Your memory changes every time you use it

Most of us treat memory as storage. Something happened, it was recorded, and remembering means retrieving the recording. On that model a vivid, detailed memory is simply a good file, and a person who remembers confidently is a person who was paying attention.

Elizabeth Loftus has spent decades showing that this is not how it works.

"Our memories are constructive. They're reconstructive. Memory works a little bit more like a Wikipedia page: You can go in there and change it, but so can other people."

Her classic experiment with John Palmer in 1974 was almost absurdly simple. Participants watched a filmed car accident and were then asked how fast the cars were going. For some, the question used the word hit; for others, smashed. The single word changed the reported speed. A week later, the participants who had been asked about the cars smashing were more likely to remember broken glass at the scene. There had been no broken glass in the film.

The mechanism is now known as the misinformation effect. Details introduced after an event, by a question, a conversation, a news report or a suggestion, can be absorbed into the memory of that event and later recalled as part of the original experience. The person is not lying and cannot feel the seam.

Later work went further. In studies on suggested memories, a share of participants came to remember childhood episodes that their own families confirmed had never taken place, and they supplied details, emotions and context of their own. Suggestion did not merely edit a memory; it produced one.

The uncomfortable finding is not that memory is unreliable, but that we have no internal signal warning us which memories to distrust. Vividness, detail, emotional weight and confidence all feel like evidence of accuracy, and none of them are. This is why Loftus's research reshaped how courts think about eyewitness testimony: a witness can be entirely sincere, thoroughly convinced, and wrong.

It would be a misreading to conclude that memory is worthless. It is good enough for almost everything we ask of it, and reconstruction is part of what makes it useful, since we store meaning rather than raw footage. The problem appears only where precision matters and where the cost of an error is high.

That suggests a few habits. Write things down close to the event, before conversation has had time to reshape them. When something matters, look for independent corroboration rather than a stronger feeling of certainty. Be careful how questions are asked, because a question can plant what it appears to be asking about. And treat your own most rehearsed stories with mild suspicion, since each retelling is another chance to revise the source.

We do not remember events. We remember our last reconstruction of them, and then we build on that.

Source:

— Elizabeth Loftus