You were in the kitchen. You know you were in the kitchen, because you can still see it: the phone face-up on the counter, the slant of afternoon light, the exact words in the text. The day someone died, or the day they left, or the day you got in. Whatever your version is, it plays back like footage. You'd swear on it.
Here's the uncomfortable part, and the whole point of this essay: the vividness of that memory says almost nothing about its accuracy. Psychologists call these flashbulb memories, and decades of research show they drift and distort exactly like ordinary memories do. The only thing that stays frozen is your confidence. The feeling of certainty and the fact of accuracy turn out to be two different systems, and the first one keeps running long after the second has quietly failed.
The term comes from Roger Brown and James Kulik, who in 1977 studied how people remembered learning of JFK's assassination: where they were standing, what they were doing, who told them. More than a decade on, people still reported these moments in photographic detail, as if a flashbulb had gone off and exposed the whole scene onto film. Brown and Kulik noticed the memories only formed under certain conditions. The news had to be surprising, and it had to matter to you personally. In their data, about 90% of Black participants had a vivid memory of learning about Martin Luther King Jr.'s assassination, compared with roughly 75% of white participants. The closer the news cut, the deeper the stamp.
So far, so intuitive. Important things get remembered harder. The trouble started when researchers figured out how to actually check.
The handwriting that disagrees with you
The morning after the Challenger exploded in 1986, Ulric Neisser and Nicole Harsch handed Emory students a questionnaire: where were you, who were you with, how did you hear. Then they waited two and a half years and asked again. The average accuracy score was 2.95 out of 7. The average confidence score was 4.17 out of 5. Read those numbers again. The memories were mostly wrong, and the people holding them were mostly sure.
The detail I can't stop thinking about: when students were shown their original questionnaires, written in their own handwriting the day after the explosion, some were stunned to find the paper contradicting them. The new memory, the wrong one, felt truer than their own ink.
That's not a quirky lab anecdote. That's how your memory of the worst phone call of your life works too.
You might object that the Challenger was, for most students, somebody else's tragedy. Fair. So consider the study Jennifer Talarico and David Rubin ran at Duke. On September 12, 2001, they had students record their memory of hearing about the attacks and, crucially, a memory of some ordinary event from the same few days. Then they retested at 7, 42, and 224 days. Consistency declined at the same rate for both memories. The flashbulb memory was no more accurate than the memory of a random Tuesday. The difference was that confidence and vividness stayed sky-high only for 9/11. The ordinary memory faded and felt faded. The flashbulb memory faded and felt like HD.
And this isn't a short-term glitch that time corrects. A consortium led by William Hirst and Elizabeth Phelps followed roughly 3,000 Americans for ten years after 9/11 and found rapid forgetting within the first year, then a plateau, with confidence holding high across the entire decade. The darkest finding is buried in the details. When people got the facts of the event wrong (which plane, which tower), they tended to correct those errors over time, because the facts were everywhere. But when they got their personal story wrong, the version of where they were and who told them, the error tended to get repeated, not fixed. Every retelling entrenched it. Nobody fact-checks you on your own kitchen.
Why the feeling is so convincing
None of this means the emotion was fake. The charge on these memories is physiologically real: the amygdala tags emotionally arousing events, which is why they come back saturated when an ordinary Wednesday comes back gray. But that emotional stamp certifies importance, not accuracy. It makes you believe the memory harder. It doesn't make the memory better.
Being closer to an event does help, a little. In a study after the 1989 Loma Prieta earthquake, Californians who actually lived through it recalled it with near-perfect accuracy, and even Atlantans with family in California outperformed those without. Personal stakes raise the floor. But it's a matter of degree, nothing like the pristine photograph your mind insists it took. The flashbulb metaphor was always wrong. Memory isn't film. It's a story you re-draft every time you tell it, and the drafts overwrite each other without leaving tracked changes.
Which means something a little vertiginous: the version of the worst day of your life that you carry right now is not the day. It's the day plus every retelling since — smoothed, re-lit, bent into better narrative shape. The kitchen might have been the hallway. The text might have been a call. And you will never feel the difference, because the feeling is the one part your brain refuses to revise. It's a cousin of the thing that happens with grief, where the emotional processing arrives on its own schedule regardless of what the calendar says. Your mind files the moment first and does its real work on it later, quietly, in the dark.
I don't think the right response is despair, or some grim resolution to distrust everything you remember. Your flashbulb memories are still yours. They carry true information about what mattered to you and how much. They're just testimony, not footage.
But there is one practical conclusion, and the researchers basically handed it to us: the only versions of these memories anyone could verify were the ones written down close to the event. The students' day-after questionnaires were the ground truth. Their own handwriting was the only honest witness in the room, including against themselves, two years later.
You can give yourself that witness. Not for national tragedies — for your own. The week of the breakup. The night you got the call about your grandmother. If you put the moment into words while it's still raw and specifically itself, before memory sands it into a better story, you've kept something no amount of vividness can fake. It's part of why talking it out counts as journaling, and honestly part of why we built Joice: on the days when something enormous happens, nobody wants to compose paragraphs, but almost everyone can talk. The transcript that comes out of it is your day-after questionnaire, in your own words, waiting.
Here's the small challenge. Pick the flashbulb memory you're most certain of, the one you'd swear on, and write down or say out loud every detail you're sure about. Where you stood. Who spoke first. Date it. Then, the next time life detonates something near you, do it within a day, while the dust is still in the air.
Ten years from now, one of those two records will be a story.
The other one will be what happened.