We have now to determine the nature of our problem more closely by reference to the history of modern philosophy. For this purpose I shall discuss and criticize in this and the succeeding chapter the Critical philosophy of Immanuel Kant. There are two reasons in particular which make this essential. At that point in history in the first place as at our own we find a break with tradition and the emergence of a new problem. In the second place the Critical philosophy is the most adequate of modern philosophies and indeed relatively to all the others completely adequate; so that in discussing Kant we discuss in principle all modern philosophy.

The term ‘adequate’ is here employed in a somewhat technical sense. The adequacy of a philosophy depends upon its range; upon the extent to which it succeeds in holding together the various aspects of human experience and exhibiting their unity. Kant is unique in the comprehensive unity of his thought. He does full justice to the first Cartesian phase of modern philosophy. As to the second it has been said with truth that all subsequent philosophies have been built out of the ruins of the Critical philosophy. I should prefer to say that until we come to those new tendencies which have been generated by the breakdown of tradition in our own time every significant movement in philosophy since Kant can be derived from the Critical philosophy by rejecting parts of it; and by reasserting what any of them has rejected the premisses for its refutation can also be found. Modern philosophies have often gained in coherence by this selectiveness; they have invariably gained this greater consistency at the expense of adequacy.

It was particularly unfortunate that by the accident of history objective idealism was the first type of philosophy so to be derived. We have come to regard Kant as the precursor of Hegel. It would be more true to say that the Critical philosophy was written to nip Hegelianism in the bud. The effect of seeing Kant through Hegelian spectacles is to shift the centre of gravity of Kant's thinking so that it falls within the Analytic of the Critique of Pure Reason. For Kant himself it falls within the Critique of Practical Reason: while the most important section of the first Critique is the Dialectic. Indeed the vital conclusion of the Critical philosophy as a whole—and it is one which points beyond Kant's own achievement—is that reason is primarily practical.

The immediate background of the Critical philosophy is the Romantic movement. It is a mistake to think of the Romantic movement as a literary and artistic revolution only. It is this of course; but it is also a revolution in social outlook and in thought. The French revolution was its most volcanic manifestation in the political field. Rousseau was its fountain-head. But in Germany the Romantic movement became the starting-point of an indigenous culture and a major factor in the creation of the German nation. Here the Romantics became philosophers. Lessing dramatist and art-critic produced the first of the philosophies of history in Religion as the Education of the Human Race; Herder the poet created the idea of Nature as a developing organic system in his Ideas for a Philosophy of History. The German romantics combined art and philosophy; they were thinkers as well as men of letters. The result was a process of creative philosophical development which culminated in Hegel. Hegelianism is the mature expression of Romantic philosophy.

The pioneers in this philosophical movement were Lessing Hamann and Herder. Between them they enunciated the main structural ideas of romantic thought. Lessing contributed the ideas of productive imagination and of development; Hamann the notion of reality as a tension of contradictions; Herder that of Nature as an organic unity. Of the three it is Hamann who is the most significant for our purpose. He was the founder of the Faith philosophy with Herder as his younger collaborator. These two were so far as I know the first philosophers to appeal to Hume against Kant reckoning Hume perhaps not without reason as a romantic like themselves.1 In his own day Hamann was an intellectual force of great influence and was known as the Magus of the North. He had had a revelation while reading his Bible in Bloomsbury on a business trip to London. He returned to Königsberg and spent the rest of his life propagating the new idea in the same city as Kant who knew him well.

The Faith philosophy rests upon a radical opposition between faith and reason. Hamann maintained that reason is an illusory guide to knowledge: we can know reality only by means of faith. The ground for this judgement is that reason works by the law of contradiction and so uses the absence of contradiction as the guarantee of truth. Faith on the contrary reveals reality as a coincidentia oppositorum a tension of contradictory elements. The further the process of reason is carried therefore the farther we are from knowledge. Now this antithesis of ‘faith’ and ‘reason’ does not carry its meaning on its face though it marks clearly and unambiguously a complete break with tradition. We need to know to what precisely the terms ‘reason’ and ‘faith’ refer. By ‘reason’ Hamann means all discursive thinking and in particular Cartesian rationalism and the mathematical methods on which physical science depends. In contrast faith refers to an inner experience a living apprehension which carries with it an immediate conviction of its own validity. Terminology is important here because in the development of the Faith philosophy after Hamann other terms were substituted. After Kant's analysis in the first Critique ‘reason’ was replaced by ‘understanding’. ‘Faith’ proved more difficult and various expressions were employed as synonyms for it—‘inner experience’ is one; ‘imagination’ is another. But in the end ‘faith’ is replaced by ‘reason’ which had lost its function. In the full development of Romantic philosophy therefore what began as a contrast between ‘faith’ and ‘reason’ appears as a contrast between ‘reason’ and ‘understanding’. In Objective idealism ‘Reason’ has come to denote what Hamann had called ‘faith’ and had contrasted with ‘reason’; though in the development its connotation has been enriched.

