1. The Subject of the Critique of Teleological Judgement

THE transference of general methodological assumptions which have proved useful in one field of inquiry to others seems reasonable policy even if its success or failure can rarely be predicted. Important historical examples of such transplantation are on the one hand the spread of teleological explanation which is typical of much medieval thought; on the other hand the spread of mechanistic explanation which characterizes the rise of natural science.

There can be little doubt that Aristotle's conception of explanation as being mainly teleological arose from his great interest in the phenomena of organic life and social organization and from the fact that he found this mode of interpreting them intellectually satisfying. For him and his medieval followers the understanding of nature was not as intimately linked with the power of making new things as it is for the modern scientist. They, consequently could be content with teleological explanations in their thought not only about man but about any other subject matter including inorganic things : and processes. .

After the great success of the mechanistic principle in physics which culminated in the Newtonian system this new mode of thinking was with ever increasing frequency applied in other inquiries. Thus to give an outstanding example, Hobbes aimed at a science of man and society deliberately modelled on physics. Individual men are regarded for the purposes of this science as atoms, all alike in being selfish, and in being governed by the same psychological laws of motion. If  we know the 'initial positions' of the atoms of any society, we can predict its future' states'. This is one of Hobbes's central assumptions however he may have obscured it at times by other considerations. It is still with us in a great deal of psychological and economic thought. The greatest achievement of mechanistic explanation outside physics is Darwin's theory of natural selection and the evolution of species which centres in the mechanistic conception of the survival of the fittest, i.e. the survival of individuals and species in their environment.

Compared with the rapid progress of the physical sciences which was a direct result of the new non-Aristotelian ways of interpreting nature, the progress of the biological and in particular the social sciences has been slow. As a consequence of this the superiority of mechanistic over teleological explanation in these sciences has never been generally recognized by  all the workers in these fields. No respectable physicist would dream of a teleological physics, however much his faith in the Newtonian system were shaken by recent developments. But the possibility of a teleological biology is by no means regarded as absurd and is at least thought worthy of refutation by biologists and philosophers.

Kant lived at a time when mechanistic explanation had already achieved its greatest triumphs in physics. In biology on the other hand and in the social sciences it was at best a scientific programme and certainly not a scientific reality. To Kant it seemed absurd to hope' that one day there would arise a second Newton who would make intelligible the production of a single blade of grass in accordance with laws of nature the mutual relations of which were not arranged by some intention.  “Such insight”, he says, “must be utterly denied to man.”

Kant believed both that a scientific explanation must be mechanistic (Newtonian) and that biological phenomena - though in some respects open to it - do not admit completely of a mechanistic explanation. It therefore becomes an important task of the critical philosophy to examine the notion of purpose and the manner of its legitimate and illegitimate employment in science, More particularly it becomes necessary "to examine the relation between mechanism and teleology.

In our own day the situation with regard to the problem treated in the Critique of Teleological Judgement has become more complex. :Mechanistic explanation has in many cases, even in physics, been unsuccessful and other quite different" types of non-teleological explanation have been successfully tried there. The issue is no longer simply between mechanistic and teleological explanation. Nevertheless while mechanistic explanations are sought and employed even now in every science, the nature and function of teleological judgements in the biological and social sciences is still under discussion.  We have here another case in which Kant's examination of a philosophical problem has more than a merely historical interest.

In the widest sense of the term all judgements involving a notion of purposiveness or purpose are teleological. The differences between their various types are so great that it would not be profitable to try to deal with all of them at the same time. In the second part of the third Critique Kant is mainly concerned with a class of teleological judgements which can be fairly well distinguished from others by two characteristic features: (I) they are intended as explanations of the existence of things and (2.) they are not explanations in terms of such human purposes and desires as vary from person to person. They are judgements about purposes in nature. Examples would be 'Small animals exist for the purpose of being eaten by larger ones' or ' The parts of any organism exist for the purposes of and by means of each other and of the whole to which they belong'.

Judgements about specific human purposes, say 'The Eskimos build igloos for the purpose of shelter', do not give rise to any new problems beyond those raised by judgements about desires. These have to some extent been discussed in the Critique of Practical Reason, which, among other things, is an account of the conflict between desire and duty. There it has been shown that an explanation in terms of human desires and purposes does not differ from any ordinary causal explanation.

