By Thomas Bartlett Whitaker
Simonides of Ceos was what you might call a pretty lucky guy. While attending a large banquet in the 5th Century BC, the poet stepped outside the hall for a moment (for reasons which aren't exactly clear but which probably amount to the most fortunate relieving of one's bowels in the history of mankind), when the entire building collapsed behind him. He was the only survivor. When asked to give an account of who was buried in the debris and where, Simonides discovered a curious thing: while totally unable to recall names in list-form, he was able to recall each person and where they sat by first remembering the exact layout of the dining hall. He later realized that by using this same layout, he could cement other people and concepts - all of the great Greek dramatists, say - in his memory in a way unavailable to him otherwise. These mental buildings became known as "memory palaces", and the art of memorizing entire ballads was passed down to modern readers in a short Latin rhetoric textbook called "Rhetorica ad Herennium.”
I've been trying to use both memory palaces and other mnemonic systems like the Major System for years, because, well, my memory stinks. Memory palaces work because human beings have very good spatial memory: we evolved through natural selection to benefit greatly from being able to recall where food and danger were located. Needing to memorize 1000 names (which no tribe of early humans had a need for) or long lists of telephone numbers are entirely recent phenomena, and the brain hasn't yet caught up with modern necessities. By pegging mundane lists of things like groceries you need to pick up at the store to visually ridiculous images (thinking of the Queen of England sitting at your kitchen table drinking a cup of tea, in order to remember to buy said tea) and then placing these images in your palace, you activate the portions of our brains responsible for better types of our memory. Spatial memory is the cute blonde who gets the puny weakling of the rest of our recall past the bouncer of forgetfulness, in other words.
The pros (yes, there are memory contests; the participants call themselves "Mental Athletes") can use these systems to memorize entire sections of the Oxford English Dictionary or huge sequences of random numbers after only a few minutes of very intense concentration. Usually, they wear earplugs plus aircraft carrier deck ear protection helmets plus dark glasses, and the audiences at the competitions are ordered to be completely silent. (Who goes to such competitions as spectators is not a question I have any answer to....) I am devoid of such tools, but I have had some success with the system nonetheless. My palace is my childhood home. I can usually fit things into the entryway, the dining room, and the kitchen. When I round the corner into the living room, to continue the metaphor, my palace collapses pretty much like the one that very nearly squished Simonides. Still, engaging in 15 or 20 minutes of memory training a day has helped me to see that the human brain is more powerful than we usually realize. With practice, it is capable of some pretty amazing feats.
Untrained and unmonitored, left to its own devices, however, memory is extremely undependable. This is what makes these Mental Athletes so incredible, seeing how different they are from the rest of us. It can make us feel very uncomfortable to realize how prone to error our memories are, for they are fundamental in constructing our identities. Any hint that what we recall may not be accurate instantly makes us feel somehow unmoored. Still, if we are honest with ourselves we will admit that we are often assailed by memories we wish we could forget but which won't go away, names or places we just thought about but can no longer get off the tips of our tongues, and, most troubling by far, being absolutely certain about some past events only to recognize that we were dead wrong. All of these events happen to us all of the time. We may pretend that memory is some sort of tape that we can rewind at our pleasure, but the reality is much more complicated
I recently read a book by Charles Fernyhough called Pieces of Light: How the New Science of Memory Illuminates the Stories We Tell About Our Pasts. This "new science" is actually not so new if you happen to work in the fields of neuroscience, evolutionary biology, or psychology, but much of it was new to me so the title was accurate. Fernyhough states that, "when you form a memory, you don't simply record a mental DVD of events that get played back at the moment of remembering. Memories are constructions, made in the present moment; they are not direct lines to the events themselves...when those memories are so central to our own sense of identity, we are naturally resistant to the idea that we could get them wrong. But we do get them wrong and probably more often than we think...it is well established that a memory is not a faithful, unchanging representation of an event. An event plays itself out in our heads, and we decide, on the basis of many different kinds of information, whether it actually happened or not." Memory, it turns out, is more mutable than popularly known. Each time you dredge up an event from the past, you are always going to be viewing it through the context of the present; each retelling subtly alters the memory in question. Each attempt at recall gives us a progressively smudgy copy of a copy of a copy, over which our present attitudes, beliefs, and "certainties" have drawn a new outline and which dares us to disagree. Merely imagining a potential event can also encode it like memory, so you falsely "remember" it happening much later. The science in the book is very interesting, but most intriguing to me was the uncomfortable feeling it left me with. We trust our memories, which is probably the last thing we should do.
