Behavioral Science Dictionary

Flashbulb memory

Memory & Perception

A vivid, confident memory of where you were when shocking news broke.

What it means

A flashbulb memory is an unusually vivid, detailed and confidently held recollection of the circumstances of learning about a shocking public event. Brown and Kulik, who named it in 1977, proposed a special 'Now Print!' mechanism recording the scene wholesale. Test-retest studies since have generally found that such accounts drift and shed detail while confidence stays high and flat, and that dissociation between confidence and consistency is a consistently replicated finding. The one study comparing decline against an everyday memory found no difference, but was too small to rule one out. Whether these memories are otherwise special remains contested: in the one cohort followed for a decade, consistency fell and then flattened, and a 2024 review of fifty-seven such studies argues that conclusions about specialness can turn on which events, populations and consistency criteria a study uses. Nearly all of this work measures consistency against the rememberer's own earlier report, not accuracy, so an account can be wrong from the first telling and stay wrong.

The original demonstration

Roger Brown and James Kulik named the phenomenon in 1977 after asking eighty Americans, a sample deliberately split between Black and white respondents, what they remembered about learning of nine shocking public events, most of them assassinations, plus one personally shocking event that each respondent chose. Nearly all produced a circumstantial account of hearing that President Kennedy had been shot, converging on recurring slots: where they were, what they were doing, who told them, how they felt, what followed. Black respondents were far likelier to report such memories for the killings of Black public figures, which the authors read as evidence that personal consequentiality, not public magnitude, triggered the effect. To explain it they borrowed the neurophysiologist Robert Livingston's 'Now Print!' mechanism, which, once surprise and consequentiality cross a threshold, records the surrounding scene wholesale. The design could not test it: every account was collected once, years after the fact, with no independent record. The study established that people believe they hold unusually clear memories of such moments; it said nothing about whether they were right.

What the evidence shows

Testing the claim required catching an event as it happened and returning later. David Pillemer did that first: forty-four people questioned a month after the 1981 attempt on President Reagan's life, and again six and a half months on, gave mostly consistent accounts. Michael McCloskey, Cynthia Wible and Neal Cohen ran a similar design on the 1986 Challenger explosion, surveying people within a week and retesting fewer than thirty of them nine months later. Retention was high: later readings of their table put it at 89 percent or better for location, activity, source and the respondent's own reaction. But the accounts were not the permanent photographic records Now Print implied, and the authors argued on logical as much as empirical grounds that ordinary memory processes sufficed. Critics answered at once that the same data supported the original hypothesis. What changed the picture was the retention interval, not a repeat of the result. Ulric Neisser and Nicole Harsch collected their Challenger baseline the morning after the explosion and retested forty-four students two and a half years on. Average agreement with the original account came to under three points on a seven-point scale, a substantial minority matched on nothing at all, and confidence in the later account still averaged above four on a five-point scale. On 12 September 2001, fifty-four students recorded their memory of hearing about the attacks alongside one of a recent ordinary event, and were retested one, six or thirty-two weeks later. Consistency fell for both and no difference in decline was detected, though with roughly eighteen people per interval a modest difference could not have been ruled out. The separation was subjective: vividness, sense of reliving and belief in accuracy held up for the flashbulb memory and fell for the everyday one. The largest study began within a week of the same attacks with 3,246 respondents across seven United States locations, retesting at eleven, thirty-five and 119 months. By then only 202 had completed all four waves, the confidence measure rests on 92 of those, and attrition was selective: those who stayed were likelier to have suffered personal loss or inconvenience. Within that group, consistency ran near .63 at eleven months and .56 at thirty-five, then flattened, ticking up a non-significant .03 over the remaining seven years, while confidence held above four on a five-point scale. Inconsistent details were more often repeated than corrected, and none of five predictors, among them emotional intensity, media attention and frequency of discussion, forecast consistency, though the last two predicted accuracy about the public facts.

