Hume’s problem of induction

In short: The problem of induction asks what entitles us to infer from past observations to future or not yet observed cases. The fact that the sun has risen every morning so far does not logically prove that it will rise tomorrow.

In more detail: The Scottish philosopher David Hume formulated the problem particularly sharply in the 18th century. It affects everyday life and science alike, because both constantly rely on generalisations from experience.

In detail

• Induction and deduction: In a deductive argument the conclusion follows necessarily from the premises; in an inductive argument one infers from individual cases to a general rule, which may also be false.
• Hume’s objection: Induction assumes that nature stays uniform, so that the future resembles the past; but this assumption can neither be proven logically nor justified by experience without going in a circle.
• The circle: Anyone who says that induction has always worked so far is once again justifying induction with an inductive argument.
• Habit instead of reason: Hume held that we expect experiences to repeat themselves out of habit, not because reason proves it to us.
• The turkey example: A well-known image describes an animal that is fed every day and concludes that it will always go on like this, until one day it is slaughtered; Bertrand Russell used a similar example with a chicken.
• Popper’s answer: Karl Popper proposed not basing science on induction but boldly putting forward theories and testing them through attempts to refute them.
• Probability: Other approaches try to rate inductive inferences as more or less probable instead of demanding absolute certainty.
• Practical relevance: Nobody gives up induction in everyday life, but the problem is a reminder that knowledge from experience always remains provisional (logik-grundlagen-schluesse, rationalismus-und-empirismus, erkenntnistheorie-wissen-und-glauben).