Ask someone where they were when they heard the news, and they won't hesitate. The kitchen, the sunlight on the counter, the exact song on the radio. They'll tell you who called, what they were wearing, how their stomach dropped. They won't say "I think." They'll say "I'll never forget it."

Then, sometimes, a sibling says: you weren't even home that day.

That small, vertiginous moment, when a memory you would have staked your name on turns out to be wrong, is one of the strangest things the mind can do to you. It also points to something that matters well beyond the news. The feeling of remembering and the fact of remembering are two separate systems, and they come apart far more often than we'd like.

The photograph that isn't

In 1977, psychologists Roger Brown and James Kulik gave these memories a name: flashbulb memories. They noticed that people seemed to hold startlingly detailed recollections of where they were when they learned of events like the assassination of John F. Kennedy. The metaphor was a camera flash, a sudden, total exposure that burns the scene in permanently.

The observation was real. These memories are special, though not in the way the name implies. They're special in how they feel: vivid, rich in sensory detail, accompanied by a sense of near-certainty, and returned to again and again.

The open question is whether they're special in being accurate. When researchers finally got a chance to test that, the answer was humbling.

The morning after the Challenger

The day after the Space Shuttle Challenger exploded in January 1986, psychologists Ulric Neisser and Nicole Harsch asked a group of students to write down exactly how they had heard the news: where they were, what they were doing, who told them, how they felt.

About two and a half years later, they asked the same students the same questions.

The results weren't subtle. Many of the later accounts didn't match the originals, and roughly a quarter of the students got nearly every detail wrong. Yet their confidence was high. When shown their own handwritten account from the day after, some students were disturbed to discover it contradicted what they now "clearly" remembered. One famously insisted that his current memory was the correct one, even while looking at his own handwriting.

That is the heart of it. Confidence did not track accuracy. The students weren't lying. They were sincerely, vividly, wrongly sure.

What actually lasts: the feeling, not the facts

A later study sharpened the picture. After September 11, 2001, Jennifer Talarico and David Rubin had students record their memory of hearing the news, along with a memory of an ordinary event from the same period. They retested at intervals of a week, six weeks, or 32 weeks.

The finding: the consistency of details decayed at about the same rate for the dramatic event and the mundane one. Over time, people forgot the specifics of 9/11 just as they forgot the specifics of an unremarkable Tuesday.

What differed was everything around the facts. For the flashbulb event, people's vividness, their sense of reliving it, and their belief in its accuracy stayed high. For the ordinary memory, those faded along with the details.

So the emotional event doesn't give you a better recording. It gives you a stronger conviction about a recording that is decaying like any other.

Why the feeling survives when the facts don't

Several well-understood mechanisms conspire here.

Emotion strengthens the core, not the periphery. Arousal engages the amygdala, which enhances consolidation of the central emotional gist. But that boost is selective. The gist gets stronger; the surrounding details, such as what you were wearing, don't get the same protection.

Rehearsal rewrites. Flashbulb events are talked about, watched on screens, and replayed for years. Each retelling is a retrieval, and retrieval is reconstructive: the memory is rebuilt from fragments and a bit of inference each time it is brought up. The details you borrowed from a news broadcast or a friend's version can quietly slide into your own account.

Vividness gets mistaken for truth. We use the richness of a memory as evidence that it's real. That's usually a decent heuristic, but a memory can be rich because you've rebuilt it many times, not because it is faithful.

A story needs a setting. The brain likes a coherent narrative. If you remember the event with certainty, your mind fills in a plausible where and who, and then you cannot tell the filled-in parts from the recorded parts. This is source monitoring failing quietly.

The practical lesson: your certainty is not a measurement

This is where the idea stops being about disasters and starts being about you.

The same gap between confidence and accuracy shows up in the memories that run your life: the argument you're sure you won, what someone promised you, what you read in that article, the textbook chapter you "know cold." The strong feeling of knowing is generated by familiarity and fluency, not by a check against reality.

That has a gentle, useful implication. You don't need to distrust your memory. You need to stop using how sure you feel as the test of whether you're right, and start using something outside your head.

Your next moves

  • Write down one significant moment today, while it is fresh. A hard conversation, a piece of news, a decision. Note the date, the place, who was there, and what was said, in plain sentences. In a year, you'll have what Neisser and Harsch's students had: a record that can disagree with your memory.
  • Before arguing about a past event, ask for the other person's version first. Two accounts that differ aren't proof someone is lying. Treat the difference as data about how reconstruction works, not as an accusation.
  • Check one "I'm certain" fact this week against a source. Pick a date, a quote, or a statistic you would have bet on, then look it up. Notice how it feels when you're wrong; the feeling is the lesson.
  • Separate "I remember it vividly" from "I remember it accurately" out loud. Try the phrase "I remember it this way" instead of "it happened like this." It costs nothing, and it makes room for correction.
  • For anything you need to know rather than just feel, test it cold. Close the notes and try to produce the answer from nothing. If you can't, the feeling of familiarity was misleading you, which is far better to discover now than during the exam or the meeting.

What this means for the things you want to keep

There's a tender side to all this. The memories we feel surest of are often the ones that matter most to us, and learning they may be partly reconstructed can feel like a loss. It doesn't have to. The emotional truth of those moments, what they meant, how they changed you, is intact. It's only the camera that's a myth.

Which is why, for the knowledge you actually need to hold on to, it helps to build a system that doesn't rely on feeling sure. That's the idea behind Recall: instead of trusting the sense that you've learned something, it asks you to retrieve it, then uses the FSRS spaced repetition model to schedule the next check just as the memory starts to fade, so your confidence is always earned by an actual answer. It works fully offline and imports your existing Anki and Quizlet decks, if you'd like to try it at https://recall.lumenlabs.works.

Either way, keep the notebook. The version of you who wrote it down will be glad of the evidence.