How Far Can You Trust Your Memory?

Prof. Daniel Reisberg’s expertise on perception and memory has impact far beyond the classroom.

It was dark outside when she was shot. Ten at night, roughly. The gunman stood outside her camper in Oregon’s Umpqua National Forest on that August night in 2003 and nailed her in the chest as she reached to close the window. Sherl Hilde, age 45, crumpled, unable to move her legs. A few moments later, the attacker shot her husband, Noris Hilde, as he called 9-1-1 on her cell phone.

Noris died, but Sherl survived. Bleeding profusely, she was flown by helicopter to the hospital, where she recuperated for a month. Meanwhile, local detectives developed a theory. The killer, they believed, was troubled 28-year-old Samuel Adam Lawson, who had recently stolen a .357 Marlin rifle from his father’s gun case. Lawson had met the Hildes hours before the killing, when the couple drove up to their reserved campsite, only to find it occupied—by Lawson, who gathered his gear and shuffled away.

Two years later, at a pretrial hearing, Sherl Hilde pointed at Samuel Lawson and proclaimed, “I’ll never forget his face as long as I live.”

Lawson was ultimately convicted of aggravated murder and is now serving a life sentence in the Oregon State Penitentiary. But there is a problem with this case. The most compelling evidence against Lawson lies in the memory of Sherl Hilde, and there is reason to think that her memory may be wrong.

Professor Daniel Reisberg [psychology 1986–] is an authority on cognition and on the complex ways in which people remember—and sometimes fail to remember—the details of their lives, including the traumatic moments they have experienced. The author of scores of articles and a dozen books, he is a leading expert on eyewitness IDs. He has testified for the defense in many trials in Oregon, Washington, and Idaho, some of them involving murder or sex abuse. At Lawson’s 2005 trial, he held the witness stand for a full day and declared, “I honestly cannot think of a single case I have ever seen or heard about that had this many concerns about one or another factor that could compromise somebody’s capacity to remember.”

The problems in the Lawson case began within minutes of the shooting. As the helicopter flew her to the hospital, Hilde was certain who the killer was. She fingered, of all people, the pilot. Initially, she told police that she never saw the shooter. Shown a photo throwdown, she did not pick out Lawson. Over time, however, as investigators attempted to reconstruct the shooting, her memory began to evolve. She recalled that the shooter had shoved a chair cushion over her eyes to prevent her from seeing him, but that she caught a glimpse of him in profile. After she had seen Lawson’s photo on half a dozen occasions, the memory of that glimpse became clearer and clearer, until finally she was certain that Lawson was the killer.

Professor Reisberg’s testimony didn’t persuade the jury, but Lawson’s attorneys [who include Laura Graser ’73, Bronson James ’94, and Bear Wilner-Nugent ’95] appealed his conviction to the Oregon Supreme Court, which will likely rule some time this year. Even if it refuses to grant a retrial, it could update the way Oregon courts handle eyewitness IDs garnered by investigators using suggestive tactics—in other words, by inadvertently leading the witness into fingering the prime suspect.

Oregon’s standard for reconstructed IDs was set in 1979, in State v. Classen, when the Oregon Supreme Court (in an opinion written by Justice Hans Linde ’47) ruled that such IDs could be shared with juries, but only if they stood a reasonable chance of being reliable. The court cited several factors that judges could weigh in allowing such IDs to be admitted as evidence—did the witness get a good view of the alleged criminal? Was he or she certain that the ID was accurate?

The Classen rules (and the federal rules on which they are based) were never perfect, but in recent decades grave doubts have begun to sprout around them. In fact, the New York–based Innocence Project reckons that eyewitness misidentification is the single greatest cause of wrongful convictions nationwide. Since 1989, nearly 300 U.S. prisoners have been exonerated by DNA evidence; about 75% of them had been mistakenly sent to prison because of faulty eyewitness evidence.

The 180-page transcript of Reisberg’s testimony explains why an honest eyewitness might make a faulty ID. Exuding at turns both the drama of Stephen King and the explanatory tone of a psychology textbook, Reisberg’s narration dwells, at first, on a white North Carolina woman, Jennifer Thompson, who was raped in 1984. Police suspected a black man, Ronald Cotton. Thompson picked Cotton out of a lineup, and the officer told her she’d done a “good job.” Cotton was duly convicted and imprisoned for 11 years—until DNA evidence conclusively proved him innocent. Thompson had identified the wrong man. Still, the effect of the cops’ certainty was so powerful, Reisberg told jurors, “that even after Jennifer was shown irrefutable [DNA] evidence that they had the wrong guy, she has testified that when she plays the videotape in her head of that awful night, the face she still sees is Cotton’s.” [Since then, Jennifer Thompson has become an outspoken advocate for eyewitness reform. Miraculously, she and Cotton have become friends, and have made several joint appearances. She no longer sees Cotton’s face in her mental video.]

