Why Do You Remember the Fact and Not Where You Got It? They Are Stored Separately
By the BrainSnail editorial team. How these articles are written and checked, and how to tell us when one is wrong.
People retain information reliably while forgetting where it came from, and that separation makes a claim heard from an unreliable source gradually indistinguishable from one that was checked.
What is being separated
Remembering something involves at least two different kinds of information, the content itself and the circumstances in which it was acquired, including who said it, where, when and how credible they seemed. Those are stored and retrieved by partly different systems, and the contextual part decays faster and is more fragile. The result is a familiar state in which somebody knows a fact confidently and cannot say how they know it, and a less obvious state in which they attribute it to the wrong source entirely, which is misattribution rather than simple forgetting.
Why it matters for what people believe
The consequences for judgement are direct:
- •A claim heard from a poor source loses its poor source over time
- •What remains is a familiar claim with no attached warning
- •Familiarity is itself used as a cue that something is true
- •So repetition from any source increases belief
- •Corrections can strengthen the original claim by repeating it
- •The effect operates without anybody being persuaded of anything
The problem with corrections
Attempting to correct a false claim runs directly into this mechanism and the research on how to do it well is substantial. Repeating the falsehood in order to deny it increases its familiarity, and the denial attached to it may be forgotten while the claim remains, so a correction can leave people more confident in the thing being corrected. The practical guidance that has emerged is to lead with the true statement rather than the false one, to repeat the falsehood as little as possible, to explain why the false claim was believed and what interest promoted it, and to supply a complete alternative account rather than simply removing a piece of somebody's understanding.
Why the system works this way
The separation looks like a design flaw and is better understood as a consequence of what memory is for. Storing every circumstance of every piece of learning would be enormously expensive and mostly useless, since the value of knowing that water boils at a particular temperature does not depend on remembering the lesson in which it was learned. Discarding the context and keeping the content is therefore efficient for almost everything people learn. The failure mode only appears in an environment saturated with claims of wildly differing reliability, which is a recent condition rather than the one the system developed in.
Where else it shows up
The separation explains several otherwise puzzling things. Unintentional plagiarism occurs when somebody recalls an idea without recalling that they read it, and the phenomenon is well documented in music and writing. Advertising works partly through it, since a brand name becomes familiar without the advertisement being remembered or believed. Eyewitness memory is corrupted when a detail learned after the event from another witness or from a questioner is later recalled as having been seen. And the effect is more pronounced in older adults and after sleep deprivation, which is why both are associated with increased susceptibility to misleading information.
The takeaway
Content and the circumstances of learning are stored partly separately, and the circumstances decay faster, so a claim from a poor source eventually becomes a familiar claim with no warning attached. Familiarity is then used as evidence of truth. Corrections that repeat the falsehood can strengthen it, so guidance is to lead with the truth and supply a complete alternative account.