Even People With “Perfect Memory” Can Be Tricked Into Recalling Fake Events

Those who can remember what they ate on a day ten years ago can be fooled by tests that distort memories

When neurons misfire
When neurons misfire: Those who can remember what they ate for lunch on a day ten years ago can be fooled by tests that distort memories. Image via razorsmile

Among us mortals who forget where we just left our keys and what we ate for lunch yesterday, there are a handful of  remarkable people with hyperthymesia—also known as highly superior autobiographical memory (HSAM)—who can remember astounding details about every day of their life going back to childhood.

Given any particular date in the last few decades, Brad Williams can remember both what he did that day and what significant world events occurred. Bob Petrella can recall the date he met every one of his friends and acquaintances. Jill Price can remember the exact day that any episode of any TV show she’s ever seen first aired—including, in some cases, that day’s weather.

But some of the same scientists who first documented this condition in 2006 have some news that might bring the hyperthymestic back to Earth: In recent experiments, they showed that even these people with extraordinary memory can be tricked into recalling events that didn’t happen.

“Researchers have previously found that memories can be distorted in every group they’ve looked at: in the young, the old, those with high intelligence and those with low intelligence,” says Lawrence Patihis, a psychologist at UC Irvine and the lead author of the new study, published today in Proceedings of the National Academy of Sciences. “And when we looked at people with HSAM, in all measures, we found memory distortions too.”

To test whether these people could be fooled into recalling false or distorted memories, the researchers recruited 20 people with hyperthymesia and compared their performance in a series of standard memory implantation tests—which attempt to trick the participant into recalling false details or events that didn’t happen—to 38 people with normal memory. These sorts of tests have frequently been used to argue for reducing the reliance on witness testimony in the criminal justice system, as they show that people frequently misremember events and details because of leading questions.

In the first test, the participants were shown a series of 15 related words one by one a screen (light, shade, bulb, table and others, for instance). Then, afterward, they were quizzed on whether they’d seen specific words, including a key “lure” word that was related to the rest but not actually shown (in this example, “lamp”). Most non-hyperthymestic people taking the test would report seeing the lure word even though it never appeared, simply because they’d associate it with the words actually shown.

But what about hyperthymestic people? In the experiment, both the control group and the participants with hyperthymesia said they saw the lure 70 percent of the time.

The second test involved a slideshow of photographs that depicted a crime. For example, the pictures showed a man bumping into a woman, knocking everything out of her bag, and pretending to help her clean her things up while putting her wallet in his jacket pocket. Forty minutes later, the participants read a text narrative that described the same event, but introduced six specific errors—such as the man putting the wallet in his pants pocket instead of jacket.

Afterward, the participants were given multiple choice questions on what they saw in the original photos. Both groups made errors, swayed by the text they read after seeing the photos, but the people with hyperthymesia actually made about 73 percent more errors than the control group, perhaps indicating a heavier reliance of textual cues.

The results of the third test were perhaps the most jarring. Scattered among dozens of irrelevant facts about recent history that were merely intended to distract, the participants were specifically told that, on 9/11, someone had captured footage of United Flight 93 crashing in Pennsylvania. Then, on a written questionnaire, they were asked if they’d seen that “well-publicized footage.” Twenty-nine percent of the control group and 20 percent of the hyperthymestic participants reported that they’d seen the footage—an impossibility, because no such video exists.

For Patihis, the results of the experiments are unsurprising, and serve as a reminder that it’s foolish to rely on witness testimony to put someone behind bars, even if they claim to have remarkable memory. “This study is a rather simple way of showing—to lawyers, to the police, to juries—that all people are likely susceptible to misinformation,” he says.

But the study also serves as a broader window into the memory processes of people with hyperthymesia and how they might—or might not—differ from the rest of us. “It’s been a puzzle as to how anyone can remember so much information, so the advantage of these memory distortion techniques is that you can get a glimpse into how memory works,” Patihis says.

One of this study’s authors, Aurora LePort, has previously conducted neurological research into the same people with HSAM tested in this study, and found structural differences in their brains, including more robust white matter in areas specifically linked to autobiographical memory—that is, recall of events that happened to oneself—rather than areas tied to broader applications of memory. The fact that these people can’t perform any better on memory distortion tests similarly suggests a core difference between autobiographical memory (at which they obviously excel) and memory as a whole.

Get the latest Science stories in your inbox.