Part II — Methodology of Teleological Judgment

Appendix. A 364 416 Methodology of Teleological Judgment.

79 Whether Teleology Must Be Treated as Belonging to the Study of Nature.

Every science must have its determinate place in the encyclopedia of all sciences. If it is a philosophical science, its place must be assigned within philosophy’s theoretical or practical part; and if within the former, either in the study of nature, insofar as this considers what can be an object of experience—and thus in the study of bodies, of the soul, or of the world in general—or in theology, the study of the original ground of the world as the sum of all objects of experience.

The question now is: What place belongs to teleology? Does it belong to natural science, properly so called, or to theology? It must be one or the other; for no science can belong to the transition from one to the other, since this signifies only the articulation or organization of the system, not a place within it.

That teleology does not belong to theology as one of its parts, although A 365 theology can make the most important use of it, is clear in itself. For its objects are natural products and their cause; and although it points beyond nature to this cause as a ground situated outside and above nature—a divine author—it does so not for determining judgment but only for reflective judgment in the study of nature, so that this idea may guide our judging of things in the world, as a regulative principle suited to the human understanding.

417 Yet it seems to belong no more to natural science, which requires determining and not merely reflective principles in order to state objective grounds of natural effects. Indeed, nothing is gained for the theory of nature, or the mechanical explanation of its phenomena through their efficient causes, by considering them according to the relation of ends to one another. Setting out nature’s ends in its products, insofar as they form a system according to teleological concepts, properly belongs only to a description of nature composed under a particular guiding thread. Here reason carries out a splendidly instructive task, practically purposive in many respects, yet gives no A 366 insight into the origin and inner possibility of these forms, which is precisely what theoretical natural science is concerned with.

Teleology as a science therefore belongs to no doctrine, but only to critique, and specifically to the critique of a particular cognitive faculty: the power of judgment. But insofar as it contains a priori principles, it can and must specify the method by which nature is to be judged according to the principle of final causes. Its methodology thus has at least a negative influence on the procedure of theoretical natural science, and also on the relation that this science, within metaphysics, can have to theology as a preparation for it.

80 On the Necessary Subordination of the Principle of Mechanism to the Teleological Principle in Explaining a Thing as a Natural End.

The right to pursue a merely mechanical explanation of all natural products is in itself wholly unrestricted. But the capacity to suffice with this alone is, given the character of our understanding insofar as it deals with things as natural ends, not only very restricted but also distinctly bounded: A 367 according to a principle of judgment, the former procedure alone can accomplish nothing toward explaining the latter. Hence our judging of such products must always also be subordinated to a teleological principle.

418 It is therefore reasonable, indeed meritorious, to pursue natural mechanism in explaining nature’s products as far as this can be done with probability, and not to abandon the attempt on the ground that meeting with nature’s purposiveness along this path is impossible in itself, but only because it is impossible for us as human beings. For it would require an intuition other than sensible intuition, and a determinate cognition of nature’s intelligible substrate, from which a ground could be given even for the mechanism of appearances according to particular laws. All this wholly exceeds our capacity.

If the natural investigator is not to labor in vain, then, in judging things whose concept as natural ends is indubitably grounded—organized beings—he must always take some original organization as his basis. This organization employs that very mechanism to produce other organized forms or to develop its own into new shapes, which nevertheless always arise from that end and in accordance A 368 with it.

It is commendable to survey, through comparative anatomy, the great creation of organized natures, to see whether something resembling a system may be found in them, specifically in their principle of generation. We need not stop at the mere principle for judging them—which gives no insight into their generation—and faintheartedly renounce every claim to insight into nature in this field. The agreement of so many animal genera in a certain common scheme, which seems to underlie not only their skeletal structure but also the arrangement of their other parts, allows a ray of hope, however faint, to enter the mind: here the admirable simplicity of the basic plan has been able to produce so great a diversity of species by shortening some parts and lengthening others, enfolding these and unfolding those. Something might therefore be accomplished here through the principle of nature’s mechanism, without which there can be no natural science at all. This analogy of forms, insofar as despite all their diversity they seem to have been generated according to a common original model, strengthens the conjecture of an actual kinship among them in their generation from a common original mother. This is suggested by the gradual approximation of one animal genus A 369 419 to another, from the one in which the principle of ends seems most fully established, namely the human being, down to the polyp; from this even to mosses and lichens; and finally to the lowest level of nature perceptible to us, raw matter. From this matter and its powers, according to mechanical laws like those by which it acts in producing crystals, the whole technique of nature seems to descend—a technique so incomprehensible to us in organized beings that we believe ourselves compelled to conceive another principle to account for it.

Here the archaeologist of nature is free to derive that great family of creatures—for so we would have to represent them if the thoroughgoing kinship just described is to have a ground—from the surviving traces of nature’s earliest revolutions, according to whatever mechanism of nature he knows or conjectures. He may have the womb of the earth, just emerging from its chaotic condition, give birth, like a great animal, first to creatures of less purposive form, and these in turn to others that developed more suitably to their place of generation and their relations to one another; until this womb itself stiffened and ossified, limiting its offspring to determinate species that would no longer depart from their kind, so that their diversity A 370 remained as it had emerged at the end of that fertile formative power’s operation.—Yet for this purpose he must still ascribe to this universal mother an organization purposively arranged for all these creatures; otherwise the purposive form of the products of the animal and plant kingdoms cannot even be conceived as possible.²⁹ But then 420 he has only deferred the ground of explanation and cannot claim to have made the generation of those two kingdoms independent of the condition of A 371 final causes.

Even the change to which certain individuals of organized genera are contingently subject, if we find that their altered character is hereditary and incorporated into their reproductive power, can hardly be judged otherwise than as the occasional development of a purposive predisposition originally present in the species for the preservation of its kind. For, given the thoroughgoing inner purposiveness of an organized being, the generation of its like is very closely bound to the condition that nothing be incorporated into its reproductive power unless it also belongs, within such a system of ends, to one of the undeveloped original predispositions. If we depart from this principle, we cannot know with certainty whether several parts of the form now found in a species might not likewise be of contingent, purposeless origin. The principle of teleology—that nothing preserved in an organized being’s propagation is to be judged unpurposive—would thereby become very unreliable in application and valid solely for the original stock, which we no longer know.

Hume objects to those who find it necessary, for all such natural ends, A 372 to assume a teleological principle of judgment, that is, an architectonic understanding: one could with equal right ask how such an understanding is possible: how the manifold capacities and properties that make possible an understanding possessing executive power as well could have come together so purposively in one being. But this objection is empty. For the entire difficulty surrounding the question of the first generation of a thing that contains ends within itself and is comprehensible through them alone rests on the demand for unity in the ground connecting the manifold elements external to one another within this product. If this ground is placed in 421 the understanding of a productive cause as a simple substance, then the question, insofar as it is teleological, is sufficiently answered. But if the cause is sought merely in matter, as an aggregate of many substances external to one another, unity of principle for the internally purposive form of their formation is wholly lacking. The autocracy of matter in productions that our understanding can comprehend only as ends is a phrase without meaning.

This explains why those who seek a supreme ground of the possibility of matter’s objectively purposive forms A 373 without granting it an understanding nevertheless like to make the whole world into one all-embracing substance—pantheism—or, what is merely a more determinate explanation of the former, into a sum of many determinations inhering in one simple substance—Spinozism. Their sole aim is to obtain that condition of all purposiveness, the unity of its ground. Through the merely ontological concept of a simple substance, they do indeed satisfy one condition of the problem, namely unity in the relation to ends. But they provide nothing for the other condition, namely the relation of that substance to its consequence as an end, by which the ontological ground is to be more closely determined for the question at hand; and thus they by no means answer the whole question. It also remains absolutely unanswerable for our reason unless we represent that original ground of things as a simple substance; its property, in relation to the specific character of the natural forms grounded upon it, namely their unity of purpose, as the property of an intelligent substance; and its relation to those forms as one of causality, because of the contingency we find in everything we can conceive as possible only as an end.

81 A 374 On Joining Mechanism to the Teleological Principle in Explaining a Natural End as a Natural Product.

Just as nature’s mechanism, according to the preceding section, cannot alone suffice for conceiving the possibility of an organized 422 being, but must originally be subordinated to a cause acting intentionally, at least given the character of our cognitive faculty, so too the mere teleological ground of such a being does not suffice for considering and judging it also as a product of nature unless nature’s mechanism is joined to that ground. It serves, as it were, as the instrument of a cause acting intentionally, to whose ends nature, in its mechanical laws, is nevertheless subordinated. Our reason does not comprehend the possibility of such a union of two wholly different kinds of causality: nature in its universal lawfulness, and an idea that restricts it to a particular form for which nature in itself contains no ground. This possibility lies in nature’s supersensible substrate, of which we can determine nothing affirmatively except that it is the being in itself of which we know only the appearance. Nevertheless, the principle that everything we assume to belong to this nature (phenomenon) A 375 and to be its product must also be conceived as connected with it according to mechanical laws remains fully in force. For without this kind of causality, organized beings, as natural ends, would still not be natural products.

If, then, the teleological principle of the generation of these beings is assumed—as it must be—one can take as one’s basis either the occasionalism or the prestabilism of the cause of their internally purposive form. According to the former, on the occasion of every mating, the supreme cause of the world would directly give organic formation, in accordance with its idea, to the matter mingling in it. According to the latter, it would have placed in the initial products of its wisdom only the predisposition by means of which an organic being produces its like and the species continually preserves itself, while the loss of individuals through their nature, which also works toward their destruction, is continuously replaced. If one assumes occasionalism in the production of organized beings, all nature is entirely lost here, and with it every use of reason in judging the possibility of such products. We may therefore presume that no one with any concern for philosophy will adopt this system.

Prestabilism, in turn, can proceed in two ways. A 376 It regards every organic 423 being generated by its like either as the educt or as the product of the former. The system of generation as mere eduction is called individual preformation, or also the theory of evolution; that of generation as production is called the system of epigenesis. The latter can also be called the system of generic preformation, since the productive capacity of the progenitors, and thus the specific form, was nevertheless virtually preformed according to the inner purposive predispositions allotted to their stock. Accordingly, the opposing theory of individual preformation might better be called the theory of involution, or of encasement.

The defenders of the theory of evolution, who exempt every individual from nature’s formative power in order to have it come directly from the Creator’s hand, nevertheless did not venture to let this happen according to the hypothesis of occasionalism. On that hypothesis, mating would be a mere formality under which a supreme intelligent cause of the world had resolved each time to form an offspring directly, leaving its mother only to unfold and nourish it. They declared themselves for preformation, A 377 as though it were not all the same whether such forms arose supernaturally at the beginning of the world or during its course. Indeed, occasional creation would spare a great many supernatural arrangements otherwise required to preserve an embryo formed at the beginning of the world, unharmed by nature’s destructive powers throughout the long period before its development. It would likewise render unnecessary and purposeless an immeasurably greater number of such preformed beings than were ever to develop, and with them as many acts of creation. Yet they wished to leave at least something here to nature, so as not to lapse into a complete hyperphysics that could dispense with every natural explanation. They still clung to their hyperphysics even when they found an admirable purposiveness in monstrosities—which cannot possibly be regarded as natural ends—even if its sole aim were that an anatomist should one day take offense at it as a purposeless purposiveness and feel a disheartening wonder. But they could not possibly fit the generation of hybrids into the system of preformation. 424 They therefore had to grant to the seed of male creatures, to which they had otherwise conceded only the mechanical property of serving as the embryo’s first nourishment, a purposively formative A 378 power as well—the very power that, with regard to the entire product generated by two creatures of the same kind, they wished to allow neither.

By contrast, even if we did not know the great advantage that the defender of epigenesis possesses over the former in the experiential grounds supporting his theory, reason would already be especially favorable to his explanation. For, with respect to things that can originally be represented as possible only through the causality of ends, it nevertheless regards nature, at least in propagation, as itself productive, not merely as unfolding. With the least possible expenditure of the supernatural, it thus leaves to nature everything that follows from the first beginning, without, however, determining anything about that first beginning itself, at which physics altogether founders, whatever chain of causes it may attempt.

With regard to this theory of epigenesis, no one has contributed more both to its proof and to establishing the genuine principles of its application, in part by restricting an overbold use of it, than Court Councilor Blumenbach. He begins every physical explanation of these formations A 379 with organized matter. He rightly declares it contrary to reason that raw matter should originally have formed itself according to mechanical laws, that life should have sprung from the nature of the lifeless, and that matter should spontaneously have arranged itself into the form of a self-preserving purposiveness. Yet at the same time, under this principle of an original organization, which is inscrutable to us, he allows natural mechanism an indeterminable yet unmistakable share. To this end he calls matter’s capacity in an organized body a formative drive, as distinct from the merely mechanical formative power generally inherent in matter: a drive standing, as it were, under the higher direction and instruction of that organization.

82 425 On the Teleological System in the External Relations of Organized Beings.

By external purposiveness I understand that in which one natural thing serves another as a means to an end. Now things that have no inner purposiveness, nor presuppose any for their possibility—earth, air, water, and so on—can nevertheless be highly purposive externally, that is, in relation to other beings. But these must A 380 always be organized beings, that is, natural ends; otherwise the former could not be judged means either. Thus water, air, and earth cannot be regarded as means to the piling up of mountains, since mountains contain nothing in themselves that would require a ground of their possibility according to ends. In relation to them, therefore, their cause can never be represented under the predicate of a means useful to that end.

External purposiveness is an entirely different concept from inner purposiveness, which is bound up with the possibility of an object, regardless of whether its very actuality is an end or not. Of an organized being we can still ask: What is it there for? But we cannot readily ask this of things in which we recognize merely the effect of nature’s mechanism. For in the former we already represent a causality according to ends for their inner possibility, a creative understanding, and relate this active capacity to its determining ground, intention. There is only one external purposiveness connected with the inner purposiveness of organization that nevertheless serves, in the external relation, as a means to an end, without our having to ask for what end this organized being itself had to exist. This is the organization of the two sexes A 381 in relation to one another for the propagation of their kind. For here we can still ask, just as with an individual: Why did such a pair have to exist? The answer is: Only this pair constitutes an organizing whole, though not one organized within a single body.

If we now ask what a thing is there for, the answer is either that its existence and generation bear no relation at all to a cause acting intentionally—in which case we always 426 understand its origin to lie in nature’s mechanism—or that there is some intentional ground of its existence as a contingent natural being. We can hardly separate this latter thought from the concept of an organized thing. For once we must place beneath its inner possibility a causality of final causes and an idea underlying that causality, we can conceive the existence of this product only as an end. For a represented effect, whose representation is at the same time the determining ground of the intelligent efficient cause for producing it, is called an end. In this case, then, we can either say that the end of such a natural being’s existence lies within itself, that is, that it is not merely an end but also a final end; or that this end lies outside it in other natural beings, that is, that it exists purposively not as a final end but necessarily also A 382 as a means.

Yet if we survey the whole of nature, we find in it, as nature, no being that could claim the distinction of being the final end of creation. Indeed, we can prove a priori that whatever might still be an ultimate end for nature could never, as a natural thing, be a final end, no matter what conceivable determinations and properties we might equip it with.

When we consider the plant kingdom, the immeasurable fertility by which it spreads over almost every soil might at first lead us to regard it as a mere product of nature’s mechanism, such as nature displays in the formations of the mineral kingdom. But a closer knowledge of its indescribably wise organization prevents us from resting in this thought and prompts the question: What are these creatures there for? If we answer: For the animal kingdom, which they nourish so that it could spread across the earth in such diverse genera, the question returns: What, then, are these plant-eating animals there for? The answer might be: For the beasts of prey, which can feed only on what has life A 383. Finally comes the question: What are these, together with the preceding natural kingdoms, good for? For the human being, for the manifold use that his understanding teaches him to make of all those creatures. And he is the ultimate end of creation here on earth, because he is the only being on 427 it who can form a concept of ends and, through his reason, make a system of ends out of an aggregate of purposively formed things.

One could also take the apparently opposite path with the Knight Linnaeus and say: Plant-eating animals exist to moderate the luxuriant growth of the plant kingdom, which would otherwise choke out many of its species; beasts of prey exist to set limits to their voracity; and finally the human being exists so that, by pursuing and reducing these, he may establish a certain balance between nature’s productive and destructive powers. Thus the human being, however highly he might be valued as an end in one respect, would in another hold only the rank of a means.

If one makes objective purposiveness in the diversity of the genera of earthly creatures and in their external relations to one another, as purposively constructed beings, one’s principle, then it accords with reason to conceive in this relation a certain A 384 organization in turn, and a system of all nature’s kingdoms according to final causes. Yet here experience seems loudly to contradict reason’s maxim, especially regarding an ultimate end of nature, which is nevertheless required for the possibility of such a system and which we can place nowhere but in the human being. For with regard to him, as one among the many animal genera, nature has made not the slightest exception, in either its destructive or its generative powers, to subjecting everything to its mechanism without an end.

The first thing that would have to be intentionally arranged in an ordering of earthly natural beings into a purposive whole would presumably be their dwelling place: the ground and the element on and in which they were to thrive. Yet closer knowledge of the character of this foundation of all organic generation indicates none but wholly unintentional causes, indeed causes more destructive than favorable to generation, order, and ends. Land and sea contain not only monuments of ancient, mighty devastations that befell them and all the creatures on and in them; their entire structure, the strata of the one and the boundaries of the other, bears wholly the appearance of a product of wild, all-powerful forces of a nature working in a chaotic condition. 428 However purposively the shape, structure, and slope of A 385 the lands may now seem arranged to receive water from the air, to supply veins of springs between varied strata of earth for manifold products, and to direct the course of rivers, closer investigation nevertheless proves that they arose merely as the effect of eruptions partly of fire, partly of water, or of upheavals of the ocean. This holds both for the first production of this shape and, especially, for its later transformation together with the destruction of its first organic products.³⁰ If, then, the dwelling place, the maternal soil of the land and the maternal womb of the sea, for all these creatures indicates nothing but a wholly unintentional mechanism of its generation, how and A 386 with what right can we demand and assert a different origin for these latter products? Even if the human being was not caught up in these revolutions, as the most exact examination of the remains of those natural devastations seems, in Camper’s judgment, to prove, he is still so dependent on the earth’s other creatures that, if a mechanism of nature universally ruling over the others is admitted, he must be regarded as included under it, even though his understanding has been able, at least in large part, to save him amid their devastation.

But this argument seems to prove more than the intention for which it was proposed: not merely that the human being can be no ultimate end of nature, and for the same reason that the aggregate of organized natural things on earth cannot be a system of ends, but even that the natural products previously regarded as natural ends have no other origin than nature’s mechanism.

429 Yet in the foregoing resolution of the antinomy between the principles of the mechanical and teleological generation of organic natural beings, we saw that, regarding nature as formative according to its particular laws, for whose systematic connection we lack the key, these are merely principles of reflective judgment A 387. They do not determine the origin of these beings in itself, but say only that, given the character of our understanding and our reason, we can conceive their origin only according to final causes. We saw that the greatest possible effort, indeed boldness, in attempts to explain them mechanically is not only permitted but commanded by reason, although we know that this can never suffice, for subjective reasons arising from the particular kind and limitation of our understanding, not because a mechanism of generation would in itself contradict an origin according to ends. Finally, we saw that the supersensible principle of nature, both outside us and within us, may well contain the compatibility of both ways of representing nature’s possibility. For representation according to final causes is only a subjective condition of our use of reason when reason wishes objects to be judged not merely as appearances, but requires these appearances themselves, together with their principles, to be referred to the supersensible substrate, in order to find possible certain laws of their unity that it can represent only through ends—including ends of reason that are themselves supersensible.

83 A 388 On the Ultimate End of Nature as a Teleological System.

We have shown above that we have sufficient reason, according to principles of reason, to judge the human being not merely as a natural end, like all organized beings, but also as the ultimate end of nature here on earth, in relation to whom all other natural things constitute a system of ends. This holds, if not for determining judgment, then for reflective judgment. Now if what is to be advanced as an end through his connection with nature must be found in the human being himself, then either the end 430 must be such that nature itself, in its beneficence, can satisfy it, or it must be his aptitude and skill for all sorts of ends for which he might use nature, externally and internally. Nature’s first end would be human happiness, its second human culture.

The concept of happiness is not one the human being abstracts from his instincts and thus draws from the animality within himself. It is a mere idea of a state to which he wishes to make that animality adequate under merely empirical conditions—which is impossible. A 389 He devises this idea for himself, and in such varied ways through his understanding, entangled with imagination and the senses; indeed, he changes it so often that nature, even if wholly subject to his choice, could adopt absolutely no determinate, universal, and fixed law to agree with this wavering concept, and thus with the end each arbitrarily sets himself. But even if we either reduced this concept to genuine natural need, in which our species is in thorough agreement with itself, or, on the other hand, raised the skill for attaining imagined ends as high as we pleased, what the human being understands by happiness, and what is in fact his own ultimate natural end, not an end of freedom, would still never be attained by him. For his nature is not such as to come to rest and find satisfaction anywhere in possession and enjoyment. On the other hand, far from having adopted him as its special favorite and favored him above all animals with its beneficence, nature spares him no more than any other animal in its destructive effects: plague, famine, floods, frost, attacks by other animals, large and small, and the like. Still worse, the contradictory character of the natural predispositions within him exposes him to torments of his own devising, and to others inflicted by his own species through the oppression of A 390 domination, the barbarity of wars, and so on. He himself works, as far as he can, toward the destruction of his own species. Thus, even with the most beneficent nature outside us, nature’s end, if directed toward the happiness of our species, would not be attained in a system of nature on earth, because the nature within us is not receptive to it. He is therefore always only a link in the chain of natural ends: a principle, indeed, with respect to many an 431 end for which nature seems to have destined him in its predispositions, since he makes himself such; yet also a means for preserving purposiveness in the mechanism of the remaining links. As the only being on earth who has understanding, and thus a capacity to set himself ends of his own choice, he is indeed titular lord of nature and, if nature is regarded as a teleological system, is by his vocation its ultimate end. But this is always conditional: he must understand how, and have the will, to give nature and himself a relation to ends that, independently of nature, can be self-sufficient and thus be a final end. This final end, however, must not be sought in nature at all.

But to discover where in the human being we must place at least that ultimate end of nature, we must seek out what A 391 nature can accomplish to prepare him for what he himself must do in order to be a final end, and separate it from all those ends whose possibility rests on conditions that may be expected from nature alone. Of the latter kind is happiness on earth, by which is understood the sum of all his ends possible through nature outside and within him. This is the matter of all his ends on earth; if he makes it his entire end, it renders him incapable of setting a final end to his own existence and bringing himself into harmony with it. Thus, of all his ends in nature, only the formal, subjective condition remains: his aptitude for setting himself ends in general and, independently of nature in determining them, using nature as a means in conformity with the maxims of his free ends in general. This is what nature can accomplish with a view to the final end lying outside it, and what can therefore be regarded as its ultimate end. Producing a rational being’s aptitude for any ends whatsoever, and consequently in his freedom, is culture. Thus culture alone can be the ultimate end that we have reason to ascribe to nature with respect to the human species—not his own happiness on earth, or even merely his being the foremost instrument for establishing order A 392 and concord in the nonrational nature outside him.

