More Evidence on the Unreliability of Memory

Our podcast about false memory, “Sure, I Remember That,” featured the research of Steven J. FrendaEric D. KnowlesWilliam Saletanand Elizabeth Loftus. If you enjoyed that, you may want to check out Loftus’s recent TED talk about her research on embedding false memories in U.S. soldiers.  It focused on soldiers who had recently gone through “Survival School” training, during which they are “captured,” sent to a mock prisoner of war camp, and aggressively interrogated:

Psychiatrist Charles Morgan and his collaborators have been studying the effects of Survival School for a number of years. We worked together to conduct a study with the soldiers who’d gone through the training in which some would be fed erroneous information. Some have been exposed to misinformation about the “perpetrator” who conducted the hostile interview. They were showed a photograph of a man who was identified as the one conducting the interrogation, and were asked questions such as, “Did your interrogator give you anything to eat? Did he give you a blanket?” The trick was that the photograph was of a completely different person. When the soldiers were fed this misinformation, 84% of them later on went ahead and identified the person whose photograph was shown. All of them were, of course, mistaken.

The soldiers’ memories for other details could also be affected by misinformation. For example, some of them were fed misinformation about a weapon that was supposedly present during the interrogation. Later on, 27% claimed to have seen the nonexistent weapon. Others were fed misinformation about a telephone that was in the interrogation room. Later on, over 90% of them claimed to have seen that non-existent telephone.

As Loftus points out, this research demonstrates that people misremember even highly stressful events, despite the popular belief that “when something traumatic happens, it leaves a kind of imprint on the mind.”


I once sat on a two-week murder trial jury and the judge did not allow us to take notes based on his observations that juries collectively remembered all the details better when they were paying full attention to testimony. He was right. In deliberations we rarely disagreed on a review of the many facts in the case and if someone did, others supplied the correction. So under such conditions where stress is not acute, my experience is that memory is pretty good and certainly can be enhanced in a collaborative situation.


This seems just as fraught with errors as individuals' memory is. How can you know that the group interaction isn't somehow affecting your own recollection of things?


I would ask if the problem here isn't really false memories, but the difficulty of actually identifying a particular person from a photograph.

What is your name...

It is true...
That's why we need to leave our written record in an email, or in whatever, when our memory is still fresh.


I don't even trust my own memory. I once saw someone get out of their car, and then get stabbed in a fight over the parking space. I was on the passenger side of the parking space and I would have sworn (in court) that I hadn't seen the guy park,. But when talking to someone about it later that day I realized I knew the driver's-side door was a different colour to the rest of the car, which could only be explained if I saw him turn the corner and drive into the space. But immediately after the event if I had written down what I saw I would have sworn I didn't see him drive into the parking space. I have little faith in individual eye-witnesses, and not much more in eye-witness group-think.

What is your name...

Sorry, but I didn't mean that I had little faith in individual eye-witnesses. Actually, I have faith.
But I just wanted to emphasize that, with the written records like an email or a diary, talking about something from our memory, especially the old memories, could be more trustworthy. So, please keep writing for any event happend to you.


I know this for a fact. Two of us were robbed at gun point and neither of us remembered the guy had a mustache. Not a huge handle-bar type, but still it showed clearly in the surveillance photos.


But would this be a false memory, or simply a case of you failing to note a detail that was pretty well irrelevant at the time? As an example, I had a conversation over the fence with my cute neighbor earlier this evening. We talked about our dogs, her horse that she's training, the weather... But I couldn't tell you exactly what she was wearing if my life depended on it.


Not recognising the face of the person who interrogated you is not irrelevant detail. Failing to notice which weapon he was using, again not irrelevant.