A Critical Look at the Analytic of Pure Practical Reason

By the critical examination of a science, or of a portion of it, which constitutes a system by itself, I understand the inquiry and proof why it must have this and no other systematic form, when we compare it with another system which is based on a similar faculty of knowledge. Now practical and speculative reason are based on the same faculty, so far as both are pure reason. Therefore the difference in their systematic form must be determined by the comparison of both, and the ground of this must be assigned.

By a critical examination of a science, or of a part of it that forms a system in itself, I mean an investigation and proof of why it must have one systematic form and no other. This involves comparing it with another system based on a similar faculty of knowledge. Practical and speculative reason are based on the same faculty, insofar as both are pure reason. Therefore, we must compare them to determine how their systematic forms differ and explain the basis of that difference.

The Analytic of pure theoretic reason had to do with the knowledge of such objects as may have been given to the understanding, and was obliged therefore to begin from intuition and consequently (as this is always sensible) from sensibility; and only after that could advance to concepts (of the objects of this intuition), and could only end with principles after both these had preceded. On the contrary, since practical reason has not to do with objects so as to know them, but with its own faculty of realizing them (in accordance with the knowledge of them), that is, with a will which is a causality, inasmuch as reason contains its determining principle; since, consequently, it has not to furnish an object of intuition, but as practical reason has to furnish only a law (because the notion of causality always implies the reference to a law which determines the existence of the many in relation to one another); hence a critical examination of the Analytic of reason, if this is to be practical reason (and this is properly the problem), must begin with the possibility of practical principles a priori. Only after that can it proceed to concepts of the objects of a practical reason, namely, those of absolute good and evil, in order to assign them in accordance with those principles (for prior to those principles they cannot possibly be given as good and evil by any faculty of knowledge), and only then could the section be concluded with the last chapter, that, namely, which treats of the relation of the pure practical reason to the sensibility and of its necessary influence thereon, which is a priori cognisable, that is, of the moral sentiment. Thus the Analytic of the practical pure reason has the whole extent of the conditions of its use in common with the theoretical, but in reverse order. The Analytic of pure theoretic reason was divided into transcendental Aesthetic and transcendental Logic, that of the practical reversely into Logic and Aesthetic of pure practical reason (if I may, for the sake of analogy merely, use these designations, which are not quite suitable). This logic again was there divided into the Analytic of concepts and that of principles: here into that of principles and concepts. The Aesthetic also had in the former case two parts, on account of the two kinds of sensible intuition; here the sensibility is not considered as a capacity of intuition at all, but merely as feeling (which can be a subjective ground of desire), and in regard to it pure practical reason admits no further division.

The Analytic of pure theoretical reason dealt with knowledge of objects that may have been given to the understanding. It therefore had to begin with intuition and, since intuition is always sensible, with sensibility. Only then could it move on to concepts of the objects of this intuition. It could end with principles only after dealing with both intuition and concepts. Practical reason, by contrast, does not deal with objects in order to know them. It deals with its own capacity to bring objects into existence, in accordance with its knowledge of them. In other words, it deals with a will that acts as a cause because reason contains the principle that determines it. Consequently, practical reason does not have to provide an object of intuition. It has to provide only a law. For the concept of causality always implies a law that determines how the existence of multiple things relates to one another. A critical examination of the Analytic of reason must therefore begin with the possibility of a priori practical principles, if reason is to be practical. Whether reason can be practical is the very question at issue. Only after that can the examination turn to concepts of the objects of practical reason, namely absolute good and evil. It can then determine what is good and evil in accordance with those principles. Before these principles, no faculty of knowledge can possibly identify anything as good or evil. Only then can the section conclude with its final chapter. That chapter deals with the relation of pure practical reason to sensibility and its necessary influence on sensibility. This influence can be known a priori and is called moral feeling. Thus, the Analytic of pure practical reason covers the full range of conditions for its use, just as the theoretical Analytic does. But it covers them in reverse order. The Analytic of pure theoretical reason was divided into the transcendental Aesthetic and the transcendental Logic. The Analytic of pure practical reason is divided, in reverse order, into the Logic and the Aesthetic of pure practical reason. I use these terms only for the sake of analogy, since they are not quite suitable. In the theoretical case, the Logic was further divided into the Analytic of concepts and the Analytic of principles. Here, it is divided into the Analytic of principles and the Analytic of concepts. The theoretical Aesthetic also had two parts, because there are two kinds of sensible intuition. Here, sensibility is not considered as a capacity for intuition at all, but only as feeling. Feeling can be a subjective ground of desire. In this respect, pure practical reason allows no further division.

It is also easy to see the reason why this division into two parts with its subdivision was not actually adopted here (as one might have been induced to attempt by the example of the former critique). For since it is pure reason that is here considered in its practical use, and consequently as proceeding from a priori principles, and not from empirical principles of determination, hence the division of the analytic of pure practical reason must resemble that of a syllogism; namely, proceeding from the universal in the major premiss (the moral principle), through a minor premiss containing a subsumption of possible actions (as good or evil) under the former, to the conclusion, namely, the subjective determination of the will (an interest in the possible practical good, and in the maxim founded on it). He who has been able to convince himself of the truth of the positions occurring in the Analytic will take pleasure in such comparisons; for they justly suggest the expectation that we may perhaps some day be able to discern the unity of the whole faculty of reason (theoretical as well as practical) and be able to derive all from one principle, which, is what human reason inevitably demands, as it finds complete satisfaction only in a perfectly systematic unity of its knowledge.

It is also easy to see why this division into two parts, with its subdivision, was not used here. The example of the former critique might have tempted us to use it. But here we are considering pure reason in its practical use. It therefore proceeds from a priori principles, not from empirical principles that determine the will. So the Analytic of pure practical reason must be divided like a syllogism. It begins with the universal in the major premise, the moral principle. It then moves through a minor premise that classifies possible actions as good or evil under that principle. Finally, it reaches the conclusion: the subjective determination of the will. This determination is an interest in the possible practical good and in the maxim based on it. Anyone who has convinced themselves of the truth of the positions in the Analytic will enjoy such comparisons. For these comparisons rightly suggest that we may perhaps one day discern the unity of the whole faculty of reason, both theoretical and practical. We may then be able to derive everything from one principle. Human reason inevitably demands this, since it finds complete satisfaction only in a perfectly systematic unity of its knowledge.

If now we consider also the contents of the knowledge that we can have of a pure practical reason, and by means of it, as shown by the Analytic, we find, along with a remarkable analogy between it and the theoretical, no less remarkable differences. As regards the theoretical, the faculty of a pure rational cognition a priori could be easily and evidently proved by examples from sciences (in which, as they put their principles to the test in so many ways by methodical use, there is not so much reason as in common knowledge to fear a secret mixture of empirical principles of cognition). But, that pure reason without the admixture of any empirical principle is practical of itself, this could only be shown from the commonest practical use of reason, by verifying the fact, that every man's natural reason acknowledges the supreme practical principle as the supreme law of his will- a law completely a priori and not depending on any sensible data. It was necessary first to establish and verify the purity of its origin, even in the judgement of this common reason, before science could take it in hand to make use of it, as a fact, that is, prior to all disputation about its possibility, and all the consequences that may be drawn from it. But this circumstance may be readily explained from what has just been said; because practical pure reason must necessarily begin with principles, which therefore must be the first data, the foundation of all science, and cannot be derived from it. It was possible to effect this verification of moral principles as principles of a pure reason quite well, and with sufficient certainty, by a single appeal to the judgement of common sense, for this reason, that anything empirical which might slip into our maxims as a determining principle of the will can be detected at once by the feeling of pleasure or pain which necessarily attaches to it as exciting desire; whereas pure practical reason positively refuses to admit this feeling into its principle as a condition. The heterogeneity of the determining principles (the empirical and rational) is clearly detected by this resistance of a practically legislating reason against every admixture of inclination, and by a peculiar kind of sentiment, which, however, does not precede the legislation of the practical reason, but, on the contrary, is produced by this as a constraint, namely, by the feeling of a respect such as no man has for inclinations of whatever kind but for the law only; and it is detected in so marked and prominent a manner that even the most uninstructed cannot fail to see at once in an example presented to him, that empirical principles of volition may indeed urge him to follow their attractions, but that he can never be expected to obey anything but the pure practical law of reason alone.

