I have a photographic memory, but once in a while I forget to take off the lens cap.
—Milton Berle
Do you remember the repressed-memory example discussed earlier? A policeman comes to your door, reads you your rights, and slaps on a pair of handcuffs. As you're carted off to jail, you learn that your twenty-eight-year-old daughter has accused you of molesting her when she was eight. Why does she believe it? She recently started therapy for some emotional issues, and the therapist thought that childhood abuse might be the cause of her current problems. Your daughter had no prior recollection of abuse, but when the therapist put her under hypnosis, she started to remember a number of vivid instances when you sexually molested her. The police were called in, and on the testimony of her twenty-year-old memory, you're sent to prison for thirty years—even though you know you didn't do it! It sounds crazy, but similar events have occurred in the United States. How can this happen? It all has to do with how our memory works.
IT'S THERE, I KNOW IT IS
Many of us think that our memory is a permanent store of past experiences. For example, which of the following two statements best reflects your view of how memory works?1
(1) Everything we learn is permanently stored in the mind, although sometimes particular details are not accessible. With hypnosis, or other special techniques, these inaccessible details can eventually be recovered.
(2) Some details that we learn may be permanently lost from memory. Such details would never be able to be recovered by hypnosis, or any other special technique, because these details are simply no longer there.
If you chose the first option, you're not alone. When psychologists asked people from various parts of the United States this question, approximately 75 percent selected the first description. We seem to think that our memories are literal snapshots of our experiences. Of course, we can't remember everything; in fact, we often complain about our memories. But when we say we have poor memories, we typically mean that we can't recall things at the present time. We think the memory is stored somewhere—we just can't bring it to mind right now. And, when we do recall something, and are confident in that recollection, we think the memory is quite accurate. But this is not how our memory operates.
As time passes, new experiences can change the memory we have for past experiences, without our even knowing it. In effect, our memory is a reconstruction of the past. Every time we recall a past event we reconstruct that memory, and with each successive reconstruction, our memory can get further and further from the truth. As Elizabeth Loftus and Katherine Ketcham state, our memory can be changed “by succeeding events, other people's recollections or suggestions, increased understanding, or a new context…truth and reality, when seen through the filter of our memory, are not objective facts, but subjective, interpretive realities.”2 As a result, our recall of the past is not fixed in stone. It is constantly changing—some memories are lost, while others are transformed.
To illustrate this point, two British psychologists secretly recorded a discussion that occurred at a Cambridge Psychological Society meeting. Two weeks later, the participants of the discussion were asked to write down everything they could remember. It turned out that they omitted about 90 percent of the specific points discussed, and when something was recalled, nearly half of the items were substantially incorrect. The individuals transformed off-the-cuff remarks into full-blown discussions, and remembered hearing comments that were never actually made.3
Even our memories of surprising and emotionally charged events can be in error. Do you remember how you heard about the space shuttle Challenger and Columbia disasters, or the collapse of the World Trade Center? The details of these flashbulb memories, as they are sometimes called, are often very vivid. We remember the place we heard the news, who told us, how we felt, and so on. Years after the event has occurred, these “facts” are ingrained in our memory. But are they always as accurate as we think they are?