It does make me feel a little better knowing that many really smart people in the past have been wrong about all of this. The great Descartes himself (he of the cogito, ergo sum business) realized that "subjective certainty" was a real problem. He never got around it, really, because he based the "veridicality" of sense perception on the only solid ground he could come up with, namely, on the validity of his proofs of god's existence. If the very notion of the "self" might be said to be located in the memory, knowing that we are so unfixed is a worrisome proposition.
The consequences for forgetting something or misremembering run the gamut from the harmless (forgetting to pick up milk at the store) to the serious (forgetting your wife's birthday...again), to the disastrous (an airplane mechanic neglecting to check a part). Considering that I have known men who were taken to Huntsville in chains and filled with poison based purely on eyewitness testimony, this issue interests me greatly. As far back as the 1800’s, experts have been aware that eyewitness identification was subject to error. This is yet another example of the courts and general society lagging far behind what is known by science, and when I question the reason for this, the best I can come up with is that there is something inherently comforting and satisfying about a witness pointing to a defendant and proclaiming: "He's the one; I know it as well as I know my own mother." A study by Loftus and Doyle in 1992 set up mock trials to investigate just how powerful this effect is, and the results are astonishing. Two different sets of jurors heard evidence presented, with the only difference consisting of the presence or absence of eyewitness testimony. Minus the witness, 18% of jurors gave guilty verdicts, while 72% of the jurors returned this verdict when an eyewitness was present. Here is the scary part: even when the identification was impeached, the guilty rate was still 68%. Many other studies have mirrored this result: even when factors such as bias or poor visibility are acknowledged, jurors overwhelmingly believe eyewitness identifications of suspects. And so prosecutors keep using them, even when the science directly contradicts their validity.
This is yet another indicator of a massive problem in our system of law, one that I have written about before but which seems to be avoided by virtually anyone else, namely the entire concept of the "jury of one's peers." Twelve laypeople are expected to come into court and determine objective truth; the presenters of these various "truths" are showmen, whose job it is to sell a certain verdict. In no other situation in life would you rely on twelve people right off the streets to arrive at such an important outcome. Imagine: you are at a restaurant. You feel an intense pain in your stomach and fall to the floor. A woman rushes to your side, screaming, "out of my way, I'm a doctor!" She quickly claims that your appendix is about to burst and if not removed this instant, you will perish. A man then steps forward, saying he feels it is food poisoning, and demands that twelve random diners vote on the cause instead. Given the choice, only an absolute fool would give a rat's ass about what an architect or a plumber thinks about the pain in your belly. You will always trust the experts, in whatever situation you are in. This is a farcical example, of course, but it really isn't that far off from the reality of our jury system. This is why many jurisdictions in Europe use professional jurists, because they recognize that normal people are not to be trusted in matters this weighty.
The numbers back up my argument: nationally, in cases which were later overturned thanks to DNA evidence or other hard science, eyewitness error was the cause of the original conviction 75% of the time. In a study by Wells, et al, from 1998, an analysis of 40 cases where DNA exonerated wrongfully convicted people showed that a whopping 90% involved mistaken identities. Cutler and Penrod examined this matter by setting up controlled experiments where a person goes into a convenience store and performed memorable activities such as paying for their purchases with pennies. Later, the clerk viewed a photo lineup and attempted to peg the customer. The percentage of correct identifications ranged between 34 and 40%. The percentage of false identifications fell between 34 and 38%. In other words, these clerks got it wrong as often as they got it right, and this was without the stress present in situations involving a weapon and potential violence, and also without the passage of time or the corrupting influence of flawed police identification procedures.
