Teaching as a Form of Revision

There is a counterintuitive claim at the center of this concept: the person who stands at the front of the room and explains things learns more from that experience than the people sitting in the chairs. If you accept this claim, it reshapes how you think about teaching, expertise, and the design of your own development.

The claim is not new. Aristotle noted that "teaching is the highest form of understanding." The Stoics built peer instruction into philosophical practice. The medieval apprenticeship system assumed that mastery required teaching. In the twentieth century, researchers confirmed it empirically. Edgar Dale's "cone of experience" framework, however imperfectly operationalized in some pop-education versions, captured a real finding: we retain and understand more of what we teach than what we read, hear, or observe. The "protege effect" — documented by researchers Alison Wood Brooks, Francesca Gino, and others — shows that people who believe they will have to teach material learn it more thoroughly and with better conceptual organization than people studying for personal mastery alone.

The mechanism is not mystical. It is structural. Here is what happens when you teach:

Explicitation of tacit knowledge. Most expertise is tacit — embedded in practice, not in propositional statements. You can do it, but you cannot fully say it. When you must teach it, you are forced to convert the tacit into the explicit. This conversion is demanding and often fails, which reveals where your knowledge is actually procedural and embodied versus where it is conceptually organized and transferable. Each failure of explicitation is a data point about what you do not yet fully understand.

Adversarial probing. A student's questions are not hostile, but they are adversarial in the productive sense — they probe the structure of your explanation from angles you would not choose. Because the student does not know what they do not know, they ask questions you didn't anticipate. These unanticipated questions often land in exactly the places your explanation has weak joints. You improvise an answer. The improvisation either works — in which case you've built new explicit knowledge — or it fails, leaving you with a visible gap to return to.

Structural reorganization. When you teach, you have to choose a sequence, a hierarchy of concepts, a frame. These choices are not neutral. They reveal how you have organized the knowledge internally. Often, in the act of teaching, you discover that your internal organization was poor — that you were holding things that should be connected as separate, or treating things as foundational that are actually derivative. The act of building a lesson plan is itself a cognitive audit.

Feedback on what was actually communicated. When a student demonstrates their understanding — by answering questions, attempting tasks, explaining things back — you discover the gap between what you said and what was understood. This gap is information about the quality of your explanation and often about gaps in your own clarity. The student who completely misunderstands a concept you thought you explained clearly is telling you something important about your model of communication and your assumptions about shared context.

The Revision Loop

Teaching is not a one-time revision event. It is a loop. You teach, you discover gaps, you revise your understanding, you teach again with improved knowledge, you discover new gaps. Each cycle through the loop leaves you with more accurately bounded knowledge — not just more knowledge, but knowledge you understand better.

This is why the most expert practitioners in many fields are not the ones who only practice. They are the ones who also teach. The surgeon who teaches surgery is exposed to the gaps in their procedural knowledge every time a resident asks why. The programmer who writes documentation is forced to see their code through the eyes of someone who has no implicit context. The musician who takes students discovers aspects of their playing they cannot name until a student's question makes it necessary.

The feedback loop also works in reverse, in a way that beginners' questions are often more valuable than experts' questions. Experts ask questions inside a shared framework; beginners ask questions that can crack the framework open. Feynman understood this, which is why he deliberately cultivated the ability to explain to non-physicists. Each explanation attempt was a self-audit. Each failure to be clear was a signal to revise his own understanding, not to dismiss the audience.

The Practical Architecture of Teaching as Revision

If you want to use teaching as a revision tool rather than just a giving-back activity, you need to design your teaching practice with specific elements:

Choose learners who will ask genuine questions. If everyone in the room already agrees with you and shares your assumptions, you won't get the adversarial probing that drives revision. The most valuable teaching for your own development happens with people who are genuinely naive about your subject — who will ask from real confusion rather than from shared frameworks.

Teach before you feel ready. Most people wait until they feel fully competent before offering to teach. This backward. Teaching while you are still actively learning something accelerates your learning because you encounter your gaps in real time, under the useful pressure of someone else's need to understand. The experience of not knowing while teaching is uncomfortable but extremely productive. It identifies exactly what to study next.

Treat student confusion as data, not failure. When a student consistently misunderstands the same point, the instinct is to think the student is the problem. Sometimes that's true. More often, the confusion is pointing at something genuinely ambiguous in your explanation — which points at something unresolved in your understanding. Following the confusion back is the revision work.

Write it out. Written explanations are a form of teaching, and they have a particular advantage: they are permanent, which means you can re-read them and notice the gaps. Tutorials, guides, articles, and even clear emails that explain your thinking are all teaching artifacts. They all force explicitation and they all reveal where your thinking is not yet clean. The discipline of writing for someone who doesn't share your context is one of the most rigorous cognitive audits available.

Revisit old material with new questions. After you have taught something multiple times, return to it and ask: what questions came up that I couldn't fully answer? Where did I find myself improvising explanations I wasn't confident in? What analogies didn't work? These are the revision agenda items. Teaching is not a one-time pass through material; it is a repeated exposure that should generate a running list of things to understand more deeply.

The Danger of Teaching Without Revision

There is a risk in teaching that must be named: it is possible to teach in a way that hardens incorrect understanding rather than revising it. If you teach from a position of performed authority — if you treat student confusion as the student's problem, if you never revisit what you've taught — you can become more confident in wrong beliefs, not less. The teacher who mistakes the performance of clarity for actual clarity is insulating themselves from the feedback that would improve them.

This is the phenomenon behind Dunning-Kruger at scale. Confident incompetence is partly generated by the social dynamics of teaching: when you stand in the role of expert, the social pressure discourages students from challenging you, which reduces the feedback that would reveal your gaps. The antidote is deliberate cultivation of challenge — actively inviting disagreement, treating correction as valuable rather than threatening, going back to sources when a question stumps you rather than improvising around it.

The revision-generating kind of teaching requires intellectual honesty. It requires saying "I don't know, let me find out" without embarrassment. It requires following student confusion rather than dismissing it. It requires treating your own explanations as hypotheses rather than deliverables — things to be tested against actual understanding, not things to be defended once offered.

The Deeper Point

The deepest reason teaching is a form of revision is that it requires you to hold your knowledge provisionally — as something being transmitted to a mind that does not yet have it. That posture of transmission, done honestly, is incompatible with the illusion of final mastery. You cannot truly try to transfer understanding to another person without noticing the places where your understanding is incomplete. And noticing is the beginning of every revision.

The teacher who learns is not a special type of teacher. It is the teacher who teaches honestly, which is every teacher who does it right.