Shortly after the space shuttle Challenger explosion in January 1986, researchers asked a number of students how they first heard the news.4 They asked those same students the identical question two and a half years later. Most of the students said their two-and-a-half-year-old memories were accurate—but not one of their memories was entirely correct, and more than a third were very inaccurate. In addition, the students were so confident in their memories that when they learned of their inaccuracies, they still didn't believe their revised memories were mistaken. In fact, they insisted their current memories were more accurate than what they said right after the Challenger exploded! As psychologist Ulric Neisser indicated, original memories are not just there—they are replaced with new, reconstructed realities.5
What about the recall of people who have photographic memories? Do you remember the Watergate scandal that occurred during the Nixon presidency? When White House counsel John Dean testified in front of the House committee investigating the scandal, he gave a number of highly detailed accounts of conversations with Nixon that seemed to be verbatim memories of what actually happened. At the time, people thought that Dean was amazing—he obviously had a photographic memory. But did he? Dean remembered the following from a meeting with Nixon and Robert Haldeman on September 15, 1973:
The President asked me to sit down. Both men appeared to be in very good spirits and my reception was very warm and cordial. The President then told me that Bob—referring to Haldeman—had kept him posted on my handling of the Watergate case. The President told me I had done a good job and he appreciated how difficult a task it had been and the President was pleased that the case had stopped with Liddy. I responded that I could not take credit because others had done much more difficult things than I had done. As the President discussed the present status of the situation, I told him that all I had been able to do was to contain the case and assist in keeping it out of the White House.6
Fortunately for us (not for Nixon), the President taped his conversations. When the tapes were played back, it was revealed that Nixon did not ask Dean to sit down, did not say Haldeman kept him posted, or that Dean did a good job, and did not mention Gordon Liddy. In effect, Dean remembered the gist of the conversation—that Nixon knew about the cover up—but a number of the specific details were changed or added. The bottom line is, our memory is not an exact copy of reality. We can forget some details that occurred and change others without even realizing it. Even more troubling, we may create entirely new memories that didn't actually happen, and these erroneous memories can lead to a number of severe consequences.
IF YOU SAY SO—THE POWER OF SUGGESTION
Consider the following true story. A young woman was sexually assaulted in 1987. Her assailant was caught and sentenced to eighteen months in prison, but the woman continued to be plagued by nightmares. To cope with her feelings of grief and rage, she sought the help of a therapist. In the midst of therapy, she began to believe that her parents sexually abused her as a child, and that her dreams were manifestations of those repressed memories. The woman told her sister and sister-in-law to keep their children away from their grandparents. A little worried, the sisters took their children to a therapist who specialized in childhood sexual abuse. While in therapy, one of the children started having nightmares of frightening creatures that she identified as her grandparents. The therapist diagnosed the children with post-traumatic stress disorder, supposedly caused by sexual abuse, and the grandparents were arrested.
During the trial, the children testified that their grandparents made them touch their genitals. One child also said her grandparents put her in a giant cage in the basement, and threatened to stab her mother in the heart if she told. Because of these memories, the grandparents were convicted of multiple counts of rape and indecent assault and battery. No physical evidence existed to corroborate any of the charges. However, the grandparents were sentenced to nine to fifteen years in prison because of memories that didn't exist until someone had a bad dream.7
In the 1980s and 1990s, a number of therapists were proclaiming that victims often repress childhood sexual abuse, and that these memories could be recalled by hypnosis and other suggestive techniques. These therapists believed that if a memory could not be recalled, the person must have repressed it to protect herself from an emotionally overwhelming event. They also thought that when an unpleasant memory is buried from a person's consciousness, the emotions attached to the memory can bubble up and cause havoc in her day-to-day life. To deal with these problems, the therapists thought the memories must be recovered.
A variety of methods have been used to recover these so-called lost memories. Therapists put people under hypnosis, asked them to visually imagine the event, and asked a number of suggestive and leading questions. They also had their clients read books on recovered memories, watch videotapes of talk shows on recovered memories, and participate in group counseling with others who supposedly had recovered memories. Clients usually had no memories of sexual abuse at the outset, but developed them after weeks and months of this therapy.8
These suggestive techniques have led many people to believe they were sexually abused in childhood. In fact, in 1988 Ellen Bass and Laura Davis published The Courage to Heal: A Guide for Women Survivors of Child Sexual Abuse, which sold 750,000 copies and started a recovered memory movement that involved dozens of books, talk show programs, and magazine articles. The problem seemed so pervasive that Bass and Davis estimated that as many as one third of all women were sexually abused as children.9 Surely some were abused, but do these estimates make sense? Are these recovered memories accurate? Many people think so and, in fact, a number of people are in jail because they were convicted on nothing more than a recovered memory.
