The Feynman Technique: How Teaching Something Reveals What You Don't Actually Know

Here’s a small test. Explain, out loud, exactly how a zipper works.

Not what it does - how it works. What the little slider actually does to those two rows of teeth. Why they lock going one way and release going the other.

Most people stall within three sentences. You’ve used zippers a thousand times, felt completely sure you understood them, and discovered in real time that you don’t. That collapse - from confident to clueless in a few seconds - is the most useful feeling in all of learning. This post is about how to trigger it on purpose.

Why this matters

You almost certainly believe you understand things you can’t actually explain.

That’s not a personal flaw - it’s how every human mind works. And it’s expensive, because the feeling of understanding is exactly what stops you from studying the thing you haven’t really learned. You close the book, satisfied, with a hole in your knowledge you never noticed.

The Feynman Technique is a simple fix. It’s a way of forcing that hidden hole into the open, fast, using nothing but a blank page and the act of explaining. It works whether you’re studying for an exam, learning a new skill at work, or trying to genuinely understand something you’ve been nodding along to for years.

Better still, it stacks three of the most powerful learning mechanisms into a single action - and you can run it in ten minutes, alone, tonight.

The illusion that hides in the “how”

Psychologists have a name for what happened with the zipper: the illusion of explanatory depth - our tendency to believe we understand how things work in far more detail than we really do.

Two researchers, Leonid Rozenblit and Frank Keil, pinned it down in 2002. Across 12 separate studies, they asked people to rate how well they understood everyday objects - zippers, flush toilets, helicopters, speedometers - on a scale of 1 to 7. People rated themselves fairly confident.

Then came the twist. The researchers asked them to write out a detailed, step-by-step explanation of how the thing actually works. Halfway through, people froze. They discovered, mid-sentence, that they could not explain a toilet. Asked to re-rate their understanding afterward, their scores dropped sharply.

Here’s the crucial part: the illusion showed up only for explanatory knowledge - the how and why of causes and mechanisms. It didn’t appear for plain facts or memorized procedures. You can know that a zipper zips without knowing how it zips. The gap hides precisely in the causal machinery - which is exactly the kind of knowledge most real learning is made of.

Think of your knowledge like a bathroom mirror. It looks perfectly clear - until you breathe on it. The fog reveals every smear and streak that was there the whole time. Explaining something is breathing on the mirror. Suddenly you see every gap you never noticed.

Who Feynman was, and what his technique actually is

Richard Feynman (1918-1988) was a Nobel Prize-winning American physicist, famous not just for his brilliance but for a rare gift: he could explain ferociously hard ideas in plain, everyday words. His personal test for real understanding was blunt. Could he explain a topic to a first-year university class? If he couldn’t boil it down to a freshman lecture, he decided he didn’t truly understand it himself.

From that spirit, modern writers packaged a study method now called the Feynman Technique - a four-step loop for learning anything by forcing yourself to explain it simply.

Here’s the loop:

1. Pick a concept. Write its name at the top of a blank page. One idea at a time.
2. Explain it in simple language, as if teaching a curious 12-year-old. Ban jargon. If you must use a technical word, define it in plain words right there.
3. Spot the gaps. Every “um,” every hand-wave, every moment you fall back on a fancy textbook phrase you can’t unpack - that’s a real hole. Circle it.
4. Go back to the source, relearn exactly those gaps, then simplify and add an analogy - a comparison to something familiar. Run the loop again until your explanation flows start to finish.

The magic is in step 3. Explaining doesn’t just review the material - it locates your ignorance with pinpoint accuracy. You can only explain what you actually understand, so the exact moment you stall is the exact spot to study next.

One honest note about the myth

Feynman never actually wrote down these four steps. The tidy “Feynman Technique” is a modern packaging, popularized largely by writer Scott Young and productivity authors. Feynman genuinely championed plain-language explanation, and his biographer James Gleick reports he kept a “Notebook of Things I Don’t Know.” But the neat recipe is a later invention. It’s fair to say the technique is inspired by Feynman and backed by the evidence - just don’t believe he personally handed down the steps.

Why explaining works so well

Explaining is powerful because it secretly does three things at once - three mechanisms that show up throughout the science of how learning actually sticks:

• Retrieval. To explain without notes, you have to pull the idea out of memory - and pulling strengthens the memory. This is the same engine behind why testing yourself beats re-reading.
• Elaboration. To make it simple, you connect it to familiar things and rephrase it in your own words. That builds richer, more durable understanding.
• Gap-finding. Because you can only explain what you understand, every stall maps a gap precisely.

There’s a fourth ingredient too: self-explanation - explaining new material to yourself as you study, asking “why is this true? how does this connect?” A 2018 meta-analysis by Bisra and colleagues combined 64 studies and found that prompting learners to self-explain produced moderate-to-large gains (an effect size of about g ≈ 0.55). The gains were strongest when people explained causal relationships - the why and how, exactly the knowledge the illusion hides.

But there’s a trap. If you explain by copying the source’s wording, you get none of this. It sounds fluent, but you’re just reciting - no retrieval, no elaboration, no gaps exposed. Fluency borrowed from the textbook is the illusion of competence in disguise. The whole point is to regenerate the idea in your own words.

Study to teach, not just to pass

Now widen the lens - from explaining to yourself, to teaching someone else.

