Occam’s Razor Explained: What It Means, What It Doesn’t, and Who Actually Said It

Last updated: October 21, 2026

Occam’s razor, also called the principle of parsimony, says that when several explanations account for the evidence equally well, you should prefer the one that makes the fewest assumptions. It is named after William of Ockham (c. 1287 to 1347), an English Franciscan philosopher and theologian, who wrote in his commentary on Peter Lombard’s Sentences that plurality should never be posited without necessity (numquam ponenda est pluralitas sine necessitate), a principle Britannica renders as “plurality should not be posited without necessity”. The razor does not say that the simplest explanation is true. It says that assumptions which explain nothing extra should be shaved off.

You have probably seen it used to end an argument: someone says “Occam’s razor”, as if the simplest story had just won by default. The principle is more modest than that, and it was not even coined by the man it is named after.

Ockham Never Wrote the Famous Sentence

The best-known formulation, entia non sunt multiplicanda praeter necessitatem, entities must not be multiplied beyond necessity, is nowhere to be found in Ockham’s texts, as the SEP entry on Ockham puts it.

Ockham’s own wordings are close cousins. Besides the Sentences line, he wrote in the Summa Logicae that it is futile to do with more what can be done with fewer (frustra fit per plura quod potest fieri per pauciora). The “entities” sentence is first recorded in 1639, in the Irish Franciscan John Punch’s commentary on Duns Scotus, in the form non sunt multiplicanda entia sine necessitate, which Punch called a common axiom of the schools. Ockham did not invent the idea either: versions appear in Aristotle, Ptolemy, Robert Grosseteste and Thomas Aquinas. His own razor also came with an exception clause that the modern slogan drops. According to the SEP, Ockham accepted three things that could make an entity necessary: reason, experience and infallible religious authority, meaning Scripture.

What made the razor his was how hard he used it, and on what. Ockham was a nominalist, but he did not use the razor to dispose of universals such as “humanity”: the SEP stresses that he rejected realist theories of universals because he found them incoherent. He saved the razor for Aristotle’s categories, arguing that most of the ten, quantity and relation among them, are not separate things in the world. His own example is that “Socrates is similar to Plato” can be true simply because both are white, with no extra entity called “similarity” sitting between them. His life was less parsimonious. Summoned to the papal court at Avignon to answer charges about his teaching, he fled in 1328 with the head of his order, Michael of Cesena, to the emperor Louis of Bavaria during the dispute with Pope John XXII over Franciscan poverty, was excommunicated, and died in Munich in 1347.

Why It Is Called a Razor

The name comes from a place and a tool. William came from Ockham, a village in Surrey whose name means “oak hamlet”, according to the SEP, and “Occam” is the Latinised spelling. A razor shaves off what is not needed.

The Latin label novacula Occami, Occam’s razor, is first recorded in 1649, in the Philosophia Christiana de Anima of the Flemish theologian Libert Froidmont, who claimed to have coined it, three centuries after Ockham’s death. Much later, other rules of thumb borrowed the name. Hanlon’s razor, credited to Robert J. Hanlon in Arthur Bloch’s Murphy’s Law Book Two (1980), advises never to attribute to malice what stupidity adequately explains, though Quote Investigator finds earlier relatives of the idea. Hitchens’s razor comes from a 2003 Slate column by Christopher Hitchens and was repeated in God Is Not Great (2007): what is asserted without evidence can be dismissed without evidence. They are heuristics, not laws of logic.

Ockham’s razor also had a contemporary opponent. His fellow Franciscan Walter Chatton (c. 1290 to 1343/44) answered with an “anti-razor”: if three things do not suffice to make a statement true, posit a fourth, and so on. Chatton used it to argue that some relations, such as causal ones, are real things, against Ockham, who denied most relations, as the SEP entry on Chatton explains. The anti-razor reminds us that “without necessity” cuts both ways: when the evidence needs another entity, refusing to posit one is as much a mistake as inventing one.

Occam’s Razor: Examples

The razor earns its reputation in cases where two explanations fit the facts and one of them needs extra machinery to do it.

Isaac Newton made a version of it the first of his rules of reasoning in the Principia: admit no more causes of natural things than are both true and sufficient to explain their appearances. Data scientists meet it as overfitting. A model with too many adjustable parameters can fit past data perfectly and predict new data badly, which is why statisticians penalise extra parameters with tools such as the Akaike and Bayesian information criteria (AIC and BIC), as the SEP entry on simplicity explains. The same entry gives a physics example. In the 1930s, beta decay seemed to lose spin, and physicists explained the gap by positing a single new particle per decay, the neutrino. Many hypotheses could fit the data, but positing one particle rather than two or three was, in the SEP’s words, the obvious default choice, and the neutrino was first detected in 1956.

Doctors learn a cousin of the razor, usually credited to Theodore Woodward of the University of Maryland in the 1940s, though the earliest printed version Quote Investigator found dates from 1962: when you hear hoofbeats, think horses, not zebras. Strictly, that is a rule about base rates, common things are common, rather than about counting assumptions. Medicine also has the counterweight, Hickam’s dictum, usually traced to the physician John Hickam: a patient can have as many diseases as they please. Astronomy offers the cautionary tale. Copernicus’ sun-centred model of 1543 still used epicycles, and by replacing Ptolemy’s equant with extra small circles it became, in one respect, more complex, according to the SEP entry on Copernicus. What it gained was unity, one ordered system instead of a separate model for each planet. The decisive simplification came with Kepler’s elliptical orbits in 1609, which shows that the simpler theory is sometimes the one nobody has written yet.

What Occam’s Razor Doesn’t Mean

A common misreading is “the simplest explanation is usually the correct one”. The razor makes a narrower claim: among explanations that account for the evidence equally well, do not add assumptions that do no work.

That condition matters. A simpler explanation that fails to explain the evidence gets no help from the razor, because it is simply wrong. A line often attributed to Albert Einstein, that everything should be as simple as possible but not simpler, makes the point well, but it cannot be found in his writings in that form; Quote Investigator traces the aphorism to a 1950 article by the composer Roger Sessions. In his Herbert Spencer Lecture at Oxford in 1933, Einstein said something more careful: theory should make its basic elements as simple and as few as possible without giving up an adequate representation of experience.

The razor also cuts against much of what people use it for online. Many conspiracy theories add entities, a secret plan, hidden coordination, a cover-up, to explain facts a shorter account already covers. Their appeal is often a different virtue, explanatory unity, one hidden cause behind everything, as the philosopher Brian Keeley noted in “Of Conspiracy Theories” (Journal of Philosophy, 1999). The razor counts assumptions, though, and a secret that requires thousands of silent participants is an expensive assumption.

Why prefer simpler theories at all? Philosophers of science have offered several answers, from Bayesian arguments that simpler hypotheses deserve higher prior probability to the statistical point about overfitting, and none of them is uncontested, as the SEP entry on simplicity lays out. A more defensible view treats the razor as a policy for searching, not a guarantee of truth: start with fewer assumptions and add more only when the evidence forces you to. The SEP calls Ockham’s own version merely a cautionary methodological principle, and since it allowed reason, experience and Scripture to supply the necessity, it is also closer to his text than “the simplest answer is right”. For a related puzzle about how far justification can go, see our post on the Münchhausen trilemma.

For a philosopher who took the opposite route and doubted everything first, read What is Cartesian doubt? Descartes explained.