Why it happens

Two things need explaining, and they are not the same. The drift is unremarkable. Each retelling is an act of reconstruction, and these events are retold constantly, so details from later coverage, from others' versions and from one's own earlier narrations get absorbed and misattributed to the original moment. That account is imported from general work on source monitoring rather than demonstrated inside the flashbulb literature: the ten-year consortium measured media attention and amount of discussion, and neither predicted consistency, so either self-reported retelling is too crude a measure, or the mechanism does less work than the story implies. The persistence of confidence is harder. Confidence tracks the felt properties of retrieval, vividness, charge, the ease with which the scene arrives, all sustained by arousal and rehearsal. Fidelity is not observable from the inside, so the mind falls back on how the memory feels and reads that as evidence about the facts. None of this requires the strong photographic mechanism Brown and Kulik borrowed, though weaker claims about privileged encoding remain live: emotional arousal does influence consolidation, but generically.

Limits and caveats

Almost everything above measures consistency with the person's own earlier report, not accuracy: no external record exists of where someone was standing when the news arrived. Consistency bounds accuracy from above: an account can be wrong from the first telling and consistently wrong thereafter. Methods vary far more than the tidy summaries suggest. A 2024 review of fifty-seven test-retest studies found the events, populations, scored attributes and consistency criteria all varying widely, argued that conclusions about whether these memories are special can turn on those choices, and calls the debate unsettled. The strongest counter-result is not a null. A multinational study of a British prime minister's resignation, tested within a fortnight and again eleven months later, found close to 86 percent of its 215 United Kingdom respondents still meeting a strict consistency criterion, against under 29 percent of 154 respondents elsewhere, and read that gap as evidence that personal consequentiality affects encoding, not just rehearsal. Current models are correspondingly hybrid: emotional reaction, surprise and perceived importance predict whether such a memory forms, with rehearsal and social sharing maintaining it. What has not survived is the photographic version, not the claim that these events are encoded differently from a forgettable Tuesday. A meta-analysis pooling sixteen studies and roughly 1,900 participants reported a small-to-moderate age-related deficit in flashbulb memory scores, but that headline effect rests on fourteen studies after an outlier was dropped, and the consistency-over-time comparison on only seven, a thin base. Samples lean heavily on undergraduates in a few countries and a handful of heavily covered events.

Where it shows up

The practical exposure is anywhere an organization treats a confident, circumstantial recollection as a record. Incident investigations run months after the fact collect vivid testimony about who called whom and in what order, internally coherent and partly wrong. Hiring panels, disciplinary hearings and legal proceedings read certainty and specificity as signals of reliability; both are produced by processes that do not track truth. The corrective is structural. Contemporaneous records keep later recall from becoming load-bearing: timestamped logs and short same-day debriefs. Where recall must be used, early collection helps, with confidence recorded separately and read as how the memory feels, not how good it is. Repeated interviewing deserves caution, since inconsistent details tend to be repeated rather than repaired. One implication is easy to get backwards. A discrepancy between someone's account today and their account last year is weak evidence of dishonesty; it is what a normal memory does, and treating it as a credibility problem rewards whoever tells the most rehearsed story.

Examples

Many people are certain they recall exactly where they were on hearing of a national tragedy, yet their accounts shift over the years.

A regional utility runs a post-incident review six months after a major outage. Engineers describe with great specificity where they were standing and who called them when the first alarm sounded, but the on-call logs place two of those calls at times when no call was made.

An employee survey asks staff to recall the moment they learned a merger had been announced two years earlier. The narratives are rich and emotionally consistent, yet a meaningful share name the wrong channel, remembering an all-hands meeting where the internal timeline records an email.

A juror finds a witness credible because the account includes the weather, the clothing and the sound of a closing door, delivered without hesitation. The precision and the certainty are properties of how the memory feels on retrieval, not evidence that the details are correct.

Two managers who were in the same room when a closure was announced give accounts a year later that agree on the mood and diverge on nearly everything else: the day of the week, who spoke first, whether anyone walked out. Nothing was recorded, both are certain, and the review has no way to tell which version, if either, is nearer to what happened.

First described in Brown & Kulik (1977); critique by Neisser & Harsch (1992).

