QuestionsCan we be certain of anything?The problem of induction
Can we be certain of anything?We must doubt

The problem of induction

David Hume 1711 · 1776Neither causation nor induction can be proved: custom guides us.
See profile
What it claims

Our inferences about the future rest not on reason but on habit.

Why

Predicting the future assumes it will resemble the past, which can be neither proven by logic nor by experience without going in circles.

In short

"The problem of induction" is an argument by David Hume (Enquiry, section 4) on the question "Can we be certain of anything?" (answer: we must doubt). Its claim: Our inferences about the future rest not on reason but on habit. According to AskPhilosophy's logic check (Z3 solver), the conclusion does follow from its 4 premises. AskPhilosophy presents 3 objections, 1 of them with a documented reply, and the original text behind each premise, checked word for word.

Enquiry, section 4
Valid

The sun has always risen: we are sure it will rise tomorrow. But what grounds this certainty? Hume shows that no reasoning justifies it.

P1Every inference from experience assumes that the future will resemble the past.
P2There are only two kinds of reasoning: demonstration (as in mathematics) and reasoning from experience.
P3We cannot demonstrate that the future will resemble the past: that it might change involves no contradiction.
P4Nor can we prove it by experience: that would be circular, since every proof from experience already assumes it.
Therefore our inferences about the future rest not on reason but on habit.

click a premise to reject it

If you accept every premise, logic forces you to accept the conclusion. So the debate is about the premises.
Checked by a logic solver: the premises cannot all be true while the conclusion is false.
The argument is valid as Hume writes it: no hidden premise is needed, and each one is essential. That is why the debate has focused for almost three centuries on the premises, especially P2 (are there really only two kinds of reasoning?).
Analogy. A chicken is fed every day of its life by the same man; each day confirms it will be fed tomorrow. One day the man wrings its neck instead. The image is Bertrand Russell's (The Problems of Philosophy, 1912), not Hume's.

Objections

3

Each objection is a conversation: click to see who replies and where the debate stands.

Objection anticipated by Hume attacks premise 4

But I have always found, in every past instance, that these qualities go with these effects: experience is enough to prove it.

Hume replies
Hume
That proves it was true in the past, not that it will be tomorrow. Every proof from experience presupposes precisely the resemblance of future to past.
Kant 1781 · attacks premise 2

There is a third kind of knowledge: principles that do not come from experience but make it possible, such as causality. They are known a priori without being mere definitions.

no reply
summary: their text is not yet in our corpus
No reply possible: objection raised in 1781, after the author's death (1776).
Peter Strawson 1952 · attacks premise 1

Asking for a justification of induction by something other than itself makes no sense: relying on experience is precisely what "being reasonable" means about matters of fact.

no reply
summary: their text is not yet in our corpus
No reply possible: objection raised in 1952, after the author's death (1776).
What today's voices say
Defend it
Steven "Destiny" BonnellWhen Alex O'Connor says we trust unprovable intuitions about the outside world, so we could trust moral ones, he sums up his own view: no provable moral fact.
Karl PopperHe accepts Hume's critique: experience can never verify a universal law. But he denies that science really uses induction, so the problem does not lead to scepticism.