Definition · Plain-language
Occam’s razor
Occam’s razor is the principle that among competing explanations, the one with the fewest unnecessary assumptions should be preferred.
The step most authors miss
Doing CRediT right? Don’t stop at the statement.
A CRediT statement credits you inside one paper. The recognition CRediT was built for happens when those roles are tied to you, persistently. Sign in with your ORCID — free — and claim your CRediT contributions on casrai.org, the home of the standard. They become a verified, portable part of your identity, not a line that disappears into one PDF.
Free: claim your contributions, then export a journal-ready CRediT statement, schema.org structured data, JATS XML, CSV or BibTeX — and preview your public profile. A membership publishes that profile publicly and verifies the journals you serve.
William of Ockham and the Latin formulation
William of Ockham (~1287–1347) was a Franciscan friar and philosopher whose logical and theological writings gave rise to the principle now bearing his name, though he stated it in several related forms. The most quoted Latin formulation — entia non sunt multiplicanda praeter necessitatem ("entities must not be multiplied beyond necessity") — does not appear verbatim in his surviving works but accurately captures his recurring methodological commitment to parsimony. The medieval principle had antecedents in Aristotle and Ptolemy, but Ockham applied it so consistently and forcefully that it became associated with him. It was later invoked by Newton ("We are to admit no more causes of natural things than such as are both true and sufficient to explain their appearances") and was implicit in the Copernican preference for a heliocentric model over increasingly complex epicycles.
AIC/BIC model selection and Hickam’s dictum
In statistics and machine learning, parsimony is formalised in model-selection criteria. Akaike’s Information Criterion (AIC) and the Bayesian Information Criterion (BIC) penalise models for each additional parameter, rewarding good fit while discouraging over-fitting. In medicine, Occam’s razor translates into diagnostic parsimony: prefer the single diagnosis that explains all symptoms before invoking multiple independent diagnoses. The counterpoint is Hickam’s dictum (attributed to John Hickam): "A patient can have as many diseases as they damn well please", reminding clinicians that in older or complex patients, multiple concurrent conditions are common and the simpler single-disease explanation may be wrong.
Hitchens’s razor and the limits of parsimony
Christopher Hitchens extended the principle to epistemology with Hitchens’s razor: "What can be asserted without evidence can be dismissed without evidence." This shifts the burden of proof rather than selecting among explanations. Occam’s razor is a heuristic, not a law of nature. Simpler explanations are preferred because they are easier to test, make fewer assumptions that could be false, and have lower prior probability of being wrong by accident — but nature is under no obligation to be simple. Complex realities do exist, and forcing a simple explanation onto genuinely complex phenomena produces errors. Occam’s razor guides theory choice under uncertainty; it does not guarantee that the simplest theory is correct.
Key facts
At a glance
- Latin formulation: entia non sunt multiplicanda praeter necessitatem
- Named after: William of Ockham (~1287–1347), Franciscan philosopher
- Principle: prefer the explanation with the fewest unnecessary assumptions
- Status: a heuristic for theory choice, not a law of nature
- AIC/BIC: statistical formalisation penalising unnecessary model parameters
- Medical counterpoint: Hickam’s dictum — patients can have multiple diseases
- Hitchens’s razor: extension to burden of proof in epistemology
Common misconceptions
What people often get wrong
Often heard: Occam’s razor says the simplest explanation is always correct.
Actually: It says the simplest explanation should be preferred as a starting hypothesis, all else being equal. Nature is not obligated to be simple. When evidence demands a more complex explanation, parsimony must yield to the data.
Often heard: Occam’s razor means you should always favour fewer entities or causes.
Actually: The principle targets unnecessary entities — those not needed to explain the evidence. When additional elements are genuinely required to fit the data, they are not "multiplied beyond necessity" and should be included. Hickam’s dictum illustrates this in medicine.
Often heard: Occam’s razor is a modern scientific principle.
Actually: The core idea traces to Aristotle and was widely used in medieval scholastic philosophy. Ockham applied it systematically in the 14th century. Newton, Copernicus and later scientists adopted it as a methodological guide long before modern philosophy of science formalised it.
Common questions
FAQ
What does Occam’s razor mean in everyday terms?+
When you have two explanations for the same observation, prefer the one that requires fewer unsupported assumptions. If your keys are not where you left them, "I misplaced them" is preferred over "someone broke in and moved them" — the first explanation needs fewer additional assumptions.
How is Occam’s razor used in machine learning?+
It is formalised in model-selection criteria such as AIC and BIC, which penalise a model for each additional parameter. A more complex model that fits training data marginally better than a simpler model may over-fit: Occam’s razor guides the preference for the simpler model that generalises better to new data.
What is the difference between Occam’s razor and Hitchens’s razor?+
Occam’s razor selects among competing explanations, preferring the one with fewest unnecessary assumptions. Hitchens’s razor addresses the burden of proof: claims made without evidence can be rejected without evidence. Both are heuristics for managing uncertainty, but they operate at different stages of inquiry.
Going deeper