But not every culture suffices for this ultimate end of nature. The culture of skill is indeed the foremost subjective condition of aptitude for advancing ends in general, yet 432 it does not suffice to advance the will in determining and choosing its ends, which nevertheless belongs essentially to the full scope of an aptitude for ends. This latter condition of aptitude, which might be called the culture of training, or discipline, is negative. It consists in freeing the will from the despotism of desires, by which we are fastened to certain natural things and rendered incapable of choosing for ourselves. For we let ourselves be fettered by the impulses that nature gave us merely as guiding threads, so that we should not neglect or even violate the vocation of animality within us, although we are free enough to tighten or loosen, lengthen or shorten them, as the ends of reason require.

Skill can hardly be developed in the human species except through inequality among people. The greater number provide life’s necessities almost mechanically, without requiring any special art, for the comfort and leisure of others who cultivate the less necessary elements of culture, science A 393 and art. By these others they are kept in a condition of oppression, bitter labor, and little enjoyment, although something of the culture of the higher classes gradually spreads to this class as well. But as culture advances—its height being called luxury when the propensity toward the dispensable begins to encroach upon the indispensable—afflictions grow equally powerful on both sides: on the one through violence from others, on the other through inner discontent. Yet this glittering misery is bound up with the development of natural predispositions in the human species, and nature’s own end, even if it is not our end, is nevertheless attained in the process. The formal condition under which alone nature can attain this final intention is that constitution of human relations in which lawful power within a whole called civil society opposes the encroachments of freedoms in mutual conflict. For only within it can the greatest development of natural predispositions occur. Yet even if people were clever enough to discover it and wise enough to submit willingly to its constraint, this would still require a cosmopolitan whole, that is, a system of all states liable to act harmfully upon one another. In its absence, and given the obstacle 433 that ambition, lust for power, and greed—especially among those A 394 holding power—place in the way of even the possibility of such a design, war is unavoidable. In war, states either split and dissolve into smaller ones, or one state unites other, smaller states with itself and strives to form a greater whole. Although war is an unintentional human endeavor, stirred by unbridled passions, it is also a deeply hidden and perhaps intentional endeavor of supreme wisdom, if not to establish, then at least to prepare, lawfulness together with the freedom of states, and thereby the unity of a morally grounded system of them. Despite the most terrible hardships it imposes on the human race, and the perhaps still greater ones with which constant readiness for it weighs upon peace, war is nevertheless one more incentive to develop to the highest degree all talents serving culture, even as the hope of a settled condition of popular happiness recedes ever further.

As for disciplining the inclinations, whose natural predisposition is entirely purposive with regard to our vocation as an animal species, but which greatly impede the development of humanity, here too, with respect to this second requirement for culture, nature displays a purposive striving toward a cultivation that makes us receptive to higher ends than nature itself can provide. The preponderance of evils that refinement of taste, even to its A 395 idealization, and luxury even in the sciences, as nourishment for vanity, pour upon us through the insatiable multitude of inclinations they generate cannot be denied. Yet neither can nature’s end be mistaken: increasingly to overcome the coarseness and vehemence of those inclinations that belong more to our animality and most oppose our cultivation for our higher vocation—the inclinations toward enjoyment—and to make room for the development of humanity. Fine art and the sciences, which through universally communicable pleasure and through polish and refinement for society make human beings, if not morally better, at least civilized, win much ground from the tyranny of sensual propensity. They thereby prepare the human being for a dominion in which reason alone is to hold power. Meanwhile the evils with which nature in part, and the unsociable selfishness of human beings in part, afflict us call forth, heighten, and temper the powers of the soul 434 so that we do not succumb to them, and thus let us feel an aptitude for higher ends hidden within us.³¹

84 A 396 On the Final End of the Existence of a World, That Is, of Creation Itself.

A final end is an end that requires no other as the condition of its possibility.

If nature’s mere mechanism is assumed as the ground for explaining its purposiveness, we cannot ask what the things in the world are there for. For on such an idealistic system the question concerns only the physical possibility of things, and to conceive them as ends would be mere reasoning without an object. Whether we interpret this form of things as chance or as blind necessity, in either case A 397 that question would be empty. But if we assume that the connection of ends in the world is real, and assume for it a particular kind of causality, namely that of a cause acting intentionally, then we cannot stop at asking why things in the world—organized beings—have this or that form, or are placed by nature in this or that relation to others. Rather, once an understanding is conceived that must be regarded as the cause of the possibility of such forms as are actually found in things, we must also ask after the 435 objective ground within that same understanding that could have determined this productive understanding to an effect of this kind. This is then the final end for which such things exist.

I said above that the final end is not an end that nature could suffice to effect and produce in accordance with its idea, because it is unconditioned. For there is nothing in nature, as a sensible being, for which the determining ground lying within nature itself would not always be conditioned in turn. This holds not only for nature outside us, material nature, but also for nature within us, thinking nature—understanding clearly that I consider within myself only what is nature. But a thing that, because of its objective character, is necessarily to exist as the final end of an intelligent cause must be such that, in the order of ends, it depends on no further condition A 398 beyond its idea alone.

Now we have only a single kind of being in the world whose causality is teleological, that is, directed toward ends, and yet is also so constituted that the law according to which these beings are to determine their ends is represented by them as unconditioned and independent of natural conditions, but necessary in itself. The being of this kind is the human being, considered, however, as noumenon: the only natural being in whom we can nevertheless recognize, from his own constitution, a supersensible capacity—freedom—and even the law of its causality together with its object, which he can set before himself as the highest end: the highest good in the world.

Of the human being, then—and thus of every rational being in the world—as a moral being, we can no longer ask for what end (quem in finem) he exists. His existence contains within itself the highest end, to which he can subject the whole of nature as far as he is able, or at least contrary to which he may not regard himself as subject to any influence of nature.—If, then, things in the world, as beings dependent in their existence, require a supreme cause acting according to ends, the human being is creation’s final end. For without him the chain of mutually subordinated ends would lack a complete foundation; and only in the human being, and even in him only as a subject of morality, A 399 is unconditional legislation with regard to ends to be found. This 436 alone makes him capable of being a final end to which the whole of nature is teleologically subordinated.³²

85 A 400 On Physicotheology.

Physicotheology is reason’s attempt to infer the supreme cause of nature and its properties from the ends of nature, which can be known only empirically. A moral theology—ethicotheology—would be the attempt to infer that cause and its properties from the moral end of rational beings in nature, which can be known a priori.

The former naturally precedes the latter. For if we wish to infer a world-cause teleologically from things in the world, natural ends must first be given, for which 437 we must then seek a final end, and for this in turn the principle of the causality of that supreme cause.

Many investigations of nature can and must be conducted according to the teleological principle without our having reason to ask after the ground of the possibility of acting purposively that we find in various products A 401 of nature. But if we also wish to have a concept of this ground, we possess absolutely no further insight than the mere maxim of reflective judgment: even if only a single organic product of nature were given to us, the character of our cognitive faculty would allow us to conceive no ground for it other than a cause of nature itself—whether of all nature or merely of this part—which contains through understanding the causality for that product. This principle of judgment brings us no further in explaining natural things and their origin, yet it opens a prospect beyond nature, so that we might perhaps determine more closely the otherwise so barren concept of an original being.

Now I say that physicotheology, however far it may be pursued, can disclose nothing to us about a final end of creation; for it does not even reach the question of such an end. It can therefore justify the concept of an intelligent world-cause as the only concept subjectively suited to the character of our cognitive faculty for the possibility of things that we can make intelligible to ourselves according to ends. But it can determine this concept no further, either theoretically or practically; and its attempt fails in its A 402 intention of founding a theology. It always remains only a physical teleology, because within it the relation to ends is, and must always be, regarded as conditioned within nature. It therefore cannot even raise the question of the end for which nature itself exists, whose ground must be sought outside nature. Yet on the determinate idea of this end depends the determinate concept of that higher intelligent world-cause, and thus the possibility of a theology.

What things in the world are useful to one another for; what the manifold within a thing is good for in relation to that thing itself; how we even have reason to assume that nothing in the world is in vain, but everything is good for something in nature, on condition that certain things are to exist as ends; and how, therefore, our reason has at its disposal for judgment 438 no other principle for the possibility of the object of its unavoidable teleological judging than to subordinate nature’s mechanism to the architectonic of an intelligent author of the world—all this the teleological contemplation of the world accomplishes splendidly and to our utmost admiration. But the data, and thus the principles, for determining that concept of an intelligent world-cause as the highest artist are merely empirical. They therefore permit no inference to properties beyond those that experience reveals in its effects. Since experience can never encompass all of nature as a system, A 403 it must often encounter grounds of proof apparently conflicting with that concept and with one another. Yet even if we could survey empirically the whole system insofar as it concerns mere nature, experience could never raise us above nature to the end of its very existence, and thereby to the determinate concept of that higher intelligence.

If we diminish the problem that physicotheology seeks to solve, its solution seems easy. Suppose we lavish the concept of a deity on any intelligent being we conceive, whether there be one or several, possessing many very great properties, though not all those required to found a nature in general agreeing with the greatest possible end. Or suppose we think nothing of supplementing through arbitrary additions what a theory’s proofs fail to provide, and where we have reason to assume much perfection—and what is much for us?—consider ourselves entitled to presuppose all possible perfection. Physical teleology then makes considerable claims to the glory of founding a theology. But if we are required to show what impels us, and moreover entitles us, to make these additions, we will seek in vain for a ground of justification in the principles of the theoretical use of reason. For this use invariably demands that, in explaining an A 404 object of experience, we ascribe to it no more properties than empirical data can be found to establish as possible. On closer examination we would see that an idea of a highest being actually underlies our thinking a priori, resting on an entirely different use of reason, the practical. This idea impels us to complete physical teleology’s deficient representation of the original ground of ends in nature, extending it to the 439 concept of a deity. We would then not falsely imagine that we had produced this idea, and with it a theology, through the theoretical use of reason in physical knowledge of the world, much less that we had proved its reality.

We cannot greatly blame the ancients for conceiving their gods as differing widely both in their powers and in their intentions and volitions, yet all, not excepting their chief, as limited in human fashion. For when they considered the arrangement and course of things in nature, they indeed found sufficient reason to assume something more than mechanical as their cause, and to suspect behind the machinery of this world the intentions of certain higher causes that they could conceive only as superhuman. But since they A 405 found good and evil, the purposive and the counterpurposive, very mixed together in the world, at least for our insight, and could not permit themselves to assume secretly underlying wise and beneficent ends, for which they saw no proof, in support of an arbitrary idea of a supremely perfect author, their judgment of the supreme world-cause could hardly turn out otherwise, insofar as they proceeded consistently according to maxims of the merely theoretical use of reason. Others, who wished to be theologians as well as physicists, thought reason would find satisfaction if they provided for the absolute unity of the principle of natural things that reason demands through the idea of a being in which, as the sole substance, all those things would be merely inherent determinations. This substance would not be the cause of the world through understanding, yet as subject it would contain all the understanding of the beings in the world. It would thus be a being that produced nothing according to ends, but in which all things, because of the unity of the subject of which they are merely determinations, would necessarily relate purposively to one another even without end or intention. They thus introduced the idealism of final causes: instead of the causal dependence on one substance, they established the elusive unity of a multitude of purposively connected substances through their inherence in one substance. A 406 This system, subsequently called pantheism when considered from the side of the inhering beings of the world, and later Spinozism from the side of the solely subsisting subject as original being, did not so much 440 solve the question of the first ground of nature’s purposiveness as declare it void. For the latter concept, stripped of all its reality, was made into a mere misinterpretation of a universal ontological concept of a thing in general.

Thus, on merely theoretical principles of the use of reason, on which alone physicotheology rests, we can never arrive at a concept of a deity sufficient for our teleological judging of nature. For either we declare all teleology a mere deception of judgment in judging the causal connection of things, and take refuge in the sole principle of nature’s mere mechanism: nature would merely seem to contain a universal relation to ends because of the unity of the substance of which it is nothing but the manifold of determinations. Or, if instead of this idealism of final causes we wish to hold to the principle of the realism of this particular kind of causality, we may place beneath natural ends many intelligent original beings or only one. But as soon as we have at hand, for grounding the concept of such a being, nothing but principles of experience drawn from the actual connection of ends A 407 in the world, we can, on the one hand, find no remedy against the discord that nature presents in many examples with respect to unity of purpose, and, on the other, never derive from this the concept of a single intelligent cause, as justified by mere experience, with sufficient determination for any theology usable in any way, theoretically or practically.

Physical teleology does indeed impel us to seek a theology, but it cannot produce one, however far we may trace nature through experience and assist the connection of ends discovered within it through ideas of reason—which, for physical problems, must be theoretical. What use is it, one may rightly complain, to place beneath all these arrangements a great understanding, immeasurable for us, and have it order this world according to intentions, if nature tells us nothing, and can never tell us anything, of the final intention? Without this we can form no common point of reference for all these natural ends, no sufficient teleological principle either for knowing all the ends within a system or for forming a 441 concept of the supreme understanding as cause of such a nature that, for our judgment reflecting teleologically upon it, A 408 could serve as a standard. I would then indeed have an artistic understanding for scattered ends, but no wisdom for a final end, which must properly contain the determining ground of that understanding. But in the absence of a final end, which only pure reason can supply a priori—since all ends in the world are empirically conditioned and can contain only what is good for this or that contingent intention, not what is good absolutely—and which alone would teach me what properties, what degree, and what relation of nature’s supreme cause I must conceive in order to judge nature as a teleological system, how and with what right may I enlarge at will my very limited concept of that original understanding, grounded in my scant knowledge of the world, of this original being’s power to make its ideas actual, of its will to do so, and so forth, and complete it into the idea of an all-wise infinite being? If this were to be done theoretically, it would presuppose omniscience in myself, so that I could comprehend nature’s ends in their entire connection and, moreover, conceive all other possible plans, in comparison with which the present one would have to be judged, with grounds, the best. For without this complete knowledge of the effect, I cannot infer a determinate concept of the supreme cause, which can be found only in that of an intelligence infinite in every A 409 respect, that is, in the concept of a deity, and establish a foundation for theology.

Thus, however far physical teleology may be extended, according to the principle stated above we may indeed say that, given the character and principles of our cognitive faculty, we can conceive nature, in the purposive arrangements that have become known to us, only as the product of an understanding to which it is subject. But whether this understanding may also have had a final intention in the whole of nature and its production—an intention that would then not lie in the nature of the sensible world—the theoretical investigation of nature can never disclose. With all our knowledge of nature, it remains undecided whether that supreme cause is its original ground according to a final end at all, rather than through an understanding determined by the mere necessity of its nature to produce certain forms 442—by analogy with what we call artistic instinct in animals—without our having to ascribe to it even wisdom on that account, much less supreme wisdom united with all the other properties required for the perfection of its product.

Physicotheology is therefore a misunderstood physical teleology, A 410 usable only as a preparation, a propaedeutic, for theology, and sufficient for this intention only through the addition of another principle on which it can rest, not in itself, as its name would suggest.

86 On Ethicotheology.

There is a judgment from which even the commonest understanding cannot free itself when it reflects on the existence of things in the world and on the existence of the world itself: all the manifold creatures, however great their artful arrangement and however diverse their purposively interrelated connections, indeed the whole of their many systems, which we incorrectly call worlds, would exist for nothing if there were no human beings—rational beings in general—in them. That is, without the human being the whole of creation would be a mere wilderness, in vain and without a final end. Yet neither is it his cognitive faculty, theoretical reason, in relation to which the existence of everything else in the world first acquires its worth, as though merely someone needed to be there who could contemplate the world. For if this contemplation of the world still represented to him nothing but things without a final end, A 411 the world’s existence could acquire no worth from being known. One must already presuppose a final end for it, in relation to which contemplating the world itself would have worth. Nor is it the feeling of pleasure and its sum in relation to which we conceive a final end of creation as given: it is not well-being, enjoyment, whether bodily or spiritual—in a word, happiness—by which we estimate that absolute worth. For the fact that, if the human being exists, he makes happiness his own final intention gives no concept of why he exists at all, or what worth he himself has, that his existence should be made agreeable to him. 443 He must therefore already be presupposed as the final end of creation if we are to have a rational ground for why nature must harmonize with his happiness when it is considered as an absolute whole according to principles of ends.—Thus it is only the faculty of desire; but not that which makes him dependent on nature through sensible impulses, not that in respect of which the worth of his existence rests on what he receives and enjoys. Rather, it is the worth that he alone can give himself, consisting in what he does, how and according to what principles he acts, not as a member of nature but in the freedom of his faculty of desire. That is, a A 412 good will alone can give his existence an absolute worth, and in relation to it the existence of the world can have a final end.

With this the commonest judgment of sound human reason also agrees completely: the human being can be a final end of creation only as a moral being, if only we direct judgment toward this question and prompt it to attempt an answer. What use is it, one will say, that this person has so much talent, that he is even very active with it and thereby exercises a useful influence on the commonwealth, and thus has great worth in relation both to his fortunate circumstances and to the benefit of others, if he possesses no good will? Considered inwardly, he is a contemptible object. And if creation is not to be altogether without a final end, then he, who as a human being also belongs to it, must nevertheless, as an evil human being in a world under moral laws, forfeit his subjective end, happiness, in accordance with these laws, as the sole condition under which his existence can coexist with the final end.

If, then, we encounter purposive arrangements in the world and, as reason unavoidably demands, subordinate the merely conditioned ends to an unconditioned highest end, that is, a final end, we first see readily that the question is no longer one of an end of nature within A 413 itself insofar as it exists, but of the end of its existence with all its arrangements, and hence of the ultimate end of creation. And in this we are properly concerned with the supreme condition under which alone a final end—that is, the determining ground of a highest understanding for producing the beings of the world—can have a place.

444 Now since we recognize the human being only as a moral being as the end of creation, we have, first, a ground, or at least the principal condition, for regarding the world as a whole connected according to ends and as a system of final causes. Above all, however, for the reference of natural ends to an intelligent world-cause that the character of our reason makes necessary for us, we have a principle for conceiving the nature and properties of this first cause as the supreme ground in the kingdom of ends, and thus for determining its concept. Physical teleology could not accomplish this: it could occasion only indeterminate concepts of that cause, and for that very reason concepts unfit for either theoretical or practical use.

From this thus-determined principle of the original being’s causality, we must conceive it not merely as an intelligence legislating for nature, but also as the legislative sovereign in a moral kingdom of ends. In relation to the highest good possible only under its rule, A 414 namely the existence of rational beings under moral laws, we must conceive this original being as omniscient, so that even the innermost dispositions—which constitute the true moral worth of the actions of rational beings in the world—are not hidden from it; as omnipotent, so that it can make all nature conform to this highest end; as all-good and at the same time just, because these two properties, united as wisdom, constitute the conditions of the causality of a supreme cause of the world as the highest good under moral laws. We must likewise conceive in it all the remaining transcendental properties, such as eternity, omnipresence, and so on, that are presupposed in relation to such a final end—for goodness and justice are moral properties.—In this way moral teleology supplies the deficiency of physical teleology and first grounds a theology. The latter, if it did not borrow unnoticed from the former but proceeded consistently, could by itself ground nothing but a demonology, incapable of any determinate concept.

But the principle of referring the world, because of the moral determination of the ends of certain beings within it, to a supreme cause A 415 as deity does this not merely by supplementing the physicoteleological ground of proof and thus necessarily taking it as a foundation. It is also sufficient by itself for this purpose and directs our attention 445 to natural ends and our investigation to the inconceivably great art hidden behind their forms, so as to give the ideas supplied by pure practical reason incidental confirmation in natural ends. For the concept of beings in the world under moral laws is an a priori principle according to which the human being must necessarily judge himself. Furthermore, if there is any world-cause acting intentionally and directed toward an end, that moral relation must be just as necessarily a condition of the possibility of creation as the relation according to physical laws—provided that intelligent cause also has a final end. Reason likewise regards this a priori as a principle necessary for its teleological judging of the existence of things. The question now is only whether we have any ground sufficient for reason, speculative or practical, to ascribe a final end to the supreme cause acting according to ends. For that this final end, given the subjective character of our reason, and however we might conceive the reason of other beings, can be none other than the A 416 human being under moral laws, may count for us as certain a priori. By contrast, the ends of nature in the physical order cannot be known a priori at all; above all, we cannot in any way see that a nature could not exist without them.

Remark.

Consider a human being in moments when his mind is attuned to moral feeling. If, surrounded by beautiful nature, he enjoys his existence in calm cheerfulness, he feels within himself a need to be grateful to someone for it. Or suppose that on another occasion, in the same frame of mind, he finds himself pressed by duties that he can and will fulfill only through voluntary sacrifice; then he feels a need to have thereby also performed something commanded and obeyed a sovereign. Or perhaps he has thoughtlessly transgressed his duty without thereby becoming accountable to other people; his stern self-reproaches will nevertheless speak within him as though they were the voice of a judge to whom he had 446 to render an account. In a word, he needs a moral intelligence so that, for the end for which he exists, there may be a being that is, in conformity with that end, the cause of him and of the world. It is vain to contrive incentives behind these feelings, for they are immediately connected with the purest moral disposition. Gratitude, obedience, and humiliation—submission to A 417 deserved chastisement—are particular states of mind directed toward duty; and the mind inclined to expand its moral disposition here freely conceives an object not in the world, so as, if possible, to demonstrate its duty toward such an object as well. It is therefore at least possible, and its ground also lies in a moral way of thinking, to represent a purely moral need for the existence of a being under whom our morality acquires either greater strength or, at least in our representation, greater scope, namely a new object for its exercise. That is, it is possible to assume a morally legislative being outside the world, without any regard to theoretical proof, still less to selfish interest, on a purely moral ground free from every alien influence, though admittedly only subjective, at the mere recommendation of pure practical reason, which legislates in its own right. Even if such a state of mind occurred rarely or did not last long, but passed fleetingly without lasting effect, or without any reflection on the object represented in such a shadowy image and without any effort to bring it under distinct concepts, its ground—the moral predisposition within us—is nevertheless unmistakable. It is a subjective principle for not resting content, in contemplating the world, with its purposiveness through natural causes, but placing beneath it a supreme cause governing nature according to moral principles.—Added to this is that the moral law impels us to strive toward a universal highest end, while we feel ourselves and all nature powerless to attain it; and that only insofar as we strive toward it may we judge ourselves in conformity with the final end of an intelligent world-cause, if A 418 there were one. Thus there is a purely moral ground of practical reason for assuming this cause, since this can be done without contradiction, if for no other reason than to avoid the danger of regarding that striving as wholly vain in its effects and thereby letting it weaken.

447 All this is intended here to say only that, although fear could first produce gods—demons—reason, through its moral principles, could first produce the concept of God. This holds even if people were, as commonly happens, very ignorant of nature’s teleology, or very doubtful because of the difficulty of reconciling its conflicting appearances through a sufficiently established principle. The inner moral determination of the end of human existence supplied what knowledge of nature lacked. For it instructed us, with regard to the final end of the existence of all things—a principle satisfying reason only when ethical—to conceive the supreme cause with properties enabling it to subordinate all nature to that single intention, for which nature is merely an instrument: that is, as a deity.

