September 11 Reports Changed While Memories Felt Certain

Confidence did not track agreement

Reports of September 11 memories changed while confidence in those reports stayed high.

The surprise is not that memory shifted. It is that feeling sure gave no clear warning. People still called the attack memories vivid and trustworthy when parts no longer matched what they had said before.

Roger Brown and James Kulik called this kind of recall a flashbulb memory. The term covers memory for learning shocking news: where someone was, what they were doing, and how they reacted.

The name sounds like a camera flash. The studies found a different split. The memory could feel bright and whole while reported details changed.

Two memories began on the same day

Jennifer Talarico and David Rubin asked students about the attacks the next day. Each student also described a daily event from the same period. That gave the team two early accounts from each person.

Groups returned over the next few months. Both kinds of report matched the first account less well. The average drop was about the same for the attack and daily events.

The attack memories still felt different. Students rated them as more vivid and gave them more trust. Yet those strong feelings did not come with better agreement across reports.

William Hirst and a larger team followed September 11 accounts for a decade. Most of the change came early. The reports then became more stable, while average confidence stayed high.

Agreement is not the same as truth

The teams could compare a later report with an earlier one. They could not replay each person's day.

That limit changes the right words for the result. Agreement asks whether two reports match. Truth asks whether a detail fits what happened. The first report may also contain an error.

A mismatch proves only that the account changed. It cannot show which version was right. It also cannot name one cause for the change.

A mental photograph is the wrong test

Flashbulb memories stand out because of how they feel. They can seem rich, forceful, and close at hand. Those traits are real parts of recall, but they do not check each detail.

That gap matters whenever a confident account carries weight. A listener can hear certainty as proof that a memory stayed fixed. The September 11 studies show that confidence and report consistency can move on separate tracks.

An early account preserves something later confidence cannot: a record of what the person said then. Confidence tells how the memory feels now. Neither one, by itself, proves exactly what happened.