O'Hear wrote a book called Karl Popper (1980) in The Arguments of the Philosophers Series. He argued that we cannot do without both justification and induction. Since that time he has been immune to the counter-arguments of non-justificationists. The following extracts come from O'Hear's 1989 book on the philosophy of science that Papineau cited for follow-up reading as a rejoinder to the first two chapters of The Logic of Scientific Discovery.
"One of the things we want to establish in science is some idea of the basis on which we might act in the future. Popper is reluctant to see the manipulation or control of nature as a main aim of science…Nevertheless, he does admit that what he calls a ‘pragmatic’ preference for one theory over another is something we can legitimately derive from science; given that we have to act, we should act on the most rational assumptions that we can. There is no method more rational than the proposing of bold theories, and their severe testing."
"The obvious rejoinder to this is that a rational method for sifting theories with regard to their past performance is not at all the same thing as a rational method for sifting theories with regard to their future performance. For that we would need something like an inductive jump, from past to future. And it is just that which Popper believes we can do without in science." (my italics, RC)
It would help to say a conjectural jump. The term “inductive” in that context means something quite different from various other uses of the term, and it is amusing to find that the analytical philosophers, who are so concerned to be clear and precise about their terms, play ducks and ravens with terms when it comes to debating with Popper on induction.
Recapitulating the several meanings of induction – discovery by induction (repeated observations of sunrises); inductive proof (verification) which is possibly now defunct; inductive probability – a p value assigned to a scientific theory; the inductive leap to the future.
The last type of induction has nothing to do with discovery or logic, it is a metaphysical theory about the existence of regularities in the world, which we maybe able to discover or approximate if the system is not too big or too small or too complex.
So after the inductivist is confronted with the failure of inductive logic, then he or she takes refuge in a metaphysical theory about regularities in the world, and calls this induction.
What is the logic of induction that scientists use, or should be using?
It is not hard to see how they uses the deductive logic of modus tollens, to test the consequences of general theories.
It may need to be stated that the theory is interesting and worhy of testing to the extent that it provides explanations and insights into the world at large, beyond the point where the data are collected. It is also helpful to have rival theories, so the outcome of a test may shift a preference from one theory to another.
Consequently a practical decision can be derived from the best theory available, which means that it has explanatory power and has survived tests, or it is simply good enough as an instrument of calculation to give useful results even if we know or suspect that it is false (like Newton’s theory).
When the inductivists demand more than that, they demand something that they cannot provide themselves. What do they provide? Can some inductivist tell me what they have got that is better than well-tested theories?
We really should number the standard arguments on each side, then we don’t need to spend a lot of time repeating them, we can just recite the numbers and get on with something more worthwhile.
On the argument that philosophy is about wisdom rather than utility, you have to wonder about the wisdom of a program in logic that has been demonstrated to fail on logical grounds. And how many resources do you want to devote to a study which is supposed to be related to science but has no relevance to the activities of scientists?