What is this capacity to grasp reality as a unity of opposites not discursively but immediately? The proper answer I believe is ‘aesthetic intuition’. It is the faculty of the mind which Kant calls ‘judgement’ (Urteilskraft) and which forms the subject of the third Critique. Significantly Kant associated it with the apprehension of beauty on the one hand and on the other with the idea of teleology. We can best define Hamann's contrast of faith and reason by saying that ‘reason’ is that in us which enables us to produce science; while ‘faith’ is our capacity for aesthetic experience. Hamann's proposal whether he is aware of it or not is to substitute the artist's standpoint for the scientist's as the basis of our knowledge of the real.

Now this is one of the common themes of romantic literature. ‘Beauty is truth’ says Keats. Browning is even more explicit when he makes the organist Abt Vogler say

The rest may reason and welcome: 'tis we musicians know.

and in the same poem we find a fine metaphorical expression of the dialectical principle linked with it;

I know not if save in this such gift be allowed to man

That out of three sounds he frame not a fourth sound but a star.

That this is also the theme of Romantic philosophy is confirmed by Fichte. Contrasting the standpoints of science and philosophy he remarks that the standpoint of philosophy is ‘counter-natural or artistic’.

Against this background the Critical philosophy was written. It is true that Kant was older than any of the thinkers we have referred to. Lessing was five years Hamann six years younger Herder twenty. But Kant's thought matured slowly. When the Critique of Pure Reason saw the light Kant was already fifty-seven. His Inaugural Dissertation of 1770 eleven years earlier shows him still in a pre-critical stage. Lessing died before the first Critique was published. Hamann came to his philosophical position early. He lived to read and protest against the Critique of Pure Reason but died in 1788 the year in which the second Critique came from the press. Even Herder was thirty-seven years of age when the first Critique appeared and though he had been a pupil of Kant's he joined Hamann in defending the Faith philosophy against Kant whose mature work they regarded as an attack upon their own position. In spite of Kant's seniority therefore all three should be regarded not as successors but as predecessors of the Critical philosophy.

The question of Kant's relation to the Romantic movement is thus decisive for any understanding of the Critical philosophy. Is Kant in the romantic camp or is he against it? The answer must be that his attitude is ambiguous. It is an attitude of critical sympathy. For the sympathy the evidence is conclusive. Kant admired Rousseau and considered the French Revolution though not its violence justified. When he described the dialectic of reason as a dialectic of illusion; when he wrote in the Introduction to the second edition of the Critique of Pure Reason that ‘he found it necessary to deny knowledge in order to make room for faith’ is it possible to believe that he had not Hamann in mind? Kant is therefore on the side of the romantics.

But it is a critical sympathy. It may well be that when he called his own doctrine a ‘critical idealism’ Kant was thinking of the Faith philosophy as an uncritical idealism. The important question is this ‘In what respect was Kant in sympathy with the Romantic philosophy; and wherein was he critical?’ A full answer would require a detailed account of Kant's doctrine and that is clearly out of the question even if I were competent to give it. We need consider only the relation of the Critical philosophy as a whole to the emergent problem of its time. The detailed exegesis of the Critiques is often controversial but on its general intent there is little ground for disagreement.

The radical difference between the Romantic philosophy and the philosophy of the Cartesians may best be expressed by reference to the function of the imagination in knowledge. There is a characteristic contrast commonly drawn in the earlier period between ‘the white light of reason’ and ‘the colourful fancies of the imagination’. The imagination is the bugbear of rational thought. It cannot be dispensed with altogether because memory depends upon it. Its task is to reproduce faithfully what has been experienced in the past. But it has a dangerous capacity for playing fast and loose with the material committed to its charge and producing attractive fancies of its own instead of recording facts. The task of the thinker is to suppress this productive spontaneity and hold the imagination to a purely reproductive function. The Romantics reverse this by insisting that the productive spontaneity of the imagination underlies all experience and particularly all cognition. This is the radical break with tradition. This is what underlies Hamann's contrast between ‘faith’ and ‘reason’. This is what instigates the substitution of the artistic for the scientific point of view. For the activity of the productive imagination is an artistic activity and it consists in combining the elements of experience in a way that is not itself given in experience. It is an activity of synthesis. If as the Faith philosophy asserts it is the source of knowledge then all knowledge is synthetic.

