Published online 6 November 2007 | Nature | doi:10.1038/news.2007.220


False memories show up in the brain

Your brain can distinguish between real and fake memories, even if you can’t.

Tell the truth: our brain can sometimes reveal if our memories are real or false.stockbyte

It’s a common situation: you’re embroiled in an argument over a fact and you know for certain that you have the right answer. But when someone rushes to their laptop to google the correct answer, you discover that you were wrong.

Whether in a fight with a spouse or giving testimony on the witness stand, it is clear that our memories are not always trustworthy. Now, researchers have found that although those vivid false memories may seem indistinguishable from true memories to you, but they are sometimes processed by different parts of the brain1.

The results could one day be used to devise an early test for Alzheimer’s disease, or to assess the accuracy of witness testimony, says study author Roberto Cabeza, a neuroscientist at Duke University in Durham, North Carolina.

Cabeza and Hongkeun Kim of Daegu University in South Korea asked 11 people to read lists of words that fall into a certain category, such as ‘farm animals’. The subjects were later asked whether specific words had occurred on the original lists, while functional magnetic resonance imaging was used to measure the changes in blood flow to different areas of their brains. The participants were also asked to say how confident they were in their answers.

The researchers found that when study participants had confidence in their answer and they were correct, blood flow increased to the medial temporal lobes located near the dividing line between halves of the brain. This region of the brain contains the hippocampus, and is important for memory.

But when subjects had confidence in their answer but were wrong (which happened up to 20% of the time depending on the circumstances), the frontoparietal region lit up. Those regions are associated with what Cabeza calls a sense of “familiarity”.

Don’t I know you?

Familiarity is a general feeling that an event has happened in the past, even though you can’t recall the specific details, such as when you are sure you have seen someone before, but can’t remember her name or how you know her. “It’s basically an empty feeling,” says Cabeza. “In the case of true memories, you have real details to retrieve.”

Previous work with brain injury patients has shown that familiarity and recollection are separate things, which can be affected independently. True recollection declines as we age, while the capacity for remembering familiarity remains intact.

But this pattern does not hold true in Alzheimer’s patients, notes Cabeza. “In Alzheimer’s patients, we see a deficit in both kinds of memory,” says Cabeza. So he suggests that measures of brain activity connected to familiarity could be used as an early diagnosis for the disease.

Accuracy and confidence

Such brain scans may one day help to distinguish real memories from false ones on the witness stand, Cabeza says. Brain scans like these have previously been suggested for use in distinguishing people who know they are lying from those who know they are telling the truth. It could also help to distinguish accidental lies from the truth, this study suggests.


But the scans aren’t as clear cut as one might hope them to be. Another kind of memory, called ‘phantom recollection’, occurs when your brain provides you with false details appropriated from other memories: for example, if you saw a woman and not only felt certain that you had met her before, but also mistakenly ‘knew’ that her name is Sheila and she went to college with you. Phantom recollections occur less frequently than false memories, but they can activate the medial temporal lobe just as true memories do, says Cabeza.

For Valerie Reyna, a cognitive neuroscientist at Cornell University in Ithaca, New York, the results underscore the fact that judges and juries should not use a witness’s confidence in their own answers as a signal that the answers are more likely to be true. “It is really surprising, but there is a very weak relation between accuracy and confidence,” she says. 

  • References

    1. Kim, H. & Cabeza, R. Journal of Neuroscience doi: 10.1523/JNEUROSCI.3408-07.2007 (2007).
Commenting is now closed.