For a few months now I’ve been doing an exercise I rarely mention. I consider it one of the most useful ideas I’ve picked up in years, and also one of the most uncomfortable. That isn’t a coincidence. The two go together.
The exercise takes thirty seconds. You can do it anywhere — at your desk, on a train, after reading an article.
You ask yourself: could I explain that right now to a ten-year-old?
Not in theory. Actually, now: no jargon, no evasions, no “it’s complicated.” In simple sentences a child would follow.
Most of the things I believe I understand don’t survive the test.
* * *
The exercise isn’t mine, and it doesn’t entirely belong to the man whose name it carries either.
Richard Feynman took pride in making even the deepest ideas accessible to first-year students. A colleague at Caltech once asked him to explain why particles with half-integer spin obey Fermi-Dirac statistics. Feynman agreed: he would prepare a freshman lecture on it. A few days later he came back. He hadn’t managed it. He couldn’t bring it down to that level — and that, he said, meant we don’t really understand it.
That’s the documented criterion. It sounds modest and it’s radical: if you can’t explain something simply, you haven’t understood it.
What circulates today as the “Feynman technique” is not something Feynman ever called it. The name came into use through Scott Young’s self-education project, and Young himself notes that the technique only loosely resembles Feynman’s actual practice. The ten-year-old I picture is my own tightening of the screw — Feynman’s benchmark was the first semester.
Feynman distinguished between two kinds of knowledge: knowing the name of a thing, and knowing the thing itself. You can know the name of a bird in every language in the world and know nothing about the bird. You can recite the formula for gravitation flawlessly and still not understand what gravitation is.
* * *
I ran the test for the first time before a presentation. I was supposed to explain to the team how a new system worked. I’d read the documentation, studied the diagrams, followed the logic. I felt prepared.
Then I pictured a listener who knew nothing about the subject. Not a colleague, not an engineer — someone who only asked: what does this actually do?
I tried it. Out loud, alone in the office, door shut.
Two sentences. Then I stalled. The terms I was using, I couldn’t define myself. I talked around the essential thing. I didn’t know what the essential thing was.
Up until that moment I’d have told anyone I understood the subject. The feeling was there. The substance wasn’t.
* * *
Why does the exercise work?
Not because of the explaining itself, but because of the direction. When you read, recognition is doing the work: patterns surface, it’s cheap, it flows, it feels like understanding. When you explain, your brain has to produce — select, order, connect to what you already know. That’s more expensive, and it’s more honest.
Cognitive science has a name for it: the generation effect. Self-generated information is processed more deeply and remembered better than material read passively. Slamecka and Graf showed this in 1978 on simple word tasks, and it has been widely replicated since. Chi and colleagues found the jump to understanding in 1994 — eighth-graders who gave themselves an explanation after each line of text understood it more deeply than the control group, and most deeply of all the ones who explained the most.
At exactly the point where you stall, your feeling was lying. The phenomenon behind it I’ve called sham understanding elsewhere: the sense of having understood without having understood.
Not deliberately. Your brain isn’t built to tell you the truth about how it works; it’s built to keep you able to act. A person who constantly doubts their own gaps is less able to act than one who simply believes their explanations.
Writing it down helps more than thinking it through, incidentally. In your head you can slip past vague terms. On paper it shows.
* * *
Three objections that count.
The first: if I do this constantly I’ll never get anything done. Correct. So I don’t do it constantly, only at selected moments — before a decision with consequences, before a presentation, after an explanation I’ve received from a system. The point isn’t to doubt everything permanently, but to doubt in the places where it matters. And to know that those places are often not the ones you’d assume.
The second: the question is a diagnostic instrument, not a knowledge generator. If you know nothing about a subject, explaining won’t help — you’ll talk into the void. The generation effect presupposes that something is there in memory to activate. The question shows you where your understanding stops. It can’t produce knowledge that isn’t there. So: read first, then explain. Not the other way round.
The third, and to me the most important: an explanation you give yourself can turn just as smooth as one you receive. You explain something to yourself, it flows, you’re satisfied — and you’re wrong. The imagined ten-year-old always nods. A real person doesn’t. They know no more than you do; they simply stall at the place where you kept talking. The exercise gets sharper the moment someone is actually sitting across from you.
And it doesn’t apply everywhere. Asking yourself whether you could follow a recipe teaches you nothing. It’s a precision instrument for one kind of knowledge: causal relationships inside complex systems. Which is exactly what a machine hands you when you ask it how something works.
* * *
If you notice you can’t explain something — that the words run out, that you’re reaching for jargon, that you’re retreating into generalities — that’s no cause for shame. It’s a find, and a useful one. You’ve just seen something that was invisible a moment ago.
The alternative is to carry on, to nod, to feel informed. That’s easier. It’s also what our information environment systematically rewards.
Could you explain that right now to a ten-year-old?
If yes — good. If no — interesting. Then you know where the work starts.
* * *
The full essay on this topic:
