Occam's Razor

Also known as: Principle of Parsimony

technique · philosophy · philosophical-tradition

Among competing hypotheses, prefer the one with fewest assumptions.

Occam's Razor is the methodological principle that among competing hypotheses adequate to explain a phenomenon, the one requiring the fewest assumptions or entities should be preferred. The principle is named for William of Ockham (c. 1287-1347), the English Franciscan philosopher and theologian whose works frequently invoked something like 'plurality must never be posited without necessity' and 'it is futile to do with more what can be done with fewer.' Ockham did not invent the principle (similar formulations appear in Aristotle, Maimonides, and Aquinas) but his vigorous methodological use earned the eponymy. The principle is preference rather than truth criterion — simpler hypotheses are not therefore more likely true, but they are preferred when both fit the evidence equally because they make fewer commitments and are easier to refute. Modern formal versions include statistical model selection criteria (AIC, BIC, Minimum Description Length) that mathematically formalize parsimony as a tradeoff against fit. The principle is widely cited but inconsistently applied — what counts as a 'simpler' hypothesis is itself contested, with simplicity measured variously as fewer entities, fewer parameters, more general principles, or shorter description. The principle is best understood as one heuristic among several rather than as an algorithm for hypothesis choice.

Originators

William of Ockham (eponymous); precursors including Aristotle, Maimonides, Aquinas high

Year / Decade

Early 14th century (Ockham); ancient and medieval precursors high

Primary sources

Ockham, W. (14th century). Various works including Summa Logicae, Sentences commentary, Sober, E. (2015). Ockham's Razors: A User's Manual high

Core components

Preference for fewer assumptions
'Plurality must never be posited without necessity' (Ockham)
'Entities must not be multiplied beyond necessity' (later formulation, 'lex parsimoniae')
Heuristic rather than truth criterion
Modern formalizations: AIC (Akaike Information Criterion), BIC (Bayesian Information Criterion), Minimum Description Length
Connection to falsifiability (simpler theories are more easily refuted)
Distinction from anti-realist 'eliminate entities' interpretation

Primary use case

Scientific methodology and hypothesis selection; statistical model selection (AIC, BIC, MDL); pedagogical framework for distinguishing well-supported from over-elaborated explanations; widely-cited reasoning heuristic across disciplines; foundation for arguments against unnecessarily complex theoretical posits; reference in applied research design.

Common criticisms

What counts as 'simpler' is itself contested and depends on description language — a hypothesis that is simple in one framework may be complex in another
sometimes the more complex hypothesis is correct (biology has many entities and processes that elementary chemistry would have ruled out as superfluous)
the principle's preference status is sometimes overstated as a truth criterion ('the simpler hypothesis is more likely true'), which is not the original meaning
in machine learning, simplicity as measured by parameter count interacts with overfitting concerns in ways that don't simply reduce to the original principle
popular use often deploys the razor as conversation-stopper rather than as one heuristic among several
Sober and others have argued the principle should be replaced by more rigorous Bayesian or information-theoretic considerations.

Lineage

Siblings: Hanlon's Razor, Sturgeon's Law