There are, however, teleological judgements which are not meant to explain the existence of anything and are not even about existing things but merely about presentations. Aesthetic judgements, according to Kant, state a necessary connexion between 'the mere presentation of an object' as purposive and the pleasure accompanying it. This . purposiveness', we re- member, is apprehended apart from any determinate purpose. Indeed if it were related to a human purpose or a purpose in nature the experience would cease to be purely aesthetic. An aesthetic judgement, Kant insists, is based on that' which in my mind I do with a presentation' and not' on anything with respect to which I depend on the existence of the [presented] object'.

Another type of teleological judgement is found in pure mathematics. For example, the theorem that the geometrical locus of all triangles with a given base and a vertical angle opposite it is a circle, is a suitable ground for the construction of any of these triangles. Kant regards this suitability as a purposive connexion - a 'purposiveness without purpose'. In being independent of specific human desires it is similar to aesthetic purposiveness. It is, however, not the result of the harmonious interplay of the imagination and the understanding, which is necessarily accompanied by pleasure; but is a purely 'intellectual purposiveness' rooted in synthetic a priori principles and a priori constructions in space.

Teleological judgements about' purposiveness' in mathematics differ from judgements about purposes in nature, in that. they are not about existing things. . Because pure mathematics does not deal with the existence, but only with the possibility of things - namely with pure perceptions corresponding to its [a priori] concepts - . . . all purposiveness found in mathematics must be regarded as merely formal. In distinguishing between different types of teleological Judgement Kant uses the over-worked terms 'formal' and ‘material’, ‘ subjective' and' objective' in a special sense which is not easily grasped and does not, so I believe, fully agree with his own terminology in other parts of the critical philosophy. These disagreements are of little consequence. Teleological judgements are material if they refer to existing things; otherwise they are formal. Teleological judgements are, moreover, subjective if they refer to the feelings or desires of the person making them; otherwise they are objective.  

He thus seems to have the following fourfold classification in mind: (1) formal and subjective teleological judgements of which aesthetic judgements are his only example; (2) formal and objective teleological judgements of which judgements about certain connexions in mathematics are his only example; (3) material and subjective teleological judgements, i.e. judgements about human purposes; and lastly (4) material and objective teleological judgements, i.e. judgements about purposes in nature. Whether or not this classification is as neat as he believed, Kant makes it quite clear why he regards all aesthetic judgements, some mathematical judgements, and all judgements about human and non-human purposes as teleological. In the illuminating modem fashion of speaking, introduced by Wittgenstein, we might say that by calling all the above- mentioned types of judgement' teleological' Kant has drawn attention to important general family resemblances between all of them and to even closer resemblances between those which he classifies under the same heading.

2. The Notion of Purpose in Nature

According to the doctrine of the Critique of Pure Reason all phenomena stand under the synthetic a priori principles of the Understanding. Thus in particular every change happens to a substance or substances and must have a cause. To give a causal explanation is, schematically speaking, to state that an event of type A involving certain things causes an event of type B involving these or other things. Consider, to give a very simple example, the sequence of events in which a weaker animal, such as an antelope, comes near a stronger carnivorous animal, such as a lion, and is subsequently killed and fed upon by the lion. A causal explanation of these events, if we can £nd one, would consist in showing that the two events belong to two distinct types and that an event of the first type leads with causal necessity to an event of the second type. The principle of causality does not, as Kant has made quite clear, in any way guarantee that we should be able to provide causal explanation in any particular case.

On the other hand a teleological explanation of the events described in our example might be given by the statement that antelopes exist for the purpose of sustaining lions. Generally speaking a teleological explanation of the sequence of two events consists in a statement that all or some of the things involved in the first exist for the purpose of all or some of the things involved in the second. (While a causal explanation pre- supposes at least a sequence of two distinguishable events a teleological explanation is possible in the case of a single event involving more than one thing. It would consist in stating that one or more of them exist for the sake of some other.)

