Gordon Clark: Questions on Kant (The Review of Metaphysics)

1952. Questions on Kant. The Review of Metaphysics. Vol V, No. 3, March.

QUESTIONS ON KANT

In the Review of Metaphysics for June 1951*, Professor George A. Schrader has given us an exquisite article on Kant. The construction of the thought is delightful to follow, and the content is instructive and stimulating. The purpose of this note is not to refute any of the six major contentions listed summarily at the end of Professor Schrader’s article, though perhaps slight modifications of the sixth conclusion might be implied; rather the purpose is to question some of the minor points of his argument and to add some minor reflections that may clarify the total picture to a small degree.

On page 520 Schrader says, “So far Kant has established the universality but not the necessity of space and time.” The context of this sentence does not seem to alter its sense as quoted in isolation, but for reasons I shall try to develop, it seems to me to be a false statement. Surely Schrader’s preceding paragraph does not offer an argument sufficient to establish conclusively the universality of space and time. However, on page 521 Schrader writes, “Kant’s arguments in the Aesthetic do not establish that space and time are necessary or even that they are universal for sensible intuition.” Just how this latter sentence modifies the former is not the point of interest. Even if the first statement makes a broader impression than Schrader intended, still he goes on to say that Kant did not rest his own case on examples of sense perception in which the awareness of space and time is involved, but in ruling out any possible intuitions which are not spatio-temporal made a far stronger contention which is “in no way based upon inductive argument.”

Now, Schrader not only repudiates the psychological interpretation of the forms of intuition, but suggests a realistic rather than an idealistic interpretation of Kant. This interpretation merits careful consideration. However, if Kant is to be taken as a realist, then, I wish to submit, his arguments must be based on induction. And if Kant is not to be taken as a realist, still his arguments, as written in the Critique, are as a matter of fact inductive. The following four reasons, I believe, at least point toward such a conclusion.

First: Schrader rightly insists that Kant’s theory of space and time is not completely put in the Aesthetic. One of the chief values of his article is to impress us with the bearing of widely separated passages on the theory of space and time. For example, though Schrader does not explicitly cite this page, Kant, in the introductory paragraph to the System of All Principles of the Pure Understanding, B 188-189, professes to have done with intuition, even though the argument will make the principles of mathematics and intuition conceivable and will deduce the possibility of such cognitions. And surely, even if we disregard the several hints on the intervening pages, the Axiom of Intuition, B 202-203, which Schrader cites, explicitly concerns intuition. If Kant wants to show that all intuitions are extensive quantities, he perforce discusses space. Adickes notes that the material belongs in the Aesthetic, paragraph 3, i.e., B 40. But unless it be taken as unsupported assertion the material constitutes an inductive argument.

Second: Schrader distinguishes, in the first sentence quoted, between universality and necessity. He then adds, p. 520, “Perhaps one must know [italics his] that something is necessary in order to know that it is universal, but a thing may be universal without being necessary.” Aside from the fact that Kant denies that a thing may be universal without being necessary – see point three below – the main theme raises problems of infinite complexity. If Kant is a realist, at least if he is more of an empiricist than a rationalist, would he not also be a nominalist? But on nominalistic principles, contrary to Schrader’s assertion, one would have to know that something was universal before knowing it was necessary. And on realistic, empirical principles experience could never arrive at universality. Hence Kant’s arguments must be inductive, and their conclusions no more than a posteriori probabilities.

But perhaps one should not depend too greatly on the thesis that Kant was a realist. Let judgment be suspended on this point. Still, in the third place, the idea of necessity in Kant is a stumbling block. Suppose it is true, with Schrader, that one must know necessity before universality. What is necessity and how is it known? In the Einleitung, B 3ff, necessity is mentioned as an infallible criterion of the a priori. The knowledge of necessity therefore must precede knowledge of the a priori. One might wonder whether there would be fewer difficulties in a theory that made knowledge of the a priori the condition of knowledge of the necessary. However that may be, if necessity is the criterion of the a priori, one must know the meaning of necessity. Now, much later, in the analysis of the arguments for God’s existence, Kant has some thing to say about necessity. All examples of necessity, he reminds us, are drawn from judgments, not from things. And judgments are necessary when the denial of the predicate forms a contradiction with the subject. But this type of necessity, the ordinary logical necessity of an analytic judgment, is insufficient for Kant’s purposes in two respects. The a priori synthetic judgments are necessary, but their necessity cannot be of this analytic type. And also, if we have no idea of what necessity might mean in the case of God, have we any idea of how space, which is also individual, can be necessary? The possibility of a priori synthetic judgments, Schrader insists, depends on the theory of space; but the theory of space requires a type of necessity which Kant either repudiates when discus sing the ontological argument or at best leaves entirely unexplained.

Underlying this difficulty there seems to me to be a shifting of Kant’s thought between an intensional and an extensional logic. As an empirical opponent of dogmatic rationalism, his universality and necessity, which, contrary to Schrader’s statement above, “gehoren auch unzertrennlich zu einander,” B 4, should be the probable results of induction; but as an apriorist he needs a more necessary necessity.

In the fourth place, and connected with the difficulties in logic, Kant makes space a pure intuition. He never tires of repudiating intellectual intuition. Intellectual intuition might have given him the necessity and the intensional logic he needed; but in its place he puts the pure sensory intuition of space and time. Schrader, page 528, asserts, “I have suggested that space and time are given in intuition and I mean to insist that they are, in this regard, on the same level empirically with intuited data.” The most apparent sense of these some what ambiguous words is difficult to accept, for no one has seen space as he has seen color. Space and color are not, obviously, on the same level in every regard. They are not on the same level with regard to being given in intuition. The color seen is seen in its totality,”but the infinity of that individual which is space forever prevents it from being thus given in intuition. If space were given in intuition on the same empirical level with color, then surely Kant’s four points in the Aesthetic, B 38-40, would have to be altered beyond recognition. At any rate perhaps it will be granted that Kant’s theory of pure sensory intuition is not easy to conceive. In fact, could we not say that a pure sensory intuition is more vague than an intellectual intuition is doubtful?

In conclusion, the real trouble in expounding Kant is that he is hopelessly self-contradictory. Schrader, to maintain some of his assertions, is quite aware of the fact that he must discount or delete certain portions of Kant. In fact Schrader seems to rely more on the earlier than on the later sections. Kant’s views were developing as he wrote; the final introductory paragraphs by which he pieced together his earlier papers do not always, in fact usually do not, present the same theory as the sections they introduce. The earlier sections are richly suggestive, for example the Transcendental Deduction; but we are not justified in pointing to such sections as definitive Kantianism. Are they indeed anything more than a passing phase of a mind that never came to rest?

Gordon H. Clark

Butler University

* George A. Schrader, “The Transcendental Ideality and Empirical Reality of Kant’s Space and Time,” The Review of Metaphysics, IV (1951), 507-36.