Retrieval, not rereading, is what makes it stick

Highlight a chapter, reread it twice, and you will feel like you know it. Close the book and try to write down its argument from memory, and the gap shows up fast. Rereading produces fluency — the words look familiar — and fluency is easy to mistake for knowledge. The recognizing is doing the work, not the remembering.

One alternative is to retrieve: put the source away and pull the idea back out before you check. The reaching seems to be what lays down a durable trace. Henry Roediger and Jeffrey Karpicke documented this as the testing effect, or retrieval practice — students who self-tested remembered markedly more later than students who restudied the same material for the same time. The effort that makes recall feel worse may be what makes it work.

If that holds, it argues for building recall into learning instead of bolting a test on at the end. Try posing the question first, struggling a moment, then reading — and see whether it sticks. Because forgetting works against you, retrieval seems to pair with timing — Spacing the practice fights the forgetting curve — and to do best in a medium that prompts you on its own, since A page can remember what you learn.

See our memory systems guide. Want to test it? Next time you finish a page, close it and try to say what it claimed.