If we also consider the knowledge we can have of pure practical reason and through it, as shown in the Analytic, we find a striking analogy with theoretical reason. But we also find equally striking differences. In the case of theoretical reason, examples from the sciences could easily and clearly demonstrate our capacity for pure rational knowledge a priori. The sciences test their principles in so many ways through systematic use that there is less reason to fear a hidden mixture of empirical principles of knowledge than there is in ordinary knowledge. But only the most ordinary practical use of reason could show that pure reason is practical on its own, without any mixture of an empirical principle. It could show this by confirming that every person’s natural reason recognizes the supreme practical principle as the supreme law of the will. This law is completely a priori and does not depend on any data from the senses. Its pure origin had first to be established and confirmed, even in the judgment of ordinary reason. Only then could science take up the law as a fact, before any debate about its possibility or any consequences drawn from it. What has just been said readily explains why this was necessary. Pure practical reason must begin with principles. These must therefore be the first data and the foundation of all science. They cannot be derived from science. A single appeal to the judgment of common sense could confirm moral principles as principles of pure reason quite well and with sufficient certainty. This is because any empirical element that might enter our maxims as a determining principle of the will can be detected at once. A feeling of pleasure or pain necessarily accompanies such an element when it arouses desire. Pure practical reason, by contrast, expressly refuses to admit this feeling into its principle as a condition. The difference in kind between empirical and rational determining principles is clearly revealed by reason’s resistance, when it gives practical laws, to any mixture of inclination. It is also revealed by a distinctive feeling. This feeling does not come before practical reason gives its law. On the contrary, the law produces it by constraining us. It is a feeling of respect, which no one has for inclinations of any kind, but only for the law. The difference is so clear and striking that even the most uninstructed person cannot fail to see it at once in an example. Empirical principles of volition may urge the person to give in to their attractions. But the person can never be expected to obey anything except the pure practical law of reason alone.

The distinction between the doctrine of happiness and the doctrine of morality, in the former of which empirical principles constitute the entire foundation, while in the second they do not form the smallest part of it, is the first and most important office of the Analytic of pure practical reason; and it must proceed in it with as much exactness and, so to speak, scrupulousness, as any geometer in his work. The philosopher, however, has greater difficulties to contend with here (as always in rational cognition by means of concepts merely without construction), because he cannot take any intuition as a foundation (for a pure noumenon). He has, however, this advantage that, like the chemist, he can at any time make an experiment with every man's practical reason for the purpose of distinguishing the moral (pure) principle of determination from the empirical; namely, by adding the moral law (as a determining principle) to the empirically affected will (e.g., that of the man who would be ready to lie because he can gain something thereby). It is as if the analyst added alkali to a solution of lime in hydrochloric acid, the acid at once forsakes the lime, combines with the alkali, and the lime is precipitated. Just in the same way, if to a man who is otherwise honest (or who for this occasion places himself only in thought in the position of an honest man), we present the moral law by which he recognises the worthlessness of the liar, his practical reason (in forming a judgement of what ought to be done) at once forsakes the advantage, combines with that which maintains in him respect for his own person (truthfulness), and the advantage after it has been separated and washed from every particle of reason (which is altogether on the side of duty) is easily weighed by everyone, so that it can enter into combination with reason in other cases, only not where it could be opposed to the moral law, which reason never forsakes, but most closely unites itself with.

The first and most important task of the Analytic of pure practical reason is to distinguish the doctrine of happiness from the doctrine of morality. Empirical principles form the entire foundation of the doctrine of happiness. They form no part at all of the foundation of morality. The Analytic must make this distinction as precisely and, so to speak, as scrupulously as a geometer does in his work. The philosopher faces greater difficulties here. This is always true when reason works through concepts alone, without construction. The philosopher cannot base his work on any intuition, since the subject is a pure noumenon. But the philosopher has one advantage. Like a chemist, he can experiment at any time with any person's practical reason. He can use this experiment to distinguish the moral (pure) principle that determines the will from the empirical principle. He does so by introducing the moral law as a determining principle to a will affected by empirical considerations. Consider, for example, a person who is ready to lie because he can gain from it. The experiment is like adding alkali to a solution of lime in hydrochloric acid. The acid immediately leaves the lime and combines with the alkali, and the lime settles out. In just the same way, suppose we present the moral law to a person who is otherwise honest. Or suppose he imagines himself, for this occasion, in the position of an honest person. Through the moral law, he recognizes that a liar is worthless. In judging what he ought to do, his practical reason immediately leaves the advantage behind. It joins instead with truthfulness, which maintains his respect for himself. The advantage is now separated and washed clean of every particle of reason, which is entirely on the side of duty. Everyone can then easily weigh the advantage. It can combine with reason in other cases, but not where it could conflict with the moral law. Reason never leaves the moral law, but joins itself most closely to it.

But it does not follow that this distinction between the principle of happiness and that of morality is an opposition between them, and pure practical reason does not require that we should renounce all claim to happiness, but only that the moment duty is in question we should take no account of happiness. It may even in certain respects be a duty to provide for happiness; partly, because (including skill, wealth, riches) it contains means for the fulfilment of our duty; partly, because the absence of it (e.g., poverty) implies temptations to transgress our duty. But it can never be an immediate duty to promote our happiness, still less can it be the principle of all duty. Now, as all determining principles of the will, except the law of pure practical reason alone (the moral law), are all empirical and, therefore, as such, belong to the principle of happiness, they must all be kept apart from the supreme principle of morality and never be incorporated with it as a condition; since this would be to destroy all moral worth just as much as any empirical admixture with geometrical principles would destroy the certainty of mathematical evidence, which in Plato's opinion is the most excellent thing in mathematics, even surpassing their utility.

But the distinction between the principle of happiness and the principle of morality does not mean they are opposed. Pure practical reason does not require us to give up all claim to happiness. It requires only that we disregard happiness when duty is at stake. In some respects, providing for our happiness may even be a duty. One reason is that happiness, including skill, wealth, and riches, provides means for fulfilling our duty. Another is that a lack of happiness, such as poverty, brings temptations to violate our duty. But promoting our happiness can never be a direct duty, much less the principle of all duty. The law of pure practical reason, or the moral law, is the only non-empirical principle that determines the will. All other principles that determine the will are empirical and, as such, belong to the principle of happiness. So all of them must be kept separate from the supreme principle of morality. They must never be made a condition of that principle. Doing so would destroy all moral worth. In the same way, mixing any empirical element into geometrical principles would destroy the certainty of mathematical proof. In Plato's view, this certainty is the finest thing about mathematics, even more valuable than its usefulness.

