You Must Remember This

At first, Chris had only the foggiest recollection of getting lost in the Spokane shopping mall at the age of 5. His older brother, a psychology student at the University of Washington, was doing an experiment on memory; would Chris, 14, read this brief account of an incident from his childhood, and write down everything he could remember about it? After reading about getting lost in the mall, Chris wrote variations on "I sort of remember the stores." But a few weeks later, pressed by his brother, he offered a richly detailed narrative: "I think I went over to look at the toy store, the Kay-Bee store and, uh, we got lost . . . And then this old man, I think he was wearing blue flannel, came up to me . . . he was kind of old. He was kind of bald on top . . . he had like a ring of gray hair . . ."

Chris had never been lost at a mall. But now he was sure that he had.

The fallibility, and malleability, of memory is old news to anyone who has ever had his vivid recall of a third-birthday party shattered by a grainy old home movie. Hundreds of experiments have shown that people easily slip false details (from a TV report, for example) into their recollection of a crime they witnessed. They even "remember" events they have only heard about: after a sniper fired into a Los Angeles school playground in 1984, one boy had vivid memories of seeing someone lying on the ground, hearing shots and running home. He had been on vacation at the time. "It is possible," concludes psychologist Elizabeth Loftus in a disturbing new book, "The Myth of Repressed Memory" (290 pages. St. Martin's. $22.95), "to create an entire memory for a traumatic event that never happened." And now, says Loftus, "some of the best minds in neuroscience are interested in how that can happen." If they succeed, they will have brought some much-needed science to the bitter and heated debate over "recovered memory" in cases of childhood sexual abuse, satanism and even UFO abductions.

Glimmers of how memories might be created out of mere wisps of suggestion began to emerge this summer, after a May conference at Harvard Medical School on the neurological bases of false memories. Start with how normal memory is thought to work: bits and pieces of an experience are parceled out to different regions of the brain. Memories of its sound settle in the auditory cortex, and memories of its appearance into the visual cortex (diagram). "Each neuron represents a little bit of the memory," explains psychologist James McClelland of the Center for the Neural Basis of Cognition in Pittsburgh. But all the scattered fragments remain physically linked. The job of assembling them falls to a part of the brain called the limbic system. Like a neural file clerk, it pulls disparate aspects of each memory from the separate file drawers scattered throughout the cortex, gathering them into a cohesive whole.

Whatever reason evolution had for creating this scheme, it practically guarantees people will "remember" things that never happened. For one thing, people routinely take perfectly accurate snippets and assemble them incorrectly. Suppose you saw an accident in which a car ran a red light; a year later someone asks, "Remember when that guy ran the stop sign?" Neurons connected to the memory of the accident click in -- but so do memories of stop signs. Time goes by. The next time you recall the accident, neurons holding memories of stop signs get activated, too -- and suddenly you remember a car running a stop sign. Most chilling, says psychologist Michael Nash of the University of Tennessee, is that "there may be no structural difference between" a memory of a true event and a false one.

False memories commonly arise in a condition called "source amnesia." Thanks to the brain's frontal lobes, most people can distinguish the memory of a dream from the memory of a real event: these gelatinous folds of gray matter tag each incoming image with how, when and where it was acquired. But if the frontal lobes are damaged, their owners cannot remember where a memory came from, explains psychologist Daniel Schacter of Harvard. The patients retrieve bits and pieces of a memory -- the face of an old teacher, a cinematic rape -- and cannot distinguish which fragment came from where. "You could be remembering a dream, a fear or something someone talked about," says Schacter. "What gives the memory a feeling of authenticity is that authentic parts are included." Even in people with perfectly fine frontal lobes, a memory's origin deteriorates more quickly than its other aspects, says psychologist Stephen Ceci of Cornell University. (Do you remember your first day at kindergarten? Or is it your mother's endless recounting that you recall?) Source memory is also highly prone to suggestion -- that the vision of a man by your bed is a real memory, for instance, and not an image drawn from hearing tales of abuse. "If you've imagined it enough and you lose the source of the information [your imagination]," says Ceci, "then you have a false memory."

Although not everyone is so suggestible -- 75 percent of the people in Loftus's lost-in-the-mall study refused to "remember" what had never been -- certain conditions can make them so. "Severe emotional stress overcomes internal checks on plausibility, and you are left with a false memory," says neurologist Marsel Mesulam of Northwestern University. From transcripts and session notes, Loftus, a psychologist at the University of Washington, found therapists who routinely told patients that if they cannot remember abuse, they should imagine it as an aid to memory. "Picture [your father] approaching the bed," one told her patient, Loftus reports. "Think about what happened in the bathtub," said another; "If you can't remember any details, just try to imagine what it must have been like." Many therapists also use hypnosis, which makes patients willing to internalize suggestions and to call any image they see a memory. "Under hypnosis," says psychiatrist David Spiegel of Stanford University, "you can create [memories] rather than retrieve them."

It was through hypnosis that a high priestess of Satan was born. "Anne," seeking help for postpartum depression, became convinced by her therapist that she had been repeatedly raped as a child by her father and his co-workers, and had been a priestess in a satanic cult that ritually murdered and cannibalized thousands of babies. She "had mastered the ability to immediately incorporate the content of [a] question into her memory," write Richard Ofshe, a social psychologist at the University of California, Berkeley, and Ethan Watters in their new book "Making Monsters: False Memories, Psychotherapy, and Sexual Hysteria" (340 pages. Scribner's. $23). Once Anne left the hothouse of therapy, she disavowed her accusations. Last year she filed suit against her therapist. More than 300 people have recanted their retrieved memories.

Despite assertions that traumatic events are repressed, strong emotions create strong memory. Among children who witnessed the murder of a parent, for instance, many distorted the memory -- but not one lost it entirely, say Ofshe and Watters. Experiments confirm only one way for extreme trauma to impair memory: by preventing it from forming. That which has never been stored cannot be retrieved.

Such skepticism about retrieved memories rankles people who counsel survivors of abuse. Harvard psychiatrist Judith Herman, for instance, says "scientists have no business using the term false memory. We have no way of judging independently [reports of] childhood experiences." That's OK if memories are used solely in therapy, and treated as mere expressions of the mind. But "memories" are sending people to jail: 700 civil and criminal cases have been filed based on retrieved memories of childhood abuse, according to the False Memory Syndrome Foundation. Some of the accused have been convicted and sentenced to 40 years, as in a case decided last week, based on nothing more substantial than retrieved memories. At least, the accusers thought they were memories.

Memories are stored, by category, in different parts of the brain. For instance, the memory of a single experience might be scattered in five places.

The brain can confound snippets of memory from a real event with snippets from an imagined one.

The mere suggestoin that you were once lost in mall can leave a memory trace in the brain.

The brain tags the idea of being lost as a sugestino, not a historic fact. It can then become linked to other aspects of being lost, from recollections of hansel and Gretel to a friend's tale of being lost, as well as to real memories of malls.

Under stress and over time, the knowledge that being lost in a mall was only a suggestion deteriorates. The memory of a real event, visiting a mall, becomes confounded with the suggestion that you were once lost in one.

If you are asked again whether you were ever lost in a mall, your brain will activate images of malls and those of being lost. The mind can even embellish the memory with snippets from actual events, such as people onece seen in a mall. Now you "remember" being lost in a mall as a child.