Thursday, January 31, 2008

A Brief Word on Eyewitness Memory


Fig. 1-1: Your new masters.

Your ability to remember things is slightly worse than a chimp’s and only slightly better than a goldfish, who will still be better at complex motor memory than you. Don’t even get me started on how we stack up to birds – it’s just depressing.

Sorry to get off on the wrong foot like this, but it’s true. Don’t worry, I’m just as stupid as you! In fact, due to my affinity for video games, comic books, and other things designed to inspire ADD in America’s youth, I probably have a worse memory than you. I don’t remember what I had for breakfast this morning. What month is it?

In a recent issue of Forensic Examiner (a forensic publication of some note), an experiment was described where people falsely mistook memories of an unarmed person for memories of a person holding a pistol, even after getting to view pictures involving the person in concentration for several seconds (staring at a photo for five seconds is longer than you think, try it). Even worse, they routinely confused memories of pistols with memories of assault rifles, and if it was suggested to them that an empty picture displayed something somehow menacing (for example, having a street corner described as a "crime scene"), they would later invent new details to confirm that implanted memory.

The same effect was shown to happen in even more extreme circumstances. In one experiment, people were shown pictures of famous events that had been altered, and people quickly responded by saying that that's how they themselves were 'positive' the event took place. At the same time, people who had attended a peaceful protest were shown pictures of the protest edited to contain violent content and images that implied that it was not peaceful. Protestors shown those photographs quickly changed their own memories of the events to suit what they were shown.

In other words, concerning my earlier quandary about what I ate for breakfast this morning, well, if you told me my breakfast consisted entirely of cat food and asbestos enough times, eventually I’d say, “sounds about right!”

In an article I read earlier this week – I could only find the abstract online – the authors cited several studies that indicate just how absurd it is to believe that people can hold accurate memories of events in the long run. One particularly alarming segment, citing such psychologists as LL Jacoby and CM Kelley, read,

“People can misattribute the familiarity caused by previous experience to a current event’s perpetual salience, but under different circumstances can also misattribute the salience of a current event to its past history.”


For example, if you’re, say, William Rodriguez and join a movement whose goal it is to promote the idea that there were explosives used in a supposed ‘controlled demolition’ of the World Trade Center towers, and these people decide to befriend you because they want you to help them sell DVDs, you just might start to really, really believe what they want you to have experienced. Essentially, after you’ve been involved in an emotional, vocal organization like the 9/11 denier movement for several years, it doesn’t matter what really happened: your emotional, gut reaction takes over and tells your brain how to interpret events. Will tweaking this memory help “the cause?” Well, so be it, says your dastardly subconscious.

In another study (I couldn't find a version online, so its full citation is available at the end of this article), on the day the OJ Simpson verdict was returned, people were asked to write down where they were. Thirty-two months later, they were asked to recall that same information. 71% were at least slightly wrong; 40% were completely wrong. And, regardless of being right, kinda wrong, or totally wrong, people were equally confident of their memories. Studies like this have been going on since 1977 at the latest (see end of article).

Several similar studies have been conducted concerning what people remember about September 11, 2001. I could only find one that showed that people could recall what they did and where they were on that day with a consistently good accuracy – the rest (P. Lee, N. Brown 2003 ; D. Shachter et al., 2004 for example ) showed a distinct loss not only in information accuracy, but overall information content. And even that first study achieved consistency only by using a method of assessment that was keyed to each individual subject.

This is because, as the article above demonstrated and as another gets into more technically, memory is contextual and associational. That is, it becomes associated with an emotional state as it sloshes around in your lateral nucleus in search of contextual meaning. And, as the article cited at the beginning of this paragraph reads, once a memory becomes part of a conditioned emotional response it becomes “relatively permanent.”

I took part in what is now one of my all-time favorite demonstrations of human fallibility. The experiment is a brief video, which you watch and then describe according to the assigned goal. You have only one goal while watching this video: count the number of passes of the basketball by people wearing white t-shirts. You can watch it here.

Did you count seventeen passes? Congratulations – you fell for it! Watch again.

You kind of miss the effect on a tiny little Java applet, but if you can get the video onto a larger screen, like a classroom projector, you’ll find that about 90% of people will fail to notice. Why? Because, even in the formation of memories, you’re only really going to commit things to memory that were relevant to what you felt you had to pay attention to at the time – for example, counting basketballs rather than looking for gorillas. Or if you’re running down a flight of stairs to get out of a burning building, any sort of rumbling, booming, or banging noises you’re going to hear are going to be secondary to any immediately-relevant information that will help you get the f*** out of that building.

This is why there is simply no way to be nice to people who base claims of 9/11 conspiracies on the eyewitness memories of people who were there (never mind the quotes that have been taken out of context or simply lied about). Someone with an associational, vague, attention-based memory of a traumatic event that happened almost seven years ago is going to remember things about as well as I remember what I had for breakfast this morning.

The bottom line is that, in an investigation replete with forensic, video, and photographic information pertaining to what happened like 9/11, it should be regarded as absurd to pretend that postulations based solely on later-recalled eyewitness testimony are of any real value. There is no doubt that people had vivid, horrible, and desperate experiences on 9/11 that help demonstrate the heroism that ordinary people are capable of in extraordinary circumstances. However, that does not mean that anecdotes are more imporant to a forensic investigation than what can be ascertained by sources less faulty than the highly fallible human mind. Many people became heroes on 9/11, but none of them became supermen. As vital and necessary as the stories of 9/11 are to all of us, none of them should be taken as perfectly-accurate, 100%-recalled play-by-play retellings of what happened. Eyewitness memory is secondary to evidence that can be assessed empirically and categorized by actuarial methods into predictive models.

Other cited experiments:

-The OJ Simpson memory experiment: Schmolck, H., Buffalo, E. A., & Squire, L. R. ( 2000). Memory distortions develop over time: Recollections of the O. J. Simpson trial verdict after 15 and 32 months. Psychological Science , 11 , 39-45.
-Original flashbulb memory study: Brown, R., & Kulik, J. (1977). "Flashbulb memories". Cognition 5: 73–99.
-From Scientific American MIND: LeDoux, J. (1994). "Emotion, Memory, and the Brain." Memory and Learning: 105-117.

No comments: