How people remember things that never happened
By William J. Cromie
The woman remembered that a man who looked exactly like Donald Thompson had raped her. But the psychologist, who, ironically, studies memory distortion, couldn't remember ever meeting the woman.
One of them was having bad memories.
The dumbfounded Thompson told police that, just before the rape, he did a live television interview, ironically again, on how people can improve their memory of faces. During questioning, the woman admitted she had watched the program. Apparently, she confused her memory of Thompson with that of the rapist.
Psychology Professor Daniel Schacter has been testing the brains of people like Thompson's accuser to determine what goes on when they have such errant recollections. In the process, he and his colleagues recently took the first pictures of brain activity during true and false memories.
"Memory isn't a videotape," he says. "Rather, it's a reconstruction using bits of sound, sights, words, and even tastes stored in different parts of the brain. Gaps in such reproductions, filled by imagination, cause error and distortions in eyewitness recollections and other aspects of everyday memory."
Part of the reconstruction takes place in the middle of the brain, between the temples. When people try to recall words memorized during laboratory experiments, this region, dominated by a curved, seed-size structure known as the hippocampus, shows increased blood flow and oxygen use. When a memory of spoken words rings true, both the hippocampal region and the area where auditory memories are processed "light up" on brain scans.
If the memory of a word is false in these experiments, the sound-sensing area stays mute, because no sound for that word was originally recorded. However, parts of the front of the brain, where efforts to recover memories are centered, show increased activity.
"Our theory is that the hippocampal region tells the frontal lobes that there's something familiar out there," Schacter explains. "The latter may then search around until it finds something that fits the recollection, whether or not it's correct."
Evidence for such a scenario comes from Schacter's experiments with amnesiacs who suffered damage to the hippocampal, or medial temporal, region. "These patients show reduced levels of both false and true memories," he explains. "They do not retain in their consciousness information recently given to them, say a list of words. Their frontal lobes do not get tricked into making associations that don't exist and so into building false memories."
When the first images of false memories made by Schacter and his colleagues were published, it led to speculation that they might be used to detect false memories of rape, abuse, and murder. George Franklin, for example, spent six years in prison after his daughter recovered a memory, apparently false, that he murdered her playmate in 1969.
However, "the idea of using this technique as a lie detector is preposterous," Schacter protests. His group did brain scans of students who listened to lists of related words, then were asked later whether or not certain words were on the lists. They heard "candy," "sugar," "chocolate," "sour," "bitter" and other words related in meaning. Afterward, they had to recall, if "sweet" was on this list.
About 60 percent of them remembered hearing "sweet." These students showed increased brain activity in the hippocampal area and frontal lobes. Those with true memories of words actually on the lists showed increased activity in the region behind the left ear (left superior temporal cortex), where memories of sounds are processed.
"I would be shocked," Schacter admits, "if this one area always lit up more when a memory is true rather than false. That's too simplistic. We must be very careful to separate laboratory experiments using lists of words from memories of real events. In addition to sounds, such memories involve fears, expectations, desires, and other emotions."
Additional obstacles exist. Schacter did the brain scans at the Good Samaritan Regional Medical Center in Phoenix, Arizona, using a technology known as positron emission tomography (PET). Such scans cost about $2,000 each.
But a greater limitation is that the machine reveals no absolute differences in brain activity, only comparisons. It shows what the brains of 12 students look like when they have true or false memories. It will not tell you whether the experiences of someone who claims she has been abducted by aliens are real or imagined. Nor will it determine the veracity of recovered memories of rape and murder without additional collaboration.
Such "memories" may be rehearsed over and over in the minds of victims until they are familiar enough to light up areas of the brain associated with real sights and sounds. "We have nothing with which to compare such activity," Schacter says. "The imagination can be as powerful as true memory."
In a recently published and highly readable book, Searching for Memory (Basic Books), Schacter covers various aspects of how people make, lose, and distort their memories. In it, he relates a number of cases of false memories formed with the help of psychotherapy, hypnosis, guided imagery, or support groups. For example, with the help of a therapist, a young woman named Ann came to believe that she had suffered vicious ritualistic sexual abuse at the hands of her mother and father. It never happened.
Schacter also tells the story of Diana Halbrooks who, during psychotherapy, became convinced she had been raised in, and ritually abused by, a satanic cult. She believed that her baby sister had been sacrificed by the cult. Halbrooks no longer believes those things and is trying to repair the shocking tears in her family's life caused by these implanted memories.
Not all recovered memories are false, of course. Schacter describes the case of a college professor who, with the help of therapy, remembered, then confirmed, that a camp counselor had molested him as a child.
Schacter concludes, "There is no question that some survivors of childhood sexual abuse forget about single abusive incidents, and some evidence that they may forget multiple episodes of abuse. But there is as yet little or no scientifically credible evidence that people who have suffered years of violent or horrific abuse . . . can indefinitely forget about the abuse."
Aging and False Memories
The amazing brain pictures of false memories will not change this situation in the foreseeable future. The only thing that will, Schacter says, is "careful research into what the brain does when it records and retrieves memories."
"We have to start off with tractable experiments before we can go for memories of whole events," he continues. In the previously mentioned tests, done with Eric Reiman of the University of Arizona, students heard words that they had or had not heard before. True memories produced activity in brain areas concerned with speech.
"The next step would be to do the same experiment with words that people read to see if areas that process visual input light up," Schacter explains. "We also want to add tests of physically similar words. If subjects read or see 'lake,' 'sake,' 'make,' and 'fake,' will they have false memories for 'snake'?"
Schacter also will try to prove his theory that people "buy into a false memory if it is associated with a true past experience or knowledge." One way of doing this would be to probe the brains of older people, who usually find it more difficult than the young to recognize words from previously studied lists. This happens even if their hippocampal regions exhibit normal activity.
Based on studies done in association with colleagues at Massachusetts General Hospital, Schacter concludes this poor memory stems from aging frontal lobes that can't put forth the effort needed to recognize the words. He predicts that the combination of normal hippocampal and low frontal-lobe activity will produce more false memories as people age.
"We might see no extra frontal activity for words that are false than for those that are true," he says. "That would mean less ability to think hard about whether or not a word that rings familiar was really on a list."
Overall, our memories are reliable. "Remembering a word like 'sweet' when it didn't occur on a list of associated words is not a complete illusion," Schacter argues. "On a certain level, it represents the gist of a list of words that include 'sugar,' 'chocolate,' and 'cake.' "
Getting the gist but not the word shows that memories aren't black and white. There are many shades of gray introduced by previous knowledge, the circumstances under which we record and retrieve a memory, and emotions such as fear and desire. "Our research goal," Schacter says, "is not just to separate legal fact from fantasy, but to uncover why people have false or distorted memories in the first place."
In Searching for Memory, he cites the hypothetical example of a woman who "was emotionally brutalized by a neglectful parent . . . and then remembers incest when none occurred. The memory is illusory . . . but it may capture something important about the past that should not be dismissed. Contrary to what some have said, there is a middle ground in the recovered-memories debate; the problem is to identify it. I believe that this is our best hope for resolving the bitter and divisive arguments that continue to rage among patients, families, and professionals."
Copyright 1998 President and Fellows of Harvard College