Key references

  1. Brown, R., & Kulik, J. (1977). Flashbulb memories. Cognition, 5(1), 73-99. doi.org/10.1016/0010-0277(77)90018-X
  2. Pillemer, D. B. (1984). Flashbulb memories of the assassination attempt on President Reagan. Cognition, 16(1), 63-80. doi.org/10.1016/0010-0277(84)90036-2
  3. McCloskey, M., Wible, C. G., & Cohen, N. J. (1988). Is there a special flashbulb-memory mechanism? Journal of Experimental Psychology: General, 117(2), 171-181. doi.org/10.1037/0096-3445.117.2.171
  4. Schmidt, S. R., & Bohannon, J. N. (1988). In defense of the flashbulb-memory hypothesis: A comment on McCloskey, Wible, and Cohen (1988). Journal of Experimental Psychology: General, 117(3), 332-335. doi.org/10.1037/0096-3445.117.3.332
  5. Pillemer, D. B. (1990). Clarifying flashbulb memory concept: Comment on McCloskey, Wible, and Cohen (1988). Journal of Experimental Psychology: General, 119(1), 92-96. doi.org/10.1037/0096-3445.119.1.92
  6. Neisser, U., & Harsch, N. (1992). Phantom flashbulbs: False recollections of hearing the news about Challenger. In E. Winograd & U. Neisser (Eds.), Affect and accuracy in recall: Studies of "flashbulb" memories (pp. 9-31). Cambridge University Press. doi.org/10.1017/CBO9780511664069.003
  7. Conway, M. A., Anderson, S. J., Larsen, S. F., Donnelly, C. M., McDaniel, M. A., McClelland, A. G. R., Rawles, R. E., & Logie, R. H. (1994). The formation of flashbulb memories. Memory & Cognition, 22(3), 326-343. doi.org/10.3758/BF03200860
  8. Finkenauer, C., Luminet, O., Gisle, L., El-Ahmadi, A., Van Der Linden, M., & Philippot, P. (1998). Flashbulb memories and the underlying mechanisms of their formation: Toward an emotional-integrative model. Memory & Cognition, 26(3), 516-531. doi.org/10.3758/BF03201160
  9. Talarico, J. M., & Rubin, D. C. (2003). Confidence, not consistency, characterizes flashbulb memories. Psychological Science, 14(5), 455-461. doi.org/10.1111/1467-9280.02453
  10. Hirst, W., Phelps, E. A., Buckner, R. L., Budson, A. E., Cuc, A., Gabrieli, J. D. E., Johnson, M. K., Lustig, C., Lyle, K. B., Mather, M., Meksin, R., Mitchell, K. J., Ochsner, K. N., Schacter, D. L., Simons, J. S., & Vaidya, C. J. (2009). Long-term memory for the terrorist attack of September 11: Flashbulb memories, event memories, and the factors that influence their retention. Journal of Experimental Psychology: General, 138(2), 161-176. doi.org/10.1037/a0015527
  11. Hirst, W., Phelps, E. A., Meksin, R., Vaidya, C. J., Johnson, M. K., Mitchell, K. J., Buckner, R. L., Budson, A. E., Gabrieli, J. D. E., Lustig, C., Mather, M., Ochsner, K. N., Schacter, D., Simons, J. S., Lyle, K. B., Cuc, A. F., & Olsson, A. (2015). A ten-year follow-up of a study of memory for the attack of September 11, 2001: Flashbulb memories and memories for flashbulb events. Journal of Experimental Psychology: General, 144(3), 604-623. doi.org/10.1037/xge0000055
  12. Kopp, S. J., Sockol, L. E., & Multhaup, K. S. (2020). Age-related differences in flashbulb memories: A meta-analysis. Psychology and Aging, 35(4), 459-472. doi.org/10.1037/pag0000467
  13. Frinco, R., Muzzulini, B., Tinti, C., Luminet, O., & Schmidt, S. (2024). A review of longitudinal studies on flashbulb memories. Where we started, are, and are going? Applied Cognitive Psychology, 38(4), e4233. doi.org/10.1002/acp.4233

Related terms

← All 1001 terms