Instead, however, of the deduction of the supreme principle of pure practical reason, that is, the explanation of the possibility of such a knowledge a priori, the utmost we were able to do was to show that if we saw the possibility of the freedom of an efficient cause, we should also see not merely the possibility, but even the necessity, of the moral law as the supreme practical law of rational beings, to whom we attribute freedom of causality of their will; because both concepts are so inseparably united that we might define practical freedom as independence of the will on anything but the moral law. But we cannot perceive the possibility of the freedom of an efficient cause, especially in the world of sense; we are fortunate if only we can be sufficiently assured that there is no proof of its impossibility, and are now, by the moral law which postulates it, compelled and therefore authorized to assume it. However, there are still many who think that they can explain this freedom on empirical principles, like any other physical faculty, and treat it as a psychological property, the explanation of which only requires a more exact study of the nature of the soul and of the motives of the will, and not as a transcendental predicate of the causality of a being that belongs to the world of sense (which is really the point). They thus deprive us of the grand revelation which we obtain through practical reason by means of the moral law, the revelation, namely, of a supersensible world by the realization of the otherwise transcendent concept of freedom, and by this deprive us also of the moral law itself, which admits no empirical principle of determination. Therefore it will be necessary to add something here as a protection against this delusion and to exhibit empiricism in its naked superficiality.

Instead, we could not derive the supreme principle of pure practical reason or explain how such knowledge is possible a priori. The most we could do was show what would follow if we understood how a cause could act freely. We would then see not only that the moral law is possible, but that it is necessary. We would see that it is the supreme practical law for rational beings whose wills we regard as free to cause actions. This is because freedom and the moral law are inseparably linked. We might even define practical freedom as the will’s independence from everything except the moral law. But we cannot understand how a cause can act freely, especially in the world of sense. We are fortunate if we can be sufficiently assured that no proof of its impossibility exists. And because the moral law requires this freedom, we are compelled, and therefore entitled, to assume it. Yet many still think they can explain this freedom through empirical principles, like any other physical faculty. They treat freedom as a psychological property. They think that explaining it requires only a closer study of the nature of the soul and the motives of the will. They do not treat freedom as a transcendental predicate of the causality of a being belonging to the world of sense. That is the real point. By doing so, they deprive us of the great revelation that practical reason gives us through the moral law. This is the revelation of a supersensible world, made possible by giving reality to the otherwise transcendent concept of freedom. They also deprive us of the moral law itself, which allows no empirical principle to determine the will. We must therefore say more to guard against this delusion and expose empiricism in its naked superficiality.

The notion of causality as physical necessity, in opposition to the same notion as freedom, concerns only the existence of things so far as it is determinable in time, and, consequently, as phenomena, in opposition to their causality as things in themselves. Now if we take the attributes of existence of things in time for attributes of things in themselves (which is the common view), then it is impossible to reconcile the necessity of the causal relation with freedom; they are contradictory. For from the former it follows that every event, and consequently every action that takes place at a certain point of time, is a necessary result of what existed in time preceding. Now as time past is no longer in my power, hence every action that I perform must be the necessary result of certain determining grounds which are not in my power, that is, at the moment in which I am acting I am never free. Nay, even if I assume that my whole existence is independent on any foreign cause (for instance, God), so that the determining principles of my causality, and even of my whole existence, were not outside myself, yet this would not in the least transform that physical necessity into freedom. For at every moment of time I am still under the necessity of being determined to action by that which is not in my power, and the series of events infinite a parte priori, which I only continue according to a pre-determined order and could never begin of myself, would be a continuous physical chain, and therefore my causality would never be freedom.

Causality understood as physical necessity, rather than freedom, concerns only the existence of things as determined in time. It therefore concerns things as phenomena, not their causality as things in themselves. Now suppose we take the features of things existing in time to be features of things in themselves. This is the common view. If so, we cannot reconcile the necessity of the causal relation with freedom. They contradict each other. For causal necessity implies that every event necessarily results from what existed earlier in time. This applies to every action that takes place at a particular time. Since the past is no longer within my power, every action I perform must necessarily result from determining grounds outside my power. So, at the moment I act, I am never free. Suppose, however, that my whole existence is independent of any external cause, such as God. On this assumption, the principles determining my causality and even my whole existence would not lie outside me. Yet this would not turn physical necessity into freedom in the least. At every moment, I would still have to be determined to act by something outside my power. The series of events would extend infinitely into the past. I would only continue it according to a predetermined order and could never begin it myself. It would be a continuous physical chain, so my causality would never be freedom.

If, then, we would attribute freedom to a being whose existence is determined in time, we cannot except him from the law of necessity as to all events in his existence and, consequently, as to his actions also; for that would be to hand him over to blind chance. Now as this law inevitably applies to all the causality of things, so far as their existence is determinable in time, it follows that if this were the mode in which we had also to conceive the existence of these things in themselves, freedom must be rejected as a vain and impossible conception. Consequently, if we would still save it, no other way remains but to consider that the existence of a thing, so far as it is determinable in time, and therefore its causality, according to the law of physical necessity, belong to appearance, and to attribute freedom to the same being as a thing in itself. This is certainly inevitable, if we would retain both these contradictory concepts together; but in application, when we try to explain their combination in one and the same action, great difficulties present themselves which seem to render such a combination impracticable.

If we attribute freedom to a being whose existence is determined in time, we cannot exempt that being from the law of necessity. That law applies to every event in the being’s existence and therefore to its actions as well. To exempt it would be to leave it to blind chance. The law of necessity applies inevitably to all causality in things whose existence can be determined in time. So, if we also had to understand things in themselves as existing in this way, we would have to reject freedom as an empty and impossible concept. If we still want to preserve freedom, only one way remains. We must regard a thing’s existence, insofar as it can be determined in time, as belonging to appearance. Its causality under the law of physical necessity must also belong to appearance. We must then attribute freedom to the same being as a thing in itself. This approach is certainly unavoidable if we want to retain both contradictory concepts together. But when we try to explain how they combine in one and the same action, we encounter great difficulties. These difficulties seem to make such a combination impossible in practice.

