THE FACTS ABOUT INDUCTION

There is no induction, says Popper; and he explains why. But very few hear him. The few who do notice Popper’s philosophy in their writings about induction, notice him not because he says, there is no induction, but because he speaks of confirmation or corroboration in a manner which they find more interesting.

This has made me stop and think. What is it that these philosophers think so obvious that they are stunned by Popper’s oversight? How come both Bar-Hillel and Isaac Levi say that it is not possible for the intelligent person that Popper is to say what he says and mean it literally? What is so obvious, they all say explicitly (and I do them the courtesy of taking them to mean what they say), is that science and technology obviously exist and flourish.

Arguments from the existence of science are transcendental arguments: if my theory were false then science would be impossible; but science exists. This is the logical form of Kant’s argument which he christened ‘transcendental’ and this is the logical form of Russell’s argument, and Sir Harold Jeffreys, and of lesser lights. The way to avoid transcenden­talism is to take the existence and success of science and technology as a datum to be explained.

We may then read the complaint against Popper as the claim that he ignores the chief datum we have. To some extent I do agree. I do declare that the success of science and of technology are empirical data that many speak of as if they have been empirically observed, as if they are so obvious that no one can fail to observe them. I do declare that these em­pirical data - the success of science and technology - are begging to be observed, with some attention to detail.

What they suggest, I think, is first and foremost that we cannot view the success of science and technology, given its range and duration, as a mere accident. If no accident, then it follows certain rules. Call these rules induction and there you are. Certainly, Popper’s is not allowed, by his own dicta, to oppose this meaning or that. So much for the current criti­cism of Popper’s claim that induction simply does not exist.

I think I have this right. If not, I too, like Popper’s, fail to see the ob­vious; I have a blind spot. I was willing to consider this seriously, but it is a depressing thought and a blind alley. As Bacon has said, as even Witt­genstein has said, no one is a good judge of his own blind spots. And so I shall assume, until corrected, that my presentation of the criticism of Popper’s denial of the existence of induction is right. Let us return to the chief datum, scientific and technological success, and its explanation: the success is due not to mere luck but to some method. Ah! says Popper’s. You believe in the sausage-machine of science, in the view that there is a sure science-making algorithm. You all do! Even great scientists often say, even Robert Oppenheimer said, give us more money and we shall buy better people and equipment and produce more and better science. But science is no sausage-making product; there is no science-producing algorithm; hence there is no guarantee that more money - or anything else - will bring about more or better science.

This, I think, is the crux of the disagreement. Popper’s critics are not very articulate, and so this paragraph is my attempt at an articulation of their views and criticisms.

And so there is some method, and it leads to more than some success with more than occasional regularity. Such is, still, our datum. And so the philosopher is frustrated in his attempt to locate this obvious datum. He may easily identify his some-induction with what I have earlier called tempered inductivism. Since tempered inductivism is not very problematic, some- induction may, perhaps, be easily vindicated. Somehow this sounds otiose.

We are closing in on the scandal in philosophy - in the sense of frustra­tion, of course. We still have as a datum some-induction. Some successful method. To ground this success in some synthetic a priori principle is to beg the question, to shift the explanation in a most frustrating manner.

The old view of science as a sausage-machine, as a total algorithm, is erroneous but very interesting and even a very cogent solution to certain problems as they appear historically. The partial view is problematic. The view that science and technology are totally algorithmic whereas the fine arts are totally spontaneous, for example, is doubly false but very clear and satisfactory. Now that we know that both scientific and artistic activity is partly routine partly inspired we may ask, how are they to be demarcated?

I have not seen all this ever expressed like this, yet I do contend that this is at the back of many writers’ minds. When they say that science is predictive, they speak of the material success of science and thus over­stress the importance of the problem of induction. But they are not there­by vulgar materialists; they clumsily try to demarcate science by its suc­cess. If so, their ill success can be removed by regaining balance and re­membering some failures too.

IV.