Ask anyone old enough where they were when they heard about a national tragedy and you will get an unusually detailed answer. The room, the light, who spoke, what was in their hands. The memory arrives whole and feels different in kind from remembering what you did last Tuesday — less like recall and more like playback.
Brown and Kulik named this flashbulb memory in 1977, and proposed that emotionally shocking, consequential news triggers a special mechanism that prints the surrounding scene with unusual completeness. The metaphor was photographic: a flash goes off, and the whole moment is captured.
The metaphor turned out to be wrong in the most useful possible way.
What the follow-up studies found
The test that broke it is simple: ask people about a shocking event immediately, then ask them again years later, and compare.
Neisser and Harsch did this with the 1986 Challenger disaster, collecting accounts the morning after and again about two and a half years later. The accounts frequently disagreed — sometimes about everything: where the person was, who told them, what they were doing. What did not waver was confidence. Students shown their own handwritten account from the day after occasionally insisted the earlier version was wrong.
Talarico and Rubin ran the cleaner version of the experiment after September 11th, comparing flashbulb memories against everyday memories from the same period. The result gave the field its enduring summary: consistency declined at roughly the same rate for both kinds of memory, but belief in accuracy stayed high only for the flashbulb ones. Hirst and colleagues' ten-year follow-up found the same shape at scale — rapid error in the first year, then relative stability of a partly incorrect account, with confidence remaining near ceiling throughout.
So the distinguishing feature of a flashbulb memory is not accuracy. It is certainty.
Why the feeling is so convincing
Several ordinary mechanisms combine to produce something that feels extraordinary:
- Emotional arousal sharpens the centre and blurs the edges. Strong emotion tends to improve memory for the core of an event while degrading peripheral detail, so the vivid bit really is better remembered — just not the surrounding scene you are most confident about.
- Rehearsal, repeatedly. You told the story ten times in the following week. Each retelling is a reconstruction, and details drift, harden and get borrowed from other people's versions. The story becomes fluent, and fluency reads as accuracy.
- Media contamination. Footage you saw later merges with what you witnessed. People frequently "remember" watching live coverage that did not exist yet.
- Vividness is mistaken for veracity. The brain has no reliable signal for truth, so it uses proxies: detail, ease, emotional charge. All three are high here, and none is diagnostic.
This is the same confidence-accuracy gap that makes the illusion of competence so hard to shake while studying, and it is why eyewitness confidence turned out to be a poor guide in the legal literature.
Memory is reconstruction, not retrieval
The deeper point is that flashbulb memories are not a broken special case. They are an unusually clear demonstration of how all remembering works.
Recall is not reading a stored file. It is rebuilding an event from fragments, plausible inference and later knowledge, every time — which is why each retrieval can subtly change the memory. See how memory works and memory consolidation.
Ordinary memories get corrected because we notice when they conflict with a diary, a photograph or someone else's version. Flashbulb memories rarely get corrected, because nobody writes down where they were standing when they heard the news, and everyone's account has already drifted. The error is invisible for lack of a record.
The practical lesson
If the most vivid memories you have are unreliable at roughly the rate of ordinary ones, the confidence you feel about anything you have not written down is not evidence.
This is an argument for records rather than for distrust. Not because your memory is unusually bad, but because everyone's works this way:
- Write it near the time. A contemporaneous note is the only version that has not been rewritten by retelling. This is the strongest case for the capture habit.
- Separate what happened from what you concluded. The two merge fast, and a year later the interpretation feels like observation.
- Date things. Memory for content survives far better than memory for when, and sequence is usually what a decision hinges on.
- Expect your own confidence to be uninformative. Certainty tells you a memory is fluent, not that it is true.
A written note does not degrade, does not get contaminated by a documentary you watched afterwards, and does not gain confidence with age. That is the whole reason external memory is worth the effort, and why writing things down changes what you can trust about your own past.
More in the doorway effect, context-dependent memory, and memory and focus.