When I say of a man who commits a theft that, by the law of causality, this deed is a necessary result of the determining causes in preceding time, then it was impossible that it could not have happened; how then can the judgement, according to the moral law, make any change, and suppose that it could have been omitted, because the law says that it ought to have been omitted; that is, how can a man be called quite free at the same moment, and with respect to the same action in which he is subject to an inevitable physical necessity? Some try to evade this by saying that the causes that determine his causality are of such a kind as to agree with a comparative notion of freedom. According to this, that is sometimes called a free effect, the determining physical cause of which lies within the acting thing itself, e.g., that which a projectile performs when it is in free motion, in which case we use the word freedom, because while it is in flight it is not urged by anything external; or as we call the motion of a clock a free motion, because it moves its hands itself, which therefore do not require to be pushed by external force; so although the actions of man are necessarily determined by causes which precede in time, we yet call them free, because these causes are ideas produced by our own faculties, whereby desires are evoked on occasion of circumstances, and hence actions are wrought according to our own pleasure. This is a wretched subterfuge with which some persons still let themselves be put off, and so think they have solved, with a petty word- jugglery, that difficult problem, at the solution of which centuries have laboured in vain, and which can therefore scarcely be found so completely on the surface. In fact, in the question about the freedom which must be the foundation of all moral laws and the consequent responsibility, it does not matter whether the principles which necessarily determine causality by a physical law reside within the subject or without him, or in the former case whether these principles are instinctive or are conceived by reason, if, as is admitted by these men themselves, these determining ideas have the ground of their existence in time and in the antecedent state, and this again in an antecedent, etc. Then it matters not that these are internal; it matters not that they have a psychological and not a mechanical causality, that is, produce actions by means of ideas and not by bodily movements; they are still determining principles of the causality of a being whose existence is determinable in time, and therefore under the necessitation of conditions of past time, which therefore, when the subject has to act, are no longer in his power. This may imply psychological freedom (if we choose to apply this term to a merely internal chain of ideas in the mind), but it involves physical necessity and, therefore, leaves no room for transcendental freedom, which must be conceived as independence on everything empirical, and, consequently, on nature generally, whether it is an object of the internal sense considered in time only, or of the external in time and space. Without this freedom (in the latter and true sense), which alone is practical a priori, no moral law and no moral imputation are possible. Just for this reason the necessity of events in time, according to the physical law of causality, may be called the mechanism of nature, although we do not mean by this that things which are subject to it must be really material machines. We look here only to the necessity of the connection of events in a time-series as it is developed according to the physical law, whether the subject in which this development takes place is called automaton materiale when the mechanical being is moved by matter, or with Leibnitz spirituale when it is impelled by ideas; and if the freedom of our will were no other than the latter (say the psychological and comparative, not also transcendental, that is, absolute), then it would at bottom be nothing better than the freedom of a turnspit, which, when once it is wound up, accomplishes its motions of itself.

When I say that a man’s theft necessarily resulted, by the law of causality, from causes that determined it earlier, I mean that the theft could not possibly have failed to happen. How, then, can a judgment according to the moral law change anything? How can it suppose that he could have refrained from stealing because the law says he ought to have done so? In other words, how can a man be called completely free at the very moment, and in the very action, in which he is subject to unavoidable physical necessity? Some try to evade this problem by saying that the causes determining his action are compatible with a relative sense of freedom. In this sense, an effect is sometimes called free when its determining physical cause lies within the thing acting. A projectile in free motion, for example, is said to move freely because nothing external pushes it while it is in flight. We also call a clock’s motion free because it moves its hands itself. The hands do not need to be pushed by an external force. In the same way, although human actions are necessarily determined by earlier causes, we still call them free. We do so because those causes are ideas produced by our own faculties. Circumstances give rise to desires through these ideas, and we then act as we please. This is a miserable evasion. Some people still allow themselves to be satisfied with it. They think that, through petty wordplay, they have solved a difficult problem that centuries have tried in vain to solve. The solution can hardly lie so completely on the surface. The freedom at issue must be the foundation of all moral laws and of the responsibility that follows from them. For this question, it does not matter whether the principles that necessarily determine causality by a physical law lie inside or outside the person. If they lie inside, it does not matter whether they are instinctive or conceived by reason. This is so if, as these people themselves admit, the determining ideas exist because of an earlier state in time, which itself exists because of a still earlier state, and so on. It does not matter, then, that these principles are internal. Nor does it matter that they have psychological rather than mechanical causality, producing actions through ideas rather than bodily movements. They are still principles that determine the causality of a being whose existence is determined in time. That being is therefore subject to the necessity imposed by conditions in the past. When the person must act, those conditions are no longer under his control. This may imply psychological freedom, if we choose to use that term for a merely internal chain of ideas in the mind. But it involves physical necessity. It therefore leaves no room for transcendental freedom, which must be understood as independence from everything empirical and, consequently, from nature generally. This applies whether nature is an object of the internal sense, considered in time alone, or of the external sense, considered in time and space. Without freedom in this latter, true sense, no moral law and no moral imputation are possible. This freedom alone is practical a priori. For this very reason, the necessity of events in time, according to the physical law of causality, may be called the mechanism of nature. We do not mean that things subject to this law must actually be material machines. We are concerned only with the necessary connection of events in a sequence in time as that sequence unfolds according to physical law. This holds whether we call the subject in which the sequence unfolds an automaton materiale, a mechanical being moved by matter, or, with Leibnitz, an automaton spirituale, one driven by ideas. If the freedom of our will were no more than the latter kind, psychological and relative, rather than also transcendental, that is, absolute, it would be no better in the end than the freedom of a turnspit. Once wound up, a turnspit carries out its motions by itself.

Now, in order to remove in the supposed case the apparent contradiction between freedom and the mechanism of nature in one and the same action, we must remember what was said in the Critique of Pure Reason, or what follows therefrom; viz., that the necessity of nature, which cannot co-exist with the freedom of the subject, appertains only to the attributes of the thing that is subject to time-conditions, consequently only to those of the acting subject as a phenomenon; that therefore in this respect the determining principles of every action of the same reside in what belongs to past time and is no longer in his power (in which must be included his own past actions and the character that these may determine for him in his own eyes as a phenomenon). But the very same subject, being on the other side conscious of himself as a thing in himself, considers his existence also in so far as it is not subject to time-conditions, and regards himself as only determinable by laws which he gives himself through reason; and in this his existence nothing is antecedent to the determination of his will, but every action, and in general every modification of his existence, varying according to his internal sense, even the whole series of his existence as a sensible being is in the consciousness of his supersensible existence nothing but the result, and never to be regarded as the determining principle, of his causality as a noumenon. In this view now the rational being can justly say of every unlawful action that he performs, that he could very well have left it undone; although as appearance it is sufficiently determined in the past, and in this respect is absolutely necessary; for it, with all the past which determines it, belongs to the one single phenomenon of his character which he makes for himself, in consequence of which he imputes the causality of those appearances to himself as a cause independent on sensibility.

To resolve the apparent conflict between freedom and the mechanism of nature in one and the same action, we must remember what was said in the Critique of Pure Reason, or what follows from it. Natural necessity cannot coexist with the freedom of the subject. But it applies only to the attributes of a thing subject to the conditions of time. It therefore applies only to the attributes of the acting subject as a phenomenon. In this respect, the determining principles of every action lie in the past and are no longer within the subject’s power. These include his own past actions and the character they may give him in his own eyes as a phenomenon. But the same subject is also conscious of himself as a thing in himself. He therefore also considers his existence insofar as it is not subject to the conditions of time. He regards himself as subject only to laws he gives himself through reason. In this aspect of his existence, nothing comes before the determination of his will. In his consciousness of his supersensible existence, every action is only a result of his causality as a noumenon. The same is true of every change in his existence as it varies according to his inner sense. It is even true of the entire series of his existence as a sensible being. None of these can ever be regarded as the determining principle of his causality as a noumenon. From this standpoint, a rational being can rightly say of every unlawful action he performs that he could have left it undone. As an appearance, the action is sufficiently determined by the past and is therefore absolutely necessary in that respect. Yet the action and all the past that determines it belong to the single phenomenon of the character he makes for himself. Consequently, he attributes the causality of those appearances to himself as a cause independent of sensibility.

