Skip to content
mentua.org

What Is Occam's Razor?

Line drawing of two points joined by a straight line and by two dotted detours through extra points, with a razor blade cutting across the detours and the end point filledLine drawing of two points joined by a straight line and by two dotted detours through extra points, with a razor blade cutting across the detours and the end point filled

5 min read≈ 1,200 words3 questions · up to 3 points

What this is about

Your bike tyre has been flat every morning for a week, and two quite different explanations fit what you have seen. You learn which one to check first, why that choice is usually right and when the rule fails, and also which famous sentence William of Ockham never wrote.

In a nutshell· skip ahead

Occam's razor is a rule for choosing between explanations: when two of them fit the evidence equally well, you prefer the one that needs fewer assumptions, because every extra assumption is one more way to be wrong. The razor tells you which explanation to test first, but it does not prove that the simpler one is true.

For a week, your bike tyre has been flat every morning, although you pump it up every evening. Your brother thinks someone lets the air out every night, while you suspect a small hole in the tube. Both explanations fit what you have seen, so which one should you check first?

What is Occam’s razor?

Occam’s razor says that when two explanations fit the evidence equally well, you prefer the one that needs fewer assumptions, where an assumption is anything the explanation needs to be true that you do not already know. The rule is named after William of Ockham (also spelled “Occam”), an English philosopher of the 14th century, and the “razor” cuts away whatever an explanation does not need.

Applied to the tyre, the razor works in four steps.

  1. List the explanations that fit. A slow leak and a person who lets the air out at night both fit the facts.
  2. Count what each one assumes. The leak needs only a small hole, and tyres get small holes all the time. The other explanation needs a person who knows your bike, comes every night without being seen, and has some reason to do it.
  3. Prefer the explanation with fewer assumptions. So you test the leak first by holding the pumped-up tube under water, where a hole shows as a thin line of bubbles.
  4. Let the test decide. If you find no bubbles, the leak explanation has failed, and the night visitor becomes the next one to test.

The razor did not prove that your brother was wrong. It only told you which explanation to test first, and the test, not the razor, settles the question.

Why should fewer assumptions make an explanation more likely?

Philosophers still argue about why simplicity should matter, but the razor works in practice for two reasons.

The first reason is arithmetic, because an explanation is only true if all its assumptions are true. Suppose that each of the night visitor’s four assumptions has a chance of one in two and that they do not depend on each other. Together they then have a chance of only one in sixteen, and every extra assumption can lower that chance but never raise it.

This also gives you a warning sign. If you watch your bike one night and see nobody, your brother could say the visitor comes after three in the morning, which saves his theory but lowers its chance. An explanation that survives only by adding excuses deserves less trust, and Bayes’ Theorem, Explained shows how to do this reasoning with real numbers.

The second reason is that real data always contains small errors. Imagine you measure the temperature every hour for six hours. A straight line misses each point a little, while a curve with enough bends hits all six exactly. But the curve has copied the measuring errors too, so its guess for the seventh hour is usually worse than the line’s. Statisticians call this overfitting: a model fits the past so closely that it predicts the future badly. In the 1970s the statistician Hirotugu Akaike built a mathematical form of the razor, a measure that charges a price for every adjustable number in a model.

Isaac Newton made the same principle the first rule of reasoning in his Principia, the book on motion and gravity.

We are to admit no more causes of natural things than such as are both true and sufficient to explain their appearances.

Isaac Newton, Principia, Book III, Rule I, tr. Andrew Motte (1729)

In plain words, you should accept only causes that really exist and are enough to explain what you observe.

What Occam’s razor teaches you

The razor teaches you to ask what an explanation needs to be true that you do not already know, and to compare that list with the lists for the other explanations.

  • Health. Medical students learn the saying “When you hear hoofbeats, think of horses, not zebras.” A cough in winter is more likely a cold than a rare lung disease, so a doctor tests for the common cause first.
  • The news. A conspiracy theory often fits the known facts, but it usually needs many people to keep a secret for years, and every extra person is an assumption. A mistake or bad luck often fits the same facts with far fewer.

Where does the razor go wrong?

The most common mistake is to read the razor as “the simplest explanation is usually right”. In fact it only compares explanations that fit the evidence equally well, so a simple explanation that does not fit the facts loses to a complex one that does.

The history of astronomy shows both sides. By the 1840s it was clear that Uranus did not move exactly as Newton’s laws predicted. The French astronomer Urbain Le Verrier kept the laws and assumed an unseen planet instead, and in 1846 the German astronomer Johann Galle found Neptune within about one degree of the predicted spot.

