The Feynman Technique: Learn Anything by Explaining It

What is the Feynman technique?

The Feynman technique is a study method built on one observation: you understand something when you can explain it simply, and trying to explain it simply is the fastest way to find out whether you do. You choose a concept, write or say an explanation for someone new to the subject, and pay close attention to the places where the explanation falls apart. Those places become your study list.

It is named after Richard Feynman, the American physicist who shared the 1965 Nobel Prize in Physics for work on quantum electrodynamics and who was widely admired as a lecturer. The four-step version now called the Feynman technique is a later summary inspired by his approach to learning and explaining, not a method he set down himself. The name stuck because the approach reflects his insistence on understanding an idea rather than knowing its name.

The four steps of the Feynman technique

The steps fit on an index card; the work is in being honest at step three. You repeat the last three steps until you can give the whole explanation without stalling and can answer a follow-up question about any sentence in it. The finished version is usually shorter than your first attempt, because you have stripped out everything you did not need.

1. 1Choose one concept and write its name at the top of a blank page. Keep it narrow: “why the seasons happen,” not “astronomy.”
2. 2Explain it in plain words, in writing or aloud, as if to a curious friend who has never studied the subject. Use short sentences, everyday vocabulary and at least one example.
3. 3Mark every place where you got stuck, skipped a step, said “basically” or “somehow,” or used a technical term you could not unpack. Each mark is a gap.
4. 4Go back to your textbook, notes or lecture for each gap. Then rewrite the explanation more simply, with an analogy where it helps, and explain it again from the start without looking.

A Feynman technique example: why we have seasons

Take a concept most people think they already understand. A first attempt at explaining the seasons often goes like this: summer is when Earth is closer to the sun, and winter is when it is farther away. Written out plainly, it looks complete. Step three asks what the explanation predicts, and the trouble appears. If distance caused the seasons, both hemispheres would have summer at the same time. They do not: when it is summer in Australia, it is winter in Canada.

Back to the source. The real cause is the tilt of Earth's axis, about 23.5 degrees. As Earth orbits the sun, the axis keeps pointing the same way in space, so for part of the year the Northern Hemisphere leans toward the sun and gets more direct sunlight and longer days, while the Southern Hemisphere leans away. Six months later the situation reverses. Earth's distance from the sun does vary a little, but Earth is actually closest in early January, during the northern winter.

The rewritten explanation is shorter than a textbook paragraph, uses no unexplained terms, and could be demonstrated with a lamp and a tilted ball. It also contains the reason the first version was wrong, which is the part a teacher or exam is most likely to probe.

How to spot the gaps in your explanation

Gaps rarely announce themselves. An explanation can sound fluent and still be hollow, especially when it borrows phrases straight from the textbook. Rereading your explanation the next day, or listening back to a recording of it, makes the hollow parts easier to hear.

When you find a gap, write the question it raises in the margin instead of a cross. “Why does the axis keep pointing the same way?” is something you can look up and answer; a cross is not. These are the signs to look for:

• A technical term you used but could not define in plain words if someone asked.
• Filler standing in for a step: “basically,” “somehow,” “it just,” “and so on.”
• A jump between two sentences where you cannot say why the second follows from the first.
• A circular definition, such as describing supply as “what suppliers supply.”
• No example, or only the example from the textbook and none of your own.
• An analogy whose limits you cannot name.

Using analogies without being misled by them

Step four usually calls for an analogy, because an analogy connects a new idea to something your listener already understands. Electric circuits are often explained with water in pipes: voltage is like water pressure, current is like the rate of flow, and resistance is like a narrow section of pipe. The picture makes the relationship between the three feel intuitive, which is why teachers keep reaching for it.

Every analogy has a point where it breaks, and knowing that point is part of understanding the concept. Open a switch and the current stops entirely; cut a water pipe and the water pours out. A good Feynman explanation uses the analogy for what it shows, then names one thing it gets wrong. If you cannot find a single limit, you probably have not tested the analogy hard enough.

When the Feynman technique works, and what to pair it with

The technique is strongest where understanding is the hard part: mechanisms in biology, principles and proofs in physics and math, arguments in philosophy and economics, causes in history. It is weaker on its own where the hard part is holding on to many separate items, such as vocabulary or a list of dates. You can explain why a word means what it means, but you still have to remember thousands of words.

The technique is itself a form of active recall, since you rebuild the idea from memory before checking it. For material where it fits poorly, use it to give facts meaning and let spaced self-testing do the remembering.

Explaining to a real listener

Explaining to a blank page has one weakness: the page never asks “why?” A listener does. Explaining to a classmate, a friend outside the subject or a family member adds the follow-up questions that expose the gaps you skimmed over. Ask them to interrupt whenever a sentence stops making sense, and treat every interruption as a gap to fix, not a failure on their part.

If nobody is around, record yourself and listen back the next day, when the explanation no longer sounds like your own thoughts. A listener who knows the material is stronger still, because they can tell you not only that a step is unclear but that it is wrong. That is the idea behind prompting an AI to question you instead of explaining for you, and behind calling an AI tutor, where you talk an idea through instead of typing it.