Now Kant fully shares this position. He not merely shares it he establishes it by careful analysis and demonstration. He goes farther in maintaining that our primary perceptual experience itself presupposes a transcendental synthesis even though at first sight it seems to be sheerly ‘given’. The root of all knowledge is the productive imagination. It is ‘a blind art hid in the depths of the soul’.

In what then does Kant's criticism of the Faith philosophy consist? Simply in this; that while the other Romantics were cheering madly because they had discovered the solution to the problem of knowledge Kant realized that they had made the problem a hundred times more difficult than it had been before. He discovery that the productive synthesis of the imagination is the root of all knowledge makes knowledge itself problematical. For it means that we invent our knowledge; that knowledge in some sense is fictional. If this is so how can it be knowledge? For knowledge is the discovery of what is already there; it is factual not fictional a receptivity and not a spontaneity of the mind. If the knower is the imaginative artist how are we to distinguish between fact and fiction? Thus the new discovery which he accepts becomes for Kant the new problem for philosophy. Characteristically he generalizes it concentrates upon its purely formal and universal aspect and formulates it as the problem of the Critical philosophy ‘How are synthetic judgements a priori possible?’

Instead of attempting to summarize Kant's abstract argument I shall risk the raising of eyebrows among the pundits and offer a concrete interpretation of its meaning so far as it concerns us here. If we take the new discovery at its face value without criticism as the romantics do what must happen? We shall subsume all aspects of our experience under the form of aesthetic activity. In particular we shall develop an aesthetic science and an aesthetic morality. Both theoretical and practical experience will be determined by artistic standards. Both the good and the true will become ‘that which satisfies the mind’ the one in the practical the other in the theoretical field. This can only result in the destruction both of science and of morality. What satisfies the mind is the beautiful. The good however is what we ought to do whether we like it or not. The true is what we ought to believe however much it goes against our inclinations. If then all experience is shown to rest on the blind art of an imaginative synthesis the problem it sets philosophy is to distinguish science and morality from art; and! indeed to distinguish also the mere play of fancy from art as a serious and deliberate activity of mature human beings. For this reason Kant had to write three Critiques the first about science the second about morality and the third about art; and throughout he was concerned to defend rationality against the romantic attack. If he had to deny knowledge he would not with Hamann repudiate reason.

Let us begin with the denial of knowledge. Knowledge is in some sense the discovery of what exists independently of any activity of ours. If we construct our knowledge if it depends at all upon a spontaneous inventive activity of the mind then there is no escape from the conclusion that we can never know the world as it is in itself independently of our ways of apprehending it. A sheerly passive receptivity in which Reality should impress itself upon the blankness of our ignorance is completely ruled out. Reality as it is in itself is unknowable. This is the famous doctrine of the Thing-in-itself of the noumenal world and it is Kant's denial of knowledge. The discovery of a law of nature by science is the construction of a mathematical formula and mathematical formulae are human inventions. So far Hamann was right in his attack upon discursive thought.

What Hamann failed to notice however is that this applies equally to what he called ‘faith’. Aesthetic intuition for all its immediacy is a synthetic activity. Suppose we rule out thought altogether what remains? A pure immediacy of sense perception in which a world of objects is revealed to us? Not at all. The artist who really does try to see the world with an innocent eye discovers simply a pattern of coloured shapes which changes as we watch it; or a pattern of sounds in which a theme repeats itself with variations. For pure intuition there are no ‘objects’; these have to be discriminated against a background. When we become as little children who have not yet learned to discriminate objects fade into their background and leave only a pattern of differences in space and time.

At this level of pure intuition however the distinction between perception and imagination has disappeared. What we see with closed or open eyes is the same thing—a spatio-temporal pattern of coloured shapes. And the space and time which holds the elements together is the same space and the same time in both cases. When we shut our eyes and imagine a scene what we see is a pattern in space and time. When we open them different colours and shapes take their place in the same space and time. So space and time are the forms not of our perception but of our intuition whether perceptual or imaginary. How then do we come to perceive not a pattern of sense-data but a world of objects? The answer must be that we do this by a further activity of discriminating within the pattern certain groups of elements which we isolate and hold together against the rest of the pattern as a background. We now do this automatically by recognizing objects which we have often discriminated in the past; but once upon a time in our infancy we had to learn to do it. At the other end of the scale the scientist returns as it were to the original pattern and discriminates it afresh in different groupings of its elements to produce a picture of the world very different from the one that our ordinary perceptual discrimination offers us.