Moving to the particulars of the Lawson case, Reisberg cast doubt on whether Sherl Hilde—shot in the dark, and then bleeding, traumatized, and heavily medicated—ever had the wherewithal to consolidate a vivid memory of the crime. He noted that, inside the brains of trauma victims, glucocorticoids work overtime. They manage stress but also, Reisberg noted, shut down memory consolidation. “It’s like never putting the book on the bookshelf,” Reisberg said. “You’re not going to be able to find it later.”

Reisberg also discussed a study in which researchers successfully planted memories in subjects by showing them a movie of a car driving through a quiet countryside devoid of manmade structures and then asking, “Can you tell me how fast was that car going when it went past the barn?” The subjects then remembered the fictional barn. Reisberg suggests that by repeatedly showing Hilde Lawson’s photograph, the police planted a memory in her mind. “Eventually,” he said, “you’re going to look at that photo and say, boy, that’s familiar.”

Reisberg was playing to a tough audience (rural, conservative Douglas County is among the most pro–death penalty jurisdictions in Oregon), but he is a nimble and animated performer—bouncy on his feet, his hands a flurry of emphatic gestures, and his voice rises to an excitable trill as he drives home his points. Students at Reed, where he’s been teaching since 1986, gave him an award last year for being the “most likely to jump on his desk.” In court, he explained the most abstruse psychological studies in everyday terms (by, for instance, telling a story about a bickering couple), and in the right moments he was not above humor. Consider this discussion of leading questions seven or so hours into Reisberg’s testimony:

Prosecutor: You want to go home pretty quick, right? Leading question?

Reisberg: “Pretty quick” isn’t essential, but I’d love to be in Portland before dark.

Judge: When does it get dark in Portland? And how fast do you drive?

Reisberg: Legally, Your Honor.

Judge: Of course.

Reisberg: That was a leading question.

When Reisberg matriculated at Swarthmore in the 1970s, he was torn between philosophy and biology. “I learned,” he says, “that philosophy had a lot of interesting questions but very few answers. And biology—I loved the intellectual muscle of hard facts, but biology didn’t ask the questions I wanted to pose.”

Reisberg found a happy medium in psychology. He was especially enchanted with the study of cognition, a close cousin of epistemology, and he fell under the spell of an austere professor, Hans Wallach, a German émigré who pronounced himself (in a daunting one-on-one meeting with Reisberg, who was just 16) “one of the best experimentalists in the world.” Working in a predigital age, Wallach concocted ingenious lever-and-pulley gadgetry to conduct elegant studies of human perception. To test how the right ear and the left ear respectively process a given snippet of sound, he had his subjects wear a gutted welders’ helmet attached to a giant overhead wheel. Subjects dialed the wheel by turning their heads—and thereby directed their ears towards any one of a surrounding bank of stereo speakers.

“Hans never threw his gadgets out,” says Reisberg. “His lab was like a science museum.” Eventually, as a grad student at the University of Pennsylvania, Reisberg aimed to set up psychology experiments framed by Wallach’s “wonderful intuition and inventive research methods.” Mostly, he studied perception, writing papers with titles such as Looking Where You Listen: Visual Cues and Auditory Attention.

Later, as a young professor at the New School in New York, Reisberg met graduate student Friderike Heuer, now his wife. She was researching how the mind records memory amid traumatic emotional events, and Reisberg joined her to coauthor several papers. Today, Reisberg still does research, but he always pairs with students. On principle, he gives them first-author credit on papers. He has written several leading psychology textbooks on learning, perception, cognition, and memory.

Teaching is his true calling, however; for two decades, he’s focused on delivering the polysyllabic news contained in psych journals to Reed students, lawyers, jurors, and lay readers worldwide. He has given workshops on eyewitness IDs to police detectives and lawyers. He’s also brought police officers to Psych 350, Psychology and the Courts, to enlighten Reed students on the challenge of obtaining witness IDs. Currently on sabbatical, he is writing a handbook that lawyers can use in cases involving perception and memory.

