's objective reality always remains. It can also A 95 be used of noumena, but without our being able to determine this concept theoretically in the slightest and thereby produce cognition. For it was proved that this concept contains nothing impossible, even in relation to an object, by securing its place in the pure understanding in all its application to objects of the senses. Afterward, when referred to things in themselves, which cannot be objects of experience, it may be incapable of being determined so as to represent a determinate object for theoretical cognition. Yet it could still be capable of being determined for application to some other purpose, perhaps a practical one. This would not be so if, as Hume held, the concept of causality contained something entirely impossible to think.

To find this condition for applying that concept to noumena, we need only look back at why we are not satisfied with applying it to objects of experience, but would also like to use it of things in themselves. It soon becomes clear that what makes this necessary for us is a practical, not a theoretical, purpose. V:55 For speculation, even if we succeeded, we would A 96 gain nothing genuine in cognition of nature or of any objects that might be given to us. At most we would make a great leap from what is sensibly conditioned to the supersensible, in order to complete and limit our cognition on the side of its grounds. Yet we already have enough to do staying with the sensibly conditioned and diligently following the chain of causes. An infinite gulf would still remain unfilled between that limit and what we know. We would have listened to an idle urge to question rather than to a serious desire for knowledge.

But besides the relation in which the understanding stands to objects in theoretical cognition, it also has a relation to the faculty of desire, which for this reason is called the will. It is called the pure will insofar as the pure understanding, which in such a case is called reason, is practical through the mere representation of a law. The objective reality of a pure will, or what is the same thing, of pure practical reason, is given a priori in the moral law, as if through a fact. For that is what we may call a determination of the will that is unavoidable, though not based on empirical principles. But the concept of a will already contains the concept of causality. Thus the concept of a pure will contains that of a causality with freedom: one that cannot A 97 be determined according to laws of nature, and so cannot have any empirical intuition as proof of its reality. Yet in the pure practical law it fully justifies its objective reality a priori, though, as is easy to see, not for the theoretical but only for the practical use of reason. Now the concept of a being with a free will is the concept of a noumenal cause. We are already assured that this concept does not contradict itself. For the concept of a cause originates entirely in the pure understanding. The deduction also secures its objective reality with regard to objects in general. In its origin it is independent of all sensible conditions, and thus in itself is not restricted to phenomena, except where a determinate theoretical use is intended. It can therefore certainly be applied to things as pure beings of the understanding. But this application cannot be grounded in any intuition, since intuition can always be only sensible. So, regarding the theoretical use of reason, V:56 a noumenal cause is a possible and thinkable, yet empty, concept. But I do not claim through it to know theoretically the character of a being insofar as it has a pure will. It is enough for me A 98 simply to designate it as such a being, and thus only to connect the concept of causality with that of freedom, and inseparably with the moral law as its determining ground. I certainly have the right to do this because of the pure, nonempirical origin of the concept of cause. For I consider myself entitled to no use of it except in relation to the moral law that determines its reality: only a practical use.

If, with Hume, I had denied the concept of causality objective reality in theoretical use not only regarding things in themselves, the supersensible, but also regarding objects of the senses, it would have lost all meaning. It would have been declared completely useless as a theoretically impossible concept. Since no use can be made of nothing, the practical use of a theoretically void concept would have been completely absurd. But as matters stand, the concept of an empirically unconditioned causality is theoretically empty, lacking an appropriate intuition, yet remains possible and refers to an undetermined object. Instead, it is given meaning through the moral law, and thus in a practical respect. So although I have no intuition that would determine its objective theoretical reality, it nevertheless has an actual A 99 application that can be presented concretely in dispositions or maxims. That is, it has a practical reality that can be exhibited, and this is sufficient to justify it even regarding noumena.

Once this objective reality of a pure concept of the understanding has been introduced into the field of the supersensible, it also gives objective reality to all the other categories, though always only insofar as they stand in a necessary connection with the determining ground of the pure will, the moral law. This reality is applicable only practically. It has not the slightest influence toward expanding theoretical cognition of these objects, as insight through pure reason into their nature. As we will also find later, these categories always refer only to beings as intelligences, and even in them only to the relation of reason to V:57 will, thus always only to the practical. They claim no cognition of these beings beyond that. Any other properties that might be brought in along with them, belonging to the theoretical way of representing such supersensible things, must all then be counted not as knowledge but only as things we are entitled—and for practical purposes even required—to assume and presuppose. This holds even when we assume supersensible A 100 beings, such as God, by analogy: that is, according to the pure relation of reason that we use practically regarding sensible beings. In this way, applying pure theoretical reason to the supersensible, but only for practical purposes, gives it not the slightest encouragement toward fanatical flights beyond its bounds.

Of the Analytic of Practical Reason

7 Propositions called practical in mathematics or natural science should properly be called technical. For these disciplines are not concerned with determining the will at all. They only indicate the variety of possible action sufficient to produce a certain effect. So they are just as theoretical as all propositions stating the connection of a cause with an effect. Anyone who wants the latter must also accept being the former.