We must confine ourselves however to those general features of Kant's doctrine which are essential to our particular purpose. The major point which applies both to the theoretical and to the practical problem both to science and to morality is this. The pattern which underlies all aspects of our experience depends upon our sensuous equipment. Its elements are our sense-data. But it is a pattern; it has a structure which unifies it; and that structure is spatio-temporal. That structure too is ours. It is the form of our imagination common to the world of fancy and of fact. So it follows that if we know the world at all it is the world as it appears to us taking upon it the forms of our mode of intuition and not the world as it is in its independent being. How then can our imaginative construction ever be true or false? How can our actions ever be right or wrong? How can we ever get beyond a synthesis that satisfies us?

Kant answers by reference to law. There are two types of law which we recognize—the law of Nature and the moral law. The first prescribes how things must happen: it is descriptive. The second is normative and prescribes how things ought to happen. The synthesis of art is a free synthesis; we can combine the elements of intuition—sounds and colours—as we please to satisfy ourselves. But when we think in order to discover the truth or when we act with a view to doing what is right the synthesis is not free; it must conform to a law which dictates its form. We can indeed think what pleases us we can do what satisfies our inclination; but so neither knowledge nor morality is possible. They become possible through conformity to a rule.

Consider knowledge first. How is the spatio-temporal pattern transformed in perception into a world of objects? By combining elements selected from the pattern in the right way. What is the right way? The way that does transform the pattern into a world of objects. This doesn't seem to mean anything; it appears to be simply ringing the changes. But is it? When I say that one and one make two I enunciate a rule for counting. If you ask why I shouldn't say one and one make three the answer is that you can say what you please; but if you want to count that is the rule and if you don't count in that way then you just aren't counting. There is only one way to count things. In that case the rule both prescribes the method of counting and describes the process of counting. So in thinking. You can think as you please of course; but if you are trying to get at the truth then there is only one way to think. Any other way is not really thinking. So thought is an activity which conforms to a law—or a system of laws which is one law—and we can call the law either a rule which prescribes how we are to think or a description of the process of thinking. It is at once the rule for thinking and the form of thought.

Knowledge says Kant is the determination of an object by means of concepts. To determine an object we can see is to combine the elements of the sensuous pattern—or rather a selection of them—so as to constitute an object; or more generally to break up and recombine the elements so that we perceive a world of objects. But what is a concept? It is a rule for doing this. The concept of a triangle says Kant is a rule for the construction of any triangle. It is equally the form of a triangle the means by which we recognize a triangle when we see it. Now the process of determining an object is a process of under-standing; it is a process of making sense of the meaningless pattern of sense-data. If it seems that Kant is making mountains out of molehills one should ask this question. What is there that is common to a coloured shape and a sound that could link them together when we see a blackbird on the apple-tree and hear it singing? Surely nothing at all. Sound and colour are sheerly different. If they are to be combined there must be a rule for their combination if any object is to emerge. The rules are concepts; concepts belong to the understanding; and the principle of understanding which makes knowledge possible is that there is only one way in which we can make sense of our experience; that is to say determine it by thought as a single world of objects in systematic relation with one another.

The concepts of the understanding are either pure or empirical. Empirical concepts are derived from experience. They are the forms of things—that is combinations of sense-data—which we have learned to distinguish and which recur again and again. As rules they are rules for recognizing new instances of a type with which we are already familiar. Because they are formed in this way there is always a possibility of error in their use. I may perceive a rabbit in a field on a country walk and it may turn out as I come nearer to be the stump of a tree. I saw an object and misjudged it by subsuming the sense-data under the wrong empirical concept. Consequently all empirical determinations of things require verification by reference to further experience. But these empirical determinations are possible only because there are pure concepts which are not produced in this way but are completely general and presupposed in all experience. For example there is nothing in the spatio-temporal pattern in intuition to suggest that its elements should be grouped into objects. This implies that if we do discover objects in it we must already possess the idea of an object as a universal form or a universal rule for determining any object whatever. These pure concepts Kant calls the categories of the understanding.

The categories themselves form a compact system. Each implies the others and there is no place for any extra. Together they constitute the law to which thought must conform; or—for it is the same thing—the form of the process of thinking. Thinking again is the process of imaginative synthesis in so far as it conforms to the rule. It organizes the data of intuition in the form dictated by the categories. Now the data of intuition have their own form which is spatio-temporal. So when thought determines objects the form of thought combines with the spatio-temporal form of intuition to produce a schema or a schematic system and this schematic system is the form of the world which we perceive. When the productive imagination conforming to the rules of the understanding generates this spatio-temporal schematism it provides a form which is both the form of our knowing and the form of the world we know. This is why Kant says that the understanding gives laws not merely to our search for knowledge but to Nature. For the structure of thought and the structure of Nature are necessarily identical.

