The brain doesn’t actually remember what it sees, but what it thinks it sees. So what should the justice system do about it?
Humans place a great deal of emphasis on their memories. Memorization has served for millennia to help mankind avoid dangers, identify friends, track and map resources and civilize the world. The power of memory — in conjunction with the power of intelligence — has defined humans as the dominant species on the planet.
But for Ronald Cotton, the power of memory betrayed him. In 1984 — as reported by the Innocence Project — Jennifer Thompson, a White college student, was raped in her apartment in Burlington, N.C. The rapist broke into Thompson’s house through the rear door at around 2:30 A.M., cut the telephone line, held her by a knife to the throat and forced her to submit to intercourse. Thompson escaped and called the police from a neighbor’s home.
The assailant left and went on to rape a second college student, Elizabeth Watson.
Thompson stated that she recalled her rapist’s face. Watson, on the other hand, could not recall any facial details definitively and could only say he was an African-American man in his twenties. From a set of mug shots, Thompson picked out Ronald Cotton after examining the images for “four of five minutes.” Thompson was unsure at the time of the initial selection, but had her confidence bolstered by the reassurance of the investigators. During the live lineup, Cotton was the only person from the mugshots to be included. It is unknown if Thompson picked Cotton because she remembered him from the rape or from the mugshot. She told the investigator that Cotton “looks the most like him.”
Cotton was sentenced to life plus 54 years imprisonment. Evidence that pointed out that the second victim, Watson, did not pick out Cotton from a lineup was not allowed. Cotton was convicted on the basis of Thompson’s positive identification, the fact that a sole print at Thompson’s apartment was similar to shoes owned by Cotton and the fact that Cotton owned a flashlight similar to the one seen held by the rapist.
In 1987, the North Carolina Supreme Court found that since the recounts of both victims did not match, the police were too hasty in their assumption of Cotton’s guilt and ordered a retrial. By this time, Watson became convinced that Cotton was her rapist, and despite the fact that another man confessed to the rapes, Cotton was convicted of both crimes.
In 1995, DNA testing exonerated Cotton and pointed to the man that previously confessed to the rapes. On May 1995, the district attorney of Alamance County formally asked for the charges against Cotton to be dropped. Cotton was released from prison June 30, 1995 and in July 1995, the governor of North Carolina pardoned Cotton — erasing the conviction from his record and making it impossible that he will ever be challenged with these charges again. Despite all of this, the whole of North Carolina’s initial compensation for Cotton’s false imprisonment was $5,000 — $500 for every full year he served (although, a settlement of almost $110,000 was reached).
“You believe that person because they have no reason to lie,” said Gary Wells, a professor of psychology at Iowa State University. “The legal system is set up to kind of sort between liars and truth tellers. And it’s actually pretty good at that. But when someone is genuinely mistaken, the legal system doesn’t really know how to deal with that. And we’re talking about a genuine error here.”
Wells argue that Thompson’s hesitation in picking her assailant from the mug shots should have been the first hint that she would be an unreliable eyewitness. “Recognition memory is actually quite rapid. So we find in our studies, for example, that if somebody’s taking longer than 10, 15 seconds, it’s quite likely that they’re doing something other than just using reliable recognition memory.”
The inadequacy of firsthand recounting
Of Cotton and the 250 other DNA exonerees that the Innocence Project has helped free, 75 percent were convicted at least in part due to eyewitness identification. Of these cases, 88 percent of the eyewitness testimony were found to be unreliable or suggestive. 28 percent of these misidentifications were due to suggestive remarks from investigators, 34 percent came from a stacked or suggestive lineup and 63 percent came from unresolved discrepancy in descriptions.
More than a third of these cases involved multiple eyewitnesses in which as many as five separate witnesses misidentified the same person, and more than half were cross-racial identifications.
Scientists have recently discovered the true nature of memory, and it is different from what the legal system assumed. It would seem that the brain doesn’t actually remember what it sees, but what it THINKS it sees.
The nature of human thought
The human brain is one of the world’s most complex and intricate computational machines. Despite this, it is vastly limited. Drawing an operational charge of just 10 watts — about 1/8 the charge drawn by a typical bathroom lamp — and with a reaction processing speed clocked by Fermin Moscoso del Prado Martin of the Université de Provence in France at 60 bits per seconds, or 1/400 the speed of a desk calculator, the information the brain takes in, parse and memorizes 1) cannot be reflective of reality, as the brain is not fast enough, big enough or powerful enough to record moment-to-moment remembrances, 2) cannot be based on raw data or sensory inputs and 3) must be relevant to the person.