But can false memories be created? A considerable amount of research indicates that memories can be created by the suggestions of others, especially when hypnosis and other suggestive techniques are used. For example, Martin Orne, one of the world's leading experts on hypnosis, put subjects in a hypnotic state after they were asleep for the entire previous night. While under hypnosis, he asked the subjects if they heard two loud noises during the night (the noises didn't actually occur). The subjects typically said that they heard the noise, awoke, and went to investigate what happened. If Orne asked when the noise occurred, they gave a specific time. Thus, Orne obtained very specific responses to events that didn't happen just by asking leading questions during hypnosis. And, when the subjects came out of their hypnotic state, they actually believed the events occurred. In essence, Orne's leading questions produced erroneous memories.10
In another series of studies, adults were put under hypnosis and told that they lived in an exotic culture, and were a different sex and race, in a previous life. A significant number of them actually developed “past life identities” that reflected the experimenter's suggestions. When other people were told that they were abused as children, they reported more abuse than people who were not given the suggestion.11 And so, false memories can easily be implanted with the use of hypnosis, leading questions, and other suggestive techniques.
These suggestive techniques are so powerful that the accused may actually start to believe they committed the crime. Consider, for example, another amazing case in which a young woman accused her father of sexually abusing her when she was a child.12 The woman gave a detailed story to police investigators, claiming that the abuse began in primary school, and that her father would have sex with either her or her sister. Her father had no memory of abusing his children. During his interrogation, however, detectives told him that he buried his memories because he couldn't face the fact of what he did to his own children. They offered him bits and pieces of information from his daughters' statements, hoping to stimulate his memory, and kept repeating three statements: (1) His daughters would not lie about something like this; (2) Sex offenders often repress their own crimes; and (3) If he admitted the allegations, his memories would return.13 After hours of interrogation, he began to remember events similar to those the detectives described. The visions were sometimes distant, but when he produced an image, the detectives (or attending therapist) would ask a leading question to bring them into focus. In the end, he confessed to sexually molesting his daughters on numerous occasions, and impregnating one of them when she was fifteen, later arranging for an abortion.
Over the next two months, the accusations grew from sexual abuse to satanic ritual abuse, alleging blood drinking, cannibalism, ritual abortions, sadistic torture, and the murder of twenty-five babies. The two girls then accused their mother and two others of belonging to a cult and having sex with the girls on numerous occasions. There was no physical evidence that the girls had been abused, sexually or otherwise. There also was no evidence of infant murders or animal mutilations, and the two daughters' memories often conflicted with one another.14
Because the girls' stories were becoming increasingly more bizarre, Richard Ofshe, an expert on cults and mind control from the University of California at Berkeley, was brought in to evaluate the case. Ofshe conducted a field experiment, telling the father, “I was talking to one of your sons and one of your daughters, and they told me about…a time when you made them have sex with each other while you watched. Do you remember that?” He did not. Why should he—Ofshe made up the story. However, Ofshe assured him it happened, and that both children remembered it. Ofshe said, “Try to think about the scene, try to see it happening.” In time, the father began to recover a “memory” of the event, and later handed Ofshe a three-page handwritten confession, complete with dialogue about the forced incestuous relation between his son and daughter. When Ofshe told him that he fabricated the entire story, the father became agitated and insisted that the images were real—as real as all of the other images he remembered.15
The court decreed a sentence of twenty years for six counts of rape. The other charges of satanic ritual abuse were dropped due to lack of evidence, as were the charges against his wife and two friends. At the sentencing hearing, the defendant denied ever sexually abusing his children, crimes he confessed to just one year earlier. Unfortunately, as Loftus and Ketcham state, “Confessions, unlike memories, do not fade with time. Tape recorded, signed, and sealed, they stay on the books, uncontaminated and intact, forever.”16
During the Salem witch trials of 1692, nineteen people were hanged, one was pressed to death, and hundreds were jailed. We like to think that things like that can't occur today, but unbelievable things are happening. We have our own brand of witch trials in the form of repressed memory cases. Even without physical evidence of abuse or other criminal behavior, people are sitting in jail cells because of recovered false memories. Fortunately, these witch hunts typically run their course. Science eventually weighs in and lawsuits are filed. For example, the Associated Press reported that a jury ordered therapists and an insurance company to pay $5.08 million to the family of a woman for making her falsely believe she was abused by her relatives.17 After such events, movements like recovered memory typically die out, but many people are left to pick up the pieces of their destroyed lives.
