Book II, Chapter II: The dialectic of pure reason in determining the highest good

Chapter II. A 198 On the Dialectic of Pure Reason in Determining the Concept of the Highest Good.

The concept of the highest already contains an ambiguity that, if overlooked, can give rise to needless disputes. The highest can mean either the supreme (supremum) or the complete (consummatum). The former is the condition that is itself unconditioned, that is, subordinate to no other (originarium); the latter is the whole that forms no part of a still greater whole of the same kind (perfectissimum). That virtue (as worthiness to be happy) is the supreme condition of everything that can seem desirable to us, and thus of all our pursuit of happiness as well, and is therefore the supreme good, has been proved in the Analytic. But this does not yet make it the whole and complete good, as the object of the faculty of desire of rational finite beings; for that, happiness is also required—not merely in the partial eyes of the person who makes himself his end A 199, but even in the judgment of an impartial reason that regards that person, in the world generally, as an end in himself. For to need happiness, to be worthy of it as well, and yet not to share in it cannot possibly be consistent with the perfect willing of a rational being who at the same time possessed all power, even if we think of such a being only as an experiment. Now insofar as virtue and happiness together constitute a person's possession of the highest good, and, when happiness is distributed in exact proportion to morality (as the person's worth and worthiness to be happy), constitute the highest good of a possible world, this means the V:111 whole, the complete good. Within it, however, virtue, as the condition, is always the supreme good, because it has no further condition above it; happiness is always something that, though agreeable to its possessor, is not by itself good absolutely and in every respect, but always presupposes morally lawful conduct as its condition.

Two determinations necessarily united in one concept must be connected as ground and consequence, either so that this unity is regarded as analytic (a logical connection) or as synthetic (a real combination): the former according to the law of identity, the latter according to that of A 200 causality. The connection of virtue with happiness can therefore be understood in either of two ways. The endeavor to be virtuous and the rational pursuit of happiness might be not two different actions but wholly identical ones, so that the former would need no other maxim as its basis than the latter. Or that connection is taken to mean that virtue produces happiness as something distinct from the consciousness of virtue, as a cause produces an effect.

Among the ancient Greek schools, only two properly followed the same method in determining the concept of the highest good to this extent: they did not allow virtue and happiness to count as two different elements of the highest good, and thus sought the unity of the principle according to the rule of identity. Yet they differed again in which of the two they chose as the basic concept. The Epicurean said: to be conscious of one's maxim leading to happiness is virtue; the Stoic: to be conscious of one's virtue is happiness. For the former, prudence amounted to morality; for the latter, who chose a loftier name for virtue, morality alone was true wisdom.

It is regrettable that these men's acuteness (though A 201 we must also admire them for having attempted, so early, every conceivable path of philosophical conquest) was unhappily employed in reasoning out an identity between utterly heterogeneous concepts, those of happiness and virtue. But it suited the dialectical spirit of their age—a spirit that even now sometimes misleads subtle minds—to abolish essential differences between principles that can never be reconciled by trying to turn them into verbal disputes, thus contriving a semblance of conceptual unity under merely V:112 different names. This generally happens where the union of heterogeneous grounds lies so deep or so high, or would require so complete a transformation of the doctrines otherwise accepted in the philosophical system, that people shrink from entering deeply into the real distinction and prefer to treat it as a disagreement over mere formalities.

Although both schools tried to reason out an identity between the practical principles of virtue and happiness, they did not agree on how to force this identity into being. Rather, they diverged infinitely from one another: one placed its principle on the aesthetic side, the other on the logical; the former in consciousness of sensible need, the latter in the independence A 202 of practical reason from all sensible determining grounds. According to the Epicurean, the concept of virtue already lay in the maxim of promoting one's own happiness; according to the Stoic, by contrast, the feeling of happiness was already contained in the consciousness of one's virtue. But what is contained in another concept is indeed identical with a part of what contains it, not with the whole; and two wholes can moreover differ specifically from one another, though they consist of the very same material, if the parts are combined into a whole in entirely different ways in each. The Stoic maintained that virtue was the entire highest good, and happiness only the consciousness of possessing it, as belonging to the subject's condition. The Epicurean maintained that happiness was the entire highest good, and virtue only the form of the maxim for pursuing it, namely the rational use of the means to it.

Now the Analytic makes it clear that the maxims of virtue and those of one's own happiness are entirely heterogeneous as regards their supreme practical principle. Far from being in agreement, although they belong to one highest good, they greatly restrict and impair one another within the same subject when it comes to making that good possible. Thus the question, How is the highest good practically A 203 possible? remains an unsolved problem despite all previous attempts at coalition. What makes it difficult to solve, however, is given in the Analytic: happiness and morality are two specifically quite different elements of the highest good, and their connection therefore cannot be cognized analytically V:113 (as though someone seeking his happiness would, simply by analyzing his concepts, find himself virtuous in this conduct, or someone pursuing virtue would find himself already happy by that very fact in his consciousness of such conduct). Rather, it is a synthesis of concepts. But because this connection is cognized as a priori, and thus as practically necessary, consequently not as derived from experience, and because the possibility of the highest good therefore rests on no empirical principles, the deduction of this concept will have to be transcendental. It is a priori (morally) necessary to bring about the highest good through freedom of the will; therefore the condition of its possibility must likewise rest solely on grounds of cognition a priori.

I. A 204 The antinomy of practical reason.

In the highest good that is practical for us, that is, to be made actual through our will, virtue and happiness are thought as necessarily connected, so that pure practical reason cannot accept the one without the other's also belonging to it. Now this connection, like any connection at all, is either analytic or synthetic. Since the connection in question cannot be analytic, as was just shown, it must be thought as synthetic, specifically as the connection of cause with effect, because it concerns a practical good, that is, something possible through action. Therefore either the desire for happiness must be the motive for maxims of virtue, or the maxim of virtue must be the efficient cause of happiness. The first is absolutely impossible, because, as was proved in the Analytic, maxims that place the determining ground of the will in the desire for one's happiness are not moral at all and cannot ground virtue. But the second is also impossible, because every practical connection of causes and effects in the world resulting from the determination of the will conforms not to A 205 the will's moral dispositions, but to knowledge of natural laws and the physical capacity to use them for one's purposes. Consequently, even the most exact V:114 observance of the moral laws cannot be expected to produce a necessary connection of happiness with virtue in the world sufficient for the highest good. Now, since promoting the highest good, whose concept contains this connection, is an a priori necessary object of our will and is inseparably bound up with the moral law, the impossibility of the former must also prove the falsity of the latter. If, then, the highest good is impossible according to practical rules, the moral law, which commands us to promote it, must also be fanciful and directed toward empty imaginary ends, and thus in itself false.

II. Critical resolution of the antinomy of practical reason.

In the antinomy of pure speculative reason, a similar conflict occurs between natural necessity and freedom in the causality of events in the world. It was resolved by proving that there is no true conflict if events A 206, and even the world in which they occur, are regarded—as indeed they should be—only as appearances. For one and the same acting being, as appearance (even before its own inner sense), has a causality in the sensible world that always conforms to the mechanism of nature; yet with regard to the same event, insofar as the acting person also regards himself as a noumenon (as pure intelligence, in his existence not determinable in time), he can contain a determining ground of that causality according to natural laws which is itself free from every natural law.

The same holds for the present antinomy of pure practical reason. The first of the two propositions, that the pursuit of happiness produces a ground for a virtuous disposition, is absolutely false. But the second, that a virtuous disposition necessarily produces happiness, is not absolutely false. It is false only insofar as that disposition is regarded as the form of causality in the sensible world, and hence if I assume existence in that world to be the only kind of existence a rational being has; thus it is false only conditionally. But since I am not merely entitled to think of my existence also as a noumenon in a world of understanding, but even V:115 possess in the moral law a purely intellectual determining ground of my causality (in the sensible world), it is not impossible A 207 that morality of disposition should have a connection as cause with happiness as effect in the sensible world—a connection that, if not immediate, is nevertheless mediated (through an intelligible author of nature), and indeed necessary. In a nature that is merely an object of the senses, this connection can never occur otherwise than contingently and cannot suffice for the highest good.

Thus, despite this apparent conflict of practical reason with itself, the highest good is the necessary highest end of a morally determined will, a true object of practical reason. For it is practically possible, and the maxims of that will, which relate to it as regards their matter, have objective reality. This reality was at first assailed by that antinomy in the connection of morality with happiness according to a universal law, but only through a misunderstanding: the relation between appearances was taken for a relation of things in themselves to these appearances.

If we find ourselves compelled to seek the possibility of the highest good—this goal of all moral wishes that reason sets before all rational beings—so far afield, namely in connection with an intelligible world, it must seem strange A 208 that philosophers of both ancient and modern times have nevertheless been able to find happiness in an entirely fitting proportion to virtue already in this life (in the sensible world), or to persuade themselves that they were conscious of it. For both Epicurus and the Stoics exalted above everything the happiness that arises in life from consciousness of virtue. And the former was not so base in his practical precepts as one might infer from the principles of his theory, which he used for explanation, not for action, or as many interpreted them, misled by the expression “sensual pleasure” for contentment. Rather, he included the most unselfish practice of goodness among the ways of enjoying the deepest joy; and contentment with little and restraint of the inclinations, such as the strictest moral philosopher might ever demand, also belonged to his plan of pleasure (by which he meant the ever-cheerful heart). In this he differed from the Stoics chiefly only by making this pleasure the motive, which they refused to do, and rightly. For on the one hand V:116 the virtuous Epicurus, like many morally well-disposed men even now who do not reflect deeply enough on their principles, fell into the error of already presupposing a virtuous disposition in the persons for whom he wished first to specify the incentive to virtue. (And in fact an upright person cannot find himself happy A 209 unless he is first conscious of his uprightness: given such a disposition, the reproaches his own way of thinking would compel him to make against himself for transgressions, and his moral self-condemnation, would rob him of all enjoyment of the pleasant things his condition might otherwise contain.) But the question is: what first makes possible such a disposition and way of thinking, which values one's existence in this manner, since before it the subject would have no feeling at all for moral worth in general? Certainly, if a person is virtuous, he will not enjoy life without being conscious of his uprightness in every action, however favorable fortune may be to the physical circumstances of his life. But to make him virtuous in the first place, and thus before he rates the moral worth of his existence so highly, can one really commend to him the peace of soul that will arise from consciousness of an uprightness for which he has as yet no feeling?

On the other hand, however, there always lies here a ground for an error of illicit substitution (vitium subreptionis) and, as it were, an optical illusion in our self-consciousness of what we do, as distinct from what we feel, which even the most experienced person cannot entirely avoid A 210. A moral disposition is necessarily connected with consciousness that the will is determined immediately by the law. Now consciousness of a determination of the faculty of desire is always the ground of satisfaction in the action thereby produced. But this pleasure, this satisfaction with oneself, is not the determining ground of the action; rather, the immediate determination of the will by reason alone is the ground of the feeling of pleasure, and that determination remains a purely practical, not an aesthetic, determination of the faculty of desire. Since this determination produces inwardly precisely the same effect of an impulse to activity as would a feeling of agreeableness expected from the desired action, we easily mistake what we ourselves do for something we merely passively feel, and take the moral V:117 incentive for a sensible impulse, as always happens in so-called illusion of the senses (here, of inner sense). There is something very sublime in human nature's being determined to act immediately by a pure law of reason, and even in the illusion of taking the subjective aspect of this intellectual determinability of the will for something aesthetic and the effect of a special sensible feeling (for an intellectual feeling would be a contradiction). It is also of great importance to call attention to this property of our personality A 211 and to cultivate reason's effect on this feeling as fully as possible. But we must also beware of degrading and disfiguring the true, genuine incentive, the law itself, as though by a false foil, through spurious praise of this moral determining ground as an incentive, when we make feelings of special joys its grounds (though they are only consequences). Respect, then, and not pleasure or the enjoyment of happiness, is something for which no antecedent feeling laid at reason's foundation is possible (since such a feeling would always be aesthetic and pathological). As consciousness of the will's immediate necessitation by law, it is scarcely an analogue of the feeling of pleasure, although in relation to the faculty of desire it does precisely the same thing, but from other sources. Yet only by this way of representing the matter can we achieve what we seek: that actions be done not merely in conformity with duty (as a result of agreeable feelings), but from duty, which must be the true aim of all moral cultivation.

But is there not a word that designates, not an enjoyment such as happiness, but nevertheless a satisfaction with one's existence, an analogue of happiness that must necessarily accompany consciousness of virtue? Yes! The word is self-contentment A 212, which in its proper meaning always signifies only a negative satisfaction with one's existence, in which one is conscious of needing nothing. Freedom, and consciousness of it as a capacity to follow the moral law with an overriding disposition, is independence from inclinations, at least as determining motives of our desire (even if not as motives that affect it). Insofar as I am conscious of this freedom in following my moral maxims, it is the sole source of an unchanging V:118 contentment necessarily connected with it, resting on no special feeling; and this contentment may be called intellectual. Aesthetic contentment (which is improperly so called), resting on the satisfaction of inclinations, however subtly they may be devised, can never be adequate to what we think of it. For inclinations change, grow with the indulgence shown them, and always leave behind a still greater emptiness than we had intended to fill. Hence they are always burdensome to a rational being, and even if he cannot cast them off, they nevertheless force from him the wish to be rid of them. Even an inclination toward what conforms to duty (for example, beneficence) can greatly facilitate the effectiveness of moral maxims, but can produce none. A 213 For in moral action everything must be directed toward the representation of the law as determining ground, if the action is to contain not merely legality, but also morality. Inclination is blind and servile, whether good-natured or not; and where morality is at issue, reason must not merely play its guardian, but, without regard to it, must as pure practical reason attend entirely to its own interest alone. Even this feeling of compassion and tenderhearted sympathy, if it precedes reflection on what duty is and becomes a determining ground, is itself burdensome to right-thinking people, confuses their considered maxims, and produces the wish to be rid of it and subject only to legislative reason.

From this we can understand how consciousness of this capacity of a pure practical reason, through its exercise in deed (virtue), can produce consciousness of mastery over one's inclinations, and thus of independence from them and consequently also from the discontent that always accompanies them; hence a negative satisfaction with one's condition, that is, contentment, which at its source is contentment with one's person. In this way freedom itself becomes capable (namely indirectly) of an enjoyment that cannot be called happiness, because it A 214 does not depend on the positive addition of a feeling, nor, strictly speaking, bliss, because it does not contain complete independence from inclinations and needs. Yet it resembles the latter, insofar as at least the determination of one's will can remain free of their influence, and is thus, at least in its origin, analogous to the self-sufficiency that can be attributed only to the highest being.

V:119 From this resolution of the antinomy of pure practical reason, it follows that in practical principles a natural and necessary connection between consciousness of morality and the expectation of happiness proportionate to it, as its consequence, can at least be thought as possible (though certainly not yet cognized and understood on that account). By contrast, principles for pursuing happiness cannot possibly produce morality. Thus morality constitutes the supreme good (as the first condition of the highest good), while happiness indeed constitutes its second element, but in such a way that it is only the morally conditioned, yet necessary, consequence of the former. Only in this subordination is the highest good the entire object of pure practical reason, which must necessarily represent it as possible, because it commands us to contribute everything possible to bringing it about. But the possibility of such a connection A 215 of the conditioned with its condition belongs wholly to the supersensible relation of things and cannot be given at all according to laws of the sensible world, although the practical consequence of this idea, namely the actions aimed at making the highest good actual, belongs to the sensible world. We shall therefore try to set out the grounds of that possibility, first with regard to what lies immediately within our power, and then, second, in what reason offers us as a supplement to our incapacity for making the highest good possible (a supplement necessary according to practical principles), and which is not within our power.