While the human brain’s processing capability can be radically scaled up to meet the task before it and scaled down at rest, the realities lie in the fact that for the average person, approximately only seven items can be held in short-term memory. Anyone who has tried to remember a telephone number can attest to this; the longer the person tries to hold on to the number, the “cloudier” the number becomes.
The human eye has an image resolution equivalent to a 576 megapixel camera at an 120 degrees field of view. However, this resolution is limited to the fovea centralis or the foveal pit, which is a highly-compact cluster of ganglion cells and cones about 0.2 millimeter in diameter and located between four to eight degrees off the optical axis. The foveal pit is how the brain gets around its low processing capabilities, as only information recorded by the foveal pit is actually processed into memory. The foveal pit serves as a spotlight: whatever it is looking at is the center of the person’s active viewing and attention.
The remaining 99 percent of the retina is broken up into multiple zones. The area around where the optical nerve leaves the eye is the optic disc. This is also known as “the blind spot” as there are no light receptors there. The retina is designed so that the fovea has the most cones, or color-detecting photoreceptors, and the density slowly reduce until, at the edge of the retina, there are no cones. While there is no consistent resolution away from the foveal pit, most professional estimates put the resolution of the non-foveal retina at about two to four megapixels and it serves mostly for motion detection.
What this means is, with the exception of a very small angle of ultrahigh resolution sight, most people’s vision has the same accuracy of a nineties digital camera with a dead spot on the sensor.
Limitations of memory
This doesn’t speak of what happens with the information that does reach the brain. Memory can best be defined by a quote from memory researcher Edward Bolles, “We remember what we understand; we understand only what we pay attention to; we pay attention to what we want.” As stated before, the brain can only remember about seven things at a time in short-term memory. To place the memory into long-term memory, the brain must feel that the information is important — either through repetition (a repeated route to work will eventually become rote, for example) or through an understanding of the importance (such as, remembering the face of someone that attacked you).
The problem lies in the fact that a person is not remembering the face when he recalls a person, but the details of the face, which the brain reimagines as an image. As a person develops new memories, pathways to old memories are lost and key details may go missing. It is estimated that in the very first day of remembrance, 46 percent of all that is remembered is lost. A person may integrate elements of other memories, suggestions from other people or general assumptions to form this new mental image, which would carry the weight of memory.
This is the problem with witness identification: the brain is not a reliable recorder. The brain is an analyzer. It takes input and perceptions from the outside world and interprets them in a way that makes sense for the person, but may not be reflective of fact.
For Thompson, the realization that her memories lied to her and sent an innocent man to prison for nearly 11 years almost collapsed her from the inside. Her initial reaction of “No, that can’t be true. It’s not possible. You know? I know Ronald Cotton raped me. There’s no question in my mind,” was shared by original case investigator Det. Mike Gauldin to CBS’s “60 Minutes.”
“I started to cry immediately. And I looked at him, and I said, ‘Ron, if I spent every second of every minute of every hour for the rest of my life telling you how sorry I am, it wouldn’t come close to how my heart feels. I’m so sorry.’ And Ronald just leaned down, he took my hands…and he looked at me, he said, ‘I forgive you,'” Thompson remembered of her first face-to-face with Cotton after his acquittal.
Based on the work of memory researchers, such as Elizabeth Loftus, the legal system is starting to address the inadequacies of eyewitness testimony. The New Jersey Supreme Court, for example, has issued a rule change stating that jurors must be made aware of the imperfect nature of memory and the fallibility of eyewitness testimony. Wisconsin has instituted a set of “best practices,” including lineups administered by an agent unfamiliar to the case, confidence assessments and the use of non-suspect fillers to minimize suggestions. North Carolina, Northampton, Mass., Suffolk County, Mass. and Santa Clara County, Calif. have all also issued new rules for handling eyewitness testimony.
“While eyewitness testimony can be persuasive evidence before a judge or jury, 30 years of strong social science research has proven that eyewitness identification is often unreliable,” the Innocence Project wrote. “Research shows that the human mind is not like a tape recorder; we neither record events exactly as we see them, nor recall them like a tape that has been rewound. Instead, witness memory is like any other evidence at a crime scene; it must be preserved carefully and retrieved methodically, or it can be contaminated.
This article originally appeared Aug. 16, 2013.