87 On the Moral Proof of the Existence of God.

There is a physical teleology that supplies a ground of proof sufficient for our theoretically reflective judgment to assume the existence of an intelligent world-cause. But we A 419 also find in ourselves, and still more in the concept of a rational being endowed with freedom in its causality in general, a moral teleology. Since the relation to ends within ourselves can be determined a priori together with its law, and thus known as necessary, this teleology requires no intelligent cause outside us to account for this inner lawfulness. Nor, when we find purposiveness in the geometrical properties of figures for all kinds of possible artistic practice, may we look beyond them to a supreme understanding imparting it. Yet this moral teleology concerns us as beings in the world, and thus as beings connected with other things in the world. Those same moral laws prescribe that we direct our judgment to these other things either as ends or as objects in relation to which we ourselves are the final end. From this moral teleology—which concerns the relation of our own causality to ends, and even to a final end that we must intend in the world, as well as the reciprocal relation of the world to that moral end and the 448 external possibility of accomplishing it, for which no physical teleology can guide us—there now arises the necessary question whether it compels our rational judgment to go beyond the world A 420 and seek an intelligent supreme principle for that relation of nature to the moral within us, so that we may represent nature as purposive also in relation to inner moral legislation and its possible realization. Consequently, there certainly is a moral teleology; and it is necessarily connected with the nomothetic of freedom on the one hand and that of nature on the other, just as civil legislation is connected with the question of where executive power is to be sought, and as there is connection generally wherever reason is to supply a principle for the actuality of a certain lawful order of things possible only according to ideas.—We will first present reason’s advance from that moral teleology, and its relation to physical teleology, to theology, and afterward consider the possibility and cogency of this mode of inference.

If one assumes the existence of certain things, or even only certain forms of things, to be contingent, and thus possible only through something else as cause, one can seek the supreme ground for this causality, and thus the unconditioned ground for the conditioned, either in the physical or in the teleological order, according to the nexus of efficient causes or that of final causes. That is, one can ask: What is the supreme productive cause? Or: What is its supreme, absolutely unconditioned end, A 421 that is, the final end of its producing these products, or all its products in general? Here, of course, it is presupposed that this cause is capable of representing ends and is therefore an intelligent being, or at least must be conceived by us as acting according to the laws of such a being.

Now if we follow the latter order, there is a principle to which even the commonest human reason is compelled to assent immediately: if there is to be any final end that reason must specify a priori, it can be none other than the human being—every rational being in the world—under moral laws.³³ For everyone judges thus: if the world consisted A 422 449 entirely of lifeless beings, or partly of living but nonrational beings, the existence of such a world would have no worth at all, since no being would exist within it that had the least concept of worth. If, on the other hand, there were also rational beings, but their reason could place the worth of the existence of things only in nature’s relation to them and their well-being, and were unable originally A 423 to procure such worth for itself through freedom, then there would indeed be relative ends in the world, but no absolute final end, since the existence of such rational beings would still always be purposeless. Moral laws, however, have the distinctive character of prescribing something to reason as an end without condition, and thus precisely as the concept of a final end requires. The existence of a reason that can itself be the supreme law in the relation to ends—in other words, the existence of rational beings under moral laws—can therefore alone be conceived as the final end 450 of a world’s existence. If this is not the case, then either no end at all in its cause underlies that existence, or ends without a final end underlie it.

The moral law, as the formal rational condition of the use of our freedom, obligates us by itself alone, without depending on any end as a material condition. Yet it also determines for us, and does so a priori, a final end that it makes us bound to pursue: the highest good in the world possible through freedom.

The subjective condition under which the human being—and, according to all our concepts, every finite rational being as well—can set himself a final end under the foregoing law is happiness. Consequently, A 424 the highest physical good possible in the world, which we are to advance as far as we can as a final end, is happiness, under the objective condition of the human being’s agreement with the law of morality, as his worthiness to be happy.

But according to all our rational capacities, we cannot possibly represent these two requirements of the final end assigned to us by the moral law as united through mere natural causes and in conformity with the idea of that final end. Thus the concept of the practical necessity of such an end through the application of our powers does not agree with the theoretical concept of the physical possibility of bringing it about, if we connect our freedom with no causality of a means other than that of nature.

Consequently, we must assume a moral world-cause, an author of the world, in order to set ourselves a final end in conformity with the moral law. And insofar as the latter is necessary, so far—that is, to the same degree and on the same ground—is it necessary to assume the former: namely, that there is a God.³⁴

This proof, which can readily be given the form of logical precision, A 425 does not mean that assuming the existence of God 451 is just as necessary as recognizing the validity of the moral law, and hence that someone unable to convince himself of the former could judge himself released from obligations under the latter. No! Only the intention to realize in the world the final end to be brought about through obedience to that law—the happiness of rational beings harmoniously coinciding with obedience to moral laws, as the highest good of the world—would then have to be abandoned. Every rational person would still have to recognize himself as strictly bound by morality’s precept; for its laws are formal and command unconditionally, without regard to ends as the matter of willing. But one requirement of the final end, as practical reason prescribes it to beings in the world, is an irresistible end implanted in them by their nature as A 426 finite beings. Reason wishes only that this end be subjected to the moral law as an inviolable condition, or also made universal in accordance with it, and thus makes the advancement of happiness in agreement with morality the final end. The moral law commands us to advance this as far as lies within our power, so far as those former ends are concerned, whatever the outcome of the effort may be. Fulfillment of duty consists in the form of an earnest will, not in the intervening causes of success.

Suppose, then, that a person, moved partly by the weakness of all the much-praised speculative arguments and partly by various irregularities he encounters in nature and the moral world, persuaded himself of the proposition that there is no God. He would nevertheless be worthless in his own eyes if he therefore wished to regard the laws of duty as merely imagined, invalid, and nonbinding, and resolved to transgress them without scruple. Such a person, with that cast of mind, would remain worthless even if he could subsequently become convinced of what he had initially doubted, although he fulfilled his duty out of fear or desire for reward, without a 452 disposition of reverence for duty, as punctiliously in its effects as could ever A 427 be demanded. Conversely, if, as a believer, he follows those laws sincerely and unselfishly to his own consciousness, yet whenever he supposes for the sake of experiment that he might one day be convinced there is no God, immediately believes himself free of every moral obligation, then his inner moral disposition must be in a poor state indeed.

We can therefore suppose an upright man, such as Spinoza, who holds himself firmly persuaded that there is no God and, since the consequence is the same regarding the object of morality, no future life either. How will he judge his own inner determination of ends through the moral law, which he actively reveres? He desires no advantage for himself from following it, either in this world or another; rather, he wishes unselfishly only to establish the good toward which that holy law directs all his powers. But his endeavor is limited. From nature he may indeed expect occasional contingent assistance, but never a lawful agreement, occurring according to constant rules, as his maxims inwardly are and must be, with the end he nevertheless feels bound and impelled to bring about. Deceit, violence, and envy will always flourish around him, though A 428 he himself is honest, peaceable, and benevolent. And the other upright people he encounters, despite all their worthiness to be happy, will nevertheless be subjected by nature, which takes no account of this, to all the evils of want, disease, and untimely death, like the earth’s other animals. They will remain so until a vast grave swallows them all, honest or dishonest—it makes no difference here—and casts them, who could believe themselves the final end of creation, back into the abyss of matter’s purposeless chaos from which they were drawn.—Thus this well-disposed man would indeed have to abandon as impossible the end that he had, and ought to have had, before his eyes in following moral laws. Or, if he wishes to remain faithful here too to the call of his inner moral vocation, and not weaken the respect that the moral law immediately inspires in him for obedience by the nullity of the sole ideal final end adequate to its lofty demand—which cannot happen without injury to his moral disposition—⁴⁵³ he must—and can very well do so, since it is at least not contradictory in itself—assume for practical purposes, that is, so as at least to form a concept of the possibility of the final end morally prescribed to him, the existence of a moral author of the world, that is, A 429 God.

88 Limitation of the Validity of the Moral Proof.

Pure reason as a practical faculty—that is, as the capacity to determine the free use of our causality through ideas, pure concepts of reason—not only contains in the moral law a regulative principle of our actions, but thereby also supplies a subjectively constitutive principle in the concept of an object that reason alone can conceive and that is to be made actual in the world through our actions according to that law. The idea of a final end in the use of freedom according to moral laws therefore has subjective-practical reality. Reason determines us a priori to advance with all our powers the greatest good of the world, which consists in uniting the greatest well-being of rational beings in the world with the supreme condition of goodness in them: universal happiness with morality most fully conforming to law. In this final end, the possibility of one part, happiness, is empirically conditioned, that is, dependent on the character of nature—whether it agrees with this end or not—and theoretically A 430 problematic. The other part, morality, with respect to which we are independent of nature’s cooperation, is established a priori as possible and is dogmatically certain. Thus the objective theoretical reality of the concept of the final end of rational beings in the world requires not only that we have a final end set before us a priori, but also that creation, that is, the world itself, have a final end in its existence. If this could be proved a priori, it would add objective reality to the subjective reality of the final end. For if creation has a final end at all, we can conceive it only as necessarily agreeing with the moral end, which alone makes the concept of an end possible. 454 Now we do indeed find ends in the world, and physical teleology displays them so abundantly that, if we judge in accordance with reason, we ultimately have grounds to assume, as a principle for investigating nature, that nothing in nature is without an end. But we seek nature’s final end within nature itself in vain. Hence, just as its idea lies only in reason, this end can and must be sought, even as regards its objective possibility, only in rational beings. Their practical reason, however, not only supplies this final end but also determines its concept with respect to the conditions under which alone a final end of creation can A 431 be conceived by us.

The question now is whether the objective reality of the concept of a final end of creation can also be demonstrated sufficiently for the theoretical demands of pure reason: if not apodictically for determining judgment, then sufficiently for the maxims of theoretically reflective judgment. This is the least one can ask of speculative philosophy, which undertakes to connect the moral end with natural ends through the idea of a single end. Yet even this little is far more than it can ever accomplish.

According to the principle of theoretically reflective judgment, we would say: If we have grounds to assume, for nature’s purposive products, a supreme cause of nature whose causality in respect of nature’s actuality—creation—must be conceived as different in kind from that required for nature’s mechanism, namely as the causality of an understanding, then we will also have sufficient grounds to conceive in this original being not merely ends everywhere in nature but also a final end. This would not demonstrate the existence of such a being, but would at least, as occurred in physical teleology, convince us that we can make the possibility A 432 of such a world comprehensible to ourselves not merely according to ends, but only by placing a final end beneath its existence as well.

But a final end is solely a concept of our practical reason and can neither be inferred from any data of experience for the theoretical judging of nature nor be referred to knowledge of nature. No use of this concept is possible except solely for practical 455 reason according to moral laws. The final end of creation is that constitution of the world which agrees with what alone we can determinately specify according to laws, namely the final end of our pure practical reason, insofar, indeed, as it is to be practical.—Now through the moral law, which imposes this latter end upon us, we have a ground for practical purposes, namely for applying our powers to realizing it, to assume its possibility and practicability, and hence also a nature of things agreeing with it. For without nature’s cooperation in a condition of that end which is not in our power, its realization would be impossible. Thus we have a moral ground for conceiving in a world also a final end of creation.

This is not yet the inference from moral teleology A 433 to a theology, that is, to the existence of a moral author of the world, but only to a final end of creation determined in this way. That for this creation—that is, for the existence of things in conformity with a final end—we must assume as author of the world, first, an intelligent being, but, second, not merely an intelligent being, as required for the possibility of natural things we were compelled to judge as ends, but one that is also moral, and thus a God: this is a second inference. Its character makes clear that it is drawn solely for judgment according to concepts of practical reason and, as such, for reflective, not determining judgment. For we cannot claim to see that, although morally practical reason in us differs essentially in its principles from technically practical reason, this must also be so in the supreme world-cause, if assumed as an intelligence, and that a particular and distinct kind of its causality is required for a final end, as against mere natural ends. Hence we cannot claim to see that in our final end we have not merely a moral ground for assuming a final end of creation as effect, but also for assuming a moral being as creation’s original ground. We can, however, say that given the character of our rational faculty, A 434 we cannot at all make comprehensible to ourselves the possibility of such purposiveness related to the moral law and its object as is contained in this final end, without an author and ruler of the world who is also a moral legislator.

456 The actuality of a highest morally legislative author is therefore sufficiently demonstrated solely for the practical use of our reason, without determining anything theoretically regarding its existence. For the possibility of its end—which is already assigned to us in any case through its own legislation—reason requires an idea that removes, sufficiently for reflective judgment, the obstacle arising from our inability to pursue that end on the merely natural concept of the world. This idea thereby acquires practical reality, even though speculative cognition wholly lacks every means of giving it such reality for theoretical purposes in explaining nature and determining the supreme cause. For theoretically reflective judgment, physical teleology sufficiently proved an intelligent world-cause from natural ends; for practical judgment, moral teleology accomplishes this through the concept of a final end that it is compelled, for practical purposes, to ascribe to creation. The A 435 objective reality of the idea of God as moral author of the world cannot indeed be demonstrated through physical ends alone. Yet when knowledge of these ends is united with knowledge of the moral end, the former are of great importance, by virtue of pure reason’s maxim of pursuing unity of principles as far as possible, in assisting the practical reality of that idea through the reality it already has for judgment in theoretical respects.

To prevent a readily arising misunderstanding, it is most necessary here to note, first, that we can conceive these properties of the highest being only by analogy. For how could we investigate its nature, to which experience can show us nothing similar? Second, through analogy we can only conceive it, not cognize it accordingly and ascribe these properties to it theoretically. For that would be for determining judgment in our reason’s speculative intention, in order to understand what the supreme world-cause is in itself. Here, however, the sole concern is what concept we must form of it, given the character of our cognitive faculties, and whether we must assume its existence in order to give likewise only practical reality to an end that pure practical reason, without any such presupposition, imposes upon us a priori to realize with all our powers: A 436 that is, merely to be able to conceive an intended effect as possible. That concept may well be transcendent 457 for speculative reason; the properties we ascribe to the being conceived through it may also conceal an anthropomorphism if used objectively. Yet their use is not intended to determine its nature, which is inaccessible to us, but to determine ourselves and our will accordingly. We name a cause according to the concept we have of its effect, but only in respect of its relation to that effect, without thereby wishing to determine its inner constitution through properties that alone are known to us in such causes and must be given through experience. Thus, for example, we ascribe to the soul, among other things, a power of locomotion, because movements of the body actually arise whose cause lies in its representations, without wishing thereby to ascribe to it the only way we know moving forces: through attraction, pressure, impact, and thus motion, which always presupposes an extended being.—In just this way we must assume something containing the ground of the possibility and practical reality, that is, the practicability, of a necessary moral final end. According to the character of the effect expected from it, we can conceive this as a wise being governing the world according to moral A 437 laws; and, in conformity with the character of our cognitive faculties, we must conceive it as a cause of things distinct from nature, merely to express the relation of this being, which surpasses all our cognitive faculties, to the object of our practical reason. We do not thereby wish to ascribe to it theoretically the only causality of this kind known to us, namely understanding and will, nor even objectively to distinguish the causality conceived in it with respect to what is a final end for us from its causality with respect to nature and its determinations of ends generally. We can assume this distinction only as subjectively necessary given the character of our cognitive faculty, and valid for reflective, not objectively determining judgment. But where practice is concerned, such a regulative principle for prudence or wisdom—to act in conformity with what, given our cognitive faculties, we can conceive as possible only in a certain way, as an end—is also constitutive, that is, practically determining. Yet this very same principle, as a principle for judging the objective possibility of things, is by no means theoretically determining, as though 458 the object itself possessed only the kind of possibility that belongs to our A 438 capacity for thinking it; it is a merely regulative principle for reflective judgment.

Remark.

This moral proof is not a newly invented ground of proof, but at most only a newly elucidated one. For it already lay within the human rational faculty before its earliest germination and is only developed further as that faculty’s cultivation proceeds. As soon as human beings began to reflect on right and wrong, at a time when they still passed indifferently over nature’s purposiveness and used nature without thinking of anything but its customary course, the judgment inevitably had to arise that in the outcome it could never be all the same whether a person had behaved honestly or falsely, justly or violently, even if to the end of his life he had encountered, at least visibly, no good fortune for his virtues or punishment for his crimes. It is as if they perceived within themselves a voice saying that things must turn out differently. Thus there must also have lain hidden a representation, however obscure, of something toward which they felt bound to strive and with which such an outcome could not be reconciled; or, if they regarded the course of the world as the sole order of things, they could not reconcile with it that inner determination of their mind’s ends. They might represent the way in which such an irregularity—an irregularity far more revolting to the human mind than the blind chance one might wish to place beneath the judging of nature as a principle—could be balanced out in any number of crude A 439 ways. Yet they could never devise any other principle for the possibility of uniting nature with their inner moral law than a supreme cause governing the world according to moral laws. For a final end assigned within them as duty, and a nature outside them without any final end, in which that end is nevertheless to become actual, stand in contradiction. They might brood over much nonsense concerning the inner constitution of that world-cause; yet that moral relation in the government of the world always remained the same, universally comprehensible to even the least cultivated reason insofar as it considers itself practical, whereas speculative reason 459 cannot nearly keep pace with it.—In all probability, this moral interest also first awakened attention to nature’s beauty and ends. These then served excellently to strengthen that idea, yet could neither ground it nor, still less, dispense with that interest. For even the investigation of natural ends acquires only in relation to the final end that immediate interest which manifests itself so strongly in admiration for them without regard to any advantage to be drawn from them.

89 On the Usefulness of the Moral Argument.

Restricting reason, with respect to all our ideas of the supersensible, to the conditions of its practical use has, regarding A 440 the idea of God, this unmistakable benefit: it prevents theology from soaring into theosophy, into transcendent concepts that bewilder reason, or sinking into demonology, an anthropomorphic representation of the highest being; and it prevents religion from lapsing into theurgy, the extravagant delusion that we can feel other supersensible beings and in turn influence them, or into idolatry, the superstitious delusion that we can make ourselves pleasing to the highest being through means other than a moral disposition.³⁵

For if we allow the vanity or presumption of reasoning to determine even the least thing theoretically, and thus to expand knowledge, about what lies beyond the sensible world; if we permit it to make a great show of insights into the existence and constitution of the A 441 divine nature, into its understanding and will, the laws of both, and the properties flowing from them into the world, then I would like to know where and at what point we are to limit 460 reason’s pretensions. For from the very source from which those insights were drawn, still more may surely be expected, if only, as one supposes, we exert our reflection. Such claims would have to be limited according to some principle, not merely because we find that all attempts with them have hitherto failed, since this proves nothing against the possibility of a better outcome. But here no principle is possible except either to assume that absolutely nothing concerning the supersensible can be determined theoretically, save in a merely negative way, or to assume that our reason contains a still unused mine of who knows how vast an expansion of knowledge, held in reserve for us and our descendants.—As for religion, however, that is, morality in relation to God as legislator, if theoretical knowledge of him had to precede it, morality would have to conform to theology. Not only would the external, arbitrary legislation of a supreme being be introduced in place of reason’s internal, necessary legislation, but everything deficient in our insight into A 442 that being’s nature would also extend to the moral precept, thus making religion immoral and perverting it.

Regarding the hope of a future life, suppose that, instead of taking as our guide for reason’s judgment of our vocation the final end that we ourselves must fulfill according to the moral law’s precept—a judgment therefore regarded as necessary or worthy of acceptance only in a practical relation—we consult our theoretical cognitive faculty. In this regard, the study of the soul, like theology above, supplies nothing more than a negative concept of our thinking being: none of its actions or the appearances of inner sense can be explained materialistically. Thus, through our entire theoretical cognitive faculty, no judgment expanding or determining knowledge of its separate nature, or of the continuance or noncontinuance of its personality after death, is possible for us on speculative grounds. Since everything here is therefore left to the teleological judging of our existence in a practically necessary respect, and to assuming our continued existence as the condition required for the final end that reason unconditionally assigns to us, a benefit becomes apparent, though at first it seems a loss. Just as theology can never become theosophy A 443 461 for us, rational psychology can never become pneumatology as a science expanding knowledge, while on the other hand it is also secured against falling into materialism. Rather, it is merely an anthropology of inner sense, that is, knowledge of our thinking self in life, and as theoretical knowledge remains merely empirical. Rational psychology, by contrast, as regards the question of our eternal existence, is no theoretical science at all, but rests on a single inference of moral teleology, just as its entire use is necessary solely for the sake of the latter as our practical vocation.

90 On the Kind of Assent in a Teleological Proof of the Existence of God.

First, every proof—whether conducted through direct empirical presentation of what is to be proved, as in proof through observation of the object or experiment, or through reason a priori from principles—requires that it not persuade but convince, or at least work toward conviction. That is, the ground of proof or inference must not be merely a subjective A 444 aesthetic determining ground of assent, mere semblance, but must be objectively valid and a logical ground of knowledge; otherwise the understanding is beguiled, not convinced. Of the former kind of apparent proof is one conducted in natural theology perhaps with good intentions, yet with deliberate concealment of its weakness: one invokes the great multitude of evidences for an origin of natural things according to the principle of ends, and exploits a merely subjective ground of human reason. This is reason’s characteristic propensity, wherever it can do so without contradiction, to substitute a single principle for many; and where only some or even many requirements for determining a concept are found in this principle, to add the rest in thought, so as to complete the concept of the thing through arbitrary supplementation. For indeed, if we encounter so many products in nature indicating to us an intelligent cause, why should we not prefer to conceive one cause instead of many, and in it not merely great understanding, 462 power, and so on, but rather all-wisdom and omnipotence—in a word, conceive it as containing the sufficient ground of such properties for all possible things? And why not ascribe to this single all-powerful original being not merely, for natural laws and products, A 445 understanding, but also, as a moral world-cause, supreme moral practical reason, since this completion of the concept supplies a principle sufficient both for insight into nature and for moral wisdom, and no even somewhat well-grounded objection can be raised against the possibility of such an idea? If at the same time the mind’s moral incentives are set in motion and a lively interest in them is added with rhetorical force, of which they are indeed worthy, there arises a persuasion of the proof’s objective sufficiency and an illusion—salutary in most instances of its use—that dispenses entirely with any examination of its logical rigor and even feels abhorrence and aversion toward such examination, as though it rested on impious doubt.—There may be nothing to object to here insofar as popular usefulness is the proper concern. Yet dividing the argument into its two heterogeneous parts, what belongs to physical and what to moral teleology, cannot and must not be prevented. Fusing them conceals where the proof’s true strength lies, and in which part and how it would have to be worked upon to sustain its validity before the strictest scrutiny, A 446 even if we should be compelled to acknowledge the weakness of our rational insight in one part. Thus it is the philosopher’s duty—even supposing he made nothing of the demand for sincerity—to expose the illusion, however salutary, that such a mingling can produce, and to separate what belongs merely to persuasion from what leads to conviction. These are determinations of assent differing not merely in degree but in kind. In this way he can openly present the mind’s attitude in this proof in all its purity and submit it candidly to the strictest examination.

But a proof aimed at conviction can in turn be of two kinds. It may seek to establish what the object is in itself, or what it is for us, for human beings in general, according to the 463 rational principles necessary for our judging of it: a proof κατ' αληθειαν or κατ' ανθρωπον, the latter word taken in its general meaning for humanity as such. In the former case it rests on sufficient principles for determining judgment; in the second, merely for reflective judgment. In the latter case, if it rests on merely theoretical principles, it can never work toward conviction. But if A 447 it takes a practical principle of reason as its basis, which thus holds universally and necessarily, it may well claim conviction sufficient for a purely practical purpose, that is, moral conviction. A proof works toward conviction without yet convincing when it is merely conducted on the path toward it: that is, it contains only objective grounds for conviction which, though not yet sufficient for certainty, are nevertheless of a kind that does not merely serve persuasion as subjective grounds of judgment.