III. On the primacy of pure practical reason in its union with speculative reason.

By primacy among two or more things connected through reason, I understand the precedence of one in being the first determining ground of the connection with all the others. In a narrower, practical sense, it means the precedence of the interest of one, insofar as the interest of the others is subordinated to it (an interest that cannot itself be ranked below any other). To every faculty of the mind we can A 216 attribute an interest, that is, a principle containing the condition under which alone its exercise is promoted. Reason, as the faculty of principles, determines the interest of all V:120 powers of the mind, but determines its own interest for itself. The interest of its speculative use consists in the cognition of the object up to the highest principles a priori; that of its practical use, in the determination of the will with regard to the final and complete end. What is required for the possibility of any use of reason at all, namely that its principles and assertions must not contradict one another, forms no part of its interest, but is the condition of having reason at all. Only its extension, not its mere agreement with itself, counts as its interest.

If practical reason may accept and think as given nothing beyond what speculative reason could independently offer it from its own insight, then the latter holds primacy. But suppose practical reason had original principles a priori of its own, with which certain theoretical positions were inseparably connected, although these eluded every possible insight of speculative reason (though they need not contradict it either). The question then is which interest is supreme A 217 (not which must yield, for the one does not necessarily conflict with the other): whether speculative reason, which knows nothing of all that practical reason offers it for acceptance, must take up these propositions and, though they transcend its powers, seek to unite them with its concepts as a foreign possession transferred to it; or whether it is entitled to follow its own separate interest obstinately and, according to the canon of Epicurus, reject as empty rationalizing everything that cannot authenticate its objective reality through evident examples available in experience—even if it were most intimately interwoven with the interest of the practical (pure) use of reason and did not in itself contradict theoretical reason—merely because it actually impairs the interest of speculative reason to this extent: it removes the boundaries speculative reason has set for itself and exposes it to every absurdity or delusion of the imagination.

Indeed, if practical reason were taken as pathologically conditioned, that is, as merely administering the interest of the inclinations under the sensible principle of happiness, no such demand could be made of speculative reason at all. Muhammad's paradise, or the theosophists' and mystics' melting union with the deity, as each one's fancy takes him, would thrust their V:121 monstrosities upon reason, and it would be just as well to have no reason at all as A 218 to abandon it in this way to every fantasy. But if pure reason can be practical of itself, and actually is so, as consciousness of the moral law demonstrates, then it is always only one and the same reason that judges according to principles a priori, whether for a theoretical or a practical purpose. And then it is clear that, even if its capacity in the former does not suffice to establish certain propositions affirmatively, provided they do not contradict it either, it must accept these very propositions as soon as they belong inseparably to the practical interest of pure reason. It must accept them, admittedly, as a foreign offering that has not grown on its own soil, but is nevertheless sufficiently authenticated, and must seek to compare and connect them with everything in its power as speculative reason. It must, however, acknowledge that these are not its insights, but nevertheless extensions of its use for some other purpose, namely a practical one; and this is not in the least contrary to its interest, which consists in restraining speculative presumption.

Thus, in the union of pure speculative with pure practical reason in one cognition, the latter holds primacy, provided, that is, that this union is not contingent and arbitrary, but grounded a priori in reason itself, and hence necessary A 219. For without this subordination a conflict of reason with itself would arise: if they were merely placed side by side (coordinated), the former would strictly close its boundary on its own account and admit nothing from the latter into its domain, while the latter would nevertheless extend its boundaries over everything and, where its need required, would seek to include the former within them as well. But pure practical reason cannot possibly be asked to subordinate itself to speculative reason and thus reverse the order, because all interest is ultimately practical, and even that of speculative reason is only conditional and is complete solely in practical use.

IV. V:122 The immortality of the soul as a postulate of pure practical reason.

Bringing about the highest good in the world is the necessary object of a will determinable by the moral law. Within this good, however, the complete conformity of dispositions to the moral law is the supreme condition of the highest good. It must therefore be possible just as its object is, since it is contained in the same command to A 220 promote that object. But the complete conformity of the will to the moral law is holiness, a perfection of which no rational being in the sensible world is capable at any moment of its existence. Yet since it is nevertheless required as practically necessary, it can be found only in a progress toward that complete conformity extending to infinity; and according to principles of pure practical reason, it is necessary to assume such practical advancement as the real object of our will.

This infinite progress, however, is possible only on the presupposition of an existence and personality of the same rational being enduring to infinity (which is called the immortality of the soul). Thus the highest good is practically possible only on the presupposition of the soul's immortality; consequently this immortality, as inseparably connected with the moral law, is a postulate of pure practical reason (by which I understand a theoretical proposition, though not demonstrable as such, insofar as it is inseparably attached to a practical law that holds unconditionally a priori).

The proposition concerning our nature's moral vocation—that we can attain complete conformity with the moral law only in a progress extending to infinity—is of the greatest use A 221, not merely with regard to our present supplementation of speculative reason's incapacity, but also with regard to religion. Without it, either the moral law is wholly degraded from its holiness, by being artificially represented as indulgent and thus suited to our convenience; or else we stretch our vocation, and with it our expectation, toward an unattainable destiny, namely a hoped-for complete acquisition of holiness of will V:123, and lose ourselves in extravagant theosophical dreams that utterly contradict self-cognition. Both merely hinder the unceasing striving to observe, exactly and throughout, a strict and unyielding command of reason that is nevertheless not imaginary, but real. For a rational but finite being, only progress to infinity from lower to higher stages of moral perfection is possible. The Infinite One, for whom the condition of time is nothing, sees in this series, endless for us, the whole of conformity with the moral law; and the holiness that his command unremittingly requires, in order to accord with his justice in the share he assigns each in the highest good, is to be found entire in a single intellectual intuition of the existence of rational beings. A 222 All that can belong to the creature with regard to hope of this share would be consciousness of his tested disposition: on the basis of his previous progress from the worse to the morally better, and of the unalterable resolve made known to him through it, he may hope for a further uninterrupted continuation of that progress, however far his existence may extend, even beyond this life¹². And so, never here or at any foreseeable future A 223 point in his existence, but only in the infinity (surveyable by God alone) V:124 of his continued duration, may he hope to be fully adequate to God's will (without indulgence or remission, which cannot be reconciled with justice).

V. The existence of God as a postulate of pure practical reason.

In the preceding analysis, the moral law led to the practical task prescribed by pure reason alone, without any contribution from sensible incentives: the necessary completion of the first and foremost part of the highest good, morality. And since this task can be fully accomplished only in an eternity, it led to the postulate of immortality. This same law must also lead to the possibility of the second element of the highest good, namely happiness proportionate to that morality, just as unselfishly as A 224 before, from mere impartial reason. It must, that is, lead to the presupposition of the existence of a cause adequate to this effect, or postulate the existence of God as necessarily belonging to the possibility of the highest good (an object of our will necessarily connected with the moral legislation of pure reason). We shall set forth this connection convincingly.

Happiness is the condition of a rational being in the world for whom, in the whole of his existence, everything goes according to his wish and will. It therefore rests on the harmony of nature with his entire end, and likewise with the essential determining ground of his will. Now the moral law, as a law of freedom, commands through determining grounds that are to be wholly independent of nature and of its harmony with our faculty of desire (as incentives). But the acting rational being in the world is not at the same time the cause of the world and of nature itself. Thus the moral law contains not the slightest ground for a necessary connection between morality and proportionate happiness in a being that belongs to the world as a part and is therefore dependent on it—a being that, for this very reason, cannot through its will be the cause of this nature, nor by its own powers make nature, as regards its happiness, agree throughout A 225 V:125 with its practical principles. Nevertheless, in the practical task of pure reason, that is, the necessary work toward the highest good, such a connection is postulated as necessary: we ought to seek to promote the highest good (which must therefore, after all, be possible). Thus the existence of a cause of all nature, distinct from nature, which contains the ground of this connection, namely the exact harmony of happiness with morality, is also postulated. But this supreme cause must contain the ground of nature's harmony not merely with a law of the will of rational beings, but with the representation of this law, insofar as these beings make it the supreme determining ground of the will: thus not merely with their conduct in its form, but also with its morality as its motive, that is, with their moral disposition. Hence the highest good in the world is possible only insofar as a supreme cause of nature is assumed that has a causality in keeping with moral disposition. Now a being capable of actions according to the representation of laws is an intelligence (a rational being), and such a being's causality according to this representation of laws is its will. Therefore the supreme cause of nature, insofar as it must be presupposed for the highest good, is a being that A 226 is the cause (and consequently the author) of nature through understanding and will, that is, God. Consequently, the postulate of the possibility of the highest derived good (the best world) is at the same time the postulate of the actuality of a highest original good, namely the existence of God. Now it was our duty to promote the highest good; hence there was not merely an entitlement, but also a necessity, connected with duty as a need, to presuppose the possibility of this highest good. Since this possibility obtains only on the condition of God's existence, it inseparably connects the presupposition of that existence with duty; that is, it is morally necessary to assume the existence of God.

Here it must be carefully noted that this moral necessity is subjective, that is, a need, and not objective, that is, itself a duty; for there can be no duty at all to assume the existence of a thing (since this concerns only the theoretical use of reason). Nor is it meant that assuming the existence of God as a ground of all obligation in general is necessary (for this ground rests, as has been sufficiently proved, solely on the V:126 autonomy of reason itself). What belongs to duty here is only the work of bringing about and promoting the highest good in the world, whose possibility can therefore be postulated, but which our reason A 227 finds conceivable only on the presupposition of a supreme intelligence. Thus assuming its existence is connected with consciousness of our duty, although this assumption itself belongs to theoretical reason. With regard to theoretical reason alone, considered as a ground of explanation, it can be called a hypothesis; but in relation to making intelligible an object nevertheless set before us by the moral law (the highest good), and hence to a need for a practical purpose, it can be called faith, and indeed pure rational faith, because pure reason alone (in both its theoretical and its practical use) is the source from which it springs.

From this deduction we can now understand why the Greek schools could never solve their problem of the practical possibility of the highest good: they always made the rule governing the human will's use of its freedom the sole and by itself sufficient ground of that possibility, without, in their opinion, needing God's existence for it. They were right, certainly, to establish the principle of morals independently of this postulate, on its own, from reason's relation to the will alone, and thus to make it the supreme practical condition of the highest good. But it was not on that account the entire condition of its possibility. The Epicureans had indeed A 228 adopted an entirely false principle of morals as supreme, namely that of happiness, and had substituted for a law a maxim of arbitrary choice according to each person's inclination. Yet they proceeded consistently enough in likewise degrading their highest good in proportion to the baseness of their principle, expecting no greater happiness than can be acquired through human prudence (which also includes abstinence and moderation of inclinations). This, as we know, must turn out meager enough, and very differently according to circumstances—not even counting the exceptions their maxims continually had to allow, which make them unfit to be laws. The Stoics, by contrast, had quite rightly chosen their supreme practical principle, namely virtue, as the condition of the highest good. But by representing the degree of virtue required by its pure V:127 law as fully attainable in this life, they not only exalted the moral capacity of the human being, under the name of a sage, beyond all limits of his nature, and assumed something contradicting all knowledge of humanity; above all, they also refused to recognize the second component belonging to the highest good, namely happiness, as a distinct object of the human faculty of desire. Instead, they made their A 229 sage, like a deity, wholly independent of nature (as regards his contentment) through consciousness of his personal excellence. They exposed him to the ills of life, but did not subject him to them (at the same time representing him as free from evil). Thus they actually left out the second element of the highest good, one's own happiness, by placing it merely in action and in contentment with one's personal worth, and so including it within consciousness of a moral way of thinking. Yet the voice of their own nature could have refuted them sufficiently on this point.

The doctrine of Christianity¹³, even when not yet considered as a religious doctrine, supplies on this point a concept of the highest A 230 V:128 good (the kingdom of God) that alone satisfies the strictest demand of practical A 231 reason. The moral law is holy (unyielding) and demands holiness of conduct, although all the moral perfection a human being can attain is always only virtue: that is, a disposition conforming to law out of respect for the law, and hence consciousness of a continual propensity to transgression, or at least to impurity, that is, the admixture of many spurious (nonmoral) motives in following the law. Consequently it requires a self-appraisal joined with humility, and thus, with regard to the holiness the Christian law demands, leaves the creature nothing but progress to infinity; but for that very reason it also entitles the creature to hope for its continued existence to infinity. The worth of a disposition fully conforming to the moral law is infinite, because all possible happiness, in the judgment of a wise and all-powerful distributor of it, has no other limitation than rational beings' lack of conformity to their duty. Yet the moral law by itself promises no happiness, for according to concepts of a natural order in general, happiness is not necessarily connected with following it. Christian moral teaching now makes up for this lack (of the second indispensable component of the highest good) by representing the world in which rational beings dedicate themselves with all their soul to the moral law as a kingdom of God. In it, nature A 232 and morals enter into a harmony, foreign to each in itself, through a holy author who makes the derived highest good possible. Holiness of conduct is already prescribed to them in this life as their V:129 guiding rule; but well-being proportionate to it, bliss, is represented as attainable only in an eternity. For the former must always be the archetype of their conduct in every condition, and progress toward it is already possible and necessary in this life; the latter, however, cannot be attained in this world under the name of happiness (so far as our powers are concerned), and is therefore made solely an object of hope. Despite this, the Christian principle of morality itself is not theological (and thus heteronomy), but the autonomy of pure practical reason on its own. For it makes cognition of God and his will the ground not of these laws, but only of attaining the highest good under the condition of following them; and it places even the proper incentive to following the former not in their desired consequences, but in the representation of duty alone, since only in its faithful observance does worthiness to acquire the latter consist.

In this way the moral law leads, through the concept of the A 233 highest good as the object and final end of pure practical reason, to religion, that is, to the cognition of all duties as divine commands—not as sanctions, that is, arbitrary ordinances of an alien will, contingent in themselves, but as essential laws of every free will in itself. Nevertheless, they must be regarded as commands of the highest being, because only from a morally perfect (holy and benevolent), and at the same time all-powerful, will can we hope for the highest good, which the moral law makes it our duty to set as the object of our striving, and thus hope to attain it through agreement with this will. Here too, therefore, everything remains unselfish and founded solely on duty, without fear or hope being allowed as underlying incentives; for if these become principles, they destroy the entire moral worth of actions. The moral law commands me to make the highest possible good in a world the ultimate object of all my conduct. But I cannot hope to bring this about except through the agreement of my will with that of a holy and benevolent author of the world. And although the concept of the highest good, as a whole in which the greatest happiness is represented as combined in the most exact proportion with the greatest degree of moral perfection (possible in creatures) A 234, also contains V:130 my own happiness, it is nevertheless not this happiness, but the moral law—which rather strictly restricts my unlimited desire for it to conditions—that is the determining ground of the will directed to promote the highest good.

Hence morality is not, properly speaking, the doctrine of how we are to make ourselves happy, but of how we are to become worthy of happiness. Only when religion is added does the hope also arise of someday sharing in happiness to the extent that we have taken care not to be unworthy of it.