From what has been said so far it is clear already that a teleological explanation is quite different from a causal one. The latter does not, strictly speaking, explain why (= for what purpose) events happen or things exist, but only how. Indeed from a completely mechanistic point of view the question why events happen or things exist would be illegitimate, because it is in principle unanswerable. The same sequence of events may be capable both of a causal and a teleological explanation. The latter is then, as it were, superimposed upon the former. The antelope's coming close to the lion and its subsequently being killed is then first explained as an instance of a causal connexion which in turn is teleologically explained by the statement that antelopes exist for the purpose of being food for lions. Frequently, however, it is the sequence of events itself and not their causal connectedness that is teleologically explained.

Cases which are likely to remain incapable of causal explanation often seem to urge us to attempt direct teleological explanation. Whether we resist this urge or succumb to it is here not relevant. 'Experience', as Kant puts it, 'leads our judgement . . . to the notion of a purpose in nature only if what is to be judged is a relation between cause and effect and if we are unable to see their necessary connexion [als gese~/ich einzusehen] by any means except by  making the Idea of the effect into the condition of the possibility of the cause.'

The purposiveness which we thus see in, or project into, nature, is either an outer purposiveness, as in our example, where one thing serves the purpose of a different thing; or else it is an inner purposiveness which is characteristic of one whole thing. The former' is merely relative and. . . accidental to the thing to which it is attributed '2 and is expressed by 'hypothetical' judgements. The latter is ascribed to living organisms. An organism is, so to speak, a purpose unto itself - a purpose in nature(Naturzweck) without qualification, and not merely a relative one. The peculiarly intimate interdependence between the parts and the whole of an organism is expressed by 'absolute teleological judgements '. The notion of an organism is characterized as follows: First, a thing is an organism only if 'the existence and form of its parts. .. [are] possible only through their relation to the whole'. The description has a verisimilitude. The existence and form of a hand, say, may indeed quite plausibly be considered impossible in separation from the body to which it belongs. Second - and again Kant's description seems convincing - the parts of an organism constitute' the unity of a whole by being mutually the cause and effect of their form'. Lastly, many organisms, and only organisms, have the power of reproducing themselves.

An organism is thus not only an organized thing but also a thing organizing itself. In a mechanism the parts are conditions of each other's function. In an organism they also exist through each other and in a sense produce each other. A blade of grass is quite different from a clock, which does not grow or produce other clocks. An organism' is thus not merely a machine: for that has only moving power; but it also has within itself a formative power with which it endows the various kinds of matter which lack it (thereby organizing them) . . .' 

So far the argument has only shown (1) that the notion of purposes in nature is in fact being employed, and that (2.) our incapacity, temporary or perhaps permanent, of providing a satisfactory causal explanation of certain natural phenomena inclines us to make use of it. It is now time, as we have done so often before, to leave these evidences of our' possession' of the notion and to consider whether or in what way our use of it is justifiable. We have to recall the well-documented warning of the Dialectic of Pure Reason that an established and logically quite clear use of a general notion may yet lack justification and be the source of contradictions and confusions.

Kant, we remember, distinguishes between a posteriori concepts which are abstracted from and exemplified in sense- experience; a priori concepts which characterize the structure of the a priori particulars of space and time; Categories (and concepts deducible from them) which though not abstracted from sense-experience are yet through their schemata applicable to it; and lastly Ideas which ate a priori but incapable of being schematized and thus are inapplicable to what is given in experience. Now the notion of a purpose in nature is certainly neither an a posteriori concept nor a characteristic of the structure of space or time. Just as in a sequence of events we perceive one event and then another but do not also perceive the one as causing the other, so when the things or parts involved in a sequence exist for the purpose of each other we perceive only the things, not the purpose. In this respect the notions of causal and of teleological connexion are similar.

. There is, however, a very important point in which they differ. The Category of causal connexion can be schematized; the notion of a purpose in nature cannot. To schematize a Category is, we remember, to exhibit the temporal conditions of Its applicability - permanence in the case of Substance, existence at a certain time in the case of Reality, etc. The schema of Causality consists' in the succession of the manifold in so far as it is subject to a rule'.l Regular succession, in other words, is a perceptual feature in the absence of which there can be no causal connexion. Teleological connexion is not similarly linked to a temporal condition or, for that matter, to any other definite feature of pure or empirical perception. We cannot indicate any necessary perceptual condition for the application of 'purpose in nature'. The notion has no schema.

