Why you cannot explain your own work simply
In 1990 a Stanford student had people tap out songs on a table. The tappers thought half their listeners would guess the tune. Three in a hundred and twenty did.
Because once you know something, you cannot reconstruct what it was like not to know it. Psychologists call this the curse of knowledge, and it is not a character flaw or a lack of empathy — it is a limit on introspection that applies to everyone, including the people who study it.
The practical consequence: your estimate of how clear you are being is unreliable in one direction. You will always think you are clearer than you are, and the more expert you are, the larger the error.
There is a particular kind of bad presentation that comes from the most competent person in the room. The work is excellent. The speaker knows the material completely. And the audience follows for about ninety seconds before quietly falling off.
It is worth understanding why that happens, because the usual diagnosis — "too technical", "knows too much detail" — is close but points at the wrong repair.
The tapping experiment
In 1990 Elizabeth Newton ran a study at Stanford that has since become the standard illustration. Participants were split into tappers and listeners. Each tapper was given a well-known tune — a national anthem, a nursery rhyme — and asked to tap the rhythm on a table. The listener tried to name it.
Before tapping, each tapper predicted how likely their listener was to get it. The average guess was around 50%.
The actual rate was 2.5%: three correct identifications out of a hundred and twenty.
The mechanism is the interesting part. While tapping, the tapper hears the full tune — melody, words, orchestration. It is not possible to tap "Happy Birthday" without hearing it. The listener has knocks on a table. The tapper knows this abstractly, and cannot use the knowledge, because the experience of not knowing is unavailable to them.
Why it gets worse with expertise
A second finding sharpens it. In 2002 Rozenblit and Keil asked people to rate how well they understood everyday objects — a zip, a flush toilet, a bicycle. Ratings were high. They were then asked to write a step-by-step explanation of how the thing worked.
Almost everyone failed, and when asked to rate their understanding again, they revised it sharply downward. Rozenblit and Keil called this the illusion of explanatory depth: familiarity with something reads, from the inside, as understanding of it.
Put the two findings together and you have the position an expert is actually in. Familiarity feels like clarity; and the clearer something feels, the less able you are to hear it as a stranger. Expertise does not just fail to help — it moves the error in the wrong direction.
Where this shows up in a real presentation
- Compression that is invisible to you. "We reweighted the sample" is a complete sentence to the person who did it and a closed door to everybody else. It is not jargon; the words are ordinary. The compression is what breaks it.
- Steps that were never in the room. You reached the conclusion over six weeks. In the retelling, four of the steps are so obvious to you that you skip them — and the conclusion arrives without visible support.
- Answering a question you were not asked. Someone asks something basic; you hear the sophisticated version of it, because that is the version that lives in your head. Your answer is correct and useless.
What does not fix it
Reading the deck again. You will read it exactly the way you wrote it, which is the problem. Introspection is the tool that is broken here, so more of it changes nothing.
Nor does "simplify". It is too vague to act on, and it pushes people toward removing detail rather than restoring the connective steps — which usually makes the argument harder to follow, not easier.
What does
Say it to someone who does not do your job
Not a colleague — a colleague shares the curse. A friend, a partner, someone in a different department. Sixty seconds, out loud, no slides. The exact moment their attention slips is the exact place the explanation broke, and you will feel it happen.
This is the highest-yield preparation step available and almost nobody does it, because it is uncomfortable in a way rehearsing alone is not.
Write the question before the answer
For each section, write the question that section answers, in the words the audience would use. If you cannot write the question, the section is there because the work happened, not because anyone needed it.
Stop trying to sound expert
In 2006 Daniel Oppenheimer tested this directly: the same content, written plainly or with unnecessarily complex vocabulary. Readers rated the authors of the simpler versions as more intelligent, not less. The complexity did not buy credibility. It cost it.
Which is a relief, because the simpler version is also the one that survives the curse. The two goals do not compete.
A limit on all of this. The curse of knowledge cannot be removed, only routed around. Nobody becomes able to hear their own material freshly; even researchers who have spent careers on this effect fall into it in their own lectures. Everything above is a workaround — borrowing somebody else's ears, because your own are compromised and will stay that way.
Sources
- Newton, E. L. (1990). The rocky road from actions to intentions. Doctoral dissertation, Stanford University.
- Rozenblit, L., & Keil, F. (2002). The misunderstood limits of folk science: an illusion of explanatory depth. Cognitive Science, 26(5), 521–562.
- Oppenheimer, D. M. (2006). Consequences of erudite vernacular utilized irrespective of necessity: problems with using long words needlessly. Applied Cognitive Psychology, 20(2), 139–156.
- Camerer, C., Loewenstein, G., & Weber, M. (1989). The curse of knowledge in economic settings: an experimental analysis. Journal of Political Economy, 97(5), 1232–1254.