IX Coach

The Curse of Knowledge: Why Experts Struggle to Teach

Why knowing too much makes you a worse teacher — and the practices that fix it

What is the curse of knowledge and how do you overcome it?

The curse of knowledge is a cognitive bias in which knowing something makes it genuinely hard to imagine not knowing it — so experts communicate at a level that loses their audience, without noticing they have done it. The term was coined by the economists Colin Camerer, George Loewenstein and Martin Weber in a 1989 paper on how better-informed agents fail to ignore what they know when predicting less-informed people, and it was carried to a general readership by Chip and Dan Heath in Made to Stick. The key point is that it is not arrogance or laziness: once information is encoded, you lose reliable access to the state of not having it, which is why the gap is invisible from the inside and obvious to everyone listening. The bias has been observed across several research literatures — hindsight bias, perspective-taking, and the tapper-listener studies — and the practical remedies are correspondingly indirect: rather than trying to remember ignorance, you get feedback from someone who still has it, or you constrain yourself with concreteness rules that make abstraction harder to hide behind.

Once you truly understand something, you lose reliable access to what it felt like not to know it. That loss is the curse of knowledge. It is not arrogance and it is not laziness — it appears to be a structural consequence of how expertise gets encoded, which is why it survives good intentions and why the people it afflicts most are usually the ones who care most about being understood. Camerer, Loewenstein and Weber gave it its name in 1989, studying how better-informed people fail to set aside what they know when predicting the less-informed; Chip and Dan Heath carried it to a wide readership in Made to Stick. Related findings converge from several directions — hindsight bias, perspective-taking research, and Elizabeth Newton’s tapper-listener study, where tappers expected half their listeners to name the tune and about one in forty did. The practices below are the working antidotes, and they share a shape worth noticing: none of them ask you to remember what ignorance felt like, because that is the one thing the bias makes unavailable. They rebuild the novice’s perspective from the outside in — through feedback, constraint, and concreteness.

Practices

Practice this with IX Coach

Practice this with IX Coach

Research

Related concepts

Frequently asked questions

What is the curse of knowledge and how do you overcome it?

The curse of knowledge is a cognitive bias in which knowing something makes it genuinely hard to imagine not knowing it — so experts communicate at a level that loses their audience, without noticing they have done it. The term was coined by the economists Colin Camerer, George Loewenstein and Martin Weber in a 1989 paper on how better-informed agents fail to ignore what they know when predicting less-informed people, and it was carried to a general readership by Chip and Dan Heath in Made to Stick. The key point is that it is not arrogance or laziness: once information is encoded, you lose reliable access to the state of not having it, which is why the gap is invisible from the inside and obvious to everyone listening. The bias has been observed across several research literatures — hindsight bias, perspective-taking, and the tapper-listener studies — and the practical remedies are correspondingly indirect: rather than trying to remember ignorance, you get feedback from someone who still has it, or you constrain yourself with concreteness rules that make abstraction harder to hide behind.

How do you take the tap test before you present?

The tap test (from Elizabeth Newton’s 1990 dissertation, popularized by the Heaths) makes the curse of knowledge visceral: tappers who know the song predict 50 % of listeners will recognize it; actual recognition is roughly 2.5 %. The experience forces an expert to feel the gap between their internal signal and what others receive. That felt gap is the calibration data experts normally lack.

How do you find a genuine novice to test your explanation on?

Imagining a novice perspective is unreliable because the expert’s knowledge is always available to contaminate the simulation. Using a real novice provides uncontaminated signal: their confused expressions, pauses, and questions map the exact gaps your mental model of them does not contain. This is deliberate error-seeking rather than comfort-seeking.

How do you open with a concrete story, not with your framework?

Abstractions require the listener to already hold the conceptual scaffolding needed to place the idea. A concrete story bypasses that requirement: it activates narrative processing, which is evolutionarily older and does not depend on pre-existing domain knowledge. Once the story creates a shared mental event, the abstract principle can be layered onto it rather than presented in a vacuum.

How do you run a jargon audit on every key explanation?

Jargon is invisible to its users because it has become the natural label for a concept they no longer have to decode. Experts stop hearing the words as words. A jargon audit forces an explicit pass through the explanation and treats each technical term as a potential break point — restoring the deliberate attention that fluency erased.

How do you ask what the audience already knows before explaining?

The curse of knowledge is partly a failure of audience modeling: experts construct a mental model of the listener that is systematically too rich. Asking directly — "What do you know about X?" — grounds the explanation in actual prior knowledge rather than assumed prior knowledge. It also activates what the audience does know, which reduces cognitive load during the explanation.

How do you build a bridging analogy from what they know to what they don’t?

An analogy transfers the relational structure of a familiar domain onto the unfamiliar one, allowing the learner to borrow inferences they have already built. For a cursed-by-knowledge expert, constructing the analogy forces them to identify what makes their concept structurally unique — which is exactly the distinction a novice needs. The failure mode is a surface analogy that imports the wrong inferences.

How do you work backwards from what the learner needs to do?

Experts organize knowledge by structure; novices need it organized by use. When an expert works backwards from a concrete outcome — "they need to be able to do X" — it filters the curriculum naturally: everything that does not serve X is cut, and everything that does serve it is prioritized. This reverses the curse’s natural direction of information flow.

How do you deliberately slow down at conceptual transitions?

Experts move between conceptual chunks at a speed calibrated by their own processing, not by the listener’s. Novices often lose the thread at transitions — the moment a sub-concept closes and a new one opens — because they have not yet consolidated what came before. Naming the transition aloud ("So, now that we have X, let’s see how it leads to Y") gives the novice a map signal and reduces the working memory cost of tracking structure alongside content.