The a Priori notion of a purpose in nature must therefore be an Idea. Indeed in saying that a particular thing is an organism we are referring to a relationship between the parts and the whole which is to us an utter mystery. It is not causal and is 'strictly speaking not analogous to any causality which we know'. We can in a similar manner say that man qua noumenon is morally free; but the Idea of moral freedom must remain mysterious to us - even if we can prove that it is internally consistent and implied by the categorical imperative. Although Ideas are thus not applied to experience they may yet be usefully employed in our thought. This has been shown in the first Critique for the Ideas of Pure Reason and can also be shown for the Idea of purpose in nature.

The Categories of the Understanding are governed by synthetic a priori principles which are the conditions of objective empirical judgements and indeed of the experience of objects. Ideas are not in this. sense constitutive (i.e. constitutive of objects). They may, however, have a regulative function which is expressed not by synthetic a priori judgements but by maxims which 'are subjective principles based not on the characteristics of an object but on the 'interest of reason to confer a certain possible perfection on the knowledge of the objects '. The maxims which express the regulative function of the Ideas of Pure Reason are rules the observation of which enables us to progress towards an increasing unification of our empirical judgements into a system.'

 The Idea of a purpose in nature also can have a regulative function in so far as it is employed in accordance with' a remote analogy to our causality with respect to our own purposes [mit unserer Causalitat nach Zwecken]';  in other words, in so far as its use is remotely analogous to the use of the notion of a human purpose. The maxim governing the regulative use of the Idea is called by Kant' the maxim for judging the inner purposiveness of organized beings '. This maxim, which ac- cording to him is at the same time the definition of an organism, he formulates thus: 'An organized product of nature is that in which everything is reciprocally both means and end.'

From it there follow two corollaries which as methodological rules are, at least for heuristic reasons, observed by biologists, namely (a) the maxim that 'nothing in such a being exists in vain' and (b) the maxim that 'nothing happens merely by accident [von ungefahr, without purpose]'. Since maxims are rules of procedure it would be more precise to say that they enjoin us to proceed on the assumption that, or better still as if, nothing in an organism existed in vain or happened merely by accident. All this does not prejudge the possibility of explaining some parts or functions of an organism by merely mechanistic laws.

In judging things as organisms or purposes in nature we are transcending possible sense-experience: for such experience can be characterized only by means of a posteriori or schema- tiled a priori concepts. If we use teleological explanation in situations where the other is unavailable we can - limited by the same provisos - go further and judge even those products of nature' which do not require us to go beyond the mechanism of blindly efficient causes. . . as belonging to a system of purposes '. That is to say, if and in so far as a direct teleological explanation of natural phenomena is possible the superimposition of a teleological on a causal explanation is also possible.

Teleological Judgement thus enables us to unite the realm of nature and the realm of purposes into one system. It enables us to bridge the gap between the conception of man as a phenomenal and causally determined being and as a morally free agent. 'Neither of man nor indeed of any rational being in the world as a moral being can we [as we can in the case of other organisms] ... ask why (quem ad finem) he exists... Only in man as the subject of morality do we find unconditioned legislation with respect to purposes - a legislation which makes him capable of being an ultimate purpose to which the whole of nature is subordinated.’

Moreover, once ‘such a clue for the study of nature', namely teleological explanation, 'has been adopted and found useful [bewahrt] we must at least try to apply this maxim of Judgement in our reflection about nature as a whole. . .' We are then, whether philosophers or not, led to the assumption that the universe' depends on and has its source in an intelligent being. . . which exists outside the world: that teleology thus can find no completion of its inquiries except in a theology'.

Many important metaphysical systems based on various notions of natural purpose, final cause, or organism do indeed require that God's existence be assumed. This is so in the meta- physics of Aristotle and Aquinas and in Whitehead's' philosophy of organism'. As Whitehead points out,' it is an especially important fact in the history of metaphysics that ' Aristotle found it necessary to complete his metaphysics by the introduction of a Prime Mover - God'. It is important not only because he was a very great thinker but mainly because' in ~s consideration of this metaphysical question he was entirely dispassionate; and he is the last European metaphysician of first-rate importance for whom this claim can be made'.