In 1984 a woman named Jennifer Thompson was raped by a man who put a knife to her throat. During the incident, Thompson studied her rapist with great care in order to memorize his facial features. "I studied every single detail on the rapist's face. I looked for scars, for tattoos, for anything that would help me identify him. When and if I survived the attack, I was going to make sure he was put in prison and he was going to rot." Later that day, Thompson went to the police station and worked up a composite sketch of her attacker. The police produced a photographic lineup, and Thompson pinpointed a man named Ronald Cotton. She was positive it was he, and she testified against him at trial. "I was sure. I knew it. I had picked the right guy, and he was going to go to jail. If there was the possibility of a death sentence, I wanted him to die. I wanted to flip the switch."
Jennifer Thompson was as certain as she could be. She was also as wrong as she could be, as DNA later showed. Ironically, she was presented with the real rapist in a trial proceeding a year after the attack, but she swore she'd never seen him before. Ronald Cotton served eleven years in prison before DNA cleared him. If this had been a death sentence in Texas, he would have been long dead before DNA had had a chance to free him.
In March of 1998, 21-year-old Alfonso Gomez was convicted of being the gunman who fired multiple shots from a passing car that killed Martha Gonzalez and seriously wounded another victim. Two witnesses positively identified the young man, and their identifications were virtually the only evidence presented by the prosecutor. It was enough: Gomez was sentenced to 41 years to life in prison. Orange County DA Howard Gundry claimed that the identifications were powerful evidence, describing them as "righteous" and "certain." "Is he the right front passenger?" Gundry asked rhetorically in his closing argument. Absolutely... I mean it was a dead-bang identification." Dead-bang wrong, as it later turned out. Gomez's murder conviction was later vacated and a man named Armando Parra was arrested. Needless to say, Gomez and Parra look nothing alike.
All of that is academic. Ten minutes from now, precisely zero of you will remember the names of Ronald Cotton or Alfonso Gomez. You might recall the name of Willie Johnson, however, since he writes for this site on occasion. Like Cotton and Gomez, he was convicted based on eyewitness testimony in a total absence of physical evidence. Unlike Cotton and Gomez, Willie didn't escape a death sentence: he's been waiting on a date with the needle since 1987.
On July 1, 1986, Angela Womble, her mother Mrs. Willie Womble, and Angela's 16-month-old son Terrance were at their home in Richmond, California. Angela's former boyfriend was a drug dealer and was in the habit of using her home as a stash spot for his money. At approximately 10pm, Angela heard a knock on the door, and ultimately recognized the voice of Allen Duchine, a friend of her ex-boyfriend's. Upon opening the door, Duchine and another man pushed her inside the home, armed with a rifle and a shotgun.
After discovering that that there was no cache of drug money in the home, the second suspect killed the elder Womble by shooting her in the head with the shotgun. Duchine then shot Angela with the rifle, and the pair fled the home. Angela survived her wounds and was able to point out Duchine from a photographic lineup while still in hospital. The gossip stream in Richmond had kicked into high gear after the shootings, and thanks to a series of telephone calls between several peripheral people in Willie Johnson's life, the police zeroed in on him as the second suspect. At some point prior to a lineup involving Johnson, Angela had already been told by relatives and friends in the community that "Willie Johnson" was a suspect. We will never know if this information came with a description, but I find it hard to believe that things like facial features were never discussed. At any rate, when showed a photographic lineup of him, she said she could not be positive, but that he "looked like the person" who had shot her mother. She claimed that he had an earring in his right ear and that he was tall and dark. She could not, however, even recall whether the lights in the kitchen and the living room (where she was shot) were on or not.