Reisberg sometimes receives unusual phone calls. In 2003, the Dalai Lama, aiming to forge a link between Buddhist monks and Western scholars, held a two-day conference at MIT, “Investigating the Mind,” to which he invited Reisberg and several other psychologists. Reisberg remembers a somewhat surreal encounter. “We were instructed to act like this was just a normal conference,” he says, alluding to his colleagues, “but that’s hard to do when one of the people at the table is wearing a bright orange robe and is regarded by millions of people as a deity.”

His Holiness focused part of the discussion on mental visualization, and the monks reported being able to see elaborate images as they meditated—deities in flowing robes and mandalas of intricate detail. “When the image appears,” said one monk, “it’s like a fish jumping out of the water.”

Reisberg and his colleagues were dubious. “Their level of detail,” he says, “was not consistent with our understanding of the brain tissue that makes mental imagery possible.” He found himself in the awkward position of countering the Dalai Lama. As he remembers it, he and his colleagues pointed out that visualization has to be limited by the structure of the brain; they told His Holiness, “You can’t grow a third arm no matter how hard you try.”

The Dalai Lama was resolute, reminding the psychologists of recent findings on neuroplasticity, in which the brain shows a remarkable ability to reorganize itself in light of training and experience.

Reisberg’s work with the court system has been more pragmatic—and also quite varied. In one case, he considered an account of sexual abuse, disclosed for the first time by a small child during her bath. He testified on how the human mind might process auditory data amid the hard, echoey surfaces of a typical bathroom. In another case, he testified on how pretrial publicity might bias jurors against a murder suspect. In a different case, when jurors were presented with a scratchy audio recording of a sting operation, he explained why it was problematic that the recording had been transcribed: the transcriptionist had made interpretive guesses as to the content—and had thereby biased jurors’ ears. In a Fruitland, Idaho, murder case, he spoke about the suggestive tactics used by police investigators seeking to rein in the suspect.

Reisberg is hardly the first psychologist to raise questions about police investigations. Professor Gary Wells, at Iowa State University, launched the inquiry, arguably, in 1978, with a journal paper that categorized the factors compromising eyewitness ID. Wells described “estimator variables”—that is, variables that the criminal justice system can’t control: the quality of the witness’s vision, for instance—and also “system variables,” which cops and courts can control, such as the quality of photo lineups. In the years since, the research on eyewitness ID has become deeper and more nuanced. There are now many American psychologists who give expert testimony on eyewitness ID, and courts are increasingly open to learning from them. Indeed, last year in State v. Henderson the New Jersey Supreme Court recognized the “troubling lack of reliability in eyewitness identifications” and ruled that, before admitting such IDs, courts must consider context, by asking, for instance, if testimony was induced, or if the witness was drunk or emotionally stressed.

As Reisberg waits, hoping that the Oregon Supreme Court grants Samuel Lawson a retrial, he feels that the court would do well to emulate Henderson. But he stresses that the real work on improving eyewitness IDs must happen outside the courtroom. “We need to keep doing research on two tracks,” he says. “In the lab, and also in real life. We need to ask if what we learn in the laboratory—about perception, about memory—plays out in real-world criminal cases. In some cases, it hasn’t.” Laboratory studies have shown that, when confronted with a weapon, subjects have proven less able to process peripheral sensory data and make reliable IDs. In real life, though, people confronted by weapons have not proven weak at making IDs. “We don’t know why this is,” Reisberg says. “It’s a puzzle we need to untangle.”

Even more important than research, Reisberg says, is dialogue with police. “They have complicated jobs,” he says, “and so much to learn. They have to know how to use firearms, and how to drive at high speeds and write up reports. They get about 10 minutes’ training on how to run an ID, but they really want to do it the right way: they want to keep bad IDs out of court too.”

Once, after a trial involving a Portland gang shooting, Reisberg says he lingered in the hall with a police detective. “We had a conversation about what kind of investigation would make sense to him and me. His concern was that, in gang killings, witnesses are afraid to pick the guy out of a lineup. They think, ‘If I do that, someone will kill my momma.’ He was asking, ‘How do you coax someone into making an ID?’ My concern was they shouldn’t be coached.”

“We came to an understanding of one another,” Reisberg says, “There was progress, and that conversation has to continue. Because it’s a double tragedy when the bad guy goes free and an innocent man goes to prison.”

Tags: Academics, Books, Film, Music, Editor's Picks, Professors, Research