Problem of Induction
The problem of induction is the challenge of justifying inferences from observed cases to unobserved ones without circularity.
Definition
The problem of induction is the question of whether and how we can rationally justify the assumption that the future will resemble the past, or that unobserved instances will conform to patterns found among observed instances.
Scope
This topic covers Hume's sceptical argument that inductive inference cannot be justified by either deductive reason or experience, and the principal responses: Popper's denial that science uses induction, Reichenbach's pragmatic vindication, Strawson's ordinary-language dissolution, and inductive-justification and probabilistic strategies.
Core questions
- Can induction be justified deductively without assuming its conclusion?
- Does any inductive justification of induction beg the question?
- Is the uniformity of nature an unjustified presupposition of inductive inference?
- Can science dispense with induction altogether?
Key concepts
- uniformity of nature
- inductive inference
- circularity
- rule-circularity
- pragmatic vindication
Key theories
- Humean scepticism
- Hume argues that inductive inference rests on the uniformity of nature, which can be supported neither by demonstration nor by non-circular experience.
- Falsificationist dissolution
- Popper accepts Hume's negative conclusion but holds that science is deductive and critical rather than inductive, so the problem does not arise for it.
- Pragmatic vindication
- Reichenbach argues that the inductive rule is justified pragmatically because it will succeed if any method will.
History
Hume framed the problem in 1739–48. Twentieth-century responses divided into denials that science is inductive (Popper), pragmatic vindications (Reichenbach), ordinary-language dissolutions (Strawson 1952), and probabilistic reconstructions that recast rather than refute the sceptical challenge.
Debates
- Does science need induction?
- Popper claims the problem is avoided because theories are tested by attempted falsification, while critics reply that corroboration and theory choice smuggle induction back in.
Key figures
- David Hume
- Karl Popper
- Hans Reichenbach
- P. F. Strawson
Related topics
Seminal works
- hume1748
- popper1959
Frequently asked questions
- Why can't we just say induction has always worked?
- Arguing that induction is reliable because it has worked in the past is itself an inductive inference, so it presupposes the very principle in question. Hume's point is that this kind of defence is circular.