The tapping experiment
In a study run at Stanford at the end of the 1980s, people were asked to tap out the rhythm of a well-known song on a table while a listener tried to name it. Before each attempt the tapper predicted how likely the listener was to get it. Tappers predicted around 50 per cent. The actual figure was about 2.5 per cent: three correct out of a hundred and twenty.
The tappers were not stupid or arrogant. They could hear the song while they tapped. It was playing in their heads, complete with melody and words, and they could not experience the tapping as the listener experienced it: a series of unrelated knocks.
This is the curse of knowledge, and the important part is that it is not a failure of effort. Knowing something changes what the signal sounds like from the inside, and no amount of trying harder restores the earlier state.
How it shows up in visual work
- A diagram that makes sense because you know what the boxes are.
- A video that opens on the detail, because the context is obvious to you.
- An abbreviation used on first appearance.
- A before-and-after where the difference is invisible to anybody who does not know what to look at.
- A chart whose axis label is a field term.
- A piece that answers the question the client asked you rather than the question the audience has.
The last one is the most expensive, and the hardest to see, because it is the whole brief being pitched at the wrong altitude.
Three fixes that work, in increasing cost
Write down what the reader already knows. Three bullet points, literally on the page, before you start: they know what a lathe is; they do not know what a tolerance is; they have never heard of this business. This converts an unanswerable question into a list you can check the draft against. It is crude and it catches most of the damage.
Define on first use, once, in passing. Not a glossary box, not a paragraph of background: a clause. “The tolerance, which is how much error the part is allowed, was a tenth of a millimetre.” A reader who knew skims past it in half a second; a reader who did not is now in the piece. The cost of including it is nearly zero and the cost of omitting it is the whole paragraph.
Show it to one person who does not know. Not a colleague, not another designer. Somebody from outside, given the piece with no preamble, then asked to say back what it was about. The gap between what they say and what you intended is a measurement, and it is the only one available. One person catches most of it; three catches nearly all.
The opposite failure exists, and it is worse socially
Over-explaining is also a failure, and it is more damaging to the relationship with the reader than under-explaining, because it carries an implication about them. A piece that defines terms its audience uses daily reads as condescension, and condescension makes people stop reading in a way that mild confusion does not: a confused reader might persist, an insulted one will not.
The resolution is not to aim for the middle. It is to be specific about who you are writing for, which is the brief again: one named reader, with a stated starting position. “Somebody who runs a small workshop and has never commissioned a film” tells you to define storyboard and not to define lathe. Without that sentence you are guessing in both directions at once.
The limitation
This cannot be fully fixed, only managed, and the reason is the mechanism itself. You cannot audit your own writing for the curse of knowledge, because the audit is performed by the same mind that has the knowledge. Everything you check will look clear to you, and it will look clear precisely because of what you know.
This is why the third fix (an actual person who does not know) is different in kind from the first two rather than merely better. The first two are you predicting somebody else’s state, which is the thing you are demonstrably bad at. The third is measurement.
It also means the problem gets worse over a project, not better. On day one you nearly remember not knowing. By week six you are as far inside it as your client is, which is exactly why the piece you show them will make sense to both of you and to nobody else.
Every lesson here is open. Your progress is recorded once you pass the Making Things with AI exam.
Check your understanding
Why is showing a draft to one uninformed outsider fundamentally different from carefully re-reading it yourself, rather than just more thorough?
Pick the one you would defend. Your answer stays private.
Stuck on something?Ask Meher, your AI tutor
