Curse of Knowledge

The curse of knowledge is the tendency for informed people to assume others share their background knowledge when communicating.

Example of Curse of Knowledge

  • A software engineer writes setup instructions that begin “just clone the repo and run the build script,” and is baffled when new hires take two days to get started. The engineer cannot imagine not knowing the dozens of implicit steps that have become automatic for her.
  • A professor races through a proof, saying “it clearly follows that…” while students stare blankly at the board. Years of familiarity make the inference feel self-evident, so the professor underestimates how much scaffolding beginners need.

Note

The curse of knowledge is related to but distinct from hindsight bias and the false consensus effect. Hindsight bias concerns overestimating how predictable an outcome was once you know it; false consensus concerns assuming others share your opinions and preferences. The curse of knowledge specifically involves being unable to set aside factual knowledge when modeling another person’s mind.

In more detail

Curse of Knowledge is a cognitive bias in which people who possess specialized knowledge find it extremely difficult to imagine what it is like to lack that knowledge, and therefore communicate as though their audience already understands. It matters because it quietly degrades teaching, writing, management, product design, and nearly every situation where an expert must explain something to a novice.

The bias arises because knowledge, once acquired, becomes automatic and feels obvious. Our minds have no reliable mechanism for “unlearning” on demand, so when we try to estimate what another person knows, we start from our own understanding and adjust insufficiently—a form of anchoring. Jargon, skipped steps, and unstated assumptions then feel perfectly clear to the speaker while leaving the listener lost.

A widely cited demonstration comes from Elizabeth Newton’s 1990 Stanford research, in which participants tapped out the rhythm of well-known songs such as Happy Birthday while listeners tried to name the tune. Tappers predicted that roughly half of their listeners would succeed; in fact only about 2.5 percent did. The tappers could hear the melody in their heads, and could not conceive that listeners heard only disconnected taps.

To counter the curse, test your explanation on an actual novice rather than imagining one, define terms the first time they appear, and ask listeners to paraphrase back what they understood. Concrete examples, analogies, and deliberately spelling out “obvious” steps help bridge the gap, as does remembering what confused you when you were first learning the material.