A Short History of Modern Philosophy: From Descartes to Wittgenstein, Second Edition (22 page)

What does Kant mean in referring to his philosophy as a form of ‘idealism’ (albeit ‘transcendental’)? This is one of the most puzzling questions of Kantian exegesis, in particular since Kant expressly dismisses the philosophy of Berkeley (which he labels ‘empirical idealism’), asserts that ‘transcendental idealism’ is a form of ‘empirical realism’, and appends to the second edition of the
Critique
a chapter called ‘The Refutation of Idealism’. I shall return to this difficult question later. First, however, it is necessary to grasp Kant’s important distinction between ‘phenomenon’ and ‘noumenon’. Kant’s theory of the synthetic
a priori
depends crucially upon the element of empiricism in his philosophy—the view that knowledge comes through the synthesis of concept and ‘intuition’. We can have
a priori
knowledge of reality only as ‘phenomenon’—as a possible object of empirical observation. Phenomena are those things which can be
discovered
to be thus and thus; things, in other words, which enter into causal relation with ourselves and our experience. Philosophers like Leibniz had tried to describe reality as ‘noumenon’—as the object of pure intellectual apprehension. Kant’s theory of the synthetic
a priori
and his refutation of scepticism are meant to establish the reality of the phenomenal world (the ‘world of appearance’). To try to establish the reality of the noumenal world is to attempt to achieve knowledge by pure concepts alone; it is to attempt to transcend the limits of the human understanding and so achieve knowledge of a world that could never be empirically discovered. Such an attempt involves the transformation of understanding into ‘pure reason’, and Kant regarded it as doomed to failure. Part of the meaning of the phrase ‘transcendental idealism’ is contained, then, in this robust emphasis on the empirical as the legitimate sphere of knowledge, and on the impossibility of knowing a ‘noumenon’ or ‘thing-in-itself.

But does Kant’s ‘transcendental idealism’ really contain a refutation of scepticism? There is a systematic ambiguity in the actual theory of transcendental idealism which makes it difficult to answer this question. The ambiguity is contained in the phrase, used in the last paragraph: ‘the world of appearance’. I have taken Kant’s claim that transcendental idealism is a form of realism seriously. I have assumed that he intends to assert that the world of which we have knowledge really
exists independently.
The world is a ‘world of appearance’ only in the sense that it exists in time, consisting of objects and processes which are either perceived by us, or else causally related to our perception. One might call this the ‘objective’ interpretation of Kant’s theory. It is an interpretation that makes the theory incompatible with Humean scepticism.

However, there is a rival interpretation, which we might call ‘subjective’. Until recently this was far more widely accepted, despite being compatible at least with the intentions underlying the Humean point of view. This rival interpretation emphasises not ‘empirical
realism
’ but ‘transcendental
idealism’.
It interprets the Transcendental Deduction as expounding a thesis about the nature of the human mind. It is our finite capacities that are being described, and the attack on empiricism is directed, not against scepticism, but against the impoverished concept of human mentality (and in particular the untenable concept of ‘experience’) from which empiricism departs. When Kant says that we have knowledge not of the world ‘as it is in itself’, but only of the world as it appears (‘the world of appearance’), this could be read as a complicated way of agreeing with the empiricist’s conclusions. The world of appearance marks a limit which we cannot in the nature of things transcend. Knowledge is described in subjective terms—as something generated by the understanding, through the synthesis of concept and intuition. In no sense does it, on this interpretation, reach beyond that synthesis to an independent world (the world of the ‘thing-in-itself’). This rival psychologistic interpretation of Kant can find support in the text, and has been profoundly influential. In retrospect, however, it seems to me that only the objective interpretation of the first
Critique
allows us to think of Kant’s enterprise as either worthwhile or significant.

