A student who is studying Popper in London sent this paper by David Papineau for comment.
His immediate task is to write a paper on induction so I will not dwell on other matters except to suggest that Popper deserves some credit for writing one of the great books of political philosophy in the 20th century and also his “three world” theory is exciting and fertile.
It will help to get clear about the several things which travel under the label of induction.
1. Inductive discovery of regularities (sometimes unhelpfully called laws), like the sun rises every morning in the east.
2. The (subjective) belief that the sun will rise every morning in the east.
3. Inductive proof of the law or the belief, by observing the sun rising in the east.
4. The shift to probabilities after Hume’s critique has been taken on board and people accept that you can’t actually have inductive proof. So inductive logic becomes the matter of putting probability values on theories (not to be confused with the probability of events or the probability values assigned in statistical analysis to indicate the probability that the result could have come about by accident rather than a genuine causal relationship).
5. The assertion that induction means the belief that there are regularities in the world, so the future will be like the past, as long as the laws or regularities persist. This has got nothing to do with discovery, or testing, or probabilities and it tends to be the last resort of inductivists when the other forms of induction are criticized. It is better described as a metaphysical theory about the world.
After various critiques of the CERTAINTY of scientific knowledge and induction were accepted, the core of the program of inductivism became the quest for inductive probabilities. In recent decades it seems that Carnap’s quest for objective probabilities had been given up and the core these days is Bayesian subjectivism
TURNING TO THE PAPINEAU PAPER, WHAT KIND OF INDUCTION IS HE DEFENDING?
In paragraph 5 he wrote “Popper’s philosophy of science centres on his rejection of inductive reasoning. This is the kind of reasoning by which we judge that some hitherto observed pattern will continue to hold good in the future“.
That is induction of the fifth type identified above – the belief in regularities in the world. Did Popper reject the idea that there are regularities in the word? For him the theoretical or generalizing sciences were all about the quest for (true) laws of nature, that is regularities (or propensities) for systems to behave in particular ways.
That is a metaphysical idea about the way the world works in general, it has nothing to say about the methodology of science or epistemology (how we learn and test our ideas).
Moving on to para 7 we find explicit criticism of Popper’s strategy of conjectures and refutations because it “can only deliver negative knowledge. It shows certain scientific theories are false, but it never shows that any theory is true”.
BUT CAN THE INDUCTIVISTS DO BETTER?
WHERE IS THE INDUCTIVIST STRATEGY THAN CAN SHOW US THAT THEORIES ARE TRUE?
Papineau mocks Popper by producing two examples of theories – that cigarettes cause lung cancer and matter is made of atoms. What sort of theories are these? Not all smokers contract lung cancer, so clearly the proposition that cigarettes cause lung cancer is not a universal law. Similarly, atoms are very complex phenomena and they might well be explained in terms of force fields, not the little billiard balls of classical atomic theory.
The causes of lung cancer and the structure of matter are much more complex than the picture painted by Papineau, there are many conjectural elements and more work remains to be done. The theory of conjectural knowledge is not demolished by reference to cigarettes causing cancer and the atomic theory of matter.
Moving on to para 11 and 12 Papineau attacks Popper’s demarcation of scientific propositions on the basis of testability. He did not take account of the historical problem situation that created the issue of demarcation: that was the aim of the logical positivist to eliminate metaphysics using the demarcation principle of VERIFICATION and also the hallmark of science was THE METHOD OF INDUCTION, BASED ON EMPIRICAL DATA, EVIDENCE OR SENSE EXPERIENCE.
Popper argued that the principle of verification could not work (eventually that turned out to be undeniable despite major efforts by the positivists and logical empiricists to use it) and neither would inductive logic to either PROVE theories to be true or to make them PROBABLE (with a numerical p value).
Note that we are talking about general, explanatory theories that explain how things work, not just statements about things that exist like black swans or atoms.
Popper suggested that to should use evidence to test our theories (because we cannot verify them) and this proposal has two advantages over the positivists.
1. It makes us check to see whether evidence can be brought to bear in the argument that we are having. It may be that the theory at stake is in principle not testable, that does not mean that it is meaningless or trivial, it just means that criticism has to proceed using criteria other than evidence.
2. It liberated the positivists from the quest to eliminate metaphysics (by making it meaningless) by finding an empirical criterion of meaning. But they did not want to be liberated and so wasted some decades until nowadays metaphysics has made a comeback to the point where Popper’s theory of metaphysical research programs may be taken seriously.
Papineau went on to write (para 12) that Popper could not say that physics is less speculative than astrology because he cannot claim that atomic theory is firmly established by a large amount of evidence (as the inductivists claim). And so Popper “is stuck with the non-problem of explaining why some speculations are better than others”.
That is not difficult: we favour theories that are better tested, that explain more, that predict more accurately, that lead to fertile research programs.
TO BE CONTINUED