Within the context of Aristotle's or Aquinas's thought the view is natural that we need teleology and that teleology leads to theology. It may seem strange if such a view is expressed by the author of the critical philosophy, 'the great destroyer in the realm of thought', as Heine calls him, the thinker 'whose terrorism by far exceeded that of Maximilien Robespierre'. The impression of strangeness arises, however, only if we  forget that the notion of purposes in nature is an Idea and that the principles governing its employment are only maxims of reflective Judgement and not empirical generalizations or synthetic a priori principles.

Kant himself sees to it that we do not forget this. 'What', be asks, 'can in the end be proved by even the most complete teleology? Does it perhaps prove that such an intelligent being exists? No, it proves no more than this, that - our cognitive faculties being what they are - we cannot at all form the conception of such a world [i.e. a teleologically organized one] unless we regard an intentionally acting being as its supreme cause.' Kant does not reinstate the older teleological meta- physics. He tries to show that its doctrines are not true of objective reality in the scientific sense but arise from certain modes of thought which are, so he believes, unavoidably adopted by man when he reflects about the phenomena of life and his own experience. If it be objected that even the application of the Categories does, according to the critical philosophy, remove us from the things in themselves, his answer would, I think, have to be that the employment of Ideas removes us still further from them.

3. Mechanism and Teleology

In his discussion of the relation between mechanistic and teleological explanation Kant remains faithful to his method of dealing with Ideas, a method which he has handled success- fully and elegantly in the other parts of the critical philosophy, ~specially in the Dialectic of Pure Reason. It consists in the formulation and resolution of an antinomy. Here it is the antinomy of teleological Judgement. Its thesis is the statement ' All production of material things is possible according to mere mechanistic laws'. Its antithesis says' some production of such things is not possible in accordance with merely mechanistic laws '. The two statements are straight contradictories and cannot therefore both be true.

The contradiction arises from careless formulation, or from neglecting the fundamental difference between statements of fact or possibility and maxims which are rules of procedure. As regards the latter even conflicting maxims can profitably be observed to a certain extent or 'up to a point' by the same person. If we remember this we must replace the statement of the thesis by the maxim ' All production of material things and their forms must be considered[ beurtheilt] as being possible in accordance with merely mechanistic laws '; and we must replace the statement of the antithesis by the maxim, 'Some products of material nature cannot be considered in accordance with merely mechanistic laws ( their consideration [Beurtheilung] requires anal- together different law of causality, namely, that of final causes). '

The contradiction has now vanished. The first maxim pre- scribes that I should reflect upon all events in material nature 'according to the principle of a mere mechanism of nature and consequently push my investigation with it as far as I can . . .' It does not prevent me, logically or in fact, from proceeding, when occasion arises, in accordance with the second maxim, the maxim of teleological explanation. Kant's argument here shows some similarity to a device which has forced itself on contemporary physicists who, without contradiction, consider the propagation of light in some contexts as wave motion, in others as the motion of particles.

Kant's resolution of the antinomy of reflective Judgement must be considered in the light of the first Critique. In that work, especially in the Analytic of Principles, he has expounded a system of theoretical a priori propositions which constitute the fundamenta1 conditions of Newtonian physics and, in his view, of all science. The result of the first Critique is thus, among other things, a mechanistic metaphysics; and nothing in the Critique of ]udgement indicates that Kant has in any way changed his view on this subject.  We must remember here that according to Kant all metaphysics which is not a mere illusion consists of propositions which are necessary conditions of objective experience.

To a mechanistic metaphysics there corresponds a mechanistic method of inquiry. If the phenomenal world' stands under' the synthetic a Priori principles of the Understanding which are constitutive of the empirical world; if, in other words, the empirical world is a mechanistic system, then it is reasonable to act as if it were such a system and to adopt the maxim of mechanistic inquiry.

The third Critique does not develop a teleological metaphysics. On the contrary, it shows that teleological principles are not constitutive of the empirical world, but can only be regulative for our reflection upon the empirical world. While the first Critique justifies the mechanistic method on the basis of a mechanistic metaphysic, the third Critique justifies the teleological method in spite of the impossibility of a teleological metaphysic. This impossibility is insisted upon time and again. Kant admits only a meta physic of nature and a metaphysic of morals. There is no metaphysic of purpose, but only a Critique of Teleological Judgement. He shows that there is no conflict between the maxims of mechanistic and teleological method. There can be no conflict between mechanistic and teleological metaphysics because, according to the critical philosophy, there can be no teleological metaphysics.