Someone is worthy of possessing a thing or a condition when his possessing it accords with the highest good. We can now readily see that all worthiness depends on moral conduct, because in the concept of the highest good this constitutes the condition of the rest (what belongs to one's condition), namely one's share in happiness. It follows that morality in itself must never be treated as a doctrine of happiness, that is, as instruction in how to obtain a share in happiness; for it is concerned solely with the rational condition (conditio sine qua non) of happiness A 235, not with a means of acquiring it. But once morality—which merely imposes duties and does not supply measures for selfish wishes—has been set forth completely, then, and only then, after the law-grounded moral wish to promote the highest good (to bring the kingdom of God to us), which could never previously have arisen in a selfish soul, has been awakened, and the step toward religion has been taken to support it, this moral doctrine can also be called a doctrine of happiness, because the hope for happiness begins only with religion.

From this we can also see that if we ask about God's final end in creating the world, we must name not the happiness of the rational beings in it, but the highest good. This adds to that wish of these beings a further condition, namely being worthy of happiness, that is, the morality of those same rational beings, which alone contains the standard according to which alone they can hope to share in happiness through the hand of a wise author. For since wisdom, considered theoretically, means cognition of the highest good, and practically, the V:131 conformity of the will to the highest good, we cannot attribute to a supreme self-subsisting wisdom an end founded merely on benevolence. For the effect of benevolence (with A 236 regard to the happiness of rational beings) can be thought adequate to the highest original good only under the limiting conditions of agreement with the holiness¹⁴ of his will. Hence those who placed the end of creation in the glory of God (provided this is not thought anthropomorphically, as an inclination to be praised) have surely found the best expression. For nothing honors God more than what is most estimable in the world: respect for his command, observance of the holy duty his law imposes upon us, when his splendid provision is added A 237 to crown so beautiful an order with proportionate happiness. If the latter makes him lovable (to speak in human terms), the former makes him an object of worship (adoration). Even human beings can indeed win love through doing good, but never respect through this alone, so that the greatest beneficence honors them only insofar as it is exercised according to worthiness.

That in the order of ends the human being (and with him every rational being) is an end in himself, that is, can never be used by anyone (not even by God) merely as a means without at the same time being himself an end, and that therefore the humanity in our person must be holy to us ourselves, now follows of itself. For the human being is the subject of the moral law, and hence of what is holy in itself V:132, for whose sake and in agreement with which alone anything at all can be called holy. For this moral law rests on the autonomy of his will as a free will, which, according to its universal laws, must necessarily be able at the same time to assent to that to which it is to submit.

VI. A 238 On the postulates of pure practical reason in general.

They all proceed from the principle of morality, which is not a postulate but a law through which reason immediately determines the will. Precisely because it is so determined, this will, as a pure will, requires these necessary conditions for following its precept. These postulates are not theoretical dogmas, but presuppositions from a necessarily practical standpoint. Thus, although they do not extend speculative cognition, they give objective reality to the ideas of speculative reason in general (through their relation to the practical), and entitle it to concepts whose very possibility it could otherwise not presume to assert.

These postulates are those of immortality, of freedom considered positively (as the causality of a being insofar as it belongs to the intelligible world), and of the existence of God. The first flows from the practically necessary condition that duration be adequate to the completeness of the fulfillment of the moral law; the second, from the necessary presupposition of independence from the sensible world and of the capacity to determine one's will according to the law of an A 239 intelligible world, that is, freedom; the third, from the necessity of the condition for such an intelligible world to be the highest good, through the presupposition of the highest self-subsisting good, that is, the existence of God.

The aim directed toward the highest good, made necessary by respect for the moral law, and the presupposition of its objective reality that flows from this aim, thus lead through postulates of practical reason to concepts that speculative reason could indeed present as problems, but could not solve. Thus, 1. to the problem in whose solution V:133 the latter could commit nothing but paralogisms (namely immortality), because it lacked the mark of persistence needed to complete the psychological concept of an ultimate subject, necessarily attributed to the soul in self-consciousness, into the real representation of a substance. Practical reason accomplishes this through the postulate of a duration required for conformity with the moral law in the highest good, as the entire end of practical reason. 2. It leads to that concerning which speculative reason contained nothing but antinomy, whose resolution it could ground only on a concept that could indeed be thought problematically, but whose objective reality it could neither prove nor determine: namely the cosmological idea of an intelligible world and consciousness of our A 240 existence in it, by means of the postulate of freedom (whose reality practical reason demonstrates through the moral law, and with it also the law of an intelligible world, to which speculative reason could only point, without determining its concept). 3. It gives meaning to what speculative reason could indeed think, but had to leave undetermined as a mere transcendental ideal: the theological concept of the original being. It does so for a practical purpose, that is, as a condition of the possibility of the object of a will determined by that law, as the supreme principle of the highest good in an intelligible world through authoritative moral legislation within it.

But is our cognition actually extended in this way through pure practical reason, and is what was transcendent for speculative reason immanent in practical reason? Certainly, but only for a practical purpose. For through this we do not cognize the nature of our soul, or the intelligible world, or the highest being as to what they are in themselves. Rather, we have only united the concepts of them in the practical concept of the highest good, as the object of our will, wholly a priori through pure reason, but only by means of the moral law, and solely in relation to it as regards the object it commands. Yet how even freedom A 241 is possible, and how this kind of causality is to be represented theoretically and positively, is not thereby understood. Only that there is such a causality is postulated through the moral law and for its sake. The same holds for the remaining ideas: no human understanding will ever fathom their possibility, but neither will any sophistry ever wrest from the conviction even of the most ordinary person the belief that they are V:134 true concepts.

VII. How is it possible to conceive an extension of pure reason for a practical purpose without thereby also extending its cognition as speculative?

To avoid becoming too abstract, we shall answer this question at once by applying it to the case before us.—To extend a pure cognition practically, an aim must be given a priori, that is, an end as object (of the will) which, independently of all theoretical principles, is represented as practically necessary through an imperative that immediately determines the will (a categorical imperative); here this is the highest good. But this is not possible without presupposing three theoretical concepts (for which, since they are merely pure concepts of reason, no corresponding intuition, and hence no objective reality by the theoretical A 242 path, can be found): namely freedom, immortality, and God. Thus the practical law, which commands the existence of the highest good possible in a world, postulates the possibility of those objects of pure speculative reason, the objective reality that speculative reason could not secure for them. Through this, theoretical cognition of pure reason certainly receives an increment, but one consisting only in this: those concepts, otherwise problematic for it (merely thinkable), are now assertorically declared to be concepts to which objects actually belong. For practical reason unavoidably needs their existence for the possibility of its object, the highest good, which is indeed practically absolutely necessary; and theoretical reason is thereby entitled to presuppose them. But this extension of theoretical reason is no extension of speculation, that is, one permitting a positive use of them henceforth for a theoretical purpose. For practical reason has accomplished nothing here beyond establishing that those concepts are real and actually have their (possible) objects, while giving us no intuition of them (nor can such an intuition be required). Thus no synthetic proposition is possible through this admitted reality of theirs. Consequently, this disclosure does not help us A 243 in the least for a speculative purpose, but it does extend this cognition of ours V:135 with regard to the practical use of pure reason. The three ideas of speculative reason mentioned above are not in themselves cognitions as yet; nevertheless, they are (transcendent) thoughts containing nothing impossible. Now, through an apodictic practical law, as necessary conditions of the possibility of what this law commands us to make our object, they acquire objective reality: that is, we are instructed by this law that they have objects, without being able to show how their concept relates to an object. And this is still not cognition of these objects, for through it we can make no synthetic judgment about them, nor determine their application theoretically, and thus can make no theoretical use of reason concerning them at all, in which all speculative cognition of them properly consists. Nevertheless, theoretical cognition, though not of these objects, but of reason in general, was thereby extended to this extent: through the practical postulates, objects were nevertheless given to those ideas, as a merely problematic thought thereby first acquired objective reality. Thus it was not an extension of cognition of given supersensible objects, but nevertheless an extension of theoretical reason and its A 244 cognition with regard to the supersensible in general, insofar as it was compelled to admit that there are such objects, without being able to determine them more closely, and hence without being able to extend this cognition of the objects themselves (now given to it on a practical ground, and only for practical use). Pure theoretical reason, for which all those ideas are transcendent and without an object, therefore owes this increment solely to its pure practical capacity. Here they become immanent and constitutive, since they are grounds of the possibility of making actual the necessary object of pure practical reason (the highest good). Without this, they are transcendent and merely regulative principles of speculative reason, imposing on it not the assumption of a new object beyond experience, but only the task of bringing its use within experience closer to completeness. Once reason possesses this increment, however, it will, as speculative reason (properly speaking, only to safeguard its practical use), proceed negatively with those ideas—that is, not extending but purifying them—in order to ward off, on the one hand, anthropomorphism as the source of superstition V:136, or the apparent extension of those concepts through supposed experience, and on the other, fanaticism, which promises such extension through supersensible intuition or similar feelings. All these are hindrances A 245 to the practical use of pure reason. Warding them off therefore certainly belongs to the extension of our cognition for a practical purpose, without its contradicting this extension to admit at the same time that, for a speculative purpose, reason has thereby gained nothing in the least.

Every use of reason with regard to an object requires pure concepts of the understanding (categories), without which no object can be thought. These can be applied to the theoretical use of reason, that is, to cognition of this kind, only insofar as an intuition (which is always sensible) is at the same time supplied as their basis, and thus only to represent through them an object of possible experience. Here, however, what I would have to think through categories in order to cognize it consists of ideas of reason, which cannot be given in any experience. But here we are also concerned not with theoretical cognition of the objects of these ideas, but only with their having objects at all. Pure practical reason supplies this reality, and theoretical reason has nothing more to do here than merely to think those objects through categories. As we have clearly shown elsewhere, this is quite possible without needing intuition (either sensible or supersensible), because the categories have their seat and origin in the A 246 pure understanding, independently of and prior to all intuition, solely as the faculty of thinking; and they always signify only an object in general, however it may be given to us. Now it is indeed impossible to give the categories an object in intuition insofar as they are to be applied to those ideas. Nevertheless, through an object that practical reason indubitably offers in the concept of the highest good, it is sufficiently secured for them that such an object actually exists, and hence that the category as a mere form of thought is not empty here, but has meaning: the reality of the concepts belonging to the possibility of the highest good is secured, without this increment nevertheless effecting the slightest extension of cognition according to theoretical principles.

V:137 If, moreover, these ideas of God, an intelligible world (the kingdom of God), and immortality are determined through predicates taken from our own nature, this determination must be regarded neither as rendering sensible those pure ideas of reason (anthropomorphisms), nor as a cognition of supersensible objects that exceeds our powers. For these predicates are none other than understanding and will, considered in relation to one another A 247 as they must be thought in the moral law, and thus only insofar as a pure practical use is made of them. Everything else that attaches to these concepts psychologically is then abstracted away: that is, insofar as we empirically observe these faculties of ours in their exercise (for example, that the human understanding is discursive, and its representations are therefore thoughts, not intuitions; that these succeed one another in time; that its will is always bound up with a dependence of satisfaction on the existence of its object, and so on, which cannot be so in the highest being). Thus nothing remains of the concepts through which we think a pure being of understanding beyond precisely what is required for the possibility of thinking a moral law; hence there is indeed a cognition of God, but only in a practical relation. If we attempt to extend it into a theoretical cognition, we obtain an understanding of his that does not think but intuits, and a will directed toward objects on whose existence his satisfaction does not depend in the least. (I will not even mention transcendental predicates, such as a magnitude of existence, that is, duration, which nevertheless does not occur in time, the only means possible for us of representing existence as a magnitude.) These are all properties of which we can form no A 248 concept at all suited to cognition of the object; and through this we learn that they can never be used for a theory of supersensible beings, and thus cannot ground speculative cognition on this side at all, but restrict their use solely to the practice of the moral law.

This last point is so evident, and can be proved so clearly in actual application, that we may confidently challenge all supposed natural theologians (a curious name)¹⁵ to name even a single property determining this V:138 object of theirs (beyond merely ontological predicates), say a property of understanding or of will, of which one could not irrefutably demonstrate that, if everything A 249 anthropomorphic is separated from it, only the bare word remains for us, without our being able to connect with it the slightest concept from which an extension of theoretical cognition might be hoped. In the practical sphere, however, there nevertheless remains to us, from the properties of an understanding and a will, the concept of a relation to which the practical law (which determines precisely this relation of understanding to will a priori) gives objective reality. Once this has occurred, reality is also given to the concept of the object of a morally determined will (that of the highest good), and with it to the conditions of its possibility, the ideas of God, freedom, and immortality—but always only in relation to the practice of the moral law (for no speculative purpose).

After these reminders, the answer to the important question is now also easy to find: whether the concept of God belongs to physics (and hence also to metaphysics, which contains only the pure principles a priori of physics in their general meaning) or to morality. To explain arrangements in nature or their alteration by taking refuge in God as the author of all things is at least no physical explanation, and is always an admission that one has reached the end of one's philosophy: for one is compelled to assume something of which one otherwise has no concept in itself, in order to form a concept of the A 250 possibility of what one sees before one's eyes. But to proceed through metaphysics from knowledge of this world to the concept of God and the proof of his existence by secure inferences is impossible for this reason: we would have to cognize this world as the most perfect possible whole, and hence, for this purpose, all possible worlds (in order to compare them with this one); thus we would have to be omniscient V:139 in order to say that it was possible only through a God (as we must think this concept). Moreover, to cognize this being's existence from mere concepts is absolutely impossible, because every existential proposition—that is, one that says of a being of which I form a concept that it exists—is a synthetic proposition. That is, it is a proposition through which I go beyond that concept and say more of it than was thought in the concept: namely, that to this concept in the understanding there is also posited a corresponding object outside the understanding, which obviously cannot be established through any inference. Thus only a single procedure remains for reason to attain this cognition: as pure reason, proceeding from the supreme principle of its pure practical use (since this use is in any case directed solely toward the existence of something as a consequence of reason), it determines its object. And here there appears, in its unavoidable task, namely the necessary direction of the will toward the highest good, not only A 251 the necessity of assuming such an original being in relation to the possibility of this good in the world, but, most remarkably, something entirely lacking in reason's advance along the path of nature: an exactly determined concept of this original being. Since we know this world only in small part, and are still less able to compare it with all possible worlds, we can indeed infer from its order, purposiveness, and magnitude a wise, benevolent, powerful, etc., author of it, but not his omniscience, all-benevolence, omnipotence, and so on. We can also readily concede that we are entitled to make up for this unavoidable lack through a permissible, entirely reasonable hypothesis: namely, that if wisdom, benevolence, etc., shine forth in so many particulars as present themselves to our closer knowledge, it will be the same in all the rest, and that it is therefore reasonable to attribute every possible perfection to the author of the world. But these are not inferences through which we may pride ourselves on our insight, but only permissions that may be granted us, and that still need a recommendation from elsewhere if we are to make use of them. Thus, on the empirical path (of physics), the concept of God always remains a concept insufficiently determined A 252 as to the perfection of the first being for us to hold it adequate to the concept of a deity (while metaphysics, in its transcendental part, can accomplish nothing at all).

V:140 I now try to set this concept against the object of practical reason, and find that the moral principle allows this object as possible only on the presupposition of an author of the world of highest perfection. He must be omniscient, in order to cognize my conduct down to the innermost part of my disposition, in all possible cases and for all the future; omnipotent, in order to accord it its appropriate consequences; likewise omnipresent, eternal, and so on. Thus the moral law, through the concept of the highest good as the object of a pure practical reason, determines the concept of the original being as the highest being, which the physical (and, carried higher, metaphysical), and hence the entire speculative, course of reason could not accomplish. Thus the concept of God is a concept belonging originally not to physics, that is, to speculative reason, but to morality; and the same can also be said of the other concepts of reason that we discussed above as its postulates in its practical use.

