Illusion of Explanatory Depth

People often feel they understand a mechanism in real detail until they’re asked to explain it step by step, at which point the understanding falls apart with no warning beforehand.

2 min read

What Is It?

The illusion of explanatory depth is the gap between how well people believe they understand a process or mechanism and how well they can actually explain it. The defining feature of the illusion isn’t poor explanation. It’s poor calibration: before attempting the explanation, people sincerely believe they understand the mechanism in much greater detail than they actually do. The act of explaining doesn’t merely reveal the gap to others, it reveals it to the explainer. Someone asked to simply rate their understanding of how a zipper works, or how a toilet flushes, will typically rate it high, confidently. Asked instead to explain the mechanism in writing, step by step, their own confidence drops sharply once they try, and stays lower even after the exercise ends.

The effect was documented by psychologists Leonid Rozenblit and Frank Keil in 2002, who found it held specifically for explanatory, mechanism-level knowledge, how and why something works, and much less for procedural knowledge (how to perform an action) and factual knowledge, where people’s self-ratings are reasonably well calibrated.

Why Does It Matter?

Strategic confidence inside organizations is often explanatory confidence in disguise. A leadership team can be genuinely certain it understands why a competitor is losing market share, why a process keeps failing in the same place, or why a customer segment is churning, and that certainty can be entirely sincere and entirely untested, because rating your own understanding and demonstrating it are two different acts, and only one of them exposes the gap.

Organizations also have a specific way of manufacturing the illusion: naming a phenomenon often creates the feeling of explaining it, even when no mechanism has actually been identified. “It’s a culture problem” or “people resisted the change” reads as an explanation. It’s a label, and a label is not a mechanism.

The illusion also explains a specific, recurring meeting failure: a plan gets approved on the strength of an explanation that sounded coherent, and only much later, usually after the plan fails, does anyone discover that no one in the room could actually walk through the mechanism in enough detail to have caught the flaw. A coherent story often survives group discussion much longer than a coherent mechanism does.

What Changes Once You See It?

You stop treating a confident, fluent explanation as evidence the explanation is correct. Fluency and accuracy are different things, and the illusion specifically trades on that gap.

You also start using a specific test before committing to a decision that depends on genuinely understanding why something happens: ask someone, out loud, to walk through the mechanism in detail, not just state the conclusion. The request is the test, because the illusion survives being stated but rarely survives being explained. You also become suspicious of explanations that stop at labels rather than mechanisms, since a name often feels like understanding long before it actually is.

Unlike many cognitive biases, this one is unusually responsive to self-testing. Most biases survive introspection; this one often collapses the moment someone honestly attempts a detailed explanation, which is what makes the diagnostic test above unusually powerful rather than merely well-intentioned.

Common Misunderstandings

It isn’t a claim that people are generally overconfident about everything. Rozenblit and Keil found people are reasonably well calibrated about procedural knowledge (how to ride a bike) and factual knowledge (a country’s capital). The illusion is specific to explanatory, mechanism-level understanding.
It isn’t solved by simply asking people to try harder to explain something before they act. The value is in the act of attempting a real explanation, not in encouraging more confidence about the eventual answer.
It doesn’t mean every confident explanation is wrong. Some are accurate and hold up under detailed scrutiny. The point is that confidence alone can’t tell you which kind you’re dealing with.
It isn’t a measure of intelligence. Experts are susceptible too, particularly outside their own domains or when relying on familiar conceptual labels instead of detailed causal understanding.

Diagnostic Question

If I asked you to explain, step by step, exactly why this is happening, would you still feel equally confident by the time you finished explaining it?

Explore Further

Field Notes

None yet.

Related Field Guide

Origin

Leonid Rozenblit and Frank Keil, “The Misunderstood Limits of Folk Science: An Illusion of Explanatory Depth” (2002), Cognitive Science.