False Memories for the Rest of Us
Now you may say, “I agree, if you put a person through hypnosis you can create a false memory, but that doesn't happen in our everyday lives.” The fact is, we don't have to go through therapy, hypnosis, or serious interrogation to have memories implanted in our brains. False memories can be created by simple suggestions and leading questions. For example, studies have asked adults to remember events that supposedly occurred in their childhood, some of which were true (provided by a family member), and some false (fabricated by the researcher). False events included being lost in a shopping mall or staying overnight at a hospital for a possible ear infection. In these studies, people were typically asked to think about the events for a few days or write detailed descriptions of what happened. When interviewed days later, anywhere from 20 percent to 40 percent of people believed the false event actually occurred. In fact, false memories have even been created for traumatic events like serious animal attacks or accidents in about one third of people tested. And so, it's possible to implant entirely false memories in some individuals simply by asking them to remember the event, write it down, or continue to think about it.18
Memory reconstruction does not just apply to events that occurred in early childhood. We also reconstruct our recent experiences. To illustrate this fact, students were shown films of traffic accidents and then asked, “About how fast were the cars going when they ‘smashed’ each other?” Other students responded to the question with the verb smashed changed to either hit, collided, bumped, or contacted. Those who saw the word smashed estimated the automobile speed to be 40.8 mph, while those who saw contacted estimated the speed at 31.8 mph.19 So the mere suggestion of a more extreme verb led to increased speed estimates. But did that affect memory? In a follow-up study, students again viewed a car crash and were either asked, “About how fast were the cars going when they ‘smashed’ each other?” or “About how fast were the cars going when they ‘hit’ each other?” One week later, they were asked whether they saw any broken glass in the accident film (in fact, there was no broken glass). When the question contained the word smashed, 32 percent said there was broken glass, while only 14 percent of those who saw the word hit remembered the glass. Thus, some students who thought the collision was more severe actually reconstructed their memory to include broken glass.20
In another study, subjects saw a film of a car stopped at a stop sign. Some subjects were then asked whether a second car passed the first car when it was stopped at the stop sign, while other subjects responded to the question with the words stop sign changed to yield sign. When the question mentioned a stop sign, 79 percent of the subjects correctly identified the sign as a stop sign when asked later. However, when the question mentioned a yield sign, only 41 percent accurately identified the sign as a stop sign. A simple change in the wording of subsequent questions can create inaccurate memories.21
This reconstruction from suggestive questioning can even occur for real-life events. In 1992 an El Al cargo plane crashed just after takeoff, killing forty-three people. Researchers questioned 193 individuals about the crash, asking them whether they had seen the television film that captured the moment when the plane hit the apartment building. More than half of the people (107) reported seeing the film—but there was no film of the crash!22
MIXING THINGS UP—THE PROBLEM OF MISATTRIBUTION
I was having lunch with a number of colleagues one day when my friend Dick started to tell an amazing story about something that had happened to his wife. We were all laughing at the absurdity of the situation that his wife found herself in, when another colleague at the table said, “Didn't that just happen on The Simpsons last week?” It turned out that Dick mixed up what happened on the TV show The Simpsons with his wife's account of what happened to her during the day. While it may seem hard to believe, this is a very common memory error, called misattribution.23