If in the history of Greek philosophy we find no clear traces of a pure rational theology before A 253 Anaxagoras, the reason is not that the earlier philosophers lacked the understanding and insight to rise to it by way of speculation, at least with the help of an entirely reasonable hypothesis. What could be easier, what more natural, than the thought that presents itself to everyone of its own accord: instead of indeterminate degrees of perfection in various causes of the world, to assume a single rational cause possessing all perfection? But the evils in the world seemed to them far too weighty as objections for them to consider themselves entitled to such a hypothesis. Thus they showed understanding and insight precisely in not permitting themselves this assumption, but instead searching among natural causes to see whether they might find there the constitution and capacity required for an original being. But after this acute people had advanced so far in their inquiries as to treat philosophically even moral objects, about which other peoples had never done more than chatter, only then did they discover a new need, namely a practical one. This did not fail to give them a determinate concept of the original being, while speculative reason had to look on, with at most the further merit of embellishing a concept that had not grown on its own soil. With a retinue A 254 V:141 of confirmations from the contemplation of nature, which only now came forward, speculative reason could promote not so much the concept's authority (which was already established) as merely its ostentatious display of supposed theoretical rational insight.

From these reminders, the reader of the Critique of Pure Speculative Reason will be fully convinced how utterly necessary, how fruitful for theology and morality, was that laborious deduction of the categories. For only through it can we prevent ourselves, when we place them in the pure understanding, from taking them with Plato as innate and founding on them extravagant pretensions, with theories of the supersensible to which no end can be foreseen, thereby making theology into a magic lantern of phantoms; or, if we regard them as acquired, prevent ourselves from restricting, with Epicurus, each and every use of them, even for a practical purpose, solely to objects and determining grounds of the senses. Now, however, the Critique has in that deduction proved, first, that they are not of empirical origin, but have their seat and source a priori in the pure understanding; and second, that, since they are related to objects in general, independently of their intuition, they indeed produce theoretical cognition only when applied A 255 to empirical objects, but nevertheless also serve, when applied to an object given through pure practical reason, for determinate thinking of the supersensible—although only insofar as this is determined merely through predicates necessarily belonging to the pure practical aim given a priori and to its possibility. The speculative restriction of pure reason and its practical extension first bring it into that balanced relation in which reason in general can be used purposively; and this example proves better than any other that, for us human beings, the path to wisdom, if it is to be secure and not impassable or misleading, must unavoidably pass through science. Yet we can be convinced that science leads to that goal only after its completion.

VIII. V:142 On assent arising from a need of pure reason.

A need of pure reason in its speculative use leads only to hypotheses, but that of pure practical reason to A 256 postulates. For in the first case I ascend from the derived as high in the series of grounds as I please, and need an original ground not to give objective reality to what is derived (for example, the causal connection of things and changes in the world), but only to satisfy my inquiring reason completely with regard to it. Thus I see order and purposiveness in nature before me, and do not need to proceed to speculation to assure myself of their actuality, but only to explain them, by presupposing a deity as their cause. Since an inference from an effect to a determinate cause—especially one so exactly and completely determined as we must think God—remains uncertain and precarious, such a presupposition can be carried no further than to the degree of the most reasonable opinion for us human beings.¹⁶ By contrast A 257, a need of pure practical reason is grounded on a duty to make something (the highest good) the object of my will, in order to promote it with all my powers. In doing so, however, I must presuppose its possibility, and hence also its conditions, namely God, freedom, and immortality, because through my speculative reason I can neither prove nor refute these. This duty rests on a law that is indeed quite independent of these latter presuppositions and apodictically certain in itself, namely the moral law; and to that extent it needs no further support V:143 from theoretical opinion concerning the inner constitution of things, the secret aim of the world's order, or a ruler presiding over it, in order to bind us most completely to actions unconditionally conforming to law. But the subjective effect of this law, namely the disposition conforming to it and also made necessary by it, to promote the highest good practically possible, nevertheless presupposes at least that this good is possible. Otherwise it would be practically impossible to strive toward the object of a concept that was fundamentally empty and without an object. Now the postulates above concern only A 258 the physical or metaphysical conditions—in a word, those lying in the nature of things—of the possibility of the highest good. But they do so not for an arbitrary speculative aim, but for a practically necessary end of the pure rational will, which here does not choose but obeys an unremitting command of reason. This command has its ground objectively in the constitution of things as they must be universally judged through pure reason, and is not founded on inclination, which is by no means entitled, for the sake of what we wish on merely subjective grounds, immediately to assume the means to it as possible, or even the object as actual. Thus this is a need for an absolutely necessary purpose, and justifies its presupposition not merely as a permissible hypothesis, but as a postulate for a practical purpose. And granted that the pure moral law unremittingly binds everyone as a command (not as a rule of prudence), the upright person may well say: “I will that there be a God; that my existence in this world be also, beyond the nexus of nature, an existence in a pure world of understanding; and finally, that my duration be endless. I hold fast to this and will not let this faith be taken from me; for this is the sole case in which my interest, because I may not relinquish any of it, unavoidably determines my judgment, without A 259 heeding rationalizations, however little I might be able to answer them or oppose them with others more plausible.”¹⁷

V:144 To prevent misinterpretations in using a concept still so unfamiliar as that of a pure practical rational faith, A 260 let me add one further remark.—It might almost seem as though this rational faith were itself being proclaimed here as a command, namely to assume the highest good as possible. But a faith that is commanded is an absurdity. Recall, however, the explanation above of what is required to be assumed in the concept of the highest good, and it will become clear that assuming this possibility must not be commanded at all, and that no practical dispositions are required to concede it; rather, speculative reason must grant it unasked. For no one would wish to assert that it is in itself impossible for the worthiness of rational beings in the world to be happy, in conformity with the moral law, to be connected with their possession of happiness proportionate to it. Now with regard to the first part of the highest good, namely morality, the moral law gives us only a command; and to doubt the possibility of this component would amount to calling the moral law itself into question. But as for the second part of that object, namely happiness conforming throughout to that worthiness, no command is needed at all to concede its possibility in general, for theoretical reason itself has nothing against it. Only the manner in which we are to think such a harmony of the laws of nature with those A 261 V:145 of freedom contains something about which a choice is open to us, because theoretical reason decides nothing here with apodictic certainty; and regarding this choice there can be a moral interest that tips the balance.

Above I said that, according to a mere natural course of things in the world, happiness exactly proportionate to moral worth cannot be expected and must be held impossible, and that therefore the possibility of the highest good, from this side, can be conceded only on the presupposition of a moral author of the world. I deliberately withheld the restriction of this judgment to the subjective conditions of our reason, so as to make use of it only when the kind of assent involved was to be more closely determined. In fact, the impossibility mentioned is merely subjective: that is, our reason finds it impossible for itself to render comprehensible, according to a mere course of nature, so exactly proportionate and thoroughly purposive a connection between two events in the world occurring according to such different laws. Yet, as with everything else purposive in nature, it also cannot prove the impossibility of this connection according to universal natural laws, that is, demonstrate it sufficiently on objective A 262 grounds.

Now, however, a deciding ground of another kind comes into play, to tip the balance amid speculative reason's uncertainty. The command to promote the highest good is objectively grounded (in practical reason), as is its possibility in general (in theoretical reason, which has nothing against it). But reason cannot objectively decide how we are to represent this possibility: whether according to universal natural laws without a wise author presiding over nature, or only on the presupposition of such an author. Here a subjective condition of reason now enters: the only way theoretically possible for it, and at the same time the only one conducive to morality (which stands under an objective law of reason), of thinking the exact agreement of the kingdom of nature with the kingdom of morals as the condition of the possibility of the highest good. Now, promoting this good, and hence presupposing its possibility, is objectively necessary (but only according to practical reason); yet the manner in which we choose to think it as possible lies within our choice, in which a free interest of V:146 pure practical reason decides in favor of assuming a wise author of the world. Thus the principle determining our judgment here is indeed A 263 subjective as a need, but at the same time, as a means of promoting what is objectively (practically) necessary, it is the ground of a maxim of assent for a moral purpose, that is, a pure practical rational faith. This faith, then, is not commanded. Rather, as a voluntary determination of our judgment to assume that existence and henceforth to lay it at the foundation of the use of reason—a determination conducive to the moral (commanded) aim, and moreover in agreement with reason's theoretical need—it has itself arisen from the moral disposition. It can therefore sometimes waver, even repeatedly in well-disposed people, but can never pass into unbelief.

IX. On the proportion of our cognitive faculties wisely suited to the human being's practical vocation.

If human nature is destined to strive toward the highest good, the measure of its cognitive faculties, especially their relation to one another, must also be assumed suited to this end. Now, however, the Critique of Pure Speculative Reason proves its greatest inadequacy to solve, in a way appropriate to that end, the most important problems A 264 set before it. It does not, admittedly, fail to recognize the natural and unmistakable hints of that same reason, or the great steps it can take to approach this great goal set before it, though never to attain it on its own, even with the aid of the greatest knowledge of nature. Thus nature here seems to have provided us only stingily, like a stepmother, with a faculty needed for our end.

Suppose, then, that nature had complied with our wish in this respect and granted us the capacity for insight or illumination we would gladly possess, or which some even imagine themselves actually to possess: what, by all appearances, would the consequence be? Unless our entire nature were transformed at the same time, the inclinations, which always have the first word, would first V:147 demand their satisfaction and, joined with rational deliberation, their greatest possible and lasting satisfaction under the name of happiness. The moral law would speak afterward, to keep them within their proper limits and even to subject them all together to a higher end that takes no account of any inclination. But instead of the struggle the moral disposition now has to wage against the inclinations, in which, after some defeats, the soul's moral strength can nevertheless A 265 gradually be acquired, God and eternity, with their terrible majesty, would lie incessantly before our eyes (for what we can prove completely counts for us, in point of certainty, as much as what we ascertain by direct sight). Transgression of the law would of course be avoided, and what is commanded would be done. But because the disposition from which actions ought to proceed cannot be instilled by any command, while the spur to activity here would be immediately at hand and external, reason would not first have to work its way upward to gather strength for resistance to inclinations through a vivid representation of the law's dignity. Thus most actions conforming to law would be done from fear, only a few from hope, and none at all from duty; and a moral worth of actions, on which alone the worth of the person and even of the world depends in the eyes of the highest wisdom, would not exist at all. Human conduct, as long as human nature remained as it now is, would therefore be transformed into a mere mechanism in which, as in a puppet show, everything would gesture well, but no life would be found in the figures. As things stand with us, however, it is quite otherwise. With every exertion of our reason, we have only a very dim and ambiguous A 266 prospect of the future; the ruler of the world lets us only conjecture his existence and his glory, not behold or clearly prove them. By contrast, the moral law within us, without promising or threatening us anything with certainty, demands unselfish respect from us; and only when this respect has become active and predominant, and only through it, does it further permit prospects into the realm of the supersensible, and even then only in faint glimpses. Thus a truly moral disposition, immediately dedicated to the law, can arise, and the rational creature can become worthy of a share in the highest good proportionate to the moral worth of his person and not merely to his actions V:148. Thus here too there may well be truth in what the study of nature and humanity otherwise teaches us sufficiently: that the unfathomable wisdom through which we exist is no less worthy of reverence in what it has denied us than in what it has bestowed upon us.

12 The conviction that one's disposition is unalterable in one's progress toward the good seems nevertheless impossible for a creature on its own. For this reason Christian religious doctrine also derives it solely from the same spirit that brings about sanctification, that is, this firm resolve and with it consciousness of perseverance in moral progress. But even in a natural way, someone conscious of having persisted for a long part of his life, up to its end, in progress toward the better, and indeed from genuine moral motives, may well cherish the consoling hope, though not the certainty, that even in an existence continued beyond this life he will persevere in these principles. And although in his own eyes he is never justified here, nor may ever hope to be so with the hoped-for future increase of his natural perfection and, with it, of his duties, he may nevertheless have in this progress a prospect of a blissful future. For although this progress concerns a goal removed to infinity, it nevertheless counts for God as possession. This is the expression reason uses to designate complete well-being, independent of all contingent causes in the world, which, like holiness, is an idea that can be contained only in an infinite progress and its totality, and is therefore never fully attained by the creature.

13 It is commonly held that the Christian prescription of morals has no advantage in purity over the Stoics' moral concept; yet the difference between the two is very apparent. The Stoic system made consciousness of strength of soul the pivot around which all moral dispositions were to turn. Although its adherents spoke of duties, and also determined them quite well, they nevertheless placed the incentive and the true determining ground of the will in an elevation of one's way of thinking above the base incentives of the senses, which gain power only through weakness of soul. Virtue for them was therefore a certain heroism of the sage, rising above human animal nature, sufficient to himself, prescribing duties to others but himself exalted above them and subject to no temptation to transgress the moral law. But they could not have done all this had they represented this law to themselves in the purity and strictness with which the Gospel's prescription presents it. If by an idea I understand a perfection to which nothing adequate can be given in experience, moral ideas are not on that account beyond our powers—that is, of a kind whose very concept we could not even determine sufficiently, or of which it is uncertain whether any object at all corresponds to it, as with the ideas of speculative reason. Rather, as archetypes of practical perfection, they serve as the indispensable guiding rule of moral conduct and at the same time as the standard of comparison. If I now consider Christian morality from its philosophical side, it would appear, compared with the ideas of the Greek schools, as follows: the ideas of the Cynics, the Epicureans, the Stoics, and the Christians are natural simplicity, prudence, wisdom, and holiness. As regards the path to attaining them, the Greek philosophers differed in that the Cynics found common human understanding sufficient, while the others found only the path of science sufficient; thus both found the mere use of natural powers sufficient. Christian morality, because it frames its prescription so purely and unyieldingly (as indeed it must), takes from the human being the confidence of being fully adequate to it, at least here in life. Yet it also restores this confidence by holding that, if we act as well as lies in our power, we may hope that what is not in our power will come to our aid from elsewhere, whether or not we know in what way. Aristotle and Plato differed only with regard to the origin of our moral concepts.

14 Here, to make clear what is distinctive about these concepts, I add only this remark. We attribute to God various properties whose quality we also find appropriate to creatures, except that in God they are raised to the highest degree: for example, power, knowledge, presence, goodness, etc., under the names of omnipotence, omniscience, omnipresence, all-benevolence, etc. Yet there are three that are attributed exclusively to God, without any addition of magnitude, and all are moral: he is the only holy one, the only blessed one, the only wise one, because these concepts already carry unrestrictedness with them. According to their order, he is therefore also the holy lawgiver (and creator), the benevolent ruler (and sustainer), and the just judge: three properties containing everything through which God becomes the object of religion, and to which the metaphysical perfections fittingly add themselves of their own accord in reason.

15 Erudition is properly only the sum of historical sciences. Consequently, only the teacher of revealed theology can be called a scholar of divinity. But suppose we also wished to call someone who possesses rational sciences (mathematics and philosophy) a scholar, although this would already contradict the word's meaning, which always counts as erudition only what one must actually be taught, and thus cannot discover for oneself through reason. Even then, the philosopher, with his cognition of God as a positive science, might cut too poor a figure to let himself be called a scholar on that account.