This brings us to the supreme principle of knowledge which Kant echoing Descartes calls the ‘Cogito’ the ‘I think’ which accompanies all my representations; or in his technical terminology the transcendental unity of apperception. If we abstract from thought what remains is the spatio-temporal pattern of sense data which Kant calls the manifold in intuition. Here as we saw there is no distinction between imagining and perceiving. Every element in the pattern is an intuition of mine. It is this reference to one and the same centre of experience which gives the data their primary unity as elements in our experience. They are all mine. But when we drop the abstraction this experiencing centre (the ‘I’) becomes the thinking Self—not the ‘I’ but the ‘I think’. To think is to determine an object; so the fact that the manifold in intuition stands under the unity of thought transforms it into a world of objects in systematic relation with one another in space and time. The form of thought which the pattern takes on is the form of a world that is comprehensible to us. The elements of the pattern which were limply ‘my sensations’ are referred to the object as my representation of it. The red which I sense becomes the colour of a rose; the sound I hear becomes a blackbird's song. Intuition is discriminated into imagination and perception. So I perceive the world and distinguish between the world and myself. I am the Subject; the world is the Object. My imaginings are subjective; but what I think is objective. This dichotomy of Subject and Object is the abstract form of all our knowledge. Subject and Object are correlatives or polar opposites which depend upon one another and the principle of their correlation is that the form of thought is the form of the object. For this reason Kant insisted that the transcendental unity of apperception is an objective unity. To know is to apprehend an object. Truth is objectivity.

If we ask at this point whether Kant has solved the problem of distinguishing between art and science by showing how knowledge can be at once an invention of the mind and a discovery of the truth the answer is ambiguous. In a sense he has. He has distinguished clearly between an imaginative synthesis that is fanciful that satisfies us and a synthesis that is necessary and which we are under an obligation to make and to accept. The antithesis of faith and reason in Hamann's philosophy which substitutes the inner conviction of the truth of an imaginative synthesis for discursive thought as the source of knowledge has been completely exploded. On the other hand the new idea which it embodies has been accepted and restated as the result of a more highly developed analysis. For lack of this analysis Hamann had drawn hasty and unwarrantable conclusions. But knowledge is still synthetic and the constructive activity of imagination remains its basis. Though it needs thought to discriminate and verification to confirm it the world is still dependent on the mind for its very existence. I construct the world I know even if I construct it in accordance with the laws of thought.

Yes but this is ambiguous. As a particular individual I am only one object in the world which is determined by thought. My thinking itself as a process that takes place in my mind is part of the process of events in the world. The world that I know contains me as a part of it determined by the categories and if I exist then in the same sense the world exists outside me and totally independent of me as a particular individual. That there is only one way to count does not mean that I cannot miscalculate; and my private thinking may lead me into error. My errors of judgement make no difference to the structure of the world. The objective unity of the ‘I think’ is not the unity of any particular experience. It is the unity of all possible experience; the logical correlate of the one existing world of objects.

Why then should not the Faith philosophy take up all this into itself and so reach a higher stage of its own development? There is only one reason; it is Kant's insistence that there is a real world unknowable by us of which the world constructed by thought is only the appearance. We know only phenomena. Of the real world of things-in-themselves we can only say that we know that it exists; but we cannot know at all what it is. It remains for ever undeterminable by thought. As is well known the mature development of the romantic philosophy in Fichte Schelling and Hegel depended upon accepting in principle Kant's analysis and rejecting the doctrine of the two worlds. In this way the world we know becomes again the real world as it was for Hamann. More important still knowledge remains though at a higher level fundamentally aesthetic. For truth becomes simply the comprehensive coherence of the imaginative synthesis.

It must be admitted that the doctrine of things-in-themselves is objectionable; and that if it goes then Kant's theory of knowledge can easily be transformed—as in fact it was—into a dialectical idealism like Hegel's. The result is a more coherent philosophy; but also a much less adequate one. For what we have considered of the Critical philosophy is only the first two parts of the first Critique. So far there is no good ground for believing in the existence of a noumenal world. The reason for the doctrine of things-in-themselves is foreshadowed in the final lection of the first Critique but only becomes fully apparent in me Critique of Practical Reason the function of which is to distinguish morality from art and so to defeat the romantic tendency to produce an aesthetic determination of the good.