The value of the teleological method lies not only in its use outside scientific inquiry - for the unification of nature, morals, and indeed religious faith into one system. It has also an important heuristic use within the field of mechanistic and, per- haps more generally, non-teleological science. The question for what purpose a thing exists and the teleological answer to it may indeed suggest new mechanistic explanations. 'For where purposes are considered as the conditions of the possibility of certain things, means have to be assumed. . .' and it is, as Kant points out, quite possible that the relation between means and end may' considered by itself require nothing which involves a purpose'. It may indeed be both 'mechanistic' and yet a subordinate cause of intended effects '.

The question how far purely mechanistic explanation can go cannot be decided a priori, and Kant certainly does not think that it can. In our own day when the invention of electronic 'brains' and of self-regulating and self-adapting machines is by serious thinkers1 expected to bring about another industrial revolution, we are inclined to believe that mechanistic explanation can be pushed very much further than Kant considered possible. Biologists today would not, I am told, be too surprised if the growth of a blade of grass were explained on purely mechanistic principles - not perhaps by a biological Newton, but by a team of competent research workers. We must not forget that Kant lived before Darwin. Yet, in spite of the comparatively limited success of mechanistic explanation in his own days, he does not reject the main ideas of the theory later developed by Darwin. The passage concerned with these ideas is very often quoted. Nevertheless I venture to put it down again (in Meredith's translation): 'When we consider the agreement of so many genera of animals in a certain common schema, which apparently underlies not only the structure of their bones, but also the disposition of their remaining parts, and when we find here the wonderful simplicity of the original plan, which has been able to produce such an immense variety of species by the shortening of one member and the lengthening of another, by the involution of this part and the evolution of that, there gleams upon the mind a ray of hope, however faint, that the principle of the mechanism in nature, apart from which there can be no natural science at all, may yet enable us to arrive at some explanation in the case of organic life.'

I cannot forbear to quote in addition part of a footnote which makes Kant's biological views still clearer. ' An hypo- thesis of this kind may be called a daring adventure on the part of reason; and there are probably few among the most acute scientists to whose minds it has not sometimes occurred. For it cannot be said to be absurd, like the generatio aequivoca, which means the generation of an organized being from crude inorganic matter. . .' It would not be absurd to suppose that 'certain water animals transformed themselves by degrees into marsh-animals, and from these after some generations into land animals. In the judgement of plain reason there is nothing a priori self-contradictory in this. But experience offers no example of it...’

The following seems a fair summary of Kant's theory of teleological explanation: the third Critique implies, no less than the first, that all explanation must be mechanistic, but it supplements its theses. Even within the field of scientific inquiry mechanistic explanation has its limits, although one cannot prophesy at what point it will break down. Kant, quite naturally, believes that this will happen much earlier than contemporary biologists are inclined to assume. His conjectures, how- ever, that certain phenomena, e.g. the growth of organisms or the affinity between different species, are not susceptible to mechanistic explanation, do not form part of the critical philosophy. They are obiter dicta expressing his strong interest in the science of his day and his expectation of its progress. Teleological principles, unlike mechanistic, are not constitutive of objects, but merely regulative. Within scientific inquiry they suggest mechanistic questions and answers. Their main function, however, is the unification into one system of our thinking about matters of fact and of moral and religious experience. If Kant hankered after the old teleological metaphysics, his published works show no evidence of it. I do not think that he did.

Kant's philosophy of the biological sciences is, as far as I can judge, by no means outmoded. This is best shown by quoting from the writings of contemporary biologists whose general outlook differs from his. Thus Dr Needham sub- scribes to a view which he calls Neo-mechanism. According to him 'the neo-mechanistic position. . . at one and the same time asserting the universal dominion of the mechanical sort of explanation over all nature, living and non-living, and admitting the inadequate nature of this sort of explanation as a full account of the world, resembles the old mechanisticism in maintaining the heuristic need for the machine, and differs from it in seeing nothing solely ultimate about the machine. It thus recognizes itself to be the way the scientific mind goes to work, and not the manner of thinking in philosophy,

theology, or art . . . It knows teleology to be an unquantitative

category, and banishes it from the laboratory to the domain of the philosophers, who are quite capable of dealing with it . . . It may be noted that such a standpoint as Neo-mechanism will not necessarily object to special "biological" laws or explanations, provided that they are clearly understood to possess an " interim" character and to be only awaiting expression in physico-chemical terms.' Dr Needham finds that the neo- mechanistic position stands in close relationship with the Hegelian views put forward by R. G. Collingwood in his book Speculum Mentis; they are really found before either Collingwood or Hegel in the Critique of Judgement. There is only one major point of difference: the sceptical biologist is much less sceptical than Kant was about the' interim' character of all special biological laws.