16 But even here we could not plead a need of reason unless we had before us a problematic yet unavoidable concept of reason, namely that of an absolutely necessary being. This concept now demands to be determined; and when the drive toward extension is added, this is the objective ground of a need of speculative reason: namely, to determine more closely the concept of a necessary being that is to serve as the original ground for others, and thus to make this ultimate being recognizable by some means. Without such antecedent necessary problems there are no needs, at least none of pure reason; the others are needs of inclination.

17 In the Deutsches Museum, Feb. 1787, there is an essay by a very subtle and lucid thinker, the late Wizenmann, whose early death is to be lamented. In it he disputes the entitlement to infer from a need the objective reality of its object, and illustrates his point through the example of a lover who, having become enamored of an idea of beauty that is merely a figment of his brain, would infer that such an object actually exists somewhere. I agree with him completely in all cases where the need is grounded on inclination. Inclination cannot postulate the existence of its object as necessary even for the person afflicted by it, much less contain a demand valid for everyone; it is therefore a merely subjective ground of wishes. Here, however, it is a need of reason, arising from an objective determining ground of the will, namely the moral law, which necessarily binds every rational being. It therefore entitles us a priori to presuppose the conditions in nature appropriate to it, and makes these inseparable from the complete practical use of reason. It is a duty to make the highest good actual to the utmost of our ability; therefore it must also be possible. Hence it is also unavoidable for every rational being in the world to presuppose what is necessary for its objective possibility. This presupposition is as necessary as the moral law, in relation to which alone it is also valid.

Chapter II. A 198 The Dialectic of Pure Reason in Determining the Concept of the Highest Good.

The concept of the highest already contains an ambiguity. Unless we notice it, it can cause needless disputes. The highest can mean either the supreme or the complete. The first is a condition that is itself unconditional, that is, subordinate to no other condition: the original condition. The second is a whole that is not part of any greater whole of the same kind: the most perfect whole. The Analytic proved that virtue, as worthiness to be happy, is the supreme condition of everything that can seem desirable to us. It is therefore also the supreme condition of all our efforts to gain happiness, and so the supreme good. But that does not yet make it the whole and complete good as an object of the faculty of desire of finite rational beings. For that, happiness is also needed. This is true not just in the partial view of a person who makes himself his end A 199, but even in the judgment of impartial reason, which regards that person in the world in general as an end in himself. For to need happiness and even to deserve it, yet have no share in it, cannot possibly agree with the perfect willing of a rational being who also had all power, even if we only imagine such a being as an experiment. Virtue and happiness together, then, constitute a person's possession of the highest good. And when happiness is distributed in exact proportion to morality—to the person's worth and worthiness to be happy—these two together constitute the highest good of a possible world. Here this means the V:111 whole, the complete good. Within it, however, virtue, as the condition, is always the supreme good, since it has no further condition above it. Happiness is always something pleasant for its possessor, but it is not by itself unconditionally good in every respect. It always presupposes morally lawful conduct as its condition.

Two determinations that are necessarily united in one concept must be connected as ground and consequence. This unity must be regarded either as analytic, a logical connection under the law of identity, or as synthetic, a real connection under the law of A 200 causality. The connection between virtue and happiness can therefore be understood in either of two ways. Striving to be virtuous and rationally seeking happiness might be not two different actions but entirely identical actions. In that case the first would need no maxim other than the one underlying the second. Or their connection is taken to mean that virtue produces happiness as something distinct from the consciousness of virtue, as a cause produces an effect.

Among the ancient Greek schools, only two really followed the same method in determining the concept of the highest good in this respect: they did not accept virtue and happiness as two different elements of the highest good. They therefore sought the unity of their principle under the rule of identity. But they differed in which of the two they chose as the basic concept. The Epicurean said: being conscious of your maxim that leads to happiness is virtue. The Stoic said: being conscious of your virtue is happiness. For the first, prudence amounted to morality. For the second, who chose a higher name for virtue, morality alone was true wisdom.

We must regret that these men's ingenuity was misapplied A 201 in trying to discover identity between such extremely different concepts as happiness and virtue. Yet we must also admire them for having tried every conceivable path of philosophical discovery so early. Their effort suited the dialectical spirit of their times. Even now this sometimes tempts subtle thinkers to remove essential differences between principles—differences that can never be reconciled—by trying to turn them into disputes about words. In this way they manufacture an apparent unity of concept under V:112 different names. This usually happens when the union of unlike grounds lies so deep or so high, or would require such a complete change in the other doctrines accepted in a philosophical system, that people hesitate to examine the real difference thoroughly. They prefer to treat it as a disagreement over mere formalities.

Both schools tried to establish that the practical principles of virtue and happiness were identical. But they did not agree on how to force this identity. Instead, they diverged enormously. One placed its principle on the side of feeling, the other on the side of logic. The first placed it in consciousness of sensible need; the second in practical reason's independence A 202 from all sensible determining grounds. According to the Epicurean, the concept of virtue was already contained in the maxim of promoting one's own happiness. According to the Stoic, by contrast, the feeling of happiness was already contained in consciousness of one's virtue. But what is contained in another concept is identical with part of what contains it, not with the whole. Moreover, two wholes can differ in kind even when they consist of exactly the same material, if their parts are combined into wholes in entirely different ways. The Stoic claimed that virtue was the entire highest good. Happiness was only consciousness of possessing virtue, as belonging to the subject's condition. The Epicurean claimed that happiness was the entire highest good. Virtue was only the form of the maxim for pursuing it, namely, the rational use of the means to it.

But the Analytic makes it clear that the maxims of virtue and those of personal happiness are entirely different in their supreme practical principle. Far from agreeing, they severely limit and interfere with each other within the same subject, even though they belong to one highest good and make it possible. So the question, how is the highest good practically A 203 possible?, remains unsolved despite all previous attempts to combine them. The Analytic has given us what makes this problem difficult: happiness and morality are two elements of entirely different kinds in the highest good. Their connection therefore cannot be cognized analytically V:113. Merely analyzing his concepts will not show someone who pursues happiness that he is virtuous in doing so; nor will someone who follows virtue find himself already happy by that very fact in being conscious of such conduct. Instead, their connection is a synthesis of concepts. But this connection is cognized as a priori, and thus as practically necessary, not as derived from experience. The possibility of the highest good therefore rests on no empirical principles. So the deduction of this concept will have to be transcendental. It is a priori morally necessary to bring about the highest good through freedom of the will. The condition of its possibility must therefore also rest entirely on a priori grounds of cognition.

I. A 204 The antinomy of practical reason.

In the highest good that is practical for us—that is, that our will is to make real—virtue and happiness are thought of as necessarily connected. Pure practical reason cannot accept one without the other also belonging to it. Now this connection, like any connection, is either analytic or synthetic. As we just showed, this particular connection cannot be analytic. It must therefore be thought of as synthetic, specifically as a connection of cause and effect, since it concerns a practical good, something possible through action. So either the desire for happiness must motivate maxims of virtue, or the maxim of virtue must be the effective cause of happiness. The first is absolutely impossible. As the Analytic proved, maxims that place the will's determining ground in the desire for one's happiness are not moral at all and cannot establish virtue. But the second is also impossible. Every practical connection between causes and effects in the world that results from the determination of the will depends not on A 205 the will's moral disposition, but on knowledge of natural laws and the physical ability to use them for one's purposes. Consequently, even the most exact V:114 observance of moral laws cannot be expected to produce a necessary connection between happiness and virtue in the world sufficient for the highest good. Now promoting the highest good, whose concept contains this connection, is an a priori necessary object of our will. It is inseparably connected with the moral law. So the impossibility of promoting it must also prove the moral law false. If the highest good is impossible under practical rules, then the moral law that commands us to promote it must also be fanciful and directed toward empty, imaginary ends. It must therefore be false in itself.

II. Critical resolution of the antinomy of practical reason.

The antinomy of pure speculative reason contains a similar conflict between natural necessity and freedom in the causality of events in the world. This was resolved by proving that there is no real conflict if we regard events A 206, and even the world in which they occur, only as appearances—as we should. One and the same acting being, as an appearance, even to its own inner sense, has a causality in the sensible world that always follows nature's mechanism. Yet with regard to the same event, insofar as the acting person also regards himself as a noumenon—as a pure intelligence whose existence cannot be determined in time—he can contain a determining ground of that causality under natural laws which is itself free from every natural law.

The present antinomy of pure practical reason works in exactly the same way. The first of its two propositions, that striving for happiness produces a ground for a virtuous disposition, is absolutely false. But the second, that a virtuous disposition necessarily produces happiness, is not absolutely false. It is false only insofar as virtue is regarded as the form of causality in the sensible world, and thus if I take existence in that world to be the only kind of existence a rational being has. It is therefore false only conditionally. But I am entitled to think of my existence also as a noumenon in a world of understanding. I even V:115 have in the moral law a purely intellectual determining ground of my causality in the sensible world. So it is not impossible A 207 for the morality of my disposition, as cause, to have a necessary connection with happiness, as effect, in the sensible world. This connection might be, if not direct, at least indirect, through an intelligible author of nature. In a nature that is merely an object of the senses, this connection can never be anything but contingent, and cannot suffice for the highest good.

So despite this apparent conflict of practical reason with itself, the highest good is the necessary supreme end of a morally determined will and a genuine object of that reason. For it is practically possible. The maxims of that will, whose content concerns the highest good, have objective reality. At first, the antinomy seemed to threaten this reality by challenging the connection of morality with happiness under a universal law. But that was only a misunderstanding: a relation between appearances was taken to be a relation of things in themselves to those appearances.

We find ourselves forced to seek the possibility of the highest good so far afield: in its connection with an intelligible world. This is the goal reason sets before all rational beings for all their moral wishes. It must therefore seem strange A 208 that ancient and modern philosophers alike could nevertheless believe they found happiness proportioned quite properly to virtue already in this life, in the sensible world, or persuade themselves that they were conscious of it. Both Epicurus and the Stoics praised above everything the happiness that arose in life from consciousness of virtue. In his practical rules, Epicurus was not as base-minded as one might infer from the principles of his theory, which he used to explain things, not to guide action. Nor was he as many interpreted him to be, misled by the expression sensual pleasure instead of contentment. He counted the most unselfish practice of good among the ways of enjoying the deepest joy. Contentment with little and restraint of inclinations, as demanding as the strictest moral philosopher could require, belonged to his plan for pleasure. By pleasure he meant a heart that is always cheerful. His main difference from the Stoics here was that he made this pleasure the motive. They refused to do so, and rightly. For, on the one hand, V:116 the virtuous Epicurus made a mistake also made today by many men of good moral disposition who have not thought deeply enough about their principles. He already assumed a virtuous disposition in the people for whom he wanted to identify the initial incentive to virtue. Indeed, an upright person cannot find himself happy A 209 unless he is first conscious of his uprightness. This is because, with that disposition, his own way of thinking would force him to reproach himself for his transgressions. Moral self-condemnation would rob him of all enjoyment of whatever comforts his situation might otherwise contain. But the question is: what first makes such a disposition possible, and such a way of thinking when assessing the worth of one's existence? Before it, the subject would have no feeling for moral worth at all. Certainly, if a person is virtuous, he will not enjoy life without being conscious of his uprightness in every action, however fortunate his physical condition may be. But to make him virtuous in the first place, before he values the moral worth of his existence so highly, can we really recommend the peace of mind that will arise from consciousness of an uprightness for which he has no feeling?

On the other hand, there is always a basis here for a mistake of illicit substitution, something like an optical illusion in our consciousness of what we do, as distinct from what we feel. Even the most experienced person cannot entirely avoid it A 210. A moral disposition necessarily includes consciousness that the will is determined directly by the law. Now consciousness of a determination of the faculty of desire is always the ground for taking satisfaction in the action it produces. But this pleasure, this satisfaction in itself, is not the action's determining ground. Instead, the direct determination of the will by reason alone is the ground of the feeling of pleasure. That determination remains a purely practical determination of the faculty of desire, not one based on feeling. Inwardly, this determination has exactly the same effect of prompting activity that would have been produced by a feeling of pleasantness expected from the desired action. So we easily mistake what we ourselves do for something we merely passively feel. We take the moral V:117 incentive for a sensible impulse. This is what usually happens in so-called illusions of the senses, here the inner sense. There is something very sublime in human nature's being determined to act directly by a pure law of reason, and even in the illusion of taking the subjective aspect of this intellectual determinability of the will for something belonging to feeling, the effect of a special sensible feeling. An intellectual feeling would be a contradiction. It is also very important to draw attention to this feature of our personality A 211 and to cultivate reason's effect on this feeling as much as possible. But we must also guard against falsely praising this moral determining ground as an incentive by supplying it with feelings of special joys as its grounds. Those feelings are only its results. Such praise would degrade and distort the real, genuine incentive, the law itself, as if by putting it against a false background. Respect, then—not pleasure or the enjoyment of happiness—is something for which no prior feeling underlying reason is possible. Such a feeling would always be sensible and dependent on our susceptibility to being affected. Respect, as consciousness of the will's direct constraint by law, is barely an analogue of the feeling of pleasure. In relation to the faculty of desire it does exactly the same thing, but from different sources. Yet only this way of representing the matter can achieve what we seek: actions performed not merely in conformity with duty, as a result of pleasant feelings, but from duty. That must be the true aim of all moral education.

But do we have a word that does not mean enjoyment, as happiness does, yet indicates satisfaction with one's existence—an analogue of happiness that must necessarily accompany consciousness of virtue? Yes: the word is self-contentment A 212. In its proper meaning it always indicates only a negative satisfaction with one's existence, in which one is conscious of needing nothing. Freedom, and consciousness of it as an ability to follow the moral law with an overriding disposition, is independence from inclinations. At least, it is independence from them as determining motives of our desire, though not as motives that affect us. Insofar as I am conscious of this freedom in following my moral maxims, it is the only source of an unchanging V:118 contentment necessarily connected with it, resting on no particular feeling. This can be called intellectual contentment. Contentment based on feeling, which is called contentment only improperly, rests on satisfying inclinations, however subtly these may be devised. It can never match what we imagine it to be. Inclinations change, grow as we indulge them, and always leave a still greater emptiness than the one we intended to fill. So they are always burdensome to a rational being. Even if it cannot put them aside, they force it to wish it were rid of them. Even an inclination toward what accords with duty—for example, toward helping others—can greatly ease the effectiveness of moral maxims, but cannot produce any. A 213 In these maxims everything must be based on the representation of the law as the determining ground, if the action is to contain not merely legality but also morality. Inclination is blind and servile, whether it is good-natured or not. Where morality is concerned, reason must not merely act as its guardian. As pure practical reason, it must attend entirely to its own interest, without regard to inclination. Even the feeling of pity and tender sympathy is burdensome to well-meaning people themselves if it comes before reflection on what duty requires and becomes the determining ground. It confuses their considered maxims and makes them wish to be rid of it and subject to lawgiving reason alone.

This helps us understand how awareness of this capacity of pure practical reason, put into action through virtue, can produce awareness of mastery over one's inclinations. That also means awareness of independence from them, and thus from the discontent that always accompanies them. It can therefore produce a negative satisfaction with one's condition, that is, contentment. At its source, this is contentment with oneself as a person. In this way, indirectly, freedom itself becomes capable of enjoyment. This cannot be called happiness, because it A 214 does not depend on the positive addition of a feeling. Strictly speaking, it cannot be called bliss either, since it does not include complete independence from inclinations and needs. Yet it resembles bliss insofar as at least the determination of one's will can remain free of their influence. So at least in its origin it is analogous to the self-sufficiency that can be attributed only to the highest being.