With this agree perfectly the judicial sentences of that wonderful faculty in us which we call conscience. A man may use as much art as he likes in order to paint to himself an unlawful act, that he remembers, as an unintentional error, a mere oversight, such as one can never altogether avoid, and therefore as something in which he was carried away by the stream of physical necessity, and thus to make himself out innocent, yet he finds that the advocate who speaks in his favour can by no means silence the accuser within, if only he is conscious that at the time when he did this wrong he was in his senses, that is, in possession of his freedom; and, nevertheless, he accounts for his error from some bad habits, which by gradual neglect of attention he has allowed to grow upon him to such a degree that he can regard his error as its natural consequence, although this cannot protect him from the blame and reproach which he casts upon himself. This is also the ground of repentance for a long past action at every recollection of it; a painful feeling produced by the moral sentiment, and which is practically void in so far as it cannot serve to undo what has been done. (Hence Priestley, as a true and consistent fatalist, declares it absurd, and he deserves to be commended for this candour more than those who, while they maintain the mechanism of the will in fact, and its freedom in words only, yet wish it to be thought that they include it in their system of compromise, although they do not explain the possibility of such moral imputation.) But the pain is quite legitimate, because when the law of our intelligible [supersensible] existence (the moral law) is in question, reason recognizes no distinction of time, and only asks whether the event belongs to me, as my act, and then always morally connects the same feeling with it, whether it has happened just now or long ago. For in reference to the supersensible consciousness of its existence (i.e., freedom) the life of sense is but a single phenomenon, which, inasmuch as it contains merely manifestations of the mental disposition with regard to the moral law (i.e., of the character), must be judged not according to the physical necessity that belongs to it as phenomenon, but according to the absolute spontaneity of freedom. It may therefore be admitted that, if it were possible to have so profound an insight into a man's mental character as shown by internal as well as external actions as to know all its motives, even the smallest, and likewise all the external occasions that can influence them, we could calculate a man's conduct for the future with as great certainty as a lunar or solar eclipse; and nevertheless we may maintain that the man is free. In fact, if we were capable of a further glance, namely, an intellectual intuition of the same subject (which indeed is not granted to us, and instead of it we have only the rational concept), then we should perceive that this whole chain of appearances in regard to all that concerns the moral laws depends on the spontaneity of the subject as a thing in itself, of the determination of which no physical explanation can be given. In default of this intuition, the moral law assures us of this distinction between the relation of our actions as appearance to our sensible nature, and the relation of this sensible nature to the supersensible substratum in us. In this view, which is natural to our reason, though inexplicable, we can also justify some judgements which we passed with all conscientiousness, and which yet at first sight seem quite opposed to all equity. There are cases in which men, even with the same education which has been profitable to others, yet show such early depravity, and so continue to progress in it to years of manhood, that they are thought to be born villains, and their character altogether incapable of improvement; and nevertheless they are judged for what they do or leave undone, they are reproached for their faults as guilty; nay, they themselves (the children) regard these reproaches as well founded, exactly as if in spite of the hopeless natural quality of mind ascribed to them, they remained just as responsible as any other man. This could not happen if we did not suppose that whatever springs from a man's choice (as every action intentionally performed undoubtedly does) has as its foundation a free causality, which from early youth expresses its character in its manifestations (i.e., actions). These, on account of the uniformity of conduct, exhibit a natural connection, which however does not make the vicious quality of the will necessary, but on the contrary, is the consequence of the evil principles voluntarily adopted and unchangeable, which only make it so much the more culpable and deserving of punishment. There still remains a difficulty in the combination of freedom with the mechanism of nature in a being belonging to the world of sense; a difficulty which, even after all the foregoing is admitted, threatens freedom with complete destruction. But with this danger there is also a circumstance that offers hope of an issue still favourable to freedom; namely, that the same difficulty presses much more strongly (in fact as we shall presently see, presses only) on the system that holds the existence determinable in time and space to be the existence of things in themselves; it does not therefore oblige us to give up our capital supposition of the ideality of time as a mere form of sensible intuition, and consequently as a mere manner of representation which is proper to the subject as belonging to the world of sense; and therefore it only requires that this view be reconciled with this idea.

The judgments of that remarkable faculty we call conscience agree perfectly with this account. A person may use all the ingenuity he likes to portray a wrongful act he remembers as an unintentional mistake. He may call it a mere oversight, the kind no one can entirely avoid. He may therefore see it as something in which he was swept along by physical necessity, and so consider himself innocent. Yet he finds that the advocate speaking in his favor cannot silence the accuser within. This is true as long as he knows that he was in his senses when he did wrong, that is, that he possessed his freedom. Even so, he explains his mistake by referring to bad habits that he gradually allowed to grow through lack of attention. These habits have grown so strong that he can regard his mistake as their natural consequence. But this cannot protect him from the blame and reproach he directs at himself. This is also why a person feels repentance whenever he recalls an action from long ago. Repentance is a painful feeling produced by moral sentiment. It serves no practical purpose insofar as it cannot undo what has been done. (For this reason, Priestley, as a true and consistent fatalist, declares repentance absurd. He deserves more credit for this candor than those who maintain that the will is governed by mechanism but call it free. These people still want us to think that freedom has a place in their compromise system, although they do not explain how such moral responsibility is possible.) But the pain of repentance is entirely legitimate. When the law of our intelligible [supersensible] existence, the moral law, is at issue, reason recognizes no distinction of time. It asks only whether the event belongs to me as my act. If it does, reason always attaches the same moral feeling to it, whether it happened just now or long ago. For our supersensible consciousness of our existence, that is, our freedom, our life as perceived through the senses is a single phenomenon. This life contains only expressions of our disposition toward the moral law, that is, of our character. It must therefore be judged by the absolute spontaneity of freedom, not by the physical necessity that belongs to it as a phenomenon. We may therefore grant that a person's future conduct could be calculated as certainly as a lunar or solar eclipse. This would be possible if we had a sufficiently deep insight into his character as shown in his inner and outer actions. We would have to know all his motives, even the smallest, and all the external circumstances that could influence them. Yet we may still maintain that the person is free. Suppose, in fact, that we could gain a further insight: an intellectual intuition of the same person. This is not granted to us. Instead, we have only the rational concept. With such an intuition, we would see that this entire chain of appearances, in all that concerns the moral laws, depends on the spontaneity of the person as a thing in itself. No physical explanation can be given of what determines that spontaneity. Without this intuition, the moral law assures us of a distinction. Our actions as appearances stand in one relation to our sensible nature. Our sensible nature stands in another relation to the supersensible foundation within us. This view is natural to our reason, though inexplicable. It also allows us to justify some judgments that we make with full conscientiousness, yet that at first seem quite opposed to fairness. Some people receive the same education that has benefited others, yet show depravity from a very early age. They continue to grow more depraved into adulthood. As a result, they are thought to be born villains, with characters altogether incapable of improvement. Nevertheless, they are judged for what they do or fail to do, and blamed as guilty for their faults. Indeed, even as children, they themselves regard these reproaches as justified. It is exactly as if they remained just as responsible as anyone else, despite the hopeless natural quality of mind attributed to them. This could not happen unless we supposed that whatever comes from a person's choice has a free causality as its foundation. Every intentional action undoubtedly comes from choice. From early youth, this free causality expresses its character in its manifestations, that is, in actions. Because the conduct is consistent, the actions show a natural connection. But this connection does not make the vicious quality of the will necessary. On the contrary, it results from evil principles that were voluntarily adopted and are unchangeable. These principles only make the person more blameworthy and deserving of punishment. A difficulty still remains in combining freedom with the mechanism of nature in a being who belongs to the world of sense. Even if all the preceding arguments are accepted, this difficulty threatens to destroy freedom completely. But alongside this danger is a circumstance that offers hope of an outcome favorable to freedom. The same difficulty bears much more heavily on the system that treats existence in time and space as the existence of things in themselves. In fact, as we shall soon see, the difficulty bears only on that system. It therefore does not force us to abandon our central assumption about time. That assumption is that time is ideal, merely a form of sensible intuition. Consequently, time is merely a way of representing things that belongs to us as subjects in the world of sense. The difficulty requires only that this view be reconciled with the idea of freedom.