In the second part of the
Critique of Pure Reason
Kant diagnoses the failure of ‘pure reason’, trying to show that the attempt to employ concepts outside the limits prescribed by their empirical application leads inevitably to fallacies—in the form of paradoxes, incoherencies and direct contradictions. The inevitable tendency of reason to transcend the limits of intelligibility Kant called the ‘Dialectic’ of reason (and this concept was to have a profound influence—through Hegel—on subsequent philosophy, although an influence that was at variance with Kant’s intentions). Kant tried to show that all the traditional metaphysical (specifically rationalist) arguments—arguments about the substantiality and immortality of the soul, the infinitude of the universe, the necessary existence of God and the reality of free will—were inevitably grounded in contradiction and paradox. The brilliance of his exposition of these errors, together with the fascination of his diagnosis of them, not as accidental but as inevitable diseases of the understanding, are unsurpassed in the history of philosophy. There is space, however, to mention only the most important of Kant’s conclusions, those concerning the soul and God. The account of free will must await the chapter which follows.

Kant’s view of the soul is extremely subtle. He begins once again from the notion of ‘apperception’. It is clear that there is no thought without a subject. And because that subject must have privileged access to its present mental states it is tempting to think, with Descartes, that its pure ‘subjectivity’ provides some indication of its essential nature; that the abolition of the distinction between being and seeming licenses the inference to the conclusion that the self has a purely ‘subjective’ being; hence that the self is not subject to the laws of objects, being indestructible and indivisible. Kant points out that there is a fallacy in this inference. There is no passage from the privilege of self-knowledge to the essence of what is known. The privilege of the first person presupposes the existence of the self as object; it is therefore not for selfknowledge to determine what it knows. The essence of the self remains hidden, even though its accidents are immediately ‘given’ to consciousness. Kant goes on to connect this view with a theory of practical knowledge, and of the moral being of the self, that I shall elaborate in the chapter that follows.

Just as the understanding has its categories, so does ‘pure reason’ have its ‘ideas’. These are categories that have outrun, as it were, the possibility of cognitive application—permanent delusions of the understanding, which one is constrained always to pursue but never to grasp. Among these ideas is that of infinity, construed not as indefiniteness (as in the perpetual incompletion of a mathematical series) but as a completed infinity (as in the Platonic and Boethian view of time). The principal and most compelling instance of that idea is God, and it is to the refutation of the traditional arguments for God’s existence that Kant turned his attention in much of the Dialectic. In particular, he presented a famous refutation of the ontological argument, a refutation which a great many have chosen to regard as conclusive, and also as damaging to the whole enterprise of rational theology. The argument turns on the premise that existence is not a predicate; it is therefore impossible, Kant argued, to advance from the concept of God to the existence of God. No concept can imply its own instantiation, and the logical character of existence is misrepresented by any attempt to make it part of the concept of a thing. Kant’s premise contained a premonition of one of the most important results of modern logic. This result was to change the course of philosophy once again.

Kant’s dismissal of the claims of ‘pure reason’ was subject to certain important reservations. For one thing, he regarded the ‘ideas’ of reason as having an important ‘regulative’ function. If regarded, not as autonomous instruments of knowledge, but as signposts for the understanding, their theoretical employment would lead not to error, but to the constant stimulation of fresh discovery. There was a more important use of ‘pure reason’, however, adumbrated already in parts of the Dialectic, but fully elaborated only in the
Critique of Practical Reason.
Reason finds its legitimate employment in the practical sphere, and we can understand the claims of theology, for example, if we see them, not as intellectual truths which could be stated and argued for, but, so to speak, as ‘intimations’, made manifest to our consciousness when we act in obedience to the moral law.

The 
Critique of Pure Reason
set out to curb the pretensions of speculative metaphysics while establishing
a priori
those principles which must be assumed if there is to be knowledge of an objective order. These principles enable the fundamental distinction between appearance and reality to be drawn with system and authority. The same concern for objectivity can be seen in Kant’s writings on ethics and aesthetics, both of which subjects he transformed entirely. There are two
Critiques
(1788 and 1790) which deal with these branches of philosophy, together with an earlier and in many ways more challenging work, the
Foundations of the Metaphysics of Morals
(1785). These works develop systems of value which not only purport to explore in a definitive way the entire question of the objectivity of moral and aesthetic judgement, but also to bring to completion the metaphysical speculations begun in the first
Critique.
Kant tries to rehabilitate, through the theory of ‘practical reason’, some crucial metaphysical dogmas which theoretical reason alone is unable to establish.