Now all theoretical grounds of proof suffice either: 1) for proof through logically rigorous inferences of reason; or, failing this, 2) for an inference by analogy; or, if this too is unavailable, still 3) for probable opinion; or finally, at the very least, 4) for assuming a merely possible explanatory ground as a hypothesis.—Now I say that no grounds of proof whatsoever working toward theoretical conviction can produce assent of this kind, from its highest to its lowest degree, if what is to be proved is the proposition that an original being exists as a God in the meaning adequate to the whole content of this concept: namely, as a moral author of the world, and thus in such a way that through him the final end of creation is also specified. A 448

1) As for the logically correct proof proceeding from the universal to the particular, the Critique has sufficiently shown the following. Since no intuition possible for us corresponds to the concept of a being to be sought beyond nature, and this concept itself, insofar as it is to be determined theoretically through synthetic predicates, therefore always remains problematic for us, absolutely no cognition of this being is possible that would expand the scope of our theoretical knowledge in the least. Nor can the particular concept of a supersensible being be subsumed under the universal principles of the nature of things so that we might infer 464 the former from the latter, because those principles hold solely for nature as an object of the senses.

2) With two things differing in kind, one can indeed conceive one of them, precisely at the point of their difference, according to an analogy³⁶ with the other. But one cannot, from that in which they differ, A 449 infer from one to the other by analogy, that is, transfer to the other this mark A 450 of specific difference. Thus, by analogy with the law of equality of action and reaction in the mutual attraction and repulsion of 465 bodies, I can also conceive the community of members of a commonwealth according to rules of right. But I cannot transfer those specific determinations—material attraction or repulsion—to this community and ascribe them to its citizens in order to constitute a system called a state.—Likewise we may conceive the causality of the original being with respect to things in the world as natural ends by analogy with an understanding as the ground of the forms of certain products we call works of art. For this is done only for the theoretical or practical use that, according to a certain principle, our cognitive faculty must make of this concept regarding natural things in the world. But from the fact that among A 451 beings in the world understanding must be ascribed to the cause of an effect judged artificial, we can by no means infer by analogy that a being wholly distinct from nature possesses, with respect to nature itself, the very same causality that we perceive in the human being. For this concerns precisely the point of difference conceived, in its very concept, between a cause sensibly conditioned with regard to its effects and the supersensible original being; it therefore cannot be transferred to the latter.—The very requirement that I conceive divine causality only by analogy with an understanding—a faculty we know in no being other than the sensibly conditioned human being—contains the prohibition against ascribing understanding to it in the literal sense.³⁷

3) Opinion has no place at all in a priori judgments; through them one either knows something as wholly certain or knows nothing at all. But even if the given grounds of proof from which we begin are empirical, as here with the ends in the world, we still cannot A 452 form any opinion by their means beyond the sensible world or grant such venturesome judgments the slightest claim to probability. For probability is a part of a certainty possible in a given series of grounds: its grounds are compared with the sufficient ground as parts with a whole. The insufficient 466 ground must be capable of being supplemented to reach this certainty. But since, as determining grounds of the certainty of one and the same judgment, they must be homogeneous—otherwise they would not together make up a magnitude such as certainty—one part cannot lie within the limits of possible experience and another outside all possible experience. Hence, since merely empirical grounds of proof lead to nothing supersensible, and the deficiency in their series cannot be supplied by anything, there is not the slightest approach to the supersensible in attempting to attain it and knowledge of it through them. Consequently, neither is there any probability in a judgment concerning the supersensible through arguments drawn from experience.

4) Whatever is to serve as a hypothesis for explaining the possibility of a given appearance must at least be wholly certain in its possibility. It is enough that, with a hypothesis, I renounce knowledge of actuality, which is still asserted in an opinion represented as probable. A 453 I cannot surrender more. The possibility of what I place at the basis of an explanation must at least be beyond doubt; otherwise empty imaginings would have no end. But to assume the possibility of a supersensible being determined according to certain concepts, when none of the required conditions of cognition, insofar as it rests on intuition, is given, and thus only the principle of contradiction remains as a criterion of this possibility—a principle that can prove only the possibility of thinking, not of the object thought—would be a wholly groundless presupposition.

The result is that absolutely no proof for theoretical purposes of the existence of the original being as a deity, or of the soul as an immortal spirit, is possible for human reason that could produce even the slightest degree of assent. The reason is entirely comprehensible: we have no material at all for determining ideas of the supersensible, since we would have to take that material from things in the sensible world, whereas such material is absolutely unsuited to that object. Thus, without any determination of these ideas, nothing remains but the concept of a nonsensible something containing the ultimate ground of the sensible world, A 454 which does not yet constitute cognition, as an expansion of the concept, of its inner constitution.

91 467 On the Kind of Assent through a Practical Faith.

If we consider only the way in which something can be an object of knowledge for us, according to the subjective character of our powers of representation—a knowable thing—then concepts are compared not with objects, but merely with our cognitive faculties and the use these can make of the given representation for theoretical or practical purposes. The question whether something is a knowable being or not concerns not the possibility of things themselves, but that of our knowledge of them.

Knowable things are of three kinds: matters of opinion (opinabile), matters of fact (scibile), and matters of faith (mere credibile).

1) Objects of mere ideas of reason, which cannot be presented for theoretical cognition in any possible experience, are to that extent not knowable things at all; hence we cannot even hold an opinion concerning them. Indeed, to hold an opinion a priori A 455 is already absurd in itself and is the direct path to sheer imaginings. Thus either our a priori proposition is certain or it contains nothing at all to which we can assent. Matters of opinion are therefore always objects of a cognition through experience that is at least possible in itself—objects of the sensible world—but impossible for us merely because of the degree of this capacity that we possess. Thus the ether of modern physicists, an elastic fluid permeating all other matter and intimately mixed with it, is a mere matter of opinion; yet it is still of such a kind that, if the external senses were sharpened to the highest degree, it could be perceived, though it can never be presented in any observation or experiment. To assume rational inhabitants of other planets is a matter of opinion. For if we could approach those planets, which is possible in itself, we would determine through experience whether such inhabitants exist or not; but we shall never approach them so closely, and so it remains opinion. Yet to hold the opinion that there are pure spirits thinking without bodies in the material universe—if we rightly reject certain actual appearances offered as evidence—is to invent fictions and is 468 no matter of opinion at all. It is a mere idea remaining when we remove everything material from a thinking being yet leave it its thinking. But whether the latter, which A 456 we know only in the human being, that is, in connection with a body, would then remain, we cannot determine. Such a thing is a being fabricated by reasoning (ens rationis ratiocinantis), not a being of reason (ens rationis ratiocinatae). Of the latter it is nevertheless possible sufficiently to demonstrate the objective reality of its concept, at least for the practical use of reason, because this use, which has its own distinctive and apodictically certain a priori principles, even requires—postulates—it.

2) Objects for concepts whose objective reality can be proved—whether through pure reason or experience, and in the former case from theoretical or practical data of reason, but in every case by means of a corresponding intuition—are matters of fact (res facti).³⁸ Such are the mathematical properties of magnitudes in geometry, because they are capable of an a priori presentation for the theoretical use of reason. Furthermore, A 457 things or their properties that can be demonstrated through experience, our own or that of others through testimony, are likewise matters of fact.—But, remarkably, even an idea of reason is found among matters of fact, though in itself it is incapable of presentation in intuition and hence of any theoretical proof of its possibility. This is the idea of freedom, whose reality as a particular kind of causality—a concept that would be transcendent in theoretical respects—can be demonstrated through practical laws of pure reason and, in conformity with them, in actual actions, and thus in experience.—It is the only one among all the ideas of pure reason whose object is a matter of fact and must be counted among the scibilia.

469 3) Objects that must be conceived a priori in relation to the dutiful use of pure practical reason, whether as consequences or as grounds, but are transcendent for its theoretical use, are mere matters of faith. Such is the highest good in the world to be brought about through freedom. Its concept cannot be sufficiently proved in its objective reality in any experience possible for us, and thus for the theoretical use of reason; yet its use for realizing that end as well as possible is commanded by pure practical reason, and it must therefore A 458 be assumed possible. This commanded effect, together with the only conditions of its possibility conceivable for us, namely the existence of God and the immortality of the soul, are matters of faith (res fidei), indeed the only objects among all objects that can be so called.³⁹ For although we must believe what we can learn only from the experience of others through testimony, it is not therefore in itself a matter of faith. For one of those witnesses it was still his own experience and a matter of fact, or is presupposed as such. Moreover, it must be possible to attain knowledge by this path of historical belief. The objects of history and geography, like everything in general that it is at least possible to know given the character of our cognitive faculties, belong not to matters of faith but to matters of fact. Only objects of pure reason can possibly be matters of faith, A 459 but not as objects of merely pure speculative reason; for as such they cannot even be counted with certainty among things, that is, objects of knowledge possible for us. They are ideas, concepts whose objective reality cannot be secured theoretically. By contrast, the highest final end to be realized by us, that through which alone we can become worthy of being ourselves the final end of a creation, is an idea having objective reality for us in practical respects, and is a thing. But because we cannot give this concept 470 that reality for theoretical purposes, it is a mere matter of faith of pure reason, and with it so too are God and immortality, as the conditions under which alone, given the character of our human reason, we can conceive the possibility of that effect of the lawful use of our freedom. Assent in matters of faith, however, is assent for purely practical purposes: a moral faith that proves nothing for theoretical knowledge, but only for pure rational knowledge that is practical and directed toward following one’s duties. It expands neither speculation nor practical rules of prudence according to the principle of self-love. If the supreme principle of all moral laws is a postulate, then the possibility of their highest object, and hence also the condition A 460 under which we can conceive that possibility, is thereby postulated along with it. Knowledge of this condition becomes neither knowing nor opinion concerning the existence and constitution of these conditions as theoretical modes of cognition, but merely an assumption in a practical, and indeed commanded, relation for the moral use of our reason.

Even if we could apparently ground a determinate concept of an intelligent world-cause on the natural ends that physical teleology presents to us in such abundance, the existence of this being would still not be a matter of faith. For since it is assumed not for fulfilling my duty but merely for explaining nature, it would be only the opinion and hypothesis best suited to our reason. Now that teleology by no means leads to a determinate concept of God. Such a concept is found only in that of a moral author of the world, because this alone specifies the final end to which we can count ourselves as belonging only insofar as we act in conformity with what the moral law imposes on us as a final end and thus obligates us to pursue. Consequently, only through its relation to the object of our duty, as a condition of the possibility of attaining duty’s final end, does the concept of God acquire the privilege of counting in our assent as a matter of faith. Yet that very same A 461 concept cannot establish its object as a matter of fact. For although the necessity of duty is quite clear to practical reason, attaining its final end, insofar as this is not wholly in our power, is assumed only for the practical use of 471 reason and is therefore not practically necessary in the way duty itself is.⁴⁰

Faith, as a settled disposition (habitus) rather than an act (actus), is reason’s moral way of thinking A 462 in assenting to what is inaccessible to theoretical cognition. It is therefore the mind’s abiding principle of accepting as true, because of our obligation to that end, what must necessarily be presupposed as a condition for the possibility of the highest moral final end,⁴¹ although we can comprehend neither its possibility 472 nor, equally, its impossibility. A 463 Faith, so called without qualification, is trust in attaining an intention whose advancement is duty, but whose practicability we cannot comprehend, and consequently neither can we comprehend that of the only conditions conceivable for us. Thus faith relating to particular objects that are not objects of possible knowledge or opinion—in which latter case, especially the historical, it would have to be called credulity rather than faith—is entirely moral. It is a free assent, not to that for which dogmatic proofs can be found for theoretically determining judgment, nor to that to which we regard ourselves as bound, but to what we assume for the sake of an intention according to laws of freedom. Yet it is not, like an opinion, without sufficient ground, but is grounded in reason, though only with respect to its practical use, A 464 sufficiently for reason’s intention. For without it the moral way of thinking has no firm constancy when confronted with theoretical reason’s demand for proof of the possibility of morality’s object, but wavers between practical commands and theoretical doubts. To be incredulous is to adhere to the maxim of not believing testimony in general; but an unbeliever is one who denies those ideas of reason all validity because they lack a theoretical grounding of their reality. He therefore judges dogmatically. But dogmatic unbelief cannot coexist with a moral maxim governing one’s way of thinking, for reason cannot command us to pursue an end recognized as nothing but a figment. A doubting faith, however, can: the absence of conviction through grounds of speculative reason is merely an obstacle for it. A critical insight into the limits of speculative reason can deprive that obstacle of its influence on conduct and 473 offer in its place an overriding practical assent.

If we wish to introduce another principle in place of certain failed attempts in philosophy and secure influence for it, it is a great satisfaction to understand how and why those A 465 attempts had to fail.

God, freedom, and immortality of the soul are the problems toward whose solution all the preparations of metaphysics are directed, as its ultimate and sole end. Now it was believed that the doctrine of freedom was necessary only as a negative condition for practical philosophy, whereas the doctrine of God and of the soul’s constitution belonged to theoretical philosophy and had to be demonstrated separately and by itself. Both were then to be connected with what the moral law—which is possible only under the condition of freedom—commands, and a religion thus brought about. But we can readily see that these attempts had to fail. For from mere ontological concepts of things in general, or of the existence of a necessary being, we can form absolutely no concept of an original being determined by predicates that can be given in experience and thus could serve cognition. And the concept grounded on experience of nature’s physical purposiveness could in turn supply no proof sufficient for morality, and thus for knowledge of a God. Nor could knowledge of the soul through experience, which we acquire only in this life, furnish a concept of its spiritual, immortal nature, and thus one sufficient for morality. Theology and pneumatology, as A 466 problems for the sciences of speculative reason, cannot be brought about through any empirical data or predicates, because their concept transcends all our cognitive faculties.—Both concepts, that of God and that of the soul with regard to its immortality, can be determined only through predicates which, although themselves possible only from a supersensible ground, must nevertheless prove their reality in experience. For only thus can they make cognition of wholly supersensible beings possible.—Such is the unique concept found in human reason 474 of human freedom under moral laws, together with the final end that reason prescribes through those laws. The former are suited to ascribe to the author of nature, and the latter to the human being, those properties containing the necessary condition for the possibility of both. Thus from this very idea we can infer the existence and constitution of those beings otherwise wholly hidden from us.

Thus the reason why the intention of proving God and immortality fails on the merely theoretical path is that no cognition of the supersensible is possible by this path of natural concepts. Its success, by contrast, on the moral path of the concept of freedom A 467 rests on this: here the supersensible lying at the foundation—freedom—not only supplies material for knowledge of the other supersensible objects, the moral final end and the conditions of its practicability, through a determinate law of causality arising from it, but also demonstrates its reality as a matter of fact in actions. Yet for that very reason it can provide no ground of proof valid except for practical purposes, which are also the only purposes religion requires.

It always remains highly remarkable here that among the three pure ideas of reason—God, freedom, and immortality—the idea of freedom is the sole concept of the supersensible that proves its objective reality, by means of the causality conceived in it, in nature through its effect possible there. Precisely thereby it makes possible the connection of the other two with nature, and of all three with one another into a religion. Thus we have within ourselves a principle capable of determining the idea of the supersensible within us, and thereby also that of the supersensible outside us, into cognition, although possible only for practical purposes—a task at which merely speculative philosophy, which could give only a negative concept even of freedom, had to despair. Hence the concept of freedom, as the fundamental concept A 468 of all unconditionally practical laws, can expand reason beyond those limits within which every theoretical concept of nature would have to remain confined without hope.

29 A hypothesis of this kind may be called a daring adventure of reason; and there may be few even among the most acute natural investigators to whom it has not sometimes occurred. For it is not actually absurd, like equivocal generation, by which is understood the production of an organized being through the mechanics of raw, unorganized matter. It would still be univocal generation in the most general meaning of the word, insofar as something organic were generated from something else organic, though specifically distinct from it among beings of this kind: for example, if certain aquatic animals gradually developed into marsh animals, and these, after several generations, into land animals. A priori, in the judgment of reason alone, this is not self-contradictory. But experience shows no example of it. According to experience, rather, all generation known to us is homonymous generation, not merely univocal as opposed to generation from unorganized material, but also producing something homogeneous in its very organization with its progenitor. Heteronymous generation is nowhere encountered as far as our experiential knowledge of nature extends.

30 If the already adopted name natural history is to remain in use for the description of nature, then what the former literally indicates—a representation of the earth’s former, ancient condition, about which we venture conjectures with good reason even though we cannot hope for certainty—may be called the archaeology of nature, in contrast to that of art. Petrifactions would belong to the former, as engraved stones and so on belong to the latter. For since work on such an inquiry, under the name of a theory of the earth, is in fact continually carried on, though slowly, as is proper, this name would not be given to a merely imagined investigation of nature, but to one to which nature itself invites and summons us.

31 What worth life has for us, if assessed merely according to what we enjoy—the natural end of the sum of all inclinations, happiness—is easy to decide. It sinks below zero. For who would wish to begin life anew under the same conditions, or even according to a new, self-devised plan, though one conforming to nature’s course, that likewise aimed merely at enjoyment? What worth life has according to what it contains when led in accordance with nature’s end for us, consisting in what we do, not merely enjoy, though we are still always only means to an undetermined final end, has been shown above. Thus nothing seems to remain but the worth that we ourselves give our life through what we not only do, but do purposively in such independence from nature that even nature’s existence can be an end only under this condition.

32 It would be possible for the happiness of rational beings in the world to be an end of nature, and then it would also be its ultimate end. At least we cannot comprehend a priori why nature should not be so arranged, since this effect might well be possible through its mechanism, at least as far as we can see. But morality, and a causality according to ends subordinated to it, are absolutely impossible through natural causes. For the principle determining them to action is supersensible; it is therefore the only thing possible in the order of ends that is absolutely unconditioned with respect to nature, and thereby alone qualifies its subject as the final end of creation, to which all nature is subordinated.—Happiness, by contrast, as shown in the preceding section on the testimony of experience, is not even an end of nature with respect to human beings in preference to other creatures, far less a final end of creation. Human beings may certainly make it their ultimate subjective end. But when I ask after the final end of creation—Why did human beings have to exist?—the question concerns an objective highest end, such as supreme reason would require for its creation. If we answer: So that beings may exist to whom that supreme cause can do good, we contradict the condition to which human reason itself subjects its most heartfelt wish for happiness, namely agreement with its own inner moral legislation. This proves that happiness can only be a conditioned end, and thus that the human being can be creation’s final end only as a moral being. As for his condition, happiness is connected with it only as a consequence proportionate to his agreement with that end as the end of his existence.

33 I deliberately say: under moral laws. It is not the human being according to moral laws, that is, one who behaves in conformity with them, who is creation’s final end. For by the latter expression we would say more than we know: namely, that an author of the world has the power to make the human being behave at all times in conformity with moral laws. This presupposes a concept of freedom and of nature—of the latter alone can we conceive an external author—that would have to contain an insight into nature’s supersensible substrate and its identity with what makes causality through freedom possible in the world, an insight far beyond our rational comprehension. Only of the human being under moral laws can we say, without exceeding the limits of our insight, that his existence constitutes the world’s final end. This also agrees completely with the judgment of human reason reflecting morally on the world’s course. We believe we perceive traces of a wise relation to ends even in evil, if only we see that the impious villain does not die before suffering the well-deserved punishment of his misdeeds. According to our concepts of free causality, good or evil conduct rests with us; but we place the highest wisdom in the world’s government in this: that the occasion for the former, and the outcome for both, is ordained according to moral laws. In the latter properly consists the glory of God, which theologians therefore not inappropriately call the ultimate end of creation.—It must further be noted that when we use the word creation, we understand nothing but what has been said here, namely the cause of the existence of a world, or of the things in it, its substances, as the proper concept of the word also implies: the actualization of substance is creation. Thus it does not already carry with it the presupposition of a freely acting, and consequently intelligent cause, whose existence we are only now seeking to prove.

34 This moral argument is not intended to supply an objectively valid proof of God’s existence, or to prove to the doubting believer that there is a God, but to show that, if he wishes to think consistently in moral matters, he must incorporate the acceptance of this proposition among the maxims of his practical reason.—Nor is this meant to say that assuming the happiness of all rational beings in the world in accordance with their morality is necessary for morality, but that it is necessary through morality. It is therefore a subjectively sufficient argument for moral beings.

35 Idolatry in a practical sense is still any religion that conceives the highest being with properties according to which something other than morality could be the condition, adequate in itself, for conforming to his will in what the human being can do. For however purely and free of sensible images that concept may have been framed theoretically, it is then nevertheless represented in practice as an idol, that is, anthropomorphically with respect to the character of its will.

36 Analogy, in a qualitative sense, is identity of the relation between grounds and consequences, causes and effects, insofar as it obtains despite the specific difference of the things or of those properties in themselves that contain the ground of similar consequences, that is, considered apart from this relation. Thus, comparing animals’ artful actions with those of human beings, we conceive the ground of these effects in animals, which we do not know, as an analogue of reason by comparison with the ground of similar human effects—reason—which we do know. We thereby also mean to indicate that the ground of the animal’s artistic capacity, called instinct, is in fact specifically different from reason, yet bears a similar relation to its effect, comparing the beaver’s building with human building.—But from the fact that the human being needs reason for building I cannot infer that the beaver must have it too and call this an inference by analogy. Yet from the similar mode of action of animals, whose ground we cannot immediately perceive, compared with that of human beings, of which we are immediately conscious, we can quite correctly infer by analogy that animals too act according to representations and are not machines, as Descartes holds; and that despite their specific difference they are nevertheless generically identical with human beings as living beings. The principle authorizing such an inference lies in the identity of the ground for counting animals, with regard to the determination in question, within the same genus as the human being as human, insofar as we compare them externally through their actions. The rationale is equal. Likewise, in comparing the purposive products of the supreme world-cause in the world with human works of art, I can conceive its causality by analogy with an understanding, but cannot infer these properties in it by analogy. For here precisely the principle making such an inference possible is lacking: namely the equality of rationale for counting the highest being and the human being within one and the same genus with respect to their respective causalities. The causality of beings in the world, which is always sensibly conditioned, as is causality through understanding, cannot be transferred to a being sharing with them no generic concept beyond that of a thing in general.

37 Nothing whatever is thereby lacking in the representation of this being’s relations to the world, with regard to either the theoretical or practical consequences of this concept. To wish to investigate what it is in itself is a curiosity as purposeless as it is vain.

38 Here I extend the concept of a matter of fact beyond the word’s customary meaning, rightly, as it seems to me. For when the relation of things to our cognitive faculties is at issue, it is not necessary, indeed not even feasible, to restrict this expression merely to actual experience, since a merely possible experience already suffices for speaking of them simply as objects of a determinate kind of cognition.

39 But matters of faith are not therefore articles of faith, if by the latter we understand matters of faith to whose profession, inward or outward, one can be obligated. Natural theology therefore contains none of this kind. For since, as matters of faith, they cannot be grounded on theoretical proofs as matters of fact can, assent to them is free and compatible with the subject’s morality only as such.