The difficulty is as follows: Even if it is admitted that the supersensible subject can be free with respect to a given action, although, as a subject also belonging to the world of sense, he is under mechanical conditions with respect to the same action, still, as soon as we allow that God as universal first cause is also the cause of the existence of substance (a proposition which can never be given up without at the same time giving up the notion of God as the Being of all beings, and therewith giving up his all sufficiency, on which everything in theology depends), it seems as if we must admit that a man's actions have their determining principle in something which is wholly out of his power- namely, in the causality of a Supreme Being distinct from himself and on whom his own existence and the whole determination of his causality are absolutely dependent. In point of fact, if a man's actions as belonging to his modifications in time were not merely modifications of him as appearance, but as a thing in itself, freedom could not be saved. Man would be a marionette or an automaton, like Vaucanson's, prepared and wound up by the Supreme Artist. Self-consciousness would indeed make him a thinking automaton; but the consciousness of his own spontaneity would be mere delusion if this were mistaken for freedom, and it would deserve this name only in a comparative sense, since, although the proximate determining causes of its motion and a long series of their determining causes are internal, yet the last and highest is found in a foreign hand. Therefore I do not see how those who still insist on regarding time and space as attributes belonging to the existence of things in themselves, can avoid admitting the fatality of actions; or if (like the otherwise acute Mendelssohn) they allow them to be conditions necessarily belonging to the existence of finite and derived beings, but not to that of the infinite Supreme Being, I do not see on what ground they can justify such a distinction, or, indeed, how they can avoid the contradiction that meets them, when they hold that existence in time is an attribute necessarily belonging to finite things in themselves, whereas God is the cause of this existence, but cannot be the cause of time (or space) itself (since this must be presupposed as a necessary a priori condition of the existence of things); and consequently as regards the existence of these things. His causality must be subject to conditions and even to the condition of time; and this would inevitably bring in everything contradictory to the notions of His infinity and independence. On the other hand, it is quite easy for us to draw the distinction between the attribute of the divine existence of being independent on all time-conditions, and that of a being of the world of sense, the distinction being that between the existence of a being in itself and that of a thing in appearance. Hence, if this ideality of time and space is not adopted, nothing remains but Spinozism, in which space and time are essential attributes of the Supreme Being Himself, and the things dependent on Him (ourselves, therefore, included) are not substances, but merely accidents inhering in Him; since, if these things as His effects exist in time only, this being the condition of their existence in themselves, then the actions of these beings must be simply His actions which He performs in some place and time. Thus, Spinozism, in spite of the absurdity of its fundamental idea, argues more consistently than the creation theory can, when beings assumed to be substances, and beings in themselves existing in time, are regarded as effects of a Supreme Cause, and yet as not [belonging] to Him and His action, but as separate substances.

The difficulty is this: Suppose we accept that a person, as a supersensible subject, can be free regarding a given action. As a subject who also belongs to the world of sense, the person is still subject to mechanical conditions regarding that same action. But we must also allow that God, as the universal first cause, causes the existence of substance. We can never give up this claim without also giving up the idea of God as the Being of all beings. We would also have to give up his all-sufficiency, on which everything in theology depends. Once we allow this, we seem forced to admit that the determining principle of a person's actions lies wholly outside that person's power. It lies in the causality of a Supreme Being who is distinct from the person. The person's existence and the entire determination of their causality depend absolutely on this being. In fact, a person's actions involve changes in that person over time. If these were changes in the person as a thing in itself, rather than merely as an appearance, freedom could not be saved. The person would be a puppet or an automaton, like Vaucanson's, made and wound up by the Supreme Artist. Self-consciousness would indeed make the person a thinking automaton. But the person's consciousness of spontaneity would be an illusion if mistaken for freedom. It would deserve the name of freedom only in a comparative sense. The immediate determining causes of the automaton's motion, and a long series of their determining causes, are internal. Yet the ultimate and highest cause lies in another's hand. Therefore, I do not see how those who still regard time and space as attributes of things as they exist in themselves can avoid admitting that actions are determined by fate. Others, like the otherwise perceptive Mendelssohn, allow that time and space are necessary conditions for the existence of finite, derived beings, but not for the infinite Supreme Being. I do not see how they can justify this distinction. Nor do I see how they can avoid a contradiction. They hold that existence in time is a necessary attribute of finite things in themselves and that God causes this existence. Yet God cannot be the cause of time or space itself. For time and space must be presupposed as necessary a priori conditions for the existence of things. Consequently, God's causality, as it concerns the existence of these things, must be subject to conditions, including the condition of time. This would inevitably introduce everything that contradicts the ideas of his infinity and independence. On the other hand, it is quite easy for us to distinguish divine existence, which is independent of all conditions of time, from the existence of a being in the world of sense. The distinction is between the existence of a being in itself and the existence of a thing as an appearance. Thus, if we do not accept the ideality of time and space, only Spinozism remains. In Spinozism, space and time are essential attributes of the Supreme Being himself. The things that depend on him, including ourselves, are not substances but merely accidents inhering in him. For if these things, as his effects, exist only in time, and time is a condition of their existence in themselves, then their actions must simply be his actions, which he performs in some place and time. Thus, despite the absurdity of its fundamental idea, Spinozism argues more consistently than the theory of creation can. This applies when beings assumed to be substances, and things in themselves existing in time, are regarded as effects of a Supreme Cause. Yet they are regarded not as belonging to him and his action, but as separate substances.

The above-mentioned difficulty is resolved briefly and clearly as follows: If existence in time is a mere sensible mode of representation belonging to thinking beings in the world and consequently does not apply to them as things in themselves, then the creation of these beings is a creation of things in themselves, since the notion of creation does not belong to the sensible form of representation of existence or to causality, but can only be referred to noumena. Consequently, when I say of beings in the world of sense that they are created, I so far regard them as noumena. As it would be a contradiction, therefore, to say that God is a creator of appearances, so also it is a contradiction to say that as creator He is the cause of actions in the world of sense, and therefore as appearances, although He is the cause of the existence of the acting beings (which are noumena). If now it is possible to affirm freedom in spite of the natural mechanism of actions as appearances (by regarding existence in time as something that belongs only to appearances, not to things in themselves), then the circumstance that the acting beings are creatures cannot make the slightest difference, since creation concerns their supersensible and not their sensible existence, and, therefore, cannot be regarded as the determining principle of the appearances. It would be quite different if the beings in the world as things in themselves existed in time, since in that case the creator of substance would be at the same time the author of the whole mechanism of this substance.

