CRIME & MISTAKEN MEMORY
ALMOST 50 YEARS AFTER HE WAS CONVICTED OF AN ARMED ROBBERY, Réjean Hinse was compensated a record $13.2 million in April 2011 by the federal and Quebec governments for his wrongful imprisonment in the 1960s. Before the trial, one of the victims picked out
Hinse’s case is not unique. According to the Innocence Project, a legal advocacy group, eyewitness misidentifications account for approximately 75 per cent of the roughly 300 wrongful convictions overturned in the United States by DNA evidence. “That figure tracks very well to Canada. We have about 20 or 30 documented wrongful convictions,” says Alan Young, the co-founder and director of the Innocence Project at Osgoode Hall Law School at York University in Toronto.
Psychological and neuroscience research have chipped away at the credence we give to witness accounts, and shown how memories can be unwittingly manipulated. And yet eyewitness identification remains a very important piece of evidence in many criminal cases. Some researchers are asking what role memories should have in court, if neuroscience can make criminal trials more objective and whether our legal system is keeping up with – or getting ahead of – science.
There’s a pretty good chance that you remember where you were when you heard about the 9/11 terrorist attacks on the World Trade Center in New York City in 2001. You might recall what you were doing when you learned that Princess Diana had died from the injuries she had received in a car crash in Paris in 1997 or even when the Space Shuttle Challenger broke apart and ignited shortly after takeoff in 1986.
Unlike the usual memories we collect, these flashbulb memories of traumatic or emotionally significant events become etched into our personal timelines. They are richly detailed and often contain obscure pieces of information about the clothes you were wearing, the coffee shop you were in, or other contextual elements that form backdrop for an elaborate, blowby-blow memory of the news.
The concept of a flashbulb memory was first floated in the late 1970s by Roger Brown and James Kulik, psychologists at Harvard University in Boston. The hypothesis was that the brain captures what you see, hear, smell and feel at the instant an emotional event occurs – like the flare of the camera’s flash when you point and shoot – and that it stores these special memories unadulterated for years and years.
But research from the past 25 years has shown that the flashbulb memory label is flawed. Psychologists studying the effect have tested and retested the accuracy of memories formed during traumatic and emotionally intense events and found that while people are confident in the accuracy of their recollection of the event, more often than not, their memories are unreliable.
“Memory is not a picture-perfect recollection of all details previously experienced,” says Craig Stark, a cognitive psychologist at the University of California, in Irvine, who studies the mechanisms of memory. “After some period of time, the accuracy of the memories starts to drift.”
On Sept. 12, 2001, Jennifer Talarico and David Rubin, psychologists at Duke University, in Durham, N.C., contacted Duke students and asked them what they had heard about the attacks of the previous morning. They tested the vividness and recollection of the memory, as well as whether the students believed the memory was accurate.
“Who or what first told you the information?” and “Where were you when you first heard the news?” were among questions they asked. They split the students into three groups, and followed up with them one week, six weeks or 32 weeks later.
As time wore on, the students remembered fewer and fewer details, and yet they remained as certain as the first day they were tested that the memory was accurate. Talarico and Rubin found that the details of so-called flashbulb memories were as easy to forget as regular, everyday memories, but the belief in the memory’s accuracy was exaggerated.
“The true ‘mystery,’ then, is not why flashbulb memories are so accurate for so long, as Brown and Kulik thought, but why people are so confident for so long in the accuracy of their flashbulb memories,” Talarico and Rubin wrote in the article describing the experiment.
Memory is a three-step process: First, during the encoding phase, details of the event are received and processed. Second, in the consolidating phase, the information is transferred to into long-term memory creating a permanent record of the encoded information. Third, in the retrieval phase, the stored information is recalled in response to some cue, like a question a friend asks or catching sight of someone as they climb into a taxicab.
The key structure involved in the encoding of memories is the hippocampus, buried deep in the centre of the brain. But when memories are formed in emotional situations, the neighbouring amygdala is pulled into the operation.