V:119 From this resolution of the antinomy of pure practical reason, it follows that practical principles allow us at least to think of a natural and necessary connection between consciousness of morality and the expectation of happiness proportioned to it, as its consequence. That does not yet mean that we can cognize or understand this connection. In contrast, principles for pursuing happiness cannot possibly produce morality. So morality constitutes the supreme good, as the first condition of the highest good. Happiness constitutes its second element, but only as the morally conditioned, yet necessary, consequence of morality. Only with this subordination is the highest good the whole object of pure practical reason. Reason must necessarily represent it as possible, since reason commands us to contribute everything we can to producing it. But the possibility of such a connection A 215 between what is conditioned and its condition belongs entirely to the supersensible relation of things. It cannot be given at all under the laws of the sensible world. Nevertheless, the practical consequence of this idea—the actions aimed at making the highest good real—belongs to the sensible world. We will therefore try to set out the grounds of that possibility, first with regard to what is directly within our power, and then with regard to what is not within our power. Reason offers the latter as a supplement to our inability, necessary under practical principles for making the highest good possible.

III. The primacy of pure practical reason in its connection with speculative reason.

By primacy among two or more things connected by reason, I mean the priority one has in being the first determining ground of its connection with all the rest. In the narrower, practical sense, it means the priority of the interest of one, to which the interest of the others is subordinate, while that first interest cannot be placed below any other. Each faculty of the mind can A 216 be assigned an interest: a principle containing the condition under which alone its exercise is promoted. Reason, as the faculty of principles, determines the interest of all V:120 powers of the mind, but determines its own interest for itself. The interest of its speculative use consists in cognition of the object up to the highest a priori principles. The interest of its practical use consists in determining the will with regard to the final and complete end. What is needed for any use of reason—that its principles and assertions must not contradict one another—is no part of its interest. It is the condition for having reason at all. Only extending its reach counts as its interest, not mere agreement with itself.

If practical reason may assume and think of as given nothing beyond what speculative reason can supply from its own insight, then speculative reason has primacy. But suppose practical reason had original a priori principles of its own, inseparably connected with certain theoretical positions. These positions might escape every possible insight of speculative reason, though they need not contradict it either. Then the question is which interest is supreme A 217, not which must yield, since one does not necessarily conflict with the other. Must speculative reason, which knows nothing of all that practical reason offers it to accept, take up these propositions? Must it try to unite them with its concepts as a foreign possession transferred to it, even though they go beyond its reach? Or is it entitled to stubbornly follow its own separate interest and, under Epicurus's canon, reject as empty quibbling everything that cannot establish its objective reality through clear examples presented in experience? May it do so even if these propositions are closely interwoven with the interest of the pure practical use of reason and do not themselves contradict theoretical reason? Its sole reason would be that they really do damage speculative reason's interest to this extent: they remove the limits it has set for itself and expose it to every kind of nonsense or delusion of the imagination.

Indeed, if we took practical reason as dependent on our susceptibility to feelings—that is, as merely managing the interest of inclinations under the sensible principle of happiness—we could not make this demand on speculative reason at all. Muhammad's paradise, or the melting union with the deity imagined by theosophists and mystics, each according to his taste, would force their V:121 monstrous creations on reason. It would be just as good to have no reason as A 218 to expose it to every fantasy in this way. But if pure reason can be practical on its own, and actually is, as consciousness of the moral law shows, then it is still always one and the same reason that judges by a priori principles, whether for theoretical or practical purposes. Clearly, then, even if its theoretical capacity is insufficient to establish certain propositions positively, yet they do not contradict it, reason must accept these very propositions as soon as they belong inseparably to the practical interest of pure reason. They come as a foreign offering, not grown on its own soil, but sufficiently authenticated nonetheless. It must try to compare and connect them with everything within its power as speculative reason. But it must recognize that these are not its own insights. They extend its use for another purpose, a practical one. This does not in any way conflict with its interest in restricting speculative overreach.

So when pure speculative and pure practical reason are united in one cognition, practical reason has primacy. This assumes that their connection is not contingent and optional, but grounded a priori in reason itself, and therefore necessary A 219. Without this subordination, reason would come into conflict with itself. If the two were merely placed side by side as equals, speculative reason would strictly close its boundaries and admit nothing from practical reason into its territory. Practical reason, however, would still extend its boundaries over everything and, wherever its need required, would try to include speculative reason within them. We cannot ask pure practical reason to be subordinate to speculative reason and so reverse the order. For all interest is ultimately practical. Even the interest of speculative reason is only conditional, and is complete only in practical use.

IV. V:122 The immortality of the soul as a postulate of pure practical reason.

Bringing about the highest good in the world is the necessary object of a will that can be determined by the moral law. Within this highest good, the supreme condition is complete conformity of dispositions with the moral law. So this conformity must be just as possible as its object, since it is included in the very command to A 220 promote that object. Complete conformity of the will with the moral law, however, is holiness. No rational being in the sensible world is capable of this perfection at any moment of its existence. Yet it is required as practically necessary. So it can be found only in a progress toward that complete conformity extending to infinity. Under principles of pure practical reason, we must accept such practical progress as the real object of our will.

But this infinite progress is possible only if we presuppose the existence and personality of the same rational being continuing to infinity. This is called the immortality of the soul. So the highest good is practically possible only if we presuppose the soul's immortality. Thus immortality, as inseparably connected with the moral law, is a postulate of pure practical reason. By this I mean a theoretical proposition that cannot be proved as such, insofar as it is inseparably attached to a practical law that holds unconditionally a priori.

The proposition that our nature's moral destiny is to reach complete conformity with the moral law only through progress extending to infinity is extremely useful A 221. This is true not only for the present supplement to speculative reason's incapacity, but also for religion. Without it, either we entirely strip the moral law of its holiness by contriving to make it lenient and thus suited to our convenience, or we stretch our calling and our expectations toward an unattainable destiny: the hoped-for complete achievement of holiness of will V:123. We then lose ourselves in fanatical theosophical dreams that completely contradict self-knowledge. Both approaches only obstruct our unceasing effort to follow, exactly and throughout, a strict and unyielding command of reason. That command is nevertheless not an imaginary ideal, but a real command. For a rational but finite being, only infinite progress from lower to higher degrees of moral perfection is possible. The Infinite, for whom the condition of time is nothing, sees in this series, endless for us, the whole of conformity with the moral law. The holiness that his command unrelentingly demands, in order to agree with his justice in the share of the highest good he assigns to each, is fully present in one intellectual intuition of rational beings' existence. A 222 As for the creature, all it can have concerning the hope of this share is consciousness of its tested disposition. From its past progress from worse to morally better, and from the unchanging resolve it has come to know through this progress, it can hope for further uninterrupted continuation of that progress, however far its existence may extend, even beyond this life¹². Thus it can hope to be fully adequate to God's will—not here or at any foreseeable future A 223 moment of its existence, but only in the infinity of its continued existence, which God alone can survey V:124. This involves no leniency or remission, which cannot be reconciled with justice.

V. The existence of God as a postulate of pure practical reason.

In the preceding analysis, the moral law led to a practical task prescribed by pure reason alone, without any contribution from sensible incentives. This task was the necessary completion of the first and foremost part of the highest good, morality. Since this task can be fully achieved only in an eternity, it led to the postulate of immortality. The same law must also lead to the possibility of the second element of the highest good: happiness adequate to that morality. It must do so just as unselfishly as A 224 before, from impartial reason alone, by presupposing the existence of a cause adequate to this effect. That is, it must postulate the existence of God as necessarily belonging to the possibility of the highest good, which as an object of our will is necessarily connected with pure reason's moral legislation. We want to demonstrate this connection convincingly.

Happiness is the condition of a rational being in the world for whom, throughout its existence, everything goes according to its wishes and will. It therefore rests on nature's agreement with its whole end, and also with the essential determining ground of its will. Now the moral law, as a law of freedom, commands through determining grounds that are to be entirely independent of nature and of nature's agreement with our faculty of desire as incentives. But the acting rational being in the world is not also the cause of the world and nature itself. So the moral law contains not the slightest ground for a necessary connection between morality and happiness proportioned to it in a being that belongs to the world as a part and therefore depends on it. Precisely for this reason, that being cannot through its will be the cause of nature, or by its own powers make nature agree throughout A 225 V:125 with its practical principles where its happiness is concerned. Nevertheless, pure reason's practical task—the necessary work toward the highest good—postulates such a connection as necessary. We ought to try to promote the highest good, which must therefore be possible. So we also postulate the existence of a cause of all nature, distinct from nature, which contains the ground of this connection: the exact agreement of happiness with morality. But this supreme cause must contain the ground of nature's agreement not merely with a law of rational beings' will, but with the representation of this law, insofar as they make it their will's supreme determining ground. Thus nature must agree not merely with the form of their conduct, but also with its morality as their motive—that is, with their moral disposition. So the highest good in the world is possible only if we assume a supreme cause of nature whose causality accords with moral disposition. Now a being capable of acting according to the representation of laws is an intelligence, a rational being. The causality of such a being according to this representation of laws is its will. So the supreme cause of nature, insofar as it must be presupposed for the highest good, is a being that A 226 is the cause, and thus the author, of nature through understanding and will: God. Consequently, the postulate of the possibility of the highest derived good, the best world, is also the postulate of the reality of a highest original good, namely, God's existence. Now it was our duty to promote the highest good. So we have not only permission but also a necessity, connected with duty as a need, to presuppose that this highest good is possible. Since it is possible only on the condition of God's existence, this inseparably connects the presupposition of his existence with duty. In other words, it is morally necessary to assume God's existence.

We must carefully note here that this moral necessity is subjective, a need, and not objective, itself a duty. There can be no duty to assume that a thing exists, since that concerns only the theoretical use of reason. Nor does this mean that assuming God's existence is necessary as a ground of all obligation in general. As we have sufficiently proved, that ground rests solely on the V:126 autonomy of reason itself. The only duty here is to work to produce and promote the highest good in the world. We may therefore postulate its possibility. But our reason A 227 finds that possibility thinkable only if it presupposes a supreme intelligence. So assuming the existence of this intelligence is connected with consciousness of our duty, although this assumption itself belongs to theoretical reason. With regard to theoretical reason alone, considered as a ground of explanation, it can be called a hypothesis. But with regard to the intelligibility of an object assigned to us by the moral law—the highest good—and thus to a need for practical purposes, it can be called belief, specifically pure rational belief. For pure reason alone, in both its theoretical and its practical use, is the source from which it arises.

This deduction now makes it understandable why the Greek schools could never solve their problem of the practical possibility of the highest good. They always made the rule governing the human will's use of freedom its sole and independently sufficient ground. They did not think God's existence was needed for it. They were right to establish the principle of morals on its own, independently of this postulate, from the relation of reason alone to the will. They were therefore right to make it the supreme practical condition of the highest good. But that did not make it the whole condition of its possibility. The Epicureans A 228 had accepted an entirely false principle of morals as supreme, the principle of happiness. They substituted for a law a maxim of arbitrary choice according to each person's inclination. Yet they were consistent enough in lowering their highest good in proportion to the low level of their principle. They expected no greater happiness than human prudence can acquire, including abstinence and moderation of inclinations. As we know, this happiness must turn out meager enough and vary greatly with circumstances. That is not even counting the exceptions their maxims constantly had to allow, which make them unfit to be laws. The Stoics, in contrast, quite correctly chose their supreme practical principle, virtue, as the condition of the highest good. But they represented the degree of virtue required by its pure V:127 law as fully attainable in this life. In doing so, they not only stretched the moral capacity of a human being, under the name of a sage, beyond all the limits of human nature, assuming something that contradicts all knowledge of humanity. Above all, they also refused to recognize the second component of the highest good, happiness, as a distinct object of the human faculty of desire. Instead, they made their A 229 sage, like a deity, entirely independent of nature as far as his contentment was concerned, through consciousness of his personal excellence. They exposed him to life's troubles but did not subject him to them, and also represented him as free from evil. Thus they actually omitted the second element of the highest good, personal happiness. They placed it merely in action and contentment with one's personal worth, and so included it in consciousness of a moral way of thinking. But the voice of their own nature could have refuted them well enough on this point.

The teaching of Christianity¹³, even before we regard it as religious teaching, provides on this point a concept of the highest A 230 V:128 good—the kingdom of God—that alone satisfies the strictest demand of practical A 231 reason. The moral law is holy, allowing no leniency, and requires holiness of conduct. Yet all the moral perfection a human being can reach is always only virtue: a disposition conforming to law out of respect for the law. This means consciousness of a continuing tendency to transgress, or at least of impurity: the admixture of many spurious, nonmoral motives in following the law. It therefore means self-assessment joined with humility. So, with regard to the holiness demanded by the Christian law, it leaves the creature nothing but progress to infinity. For that very reason, however, it also entitles the creature to hope for its continued existence to infinity. The worth of a disposition fully adequate to the moral law is infinite. For in the judgment of a wise and all-powerful distributor of happiness, all possible happiness has no limit except rational beings' lack of conformity with their duty. But the moral law by itself promises no happiness. Under concepts of an order of nature in general, happiness is not necessarily connected with following that law. Christian moral teaching supplies this lack—the second indispensable component of the highest good—by representing a world in which rational beings devote themselves wholeheartedly to the moral law as a kingdom of God. In it, nature A 232 and morals are brought into a harmony foreign to each of them on its own by a holy author who makes the derived highest good possible. The holiness of conduct is assigned to them as their V:129 guide already in this life. But the well-being proportioned to it, bliss, is represented as attainable only in an eternity. For holiness must always be the model of their conduct in every condition, and progress toward it is possible and necessary already in this life. Bliss, however, under the name of happiness, cannot be attained in this world as far as our own power is concerned. It is therefore made solely an object of hope. Despite this, the Christian principle of morality itself is not theological, and thus not heteronomy. It is the autonomy of pure practical reason on its own. For it makes cognition of God and his will the ground not of these laws, but only of reaching the highest good on condition of following them. It even places the real incentive to follow these laws not in their desired consequences, but in the representation of duty alone. For only in faithfully observing duty does worthiness to acquire that highest good consist.

In this way, through the concept of the A 233 highest good as the object and final end of pure practical reason, the moral law leads to religion. That is, it leads to cognition of all duties as divine commands, not as decrees—that is, arbitrary ordinances of an alien will, contingent in themselves—but as essential laws of every free will in itself. Nevertheless, they must be regarded as commands of the highest being. For only from a morally perfect will, holy and good, that is also all-powerful can we hope for the highest good. The moral law makes it our duty to set that good as the object of our efforts. So we can hope to reach it only through agreement with this will. Here too, therefore, everything remains unselfish and grounded solely on duty. Neither fear nor hope may be made underlying incentives. If they become principles, they destroy the entire moral worth of actions. The moral law commands me to make the highest possible good in a world the ultimate object of all my conduct. But I can hope to bring this about only through agreement of my will with that of a holy and good author of the world. The concept of the highest good is that of a whole in which the greatest happiness is represented as joined in the most exact proportion to the greatest degree of moral perfection possible in creatures A 234. It therefore includes V:130 my own happiness. Nevertheless, happiness is not the determining ground of the will directed to promote the highest good. The moral law is. Indeed, that law strictly limits my unlimited desire for happiness by imposing conditions on it.

So morality is not really the teaching of how we are to make ourselves happy, but of how we are to become worthy of happiness. Only when religion is added does the hope also arise of someday sharing in happiness to the extent that we have taken care not to be unworthy of it.