In the Dialectic of the Critique of Pure Reason Kant for the first time comes to grips with reason itself. It is the understanding which is concerned with the knowledge we have of the world of objects. The understanding does its work by applying the categories to the manifold of intuition. Reason must be distinguished from understanding. It has its ‘ideas’ which are concerned with the ultimate nature of things with the unconditioned; while the understanding determines objects by discovery of their conditions. We are therefore tempted to use the Ideas of reason which are pure formal concepts to provide knowledge of what lies beyond the conditioned existence of the world we know in sensuous experience. There arises in this way a dialectic of reason and the product of this dialectic is metaphysics. What then is the place of reason in knowledge? Kant's answer is twofold. First the dialectic of reason is a dialectic of illusion. All metaphysics that is all knowledge of the supersensuous is nonsense. We can know only by determining the structure of a world given in sense perception. The moment we speculate about what cannot be given perceptually we can produce theories in plenty but since they have content and are not merely formal they require verification; and they cannot be verified. Here Kant reveals himself as a positivist. Secondly the ideas of reason being concepts are rules or principles which guide the understanding in the production of knowledge. Thus reason is a higher principle than understanding. It is not itself a faculty of cognition but it provides the understanding with the practical rules for the use of the categories in the search for knowledge. So as regards reason provided again that we distinguish it from understanding Hamann is right. The use of pure reason to give us knowledge of reality is the source of illusion.

It is only when we turn to consider our practical experience as agents and not our theoretical experience as thinkers that we discover the true character of reason. This is the final and quite revolutionary conclusion of the Critical philosophy. Reason is primarily practical. It is not a faculty of cognition but a faculty of rules. If it has a secondary theoretical function that is because thinking is something that we do; so that Reason is necessary to provide the rules that guide our search for knowledge. The understanding which is theoretical is as it were the viceroy of reason in the theoretical field. Reason itself is the ultimate legislator. This is the dignity of reason. For Kant—and as a philosopher—action is more important than knowledge. If it was important to distinguish science from art it is much more important to distinguish morality from art. The major danger which Kant saw in the uncritical idealism of the romantics was this confusion—the danger of substituting aesthetic for moral standards in the determination of conduct. Indeed science itself as a human activity depends upon practical rationality.

The problem of action is set by the antinomy of freedom and determinism. Morality presupposes freedom. Science presupposes determinism. Knowledge is the determination of an object but this determination is theoretical. If the object can be determined by thought by a judgement which may be true or false then it must already be determinate. If it were not determinate no judgement of ours could be either false or true. Action however is the determination of something not in theory but in actual fact. To act is to make something other than it would have been if we had not determined it. In knowing an object we make no difference to it: in acting upon it we do make a difference to it. Now our actions as events in the world we know must be as completely determined as everything else. But in action we presuppose that we determine the world by our actions. The correlative of this freedom is that the world which we determine in action must be indeterminate capable of being given a structure that it does not already possess. We can only know a determinate world; we can only act in an indeterminate world. Therefore if we really do act if our freedom of will is not an illusion the world in which we act must be unknowable.

This is the fundamental ground of Kant's belief that there must be a real world behind the phenomenal world which we cannot know. If there is not it becomes impossible to distinguish between an aesthetic and a moral praxis. An aesthetic praxis would consist in determining a future state of affairs—an ideal—as the good and therefore as the end to which our actions ought to be the means. If we could do this then all we should need to enable us to act rightly would be the scientific understanding of the causal process which would connect the present state of affairs with the future state which is the good. We could plan our Utopia. Why not? Because the determination of the good as an object in time is necessarily aesthetic; it rests upon an inner conviction which cannot be verified. Conceptual thought could at most tell us that such a state of affairs is possible and how we can realize it—but not that it is the good. If we are none the less convinced that it is this can only mean that we have presented to our minds an imaginative synthesis which we feel to be fully satisfactory. The hall-mark of the aesthetic stand-point is that it defines both the true and the good as ‘that which satisfies the mind’. This is the proper definition of the beautiful and the satisfaction is a disinterested satisfaction. But if we confuse this with the morally good then our Utopia takes on the character of a moral determination. It becomes universally obligatory as the objective of action. Suppose then that there is a person who finds himself possessed of such a vision of the good convinced by an inner sense of felt and disinterested necessity; suppose that he happens also to possess the power to compel other men to cooperate with him in realizing this good—then it becomes not merely right but a moral duty for him so to compel them. Their refusal to accept the objective must be judged to be immoral. Being wise after the event we can say that Kant was warning the German people that if they accepted the romantic philosophy they would find themselves in the end with their freedom lost under the fanatical dictatorship of an Adolf Hitler. There is an interesting statement in Kant's Lectures on Ethics which even if it is precritical is worth a reference. He has been saying things with which we are familiar in the Fundamental Principles of the Metaphysics of Ethics; and he suddenly breaks off and warns his students against the dangers of a purely formal ethic. ‘God wants men to be made happy by men’ he exclaims ‘and if only all men united to promote their own happiness we could make a paradise of Nova Zembla.’2