Another contemporary biologist, von Bertalanffy, shows less sympathy for any kind of teleology; but his views about the limits of physico-chemical explanation are similar to those expressed in the third Critique. ' Just because we are championing exact, theoretical and quantitative biology we have', he says, 'to point out that what is expressed in the "exact" sciences as "laws" represents only a small section of reality. . . As mathematical biologists, we put the greatest emphasis on the obedience of organic forms to exact laws. . . But for precisely this reason we know only too well that it is only a small part of phenomena that can be understood in an "exact" way. Two skulls are distinguished not only by coarse differences of proportion which we can measure and calculate, but also by a wealth of characteristics that can only be described in verbal language, or even are noticed only by the morphologist's trained eye, but which he is hardly able to put into words.' The question of 'biological mechanism' relates only to 'those general traits of order which we are able to state in the form of "laws" '.

Contemporary philosophers are inclined to formulate the problem of the relation between mechanistic and teleological explanation as the question whether teleological statements can be 'reduced to' non-teleological ones. The answer, of course, depends largely on what is meant by such a 'reduction'. If the conditions for a successful reduction are sufficiently lax then almost any type of statement will be reducible to any other. If they are very strict then no type of statement will be reducible to any other. Teleological statements about human purposes and intentions do not, as we have seen, pre- sent any great difficulty since the intention or the making of a plan clearly precedes the activities which are explained as being caused by it.

R. B. Braithwaite and others have shown that in many con- texts goal-directed behaviour on the part of organisms can be explained without resort to teleology. Braithwaite argues that such behaviour can be explained in terms of plasticity which 'is a property of the organism with respect to a certain goal, namely that the organism can attain the same goal under different circumstances by alternative forms of activity making use frequently of different causal chains'. The 'reduction' of teleological statements about fina1 causes to plasticity-statements about alternative causal chains leading to the same effect is performed with admirable Clarity and in great detail. Kant's position is quite compatible with this and similar reductions - as long as it is understood that they are to be employed within the natural sciences.

In so far, however, as a philosopher must consider not only science, but other types of experience such as art and morals; and in so far as he must consider them together, teleological explanation is not' reducible' to mechanical. The inquiry into how we can for certain purposes and within a limited field of experience do without teleology must then give way to an inquiry how in our thinking we employ the notion of (non- human) purpose, and what kind of truth or objectivity, if any, we can achieve by doing so. The Critique of Teleological Judgement is such an inquiry;

4. Opinion, Knowledge, and Faith

The arduous task of the Critique of reason had been undertaken in the hope of finding some answer to the three great meta- physical problems of moral freedom, the immortality of the soul, and the existence of God. In the course of the inquiry many new and important problems had to be formulated and their solution attempted, and there is hardly a major philosophical issue which has been left without examination. The three great problems of metaphysics themselves appear in an entirely new light whatever we may think of the solution to them which Kant suggests in the Dialectic of Pure Reason and the second Critique. Towards the end of the Critique of Judgement Kant briefly considers them again by means of a distinction between three types of knowledge, in a wide sense of the term.

Kant distinguishes between matters of opinion, matters of fact, and matters of faith. Matters of opinion are' objects of an empirical knowledge which is at least in principle possible. . . [einer wenigstens an sich moglichen Erfahrungserkenntnis], but is [in fact] impossible for us because the degree to which we are capable of empirical cognition is not sufficiently high'. In order to illustrate this definition Kant gives two examples. His first example of a matter of opinion is the assumption that other planets are inhabited by rational beings. The reason why we cannot verify or falsify it lies in purely technical limitations which mayor may not one day be overcome. The difference between verifying whether the house next door, the north pole, or Mars is inhabited is one of degree.

