Why the person who explains best understands least

Ask two people how a zip works. One gives a confident thirty-second account with a diagram. The other hesitates, mentions the slider, stops, and admits they are not sure what forces the teeth together. The second person has just discovered something the first has not.

What everyone sees

Explanatory fluency is read as a proxy for depth, and it is a reasonable proxy in domains where the speaker has been tested. Untested, it inverts. Producing a smooth account requires selecting a level of description at which everything connects, and the selection is easiest for someone who does not know where the joins are.

What is actually happening

Rozenblit and Keil demonstrated an illusion of explanatory depth: participants rated their understanding of everyday mechanisms highly, then rated it much lower after attempting to produce a detailed causal account, with the drop specific to explanatory knowledge rather than to facts or procedures. Fernbach, Rogers, Fox and Sloman applied the same test to political positions and found that asking people to explain mechanisms — rather than to give reasons — reduced the extremity of their stated views and increased willingness to consider opposing arguments. The asymmetry is that attempting the explanation is what reveals the gap, so anyone who has not attempted it retains an unpunctured estimate, and confidence survives precisely where it has not been tested.

Why it stays hidden

The inversion hides because audiences reward fluency and cannot audit it. A confident account is easier to follow, easier to repeat and more satisfying to hear, and hesitation is read as ignorance rather than as the sound of a model being checked. So the speaker with the accurate self-assessment is the one who appears less competent.