Eventually Angela was shown a second photo lineup with fresh photographs of Willie. She took some time before pointing him out, saying: "His face looks familiar...but you see, it was dark." She said the photo of Johnson had the "same facial features, round head, and forehead structure." On the way to this second lineup at the county jail, the lead detective on the case had admitted to Angela that they had a suspect in custody. Willie Johnson's second set of lineup photographs was taken while he was wearing a jailhouse uniform. He was the only member of the second lineup to wear such attire. Despite the fact that Angela could have easily drawn a connection between the uniform and the detective's comments about a suspect in custody, the state courts of California found no legal problems with Angela's identification. Neither did the courts have a problem with Angela changing her testimony after learning that Willie wore an earring in his left ear, not his right. Nor did anyone seem to mind that Willie was the only person featured in both the first and second photo lineups. That Angela Womble might have been remembering the first series of photographs during the second lineup instead of the actual suspect seems to have occurred to no one. As in the case of Alfonso Gomez, the prosecutor in Johnson's trial told the jury that Angela's "identification of defendant [was] as reliable an identification as may be found in a courtroom." When appealing on this obvious fabrication, the state court claimed that "defendant also contends that prosecutor engaged in misconduct by calling [Womble's] identification of defendant as reliable an identification as could be found in a courtroom. We disagree; the statement cannot reasonably be interpreted as vouching, but would have been understood as an invitation to draw the desired inference." I very much doubt that anyone on Johnson's jury understands the tiny crack of daylight between "vouching" and "an invitation to draw the desired inference"; I'm not sure there is one, actually. What is certain is that this is the language by which a supposedly modern nation kills its own citizens.
Thanks to advances in fields like the social sciences, neuroscience, cognition, and psychology, we know far more about how humans process memory during both violent acts and the eyewitness identification process. There are two sorts of variables that the police must deal with in these situations: estimator variables which cannot be controlled (things like lighting, distance, problems in unreliability of identification when witness an perpetrator are of different races, stress, and presence of a weapon) and system variables which can be controlled (including the type of lineup used, the selection of "fillers" in the lineup), instructions to witnesses during lineup, a lack of blind or double blind administration, and confirmation of a suspect during the lineup . There are literally hundreds of ways an investigator can intentionally or unintentionally corrupt the identification process. Merely saying, "good, you identified the suspect" after a lineup can change a witness's highly tentative "maybe" to a "100% confident" identification.
Stress and weapon focus are also powerful factors. Yale psychiatrist Charles Morgan once tested the ability of highly trained military survival school students on their ability to identify interrogators following low- and high-stress scenarios. In each sort of interrogation, subjects were face-to-face with their questioners for 40 minutes in a well-lit room. When they were asked to identify their interrogator, subjects exposed to low-stress situations incorrectly pinpointed their targets only 12% of the time. Those subjects exposed to high-stress questioning, however, failed 68% of the time. That's a fail rate of more than two out of three times, involving highly trained military personnel after a 40 minute exposure. When you add a weapon into the mix, the rates get even worse. For reasons that are obvious, eyewitnesses tend to be drawn to the point of the weapon rather than the subject's face, an effect known as "weapon focus."
Without extensive memory training, the rate of memory degradation (the "forgetting curve") has been shown to be "Ebbinghausian" in nature, that is, it begins to drop off sharply within 20 minutes following the initial coding, and continues to do so exponentially until it begins to level off around day two. In the case of Willie Johnson, his tentative identification did not take place until well after the forgetting curve had finished its business.
The largest cause of error in this process is known as the "relative judgment" process, a term which fits Johnson's case in more ways than one. In the first, scientific sense, this is what happens when you are trying to select a person from a lineup and simply pick the person that most looks like the perpetrator. The actual person responsible may not be present, so the person who best fits the memory gets tagged. In Johnson's case, there is actually a very good reason why Angela Womble said he looked familiar, but couldn't be certain." In this case, "relative judgment" hits far closer to home than in most.
Since Willie Johnson's arrest and conviction, 23 witnesses have come forward to identify the real killer as Willie's brother, Timothy Johnson, who died in a drive-by shooting in 1989. Timothy was a violent, crack-cocaine user and dealer, a man feared in Richmond. He also happened to be friends with Duchine and also the direct competitor of Angela's former drug-dealing boyfriend. The stories that have come out since his death are a sad commentary on both our criminal justice system and on our class and race relations. In response to this flood of new witnesses and testimony, Hal Jewett, the Contra Costa deputy district attorney who prosecuted Johnson stated: "I have trouble believing that all of those witnesses were so devoid of a sense of morality that they would remain silent." This is typical prosecutor-speak: what does it matter what he has trouble believing? Justice is not about which issues the various parties can and cannot understand. It is about facts. His comments speak to both a total dereliction of duty to pursue justice regardless of where the path leads, and to a frightful amount of ignorance of the experiences of urban African-Americans in his own city. In any case, if one hesitant, coached witness is good enough to sentence someone to death, why aren't 23 good enough to cast doubt?
