I am an engineer and know nothing about philosophy, my friend is an economist who argues that the problem with induction invalidates all predictions based on induction.

I can understand this on a fundamental level but does this invalidate all predictions and models made from science? He argues that the scientific method is not truly deductive and so invalid. It seems to me that we have gotten pretty far with these predictions so does it matter if they are invalid?

  • 1
    The Problem of Induction is a very interesting philosophical problem. About its "real life impact" we have to consider its strong relationship with cause and effect: if every time the lion see the footprint of a gazelle he starts to ponder about the "realiability" of the inference: if footprint, then gazelle... well, lions would be all extinct since millenia. Dec 14, 2017 at 16:00
  • Your friend seems to confuse justification with formal validity. We do not have a formally valid proof that the Sun will rise tomorrow, but we are very justified to act on this induction. Pretty much any substantive argument involving practical matters is formally invalid, science is not an exception. But, as Toulmin put it, this is a strike against formal validity, not against substantive arguments, see argumentation theory. Real world can not be reasoned about by the standards of mathematics.
    – Conifold
    Dec 14, 2017 at 21:35

3 Answers 3


Science does not work directly from induction, it requires a theory in addition to observations. That provides a mechanism that forces statistics to apply. As you repeat an experiment, the odds it will ultimately fail to repeat in the future do go down. They never get to zero. You could still be wrong. But that becomes less and less likely as you successfully reuse the theory over time.

Even if you don't do the p-value computations, you are safely covered by the Rule of Succession, as long as you have a falsifiable theory, and you challenge it. You have some odds p < 1 of being wrong, and if you successfully repeatedly use your result n times the odds it will eventually fail are <= p^n, which converges to zero when n is infinite. Since 'being falsifiable' implies you would recognize when you failed, you are legitimately sampling a distribution, and this is not induction, it is math.

Each time you 'fail to falsify' you 'reject the null hypothesis' and in doing so you are building a statistical basis for believing your model and its underlying assumptions are less and less likely to be wrong to a significant degree. If you do keep actually compute the p-values, as people in very slippery sciences like psychology and sociology force themselves to do, you can go back and figure out just how unlikely ultimate failure is becoming due to seeing the outcome in various instantiations.

Statistical convergence is not really induction in the pure philosophical sense, it is deduction with probabilistic truth values, that never really reaches 100% reliability.

So he is 'right' that science is never logically watertight. And we already knew that, given that it constantly changes and evolves. But he is wrong to say the procedure is not deductive.

  • Built into the process, however is the assumption that since a problem has been well modeled statistically in the past, it will be well modeled statistically in the future.
    – Cort Ammon
    Apr 6, 2018 at 21:50
  • @CortAmmon No, always tossing the losers still predicts convergence, even if your models are really bad. Ptolemaic astronomy worked, and Alchemy did discover random things, even if their reasoning was unrelated to reality. You can misidentify losers, as Lakatos points out, but it should not be very common in any experimental system. The point is that experimentation just is not induction.
    – user9166
    Apr 6, 2018 at 22:22

...does [the problem of induction] invalidate all predictions and models made from science?

Yes. The problem says that “we are not entitled to any degree of confidence whatever, no matter how slight, in any predictions regarding what we have not observed” Lange (2011), "Hume and the Problem of Induction", p. 43.

The sticking point is the problem's underlying premise: there is no reason to believe that similar events will have similar results; instead, from any event, any result is conceivable. That premise, of course, destroys predictability.

...so does it matter if [the predictions] are invalid?

No. Induction is famously the "glory of Science" and the "scandal of Philosophy" (C.D. Broad 1926). Science has assumed, successfully, that similar events will indeed have similar results. The predictions are not invalid in the sense that they are wrong; it is just that the philosophical justification is not present. As a practical matter, the predictions are reliable.


The Problem of Induction does represent a problem for the philosophy that is often referred to as science. That empirical reasoning is not "proven" may or may not be a problem for you personally. Empiricism has lead to knowledge that is closer to true then what we had before. So empiricism is useful as the computer you are using to view this demonstrates.
Is our current understanding of empiricism "true"? Almost certainly not. That is part of the philosophy of the sciences. The assumption that our current understanding is flawed in some way. This encourages people to suggest improvements. As a result our understanding improves over time.
So no I do not have an issue with the problem of induction, it only represent another area where our understanding could improve. That science currently has flaws in its understanding of things is a strength, rather than a weakness.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .