On January 28, 1986, the space shuttle Challenger broke apart 73 seconds after launch, and much of the United States watched it happen live. The next morning, a psychology class at Cornell was handed a questionnaire: write down how you heard the news, where you were, what you were doing, who told you, and how you felt. One hundred and six students filled it out, most of them within 24 hours of the explosion. Two and a half years later, 44 of those students answered the same questions again. The two sets of answers barely matched each other.
🧠 The Questionnaire They Forgot Filling Out
Ulric Neisser and Nicole Harsch published the results as Phantom Flashbulbs: False Recollections of Hearing the News About Challenger. On seven specific details, the returning students scored an average of 2.95 out of 7. Eleven of them scored zero. Twenty-two scored two or less. Only 11 of the 44 even remembered having filled out the first questionnaire.
Then came the part that made the study famous. The researchers interviewed the participants and showed them their own handwriting from two and a half years earlier. One participant studied her original page and said: “I know that’s my handwriting, but I couldn’t possibly have written that.”
And the confidence? On a scale of 1 to 5, the average rating the students gave their own accuracy was 4.17. They were not being arrogant. They were reporting what the memory felt like from the inside.
🤔 Clear Is Not the Same as Correct
The term “flashbulb memory” was coined by Roger Brown and James Kulik in 1977, and the metaphor did a lot of work. A flashbulb fires once, floods the whole room with light, and whatever it illuminates gets burned onto the film permanently. Brown and Kulik proposed exactly that: a special biological mechanism that, once an event crosses a threshold of surprise and consequence, saves the surrounding circumstances forever.
The data came back less cinematic. When Talarico and Rubin started collecting memories on September 12, 2001, one day after the World Trade Center attack, and retested 54 Duke students at one week, six weeks, and thirty-two weeks, consistency fell for those memories at roughly the same rate as an ordinary everyday event they had also asked about. What did not fall was vividness, the sense of reliving, and belief in the accuracy. Their conclusion became the line the field now quotes: flashbulb memories are not special in their accuracy, only in their perceived accuracy.
The mechanism makes sense once you stop picturing film. A fact you study gets retrieved, checked, corrected, retrieved again. A day you lived gets told to other people, and every telling is a fresh pass over the same page. There is no second exposure to the original, so errors that enter early never get caught. Worse, the events we most want to “never forget” are the ones we retell the most, and rehearsal is exactly what hardens a mistake into a memory.
🔗 What It Means for You
The practical lesson is uncomfortable: the feeling of “I know this one” is a poor measure of whether you know it. Familiarity comes from retelling, and retelling is not retrieval. If you are studying for anything, the sense that a topic feels solid may only be measuring how many times you have read it.
The confidence is also the whole engine behind personalized readings, horoscopes, and profile reports. A vague line feels uncannily accurate on the way in, and then you repeat it to three people over the next week. Each repetition makes the fit feel tighter and the memory of being seen feel older than it is.
There is an advertising version too. Researchers have found that sharply differentiated brands can generate something that behaves like a flashbulb memory, scoring high on strength, sharpness, and vividness, while brands with muddled positioning do not. Distinctiveness does not just help a brand get remembered. It helps it get remembered with confidence.
And for anyone writing fiction: this is the most honest unreliable narrator material there is. A character who insists he remembers that night perfectly is not lying. He is describing his mind to you accurately and his evening inaccurately, which is the most human thing he could do.
🎲 The Two-Person Test
Today is September 11. The Duke study began on September 12.
Here is the exercise. Pick one public event you are certain you remember well. Write down five details: where you were, who was with you, how you found out, what you were wearing, what you were doing before. Then ask the person who was with you to write down their five, separately, before either of you compares notes.
The last detail worth keeping: the metaphor itself misled the field for decades. Brown and Kulik imagined a single exposed negative. What the research points to is closer to a manuscript that gets edited every time it is retold, with the original date still stamped in the top corner.