40 The final end that the moral law imposes on us to advance is not the ground of duty; for this lies in the moral law, which as a formal practical principle guides categorically, regardless of objects of the faculty of desire—the matter of willing—and thus regardless of any end. This formal character of my actions, their subordination to the principle of universal validity, in which alone their inner moral worth consists, is wholly within our power. I can very well abstract from the possibility or impracticability of the ends I am bound to advance according to that law, since only the external worth of my actions consists in them, as from something never wholly in my power, and attend only to what is my doing. Yet the intention to advance the final end of all rational beings—happiness insofar as it is possible in harmony with duty—is imposed by that very law of duty. But speculative reason has no insight into its practicability, either through our own physical capacity or through nature’s cooperation. Rather, as far as we can reasonably judge, it must regard the expectation of such a result of our good conduct from such causes, from nature alone within and outside us, without assuming God and immortality, as groundless and vain, however well intended. And if it could have complete certainty of this judgment, it would have to regard the moral law itself, in practical respects, as a mere deception of our reason. But since speculative reason is fully convinced that the latter can never happen, while those ideas whose object lies beyond nature can be conceived without contradiction, it must, for the sake of its own practical law and the task thereby imposed, thus in moral respects, recognize those ideas as real, so as not to contradict itself.

41 It is trust in the promise of the moral law, not, however, a promise contained within that law, but one I place in it, and indeed on a morally sufficient ground. For no law of reason can command a final end without reason’s also promising its attainability, though uncertain, and thereby authorizing assent to the sole conditions under which our reason can conceive it. The word faith already expresses this. It may seem questionable only how this expression and this particular idea enter moral philosophy, since they were first introduced with Christianity, and their adoption might perhaps appear merely a flattering imitation of its language. But this is not the only instance in which this wondrous religion, in the greatest simplicity of its presentation, has enriched philosophy with far more determinate and purer concepts of morality than philosophy had hitherto been able to supply. Once they exist, however, reason freely approves and accepts them as concepts it could and should have reached and introduced on its own.

Appendix. A 364 416 Methodology of Teleological Judgment.

79 Whether teleology must be treated as belonging to the study of nature.

Every science must have its definite place in the complete system of sciences. If it is a philosophical science, it must be assigned a place in philosophy's theoretical or practical part. If its place is in the former, it must belong either to the study of nature, insofar as this considers what can be an object of experience—and thus to the study of bodies, the study of the soul, or general cosmology—or to theology, the study of the original ground of the world as the totality of all objects of experience.

The question now is: what place belongs to teleology? Does it belong to natural science, properly so called, or to theology? It must be one of the two. No science can belong to the transition from one to the other, since this means only the articulation or organization of the system, not a place within it.

It is clear in itself that teleology does not belong to theology as one of its parts, although A 365 theology can make the most important use of it. Teleology concerns natural productions and their cause. Although it points beyond this to a ground outside and above nature—a divine author—it does so not for determining judgment, but only for reflective judgment in considering nature. Its purpose is simply to guide our judgment of things in the world through such an idea, as a regulative principle suited to the human understanding.

417 But teleology seems no more to belong to natural science. That science needs determining principles, not merely reflective ones, to give objective grounds for natural effects. In fact, considering natural phenomena in terms of the relation of ends to one another contributes nothing to the theory of nature, or to the mechanical explanation of its phenomena through their efficient causes. Identifying nature's ends in its products, insofar as they form a system according to teleological concepts, really belongs only to a description of nature drawn up along a particular guiding thread. Here reason carries out a splendidly instructive task, one that serves practical purposes in many ways. But it gives no A 366 insight into the origin and inner possibility of these forms, which is the real concern of theoretical natural science.

Teleology as a science therefore belongs to no doctrine, but only to critique: specifically, the critique of a particular cognitive power, the power of judgment. But insofar as it contains a priori principles, it can and must state the method for judging nature according to the principle of final causes. Its methodology thus has at least a negative influence on the procedure of theoretical natural science. It also influences the relation that natural science, within metaphysics, can have to theology as a preparation for it.

80 On the necessary subordination of the principle of mechanism to the teleological principle in explaining a thing as a natural end.

The right to pursue a purely mechanical explanation of all natural products is in itself entirely unlimited. But given the character of our understanding, when it deals with things as natural ends, its ability to make do with this alone is not just very limited. It also has a definite boundary. Namely, A 367 according to a principle of judgment, the former procedure alone can do nothing to explain the latter. Our judgment of such products must therefore always be subordinated at the same time to a teleological principle.

418 It is therefore reasonable, indeed worthwhile, to follow nature's mechanism as far as we plausibly can in explaining natural products. We should not abandon this attempt on the grounds that it is in itself impossible to encounter nature's purposiveness along this route, but only because it is impossible for us as human beings. For it would require an intuition other than a sensible one, and a determinate knowledge of nature's intelligible substrate. From this substrate we could even give a ground for the mechanism of appearances according to particular laws. All this completely exceeds our powers.

So, to avoid working entirely in vain, the investigator of nature must always base judgments of things whose concept as natural ends is undoubtedly well grounded—organized beings—on some original organization. This organization itself uses that mechanism to produce other organized forms, or to develop its own form into new shapes. But these shapes always arise from that end and in keeping with it A 368.

It is admirable to survey the great creation of organized natures through comparative anatomy, to see whether something like a system can be found in them, specifically in terms of the principle of their generation. We need not stop at the mere principle for judging them, which gives no insight into their generation, and lose heart, giving up every claim to insight into nature in this field. So many animal kinds agree in a certain common plan, which seems to underlie not only their bone structure but also the arrangement of their other parts. Through shortening some parts and lengthening others, enclosing these and unfolding those, the remarkable simplicity of the basic design could have produced such a great variety of species. This lets a ray of hope, however faint, enter the mind: perhaps the principle of nature's mechanism, without which there can be no natural science at all, may accomplish something here. Despite all their differences, these forms seem to have been produced according to a common original pattern. This analogy strengthens the suspicion that they are really related through generation from one common original mother. The step-by-step approach of one animal kind A 369 419 to another extends from the one in which the principle of ends seems most fully confirmed, namely the human being, down to the polyp; from that even to mosses and lichens; and finally to the lowest level of nature perceptible to us, raw matter. From this matter and its forces, acting according to mechanical laws like those by which it produces crystals, the whole technique of nature seems to descend. Yet in organized beings this technique is so incomprehensible to us that we believe we must think of another principle to account for it.

Here the archaeologist of nature is free to trace the origin of that great family of creatures from the remaining traces of nature's earliest upheavals, using every mechanism of nature known to him or conjectured by him. For we would have to picture them as such a family if the continuous relationship just mentioned is to have a ground. He may picture the earth's womb, just emerging from its chaotic state, as a great animal, so to speak. It initially gives birth to creatures with less purposive forms; these in turn give birth to others that develop forms better suited to their place of generation and their relations to one another. Eventually this womb itself stiffens and ossifies, restricting its offspring to definite species that no longer depart from their type. The diversity would then A 370 remain just as it stood when that fertile formative power finished its work.—Still, for this purpose he must attribute to this universal mother an organization purposively arranged for all these creatures. Otherwise the possibility of the purposive form of the products of the animal and plant kingdoms cannot be thought at all.²⁹ But then 420 he has only pushed the ground of explanation further back. He cannot claim to have made the generation of those two kingdoms independent of the condition of A 371 final causes.

Even changes that particular individuals of organized kinds happen to undergo must be judged in this way. If we find that their altered character is inherited and incorporated into their reproductive power, we can properly judge it only as the development, on a particular occasion, of a purposive predisposition originally present in the species for preserving the kind itself. For an organized being's production of its own kind, given its thoroughgoing inner purposiveness, is closely bound to this condition: nothing is incorporated into reproductive power unless, within such a system of ends, it also belongs to one of the undeveloped original predispositions. If we abandon this principle, we cannot know with certainty whether several features of a species' present form might likewise have originated accidentally and without an end. The principle of teleology—that in an organized being we should judge nothing preserved through its reproduction to be unpurposive—would then become very unreliable in application. It would hold only for the original stock, which we no longer know.

Hume objects to those who think it necessary, for all such natural ends, A 372 to assume a teleological principle of judgment, that is, an architectonic understanding. He says we could ask with equal right how such an understanding is possible: how the various powers and properties that make possible an understanding that also has the power to carry out its plans could have come together so purposively in one being. But this objection is invalid. The whole difficulty surrounding the question of the first production of a thing that contains ends within itself and is understandable only through them rests on asking about the unity of the ground connecting the manifold elements external to one another in this product. If this ground is placed in 421 the understanding of a producing cause as a simple substance, that question, insofar as it is teleological, is sufficiently answered. But if the cause is sought merely in matter, as an aggregate of many substances external to one another, there is no unified principle at all for the internally purposive form of their arrangement. Matter's autocracy in productions that our understanding can comprehend only as ends is a phrase without meaning.

That is why those who seek a supreme ground of the possibility of matter's objectively purposive forms A 373 without granting it an understanding nevertheless like to turn the universe into a single, all-encompassing substance—pantheism. Or they turn it into a totality of many determinations inhering in a single simple substance—Spinozism, which is only a more specific explanation of the former. Their sole purpose is to obtain that condition of all purposiveness, the unity of its ground. By the merely ontological concept of a simple substance, they satisfy one condition of the problem: unity in the relation to ends. But they offer nothing for the other condition: the substance's relation to its consequence as an end, by which that ontological ground must be more closely determined for the question at hand. So they do not answer the whole question at all. Indeed, it remains completely unanswerable for our reason unless we represent that original ground of things as a simple substance, and its property corresponding to the specific character of the natural forms grounded in it—their unity of ends—as the property of an intelligent substance. We must also represent its relation to those forms as one of causality, because of the contingency we find in everything we can think possible only as an end.

81 A 374 On joining mechanism to the teleological principle in explaining a natural end as a natural product.

As the preceding section showed, nature's mechanism alone cannot enable us to think of the possibility of an organized 422 being. Rather, at least given the character of our cognitive power, it must originally be subordinated to an intentionally acting cause. In the same way, the merely teleological ground of such a being is insufficient for regarding and judging it also as a product of nature. Nature's mechanism must be joined to that ground, as if it were the tool of an intentionally acting cause, to whose ends nature, in its mechanical laws, is nevertheless subordinated. Our reason does not comprehend the possibility of such a union of two entirely different kinds of causality: nature in its universal lawfulness, and an idea that restricts nature to a particular form for which it contains no ground by itself. This possibility lies in nature's supersensible substrate. We can determine nothing positive about this beyond saying that it is the being in itself, of which we know only the appearance. Nevertheless, the following principle remains in force: everything we assume to belong to this nature (phenomenon) A 375 and to be its product must also be thought as connected with nature according to mechanical laws. Without this kind of causality, organized beings, as natural ends, would still not be natural products.

Now if we accept the teleological principle for the generation of these beings—as we cannot avoid doing—we can base the cause of their internally purposive form either on occasionalism or on prestabilism. According to the former, on the occasion of every act of mating, the supreme cause of the world would directly give organic formation to the matter mingling in it, in keeping with its idea. According to the second, this cause would have placed in the first products of its wisdom only a predisposition through which an organic being produces its own kind, and the species continually preserves itself. In the same way, the loss of individuals, whose nature also works toward their destruction, would be continuously replaced. If we accept occasionalism in the production of organized beings, all nature in this process is completely lost, and with it every use of reason in judging the possibility of such products. We may therefore assume that no one with any concern for philosophy will accept this system.

Prestabilism, in turn, can proceed in two ways. A 376 It regards each organic 423 being generated by its own kind either as something drawn out of the parent or as the parent's product. The system that treats generations as merely drawing out what is already there is called individual preformation, or the theory of evolution. The system that treats generations as products is called epigenesis. The latter can also be called the system of generic preformation. For the productive power of the parents, and therefore the specific form, was still preformed virtually, according to the inner purposive predispositions given to their stock. Accordingly, the opposing theory of individual preformation might better be called the theory of involution, or of nesting.

The supporters of the theory of evolution exempt each individual from nature's formative power, making it come directly from the Creator's hand. Yet they did not dare to let this happen according to the hypothesis of occasionalism. On that hypothesis, mating would be a mere formality under which a supreme intelligent cause of the world had decided to form an offspring directly each time, leaving the mother only to unfold and nourish it. They declared themselves for preformation A 377, as if it made any difference whether such forms originated supernaturally at the beginning of the world or during its course. In fact, occasional creation would save a great many supernatural arrangements. Such arrangements would be needed to keep an embryo formed at the beginning of the world unharmed by nature's destructive forces throughout the long period before its development. It would also save an immeasurably greater number of preformed beings than were ever to develop, and with them as many acts of creation, which would otherwise be unnecessary and purposeless. But these supporters wanted at least to leave something to nature, rather than fall into a complete hyperphysics that could dispense with every natural explanation. They still clung to their hyperphysics even when they found an admirable purposiveness in malformed births, which cannot possibly be regarded as natural ends. This purposiveness would have to be aimed merely at an anatomist's one day being troubled by it as a purposeless purposiveness and feeling disheartening wonder. But they could not possibly fit the production of hybrids into the system of preformation 424. They had conceded to the male creatures' semen nothing but the mechanical property of serving as the embryo's first nourishment. They now also had to grant it a A 378 purposively formative power, even though, in the generation of a complete product by two creatures of the same kind, they were unwilling to concede this power to either parent.

By contrast, even if we did not know the great advantage the defender of epigenesis has over the former theory in the empirical evidence supporting it, reason would already be especially well disposed toward this way of explaining things. For with regard to things we can originally represent as possible only through causality according to ends, it at least regards nature as itself productive in reproduction, not merely as unfolding what is already there. Thus, with the smallest possible expenditure of the supernatural, it leaves everything after the first beginning to nature. But it determines nothing about this first beginning, at which physics as a whole founders, whatever chain of causes it may try.

No one has done more to prove this theory of epigenesis, or to establish the genuine principles of its application, partly by restricting an overly bold use of it, than Court Councilor Blumenbach. He begins every physical explanation of these formations with organized matter A 379. He rightly declares it contrary to reason to suppose that raw matter originally formed itself according to mechanical laws; that life could have sprung from the nature of the lifeless; or that matter could by itself have arranged itself into the form of a self-preserving purposiveness. At the same time, under this principle of an original organization, which we cannot fathom, he leaves nature's mechanism a share that cannot be determined, yet also cannot be denied. Accordingly, he calls matter's capacity in an organized body a formative drive, distinguishing it from the merely mechanical formative power universally present in matter. This drive stands, as it were, under the higher direction and guidance of the original organization.

82 425 On the teleological system in the external relations of organized beings.

By external purposiveness I mean the case in which one natural thing serves another as a means to an end. Things that have no inner purposiveness, and do not presuppose it for their possibility—for example, earths, air, water, and so on—can nevertheless be very purposive externally, that is, in relation to other beings. But those other beings must A 380 always be organized beings, or natural ends. Otherwise the former things could not be judged as means either. Thus water, air, and earths cannot be regarded as means for piling up mountains. Mountains contain nothing in themselves that requires a ground of their possibility according to ends. Their cause can therefore never be represented, in relation to them, under the predicate of a means useful for that purpose.

External purposiveness is a completely different concept from inner purposiveness, which is connected with an object's possibility regardless of whether its actuality is itself an end. We can still ask of an organized being, “What is it there for?” But we cannot readily ask this of things in which we recognize only the effect of nature's mechanism. For in organized beings we already represent a causality according to ends—a creative understanding—as required for their inner possibility, and we refer this active capacity to its determining ground, intention. There is only one external purposiveness connected with the inner purposiveness of organization that serves an end in the external relation of a means without requiring us to ask for what end this being, organized in this way, had to exist. This is the organization of the two sexes A 381 in relation to one another for reproducing their kind. For here, just as with an individual, we can still ask, “Why did such a pair have to exist?” The answer is: only the pair constitutes an organizing whole, though not an organized whole within a single body.

Now when we ask what a thing is there for, the answer is either: its existence and generation have no relation at all to a cause acting according to intentions. In that case we always mean 426 that it originates from nature's mechanism. Or the answer is: there is some intentional ground of its existence as a contingent natural being. We can hardly separate this thought from the concept of an organized thing. For once we must place a causality of final causes and an idea underlying that causality beneath its inner possibility, we also cannot think of the existence of this product except as an end. A represented effect whose representation is at the same time the determining ground of an intelligent efficient cause for producing it is called an end. In this case, then, we can say either: the end of this natural being's existence lies within itself, so it is not merely an end but also a final end. Or we can say: that end lies outside it, in other natural beings, so it exists purposively not as a final end but necessarily also A 382 as a means.

But if we survey the whole of nature, we find in it, as nature, no being that could claim the distinction of being the final end of creation. Indeed, we can prove a priori that whatever might be an ultimate end for nature, endowed with every conceivable determination and property, could still never, as a natural thing, be a final end.

When we look at the plant kingdom, its immense fertility, through which it spreads over almost every soil, might initially suggest that it is a mere product of nature's mechanism, like the mechanism nature displays in the formations of the mineral kingdom. But a closer knowledge of the indescribably wise organization within it prevents us from staying with this thought. Instead, it prompts the question: what are these creatures there for? If we answer, “For the animal kingdom, which feeds on them, so that it could spread over the earth in so many different kinds,” the question returns: what, then, are these plant-eating animals there for? The answer might be: for the predators, which can feed only on what has life A 383. Finally, the question is: what are these, together with the previous kingdoms of nature, good for? For human beings, for the many uses their understanding teaches them to make of all those creatures. The human being is the ultimate end of creation here on earth, because he is the only being on 427 it able to form a concept of ends and, through reason, to make a system of ends out of an aggregate of purposively formed things.

With the knight Linnaeus, we could also take what seems to be the opposite route and say: plant-eating animals exist to restrain the luxuriant growth of the plant kingdom, which would otherwise choke many of its species; predators exist to limit the voracity of those animals; and finally human beings exist so that, by hunting and reducing the predators, they establish a certain balance between nature's productive and destructive forces. Thus, however much the human being might deserve to count as an end in one respect, in another he would again rank only as a means.

Suppose we take as our principle an objective purposiveness in the diversity of the kinds of earthly creatures and in their external relations to one another as purposively constructed beings. It is then in keeping with reason to think in these relations of a certain A 384 organization, a system of all nature's kingdoms according to final causes. But here experience seems loudly to contradict reason's maxim, especially concerning an ultimate end of nature. Such an end is required for the possibility of this system, and we can place it nowhere but in the human being. Yet regarding the human being as one among many animal kinds, nature has made no exception at all, either in its destructive or its productive forces, to its practice of subjecting everything to its mechanism without an end.

In an arrangement of earthly natural beings into a purposive whole, the first thing that would have to be intentionally arranged would surely be their dwelling place: the soil and the element on and in which they were to thrive. But closer knowledge of this foundation of all organic generation points only to causes acting wholly unintentionally. Indeed, these causes are more destructive than favorable to generation, order, and ends. Land and sea contain not only monuments of ancient, immense devastations that affected them and every creature on and in them. Their entire structure—the strata of the one and the boundaries of the other—has every appearance of being the product of wild, overpowering forces of a nature working in a chaotic state. 428 However purposively the present shape, structure, and slope of the A 385 land may seem arranged to receive water from the air, to provide watercourses between various kinds of strata for various products, and to channel rivers, closer investigation shows that they arose merely as effects of eruptions, partly of fire and partly of water, or of upheavals of the ocean. This applies both to the first production of this shape and especially to its later transformation, together with the destruction of its first organic products.³⁰ Now if the dwelling place, the maternal soil of the land, and the maternal womb of the sea for all these creatures point to nothing but a wholly unintentional mechanism of their production, how, and A 386 with what right, can we demand and assert another origin for these later products? The closest examination of the remains of those natural devastations seems, in Camper's judgment, to show that the human being was not involved in these upheavals. Nevertheless, he depends so much on the earth's other creatures that, if we admit a mechanism of nature universally governing the others, we must regard him as included under it too, even if his understanding has largely enabled him to escape their devastations.

But this argument seems to prove more than was intended in putting it forward. It seems to prove not merely that the human being cannot be nature's ultimate end, and that for the same reason the aggregate of organized natural things on earth cannot be a system of ends, but also that the natural products previously regarded as natural ends have no origin other than nature's mechanism.

429 Yet in the solution above to the antinomy between the principles of mechanical and teleological generation of organic natural beings, we saw that these are only principles of reflective judgment A 387 concerning nature's formative activity according to its particular laws, to whose systematic connection we lack the key. They do not determine these beings' origin in itself. They say only that, given the character of our understanding and reason, we can think their origin only according to final causes. Consequently, the greatest possible effort, even boldness, in trying to explain them mechanically is not only permitted: reason also calls on us to make it. This is so although we know we can never succeed sufficiently, for subjective reasons arising from the particular kind and limitation of our understanding, not because mechanical generation would in itself contradict an origin according to ends. We also saw that the compatibility of the two ways of representing nature's possibility may well lie in nature's supersensible principle, both outside us and within us. Representation according to final causes is only a subjective condition of our use of reason when reason wants objects judged not merely as appearances. Instead, it demands that these appearances themselves, together with their principles, be referred to the supersensible substrate, to make possible certain laws of their unity that it can represent only through ends. And reason also has ends that are supersensible.

83 A 388 On the ultimate end of nature as a teleological system.

We have shown above that principles of reason give us sufficient grounds to judge the human being not just as a natural end, like every organized being, but also as nature's ultimate end here on earth. In relation to him, all other natural things form a system of ends. This holds not for determining but for reflective judgment. Now what is to be promoted as an end through his connection with nature must be found within the human being himself. So either this end 430 must be of a kind that nature itself can satisfy through its beneficence, or it must be his fitness and skill for all kinds of ends for which he could use nature, both external and internal. The first end of nature would be human happiness; the second, human culture.

The concept of happiness is not one that human beings abstract from their instincts and thus draw from the animal nature within themselves. It is a mere idea of a state to which they want to make those instincts adequate under merely empirical conditions—which is impossible A 389. Human beings devise this idea for themselves through an understanding entangled with imagination and the senses, and do so in very different ways. They even change it so often that nature, even if it were entirely subject to their choice, could adopt no definite, universal, fixed law for agreeing with this fluctuating concept, and thus with the end each person sets at will. But even if we reduced this end to genuine natural need, about which our species universally agrees with itself, or, on the other hand, raised as high as we liked our skill in obtaining imagined ends, human beings would still never attain what they understand by happiness. This is, in fact, their own ultimate natural end, not an end of freedom. Their nature is not such that it can ever stop acquiring and enjoying and be satisfied. On the other hand, far from adopting human beings as its special favorites and treating them more generously than all other animals, nature spares them no more than any other animal in its destructive effects: plague, hunger, floods, frost, attacks by other animals large and small, and the like. Still more, the conflicting natural predispositions within human beings involve them in plagues of their own devising, as well as afflictions imposed by their own kind through oppressive A 390 rule, the barbarity of wars, and so on. These bring such misery, and human beings do so much to destroy their own species, that even with the most beneficent nature outside us, nature's end, if directed toward our species' happiness, would not be attained in a system of nature on earth. For the nature within us is not receptive to it. Human beings therefore always remain only a link in the chain of natural ends. They are indeed a principle with regard to many 431 ends for which nature seems to have destined them by their predispositions, since they make themselves so. But they are also a means of preserving purposiveness in the mechanism of the other links. As the only being on earth with understanding, and therefore with the capacity to set ends for himself at will, the human being is certainly the entitled lord of nature. If we regard nature as a teleological system, he is by his vocation its ultimate end. But this is always conditional: he must understand how, and have the will, to give nature and himself a relation to ends that is self-sufficient independently of nature, and can therefore be a final end. This final end, however, must not be sought in nature at all.