The difficulty mentioned above can be resolved briefly and clearly as follows: Suppose that existence in time is merely a way thinking beings in the world represent things through their senses. In that case, existence in time does not apply to them as things in themselves. The creation of these beings is then a creation of things in themselves. This is because creation does not belong to the sensible way of representing existence or causality. It can refer only to noumena. So when I say that beings in the sensible world are created, I am regarding them as noumena. It would therefore be a contradiction to say that God is a creator of appearances. Likewise, it is a contradiction to say that, as creator, he causes actions in the sensible world, considered as appearances. This is so even though he causes the existence of the beings who act, considered as noumena. Now suppose we can affirm freedom despite the natural mechanism governing actions as appearances. We can do this by treating existence in time as belonging only to appearances, not to things in themselves. In that case, the fact that the acting beings are creatures cannot make the slightest difference. Creation concerns their supersensible existence, not their sensible existence. It therefore cannot be regarded as the determining principle of appearances. Things would be quite different if beings in the world existed in time as things in themselves. In that case, the creator of substance would also be the author of its entire mechanism.

Of so great importance is the separation of time (as well as space) from the existence of things in themselves which was effected in the Critique of the Pure Speculative Reason.

The separation of time (as well as space) from the existence of things in themselves is extremely important. This separation was made in the Critique of the Pure Speculative Reason.

It may be said that the solution here proposed involves great difficulty in itself and is scarcely susceptible of a lucid exposition. But is any other solution that has been attempted, or that may be attempted, easier and more intelligible? Rather might we say that the dogmatic teachers of metaphysics have shown more shrewdness than candour in keeping this difficult point out of sight as much as possible, in the hope that if they said nothing about it, probably no one would think of it. If science is to be advanced, all difficulties must be laid open, and we must even search for those that are hidden, for every difficulty calls forth a remedy, which cannot be discovered without science gaining either in extent or in exactness; and thus even obstacles become means of increasing the thoroughness of science. On the other hand, if the difficulties are intentionally concealed, or merely removed by palliatives, then sooner or later they burst out into incurable mischiefs, which bring science to ruin in an absolute scepticism.

One might say that the solution proposed here is itself very difficult and can hardly be explained clearly. But is any other solution that has been tried, or might be tried, easier or more intelligible? Rather, the dogmatic teachers of metaphysics may have shown more cleverness than honesty by keeping this difficult point out of sight as much as possible. They apparently hoped that, if they said nothing about it, probably no one would think of it. If science is to advance, all difficulties must be brought to light. We must even search for hidden difficulties. Every difficulty calls for a remedy, and finding one always makes science either broader or more precise. In this way, even obstacles help make science more thorough. On the other hand, if difficulties are deliberately concealed or merely eased by temporary fixes, they sooner or later erupt into incurable troubles. These troubles ruin science by leading to absolute scepticism.

Since it is, properly speaking, the notion of freedom alone amongst all the ideas of pure speculative reason that so greatly enlarges our knowledge in the sphere of the supersensible, though only of our practical knowledge, I ask myself why it exclusively possesses so great fertility, whereas the others only designate the vacant space for possible beings of the pure understanding, but are unable by any means to define the concept of them. I presently find that as I cannot think anything without a category, I must first look for a category for the rational idea of freedom with which I am now concerned; and this is the category of causality; and although freedom, a concept of the reason, being a transcendent concept, cannot have any intuition corresponding to it, yet the concept of the understanding- for the synthesis of which the former demands the unconditioned- (namely, the concept of causality) must have a sensible intuition given, by which first its objective reality is assured. Now, the categories are all divided into two classes- the mathematical, which concern the unity of synthesis in the conception of objects, and the dynamical, which refer to the unity of synthesis in the conception of the existence of objects. The former (those of magnitude and quality) always contain a synthesis of the homogeneous, and it is not possible to find in this the unconditioned antecedent to what is given in sensible intuition as conditioned in space and time, as this would itself have to belong to space and time, and therefore be again still conditioned. Whence it resulted in the Dialectic of Pure Theoretic Reason that the opposite methods of attaining the unconditioned and the totality of the conditions were both wrong. The categories of the second class (those of causality and of the necessity of a thing) did not require this homogeneity (of the conditioned and the condition in synthesis), since here what we have to explain is not how the intuition is compounded from a manifold in it, but only how the existence of the conditioned object corresponding to it is added to the existence of the condition (added, namely, in the understanding as connected therewith); and in that case it was allowable to suppose in the supersensible world the unconditioned antecedent to the altogether conditioned in the world of sense (both as regards the causal connection and the contingent existence of things themselves), although this unconditioned remained indeterminate, and to make the synthesis transcendent. Hence, it was found in the Dialectic of the Pure Speculative Reason that the two apparently opposite methods of obtaining for the conditioned the unconditioned were not really contradictory, e.g., in the synthesis of causality to conceive for the conditioned in the series of causes and effects of the sensible world, a causality which has no sensible condition, and that the same action which, as belonging to the world of sense, is always sensibly conditioned, that is, mechanically necessary, yet at the same time may be derived from a causality not sensibly conditioned- being the causality of the acting being as belonging to the supersensible world- and may consequently be conceived as free. Now, the only point in question was to change this may be into is; that is, that we should be able to show in an actual case, as it were by a fact, that certain actions imply such a causality (namely, the intellectual, sensibly unconditioned), whether they are actual or only commanded, that is, objectively necessary in a practical sense. We could not hope to find this connexion in actions actually given in experience as events of the sensible world, since causality with freedom must always be sought outside the world of sense in the world of intelligence. But things of sense are the only things offered to our perception and observation. Hence, nothing remained but to find an incontestable objective principle of causality which excludes all sensible conditions: that is, a principle in which reason does not appeal further to something else as a determining ground of its causality, but contains this determining ground itself by means of that principle, and in which therefore it is itself as pure reason practical. Now, this principle had not to be searched for or discovered; it had long been in the reason of all men, and incorporated in their nature, and is the principle of morality. Therefore, that unconditioned causality, with the faculty of it, namely, freedom, is no longer merely indefinitely and problematically thought (this speculative reason could prove to be feasible), but is even as regards the law of its causality definitely and assertorially known; and with it the fact that a being (I myself), belonging to the world of sense, belongs also to the supersensible world, this is also positively known, and thus the reality of the supersensible world is established and in practical respects definitely given, and this definiteness, which for theoretical purposes would be transcendent, is for practical purposes immanent. We could not, however, make a similar step as regards the second dynamical idea, namely, that of a necessary being. We could not rise to it from the sensible world without the aid of the first dynamical idea. For if we attempted to do so, we should have ventured to leave at a bound all that is given to us, and to leap to that of which nothing is given us that can help us to effect the connection of such a supersensible being with the world of sense (since the necessary being would have to be known as given outside ourselves). On the other hand, it is now obvious that this connection is quite possible in relation to our own subject, inasmuch as I know myself to be on the one side as an intelligible [supersensible] being determined by the moral law (by means of freedom), and on the other side as acting in the world of sense. It is the concept of freedom alone that enables us to find the unconditioned and intelligible for the conditioned and sensible without going out of ourselves. For it is our own reason that by means of the supreme and unconditional practical law knows that itself and the being that is conscious of this law (our own person) belong to the pure world of understanding, and moreover defines the manner in which, as such, it can be active. In this way it can be understood why in the whole faculty of reason it is the practical reason only that can help us to pass beyond the world of sense and give us knowledge of a supersensible order and connection, which, however, for this very reason cannot be extended further than is necessary for pure practical purposes.