Someone is worthy to possess a thing or a condition when that person's possession of it agrees with the highest good. We can now easily see that all worthiness depends on moral conduct. For in the concept of the highest good, this conduct is the condition of the rest, which belongs to one's condition: a share in happiness. It follows that morality in itself must never be treated as a teaching about happiness, that is, as guidance for coming to share in happiness. It concerns only the rational condition without which happiness is impossible A 235, not a means of acquiring it. But suppose morality has been presented completely. It merely imposes duties and does not supply measures for satisfying selfish wishes. Suppose, then, that the law-based moral wish to promote the highest good—to bring the kingdom of God to us—has been awakened, a wish that could not previously arise in any selfish soul, and that the step to religion has been taken to support it. Only then can this moral teaching also be called a teaching about happiness. For only with religion does the hope for happiness first begin.

We can also see from this that if we ask about God's ultimate end in creating the world, the answer must not be the happiness of rational beings in it, but the highest good. This adds a condition to those beings' wish for happiness: worthiness to be happy, that is, the morality of those same rational beings. Morality alone provides the standard by which they can hope to share in happiness at the hand of a wise author. For wisdom, considered theoretically, means cognition of the highest good, and practically, the V:131 will's conformity with the highest good. So we cannot assign to a supreme, independent wisdom an end grounded merely on goodness. For the effect of this goodness, in A 236 regard to rational beings' happiness, can be thought adequate to the highest original good only under the limiting conditions of agreement with the holiness¹⁴ of his will. Those who placed creation's end in the glory of God therefore chose perhaps the best expression, provided we do not think of this in human terms as an inclination to be praised. For nothing honors God more than what is most valuable in the world: respect for his command and observance of the holy duty his law imposes on us, when his magnificent arrangement A 237 crowns such a beautiful order with appropriate happiness. If the latter, speaking in human terms, makes him lovable, the former makes him an object of worship. Even human beings can earn love through doing good, but never respect through that alone. So even the greatest generosity brings them honor only by being exercised according to worthiness.

It now follows by itself that in the order of ends, a human being, and with him every rational being, is an end in itself. No one, not even God, can ever use him merely as a means without his also being an end at the same time. Thus the humanity in our person must be holy to us. For a human being is the subject of the moral law, and thus of what is holy in itself V:132. Only for the sake of this law and in agreement with it can anything at all be called holy. For this moral law rests on the autonomy of his will as a free will. Under its universal laws, that will must necessarily also be able to agree to what it is to submit to.

VI. A 238 The postulates of pure practical reason in general.

They all start from the principle of morality. This is not a postulate but a law through which reason directly determines the will. Precisely because it is determined in this way, the will, as a pure will, requires these necessary conditions for following its prescription. These postulates are not theoretical dogmas, but presuppositions from a necessarily practical standpoint. They therefore do not extend speculative cognition. But through their relation to the practical, they give objective reality to the ideas of speculative reason in general. They entitle reason to concepts whose very possibility it otherwise could not presume to assert.

These postulates are those of immortality, of freedom considered positively as the causality of a being insofar as it belongs to the intelligible world, and of God's existence. The first follows from the practically necessary condition that duration be adequate to the complete fulfillment of the moral law. The second follows from the necessary presupposition of independence from the sensible world and the capacity to determine one's will according to the law of an A 239 intelligible world, that is, freedom. The third follows from the necessity of the condition that enables such an intelligible world to be the highest good, by presupposing the highest independent good, that is, God's existence.

Respect for the moral law makes it necessary to aim at the highest good. This aim, and the resulting presupposition of the highest good's objective reality, therefore lead through postulates of practical reason to concepts that speculative reason could present as problems but could not solve. Thus: 1. They lead to the problem in whose attempted solution V:133 speculative reason could commit nothing but paralogisms, namely, immortality. For speculative reason lacked the mark of permanence needed to complete the psychological concept of an ultimate subject, necessarily attributed to the soul in self-consciousness, into the real representation of a substance. Practical reason achieves this through the postulate of a duration required for conformity with the moral law in the highest good, as the whole end of practical reason. 2. They lead to what speculative reason contained only as antinomy. Its solution could rest only on a concept thinkable as a possibility, but whose objective reality speculative reason could neither prove nor determine: the cosmological idea of an intelligible world and consciousness of our A 240 existence in it. Practical reason does this through the postulate of freedom. It establishes freedom's reality through the moral law, and with it the law of an intelligible world. Speculative reason could only point to that world, not determine its concept. 3. They give meaning to what speculative reason could think, but had to leave undetermined as a mere transcendental ideal: the theological concept of the original being. This meaning is practical: the being is a condition for the possibility of the object of a will determined by that law. Through its authoritative moral legislation in an intelligible world, it is the supreme principle of the highest good in that world.

But is our cognition really extended in this way by pure practical reason? Does what was transcendent for speculative reason become immanent in practical reason? Certainly, but only for practical purposes. For this does not give us cognition of our soul's nature, the intelligible world, or the highest being as they are in themselves. We have only united the concepts of them in the practical concept of the highest good, as the object of our will. We have done so entirely a priori through pure reason, but only through the moral law and solely in relation to it, with regard to the object it commands. We gain no insight into even how freedom A 241 is possible, or how we should represent this kind of causality theoretically and positively. Only that such a causality exists is postulated through the moral law and for its sake. The same holds for the other ideas. No human understanding will ever fathom their possibility. Yet no sophistry will ever tear from even the most ordinary person's conviction the belief that they are V:134 true concepts.

VII. How is it possible to think of an extension of pure reason for practical purposes without also extending its cognition as speculative reason?

To avoid becoming too abstract, we will answer this question at once by applying it to the case before us.—To extend pure cognition practically, a purpose must be given a priori. That means an end as an object of the will, represented as practically necessary independently of all theoretical principles through an imperative that directly determines the will, a categorical imperative. Here that end is the highest good. But this is not possible without presupposing three theoretical concepts: freedom, immortality, and God. Since these are mere pure concepts of reason, no corresponding intuition can be found for them, and thus no objective reality by a theoretical A 242 route. So the practical law commanding the existence of the highest good possible in a world postulates the possibility of those objects of pure speculative reason—the objective reality that speculative reason could not secure for them. This does indeed add to pure reason's theoretical cognition. But the addition consists solely in this: those concepts, otherwise problematic for it, merely thinkable, are now positively declared to have actual objects corresponding to them. For practical reason inevitably needs their existence for the possibility of its own absolutely necessary practical object, the highest good. Theoretical reason is thereby entitled to presuppose them. But this extension of theoretical reason is no extension of speculation that would now allow positive use of them for theoretical purposes. Practical reason has accomplished nothing more here than establishing that those concepts are real and actually have their possible objects. It gives us no intuition of them, nor can one be required. So no synthetic proposition is possible through this concession of their reality. Consequently, A 243 this disclosure does not help us in the least for speculative purposes. It does help extend this cognition of ours V:135 with regard to the practical use of pure reason. The three ideas of speculative reason mentioned above are not yet cognitions in themselves. Yet they are transcendent thoughts containing nothing impossible. Now, through an apodictic practical law, they acquire objective reality as necessary conditions for the possibility of what that law commands us to make our object. That is, the law directs us to recognize that they have objects, although we cannot indicate how their concept relates to an object. This is still not cognition of these objects. For through it we can make no synthetic judgment about them, nor determine their application theoretically. So we can make no theoretical use of reason concerning them, which is what all speculative cognition of them really consists in. Nevertheless, theoretical cognition, not of these objects, but of reason in general, was extended to this extent: through the practical postulates, objects were given to those ideas. A merely problematic thought thereby first acquired objective reality. So this was not an extension of cognition of given supersensible objects. But it was an extension of theoretical reason and its A 244 cognition with regard to the supersensible in general. For reason was compelled to admit that there are such objects, though it could not determine them more closely or extend this cognition of the objects themselves. They were now given to it on practical grounds and only for practical use. Pure theoretical reason, for which all those ideas are transcendent and without an object, thus owes this addition solely to its pure practical capacity. Here the ideas become immanent and constitutive, since they are grounds of the possibility of making real the necessary object of pure practical reason, the highest good. Without this, they are transcendent and merely regulative principles of speculative reason. They do not require it to assume a new object beyond experience, but only to bring its use within experience closer to completeness. Once reason possesses this addition, however, as speculative reason it will work negatively with these ideas, really only to secure its practical use. It will not extend them but purify them. On one side, it will ward off anthropomorphism as the source of superstition V:136, or the apparent extension of those concepts through supposed experience. On the other, it will ward off fanaticism, which promises such extension through supersensible intuition or similar feelings. All these are obstacles A 245 to the practical use of pure reason. Warding them off therefore certainly belongs to extending our cognition for practical purposes. This does not contradict admitting at the same time that reason has gained absolutely nothing thereby for speculative purposes.

Every use of reason concerning an object requires pure concepts of the understanding, or categories. Without them no object can be thought. These can be applied to reason's theoretical use, that is, to cognition of this kind, only insofar as intuition, which is always sensible, is also supplied as their basis. So they serve only to represent through them an object of possible experience. But here what I would have to think through categories in order to cognize it consists of ideas of reason that cannot be given in any experience. Yet our concern here is not theoretical cognition of the objects of these ideas, but only that they have objects at all. Pure practical reason supplies this reality. Here theoretical reason need only think those objects through categories. As we have clearly shown elsewhere, this can certainly be done without requiring any intuition, sensible or supersensible. For the categories have their seat and origin in the A 246 pure understanding independently of, and before, all intuition, simply as the faculty of thinking. They always mean only an object in general, however it may be given to us. Now insofar as the categories are to be applied to those ideas, no object can be given to them in intuition. Nevertheless, an object that practical reason unquestionably supplies in the concept of the highest good sufficiently secures the reality of the concepts required for the possibility of the highest good. It secures that such an object actually exists, and that the category, as a mere form of thought, is therefore not empty here but has meaning. Yet this addition produces not the slightest extension of cognition according to theoretical principles.

V:137 Next, when these ideas of God, an intelligible world—the kingdom of God—and immortality are determined by predicates taken from our own nature, we must not regard this determination as either rendering sensible those pure ideas of reason, through anthropomorphisms, or as cognition of supersensible objects beyond our reach. For these predicates are nothing but understanding and will, considered in their relation to each other A 247 as they must be thought in the moral law, and therefore only insofar as we make a pure practical use of them. We then abstract from everything else psychologically attached to these concepts, that is, everything we observe empirically in the exercise of our faculties. For example, human understanding proceeds through concepts, so its representations are thoughts, not intuitions; these follow one another in time; and our will always carries with it a dependence of contentment on the existence of its object, and so on. These things cannot apply to the highest being in that way. So nothing remains of the concepts through which we think a pure being of understanding except precisely what is needed to make it possible to think a moral law. This is indeed cognition of God, but only in a practical relation. If we try to extend it into theoretical cognition, we arrive at an understanding of his that does not think but intuits, and a will directed toward objects whose existence does not in the least affect his contentment. I will not even mention transcendental predicates, such as a magnitude of existence, or duration, that does not occur in time, although time is our only possible means of representing existence as a magnitude. These are all attributes of which we can form no A 248 concept fit for cognition of the object. This teaches us that they can never be used for a theory of supersensible beings. So they cannot establish speculative cognition on this side at all. They restrict their use solely to putting the moral law into practice.

This last point is so evident and can be so clearly demonstrated in practice that we can confidently challenge all supposed natural theologians—a strange name¹⁵—to name even one attribute determining V:138 their object beyond merely ontological predicates, such as an attribute of understanding or will, for which we could not conclusively prove the following. Once everything A 249 anthropomorphic is removed from it, only the bare word remains for us. We cannot connect it with the slightest concept from which we might hope to extend theoretical cognition. In practical matters, however, the attributes of an understanding and a will still leave us the concept of a relation. The practical law, which determines precisely this relation of understanding to will a priori, gives that concept objective reality. Once this has happened, reality is also given to the concept of the object of a morally determined will, the highest good, and with it to the conditions of its possibility: the ideas of God, freedom, and immortality. But this is always only in relation to putting the moral law into practice, never for any speculative purpose.

After these reminders, the answer to an important question is now easy to find: does the concept of God belong to physics—and therefore also to metaphysics, which contains only physics' pure a priori principles in a general sense—or does it belong to morality? If we resort to God as author of all things to explain arrangements in nature or their changes, that is at least no physical explanation. It is always an admission that we have reached the end of our philosophy. For we are forced to assume something of which we otherwise have no concept in itself, in order to form a concept of the A 250 possibility of what we see before us. Nor can metaphysics take us from knowledge of this world to the concept of God and proof of his existence through reliable inferences. For we would have to cognize this world as the most perfect possible whole. To do that, we would have to cognize all possible worlds so we could compare them with this one. We would therefore have to be omniscient V:139 to say that this world was possible only through a God, in the sense in which we must think this concept. Moreover, cognizing this being's existence from mere concepts is absolutely impossible. Every existential proposition—a proposition saying that a being of which I form a concept exists—is synthetic. It goes beyond that concept and says more about it than was thought in the concept: that an object outside the understanding is posited as corresponding to this concept in the understanding. Clearly, no inference can establish this. So only one procedure remains for reason to reach this cognition. As pure reason, it must determine its object by starting from the supreme principle of its pure practical use. This use is in any case directed solely toward the existence of something as a consequence of reason. In reason's unavoidable task, the necessary direction of the will toward the highest good, there appears not only A 251 the necessity of assuming such an original being in relation to the possibility of this good in the world. Most remarkably, there also appears something wholly missing from reason's progress along the path of nature: a precisely determined concept of this original being. We know only a small part of this world, and can still less compare it with all possible worlds. So from its order, purposiveness, and size we may infer a wise, good, powerful author, and so on, but not his omniscience, universal goodness, omnipotence, and so on. We can readily grant that we are entitled to fill this unavoidable gap with a permissible and entirely reasonable hypothesis. If wisdom, goodness, and so on shine out in as many parts as our closer knowledge reaches, they will do so in all the rest as well. It is therefore reasonable to attribute every possible perfection to the world's author. But these are not inferences on which we can pride ourselves for our insight. They are only permissions that can be granted to us, and they still need another recommendation to justify using them. On the empirical path of physics, then, the concept of God always remains an insufficiently determined A 252 concept of the first being's perfection to be considered adequate to the concept of a deity. And metaphysics in its transcendental part can accomplish nothing at all here.

V:140 I now try to compare this concept with the object of practical reason. I find that the moral principle allows this object as possible only if we presuppose an author of the world of supreme perfection. He must be omniscient, to know my conduct down to my deepest disposition in every possible case and for all the future; omnipotent, to give it the appropriate consequences; and likewise omnipresent, eternal, and so on. So through the concept of the highest good as the object of pure practical reason, the moral law determines the concept of the original being as the highest being. The physical path of reason, or the metaphysical path that takes it further—and thus the whole speculative path—could not accomplish this. So the concept of God originally belongs not to physics, that is, to speculative reason, but to morality. The same can be said of the other concepts of reason that we discussed above as postulates of reason in its practical use.