To discover what, in the human being, we must at least identify as nature's ultimate end, we must pick out what A 391 nature can accomplish to prepare him for what he himself must do to be a final end. We must separate this from all ends whose possibility rests on conditions we can expect only from nature. Happiness on earth is of the latter kind. By this we mean the totality of all his ends possible through nature outside and within him. This is the material of all his ends on earth. If he makes it his entire end, it makes him incapable of setting a final end for his own existence and bringing himself into agreement with it. Of all his ends in nature, then, only the formal, subjective condition remains: his fitness to set himself ends in general and to use nature as a means suited to the maxims of his freely chosen ends in general, independently of nature in determining those ends. This is what nature can accomplish with a view to the final end lying outside it, and what can therefore be regarded as its ultimate end. Producing a rational being's fitness for any ends whatsoever, and thus in his freedom, is culture. Only culture, then, can be the ultimate end that we have reason to attribute to nature with regard to the human species. It is not human happiness on earth, still less merely being the chief instrument for establishing order A 392 and harmony in the nonrational nature outside him.

But not every kind of culture is sufficient for this ultimate end of nature. The culture of skill is certainly the chief subjective condition of fitness for promoting ends in general. Yet 432 it is not sufficient to develop the will in determining and choosing its ends, which is essential to the full scope of fitness for ends. This latter condition of fitness, which we might call the culture of training or discipline, is negative. It consists in freeing the will from the despotism of desires. These attach us to certain natural things and make us incapable of choosing for ourselves. We let drives serve as fetters, although nature gave them to us only as guiding threads, so that we would not neglect or violate the vocation of the animal nature within us. Yet we are free enough to tighten or loosen them, lengthen or shorten them, as the ends of reason require.

Skill can hardly develop in the human species except through inequality among people. The majority provide life's necessities almost mechanically, without needing special art, to supply the comfort and leisure of others. These others cultivate the less necessary parts of culture, science A 393 and art, and keep the majority in a condition of oppression, hard labor, and little enjoyment. Still, some of the culture of the higher class gradually spreads to this class too. But as culture advances, afflictions increase with equal force on both sides: on one through others' violence, on the other through inner dissatisfaction. Culture's height is called luxury when the inclination toward the dispensable starts to harm the indispensable. Yet this glittering misery is bound up with the development of natural predispositions in the human species. Nature's own end, although not our end, is nevertheless attained in it. The formal condition under which alone nature can attain this final intention is a constitution governing people's relations to one another in which lawful power counters the damage caused by their mutually conflicting freedom, within a whole called civil society. Only in such a society can natural predispositions develop to the fullest. But even if people were clever enough to discover it and wise enough to submit willingly to its constraint, a cosmopolitan whole would still be needed: a system of all states that are in danger of harming one another. Without it, and given the obstacle 433 that ambition, lust for power, and greed—especially in those A 394 who hold power—put in the way of even the possibility of such a plan, war is inevitable. Sometimes states split up and dissolve into smaller ones; sometimes one state unites smaller ones with itself and tries to form a larger whole. War is an unintentional human attempt, prompted by unbridled passions. Yet it is also a deeply hidden, perhaps intentional attempt by supreme wisdom, if not to establish, then at least to prepare, lawfulness together with the freedom of states, and thereby the unity of a morally grounded system of them. Despite the most dreadful sufferings it inflicts on the human race, and perhaps the still greater burdens imposed in peacetime by constant readiness for it, war is nevertheless one more incentive to develop to the highest degree all talents serving culture. Meanwhile, the hope of a peaceful state of national happiness recedes ever further.

As for disciplining the inclinations, our natural predisposition toward them is entirely purposive in relation to our vocation as an animal kind. But they make the development of humanity very difficult. Still, with regard to this second requirement for culture too, nature shows a purposive striving toward an education that makes us receptive to higher ends than nature itself can supply. We cannot deny the preponderance of evils poured over us by the refinement of taste to the point of A 395 idealizing it, and even by luxury in the sciences as nourishment for vanity, through the unsatisfiable multitude of inclinations they produce. But neither can we overlook nature's end: increasingly to overcome the coarseness and violence of those inclinations that belong more to the animal nature within us and most oppose our development toward our higher vocation—the inclinations toward enjoyment—and to make room for the development of humanity. Fine art and the sciences, through a pleasure that can be universally communicated and through polish and refinement for society, make human beings civilized, even if not morally better. They greatly weaken the tyranny of the propensity toward the senses and thereby prepare human beings for a rule in which reason alone is to have authority. Meanwhile, the evils with which nature and people's quarrelsome selfishness afflict us summon, heighten, and harden the powers of the soul 434 so that we do not succumb to them. In this way they let us feel a fitness for higher ends hidden within us.³¹

84 A 396 On the final end of the existence of a world, that is, of creation itself.

A final end is an end that needs no other end as a condition of its possibility.

If we assume nature's mere mechanism as the ground for explaining its purposiveness, we cannot ask what things in the world are there for. According to such an idealist system, we are concerned only with the physical possibility of things. Thinking of them as ends would be mere reasoning without an object. Whether we interpret their form as due to chance or to blind necessity, in either case A 397 that question would be empty. But suppose we take the connection of ends in the world as real, and assume for it a special kind of causality, that of an intentionally acting cause. We then cannot stop with asking why things in the world—organized beings—have this or that form, or have been placed by nature in this or that relation to others. Once we think of an understanding that must be regarded as the cause of the possibility of the forms actually found in things, we must also ask about the 435 objective ground within that same understanding that could have determined this productive understanding to an effect of this kind. This ground is then the final end for which such things exist.

I said above that the final end is not an end that nature could sufficiently bring about and produce in keeping with its idea, because it is unconditioned. For there is nothing in nature, as an object of sense, whose determining ground within nature is not itself always conditioned in turn. This holds not only for nature outside us, material nature, but also for nature within us, thinking nature. It must be understood that I am considering within myself only what is nature. But a thing that, because of its objective character, is necessarily to exist as the final end of an intelligent cause must be such that, in the order of ends, it depends on no condition A 398 other than its idea alone.

Now there is only one kind of being in the world whose causality is teleological, that is, directed toward ends, yet is also of such a character that these beings themselves represent the law by which they must determine their ends as unconditioned, independent of natural conditions, and necessary in itself. This being is the human being, but considered as a noumenon: the only natural being in which, from his own character, we can nevertheless recognize a supersensible capacity—freedom—and even the law of its causality, together with its object, which he can set himself as his highest end: the highest good in the world.

Now of the human being—and thus of every rational being in the world—as a moral being, we can no longer ask for what end (quem in finem) he exists. His existence has the highest end within itself. To this end he can subordinate the whole of nature as far as he is able; at least, he must not regard himself as subject to any natural influence that conflicts with it.—If things in the world, as beings dependent in their existence, need a supreme cause acting according to ends, then the human being is creation's final end. Without him, the chain of subordinate ends would lack a complete ground. Only in the human being, and even in him only as the subject of morality A 399, do we find unconditioned legislation concerning ends. This 436 alone makes him capable of being a final end to which the whole of nature is teleologically subordinated³².

85 A 400 On physicotheology.

Physicotheology is reason's attempt to infer nature's supreme cause and its properties from nature's ends, which can be known only empirically. A moral theology, or ethicotheology, would be the attempt to infer that cause and its properties from the moral end of rational beings in nature, which can be known a priori.

The former naturally precedes the second. For if we want to infer a world cause teleologically from things in the world, ends of nature must first be given. For these 437 we must then seek a final end, and for this, in turn, the principle of that supreme cause's causality.

Many investigations of nature can and must proceed according to the teleological principle without our having reason to ask about the ground of the possibility of acting purposively that we find in various natural products A 401. But if we also want a concept of this ground, we have absolutely no further insight for that purpose beyond the maxim of reflective judgment. Namely, even if just one organic product of nature were given to us, the character of our cognitive power would prevent us from thinking any ground for it other than a cause of nature itself—whether of all nature or only this part of it—that contains the causality for the product through understanding. This principle of judgment takes us no further in explaining natural things and their origin. Yet it opens some prospect beyond nature, so that perhaps we can determine more closely the otherwise so unfruitful concept of an original being.

Now I say: however far physicotheology may be pursued, it can reveal nothing to us about a final end of creation. It does not even reach the question of such an end. It can justify the concept of an intelligent world cause as the only concept subjectively suited to the character of our cognitive power for the possibility of things that we can make understandable to ourselves through ends. But it cannot determine this concept further for either theoretical or practical purposes. Its attempt does not achieve its A 402 intention of grounding a theology. It always remains only a physical teleology, because the relation to ends within it is always, and must always be, regarded as conditioned within nature. It therefore cannot even raise the question of the end for which nature itself exists, whose ground must be sought outside nature. Yet a determinate concept of that higher intelligent world cause, and thus the possibility of a theology, depends on a determinate idea of this end.

What things in the world are useful to one another for; what the manifold parts in a thing are good for with regard to that thing itself; how we even have reason to assume that nothing in the world is in vain, but everything is good for something in nature, on condition that certain things exist as ends: all this the teleological consideration of the world accomplishes splendidly, to our utmost admiration. In doing so, our reason has at its disposal for judgment 438 no other principle of the possibility of the object of its unavoidable teleological judging than to subordinate nature's mechanism to the architectonic of an intelligent author of the world. But the data, and thus the principles, for determining this concept of an intelligent world cause as supreme artist are merely empirical. They allow us to infer no further properties than experience reveals in its effects. Since experience can never encompass the whole of nature as a system A 403, it must often encounter grounds of proof that seem to conflict with that concept and with one another. Yet even if we could empirically survey the entire system insofar as it concerns mere nature, experience could never raise us beyond nature to the end of its very existence, and thereby to a determinate concept of that higher intelligence.

If we make the problem physicotheology seeks to solve a small one, its solution seems easy. Suppose we lavish the concept of a deity on any intelligent being we think of—one or several—possessing many very great properties, but not all those required to ground a nature in general that agrees with the greatest possible end. Or suppose we think nothing of supplementing the deficiencies of a theory's evidence with arbitrary additions. Where we have grounds for assuming much perfection—and what is much for us?—we then consider ourselves entitled to presuppose all possible perfection. Physical teleology can then make impressive claims to the honor of grounding a theology. But if we are required to show what prompts us, and moreover entitles us, to make those additions, we will search in vain for justification in the principles of reason's theoretical use. This use insists that, in explaining an A 404 object of experience, we attribute to it no more properties than empirical data supporting their possibility can be found. Closer examination would show that an idea of a highest being really underlies this in us a priori, resting on an entirely different use of reason, the practical one. This idea prompts us to complete physical teleology's deficient representation of the original ground of nature's ends into the 439 concept of a deity. We would then not falsely imagine that we had produced this idea, and with it a theology, through the theoretical use of reason in our physical knowledge of the world, much less proved its reality.

We should not blame the ancients too severely for thinking of their gods as differing greatly in their powers, intentions, and purposes of will, but all still limited in human fashion, not even excepting their chief. For when they considered the arrangement and course of things in nature, they certainly found enough grounds to assume something more than mechanical as their cause. They suspected intentions of certain higher causes, which they could think of only as superhuman, behind the machinery of this world. But because A 405 they found good and evil, the purposive and the counterpurposive, very mixed in it, at least to our insight, they could not permit themselves to assume wise and beneficent ends secretly underlying it when they saw no proof of them, just to support an arbitrary idea of a supremely perfect author. Their judgment about the supreme world cause could hardly have turned out otherwise, insofar as they proceeded consistently according to maxims of the merely theoretical use of reason. Others, who wanted to be theologians as well as physicists, thought they could satisfy reason by providing the absolute unity of the principle of natural things that reason demands. They did so through the idea of a being in which, as the sole substance, all those things would be merely inhering determinations. This substance would not cause the world through understanding, but all the understanding of worldly beings would nevertheless be found in it as their subject. It would therefore be a being that produced nothing according to ends, but within which all things, because of the unity of the subject whose mere determinations they are, must necessarily relate purposively to one another even without end or intention. Thus they introduced the idealism of final causes. Instead of making the hard-to-explain unity of a multitude of purposively connected substances a causal dependence on one substance, they turned it into inherence in one A 406. Considered from the side of the inhering worldly beings, this system was called pantheism; considered from the side of the sole self-subsisting subject as original being, it was later called Spinozism. It did not so much 440 solve the question of the first ground of nature's purposiveness as declare it null. For it stripped the latter concept of all reality and turned it into a mere misinterpretation of a general ontological concept of a thing in general.

Thus, from merely theoretical principles of the use of reason, on which alone physicotheology rests, we can never arrive at a concept of a deity adequate for our teleological judgment of nature. Either we declare all teleology a mere deception of judgment in assessing the causal connection of things, and retreat to the sole principle of nature's mere mechanism. Nature would then only seem to contain a universal relation to ends, because of the unity of the substance of which nature is nothing but the manifold of determinations. Or, instead of this idealism of final causes, we remain committed to the principle of the reality of this special kind of causality. We may then place many intelligent original beings beneath nature's ends, or only one. But as soon as we have nothing at hand to ground our concept of it except principles of experience drawn from the actual connection of ends A 407 in the world, two difficulties arise. On the one hand, we have no answer to the disharmony nature presents in many examples concerning the unity of ends. On the other, from this we can never draw a sufficiently determinate concept of a single intelligent cause, as we arrive at it with justification from experience alone, for a theology useful in any way, theoretically or practically.

Physical teleology certainly prompts us to seek a theology, but cannot produce one, however far we trace nature through experience and supplement the connection of ends discovered in it with ideas of reason, which must be theoretical for physical problems. People will rightly complain: what good is it to place a great understanding, immeasurable to us, beneath all these arrangements and have it order this world according to intentions, if nature tells us nothing, and never can tell us anything, about the final intention? Without that, we can form no common point of reference for all these natural ends, no adequate teleological principle either for knowing all the ends in a system or for forming a 441 concept of the supreme understanding as the cause of such a nature that could guide our judgment as it reflects teleologically on nature A 408. I would then have an artistic understanding for scattered ends, but no wisdom for a final end, which must really contain the determining ground of that understanding. Only pure reason can supply a final end a priori, since all ends in the world are empirically conditioned and can contain only what is good for this or that contingent intention, not what is good without qualification. Such an end alone would teach me what properties, what degree, and what relation I must think in nature's supreme cause in order to judge nature as a teleological system. Without it, how and with what right may I arbitrarily expand my very limited concept of that original understanding, which I can base on my scant knowledge of the world; of this original being's power to make its ideas actual; of its will to do so, and so on, completing it into the idea of an all-wise infinite being? If this were done theoretically, it would presuppose omniscience in me. I would have to understand nature's ends in their complete connection and, moreover, be able to think of every other possible plan, in comparison with which the present one could justifiably be judged the best. Without this complete knowledge of the effect, I cannot infer a determinate concept of the supreme cause—which can be found only in the concept of an intelligence infinite in every A 409 respect, that is, the concept of a deity—and establish a foundation for theology.

So, however far physical teleology may be extended, we can indeed say, according to the principle stated above, that given the character and principles of our cognitive power, we can think of nature, in the purposive arrangements known to us, only as the product of an understanding to which it is subordinated. But theoretical inquiry into nature can never reveal whether this understanding also had a final intention for nature as a whole and its production—an intention that would then not lie in the nature of the sensible world. With all our knowledge of nature, it remains undecided whether that supreme cause is its original ground according to a final end at all. It might instead act through an understanding determined by the mere necessity of its nature to produce certain forms 442, by analogy with what we call artistic instinct in animals. There would then be no need to attribute even wisdom to it, much less supreme wisdom united with all the other properties required for the perfection of its product.

Physicotheology is therefore a misunderstood physical teleology A 410, useful only as a preparation for theology. It is sufficient for this purpose only when another principle is added on which it can rely, not by itself, as its name would suggest.

86 On ethicotheology.

Even the most ordinary understanding cannot escape a certain judgment when it reflects on the existence of things in the world and on the existence of the world itself. All the diverse creatures, however artfully constructed and however variously connected through purposive relations to one another—even the whole of their many systems, which we incorrectly call worlds—would exist for nothing if there were no human beings in them, or rational beings in general. Without the human being, the whole of creation would be a mere wilderness, in vain and without a final end. But it is not his cognitive power, theoretical reason, in relation to which the existence of everything else in the world first acquires its value, as if someone merely had to be there to contemplate the world. For if this contemplation presented him only with things without a final end, then A 411 no value could accrue to their existence from their being known. We must already presuppose a final end of the world, in relation to which contemplating it would itself have value. Nor is it the feeling of pleasure and its total sum in relation to which we think of a final end of creation as given. That is, it is not well-being, enjoyment, whether bodily or intellectual—in a word, happiness—by which we assess that absolute value. For the fact that, once human beings exist, they make happiness their own final intention gives no concept of why they exist at all, or what value they themselves have, such that their existence should be made pleasant. 443 The human being must therefore already be presupposed as creation's final end if we are to have a rational ground for nature's having to agree with his happiness when regarded as an absolute whole according to principles of ends.—So it is only the faculty of desire that matters here. But not the faculty that makes him dependent on nature through sensible impulses, not that in relation to which the value of his existence rests on what he receives and enjoys. Rather, it is the value that he alone can give himself, consisting in what he does: how and according to what principles he acts, not as a member of nature, but in the freedom of his faculty of desire. That is, a A 412 good will is what alone can give his existence absolute value, and in relation to which the existence of the world can have a final end.

The most ordinary judgment of sound human reason agrees completely with this: the human being can be creation's final end only as a moral being, if we merely direct judgment to this question and encourage it to attempt an answer. People will say: what good is it that this person has so much talent, is very active in using it, and thereby beneficially influences the community, thus having great value both for his own fortunate circumstances and for others' benefit, if he has no good will? Considered inwardly, he is a contemptible object. And if creation is not to be entirely without a final end, then he, who belongs to it as a human being, must, as an evil person in a world under moral laws, lose his subjective end, happiness, in accordance with those laws. This is the sole condition under which his existence can coexist with the final end.

If we now find arrangements of ends in the world and, as reason unavoidably requires, subordinate conditioned ends to an unconditioned supreme end, a final end, we can first easily see that we are then concerned not with an end of nature within A 413 nature, insofar as it exists, but with the end of its existence together with all its arrangements. Thus we are concerned with the ultimate end of creation, and, in this, really with the supreme condition under which alone a final end can occur: the determining ground of a highest understanding for producing beings in the world.

444 Since we recognize the human being as creation's end only as a moral being, we now have, first, a ground, or at least the chief condition, for regarding the world as a whole connected according to ends and as a system of final causes. Above all, for the relation of natural ends to an intelligent world cause that the character of our reason makes necessary for us, we have a principle for thinking the nature and properties of this first cause as the supreme ground in the kingdom of ends, and so for determining its concept. Physical teleology could not do this. It could occasion only indeterminate concepts of that cause, and precisely for that reason concepts unfit for theoretical as well as practical use.

From this precisely determined principle of the original being's causality, we must think of it not merely as an intelligence legislating for nature, but also as a legislating head in a moral kingdom of ends. In relation to the highest good possible A 414 only under its rule—the existence of rational beings under moral laws—we must think of this original being as omniscient, so that even the innermost dispositions, which constitute the actual moral value of rational worldly beings' actions, are not hidden from it; as omnipotent, so that it can make the whole of nature adequate to this highest end; as all-good and also just, because these two properties, which together constitute wisdom, are the conditions of the causality of a supreme cause of the world as the highest good under moral laws. We must likewise think in it all the remaining transcendental properties, such as eternity, omnipresence, and so on, that are presupposed in relation to such a final end, since goodness and justice are moral properties.—In this way moral teleology makes up for the deficiency of physical teleology and first grounds a theology. For if physical teleology did not borrow unnoticed from moral teleology, but proceeded consistently, it could by itself ground nothing but a demonology, incapable of any determinate concept.

But the principle that relates the world to a supreme cause A 415 as a deity, because of the moral vocation of certain beings in it, does not do this merely by completing the physicotheological proof and thus necessarily basing itself on that proof. It is also sufficient by itself for this purpose. It directs 445 attention to nature's ends and encourages investigation of the incomprehensibly great art hidden behind their forms, to give incidental confirmation in natural ends to the ideas supplied by pure practical reason. For the concept of worldly beings under moral laws is an a priori principle according to which human beings must necessarily judge themselves. Reason also recognizes a priori the following as a necessary principle for its teleological judgment of things' existence: if there is any world cause acting intentionally and directed toward an end, that moral relation must be just as necessary a condition of the possibility of a creation as the relation according to physical laws, assuming that intelligent cause also has a final end. All that matters now is whether we have any ground sufficient for reason, speculative or practical, to attribute a final end to the supreme cause acting according to ends. For if so, we can regard it as certain a priori for us that, given our reason's subjective character—and however we may think of the reason of other beings—this end can be none other than the A 416 human being under moral laws. By contrast, nature's ends in the physical order cannot be known a priori at all. In particular, we cannot in any way understand that a nature could not exist without them.

Remark.

Consider a person at moments when his mind is attuned to moral feeling. Surrounded by beautiful nature and calmly, cheerfully enjoying his existence, he feels a need within himself to be grateful to someone for it. Or suppose that, in the same state of mind on another occasion, he finds himself beset by duties that he can and will fulfill only through voluntary sacrifice. He feels a need to have thereby also carried out something commanded and obeyed an overlord. Or perhaps he has thoughtlessly violated his duty, without thereby becoming answerable to other people. His stern self-reproaches will nevertheless speak within him as though they were the voice of a judge to whom he must 446 account for it. In a word, he needs a moral intelligence, so that for the end for which he exists there is a being that is the cause of that person and the world in keeping with this end. It is futile to invent motives behind these feelings. They are directly connected with the purest moral disposition. For gratitude, obedience, and humility—submission to A 417 deserved punishment—are particular attitudes of mind toward duty. Here the mind inclined to expand its moral disposition freely thinks of an object outside the world, so as, if possible, to demonstrate its duty toward such an object too. It is therefore at least possible, and grounded in a moral way of thinking, to represent a purely moral need for the existence of a being under which our morality gains either greater strength or, at least in our representation, greater scope: a new object on which to exercise it. That is, it is possible to assume a morally legislating being outside the world, without regard to theoretical proof, still less to selfish interest, on a purely moral ground free of every foreign influence, though admittedly only subjective, solely on the recommendation of a pure practical reason that legislates by itself alone. Such a state of mind might occur rarely or not last long. It might pass fleetingly, without lasting effect, without any reflection on the object represented in such a shadowy image, and without any effort to bring it under clear concepts. Yet its ground is unmistakable: the moral predisposition within us as a subjective principle for not being satisfied, in contemplating the world, with its purposiveness through natural causes, but for placing beneath it a supreme cause governing nature according to moral principles.—Moreover, we feel driven by the moral law to strive for a universal highest end, yet feel that neither we nor nature as a whole can attain it. We may judge ourselves to accord with the final end of an intelligent world cause, if A 418 there is one, only insofar as we strive for it. Thus practical reason has a purely moral ground for assuming this cause, since this can be done without contradiction, if for no other reason than to avoid the danger of regarding this striving as entirely futile in its effects and thereby letting it weaken.

447 All this is meant to say only the following: fear may first have produced gods, or demons, but reason, through its moral principles, could first produce the concept of God. This remains true even if people were, as commonly happens, very ignorant about nature's teleology, or very doubtful because it is difficult to reconcile its contradictory appearances through an adequately established principle. The inner moral determination of the end of one's existence supplied what knowledge of nature lacked. It directed us, with regard to the final end of all things' existence—for which only an ethical principle satisfies reason—to think of the supreme cause with properties enabling it to subordinate the whole of nature to that single intention, for which nature is merely an instrument: that is, to think of it as a deity.