In his ethics and aesthetics Kant was less concerned with the demolition of speculative pretensions and more concerned with providing positive support for evaluative judgements. He wished to justify fundamental items of belief, and to provide the underpinning of thoughts which seem both vulnerable to philosophical scepticism and at the same time basic to our conception of ourselves. Once again Kant considered himself to be responding to the challenge of Hume’s scepticism, in an area where— because moral and aesthetic principles provide obstacles both to the fulfilment of natural inclination and to the exercise of choice—there is a universal motive to welcome scepticism. Moreover, this motive seems well founded. For what else can moral and aesthetic principles amount to, if not the expressions of individual preferences, powerful perhaps in their sovereignty over the mind which conceives them, but unwarranted by any independent order? They seem to be supported, if at all, by sanctions which are as arbitrary as the laws which they uphold.

Hume formulated the fundamental premises of this scepticism in two succinct but complex thoughts. First, there is no derivation of an ‘ought’ from an ‘is’, which is to say that moral judgements, since they do not describe how things are, gain no justification from natural science. Secondly, since the sole motive to action is desire, and reason itself is no motive, the only rational justification that can be offered for any action lies in showing that it contributes to the satisfaction of an agent’s wishes. All reasoning is reasoning about means, and has no authority beyond that of the desire which compels it. There is no innate power of reason to overrule desire, and hence no power of reason to determine action objectively.

The first great insight contained in Kant’s moral philosophy was the realisation that Hume’s first source of scepticism was of no real significance. Suppose that the ‘is—ought’ problem were solved, so that moral judgements could be determined with the objectivity of a natural science. That would not refute scepticism. For to refute scepticism we must also show how such judgements provide reasons for acting. In other words, Hume’s second objection would still be sufficient to refute the objectivity of morals. On the other hand, if we can show that there are objective reasons for acting, then the ‘is—ought’ problem becomes insignificant. It no longer matters that we can or cannot derive an ‘ought’ from an ‘is’, for morality will gain its rational basis independently. It seemed to Kant therefore that the ancient distinction between theoretical and practical reason should be revived, and the foundations of the second explored with the same discipline that he had devoted to the exposition of the first. This reintroduction and elaboration of the concept of practical reason was among one of the most influential of Kant’s achievements and provided the grounds not only for his own partial repudiation of the metaphysics of the first
Critique
but also for many of the insights of later German idealists.

Theoretical reason guides belief, and practical reason guides action; the first therefore aims at truth, the second at rightness. The first, when employed legitimately, Kant called understanding; when illegitimately, pure reason. Reason can, however, also be employed legitimately, but it must then be subjected to those determinations which transform it into practical reason. Understanding issues in judgements (intellectual acts which might be true or false); practical reason issues in imperatives, which may be acted on, but which cannot be called true. (Hence— though Kant does not derive this conclusion—there is no
logical
argument from an ‘is’ to an ‘ought’, no argument which proceeds by the use of principles governing truth alone.) Practical reason consists, therefore, in the justification of imperatives, and the problem is to define and validate a concept of objectivity which will both apply to such imperatives and generate a recognisable system of morality.

The categorical imperative

Imperatives, Kant noticed, are of two kinds, the hypothetical and the categorical. The first kind are distinguished by the presence of a conditional antecedent, an ‘if...’, which makes reference to some condition of need or desire. ‘If you want a drink, then go into the drawing-room’. The consequent of such an imperative states (if the whole is valid) an adequate means to the satisfaction of the want or desire mentioned in the antecedent. Such imperatives can be justified objectively, without assuming any special function of practical reason. It suffices to show that, as a matter of fact, the means referred to are adequate to the end supposed. But in an important sense hypothetical imperatives neither have nor claim objectivity: for they provide reasons for action only to people who have the desire mentioned in their antecedent. Their weight, or motivating force, depends upon the actual desires of the subject to whom they are addressed, and derives purely from the motivating force of those desires. According to Hume, there is no other practical employment of reason than in the generation of imperatives of this kind, that is, in a specific and limited application of theoretical reason to the calculation of the means to our ends.