A striking finding shows up again and again: you learn material better when you study it in order to teach it than when you study it to pass a test. Researchers call this the protégé effect - the boost you get from preparing to be someone’s mentor.

The name comes from a 2009 study by Catherine Chase and colleagues. They built software called Betty’s Brain, where 8th-graders taught biology to a computer character named “Betty.” Students who believed they were teaching Betty spent more time reading, learned more, and monitored their own errors more closely than students studying the same material for themselves. The effect was largest for lower-achieving students - the ones who usually need the most help.

Why would merely expecting to teach change anything? A 2014 study by Nestojko and the Bjorks tested it cleanly. Students studied a passage; half were told they’d be tested, half were told they’d have to teach it. Then - the sneaky part - everyone was actually tested and nobody actually taught. Yet the “expecting-to-teach” group produced more complete, better-organized recall, especially on the main points. Just the mindset changed how they studied.

Think of it like packing a suitcase. Studying to pass a test is throwing clothes in loosely. Studying to teach forces you to fold and organize everything so it fits neatly and you can find any item later. That folding-and-organizing is exactly what produced the better recall.

But intending to teach isn’t enough

Here’s where the popular story oversells itself. It’s tempting to conclude: “Great, I’ll just plan to teach and reap the rewards.” Not so fast.

A 2013 study by Fiorella and Mayer tested this directly. College students learned the Doppler effect - why an ambulance siren changes pitch as it races past. Some just prepared to teach; others prepared and actually taught, by recording a video lecture for a fictitious student.

On an immediate test, both groups did about equally well. But on a delayed test days later, only the students who both prepared and actually taught kept the strongest gains. The benefit of merely intending to teach faded. The act of actually generating the explanation locked it in.

The lesson is clean: the mindset primes you, but the generation cements it. Intention alone doesn’t survive to the delayed test. You have to produce the explanation - say it, record it, or write it down.

This echoes a much older finding. A 1982 meta-analysis of 65 tutoring studies found that tutored students improved - but so did the tutors themselves. Later work explained why: effective tutors don’t just recite facts, they generate explanations and integrate knowledge as they teach. Passive tutors who read their notes aloud gained little. Teaching isn’t magic. It’s real but conditional, and the condition is that you actively build the explanation.

The rubber duck: your zero-effort student

You don’t need a live audience. Here’s a beloved trick from computer programming, described in the 1999 book The Pragmatic Programmer: rubber-duck debugging.

Programmers keep a rubber duck on their desk. When their code won’t work, they explain it to the duck - line by line, out loud, in plain words. Astonishingly often, the bug leaps off the screen mid-explanation, before the duck has to say a thing. Why? Because articulating each step forces you to actually check each step, instead of skimming past the broken assumption you’d been glossing over.

That’s the Feynman Technique with a bath toy. The duck is a stand-in student that forces you to articulate. Your “student” can be a rubber duck, an empty chair, an imagined 12-year-old, or a “future you” who’ll read your notes next month. What matters is that you produce the explanation - out loud or on paper - not that anyone’s listening.

Common misconceptions

A few myths quietly blunt the technique. Here’s the reality.

• “If I understood the lecture, I understand the topic.” The illusion of explanatory depth makes that feeling systematically inflated. Prove it on a blank page before you trust it.
• “A smooth explanation means I’ve got it.” Sometimes you just breezed past the hard part without noticing. Stress-test the steps you skipped - ask “but why exactly?” at each stall.
• “Planning to teach is basically teaching.” The intention effect fades by the delayed test. You have to actually produce the explanation to keep the gains.
• “Big words show mastery.” Usually the opposite. Jargon hides confusion; if you can’t say it plainly, you probably don’t fully get it.
• “Reading my notes aloud to a friend counts as teaching.” It does almost nothing. The gains come from generating causal explanations from memory, not from reciting summaries.

How to use this

1. Run one full Feynman loop this week. Take a concept you’re studying. Write its name on a blank page and explain it as if to a curious 12-year-old. Circle every place you stall, relearn just those spots, then re-explain until it flows. One clean loop teaches you more than an hour of re-reading.
2. Adopt a rubber duck (or an empty chair). Pick something you think you understand - how a zipper works, how interest compounds, a concept from your field - and explain it out loud, step by step. Notice exactly where you run dry. That’s your study list.
3. Produce one teaching artifact. Don’t just intend to teach - make something. Write a short explainer, record a two-minute voice memo, or make clean notes for “future you.” Producing it is what makes the learning last.
4. Ban the jargon on purpose. Whenever you reach for a technical term, stop and define it in plain words as if to a child. If you can’t, you’ve found a gap - not a vocabulary choice.
5. Regenerate, never copy. Close the book before you explain. If your words are the textbook’s words, you’re reciting, not learning.

Conclusion

The single idea worth keeping is this: the feeling of understanding and actual understanding are two different things, and explaining is how you tell them apart. You can’t fake your way through teaching a 12-year-old. The moment you stall is the moment you’ve found what to learn next.

So the next time something feels obvious, breathe on the mirror. Explain it out loud, to a duck if you have to, and watch how quickly the smears appear.

There’s a deeper reason this works, though - explaining forces you to pull ideas out of memory instead of pushing them in, and that single act of retrieval turns out to be one of the most powerful learning tools ever measured. That’s the thread worth pulling next: why testing yourself beats re-reading every time.