Of all the ideas of pure speculative reason, only the concept of freedom greatly expands our knowledge of the supersensible, strictly speaking. Even then, it expands only our practical knowledge. I ask myself why freedom alone is so fruitful. The other ideas merely mark out an empty space for possible beings of the pure understanding. They cannot in any way define the concept of those beings. I soon find that I cannot think of anything without a category. So I must first look for a category for the rational idea of freedom, which concerns me here. The category is causality. Freedom is a concept of reason, and it is transcendent, so no intuition can correspond to it. Yet freedom demands the unconditioned for the synthesis of the concept of causality, which belongs to the understanding. A sensible intuition must be given for that concept before its objective reality can be established. Now, all the categories are divided into two classes. The mathematical categories concern the unity of synthesis in the conception of objects. The dynamical categories concern the unity of synthesis in the conception of the existence of objects. The mathematical categories, those of magnitude and quality, always involve a synthesis of things of the same kind. In this synthesis, it is impossible to find the unconditioned that precedes what is given as conditioned in sensible intuition, in space and time. For the unconditioned would itself have to belong to space and time, and would therefore still be conditioned. The Dialectic of Pure Theoretic Reason thus concluded that both opposing ways of reaching the unconditioned and the totality of conditions were wrong. The categories of the second class, those of causality and of the necessity of a thing, did not require the conditioned and its condition to be of the same kind in the synthesis. Here we need not explain how an intuition is composed from a variety of elements within it. We need only explain how the existence of the corresponding conditioned object is added to the existence of the condition. That is, we must explain how the understanding adds the object’s existence as connected with the condition’s existence. In this case, it was permissible to suppose that the unconditioned precedes the entirely conditioned in the world of sense and belongs to the supersensible world. This applied both to causal connections and to the contingent existence of things themselves. The unconditioned nevertheless remained indeterminate, and the synthesis was transcendent. The Dialectic of Pure Speculative Reason therefore found that two apparently opposed ways of finding the unconditioned for the conditioned did not really contradict each other. For example, the conditioned in the series of causes and effects in the sensible world may be conceived as having a causality with no sensible condition. The same action, insofar as it belongs to the world of sense, is always sensibly conditioned and therefore mechanically necessary. Yet it may at the same time come from a causality that is not sensibly conditioned. This would be the causality of the acting being insofar as it belongs to the supersensible world. The action may consequently be conceived as free. The only question then was how to change “may be” into “is.” We needed to show, in an actual case and as if by a fact, that certain actions imply such a causality. This causality would be intellectual and sensibly unconditioned. The actions could be either actual or merely commanded, that is, objectively necessary in a practical sense. We could not hope to find this connection in actions actually given in experience as events in the sensible world. For causality through freedom must always be sought outside the world of sense, in the world of intelligence. But only things of sense are available to our perception and observation. So we had no choice but to find an incontestable objective principle of causality that excludes all sensible conditions. In such a principle, reason does not look to anything else for the ground that determines its causality. Rather, through the principle, reason contains that determining ground itself. In this principle, then, reason itself is practical as pure reason. This principle did not need to be sought or discovered. It had long existed in the reason of all human beings and was part of their nature. It is the principle of morality. Therefore, unconditioned causality and the capacity for it, namely freedom, are no longer merely thought of indefinitely and as a possibility. Speculative reason could prove that such thought was possible. Now, however, unconditioned causality is definitely known to exist, even in terms of the law of its causality. We also know positively that a being in the world of sense, namely myself, belongs to the supersensible world as well. In this way, the reality of the supersensible world is established and definitely given for practical purposes. This definiteness would be transcendent for theoretical purposes, but it is immanent for practical purposes. We could not, however, take a similar step with the second dynamical idea, that of a necessary being. We could not reach it from the sensible world without the help of the first dynamical idea. If we tried, we would venture to leave behind everything given to us in a single bound. We would leap to something of which nothing is given to help us connect such a supersensible being with the world of sense. For the necessary being would have to be known as given outside ourselves. By contrast, it is now clear that this connection is quite possible in relation to ourselves. On the one hand, I know myself as an intelligible [supersensible] being determined by the moral law through freedom. On the other hand, I know myself as acting in the world of sense. Only the concept of freedom enables us to find the unconditioned and intelligible for the conditioned and sensible without going outside ourselves. Through the supreme and unconditional practical law, our own reason knows that it belongs to the pure world of understanding. It also knows that the being conscious of this law, our own person, belongs to that world. Moreover, reason defines how our person, as a member of that world, can act. We can therefore understand why, in the whole faculty of reason, only practical reason can help us go beyond the world of sense. Only practical reason can give us knowledge of a supersensible order and connection. But for this very reason, that knowledge cannot extend further than is necessary for pure practical purposes.

Let me be permitted on this occasion to make one more remark, namely, that every step that we make with pure reason, even in the practical sphere where no attention is paid to subtle speculation, nevertheless accords with all the material points of the Critique of the Theoretical Reason as closely and directly as if each step had been thought out with deliberate purpose to establish this confirmation. Such a thorough agreement, wholly unsought for and quite obvious (as anyone can convince himself, if he will only carry moral inquiries up to their principles), between the most important proposition of practical reason and the often seemingly too subtle and needless remarks of the Critique of the Speculative Reason, occasions surprise and astonishment, and confirms the maxim already recognized and praised by others, namely, that in every scientific inquiry we should pursue our way steadily with all possible exactness and frankness, without caring for any objections that may be raised from outside its sphere, but, as far as we can, to carry out our inquiry truthfully and completely by itself. Frequent observation has convinced me that, when such researches are concluded, that which in one part of them appeared to me very questionable, considered in relation to other extraneous doctrines, when I left this doubtfulness out of sight for a time and only attended to the business in hand until it was completed, at last was unexpectedly found to agree perfectly with what had been discovered separately without the least regard to those doctrines, and without any partiality or prejudice for them. Authors would save themselves many errors and much labour lost (because spent on a delusion) if they could only resolve to go to work with more frankness.

Let me add one more remark here. Every step we take with pure reason agrees with all the essential points of the Critique of the Theoretical Reason. This is true even in the practical sphere, where we pay no attention to subtle speculation. The agreement is so close and direct that each step seems deliberately designed to confirm the Critique. Anyone can see this if they trace moral inquiries back to their principles. The most important proposition of practical reason agrees thoroughly with the often seemingly over-subtle and unnecessary remarks of the Critique of the Speculative Reason. This agreement is completely unlooked for and plainly visible. It is surprising and astonishing, and it confirms a principle that others have already recognized and praised. In every scientific inquiry, we should proceed steadily, with all possible precision and openness. We should not concern ourselves with objections raised from outside the inquiry’s sphere. Instead, we should pursue the inquiry on its own terms, as truthfully and completely as we can. Repeated observation has convinced me of something further. During an inquiry, a point has often seemed very doubtful to me when considered in relation to other, unrelated doctrines. When I set that doubt aside for a time and focused only on completing the inquiry, the point eventually proved to agree perfectly with what had been discovered separately. This agreement was unexpected. The separate discovery had been made without any regard for those doctrines and without any preference or prejudice in their favor. Authors would save themselves many errors and much wasted labor if they would only resolve to work more openly. The labor is wasted because it is spent on a delusion.