Which was it? In the original version of a vignette, a robber hides behind a door. But some subjects remembered him hiding behind a tree.

Why Your Brain is Not a Camcorder

In TV courtroom dramas, an honest eyewitness sometimes changes his story--especially if an unscrupulous lawyer confronts him with a conflicting account of events. In fact, hundreds of psychological studies have documented that memories are surprisingly susceptible to misinformation. Now neuroscientists have gained new insights into what happens in the brain when memory goes awry.

In a study described this week in Learning & Memory, Craig Stark of Johns Hopkins University in Baltimore and his graduate student Yoko Okado recruited 20 men and women, who slid into a functional magnetic resonance imaging (fMRI) scanner. The volunteers watched eight vignettes, each consisting of 50 still images. In one, a man hides behind a door after snatching a woman's wallet. Then the subjects watched the same eight vignettes again, but this time the researchers changed a few key details, having the man hide behind a tree instead of a door, for example. Two days later the subjects returned to the lab and took a pencil and paper test that assessed their recollection of the original vignettes.

False memories cropped up in about a third of the answers, Stark says. Many subjects wrongly reported, for instance, that the man hid behind the tree in the original vignette--or in both versions. The fMRI scans hint at what happens in the brain when such false memories are created. When a subject viewing the original vignette showed increased activity in two subdivisions of the left hippocampus--a structure well-known to have a role in making memories--the subject tended to remember the original image correctly. But when activity was greater on the second "misinformation" trial, the subject tended to remember incorrectly.

These regions of the hippocampus probably encode details of the vignettes, Stark says, but that's not the whole story. Other parts of the hippocampus, as well as the prefrontal cortex, were less active during misinformation trials that led to false memories. Stark believes these areas record the source of explicit memories--information about how we know what we know. When they underperform, a subject may be more likely to report a false memory.

"It's a pretty neat study," says Ken Paller, a cognitive neuroscientist at Northwestern University in Evanston, Illinois. The findings suggest that the processes in the brain that created the false memories were the same as the ones that created the true ones, he says.

Related sites
The Learning & Memory paper
The Stark Lab
Paller's site