AN ENDLESS SERIES OF DEBATES

Popper’s Logic of Scientific Discovery comes to solve the problem of demarcation of science, and the problem of induction. Customarily, scientific theories are viewed as stable or near-stable-as true, as essentially true, or as highly probable; consequently, by the customary view scientific theories are less prone to upset and overthrow than pseudo-science and superstition.

Popper’s view, then, is that a theory is scientific if and only if it can be overthrown with the help of experience, and that we gain theoretical knowledge from experience when and only when such revolutions occur. This raises a few problems, chiefly the problems of the stability of science. What causes research to coordinate? What makes us accept empirical evidence? What causes unanimity in science and technology? What is the role of positive evidence in science and technology?

Popper’s theory of positive evidence is the better-known part of his teaching. People often identify it with his theory of learning from experi­ence: they take it for granted that he joins the multitude in viewing learn­ing from experience as finding empirical support to our theories.

Some people think Popper cannot seriously mean what he says here. Some think he is being perverse or masochistic. But what do these people say when they find Socrates make these same remarks, as for instance in. Gorgias and Greater Hippiasl Well, many scholars think that when speak­ing so Socrates is being ironical, by which they mean tongue-in-cheek and even sarcastic; others think that on these occasions he is merely being educational. But mPhaedo, which is likely to be a rather near-stenographic record, Socrates says: If all your life you have backed the wrong horse, you are better off switching horses in your last hour, than not. Dividing people into black and white is always silly, to be sure, but I confess to the folly of dividing people into those who see Socrates’ point and to those who shrug a shoulder when told of it.

Science, then, is to Popper a special case of Socratic dialogue, with experiment and observation offering new arguments, or new empirical criticisms, to use Faraday’s idiom.

I have put this point to Popper a few years ago. His present view is this.⁵ He says a good theory should be not only capable of being refuted, but also it should not be refuted too soon: we want positive evidence before we get negative evidence. We want positive evidence so as to be assured, he adds, that knowledge grows. So now, it seems, Popper has changed his view of science from a theory of conjectures and refutations to a theory of conjectures, corroborations, and refutations. In a footnote, however, he denies that any change has occurred when he says: “I feel greatly indebted to Dr. Agassi for drawing my attention to the fact that I have previously never explained clearly the distinction between” the desirability of having refutable theories and the desirability of their not being refuted in the very first test. So Popper says he has not changed his mind, merely explained clearly what he had explained earlier. It is not that Popper had not seen that the requirement for testability is different from the requirement for corroboration, of course; it is merely that he had overlooked, he says, the need to explain clearly the other requirement.

Here I disagree on general methodological grounds. Pure Popperism may be instrumentalism, or even worse; modified Popperism, whether Popperism plus the addendum that it is nice to have positive evidence from time to time, or Popperism with any other addendum, may be quite true for all we know. But the rules of the game forbid putting addenda without extensive study. This may be easily seen when we take other case-histories.

P. W. Bridgman was shocked by the Einsteinian revolution, and felt that the role of the philosopher is to insure the prevention of such mishaps in the future, by offering maximum stability to science. And he tried to reform science by introducing the maxim that all concepts of science should be operationally understood or else expurgated; e.g. ‘length’ should mean the measuring of length. Bridgman’s philosophy is admirable, not only in its immense tenacity, but also in its being an intended reform of science rather than the usual philosophical approval of the scientists. In his tenacity and reformism Bridgman did not shrink from rejecting general relativity because according to it the measuring of a length is not identical with the finding of that length, but, at best, with approaching that length asymptotically. Einstein himself was sympathetic to Bridg­man’s views, which in part he is the coauthor of, and he conceded that had operationalism worked in all other instances of science except general relativity, Bridgman would have had a somewhat significant point against it. But Bridgman’s original program does not work at all.

There is little doubt that though Bridgman’s philosophy had initially raised much opposition, practically all its original opponents dropped all their objections after Bridgman has inserted his correction. Indeed, they lost all interest in it. The reason for this is obvious: though the corrected version of Bridgman’s philosophy is true, it is no answer to Bridgman’s original quest. He had developed an austere program for science in order to keep it absolutely stable. He then relaxed his stringent requirements, and the desired stability is all gone. We are now where we were before Bridgman started, though made wiser by his failure. His correction of the failure, however, is hardly more than an admission of failure, though in a face-saving manner he has pretended to be merely modifying it.

Lest anyone think that failure is loss of face, that saving face is in any way useful, let me mention one great failure in the history of 20th- century philosophy, which is also one of the most monumental achieve­ments of the century. I am speaking of logicism, namely the view that mathematics is a part of logic. The idea is ancient, indeed a cornerstone of pre-Socratic philosophy. It became so important when modern logi­cians, Frege and Russell in particular, showed both how much of mathe­matics is a part of logic, and also, that not all of it is. I shall leave this point at that, referring the interested to the most critical and most appreci­ative recent paper by Alonzo Church on the subject.⁶ What I wish to say briefly here, is that Popper has a program akin to logicism and that he has deviated from it as Russell did, but not in the same manner, since Russell was aware of his failures.

Assume, then, that science is that kind of dialectic process. How much of science makes sense, then? This is a tricky question, since, if some of science makes no good sense according to this view, we may follow one of two opposite alternatives. The one alternative is to give up this view. The other is to reform science. Indeed, if Popper were not a reformer like Bridgman, but a justifier like Reichenbach, I doubt that he would be as interesting as he is. Yet there is a risk here of becoming a dogmatic Popperian and of always blaming science when it fails to stand up to the Popperian standards!

Indeed, this is true. Let us, then, explore this danger of dogmatism and see what it amounts to. Let us, in other words, first see how much of science is made sense of, and how much is left out, if we take Popper’s program in its rigorous version.

V.