In Part I of this series, I began recounting my recent experiences testifying as an eyewitness memory expert in an armed robbery trial. The two victims in the case, both White men, had described their assailant as a 20-something-year-old, clean-shaven Black man, 6'5" tall, with missing or broken teeth. The defendant standing trial, however, was 18 years-old, 5'10", goateed, and not the slightest bit dentally challenged.
My testimony focused on the extensive scientific literature on eyewitness accuracy, the overarching conclusion of which is that memory is not a videotape. Whether we're observing a crime in process or a mundane conversation at the grocery store, we don't capture a pristine snapshot of the events in question that we can then subsequently conjure up on demand. Rather, we usually remember the gist of what we observe, filling in the gaps with expectation, assumption, and information we learned later on.
But this constructed nature of memory doesn't stop us from being confident in our own ability to remember events perfectly. For example, I think I can recall things about the apartment I lived in as a toddler. I can even picture the railing around the balcony outside. But do I really remember any of that, or have I pieced together what feels like a recollection from old photographs, home movies, and stories my parents have told over the years?
I also have more recent memories that feel quite vivid, but for various reasons fail to stand up to the close scrutiny of reality check. Such as social interactions that I'd swear I can remember word-for-word, action-for-action, exactly as they occurred... except that my mental image of these events depicts them as if shown from a third-person security camera on high. In my memory, I can actually see myself in the conversation from a detached perspective, which clearly isn't how I experienced the event originally.
The legal complications of the constructive nature of memory are clear: being an accurate eyewitness is far more difficult that we usually give it credit for. Consider the work done by the Innocence Project, a national organization dedicated to assisting wrongfully convicted individuals through DNA analysis. Thanks to their efforts, over the past two decades, more than 250 people have been exonerated by DNA tests that either weren't available at the time of their original trial, or just weren't conducted for some reason.
Of those 250+ cases, ¾ involved eyewitness testimony. That is, in close to 200 real cases over the past 20 years, at least one eyewitness (and sometimes more) pointed at an innocent person and said–quite confidently but erroneously–"that's the guy." And these are just the cases that we know about.
Now, don't think that all these instances of mistaken identification are simply a matter of a witness only getting a fleeting glance of a masked culprit. Consider Jennifer Thompson, who, in the early 1980's was raped by a stranger in her own home. Her assailant spent an extended period of time in her apartment–even carrying on several conversations with her. In fact, Thompson spent much of her attack strategizing about how best to remember what her rapist looked like, promising herself that she'd be able to identify him later.
She did identify someone in a police lineup, and very confidently at that. But she was wrong. As detailed in the book that she later co-wrote about the incident, the man she falsely identified would spend 11 years in prison before being exonerated by DNA analysis. Her co-author for the book? Ronald Cotton–the very man she mistakenly identified years earlier, who now works alongside Thompson in her campaign to raise awareness of mistaken conviction.
Back to present day... in the case in which I testified, many of the major issues with the eyewitness identification didn't require an expert to explain. Like the disparities between how the victims described their attacker and the actual physical appearance of the defendant. And the fact that the victims had spent most of their 911 call describing in detail the sports car their assailant used to speed away from the scene, while the defendant had no car and lived just a few short blocks down the street.
But other issues weren't just a matter of common sense. Surveys of actual jurors suggest that many laypeople are unaware of the often tenuous link between an eyewitness's confidence and his or her accuracy. As I discussed in my testimony, research studies and actual case history–like Jennifer Thompson's example above–demonstrate that even the most confident of witnesses make mistakes.
Moreover, eyewitness confidence is malleable. That is, it can change over time, becoming inflated as the witness learns that the person he identified is someone whom the police like as a suspect. Or after talking with other witnesses. Or after reading media accounts of the crime or encountering the defendant in court during pre-trial hearings. Research studies have demonstrated that even the verbal response of the police officer who administers a photo array or lineup can make witnesses feel more confident in their identification.
Indeed, in the case at hand, the same witnesses who once tentatively identified the defendant in a photo array at 50-60% confidence, were now testifying in open court that they were 80-90% sure that this was the guy.
But in preparing to testify, perhaps the most troubling aspect of the case to me was the photo array itself–that sheet with 9 photos that had been presented to the victims for the purpose of eyewitness identification. When I looked at the photo array, something about it seemed fishy.
I couldn't quite put my finger on it, but to my eyes, the image of the defendant seemed to leap off the page. Perhaps because his face filled more of the frame than the photos of the other men? Perhaps because his head was cocked at an angle unlike the others? I wasn't entirely sure what it was, but I did know that something about the array in this case just didn't look right. No, it wasn't as bad as the lineup pictured at left, but I wasn't sold that this array had been a fair and balanced test of eyewitness memory.
So the week before I testified, I decided to apply some good, old-fashioned psychological sleuthing to the particulars of the photo array in this case...