In 2009, Officer Dina Hoffman dealt with a drunk in a car. Uncounted incidents later, she had to testify on that case. Her recollection was that the man was in the front seat, and that is what she testified. Surveillance video shows that the drunk was in the back seat. Officer Hoffman was charged with perjury.
At her trial, Hoffman acknowledged testifying incorrectly. But she said it was not willful.
"I made a mistake," she said this week. "When I testified [in the 2009 trial] . . . I believed I was telling the truth."
She and her attorneys said her erroneous testimony was based on a memory that had become cloudy nearly a year after the original arrest, and on having read another officer's incorrect report that was handed to her about 30 minutes before she took the stand.
Yesterday, the jury acquitted her of all charges after a prosecution by a D.A. from another jurisdiction and an eleven-hour deliberation. She has two years in on the job, months of this on administrative leave and under suspicion. She wants to return to work and get on with her life.
Memory and Crime Studies
At least two sets of assumptions permeate the research of memory. The first is that stress aids memory: it has a tendency to set events firmly in mind. Therefore, witnesses are able to remember the perpetrators of crimes they have seen. This has a lot of common utility--who else will you ask, except someone who saw? I know I have been able to identify wrongdoers in two incidents. It all stuck through grand jury or trial. And I still remember all of it.
The second is that traumatic stress causes memory to fail. I also know this first-hand--I was a pedestrian in an auto accident. I remember parts of it very clearly, and not the minutes I knew I would be hit and was hit. In a less traumatic story, in my first job, I remembered no isolated incident the first three months: just a blur. (That's what working for Dairy Queen will do to you.)
Memory Study Methodology: The T.V. Generations
In C.A. Morgan et al (2004), one of the first things the researchers note: Most studies of memory take place under lab conditions. In other words, a lot of words have been written about the memory of viewers of traumatic filmstrips. Then this work has been applied to real-life stress in regard to survivors of combat, crime, and tense situations. This is the kind of study that has been guiding law enforcement training and going against or influencing real-life wisdom.
Other industries study memory using film presentation--the advertising industry proves that non-stress memory is short. Consumers don't remember the name of the brand or what the actor said, unless there's a catchy "hook" and a lot of repetition of context along with the brand. The recognition devices repeat the "seen things" that are indeed, not very important. This research suggests that trauma aids memory. I would say preparation and cues aid memory, and one study I looked at seems to agree. You could even go so far as to say marketers are paid to increase stress for consumers and therefore build memory.
The U.S. military is also studying memory in regard to stress. U.S. Army survival school, for instance, is looking like a far better place to examine the accuracy of memory under high stress and life-threatening conditions than a laboratory with films of crime scenes. And their study suggests the opposite--that stress leads to inaccurate memory. Different arena. Different level of stress.
Better Study Conditions
In one study reviewed at link above, using abusive enemy prison guard scenarios, survival students were found to have better memory under low stress. Under high stress, they had to see more 'cues"--the same clothes on their guard/tormentor in order to identify him, or a series of memory encounters, starting with a photo array of cued or even un-cued images, before an un-cued lineup.
In addition, time has shown to change the effects of memory. High-stress memory needs to be consolidated, partly through REM sleep and partly through a chance to access all the parts of the incident. As early as 1992, Christianson reviewed memory study literature and hypothesized that "emotional events receive some preferential processing mediated by factors related to early perceptual processing and late conceptual processing". This means, I think, that the emotions have to be dealt with first, in a random stack, in the order of personal significance--before the narrative can be built for a testimony.
Plenty To Study vs. My Experience
There's a ton more articles on this, and I just parsed through a few. In my experience, memory in trauma works if you are prepared to have that trauma. Even realistic research is being conducted under controlled conditions (and what else can they do?) Therefore, I think anecdotal understanding is also very important to understanding memory and its falsities. I offer my own examples, small though they are:
In both incidents where I testified, I knew before the time of the stressful incident that I had trouble on my hands. I had time to describe my troublemakers to myself. When I was hit by a car, I had no warning, and thus do not remember. For that first job at Dairy Queen, I had no training or experience, it was busy, and I didn't remember individual events.
Somebody correct me if I am wrong--but I think stress and memory follows a kind of Bell Curve, although certainly more round than this one. At no stress and great stress, memory approaches zero. In middle degrees of stress, memory is better.
We don't remember the unimportant--and we can't remember the all-important without training, experience, and preparation/"cues". Thus, I imagine eyewitnesses who have advance knowledge of names and habits are better able to recount events because they have "cues" already embedded. In short, if you know Joe Blow beforehand, he doesn't have to be wearing the same shirt for you to recognize him. If you've been down a tough road before, you can cue yourself more quickly.
I believe that Officer Hoffman was relatively new at her job, and therefore even a routine event had a certain level of stress. I believe she may not have had time to "cue" herself in a breaking situation.
I also believe this drunk was an unmitigated jerk, and that ratcheted the stress. Because it had more significance to her than it might have had five years down the road, her memory was not as good. It was not reinforced by proper 'cues', (an inaccurate report by another; she should have been Advised of the Video by the D.A under disclosure in advance) and was made to pay for her inexperience.
Now she has this traumatic trial on her agency record and in her mind. If she can dominate that stress, and gets some luck, she should be okay. She'll probably write the best reports ever. Does she have time to do that on an ordinary shift in a hot summer?
This drunk demanded too much. He was in the wrong, and he tried to hide it by a counter-accusation. The court system on all sides failed her. [NOTE: In comments, it's considered that she also failed herself by not relying on her OWN report.] Those other parties to disaster will never know the consequences of this legal proceeding on her: if it affects her performance, her promotability, her timeliness and confidence. I am glad the jury saw clear to acquit her. I hope she got some further training on memory and preparation, even in some informal manner. Reassurance and discussion would make a difference.
If anyone sees inaccuracies in the above assessments, please write in. I will add information if it becomes available. In the meantime, best of luck to Officer Hoffman, and all her brothers and sisters out in the streets, keeping us safe.
Story in the Washington Post, linked above.
C.A. Morgan, et al, "Accuracy of eyewitness memory for persons encountered during exposure to highly intense stress," International Journal of Law and Psychiatry, 27(2004): 265-279. Linked above.
K. L. Keller, "Memory factors in advertising: the effect of advertising retrieval cues on brand evaluations. (Abstract). Journal of Consumer Research (1987). Linked here.
S. A. Christianson. "Emotional stress and eyewitness memory: a critical review." (Abstract). Psychological Bulletin, 112: 2 (1992, September), 284-309. Linked here.