WHITEHEAD’S SCANDAL
What is common to Kant and Whitehead is the deep feeling that some proof urgently needs to be given, can easily be produced, yet is not yet available.
The faint hint is that philosophers are lazy - scandalously lazy.
This is why Kant’s scandal is no scandal at all - it took so much to overthrow Berkeley’s idealism (assuming this, at least), that we cannot honestly blame Kant’s contemporaries or successors for not having succeeded to accomplish this task: the task was not as easy as they thought. With Whitehead, the feeling that the problem of induction is urgent and must be soluble of course, is what led him to echo Kant. Is this really so bad? Whitehead himself was not a man of poor intellectual abilities; why did he fail to solve the problem?Wherever there is a task which seems at once both so easy and so difficult, it may be advisable to explain the conflicting appearances first. For, as long as we are not clear whether the task seems easy or seems difficult, we are not clear even about its most superficial aspects; we do not have a clear expectation; we are, in fact, confused. We may confuse two tasks, one (seemingly) easy, one not. Or we may confuse two different settings, one in which the task is (seemingly) easy, one not. Or we may confuse two different criteria of satisfactory executions of the task, by one criterion the task is (seemingly) easy, by one not. Can this apply to induction?
It seems to me clearly to be the case. I shall not enumerate all the possible distinctions, not even all those which are exemplified in history. I shall not even mention all important distinctions along the above lines; I shall only mention one. There are two versions of inductivism, of the views of what induction is or should be. The one is a radicalist inductivism, and the other is tempered inductivism - one which recognizes its own limitations.
The radicalist version is one which denies all knowledge, and all foundation of knowledge, except the analytic a priori and the synthetic a posteriori. As Hume and Kant have proven, this version is plainly inconsistent as it rests on a principle of induction which is neither analytic a priori nor synthetic a posteriori. And so the principle of induction must count as meaningless or as transcendentally proven by the very success of science and hence count as synthetic a priori. Indeed, Wittgenstein in his radicalist Tractatus declared the very principle of induction meaningless, and Russell in his ‘Limits of Empiricism’ accepted a tempered version of inductivism by declaring the principle of induction synthetic a priori,(Here we see the force of Berkeley’s argument from radicalist inductivism. We may accept the distinction between primary and secondary qualities on some a priori grounds; this will deprive us of claims for radicalist inductivism. It is only radicalist inductivism, not necessarily any tempered one, which imposes idealism on us.)
Some transcendentalists leave no room at all for induction, some trans- cendentalists leave everything to induction except the principle of induction. The middle ground between these two extremes is a vast terra incognita. Mill assumed, a priori, a principle of simplicity of nature which Keynes stated, at least with greater precision, as the principle of limited variation. The principle cannot work because we do not know what variant is excluded from, what variant is included in, the small group of those laws which nature is permitted to adopt. But once we make the principle more specific, then hey, presto! induction becomes no problem at all. This idea, incidentally, has been discovered by Descartes, and his Principia contains an explicit principle of induction which he (quite rightly) treats as not problematic in the least.
A clear example for all this is organic chemistry or molecular biology: given a few well-known general laws and a list of atoms which a molecule may contain, and the search for the composition and even structure of any given molecule, however big or complex, is merely a question of time.
Within the accepted framework there is a finite number of given possible alternatives; we may assign each an a priori and an a posteriori probability; we can verify one of them; we can render one of them probable - especially by induction by elimination. Induction is simply no problem at all for philosophy once the setting is sufficiently, clearly, specified.And so, quite possibly, Whitehead felt the scandal because he felt that the problem was on the one hand easily soluble, on the other insoluble; yet he failed to distinguish the different settings which rightly make the problem very easy and very difficult respectively. If this is so, then the chief question is, can we solve the easy problem and ignore the difficult one? I think this problem of strategy is very alive in the literature though it is, as usual, poorly articulated. What kind of induction, we may ask, is science engaged in, and can philosophers confine their discussion to that induction which in actual fact is employed by active science? Come to think of it, do scientists, in actual fact, employ any rules of induction? Most philosophers say, but of course. Not all, however.
III.