ELIZABETH F. LOFTUS has been researching how our memories work since the early 1970s. The professor of psychology and law now teaches at the University of California, Irvine. Quick to come from her lips is contempt for the analogy that human memory works like a computer hard disk, on which data are cleanly written and from which data are accurately read back. Loftus says our memories are routinely wrong.

Indeed, we all have forgotten where we placed our keys or blanked on a name. But that is not all. Our memories can change over time. In our mental images, we often paint ourselves in rosy colors and make the good old days nicer than they really were.

Our distortions might seem harmless except for one difficult fact: they can produce serious consequences for ourselves or others. In courtrooms, eyewitnesses who incorrectly recall the color of an accused perpetrator’s shirt can send an innocent person to prison for life. Loftus’s research has proved that eyewitness testimony can be flawed and that courtroom attorneys can influence a witness’s memory of events. Her results have altered legal proceedings. Most controversial is her assertion that adults and children who have long repressed harsh memories of being sexually abused yet suddenly recall events when pressed in counseling sessions, depositions or court cases may be unwittingly fabricating the scenes. This stance has brought the prominent cognitive psychologist a great deal of trouble, but as Loftus notes in the accompanying interview, she is convinced that poor practices by therapists are indeed contributing to false memories.

Interview by Andreas Krauss

MIND: How did you become interested in cases of sexual abuse?

Elizabeth Loftus: In the mid-1990s the repressed memory theory had become very popular among psychologists. Someone would know almost nothing about the fact that they were abused as a child. But the repressed experience, years later, could then cause psychological problems. Through psychotherapy it was possible—according to those who defended the repression hypothesis—to bring back the repressed memory and treat it. Based on this idea, the craziest stories were soon in circulation. Patients suddenly, and seriously, believed that they had been sexually abused in childhood. But they weren’t being cured. And when all this was happening, therapists could point to no scientific evidence that the repressed memory theory was valid.

MIND: How widespread did the theory become?

Loftus: Debra A. Poole of Central Michigan University carried out a study in 1995 and found that in the U.S. and the U.K. about a fourth of all therapists were using methods that could be characterized as dangerous: among them were hypnosis, dream interpretation, or direct demands on patients to imagine that they had been sexually abused as children. These methods are, in part, still popular today.

MIND: Why would psychotherapists want to encourage their patients to believe they had been victims of abuse?

Loftus: According to this therapeutic system, patients could not remember their traumas anymore. But as the psychologists and psychiatrists attempted to help bring memories back to life, there was an increasing chance that they would implant false memories in patients. Many therapists traced every mental problem back to sexual abuse.

MIND: Do you believe that traumatic events cannot be repressed?

Loftus: I am not saying that it does not happen. We just don’t have any reliable way to determine if it has in a given individual. As long as that is true, we should avoid theories that, so far, have done more harm than good.

MIND: You investigated one such case yourself.

Loftus: Yes. In 1997 psychiatrist David L. Corwin and his colleague Erna Olafson, both then at the University of Cincinnati, published an article about a girl they called Jane Doe. According to the authors, she had been sexually abused by her mother when a child. As evidence, Corwin provided a video shot from 1984 in which Jane, as a six-year-old, described her experience. When the psychiatrist reinterviewed her 10 years later, she initially denied the abuse but then remembered it again. For Corwin, this was proof that memories could be repressed.

MIND: You see it differently?

Loftus: When Jane was interviewed the first time, her divorced parents had already been fighting over custody of her for five years. During the legal proceedings, the father accused his former wife of having sexually abused Jane. Child Protective Services had come, at that time, to the conclusion that Jane had been neither sexually nor physically abused. In addition, the father had served time in jail [for failure to comply with visitation orders] and was therefore far from being as trustworthy as Corwin implied. A clinical psychologist had also investigated the situation and found that it was unclear whether there really had been abuse by the mother or whether it had emerged from the imagination of the father, who then may have fed it all to Jane, who made the allegations.

MIND: Perhaps the way Jane was interviewed affected her stories. Can suggestive questioning influence memories?

Loftus: The effects can be amazing. In research studies, we have been able to convince adult volunteers that they got desperately lost in a supermarket as a child, even though their parents tell us it never happened. Many of the subjects even invent details to make the story more dramatic.

MIND: Some colleagues say such experiments are not comparable to repressed memories of abuse.

Loftus: These people are trying to discredit my research because it undermines the foundations of their worldview. But our memories are malleable—whether we apparently got lost in a supermarket as a kid, were supposedly sexually abused or, as witnesses at a trial, wrongly remember the color of the getaway car.

MIND: Can eyewitness testimony still be credible?

Loftus: Yes, if factors that could confound it are kept out. For example, different witnesses to a crime should be questioned privately, and the questions should be open-ended, not directed at eliciting a certain answer. When witnesses speak, they should not be interrupted by the questioner. At suspect lineups, the eyewitness should not be told who the suspect is, nor should it be suggested. The fact that memories can be influenced is something that should be explained, repeatedly, to both judges and jurors.

MIND: Can real and false memories be distinguished later?

Loftus: Not reliably. Real memories are usually more detailed. But the more often false memories are explicitly formulated, the livelier they become, and they thus seem more credible.

MIND: What would you advise therapists—and patients—to do?

Loftus: A therapist should not start out with the assumption that repressed sexual abuse is the only possible explanation for psychological problems. He or she should consider other causes. And they should remain aware of the power of suggestion in their own actions. If I were a patient, I would hasten to get a second opinion if my therapist seemed to employ dubious methods.

MIND: Where is your research directed right now?

Loftus: False memories can be deliberately implanted. But do they have long-lasting influence on our behavior? We are looking into this question.