87 On the moral proof of God's existence.

There is a physical teleology that provides a sufficient ground of proof for our theoretically reflective judgment to assume the existence of an intelligent world cause. But we A 419 also find a moral teleology within ourselves, and still more in the concept of a rational being endowed with freedom in its causality in general. Since the relation to ends within us can be determined a priori together with its law, and thus known as necessary, this teleology needs no intelligent cause outside us to account for this inner lawfulness. We need such a cause no more than we need to look beyond the purposiveness we find in the geometrical properties of figures, for every possible exercise of art, to a highest understanding that gave it to them. But this moral teleology still concerns us as worldly beings, and thus as beings connected with other things in the world. The same moral laws prescribe that we direct our judgment toward these other things either as ends, or as objects in relation to which we ourselves are the final end. This moral teleology concerns the relation of our own causality to ends, even to a final end we must intend in the world. It also concerns the reciprocal relation of the world to that moral end and the 448 external possibility of carrying it out, about which no physical teleology can guide us. From it now arises the necessary question: does it compel our rational judgment to go beyond the world A 420 and seek a supreme intelligent principle for this relation of nature to the moral within us, so that we may represent nature as purposive also in relation to our inner moral legislation and its possible realization? There is therefore indeed a moral teleology. It is connected with the legislation of freedom on the one hand and that of nature on the other just as necessarily as civil legislation is connected with the question of where executive power is to be sought. Such a connection exists generally wherever reason is to give a principle of the actuality of a certain lawful order of things possible only through ideas.—We will first present reason's advance from that moral teleology and its relation to physical teleology to theology, then consider the possibility and cogency of this kind of inference.

If we assume the existence of certain things, or even only certain forms of things, to be contingent, and thus possible only through something else as their cause, we can seek the supreme ground of this causality, and thus the unconditioned ground of what is conditioned, either in the physical or in the teleological order: according to the nexus of efficient causes or the nexus of final causes. That is, we can ask: what is the supreme productive cause? Or: what is its supreme, absolutely unconditioned end A 421, the final end of its production of these products or of all its products in general? Here we admittedly presuppose that this cause can represent ends and is therefore an intelligent being, or at least must be thought by us as acting according to the laws of such a being.

Now, if we follow the latter order, there is a principle to which even the most ordinary human reason is compelled to agree immediately: if there is to be any final end, which reason must specify a priori, it can be none other than the human being—every rational worldly being—under moral laws.³³ For everyone judges as follows. If the world consisted A 422 449 entirely of lifeless beings, or partly of living but nonrational beings, its existence would have no value at all, because no being would exist in it with the slightest concept of value. Suppose, instead, there were also rational beings, but their reason could place the value of things' existence only in nature's relation to them, their well-being, and could not itself originally A 423 create such value in freedom. There would then be relative ends in the world, but no absolute final end, because the existence of such rational beings would itself always be purposeless. Moral laws, however, have this distinctive character: they prescribe something to reason as an end without condition, exactly as the concept of a final end requires. Thus only the existence of a reason that can itself be the supreme law for itself in the relation to ends—in other words, the existence of rational beings under moral laws—can be thought as the final end 450 of a world's existence. Otherwise, either no end at all in its cause underlies the world's existence, or ends without a final end underlie it.

The moral law, as the formal rational condition of the use of our freedom, binds us by itself alone, without depending on any end as a material condition. Yet it also determines for us, a priori, a final end that it obliges us to pursue: the highest good in the world possible through freedom.

The subjective condition under which the human being—and, according to all our concepts, every finite rational being—can set himself a final end under the above law is happiness. Consequently, A 424 the highest physical good possible in the world, which we are to promote as a final end as far as we can, is happiness, under the objective condition of human agreement with the law of morality, as worthiness to be happy.

But with all our powers of reason, we cannot represent these two requirements of the final end assigned to us by the moral law as connected through mere natural causes and as adequate to the idea of that final end. So the concept of the practical necessity of such an end through the application of our powers does not agree with the theoretical concept of the physical possibility of bringing it about, if we connect our freedom with no causality as a means except nature's.

Consequently, we must assume a moral world cause, an author of the world, in order to set ourselves a final end in accordance with the moral law. As far as the latter is necessary, so far—that is, to the same degree and on the same ground—it is also necessary to assume the former: namely, that there is a God.³⁴

This proof, which can easily be given a logically precise form A 425, does not mean that assuming God's existence 451 is just as necessary as acknowledging the moral law's validity, and that whoever cannot convince himself of the former may therefore judge himself released from obligations under the latter. No! Only the intention of realizing the final end in the world through following the moral law would then have to be abandoned: the happiness of rational beings harmoniously coinciding with their observance of moral laws, as the world's highest good. Every rational person would still have to acknowledge himself strictly bound by morality's prescription. For its laws are formal and command unconditionally, without regard to ends as the material of willing. But one requirement of the final end, as practical reason prescribes it to worldly beings, is an irresistible end placed in them by their nature as A 426 finite beings. Reason wants only to subject this end to the moral law as an inviolable condition, or also to make it universal according to that law. It thus makes promoting happiness in agreement with morality the final end. The moral law commands us to promote this as far as lies within our power, as regards happiness, whatever outcome this effort may have. Fulfilling duty consists in the form of an earnest will, not in the intermediate causes of success.

Suppose, then, that a person persuades himself that there is no God, moved partly by the weakness of all the much-praised speculative arguments, partly by various irregularities he encounters in nature and the moral world. He would still be worthless in his own eyes if he therefore chose to regard the laws of duty as merely imagined, invalid, and nonbinding, and resolved to violate them without hesitation. With that way of thinking, he would remain worthless even if he later became convinced of what he had initially doubted. He might fulfill his duty, but from fear or desire for reward, without 452 a disposition that honors duty, as punctiliously in its effects as could ever A 427 be demanded. Conversely, suppose that as a believer he follows duty sincerely and unselfishly to the best of his awareness, yet whenever he considers the hypothetical case of one day becoming convinced that there is no God, he immediately thinks himself free of all moral obligation. His inner moral disposition would then be in a very poor state.

We can therefore suppose an upright person, such as Spinoza, firmly convinced that there is no God and no future life, since this has the same consequence regarding morality's object. How will he judge the inner determination of his own end through the moral law, which he actively honors? He seeks no advantage for himself from following it, either in this world or another. Rather, he wants unselfishly only to establish the good toward which that holy law directs all his powers. But his striving is limited. From nature he can expect occasional, accidental support, but never a lawful agreement, occurring according to constant rules—as his maxims inwardly are and must be—with the end he nevertheless feels bound and driven to bring about. Deceit, violence, and envy will always be rife around him, although A 428 he himself is honest, peaceable, and benevolent. The upright people he encounters will, despite all their worthiness to be happy, still be subjected by nature, which takes no account of this, to all the evils of deprivation, disease, and untimely death, just like the earth's other animals. They will always remain so until one vast grave swallows them all, honest and dishonest alike—it makes no difference here—and throws them, who could believe themselves creation's final end, back into the abyss of the purposeless chaos of matter from which they were drawn.—He would therefore have to abandon as impossible the end that this well-disposed person had, and ought to have had, in view when following moral laws. Or suppose he wants to remain faithful here too to the call of his inner moral vocation, and not weaken the respect that the moral law directly inspires in him for obedience by declaring null the only ideal final end adequate to its high demand. This weakening cannot occur without harming his moral disposition. 453 He must then assume, for practical purposes, the existence of a moral author of the world, that is, A 429 God, so as at least to form a concept of the possibility of the final end morally prescribed to him. He can perfectly well do this, since it is at least not contradictory in itself.

88 Limitation of the validity of the moral proof.

Pure reason, as a practical capacity—that is, the capacity to determine the free use of our causality through ideas, or pure concepts of reason—contains in the moral law not only a regulative principle for our actions. Through it, reason also supplies a subjectively constitutive principle in the concept of an object that reason alone can think and that is to be made actual in the world by our actions according to that law. The idea of a final end in using freedom according to moral laws therefore has subjective practical reality. Reason determines us a priori to promote with all our powers the world's highest good. This consists in connecting the greatest well-being of rational worldly beings with the supreme condition of goodness in them: universal happiness with the strictest conformity to morality. In this final end, the possibility of one part, happiness, is empirically conditioned. It depends on nature's character, whether nature agrees with this end or not, and is problematic theoretically A 430. The other part, morality, in relation to which we are independent of nature's cooperation, is established a priori as regards its possibility and is dogmatically certain. The objective theoretical reality of the concept of rational worldly beings' final end therefore requires not only that we have a final end set for us a priori, but also that creation, the world itself, have a final end with regard to its existence. If this could be proved a priori, it would add objective reality to the subjective reality of the final end. For if creation has any final end, we can think it only as necessarily agreeing with the moral end, which alone makes the concept of an end possible 454. Now we certainly find ends in the world. Physical teleology displays them in such abundance that, if we judge according to reason, we ultimately have grounds to adopt as a principle of natural inquiry that nothing in nature is without an end. Yet we search in vain within nature itself for nature's final end. Just as its idea lies only in reason, so this final end, even as regards its objective possibility, can and must be sought only in rational beings. Their practical reason not only specifies this final end but also determines its concept with regard to the conditions under which alone we A 431 can think a final end of creation.

The question now is whether the objective reality of the concept of creation's final end can also be demonstrated sufficiently for pure reason's theoretical requirements: if not with demonstrative certainty for determining judgment, then at least sufficiently for the maxims of theoretically reflective judgment. This is the least we can ask of speculative philosophy, which undertakes to connect the moral end with natural ends through the idea of a single end. But even this little is far more than it can ever accomplish.

According to the principle of theoretically reflective judgment, we would say: suppose we have grounds to assume, for nature's purposive products, a supreme cause of nature whose causality regarding nature's actuality—creation—must be thought as different from what nature's mechanism requires, namely as the causality of an understanding. We would then also have sufficient grounds to think in this original being not merely ends everywhere in nature but also a final end. Even if this did not demonstrate such a being's existence, it would at least, as in physical teleology, convince us that we can make the possibility A 432 of such a world understandable to ourselves only through ends, and moreover only by placing a final end beneath its existence.

But a final end is solely a concept of our practical reason. It cannot be inferred from any data of experience for the theoretical judgment of nature, or be related to knowledge of nature. No use of this concept is possible except for practical 455 reason alone, according to moral laws. Creation's final end is that character of the world that agrees with the only thing we can specify determinately according to laws: the final end of our pure practical reason, insofar as it is to be practical.—Now the moral law, which imposes this latter end on us, gives us a ground for assuming its possibility and feasibility for practical purposes, namely for applying our powers to bring it about. We may therefore also assume a nature of things agreeing with it. For realizing it would be impossible without nature's cooperation in a condition of it that is beyond our control. So we have a moral ground for thinking in a world a final end of creation as well.

This is not yet the inference from moral teleology A 433 to theology, that is, to the existence of a moral author of the world. It is only an inference to a final end of creation determined in this way. That this creation—the existence of things according to a final end—requires us to assume an author of the world who is, first, intelligent, but second, not merely intelligent as required for the possibility of natural things we were compelled to judge as ends, but also moral, and thus a God: this is a second inference. We can see from its character that it is made solely for judgment according to concepts of practical reason, and as such for reflective, not determining, judgment. For we cannot claim to understand that because moral-practical reason within us differs essentially in its principles from technical-practical reason, this must also be so in the supreme world cause, if we assume it to be an intelligence. We cannot claim that its final end requires a special and different kind of causality from that required merely for natural ends. We therefore cannot claim that our final end gives us not merely a moral ground for assuming a final end of creation as an effect, but also grounds for assuming a moral being as creation's original ground. But we can say this: given the character of our rational capacity A 434, we cannot make the possibility of such purposiveness related to the moral law and its object, as found in this final end, understandable to ourselves at all without an author and ruler of the world who is also a moral legislator.

456 The actuality of a highest morally legislating author is therefore sufficiently demonstrated only for the practical use of our reason, without determining anything theoretically about his existence. For practical reason needs an idea to make possible its end, which is assigned to us by its own legislation even without this idea. Through that idea, the obstacle arising from our inability to follow reason on the mere natural concept of the world is removed sufficiently for reflective judgment. This gives the idea practical reality, although for speculative knowledge it completely lacks every means of acquiring such reality theoretically, to explain nature and determine the supreme cause. For theoretically reflective judgment, physical teleology sufficiently proved an intelligent world cause from nature's ends. For practical judgment, moral teleology does this through the concept of a final end that it is compelled to attribute to creation for practical purposes. The A 435 objective reality of the idea of God as a moral author of the world cannot, admittedly, be demonstrated through physical ends alone. Nevertheless, when knowledge of them is combined with knowledge of the moral end, they are of great importance. By pure reason's maxim of pursuing unity of principles as far as possible, they support this idea's practical reality through the reality it already has theoretically for judgment.

To prevent a misunderstanding that can easily arise, it is essential to note two things here. First, we can think these properties of the highest being only by analogy. For how could we investigate its nature, when experience can show us nothing similar? Second, through this analogy we can only think it, not know it accordingly or attribute those properties to it theoretically. That would be a task for determining judgment in reason's speculative use, to understand what the supreme world cause is in itself. But here the question is only what concept we must form of it, given the character of our cognitive powers, and whether we must assume its existence in order to give merely practical reality to an end. Pure practical reason imposes on us a priori, without any such presupposition, the duty of realizing this end with all our powers. We need only A 436 be able to think of an intended effect as possible. That concept may well exceed the bounds of speculative reason 457. The properties we attribute to the being thought through it may also conceal an anthropomorphism if used objectively. But their use is not intended to determine its nature, which is inaccessible to us, but to determine ourselves and our will accordingly. We name a cause according to our concept of its effect, but only regarding its relation to that effect, without thereby intending to determine its inner character through properties that alone we know in causes of this sort and that must be given through experience. For example, we attribute to the soul, among other things, a power of locomotion, because bodily movements actually arise whose cause lies in its representations. We do not thereby intend to attribute to it the only kind of moving forces we know: attraction, pressure, impact, and thus motion, which always presuppose an extended being. In just the same way, we must assume something containing the ground of the possibility and practical reality—that is, feasibility—of a necessary moral final end. Given the character of the effect expected from it, we can think of this something as a wise being governing the world according to moral A 437 laws. Given the character of our cognitive powers, we must think of it as a cause of things distinct from nature. This expresses only the relation of this being, which exceeds all our cognitive powers, to the object of our practical reason. We must not thereby intend to attribute to it theoretically the only causality of this kind known to us, an understanding and a will. We must not even intend to distinguish objectively, within this being itself, its causality regarding what is a final end for us from its causality regarding nature and nature's determinations of ends in general. We can assume this distinction only as subjectively necessary given the character of our cognitive power, and as valid for reflective, not objectively determining, judgment. But in practical matters, such a regulative principle for prudence or wisdom is also constitutive, that is, practically determining: we are to act in accordance with that as an end which, given our cognitive powers, we can think possible only in a certain way. Yet this same principle, as a principle for judging things' objective possibility, is by no means theoretically determining. It does not establish that 458 the object itself has only the kind of possibility that our A 438 capacity can think. It is merely a regulative principle for reflective judgment.

Remark.

This moral proof is not newly invented; at most, its ground has only been newly clarified. For it already lay in the human rational capacity before its earliest awakening, and is only developed further as that capacity's culture progresses. As soon as people began reflecting on right and wrong, they inevitably arrived at the judgment that the final outcome could never be indifferent to whether someone had acted honestly or deceitfully, fairly or violently. This happened at a time when they still passed indifferently over nature's purposiveness, using nature without thinking of anything but its accustomed course, even if, at least visibly, they found no good fortune for their virtues or punishment for their crimes up to the end of life. It is as though they heard a voice within themselves saying that things must turn out differently. Thus there must also have been a hidden, though obscure, representation of something they felt bound to pursue, with which such an outcome could not be reconciled. Or, if they regarded the course of the world as the sole order of things, they did not know how to reconcile that inner determination of their mind's end with it. They might imagine in many very A 439 crude ways how such an irregularity could be corrected—an irregularity that must outrage the human mind far more than the blind chance someone might propose as a principle for judging nature. Yet they could never conceive any principle for the possibility of uniting nature with their inner moral law other than a supreme cause governing the world according to moral laws. For a final end assigned within them as a duty contradicts a nature outside them without any final end, in which that end must nevertheless become actual. They could hatch much nonsense about this world cause's inner character. But that moral relation in the world's government always remained the same, universally understandable even to the most uncultivated reason insofar as it regards itself as practical. Speculative reason, by contrast, cannot 459 come close to keeping pace with it.—In all probability, this moral interest also first awakened attention to nature's beauty and ends. These then served excellently to strengthen that idea, but could neither ground it nor, still less, dispense with that moral interest. For even the investigation of natural ends acquires only through its relation to the final end that immediate interest displayed so greatly in our admiration of them without regard to any advantage to be gained.

89 On the usefulness of the moral argument.

Restricting reason, in all our ideas of the supersensible, to the conditions of its practical use has an unmistakable benefit A 440 regarding the idea of God. It prevents theology from soaring into theosophy, into excessive concepts that confuse reason, or sinking into demonology, an anthropomorphic way of representing the highest being. It prevents religion from falling into theurgy, the extravagant delusion that we can feel other supersensible beings and in turn influence them, or into idolatry, the superstitious delusion that we can please the highest being through means other than a moral disposition³⁵.

For suppose we allow the vanity or presumption of reasoning to determine even the slightest thing theoretically, extending knowledge about what lies beyond the sensible world. Suppose we permit people to boast of insights into the existence and character of the A 441 divine nature, its understanding and will, the laws of both, and the properties flowing from these into the world. I would then like to know where and at what point we mean to 460 limit reason's pretensions. For wherever those insights came from, still more can surely be expected from the same source, if we only exert our reflection, as people think. Such claims would have to be limited according to some principle, not merely because we find that all attempts so far have failed. That proves nothing against the possibility of better success. But no principle is possible here except either to assume that absolutely nothing can be determined theoretically about the supersensible, other than purely negatively, or to assume that our reason contains a still unworked mine of who knows how much expanded knowledge, preserved for us and our descendants.—As for religion, morality in relation to God as legislator, if theoretical knowledge of God had to come first, morality would have to conform to theology. Not only would an external, arbitrary legislation of a supreme being replace reason's inner, necessary legislation, but every deficiency in our insight into A 442 that being's nature would also extend into moral prescription, making religion immoral and corrupting it.

As for the hope of a future life, suppose we consult our theoretical cognitive power instead of using the final end we ourselves must realize according to the moral law as the guide for reason's judgment of our vocation—a judgment regarded as necessary or worthy of acceptance only in a practical relation. For this purpose the study of the soul, like theology above, gives nothing more than a negative concept of our thinking being: none of its actions or the appearances of inner sense can be explained materialistically. Thus our entire theoretical cognitive power cannot, on speculative grounds, make any judgment extending or determining our knowledge of its separate nature or of the continuation or noncontinuation of its personality after death. Everything here is left to teleological judgment of our existence in a practically necessary respect, and to assuming our continued existence as a condition required for the final end that reason unconditionally assigns us. This also shows the benefit, though at first it seems a loss: just as theology can never become theosophy for us A 443 461, rational psychology can never become pneumatology, a science extending our knowledge of spirits. On the other hand, it is also protected from falling into materialism. Rather, it is merely an anthropology of inner sense: knowledge of our thinking self in life. As theoretical knowledge, it also remains merely empirical. By contrast, regarding the question of our eternal existence, rational psychology is no theoretical science at all. It rests on a single inference of moral teleology, and its entire use is necessary only for the sake of that teleology as our practical vocation.

90 On the kind of acceptance as true in a teleological proof of God's existence.

First, every proof must not merely persuade, but convince, or at least work toward conviction. This applies whether the proof proceeds through a direct empirical presentation of what is to be proved, as through observation of the object or experiment, or through reason a priori from principles. That is, the ground of proof or inference must not be merely a subjective A 444, aesthetic determining ground of assent, a mere semblance. It must be objectively valid and a logical ground of knowledge. Otherwise the understanding is beguiled, not convinced. One such deceptive proof is offered in natural theology, perhaps with good intentions but still with deliberate concealment of its weakness. It draws on the great mass of evidence that natural things originate according to the principle of ends and exploits a merely subjective ground of human reason: its characteristic tendency to substitute one principle for many wherever this can be done without contradiction. Where only some or even many requirements for determining a concept are found in this principle, reason adds the rest in thought, completing the concept of the thing by arbitrary supplementation. After all, if we encounter so many products in nature that indicate an intelligent cause to us, why should we not think of one such cause rather than many? And why should we not attribute to it not merely great understanding 462, power, and so on, but all-wisdom and omnipotence—in a word, think of it as containing the sufficient ground of such properties for all possible things? Moreover, why attribute to this single all-powerful original being not only understanding for nature's laws and products A 445, but also, as a moral world cause, the highest moral practical reason? Completing the concept in this way gives a principle sufficient both for insight into nature and for moral wisdom, and no even moderately grounded objection can be made against such an idea's possibility. If the mind's moral incentives are also set in motion, and a lively interest in them is added with rhetorical force, of which they are indeed worthy, the result is persuasion of the proof's objective adequacy. A semblance arises that is also beneficial in most cases of its use. It lifts itself entirely above any examination of the proof's logical rigor and even regards such examination with abhorrence and aversion, as if it rested on an impious doubt.—There may be nothing to say against this insofar as popular usefulness is the real concern. But we cannot and must not prevent the argument from being divided into its two different components: what belongs to physical teleology and what belongs to moral teleology. Fusing them makes it impossible to recognize where the proof's real force lies, and which part must be worked on, and how, to make its validity withstand the strictest examination A 446, even if we should have to acknowledge the weakness of our rational insight in one part. It is therefore the philosopher's duty, even if he thought nothing of the demand for honesty made on him, to expose the semblance such a mixture can produce, however beneficial it may be. He must separate what belongs merely to persuasion from what leads to conviction—two determinations of assent differing not merely in degree but in kind. This allows him openly to present the mind's attitude in this proof in its full purity and freely submit it to the strictest examination.

A proof intended to convince can itself be of two kinds. It can establish either what the object is in itself, or what it is for us, human beings in general, according to the 463 rational principles necessary for our judgment of it: a proof “according to truth” or “according to the human being,” the latter expression taken generally for humanity as a whole. In the first case it rests on sufficient principles for determining judgment; in the second, only for reflective judgment. In the latter case, if it rests on merely theoretical principles, it can never work toward conviction. But if A 447 it bases itself on a practical principle of reason, which therefore holds universally and necessarily, it may claim a conviction sufficient for purely practical purposes: a moral conviction. A proof works toward conviction without yet convincing when it is merely on the way there: it contains objective grounds that, although not yet sufficient for certainty, are nevertheless of a kind that they do not merely serve persuasion as subjective grounds of judgment.

Now all theoretical grounds of proof are sufficient either 1) for proof through logically strict inferences of reason; or, failing that, 2) for an inference by analogy; or, if even that is unavailable, 3) for a probable opinion; or finally, at the lowest level, 4) for assuming a merely possible ground of explanation as a hypothesis.—Now I say that no grounds of proof working toward theoretical conviction can produce acceptance as true of this kind, from its highest to its lowest degree, if what is to be proved is the existence of an original being as a God in the meaning adequate to the concept's entire content: as a moral author of the world, such that the final end of creation is also specified through him A 448.

