Perception, Memory, and Justice

Psychological research reveals a lot of deeply troubling facts about human perception and memory that should make us very skeptical of eyewitness testimony. Of course, we don’t need science to tell us that if we witness something from too far away, or if it’s too dark, or if we’re intoxicated, then our testimony is not going to be very reliable. That just seems like common sense. But common sense may also lead us astray when it comes to the reliability of eyewitness testimony, and that is what a lot of the research on this topic is telling us.

In the criminal justice system, eyewitness testimony can make or break a case. Yet our eyes can deceive us and memory can be a fickle friend. So how much can we really trust eyewitness testimony?

Psychological research reveals a lot of deeply troubling facts about human perception and memory that should make us very skeptical of eyewitness testimony. Of course, we don’t need science to tell us that if we witness something from too far away, or if it’s too dark, or if we’re intoxicated, then our testimony is not going to be very reliable. That just seems like common sense. But common sense may also lead us astray when it comes to the reliability of eyewitness testimony, and that is what a lot of the research on this topic is telling us.

Imagine the following scenario. You’re waiting in line at the bank when armed robbers enter and make everyone lie on the floor while the tellers empty their cash drawers. A tense, dramatic event like this is sure to stick in your memory, right? After all, it’s not everyday you witness a bank robbery. Perhaps while it’s happening, you’ll make a special effort to remember details, as that could prove important if you’re ever called to testify. If the event is particularly stressful, later it might even replay itself over and over in your head, when you’re desperately trying not to think about it.

You get the picture. So, the question I have for you is this: How good do you think you’d be at identifying one of the perpetrators later on in a line-up? We’ll imagine for the sake of argument that none of them was wearing a mask during the robbery. The event took place in a well lit bank in the middle of the day, and let’s assume you weren’t drunk or high, and if you have bad eyesight, you were wearing your glasses or contacts at the time.

I’m guessing many of you think you’d do a pretty good job and would be fairly confident about later identifying a suspect in a line-up. But there are some other factors you may not have considered. For example, how close would you have to be to one of the perpetrators to be able to identity him later? And what effect does the angle at which you observe the perpetrators have on this ability? Recall, you’re lying on the floor during most of the robbery. So, how good would you be at identifying a suspect in a line-up (i.e. looking at him straight ahead) when you only saw him from a weird angle on the floor? And how much time would you need to observe his face in order to recognize him again later? Even if you knew how long the average person needs (unlikely, unless you study this stuff), how good are you at gauging time, especially during a high-stress event, like a bank robbery, when (your experience of) time might slow significantly? How does stress affect your ability to remember important details? Finally, if the robbery was carried out using guns or other weapons, how would this affect your ability to identify a suspect later?

Can common sense provide answers to these questions? Sure, we might all make guesses, but we might also get the answers wrong. Luckily, the scientists who research the psychology of perception and memory are providing answers and many of them are quite surprising.

Take the last question I asked about your ability to identify a suspect later—what effect would a weapon used in the robbery have on your ability? It’s not obvious that it would have any effect, right? It turns out that when a weapon is used, witnesses have an unconscious tendency to focus their attention on the weapon rather than on the faces of those holding the weapons. So, if you’re looking at the gun and not the guy, how are you going to be able to identify the suspect later?

This list of questions I asked focused on some different factors that could affect your ability to perceive or pay attention to details, such that they could be recalled later. But the psychological research also shows how easily your memory can be pulled off track. The way witnesses are questioned by law enforcement can influence and change memories. For example, if you’re asked, “Did you see the gun?” rather than “Did you see a gun?” that can affect what you think you did or did not see. In the first case, which uses a definite article—“the”—that there was a gun is already suggested in the question, whereas in the second case, which uses an indefinite article—“a”—there is no such suggestion. As far as the police know, there may or may not have been a gun. And this can actually make a difference to a witnesses memory.

That’s because memory is something that is constructed from little bits and pieces. It’s not, as we might imagine, simply a replaying of events, like watching a movie. How we put together these bits and pieces can be manipulated by external forces, whether intentionally or not. But if law enforcement don’t know about the various ways memories can be influenced, they may ask the wrong questions and end up with witnesses whose memories are not a reliable indicator of what actually happened.

Can’t this all get sorted out in court? Isn’t that why we have an adversarial system? If a witness for the prosecution is unreliable, the defense can cross examine and cast doubt on the testimony.

In an ideal world, that would be the case. But lawyers are not scientists or psychologists and they might be just as ignorant of the research on perception and memory as the next guy. And besides, jurors have all sorts of unconscious cognitive biases that can be hard to overcome. Take, for example, how confident a witness is in identifying a suspect. We’ve already mentioned some of the ways in which we may not be in a position to get a good enough look to form the right kind of memories to begin with, even if we’re convinced we are. And then those memories can be manipulated. You could tell jurors about all this research, but the fact is that if a witness takes the stand and is very confident about what and who they saw and what they remembered, jurors tend to believe them. Yet the research shows that there is no positive correlation between confidence and accuracy. The confidence of a witness does not indicate that they are reliable or that we should believe their testimony.

This is why we now hear so many stories about convictions based on eyewitness testimony eventually being overturned because of new DNA evidence, sometimes years or even decades after the original conviction. Did you know that about 73% of all convictions overturned through DNA evidence were originally based on eyewitness testimony? It’s quite shocking and it’s definitely not justice. Innocent people’s lives get destroyed because other people think they see something, think they remember something, but they can be very, very wrong.

So what changes do we need to make to the criminal justice system in the light of this new research on memory and perception? When fingerprint and DNA evidence were introduced, it revolutionized the justice system. Maybe it’s time for another revolution, this time based on the latest psychological research. The question is: how do we go about integrating the findings of this research, especially when it does not give us black and white answers, the way that fingerprint or DNA evidence might. 

To learn more about this fascinating research, tune into this week’s show with our guest Daniel Reisberg, psychologist and author of The Science of Perception and Memory: A Pragmatic Guide for the Justice System.