We have a tendency to take past experiences and jumble them up. We attribute one person's comments to someone else, or think we did something at one time or place, when, in fact, it was actually at another time or place. This misattribution can contribute to the memory errors that we make when asked suggestive and leading questions. For example, it's possible that people recalled seeing the El Al plane crash on TV because they misattributed a film they saw for another crash to the El Al crash. People may have believed they were lost in a mall because they combined a genuine experience of being lost someplace with their actual memories of the mall in question.24
Former President Ronald Reagan had a habit of misattributing fiction to fact. When running for office in the early 1980s, he repeatedly told a story of a World War II bombing raid over Europe. After a B-1 bomber was hit by antiaircraft fire, the gunner cried out that he couldn't eject from his seat. To comfort him, the captain of the plane said, “Never mind, son, we'll ride it down together.” Reagan ended the story by noting that the commander was posthumously awarded the Congressional Medal of Honor for his heroism. A journalist curious about the story researched the incident and found no record of the award. However, he did find a scene in a 1944 movie called A Wing and a Prayer that sounded very similar to Reagan's account. In the movie, the captain of a Navy bomber rode the plane down with his wounded radioman, saying, “We'll take this ride together.” When the White House was questioned about the story Reagan presented as reality, a spokesman replied, “If you tell the same story five times, it's true.”25
Eyewitness Court Testimony
Misattribution can have severe consequences in many aspects of life. Consider the case of Timothy Hennis, a sergeant in the US Army. Hennis was convicted of murdering three people in July 1986, even though he had an airtight alibi at the time that the murders were committed.26 Why was he convicted? An eyewitness positively identified Hennis as the man walking down the victims' driveway about 3:30 am on the night of the murders. Another eyewitness remembered seeing Hennis use a bank ATM around the same time that someone with one of the victim's stolen ATM card withdrew money from her bank account.
There was a complete lack of physical evidence to convict Hennis—no fingerprints or hair samples matched him. In the opinion of the experts, the bloody footprints found in the house were made by a size 8½ to 9½ shoe, while Hennis wore a size 12. There were no traces of blood on his clothing, and no physical evidence in his car. In fact, an expert told the jury that there wasn't one piece of evidence that tied Hennis to the crime scene.
After two days of deliberation, the jury found Hennis guilty of murder and the judge sentenced him to death by lethal injection. The testimony of the two eyewitnesses sealed his fate. But did the eyewitnesses really see Hennis? Six months before the trial, one eyewitness admitted that he could have been mistaken in his identification. He even signed an affidavit to that effect. In fact, prior to seeing a photo lineup, he initially described the man as having brown hair, standing six feet tall, and weighing about 167 pounds. Hennis was blond, six-foot-four, and weighed 202 pounds. In addition, the other eyewitness initially told the police and lawyers that she didn't see anyone at the bank that day.
Why were these two people so positive at the trial when they pointed their finger at Hennis and said that he was the one? Were they lying? Not necessarily. After months of television and newspaper coverage reporting that they may have seen the murderer, they could have reconstructed their memories. In fact, the eyewitness at the ATM machine may have seen someone who looked similar to Hennis at another time, and then misattributed that recollection to her ATM memory. And when this reconstructed memory was rehearsed for police, she began to accept it as fact.27 In a similar fashion, the other eyewitness likely felt pressure to remember something for the police and lawyers, and after months of rehearsing an initially fuzzy memory, he could have firmly believed he saw Hennis walking down the driveway. Timothy Hennis was fortunate enough to receive a new trial and was found innocent due to a lack of physical evidence. Interestingly, while waiting on death row, he received several anonymous notes, thanking him for taking the rap and doing the time.
