Concept · part of How people actually believe things

A stranger emails to say you have won a lottery you never entered. Nobody had to teach you to doubt it. Something in you checked the claim against what you know and the sender against who you trust, and both checks failed.

Epistemic vigilance is the name, from the research tradition of Dan Sperber, for that built-in screening. One paper the book uses calls it a “mental immune system.” Hugo Mercier later sharpened the idea into open vigilance. It is the strongest objection to any story in which people believe whatever a feed pushes at them.

The book accepts the objection and folds it in. Each person is a checkpoint deciding what to pass on, and vigilance is one of that checkpoint’s tests: part of how selection works, and the part most tied to truth. The book adds three limits. Vigilance guards what you believe and does little to stop a threatening headline from grabbing your attention. It competes with other pulls, such as identity and belonging, so people still share things they say they do not want spread. And it may be worn down: the book suggests, as a hypothesis and not a result, that years inside engagement-driven media train people’s habits, this one included.

Read the full argument

How it connects


How people actually believe things (1 of 7) · Open vigilance (Mercier’s paired construct) →

Explore all the ideas · Read the book in order · Open the interactive map