Kant's second example is the assumption made by some physicists of his day (and, we may add, of the early twentieth century) that there is an ether, i.e. an elastic liquid which penetrates all substances. Although the existence of such a liquid cannot, and is not meant to be, verified or falsified by any experiment or observation, such a test would according to Kant be possible' if our external senses were sharpened to the highest degree',1 which, however, can never happen. The now obsolete concept of an ether belongs to the same class as the concepts' gene', 'atomic nucleus', and others which although unexemplifiable in sense-experience, are yet essential to certain empirical theories.

These' semi-empirical' concepts, as we might call them, cut across the Kantian classification of all general notions into a posteriori and a priori concepts and Ideas of reason. Kant's attempt to assimilate them to a posteriori concepts and to assimilate all statements about their imperceivable instances to matters of opinion does not seem convincing. Mter the fashion of some medieval logicians, Brentano and Russell regarded semi-empirical concepts, together with many other constituents of propositions, as syncategorematic terms (incomplete symbols). Such terms do not have instances, as have empirical concepts, but they may nevertheless contribute to the meaning of empirical propositions containing them. I cannot, for example, meaningfully say, 'This is an atomic nucleus', al- though the proposition 'The explosion of an atomic bomb is the result of changes involving atomic nuclei' has empirical meaning and although certain statements about the structure of atoms are matters of scientific opinion.

This is not the place to expound or to develop a theory of semi-empirical terms or, more generally, of incomplete symbols. But the absence of any such theory in the critical philosophy is certainly worth noting. There is no evidence that Kant ever even considered it, and it is idle to speculate how far it would have affected his system, in particular the Transcendental Analytic of Concepts. I do not, however, believe that it would have made the distinction which Kant draws between different types of concept superfluous.

Matters of fact are instances of concepts 'the objective reality of which can be demonstrated (be it through pure reason or experience) and in the former case either from theoretical or practical data of reason; in all cases, however, by means of a perception which corresponds to the concept'. To prove the reality of a concept or, what comes to the same, that an instance of it exists as a matter of fact, is thus (I) to make clear the logical content of the concept by means of empirical or a priori principles, (2) to indicate an instance of it in one's own or other people's experience. This wide sense of the term' matter of fact' which, as Kant himself admits, goes beyond its ordinary meaning, covers historical, mathematical, and scientific statements.

It also makes - what Kant considers 'very remarkable' - the statement that man is free a statement of fact. This is so because we can demonstrate the reality of the Idea of freedom (1) by means of practical a priori principles, and (2) by indicating actions, i.e. perceivable events happening in space and time, which correspond to it. But moral freedom is not a scientific fact, a matter of fact in the sense of theoretical reason. It is no more, but also no less, than a moral or practical fact in the sense explained by the second Critique.

Even Kant's wide definitions of 'matters of fact' and of the correlative 'reality of a ,concept , cover the Idea only of freedom, not any other Ideas. The objects which correspond to the others and which are not given in experience are matters of faith which' must be thought as related to the employment of pure practical reason in the service of duty' but which 'transcend the theoretical use of reason'. These Ideas are the highest good - the just proportion between virtue and happiness - and what Kant believes to be the only possible conditions of its achievement, namely the immortality of the soul and the existence of God. Faith is 'a confidence in the achievement of a purpose, the promotion of which is a duty; but we can have no insight into the manner in which this purpose could be realized. . .' Faith thus is neither a matter of knowledge, nor of opinion, but is 'wholly a matter of morality' - one of the things which Kant never tires of urging whenever occasion offers.

Aesthetic and teleological notions and judgements do not fall within the region of factual knowledge, of opinion, or of faith. They belong to man's creative powers, his capacity to become aware of beauty and to produce it, and his capacity to impose order and system upon his perception and thought. They are used in his attempts to reach the unknown and even the unknowable.

We have come to the end of this introduction to Kant's philosophy. I have tried to present its main theses and their connexion with each other and to point out their relevance to contemporary problems. One main aim of the undertaking has been to show that no one who takes these problems or indeed philosophy seriously can without much loss to himself ignore Kant's teaching. I should, therefore, like to conclude by expressing the hope that my exposition, brief and inadequate as in the nature of the case it has had to be, has not greatly misrepresented the thought of a very great thinker.