Pretty scary stuff! No one really knows how many people are sitting in a jail cell because of faulty eyewitness testimony. But consider the following. It's been estimated that every year in the United States there are over seventy-five thousand criminal trials that are decided on the basis of eyewitness testimony. In addition, a recent study analyzed forty cases where DNA evidence proved that the wrong person was imprisoned. Thirty-six of those cases, or 90 percent, involved erroneous eyewitness testimony.28
And yet, we place significant importance on eyewitness accounts. Elizabeth Loftus conducted a study where people, acting as jurors, heard a description of a robbery/murder, along with the prosecution and defense arguments. When the jurors heard only the circumstantial evidence in the case, 18 percent found the defendant guilty. However, when they heard the exact same evidence with one difference—the testimony from a single eyewitness—72 percent of the jurors found the defendant guilty. Such is the power of eyewitness testimony. As Loftus concluded, “Anyone in the world can be convicted of a crime he or she did not commit…based solely on the evidence of a witness that convinces a jury that his memory about what he saw is correct.”29
Why is eyewitness testimony so powerful? As noted, many of us tend to think that our memories are permanently recorded, like nonerasable computer disks or videotapes. But as we have seen, our memory is not a literal copy or snapshot of an event; rather, it's a fragmentary and often-distorted representation of reality.30 Unfortunately, we are particularly susceptible to misattribution errors in eyewitness accounts. Studies reveal, for example, that when people are shown pictures of two different faces, they later remember seeing a picture of a new face they had never seen before. Why? The new face had some of the characteristics of the two faces they did see. We make memory-conjunction errors, where we take attributes from different faces (e.g., eyes, nose, mouth) and combine them into a new face.31
In essence, we typically just get a general sense of familiarity of the facial features we see—which can be a recipe for disaster in eyewitness identification. Consider how police proceed in a criminal investigation. If you have some recollection of what a criminal looks like, you usually review a lineup or look through a set of photographs to make a positive ID. Psychologist Gary Wells has demonstrated that these common police procedures can actually promote misattribution, because witnesses are encouraged to rely on familiarity. Wells found that when witnesses see all the suspects and then have to identify the criminal (like in a lineup), they base their decision on relative judgments. That is, they pick the person who looks most like the suspect relative to the others in the lineup. The problem is, a witness will often select the person who looks most like the criminal even when the criminal isn't part of the lineup. One way to overcome this problem is to have witnesses make a “thumbs up or thumbs down” assessment on each suspect viewed individually. In fact, given these scientific findings, some police forces are incorporating such procedures in an effort to increase eyewitness accuracy.32
Eyewitness testimony is also powerful because eyewitnesses are often very confident in their identifications. As we've seen, however, confidence and accuracy do not necessarily coincide. In fact, confidence can be influenced by the mere suggestions of police and lawyers. For example, one study had subjects view a security video of a man entering a department store. They were told that the man murdered a security guard, and were asked to identify the person from a set of photos (the gunman was not in the set). Some of the people received confirming feedback—they were told they correctly identified the suspect. Others received disconfirming feedback or none at all. Those people receiving confirming feedback were more confident in their decisions, trusted their memories more, and said they actually had a better view of the gunman. Of course, they were wrong, but their confidence would play well in court. As psychologist Daniel Schacter says, “Eyewitness confidence bears at best a tenuous link to eyewitness accuracy: witnesses who are highly confident are frequently no more accurate than witnesses who express less confidence.”33
WHERE DOES THIS LEAVE US?
Like our perception of the external world, our memory of past events is constructive. Memories can be influenced by suggestive and leading questions, and we can mix up past experiences to create new, reconstructed memories. As with perception, the memories we retrieve can also be biased by what we want and expect to believe. For example, one study showed people a picture of a white man and a black man talking in the subway. The white man had a straight razor in his hand. When later asked to recall the picture, half of the subjects said the razor was in the black man's hand. An erroneous memory was created because of what this group of people expected to see.34 As psychologist Daniel Schacter states, “remembering the past is not merely a matter of activating or awakening a dormant trace or picture in the mind, but instead involves a far more complex interaction between the current environment, what one expects to remember, and what is retained from the past. Suggestive techniques tilt the balance among these contributors so that present influences play a much larger role in determining what is remembered than what actually happened in the past.”35
Of course, it's impossible to discuss the many different ways our memories can go wrong here.36 But I believe that the point is made. We can't just accept our memory for an event as reality. Even if we're very confident in a memory, we may still be very wrong. As with many of the topics we're exploring here, however, all is not bad. We often remember things quite well. In addition, as with decision heuristics, some of our memory problems are the result of rather useful strategies. If we remembered every detail of our past experiences, we would quickly reach information overload and have a difficult time functioning. For all its vices, our memory still allows us to function quite well. However, we have to be aware that our memories can be in error, and that those errors can have a significant influence on our beliefs and decisions.