David Deutch - The Fabric of Reality

CRYPTO-INDUCTIVIST: You say ‘one should ’ rely on the best-corroborated theory, but why, exactly? Presumably because, according to Popper, the process of corroboration has justified the theory, in the sense that its predictions are more likely to be true than the predictions of other theories.

DAVID: Well, not more likely than all other theories, because no doubt one day we’ll have even better theories of gravity …

CRYPTO-INDUCTIVIST: Now look. Please let’s agree not to trip each other up with quibbles that do not bear on the substance of what we are discussing. Of course there may be a better theory of gravity one day, but you have to decide whether to jump now, now. And given the evidence available to you now, you have chosen a certain theory to act upon. And you have chosen it according to Popperian criteria because you believe that those criteria are the ones most likely to select theories which make true predictions.

DAVID: Yes.

CRYPTO-INDUCTIVIST: So to summarize, you believe that the evidence currently available to you justifies the prediction that you would be killed if you leapt over the railing.

DAVID: No, it doesn’t.

CRYPTO-INDUCTIVIST: But dammit, you are contradicting yourself. Just now you said that that prediction is justified.

DAVID: It is justified. But it was not justified by the evidence, if by ‘the evidence’ you mean all the experiments whose outcomes the theory correctly predicted in the past. As we all know, that evidence is consistent with an infinity of theories, including theories predicting every logically possible outcome of my jumping over the railing.

CRYPTO-INDUCTIVIST: So in view of that, I repeat, the whole problem is to find what does justify the prediction. That is the problem of induction.

DAVID: Well, that is the problem that Popper solved.

CRYPTO-INDUCTIVIST: That’s news to me, and I’ve studied Popper extensively. But anyway, what is the solution? I’m eager to hear it. What justifies the prediction, if it isn’t the evidence?

DAVID: Argument.

CRYPTO-INDUCTIVIST: Argument?

DAVID: Only argument ever justifies anything — tentatively, of course. All theorizing is subject to error, and all that. But still, argument can sometimes justify theories. That is what argument is for.

CRYPTO-INDUCTIVIST: I think this is another of your quibbles. You can’t mean that the theory was justified by pure argument, like a mathematical theorem. [2]The evidence played some role, surely.

DAVID: Of course. This is an empirical theory, so, according to Popperian scientific methodology, crucial experiments play a pivotal role in deciding between it and its rivals. The rivals were refuted; it survived.

CRYPTO-INDUCTIVIST: And in consequence of that refuting and surviving, all of which happened in the past, the practical use of the theory to predict the future is now justified.

DAVID: I suppose so, though it seems misleading to say ‘in consequence of’ when we are not talking about a logical deduction.

CRYPTO-INDUCTIVIST: Well that’s the whole point again: what sort of consequence was it? Let me try to pin you down here. You admit that it was both argument and the outcomes of experiments that justified the theory. If the experiments had gone differently, the argument would have justified a different theory. So do you accept that in that sense — yes, via the argument, but I don’t want to keep repeating that proviso — the outcomes of past experiments did justify the prediction?

DAVID: Yes.

CRYPTO-INDUCTIVIST: So what exactly was it about those actual past outcomes that justified the prediction, as opposed to other possible past outcomes which might well have justified the contrary prediction?

DAVID: It was that the actual outcomes refuted all the rival theories, and corroborated the theory that now prevails.

CRYPTO-INDUCTIVIST: Good. Now listen carefully, because you have just said something which is not only provably untrue, but which you yourself conceded was untrue only moments ago. You say that the outcomes of experiments ‘refuted all the rival theories’. But you know very well that no set of outcomes of experiments can refute all possible rivals to a general theory. You said yourself that any set of past outcomes is (I quote) ‘consistent with an infinity of theories, including theories predicting every logically possible outcome of my jumping over the railing’. It follows inexorably that the prediction you favour was not justified by the experimental outcomes, because there are infinitely many other rivals to your theory, also unrefuted as yet, which make the opposite prediction.

DAVID: I’m glad I listened carefully, as you asked, for now I see that at least part of the difference between us has been caused by a misunderstanding over terminology. When Popper speaks of ‘rival theories’ to a given theory, he does not mean the set of all logically possible rivals: he means only the actual rivals, those proposed in the course of a rational controversy. (That includes theories ‘proposed’ purely mentally, by one person, in the course of a ‘controversy’ within one mind.)

CRYPTO-INDUCTIVIST: I see. Well, I’ll accept your terminology. But incidentally (I don’t think it matters, for present purposes, but I’m curious), isn’t it a strange assertion you are attributing to Popper, that the reliability of a theory depends on the accident of what other theories — false theories — people have proposed in the past, rather than just on the content of the theory in question, and on the experimental evidence?

DAVID: Not really. Even you inductivists speak of…

CRYPTO-INDUCTIVIST: I am not an inductivist!

DAVID: Yes you are.

CRYPTO-INDUCTIVIST: Hmph! Once again, I shall accept your terminology if you insist. But you may as well call me a porcupine. It really is perverse to call a person an ‘inductivist’ if that person’s whole thesis is that the invalidity of inductive reasoning presents us with an unsolved philosophical problem.

DAVID: I don’t think so. I think that that thesis is what defines, and always has defined, an inductivist. But I see that Popper has at least achieved one thing: ‘inductivist’ has become a term of abuse! Anyway, I was explaining why it’s not so strange that the reliability of a theory should depend on what false theories people have proposed in the past. Even inductivists speak of a theory being reliable or not, given certain ‘evidence’. Well, Popperians might speak of a theory being the best available for use in practice, given a certain problem-situation. And the most important features of a problem-situation are: what theories and explanations are in contention, what arguments have been advanced, and what theories have been refuted. ‘Corroboration’ is not just the confirmation of the winning theory. It requires the experimental refutation of rival theories. Confirming instances in themselves have no significance.

CRYPTO-INDUCTIVIST: Very interesting. I now understand the role of a theory’s refuted rivals in the justification of its predictions. Under inductivism, observation was supposed to be primary. One imagined a mass of past observations from which the theory was supposed to be induced, and observations also constituted the evidence which somehow justified the theory. In the Popperian picture of scientific progress, it is not observations but problems, controversies, theories and criticism that are primary. Experiments are designed and performed only to resolve controversies. Therefore only experimental results that actually do refute a theory — and not just any theory, it must have been a genuine contender in a rational controversy — constitute ‘corroboration’. And so it is only those experiments that provide evidence for the reliability of the winning theory.