1) Concerning logically correct proof, which proceeds from the universal to the particular, the Critique has sufficiently shown the following. No intuition possible for us corresponds to the concept of a being to be sought beyond nature. Thus this concept itself, insofar as it is to be theoretically determined through synthetic predicates, always remains problematic for us. There can therefore be absolutely no knowledge of it that would extend the scope of our theoretical knowledge in the least. Nor can the particular concept of a supersensible being be subsumed under the universal principles of the nature of things, so that we could infer from those principles to 464 this being. Those principles hold solely for nature as an object of the senses.

2) With two things of different kinds, we can indeed think of one by analogy³⁶ with the other precisely at the point where they differ. But we cannot A 449 infer from one to the other by analogy in what makes them different, transferring this feature A 450 of their specific difference to the other. Thus, by analogy with the law of equality of action and reaction in bodies' mutual attraction and repulsion 465, I can also think of the association of a community's members according to rules of right. But I cannot transfer those specific determinations, material attraction or repulsion, to the members and attribute them to citizens in order to constitute the system called a state.—Likewise, we may think of the original being's causality regarding worldly things as natural ends by analogy with an understanding as the ground of the forms of certain products we call works of art. We do this only for the theoretical or practical use of our cognitive power that we must make of this concept regarding natural things in the world according to a certain principle. But from the fact that, among A 451 worldly beings, understanding must be attributed to the cause of an effect judged as artfully produced, we cannot infer by analogy that the being entirely distinct from nature also possesses, in relation to nature itself, the very same causality we perceive in human beings. For this concerns precisely the point of difference between a cause sensibly conditioned in its effects and the supersensible original being, a difference thought in the very concept of the latter and therefore not transferable to it.—The requirement that I think divine causality only by analogy with an understanding, a capacity we know in no being other than the sensibly conditioned human being, itself contains the prohibition against attributing understanding to God in its proper meaning³⁷.

3) Opinion has no place at all in a priori judgments. Through them we know either something entirely certain or nothing at all. Even when the given grounds of proof from which we start are empirical, as here from ends in the world, we still cannot use them A 452 to hold opinions beyond the sensible world or grant such venturesome judgments the slightest claim to probability. For probability is a part of a certainty possible within a certain series of grounds. Its grounds are compared with a sufficient ground as parts with a whole, into which that insufficient 466 ground must be capable of being supplemented. But as grounds determining the certainty of one and the same judgment, they must be of the same kind. Otherwise they would not together make up a magnitude such as certainty is. Thus one part cannot lie within the bounds of possible experience and another outside all possible experience. Mere empirical grounds of proof lead to nothing supersensible, and their deficient series cannot be supplemented by anything. In attempting through them to reach the supersensible and knowledge of it, therefore, there is not the slightest approach toward the goal. Consequently, in a judgment about the supersensible through arguments drawn from experience, there is no probability either.

4) Anything to serve as a hypothesis explaining the possibility of a given appearance must at least be completely certain in its possibility. It is enough that with a hypothesis I renounce knowledge of actuality, which is still asserted in an opinion offered as probable A 453. I cannot give up more than this. The possibility of what I base an explanation on must at least be beyond doubt; otherwise there would be no end to empty fantasies. But assuming the possibility of a supersensible being determined according to certain concepts would be a wholly groundless presupposition. For none of the conditions required for knowledge insofar as it rests on intuition is given here. So the mere principle of contradiction remains as the criterion of this possibility, although it can prove only the possibility of thinking, not of the object thought itself.

The result is that human reason can produce absolutely no theoretical proof of the existence of the original being as a deity, or of the soul as an immortal spirit, sufficient to bring about even the slightest degree of acceptance as true. The reason is entirely understandable: we have no material at all for determining ideas of the supersensible. We would have to take such material from things in the sensible world, but it is wholly inadequate to that object. So without any determination of these ideas, nothing remains but the concept of a nonsensible something containing the ultimate ground of the sensible world A 454. This does not yet constitute knowledge of its inner character through an extension of the concept.

91 467 On the kind of acceptance as true through practical faith.

Suppose we look only at how something can be an object of knowledge for us, given the subjective character of our powers of representation—a knowable thing. We then compare concepts not with objects, but only with our cognitive powers and the use these can make of the given representation for theoretical or practical purposes. The question whether something is a knowable being is not about the possibility of things themselves, but about our knowledge of them.

Knowable things are of three kinds: matters of opinion (opinabile), matters of fact (scibile), and matters of faith (mere credibile).

1) Objects of mere ideas of reason that cannot be presented in any possible experience for theoretical knowledge are, to that extent, not knowable things at all. We therefore cannot even hold an opinion about them. Indeed, holding an opinion a priori A 455 is already absurd in itself and leads directly to sheer fantasies. Thus either our a priori proposition is certain, or it contains nothing to accept as true. Matters of opinion are therefore always objects of experiential knowledge possible at least in itself—objects of the sensible world—but impossible for us given the mere degree of this capacity we possess. Thus the ether of modern physicists, an elastic fluid permeating all other matter and intimately mixed with it, is a mere matter of opinion. It is still of such a kind that, if the outer senses were sharpened to the highest degree, it could be perceived; yet it can never be presented in any observation or experiment. Assuming rational inhabitants of other planets is a matter of opinion. For if we could approach those planets, which is possible in itself, we would determine through experience whether such inhabitants exist. But we will never get so close, so it remains opinion. Yet holding the opinion that pure spirits, thinking without bodies, exist in the material universe—if, as is proper, we reject certain alleged actual appearances of them—is invention, not 468 a matter of opinion at all. It is a mere idea left over when we remove everything material from a thinking being while still leaving it thought. But whether the latter, which A 456 we know only in human beings, in connection with a body, would then remain, we cannot determine. Such a thing is a being fabricated by reasoning (ens rationis ratiocinantis), not a rationally grounded being (ens rationis ratiocinatae). With the latter, it is possible to demonstrate its concept's objective reality sufficiently at least for reason's practical use. For this use, which has its own distinctive and demonstratively certain a priori principles, even requires or postulates it.

2) Objects for concepts whose objective reality can be proved are matters of fact (res facti)³⁸. Such proof may come through pure reason or experience, and in the former case from theoretical or practical data of reason, but always through a corresponding intuition. The mathematical properties of magnitudes in geometry are of this kind because they can be presented a priori for reason's theoretical use. Further, A 457 things or their properties that can be demonstrated through experience—our own, or others' through testimony—are likewise matters of fact.—But remarkably, even an idea of reason appears among matters of fact, although in itself it cannot be presented in intuition and thus cannot admit a theoretical proof of its possibility. This is the idea of freedom. Its reality as a special kind of causality, whose concept would exceed the bounds of theoretical reason, can be demonstrated through practical laws of pure reason and, in accordance with them, in actual actions, hence in experience.—It is the only idea of pure reason whose object is a matter of fact and must be counted among things knowable.

469 3) Objects that must be thought a priori in relation to the dutiful use of pure practical reason, whether as consequences or grounds, but that exceed the bounds of its theoretical use, are mere matters of faith. Such an object is the highest good to be brought about in the world through freedom. Its concept cannot be sufficiently proved as to its objective reality in any experience possible for us, and thus for reason's theoretical use. Yet pure practical reason commands its use to realize that end as well as possible, so it must be assumed A 458 to be possible. This commanded effect, together with the only conditions of its possibility thinkable for us—God's existence and the soul's immortality—are matters of faith (res fidei). Indeed, they are the only objects that can be so called³⁹. For although we must believe what we can learn only from others' experience through testimony, that does not yet make it a matter of faith in itself. For one of those witnesses, at least, it was personal experience and a matter of fact, or is presupposed as such. Moreover, it must be possible to reach knowledge by this path of historical belief. The objects of history and geography, like everything that our cognitive powers can at least possibly know, belong not to matters of faith but to matters of fact. Only objects of pure reason can possibly be matters of faith A 459, but not as objects of merely pure speculative reason. For as such they cannot even be counted with certainty among things, that is, objects of knowledge possible for us. They are ideas: concepts whose objective reality cannot be secured theoretically. By contrast, the highest final end to be realized by us, through which alone we can become worthy of being a creation's final end ourselves, is an idea having objective reality for us practically, and is a thing. But because we cannot give this concept that reality for 470 theoretical purposes, it is a mere matter of faith of pure reason. So too, along with it, are God and immortality, as the conditions under which alone, given the character of our human reason, we can think the possibility of that effect of the lawful use of our freedom. Acceptance as true in matters of faith is acceptance for purely practical purposes: a moral faith. It proves nothing for theoretical knowledge, but only for practical pure rational knowledge directed toward following one's duties. It does not extend speculation, or practical rules of prudence based on self-love, at all. If the supreme principle of all moral laws is a postulate, then the possibility of their highest object is also thereby postulated, and with it the condition A 460 under which we can think this possibility. Knowledge of this condition thereby becomes neither knowledge nor opinion about these conditions' existence and character as a theoretical kind of cognition. It is merely an assumption in a practical and commanded relation, for the moral use of our reason.

Even if we could apparently ground a determinate concept of an intelligent world cause on the natural ends that physical teleology presents so abundantly, this being's existence would still not be a matter of faith. For since it is assumed not to fulfill my duty but only to explain nature, it would be merely the opinion and hypothesis most suited to our reason. But that teleology by no means leads to a determinate concept of God. Such a concept is found only in the concept of a moral author of the world, because this alone specifies the final end. We can count ourselves as belonging to this end only insofar as we conform to what the moral law imposes on us as a final end, and thus obliges us to do. Consequently, only through relation to the object of our duty, as a condition of the possibility of attaining its final end, does the concept of God gain the privilege of counting as a matter of faith in our acceptance as true. Yet this same A 461 concept cannot establish its object as a matter of fact. Although the necessity of duty is clear to practical reason, attaining its final end, insofar as this is not entirely within our control, is assumed only for the practical use of 471 reason. It is therefore not practically necessary in the way duty itself is⁴⁰.

Faith, as a habitual disposition, not an act, is reason's moral way of thinking A 462 in accepting as true what is inaccessible to theoretical knowledge. It is therefore the mind's steadfast principle of accepting as true, because of our obligation to the highest moral final end, what must be presupposed as a condition of that end's possibility⁴¹, although we cannot understand either its possibility 472 or, equally, its impossibility A 463. Faith, so called without qualification, is trust in attaining an intention that it is our duty to promote, but whose realization we cannot understand as possible, and consequently neither can we understand the possibility of its only conditions thinkable for us. Thus faith relating to particular objects that are not objects of possible knowledge or opinion is entirely moral. In the latter case, especially in historical matters, it would have to be called credulity rather than faith. Faith is a free acceptance as true, not of what admits dogmatic proofs for theoretically determining judgment, nor of what we consider ourselves obliged to accept, but of what we assume to serve an intention according to laws of freedom. Yet it is not like an opinion without sufficient grounds. It is grounded in reason, though only with regard to its practical use A 464, sufficiently for reason's intention. For without it, the moral way of thinking has no firm persistence when it encounters theoretical reason's demand for proof of the possibility of morality's object. Instead, it wavers between practical commands and theoretical doubts. To be distrustful means to follow the maxim of not believing testimony in general. But an unbeliever denies all validity to those ideas of reason because they lack a theoretical grounding of their reality. He therefore judges dogmatically. Dogmatic unbelief cannot coexist with a moral maxim governing one's way of thinking, since reason cannot command pursuit of an end recognized as nothing but a fantasy. But a doubting faith can. Here the lack of conviction through grounds of speculative reason is only an obstacle. Critical insight into that reason's limits can remove the obstacle's influence on conduct and 473 replace it with a predominating practical acceptance as true.

When we want to put another principle in place of certain failed philosophical attempts and give it influence, it is deeply satisfying to understand how those attempts failed and why A 465 they had to fail.

God, freedom, and the soul's immortality are the problems toward whose solution all the preparations of metaphysics aim, as their ultimate and sole end. Now people believed the doctrine of freedom was needed only as a negative condition for practical philosophy, whereas the doctrines of God and the soul's character belonged to theoretical philosophy and had to be demonstrated by themselves and separately. They were then to be connected with what the moral law commands, which is possible only under the condition of freedom, thereby producing a religion. But we can soon see that these attempts had to fail. From mere ontological concepts of things in general, or of the existence of a necessary being, we can form absolutely no concept of an original being determined through predicates that can be given in experience and thus serve knowledge. In turn, a concept grounded on experience of nature's physical purposiveness could supply no proof sufficient for morality, and therefore for knowledge of a God. Nor could experiential knowledge of the soul, which we acquire only in this life, provide a concept of its spiritual, immortal nature sufficient for morality. Theology and pneumatology, as A 466 tasks for speculative reason's sciences, cannot be established through any empirical data or predicates, because their concept exceeds all our cognitive powers.—Both concepts, God and the soul regarding its immortality, can be determined only through predicates that, although themselves possible only from a supersensible ground, must nevertheless prove their reality in experience. Only in this way can they make knowledge of wholly supersensible beings possible.—The sole concept of this kind found in human reason is 474 that of human freedom under moral laws, together with the final end that reason prescribes through these laws. The laws allow us to attribute to nature's author, and the final end to the human being, those properties containing the necessary condition for the possibility of both. Thus, from this very idea we can infer the existence and character of those beings otherwise entirely hidden from us.

Thus the reason the intention to prove God and immortality fails on the merely theoretical path is that no knowledge of the supersensible is possible along this path of natural concepts. Its success on the moral path, that of the concept of freedom A 467, has this ground: here the supersensible underlying it, freedom, supplies material for knowledge of the other supersensible objects—the moral final end and the conditions of its feasibility—through a determinate law of causality arising from it. It also demonstrates its reality in actions as a matter of fact. But for that very reason it can provide a valid ground of proof only for practical purposes, which are also the only purposes religion needs.

It always remains remarkable that among the three ideas of pure reason—God, freedom, and immortality—freedom is the only concept of the supersensible that proves its objective reality in nature, through the causality thought in it and its possible effect within nature. Precisely through this it makes possible the connection of the other two with nature, and of all three with one another in a religion. We therefore have within us a principle capable of determining the idea of the supersensible within us, and thereby also of the supersensible outside us, into knowledge, although possible only for practical purposes. Merely speculative philosophy, which could also give only a negative concept of freedom, had to despair of this. Thus the concept of freedom, as the fundamental concept A 468 of all unconditionally practical laws, can extend reason beyond the boundaries within which every theoretical concept of nature would have to remain confined without hope.

29 A hypothesis of this kind may be called a daring adventure of reason, and few even among the most perceptive investigators of nature may never have had it cross their minds. For it is not actually absurd, as spontaneous generation is, by which we mean the production of an organized being through the mechanics of raw, unorganized matter. It would still be generation from an organism in the broadest sense of the term, provided only that something organic was generated from something else organic, though specifically different within this kind of being. For example, certain aquatic animals might gradually develop into marsh animals, and these after several generations into land animals. A priori, in the judgment of mere reason, there is no contradiction in this. But experience shows no example of it. Rather, all generation we know is generation of the same kind. It is not merely generation from an organism, as opposed to generation from unorganized material, but also brings forth a product of the same kind as its parent in its very organization. And generation of a different kind is found nowhere within the reach of our experiential knowledge of nature.

30 If the established name natural history is to remain in use for the description of nature, we can call what that name literally indicates—the representation of the earth's former, ancient state, about which, though we cannot hope for certainty, we venture well-grounded conjectures—the archaeology of nature, as distinct from that of art. Fossils would belong to the former, just as engraved stones and so on belong to the latter. For since work on such an inquiry, under the name of a theory of the earth, does in fact continue steadily, though properly slowly, the name would not designate a merely imagined investigation of nature, but one to which nature itself invites and urges us.

31 The value life has for us when assessed only by what we enjoy—the natural end of the sum of all inclinations, happiness—is easy to decide. It falls below zero. For who would choose to begin life again under the same conditions, or even according to a new, self-devised plan, though consistent with nature's course, if it too aimed only at enjoyment? The value life has through what it contains when led according to nature's end for us has been shown above. This consists in what we do, not merely enjoy, though we still remain only means to an undetermined final end. Thus nothing seems left but the value we ourselves give our lives through what we not only do, but do purposively in such independence from nature that even nature's existence can be an end only under this condition.

32 It would be possible for the happiness of rational beings in the world to be an end of nature, and it would then also be nature's ultimate end. At least we cannot understand a priori why nature should not be arranged this way, since its mechanism could, as far as we can understand, produce this effect. But morality and a causality according to ends subordinated to it are absolutely impossible through natural causes. For the principle determining them to act is supersensible. Thus it is the only possibility in the order of ends that is wholly unconditioned with regard to nature, and it alone thereby qualifies its subject as creation's final end, to which the whole of nature is subordinated.—Happiness, by contrast, as shown from experience in the preceding section, is not even an end of nature with regard to human beings in preference to other creatures, far less a final end of creation. Human beings may certainly make it their ultimate subjective end. But when I ask about creation's final end, “Why did human beings have to exist?” I am asking about an objective supreme end, such as the highest reason would require for its creation. If the answer is, “So that beings might exist to whom that supreme cause could do good,” this contradicts the condition to which human reason itself subjects its deepest wish for happiness: agreement with its own inner moral legislation. This proves that happiness can only be a conditioned end, and the human being can therefore be creation's final end only as a moral being. As for his condition, happiness is connected with it only as a consequence proportionate to agreement with that end as the end of his existence.

33 I deliberately say under moral laws. It is not the human being in accordance with moral laws, one who behaves in keeping with them, who is creation's final end. The latter expression would say more than we know: namely, that an author of the world has the power to make human beings always behave in accordance with moral laws. This presupposes a concept of freedom and nature—we can think of an external author only for the latter—that would have to contain insight into nature's supersensible substrate and its identity with what makes causality through freedom possible in the world. Such insight goes far beyond our rational understanding. Only of the human being under moral laws can we say, without exceeding the limits of our insight, that his existence constitutes the world's final end. This also agrees completely with the judgment of human reason reflecting morally on the world's course. We believe we perceive traces of a wise relation to ends even in evil, provided we see that the impious villain does not die before suffering the well-deserved punishment for his crimes. According to our concepts of free causality, good or evil conduct depends on us. But we place the highest wisdom in the world's government in this: the occasion for good conduct is provided, and for both kinds the outcome is decreed according to moral laws. The latter is what properly constitutes God's glory, which theologians therefore quite appropriately call creation's ultimate end.—It should also be noted that when we use the word creation, we mean nothing other than what has been said here: the cause of the existence of a world, or of the things, the substances, in it. This is also what the word's proper concept implies (the actualization of substance is creation). It therefore does not already carry the presupposition of a freely acting, and consequently intelligent, cause, whose existence we are only now seeking to prove.

34 This moral argument is not meant to supply an objectively valid proof of God's existence, or to prove to the doubting believer that there is a God. It is meant to show that if he wants to think morally consistently, he must include acceptance of this proposition among the maxims of his practical reason.—Nor does it mean that assuming the happiness of all rational worldly beings in proportion to their morality is necessary for morality, but that it is necessary through morality. It is therefore an argument sufficient subjectively, for moral beings.

35 Idolatry in the practical sense is still any religion that thinks of the highest being with properties according to which something other than morality could by itself be the suitable condition for conforming to his will in what human beings can do. For however purely and free of sensible images we may have conceived that concept theoretically, practically it is then still represented as an idol: anthropomorphically in the character of its will.

36 Analogy, in the qualitative sense, is the identity of the relation between grounds and consequences, or causes and effects, insofar as this relation holds despite the specific difference of the things themselves, or of the properties containing the ground of similar consequences, considered apart from this relation. Thus, when comparing animals' artful actions with human ones, we think of the ground of these effects in animals, which we do not know, as an analogue of the ground of similar human effects, reason, which we do know. At the same time, we mean to indicate that the ground of animals' capacity for art, called instinct, is actually specifically different from reason, yet has a similar relation to its effect, comparing beavers' construction with human construction.—But from the fact that human beings need reason for building, I cannot infer that beavers must also have it and call this an inference by analogy. Yet from animals' similar way of acting, whose ground we cannot directly perceive, compared with the human way, of which we are directly conscious, we can quite correctly infer by analogy that animals also act according to representations. They are not machines, as Descartes maintains, and despite their specific difference, belong with human beings to the same genus, living beings. The principle authorizing this inference lies in the identity of a ground for counting animals, regarding this determination, in the same genus as human beings as such, insofar as we compare them externally through their actions. There is an equal basis for reasoning. Likewise, by comparing the supreme world cause's purposive products in the world with human works of art, I can think its causality by analogy with an understanding. But I cannot infer by analogy that it has these properties. For the very principle making such an inference possible is missing here: the equal basis for reasoning that would place the highest being and the human being in one and the same genus as regards their causality. The causality of worldly beings, always sensibly conditioned, including causality through understanding, cannot be transferred to a being that shares with them no generic concept beyond that of a thing in general.

37 Nothing at all is thereby missing in our representation of this being's relations to the world, regarding either the theoretical or the practical consequences of the concept. Trying to investigate what it is in itself is an idle curiosity as purposeless as it is futile.

38 Here I believe I am right to extend the concept of a matter of fact beyond the word's ordinary meaning. For when speaking of things' relation to our cognitive powers, it is unnecessary, indeed unworkable, to restrict this expression to actual experience alone. A merely possible experience is already sufficient to speak of them simply as objects of a determinate kind of knowledge.

39 Matters of faith are not thereby articles of faith, if the latter means matters of faith whose profession, inward or outward, can be made obligatory. Natural theology contains no such articles. For since matters of faith cannot rest on theoretical proofs as matters of fact can, accepting them as true is free, and only as such is compatible with the subject's morality.

40 The final end that the moral law imposes on us to promote is not the ground of duty. That ground lies in the moral law, which, as a formal practical principle, guides categorically, regardless of the objects of the faculty of desire, the material of willing, and thus regardless of any end. This formal character of my actions—their subordination to the principle of universal validity—in which alone their inner moral worth consists, is entirely within our control. I can perfectly well abstract from the possibility or infeasibility of the ends I am obliged to promote according to that law, since only my actions' external value lies in them. They are never fully within my control; I can attend solely to what is my own doing. Yet the law of duty itself imposes the intention of promoting the final end of all rational beings: happiness insofar as it is possible in agreement with duty. But speculative reason has no insight into its feasibility, either through our own physical powers or through nature's cooperation. Rather, as far as we can reasonably judge, it must regard such a result of our good conduct from mere nature within and outside us, without assuming God and immortality, as a groundless and empty expectation, however well meant. If speculative reason could be completely certain of this judgment, it would have to regard the moral law itself as a mere deception of our reason practically. But speculative reason is fully convinced that the latter can never happen, while those ideas whose object lies beyond nature can be thought without contradiction. Thus, for the sake of its own practical law and the task that law imposes, hence morally, it must acknowledge those ideas as real so as not to contradict itself.

41 Faith is trust in the promise of the moral law, but not in a promise contained in that law. It is one I place in it, on a morally sufficient ground. For no law of reason can command a final end without reason also promising its attainability, though uncertainly, and thereby authorizing acceptance as true of the only conditions under which our reason can think it. The word faith itself already expresses this. It may seem questionable how this expression and this particular idea enter moral philosophy, since they were first introduced with Christianity, and adopting them might seem merely a flattering imitation of its language. But this is not the only case in which that remarkable religion, in the greatest simplicity of its presentation, has enriched philosophy with far more determinate and purer concepts of morality than philosophy had previously been able to supply. Once present, however, these concepts are freely approved by reason and accepted as ones it could and should have reached and introduced by itself.