If the good then cannot be a determinate object at which we aim if it is theoretically unknowable there is no way in which we can determine the rule of right action as the means to the good. However far science may advance knowledge brings us no nearer to an empirical determination of what we ought to do. We can determine rules of prudence through knowledge and improve them as our knowledge improves. But even if they were certain as they never can be they would only determine the means to satisfy our inclinations and these of course vary from one man to another and from time to time. The rules of moral action must be rules for achieving an end which we cannot theoretically determine. How is this possible? Only by acting in conformity to a law which is determined for us by reason. The primary function of reason as legislator is to prescribe the moral law which must then be a law which is universally and absolutely binding on all rational beings and also a law which of necessity we cannot understand. For to understand it we should have to determine the good as an object in time—as a future Utopia as the end of the process of Nature. To do this would be to substitute an aesthetic for a moral determination.

We said that all this must follow unless our belief in freedom is an illusion; and this as we see means unless there is a world in which our actions determine the good without our knowledge. But how do we know that we are free agents? Kant's reply is that we do not and cannot. In the moral field we are beyond the limits of knowledge. Then why should freedom not be illusory and determinism the truth? Kant answers that we must believe in freedom. We are in the field of faith where only belief is possible. But beliefs like judgements are not all on the same level. There are necessary beliefs; there are beliefs which though not strictly necessary are yet reasonable beliefs; and there are beliefs which are not even reasonable but merely fanciful. The belief in freedom is necessary in the strictest sense. It is a necessity of reason like the belief in the law of contradiction. To reject it is to reject reason itself not merely in the practical but also in the theoretical field. For we have already seen that the understanding with its presupposition of total determination in the world rests upon the law of reason. Indirectly therefore the belief that our judgements can be true or false depends upon our belief that our actions can be right or wrong of which indeed it is a special case. To believe something because it is true and not because we want to believe it is to act in the field of theory from duty and not from inclination. Reason is primarily practical.

On the rational necessity of a belief in freedom Kant grounds two other beliefs which though not in the same sense necessary are yet he holds eminently reasonable. They are the beliefs in God and in immortality. Rules of action are rules for the realization of an end. If then we are bound by the laws of reason to act morally it is reasonable to believe that the course of action so dictated is the means to the realization of the good even though we cannot determine what the good is. Also there is no determinable relation between the ends which we realize in this world by doing our duty and the good. The good in realization is happiness. There is no determinate relation in experience between duty and happiness. It is reasonable therefore to believe in a future life in which the good is actually realized as the end to which moral action is the means. And since the union of duty and happiness—as means and end—must be a practical synthesis it is reasonable to believe in a supreme Agent who achieves this synthesis in practice.

Having distinguished science and morality from the mere synthesis of the productive imagination through the idea of law; having established reason as the lawgiver and grounded the theoretical in the practical law of reason; Kant turns in the third Critique to the field of art. For there is clearly a distinction to be drawn between the blind art which underlies all experience and the deliberate artistry which is one of the modes of human activity with a principle of discrimination of its own. In the decade that had elapsed since the publication of the first Critique the development of the Faith philosophy had gone ahead. Herder in particular was applying its principle to Nature and the result was the conception of Nature as an organic system developing itself progressively in history. A new science of Nature based upon aesthetic intuition which claimed to supersede mathematical physics was now not merely proposed. It was in active prosecution. A Critique of (Aesthetic) Judgement was a necessary completion to the Critical Philosophy.

The third Critique is no more a philosophy of art than the second is a moral philosophy. Its purpose is to analyse the discrimination of the beautiful and so to distinguish it from the discrimination of the true. To emphasize this we might reasonably formulate the question as follows: ‘What is it that inclines us to attribute truth to the artistic synthesis and even a superior truth; and is there any justification at all for this tendency?’ We can summarize Kant's answer in a few sentences for our present purpose. Here without question we are dealing with imaginative synthesis in intuition. But we discriminate among such syntheses by judging that some are beautiful and some are not. Now this judgment is not grounded in a concept; it is not therefore an understanding of the object and consequently it is not knowledge. The ground of our judgement is a feeling of satisfaction which refers to the form of what is apprehended. But the pleasure we find in contemplating the beautiful is not the satisfaction of a particular need or interest of ours. It is a disinterested satisfaction. It appeals to us not as particular individuals but as cognitive beings. It satisfies the mind. Because of this our judgement claims to be universally valid and not a matter of private taste. This can only mean that the form of the synthesis is adapted to our mode of apprehension through the interrelation of intuition and conception. It fulfils the conditions for our cognition perfectly and is therefore an ideal object. But it is not cognized; for if it were it would be in accordance with a rule and the validity of our judgement could be demonstrated or verified by reference to experience. This explains our tendency to treat the immediate inner conviction that the beautiful occasions in us as a guarantee of truth. It also reveals the illusory character of this conviction. The fatal error is the assumption that truth is what satisfies the mind. Truth is what is determined in accordance with a law and can be guaranteed only by reference to the law.