In 1859 Le Verrier tried the same move, because Mercury’s orbit turns a little faster than the known planets could explain. Astronomers called his second unseen planet Vulcan and searched for it for decades, but it does not exist. In 1915 Einstein’s general theory of relativity explained the orbit with a new theory of gravity and no extra planet. So the same simple step gave a real planet once and an imaginary one the next time, and only observation could tell the two apart.

The strongest objection is that “simple” has more than one meaning. The philosopher Alan Baker separates parsimony, which means assuming fewer things, from elegance, which means having fewer or shorter laws. The two can conflict: Neptune added a planet but kept Newton’s laws, whereas Einstein removed the extra planet but made the laws harder. Debates about the multiverse turn on this question, as Why Does the Universe Seem Fine-Tuned for Life? shows: many unseen universes are a lot of extra things, but they may follow from simple laws.

There is also no proof that nature itself is simple, so some philosophers defend the razor as a rule of good method rather than a fact about the world. In medicine, for example, an older patient can have two or three diseases at once, so looking for one single cause can mislead.

Learn more

The idea is older than Ockham: Aristotle and Thomas Aquinas had already said that nature uses no more means than it needs. Ockham, a Franciscan friar born around 1287 in the village of Ockham near London, used the rule so often that it took his name. His own Latin words include “Pluralitas non est ponenda sine necessitate” (“plurality should not be assumed without necessity”), but the famous sentence “Entities must not be multiplied beyond necessity” is not in his writings. In 1918 William Thorburn traced its earliest close version to the Irish Franciscan John Punch, in a book printed in 1639.

Why Is the Night Sky Dark? Olbers’ Paradox, Explained shows another way to test an explanation: ask what you should already see if it were true.

Quiz

3 questions · up to 3 points

Open a question and pick one answer. Only your first answer counts: +1 if it is right, −1 if it is wrong.

  1. What does Occam's razor say?
    Show answer

    When two explanations fit the evidence equally well, prefer the one that needs fewer assumptions. This is the rule, and it applies only when both explanations fit the facts, because it compares how much each of them has to assume.

  2. A plant on your windowsill is drooping. It could need water, or it could have a rare disease and an insect pest at the same time. What does the razor suggest?
    Show answer

    Water it first, and look for disease and pests only if it does not recover. Lack of water needs one common assumption, while the other explanation needs two rarer ones at once. So you test the simpler one first and keep the other in reserve.

  3. Which statement about Occam's razor is accurate?
    Show answer

    It helps you choose between explanations that fit the evidence equally well, but the evidence still decides. The razor is a guide for what to try first, and the evidence has the last word: Le Verrier's planet Vulcan was the same kind of simple answer as Neptune, but observation showed that it was wrong.

The takeaway

Occam's razor says that when two explanations fit the evidence equally well, you should prefer the one that needs fewer assumptions. It works because every extra assumption is one more way to be wrong, and because complicated models tend to copy the errors in the data. So when your tyre is flat every morning, you look for a small hole before you suspect a person who lets the air out at night. The razor only decides what to test first, and the evidence must still decide what is true.

References

  1. Simplicity. Alan Baker, Stanford Encyclopedia of Philosophy, 2022

    The best overview of what simplicity means in science, the difference between fewer things and simpler laws, and the arguments for and against the razor.

  2. The Myth of Occam's Razor. William M. Thorburn, Mind 27 (107), 345–353, via Wikisource, 1918

    The paper that showed the famous sentence is not in Ockham's writings, with a list of the Latin phrases he did use and an account of where the name "razor" came from.

  3. William of Ockham. Sharon Kaye, Internet Encyclopedia of Philosophy

    A readable account of Ockham's life and thought, with a short section on the razor, its older roots and how he used it.

occams-razorockhams-razorlaw-of-parsimonyparsimonysimplicitywilliam-of-ockhamphilosophy-of-sciencescientific-methodcritical-thinkingexplanationhypothesis-testingoverfittingmodel-selectionisaac-newtonle-verrierneptunevulcan

Related reading

More in Philosophy →

Bayes' Theorem, Explained

Why a positive test result often means less than it seems. Bayes' theorem worked through with 1,000 women and a mammogram, step by step.

8 min read≈ 1,770 words3 questions

Plate Tectonics, Explained

How the Earth's outer shell moves in plates, how magnetic stripes on the ocean floor proved it, and what the theory still cannot explain.

5 min read≈ 1,210 words3 questions