The new witnesses are remarkably specific. Some said that Timothy Johnson confessed to the crime, and many of these people attempted to get him to turn himself in. Timothy said he could not stand imprisonment, when pressed. Under cross-examination, these witnesses gave different reasons for why they had not come forward during the first trial. Many stated that they were never approached by the police, while others said they feared Timothy. Many said that it was an unwritten rule of the neighborhood to steer clear of any matters involving drugs or drug-dealers. A few quotes from these witnesses taken from news articles and trial transcripts are very enlightening.
Willie and Timothy Johnson's aunt Betty Thomas Johnson said it was simple - although perhaps incomprehensible to Richmond outsiders. In Richmond, things are done differently from Petaluma... once I was at my mom's house [in Richmond], and there was a drive-by shooting. In Petaluma, I would have described the shooters to the police...in Richmond, members of my family would be injured or killed if I cooperated with police. My mother told me I was crazy even to think about talking to the police." She went on to admit that another nephew had told her that Timothy was the actual killer. “I immediately called my mom and told her Timothy was the killer, but she already knew; everyone in the family seemed to know much more than me. I didn’t think I could change anything by saying anything."
Willie's attorney John Philipsborn stated that "some older people felt we were foolish to believe somebody like Willie Johnson would really get a thorough review of this case... there are plenty of people who just don’t believe that people from their income level and their ethnic background will ever get a fair hearing."
It gets worse, of course, as it always seems to around this joint. Six years after Willie's conviction, Allen Duchine, the other assailant who had received a life sentence for his role in the shootings, wrote to the California Supreme Court to inform them that it was Timothy, not Willie, who had been with him on that fateful day. He admitted that it was Timothy who had killed Mrs. Womble, and that Willie was not present. “After Angela Womble identified the wrong person in connection with the shooting, I went along with that story because I feared Timothy Johnson," he wrote.
When Duchine later attended a hearing on the matter, he reversed himself again. Visibly angry, he claimed that his letter was a lie: "I gotta look out for myself now," Duchine said on the witness stand. "I'm tired of sacrificing my life to help everybody else out." Any convict could tell you what had happened here: the system leaned on him. It is common for inmates to be simultaneously threatened and promised parole when co-defendants get retrials or hearings in court; I've seen this happen at least five or six times during my stay on death row. There is no bigger carrot than freedom, no heavier a stick than being sent to solitary confinement for potentially decades over a nonsense case. Sure enough, now that years have passed and Duchine's parole situation has improved, he has gone back to admitting that Timothy was the real killer. I've heard it said that in history, first comes tragedy, then comes farce. For the life of me, I cannot figure out how to separate the two in the case of Willie Johnson.
What is a responsible, ethical person supposed to take from stories like Willie's? It's hard to tell sometimes. First, I think, we need to take a moment and pause in front of the cultural mirror and acknowledge that we are a nation that permits our fellow citizens to be legally murdered based off of the testimony of a panicked victim in a room where the light may have even been off. We need to recognize that so many of the "certainties" in our memories are anything but, even if that means admitting that we are neither as stable or as accurate as we would have preferred. Finally, many of you will one day be jurors. It isn't likely that you will ever have to sit on a death-qualified panel, but my final point works equally well whatever the nature of the case: when you sit on a jury, recognize that you are merely a member of the audience of a play. Prosecutors are actors, salesmen with the entire power of the state apparatus behind them. When they tell you that their evidence is "dead certain," understand that the only thing that is actually certain is that they want you to believe them. Willie's story may be the weathervane of where the judicial winds are blowing today, but it is you who get to determine their direction tomorrow.
San Quentin State Prison
San Quentin, CA 94974