The Problem of Induction (SACE Stage 2 Philosophy)

What this dot point is asking

You need to reconstruct Hume's argument, show why it is a circle, and evaluate at least one serious attempt to answer it.

Hume's argument

Induction is reasoning from observed cases to a general conclusion or to a prediction about unobserved cases. We rely on it constantly: bread has nourished us before, so it will nourish us again. Hume asked what justifies this. He argued that any such inference depends on the principle of the uniformity of nature, the assumption that the future will resemble the past and that unobserved cases resemble observed ones.

How could that principle be justified? Hume divides all reasoning into two kinds. Demonstrative reasoning concerns relations of ideas, like mathematics, where the denial is a contradiction. But it is not contradictory to suppose nature might change, so the uniformity principle cannot be demonstrated. The other kind, probable or factual reasoning, is itself inductive, so to justify induction by appeal to its past success is to assume the very principle in question. The justification is circular. Hume concludes that our reliance on induction is a matter of custom or habit, not reason.

Why this matters

If Hume is right, then the entire edifice of empirical science rests on a foundation that reason cannot underwrite. Laws of nature, predictions, and even the everyday expectation that the floor will support you are all inductive. The problem is not that these beliefs are false but that we cannot give a non-circular reason for holding them. This is a sceptical result about justification, not about truth.

Popper's falsificationism

Karl Popper accepted that induction cannot be justified and tried to show that science does not actually need it. On his view, science proceeds not by confirming theories through repeated observation but by attempting to falsify bold conjectures. A theory is scientific if it forbids certain observations and therefore exposes itself to refutation. We never prove a theory true; we only retain those that have survived serious attempts to refute them.

The attraction is that falsification is deductive: a single counterexample logically refutes a universal claim, so no inductive leap is required. The standard objection is that Popper has not really escaped induction. To rely on a theory that has so far survived testing, for instance to build a bridge with it, still assumes that its past survival is a guide to future performance, which is an inductive assumption in disguise.

Other responses

Pragmatic vindication, associated with Hans Reichenbach, argues that even if we cannot prove induction works, we can show that if any method works, induction will, so we lose nothing by using it. Probabilistic approaches try to reframe the issue using Bayesian conditionalisation, though critics note these still need prior probabilities that look inductive. P. F. Strawson offered a dissolving move: to ask whether induction as a whole is rational is like asking whether the law is legal; being reasonable just means proportioning belief to evidence in the inductive way, so the demand for an external justification is confused.

Evaluation

Hume's argument is widely regarded as valid: the circularity is real and no response fully closes it on Hume's own terms. The live question is whether the demand for a non-circular justification is the right demand. Strawson and the pragmatists suggest the problem may be misframed rather than solved, while Popper reorganises scientific method around deduction at the cost of arguably reintroducing induction at the point of application. A strong answer concedes the force of the sceptical argument while assessing how far each reply changes the question rather than answering it.