“The amygdala is sensitive to stress hormones, it has a lot of receptors for them,” says Jorge Armony, a neuroscientist at the Douglas Mental Health University Institute and Mcgill University, who studies the interplay of emotion with consciousness, attention and memory. “The more stress, the more the amygdala responds,” he says.
Imagine for a moment, that you are walking down the street at night and a menacing figure suddenly makes a beeline toward you. In reality, the situation would trigger your brain to switch on the sympathetic nervous system and the adrenal glands that release stress hormones and stress neurotransmitters, the chemical substances that nerves broadcast to communicate with each other.
In addition to triggering the “fight or flight” response, these hormones and chemicals also act on the amygdala, which strengthens the storage of information into long-term memory.
“The emotional arousal makes it so the memory gets laid down better. It is stronger and it will last longer,” says Stark. The recalled memory is more vivid and powerful than the humdrum memory of your walk home from work, but it may not be more accurate.
A false memory is convincing, and seems as real as a true memory. “You can’t tell them apart. But every time you remember the event, you reconsolidate and re-encode the details. The more you tell yourself the story, the more you are going to dilute the memory,” says Armony.
Research suggests that witnesses form false memories by combining bits of information from several sources, so that the memory more closely resembles a fusion of experiences rather than a recording.
“We forget that our memory is not perfect, that it is open to distortion,” says Stark. “But this is a normal memory system. We shouldn’t expect juries or eyewitnesses to have anything other than a normal memory system.” “I’ll never forget the face of the man who held the gun to my head.”
Unfortunately, the opposite may be true. Experience and laboratory studies show that people have more trouble recalling the details of what a thief looked like or what he wore if he held a gun or a knife.
Striking, dangerous or unusual details tend to grab our attention, says Armony. “You remember what you pay attention to. The gun is the threat stimulus, so that is what you are going to remember. You’re not paying attention to the colour of the shirt of the assailant or the face.”
Armony, who has studied the effect of facial expressions on the brain, has found that we are more likely to remember frightened faces than neutral or happy ones. He’s also found that women are better at remembering fearful female faces.
Armony does many of his studies using a brain scanning device called an FMRI that can reveal the regions of the brain that are at work during a particular task. During Armony’s tests of fearful faces, he finds that the amygdala – the centre of emotional memory – is hard at work.
“A fearful face is a signal of danger in the environment. It’s a sign that clearly you need to pay attention to the environment and be ready to respond,” says Armony. Scientists and legal experts are trying to find ways to improve the way witnesses identify the suspects that are brought to trial. A decade ago, a Canadian commission of inquiry brought about big changes in the way police conduct suspect lineups and identifications.
In 2000 and 2001, former Supreme Court justice Peter Cory led a commission of inquiry to look at the circumstances around Thomas Sophonow’s arrest and prosecution. Sophonow was arrested in March 1994 and charged with the murder of a teenage waitress at a Winnipeg doughnut shop.
Several eyewitnesses had seen a man leave the shop the night that the young woman was found. When police arranged a photo lineup of possible suspects, more than one pointed to Sophonow’s picture. Sophonow was tried three times for the murder, convicted and spent four years in prison before he was acquitted of the crime by the Manitoba Court of Appeal.
When Elizabeth Loftus, an expert on memory and eyewitnesses from the University of California, Irvine, reviewed the Polaroids the police had assembled for the lineup, she remarked that Sophonow’s picture stood out. The picture had a border around it, had been taken outside, and portrayed Sophonow in a cow- boy hat, worn by none of the others in the lineup.
“Sophonow was also the only man who was depicted in both the photo and the physical lineup, making it possible witnesses chose him because they had seen his picture before. Moreover, Sophonow was the tallest man shown to witnesses, and the killer had been described as a tall man,” Loftus wrote in the journal Canadian Psychology in 2003.
In his report, commissioner Cory called the lineup “startlingly unfair,” and he issued a set of recommendations that included best practices for witness identifications. For example, photo lineups should contain at least 10 subjects and should be presented to the witness sequentially, not as a group. The officer conducting the lineup should not be able to identify the suspect or know whether his photo is in the lineup and should record “verbatim” all of the comments the witness makes about each photo.