This does not however mean that aesthetic judgement has no function at all in relation to knowledge. Provided that we do not confuse art with science by confusing the judgement of beauty with the judgement of truth as the Romantics do we may find it not merely useful but even necessary to the progress of knowledge. It is clear—though this is not the important point—that the formation of a scientific hypothesis apart from its verification is analogous to the production of an artistic synthesis. But the essential issue is the relation of the judgement of beauty to the teleological idea; for the critique of teleological judgement is Kant's mature criticism of all romantic thought of which Hegelianism is the full flowering.

In analysing the teleological idea Kant has in mind any attempt to attribute purposiveness to Nature. This is the root concept of any attempt to produce a philosophy of history like Lessing's or Herder's or to represent Nature as a developing organic unity as Herder was doing. In exhibiting the intimate relation of teleology with beauty Kant was recognizing the essentially artistic standpoint of the Faith philosophy. In recognizing beauty we are recognizing a form of imaginative synthesis which is adapted to the satisfaction of our minds. It is as if the object had been constructed in order to give us a disinterested pleasure in its contemplation. In the case of a work of art this is actually so. In the case of natural beauty it seems to be so. We think of Nature as an artist painting the rose and composing the blackbird's song for our delight. The adaptation though it may be to our minds is in the object. It lies in the inner relations of its parts in the harmony and balance of their organization so that each part has a function in the felt unity of the whole.

Now if we seek to know Nature in this way we must conceive her as a work of art and therefore as an organic unity in which every part is functionally related to all the others and the whole is the embodiment of a developing purpose. But what would this purpose be? Only the production of an ideal object for our satisfaction as cognitive beings. We can only view Nature in this way by taking ourselves as the end to which creation strives. Such a purpose is wholly subjective. We are objectifying it illicitly by projecting our purpose of knowing Nature upon the world and representing it as Nature's purpose to be known by us. No doubt we shall find ways of concealing the glaring subjectivity and egoism from ourselves in mystical language. We may talk of ‘Nature’ or ‘Reality’ or ‘the Absolute’ knowing itself in us. But such language is only the regular subterfuge of transcendental metaphysics.

Nevertheless there is a necessary part to be played in knowledge by the teleological idea. Its function says Kant is heuristic and it provides a regulative principle in the search for a true knowledge of Nature. A blind investigation will be fruitless. We must bring with us an ideal—a formal notion of what would satisfy our minds if we could find it. Even in mathematical sciences we must assume that Nature is governed by a system of empirical law which is adapted to our modes of knowing. We cannot know in advance that this is so; and we may find as science progresses that the complexities of empirical structure are beyond our powers. Again the idea of Nature as an organic whole is illusory since it involves the thought that Nature is a finite totality when in fact it is infinite. Time does not end with us and we have no ground for assuming that time must have a stop. Yet within nature we do find totalities which arrest our attention with this immediate conviction of purposive functional unity. Plants and animals are only the most obvious examples. Nature does seem to produce objects which are adapted to our modes of cognition and to display a purposiveness in their organization even if it is a purposiveness without a purpose. In such cases Nature seems to designate objects appropriate to our search for knowledge. But the satisfaction our minds find in their discovery is not itself knowledge. It selects an object to be known. To know it we must accept it as a whole within Nature and analyse its construction. That analysis and the knowledge to which it leads are necessarily mathematical. Even a science of biology must therefore be a biophysics.

So Kant defines critically and negatively his relation to the Romantics. We might sum up his final argument in contemporary language quite simply. All knowledge of Nature is empirical; and all empirical knowledge must be verifiable. If it is not verifiable it is illusory. What is characteristic of the artist's productions and of artistic experience as a whole is that it does not and cannot verify its synthesis. The poet may anticipate the scientist; but even if his insight prove true it is not thereby scientific. It is at best the suggestion of a hypothesis which science may in its own systematic development find verifiable.

• 1. ‘Hume is the man for me for he at least honours the principle of belief, and includes it in his system, while our countryman (sc. Kant) is always chewing the cud of his causal whirligig, without a thought for belief. I don't call that honest.’ Hamann in a letter to Herder, 10 May 1781.
• 2. I. Kant, Lectures on Ethics (trans. L. Infield), p. 55.