In the history of Greek philosophy, we find no clear traces of a pure rational theology before A 253 Anaxagoras. This was not because earlier philosophers lacked the understanding and insight to rise to it through speculation, at least with the help of an entirely reasonable hypothesis. What could be easier or more natural than the thought that suggests itself to everyone: instead of various causes of the world with indefinite degrees of perfection, assume a single rational cause possessing all perfection? But the evils in the world seemed far too serious as objections for them to consider themselves entitled to such a hypothesis. So they showed understanding and insight precisely by not allowing themselves that assumption. Instead, they searched among natural causes to see whether they might find in them the character and capacity required of an original being. But this perceptive people advanced so far in their inquiries that they treated even moral subjects philosophically, subjects about which other peoples had never done more than chatter. Only then did they discover a new need, a practical one, which unfailingly gave them a determinate concept of the original being. Speculative reason could only watch. At most, it had the further merit of embellishing a concept that had not grown on its own soil. With a train A 254 V:141 of confirmations from the study of nature, which only then came forward, it promoted not the concept's authority, which was already established, but rather only its display of supposed theoretical rational insight.

These reminders will fully convince a reader of the Critique of Pure Speculative Reason how necessary that laborious deduction of the categories was, and how beneficial for theology and morality. For only through it can we prevent either of two errors. If we place the categories in the pure understanding, we might with Plato regard them as innate, and base on them overreaching presumptions in theories of the supersensible whose end cannot be seen. This would turn theology into a magic-lantern show of phantoms. If we regard them as acquired, we might with Epicurus restrict every use of them, even for practical purposes, solely to objects and determining grounds of the senses. But in that deduction, the Critique first proved that they are not empirical in origin, but have their seat and source a priori in the pure understanding. Second, it proved that, since they refer to objects in general independently of their intuition, they produce theoretical cognition only when applied A 255 to empirical objects. Yet, applied to an object given by pure practical reason, they also serve for determinate thought of the supersensible. This holds only insofar as the supersensible is determined solely by predicates necessarily belonging to the pure practical purpose given a priori and to its possibility. Restricting pure reason speculatively and extending it practically first bring it into that balanced relation in which reason in general can be used purposively. This example proves better than any other that for us human beings the path to wisdom must inevitably pass through science, if it is to be secure rather than impassable or misleading. But only after completing that science can we be convinced that it leads to this goal.

VIII. V:142 Assent arising from a need of pure reason.

A need of pure reason in its speculative use leads only to hypotheses, but a need of pure practical reason leads to A 256 postulates. For in the first case I ascend from what is derived as high in the series of grounds as I choose. I need an original ground not to give objective reality to what is derived—for example, the causal connection of things and changes in the world—but only to satisfy my inquiring reason completely concerning it. Thus I see order and purposiveness in nature before me. I do not need to move into speculation to assure myself of their reality, but only to explain them by presupposing a deity as their cause. Now an inference from an effect to a determinate cause is always uncertain and precarious, especially to a cause determined as precisely and completely as we must think God to be. So such a presupposition can never be raised beyond the level of the most reasonable opinion possible for us human beings.¹⁶ By contrast, A 257 a need of pure practical reason rests on a duty to make something, the highest good, the object of my will, in order to promote it with all my powers. But in doing so I must presuppose its possibility, and thus also its conditions: God, freedom, and immortality. For my speculative reason can neither prove nor refute these. This duty rests on a law entirely independent of these last presuppositions, a law apodictically certain in itself: the moral law. To that extent, it needs no further support V:143 from theoretical opinion about the inner constitution of things, the hidden purpose of the world's order, or a ruler presiding over it, in order to bind us completely to unconditionally lawful actions. But the subjective effect of this law—the disposition that agrees with it and that it makes necessary, to promote the practically possible highest good—at least presupposes that the highest good is possible. Otherwise it would be practically impossible to strive for the object of a concept that was fundamentally empty and without an object. Now the postulates above concern only A 258 the physical or metaphysical conditions—in a word, the conditions lying in the nature of things—for the possibility of the highest good. But they do so not for some optional speculative purpose, but for a practically necessary end of the pure rational will. Here the will does not choose but obeys an unrelenting command of reason. This command has its ground objectively in the constitution of things as they must be universally judged by pure reason. It does not rest on inclination. Inclination is in no way entitled, for the sake of what we wish on merely subjective grounds, to assume at once that the means to it are possible, or even that the object is real. So this is a need for an absolutely necessary purpose. It justifies its presupposition not merely as a permissible hypothesis, but as a postulate for practical purposes. Granted that the pure moral law unrelentingly binds everyone as a command, not as a rule of prudence, an upright person may well say: “I will that there be a God; that my existence in this world be, beyond the natural connection, also an existence in a pure world of understanding; and finally, that my duration be endless. I insist on this and will not let this belief be taken from me. For this is the one case in which my interest inevitably determines my judgment, since I may not give up any part of that interest. I pay no attention A 259 to specious arguments, however unable I may be to answer them or counter them with more plausible ones.”¹⁷

V:144 To prevent misinterpretations in using so unfamiliar a concept as pure practical rational belief, A 260 let me add one more remark.—It might almost seem that this rational belief is itself being announced here as a command: to assume that the highest good is possible. But a commanded belief is an absurdity. Recall, however, the explanation above of what the concept of the highest good requires us to assume. We will then see that assuming this possibility need not be commanded at all. No practical dispositions are needed to concede it; speculative reason must grant it without being asked. For surely no one would claim that the connection of rational beings' worthiness to be happy in the world, in conformity with the moral law, with a proportionate possession of happiness is impossible in itself. Concerning the first part of the highest good, morality, the moral law gives us only a command. Doubting the possibility of that component would amount to doubting the moral law itself. As for the second part of that object, happiness fully adequate to that worthiness, conceding its possibility in general needs no command either. Theoretical reason itself has no objection. Only the way in which we are to think of such a harmony of the laws of nature with those A 261 V:145 of freedom leaves something in respect of which we have a choice. For theoretical reason decides nothing about this with apodictic certainty. Regarding this choice, there can be a moral interest that turns the balance.

Above, I said that happiness exactly proportioned to moral worth is not to be expected from the mere course of nature in the world, and must be regarded as impossible. From this side, then, the possibility of the highest good can be conceded only if we presuppose a moral author of the world. I deliberately held back the qualification that restricts this judgment to the subjective conditions of our reason. I wanted to use it only when the kind of assent involved was to be determined more closely. In fact, this impossibility is merely subjective. That is, our reason finds it impossible for itself to make comprehensible, through the mere course of nature, such an exactly proportioned and thoroughly purposive connection between two events in the world occurring under such different laws. Yet, as with everything else purposive in nature, it also cannot prove that this connection is impossible under universal natural laws. It cannot sufficiently demonstrate this from objective A 262 grounds.

But now a different kind of ground for deciding comes into play, to turn the balance when speculative reason wavers. The command to promote the highest good is objectively grounded in practical reason. Its possibility in general is likewise objectively grounded in theoretical reason, which has no objection to it. But reason cannot objectively decide how we are to represent this possibility: under universal natural laws without a wise author presiding over nature, or only by presupposing such an author. Here a subjective condition of reason enters. This is the only way theoretically possible for it, and also the only way helpful to morality, which stands under an objective law of reason, to think the exact agreement of the kingdom of nature with the kingdom of morals as a condition for the possibility of the highest good. Now promoting that good, and therefore presupposing its possibility, is objectively necessary, though only according to practical reason. At the same time, the manner in which we choose to think it possible is up to us. In this choice, a free interest of V:146 pure practical reason decides in favor of assuming a wise author of the world. So the principle determining our judgment here is indeed A 263 subjective, as a need. But as a means of promoting what is objectively, practically necessary, it is also the ground of a maxim of assent for moral purposes: a pure practical rational belief. This belief, then, is not commanded. It springs from the moral disposition itself as a voluntary determination of our judgment to assume that existence and make it the basis of reason's further use. This determination supports the commanded moral purpose and also agrees with reason's theoretical need. So even in well-disposed people it may sometimes waver, but can never turn into unbelief.

IX. How the proportion of human cognitive faculties is wisely suited to our practical destiny.

If human nature is destined to strive for the highest good, then the measure of its cognitive faculties, especially their relation to one another, must also be assumed suitable for this end. But the Critique of Pure Speculative Reason proves reason's extreme inadequacy for solving its most important problems A 264 in a way suited to that end. It does not fail to recognize the natural hints of that same reason, which cannot be overlooked, or the great steps reason can take to approach the great goal set before it. Yet reason can never reach that goal on its own, even with the help of the greatest knowledge of nature. So nature seems here to have treated us like a neglectful stepmother in supplying us with a faculty needed for our end.

Now suppose nature had granted our wish here and given us the power of insight or enlightenment we would like to possess, or that some even imagine they actually possess. What would probably follow? Unless our whole nature were changed at the same time, the inclinations, which always speak first, would initially V:147 demand satisfaction. Combined with rational reflection, they would demand their greatest possible lasting satisfaction under the name of happiness. The moral law would speak afterward, to keep them within their proper limits and even subject them all to a higher end that takes no account of inclination. But instead of the struggle that the moral disposition now has to wage against inclinations—a struggle in which, after some defeats, the soul's moral strength can nevertheless A 265 gradually be acquired—God and eternity, with their terrifying majesty, would stand constantly before our eyes. For what we can prove completely has as much certainty for us as what we establish by seeing it. Transgression of the law would of course be avoided, and what is commanded would be done. But no command can instill the disposition from which actions ought to be done. Here the spur to activity is immediately at hand and external. Reason therefore does not first have to work its way upward, to gather strength to resist inclinations through a vivid representation of the law's dignity. So most lawful actions would be done from fear, a few from hope, and none from duty. There would be no moral worth in actions at all. Yet this alone determines the worth of the person, and even of the world, in the eyes of the highest wisdom. As long as human nature remained as it is now, human conduct would therefore be transformed into a mere mechanism. As in a puppet show, everything would move properly, but there would be no life in the figures. As things are, our situation is quite different. With all our reason's effort, we have only a very obscure and ambiguous A 266 view of the future. The world's ruler lets us conjecture his existence and glory, but not see or clearly prove them. In contrast, the moral law within us demands unselfish respect without promising or threatening anything with certainty. Only when this respect has become active and dominant, and only through it, does the law permit prospects into the kingdom of the supersensible, and even then only faint glimpses. In these circumstances a genuinely moral disposition, directly devoted to the law, can exist. The rational creature can become worthy of a share in the highest good proportioned to the moral worth of its person, not merely its actions V:148. So here too, what the study of nature and humanity otherwise sufficiently teaches us may well be right: the unfathomable wisdom through which we exist is no less worthy of reverence in what it denied us than in what it granted us.

12 Yet conviction that one's disposition is unchanging in progress toward the good seems impossible for a creature on its own. For this reason Christian religious teaching derives it only from the same spirit that brings about sanctification: this firm resolve and, with it, consciousness of perseverance in moral progress. But even in the natural way, someone conscious of having continued to progress toward the better for a long part of his life right to its end, and from genuine moral motives, may reasonably have a comforting hope, though not certainty. He may hope that he will persevere in these principles also in an existence continued beyond this life. In his own eyes he is never justified here. Nor may he ever hope to be justified as his natural perfection increases in the future, as he hopes it will, since his duties will increase with it. Nevertheless, in this progress he can have a prospect of a blessed future. Although the progress concerns a goal postponed to infinity, it counts for God as possession of that goal. For this is the expression reason uses to describe complete well-being independent of all contingent causes in the world. Like holiness, this is an idea that can be contained only in an infinite progress and its totality, and so is never fully attained by the creature.

13 People commonly think that the Christian moral prescription has no advantage in purity over the Stoics' moral concept. But the difference between the two is quite visible. The Stoic system made consciousness of strength of soul the pivot around which all moral dispositions were to turn. Its followers spoke of duties and defined them quite well. Yet they placed the incentive and the will's actual determining ground in raising one's way of thinking above the low incentives of the senses, which gain power only through weakness of soul. For them, then, virtue was a kind of heroism of the sage, who rises above human animal nature and is sufficient to himself. He prescribes duties to others but is himself above them and subject to no temptation to transgress the moral law. But they could not have done all this if they had represented that law in the purity and strictness with which the Gospel prescribes it. If by an idea I understand a perfection for which nothing adequate can be given in experience, moral ideas are not therefore beyond our reach. They are not like speculative reason's ideas, whose very concepts we cannot sufficiently determine, or for which it is uncertain whether any object corresponds to them at all. Instead, as models of practical perfection, they serve as an indispensable guide to moral conduct and also as a standard of comparison. If I now consider Christian morality from its philosophical side, it would appear as follows in comparison with the ideas of the Greek schools. The ideas of the Cynics, Epicureans, Stoics, and Christians are, respectively, natural simplicity, prudence, wisdom, and holiness. Concerning the path to reach these, the Greek philosophers differed in this way: the Cynics considered ordinary human understanding sufficient, the others only the path of science. So both considered the mere use of natural powers sufficient. Christian morality makes its prescription so pure and unyielding, as indeed it must, that it takes away people's confidence in being fully adequate to it, at least here in life. But it also restores that confidence in this way: if we act as well as is within our power, we can hope that what is not within our power will come to our aid from elsewhere, whether or not we know how. Aristotle and Plato differed only concerning the origin of our moral concepts.

14 Here, to make the distinctive character of these concepts clear, I add only this remark. We attribute various properties to God whose quality we also find appropriate to creatures, except that in God they are raised to the highest degree. For example, power, knowledge, presence, goodness, and so on become omnipotence, omniscience, omnipresence, universal goodness, and so on. Yet there are three properties attributed exclusively to God, without any added indication of magnitude, and all are moral: he is the only holy one, the only blessed one, the only wise one. For these concepts already include unlimitedness. In their order, he is therefore also the holy lawgiver and creator, the good ruler and preserver, and the just judge. These three properties contain everything through which God becomes the object of religion. The metaphysical perfections fitting them then add themselves in reason of their own accord.

15 Learning properly means only the body of historical sciences. Consequently, only a teacher of revealed theology can be called a scholar of God, a theologian. Suppose, however, we also wanted to call someone who possesses rational sciences, mathematics and philosophy, a scholar. This would already contradict the meaning of the word, which always counts as learning only what one must be taught, and thus cannot discover on one's own through reason. Even so, a philosopher with his cognition of God as a positive science might cut too poor a figure to let himself be called a scholar on that account.

16 But even here we could not invoke a need of reason if we did not have before us a problematic yet unavoidable concept of reason: that of an absolutely necessary being. This concept calls for determination. When the drive to extend cognition is added, this is the objective ground of a need of speculative reason: to determine more closely the concept of a necessary being that is to serve as the original ground of others, and so to identify that ultimate being by some characteristic. Without such prior necessary problems, there are no needs, at least none of pure reason. The others are needs of inclination.

17 In the Deutsches Museum, Feb. 1787, there is an essay by a very perceptive and clear thinker, the late Wizenmann, whose early death is regrettable. He disputes our right to infer from a need the objective reality of its object. He illustrates his point with the example of a person in love who became infatuated with an idea of beauty that was merely a creation of his imagination, and then wanted to infer that such an object actually existed somewhere. I agree with him completely in every case where the need rests on inclination. Even for the person troubled by it, inclination cannot postulate its object's existence as necessary, much less contain a demand valid for everyone. It is therefore a merely subjective ground of wishes. But here we have a need of reason, arising from an objective determining ground of the will: the moral law. This law necessarily binds every rational being. It therefore entitles us a priori to presuppose conditions in nature adequate to it, and makes those conditions inseparable from the complete practical use of reason. It is a duty to make the highest good real to the utmost of our ability. So it must also be possible. Consequently, every rational being in the world must inevitably presuppose what is necessary for its objective possibility. This presupposition is as necessary as the moral law, and it is valid only in relation to that law.