Television suspect lineups frequently show six sometimes-similar men parading in front of the witness at the same time. Studies show that this method tends to attract false identifications because the witness makes relative judgments about each of the suspects. Witnesses are more likely to make a true identification – and less likely to make an error – if the individuals or their photos are presented sequentially.
The U.S. National Institute of Justice has recently made recommendations similar to those issued in the Sophonow report to improve eyewitness reliability. And many – but not all – police in North America employ them.
In Quebec, each police department has its own management policy when it comes to suspect identification, says Valerie Tremblay, a crown prosecutor at the Quebec Justice Department who educates police officers on the best practices for witness interviews and suspect identification at the École nationale de police du Québec in Nicolet, near Trois Rivières. When she and her colleagues teach, they explain the Sophonow recommendations to the officers, but emphasize that each officer must follow their department’s rules.
In the end, what’s most important, is to avoid any influence on the witness’s recollection of the suspect or their choice. “Any gesture, any comment by the police officer, risks leading to a judicial mistake,” says Tremblay.
“We are far away from the practices we used to have 10 or 20 years ago,” she says. Witnesses are no longer under pressure to find the suspect, by being told that he is in the photo array.
Despite its imperfections, eyewitness identification remains an important part of many criminal cases, though experts now say that it should be considered like any other piece of circumstantial evidence.
Psychology researchers continue to look for ways to improve witness identification of suspects. In one new study, Neil Brewer, from Flinders University in Adelaide, Australia, tested a novel lineup design to reduce the rate of wrongful accusations.
Even with sequential lineups, witnesses spend as much time as they’d like pondering the faces of individual suspects. Brewer recruited more than 900 volunteers and showed them short films displaying crimes or run-of-the-mill events that focused on a single individual.
One half of the group was shown a traditional police lineup of 12 people and took their time in identifying the offender. The other group had three seconds to make their “yes” or “no” choice and to rate their choice using an 11-point scale that ran from “absolutely confident this is the culprit” to “absolutely certain this is not the culprit.” With the time limit and the scale, the witness was more likely to identify the suspect.
Other researchers are looking at how neuroscience is being used to demonstrate guilt or culpability – and whether it should be. Jennifer Chandler, a professor of law at the University of Ottawa, has launched a survey of Canadian case law for cases where lawyers have introduced – or tried to introduce – neuroscientific evidence into court.
“In the U.S. there have been a couple of attempts to introduce FMRI evidence for the purpose of lie detection, but I haven’t seen it in Canada yet,” says Chandler. She suspects that the first time these brain scans are introduced into a Canadian court, it will be to determine the severity of the sentence by focusing on the accused rather than on the reliability of eyewitness testimony.
“If you take a deterministic view of the brain, it means you are less morally blameworthy, so a lower sentence would be appropriate,” say Chandler.
But during sentencing judges also consider the prospect of the criminal’s rehabilitation or treatment, she says. Such evidence could wind up backfiring for the accused. If you have an identifiable brain condition, you might end up labelled “dangerous” or sequestered from society. “Our study will give really good insight into how judges think about this if FMRI is introduced,” she says. Elsewhere, neuroscience evidence has possibly been introduced too early – before science has had the chance to prove that it works.
In September 2008, the New York Times reported that a woman in Pune, India, had been convicted of murder based on evidence that included a scan of her brain. The test, called Brain Electrical Oscillations Signature (BEOS), is said to be able to detect small changes in electrical activity in the brain emitted when memories are retrieved from storage. According to its inventors, BEOS can identify a guilty person by detecting neurological signals that are specific to memories of acts in which the accused participated.
If true, the test could become part of a new battery of lie-detection methods. There is evidence that our brains do respond differently to familiar stimuli than to new ones, but Beos does not seem to have been put under the scrutiny of peer-review to whic h most science is subject.
With legal experts eager to ask the questions of which signals we can and cannot extract from our brains, and neuroscientists leaping at the opportunity to apply their knowledge, the field of neuro-law will continue to grow. Ultimately, it could lead to better-founded cases and eliminate potentially costly mistakes.