The Memory Illusion

by Dr Julia Shaw

  Porter and Peace found that the memories of the traumatic events were highly consistent over time, remaining virtually unchanged in almost all features. It also turned out that in comparison to the highly positive life experiences, the memories of negative experiences were significantly more stable over time. These findings suggest, as does Porter and Birt’s research, that memories of trauma are special, just not special in the way that many people assume, indeed often being better than other kinds of memories.

  This is supported by a 2012 review of the literature by the Svein Magnussen and Annika Melinder from the University of Oslo in 2012:8 ‘The current evidence from systematic and methodologically sound studies strongly suggests that memories of traumatic events are more resistant to forgetting than memories of mundane events.’ This is both good news and bad news, as it seems to increase the likelihood of accuracy for eyewitness and victim testimonies (though they will still be up against all the problems discussed previously), but it also means that traumatic memories we might rather forget could haunt us forever. Interestingly, this persistence of memory seems to occur not only with traumatic events we directly experienced ourselves, but also indirectly experienced traumatic events that we have heard about repeatedly through the media.

  Flashbulbs

  ‘Where were you when 9/11 happened?’ was a tremendously popular question in the 2000s. Similarly, in the decades before, people might ask each other’s whereabouts when the Challenger shuttle exploded or when JFK was shot.

  What these types of questions imply is that we have the capacity for immediate powerful recollections of the circumstances we were in at particular significant moments. These are sometimes referred to as flashbulb memories. They are detailed and vivid, and typically involve recalling the situation in which a piece of historically important news was heard, along with a detailed recollection of the event itself. Rememberers often mention the informant who shared the news, what they were doing when they heard the news, what they were wearing, and what they were thinking, feeling and saying. I mentioned this phenomenon in a university lecture recently, and one of the students in the audience said that on finding out that the first plane had flown into the twin towers in New York on 11 September, 2001, she had made a crude joke about the inability of pilots to fly straight – a joke she had felt self-conscious about ever since.

  In 1977 Harvard University researchers Roger Brown and James Kulic9 investigated these kinds of memories. They sent out a questionnaire to 80 people to ask about what made them remember important historical events such as assassinations, highly newsworthy occurrences and personally important experiences. From the questionnaire responses, they concluded that many people have memories with considerable perceptual clarity for important historic events. In other words, people could report more correct details with higher confidence for certain kinds of events, with these events having three main characteristics.

  First, the event needed to generate a high level of surprise. It could not be a trivial or expected event. Second, the event needed to carry important consequences for the person or for people in general – referred to as having a high level of consequentiality. This can be an important consequence for them personally, or for society. For example, 9/11 may not be personally relevant to a particular individual, but it was quickly deemed important for society, making it an event with a high level of consequentiality. Finally, the event had to generate high levels of emotional arousal – the individual needed to experience fear, sadness, anger or some other strong emotion. Brown and Kulic argued that without these three conditions being fulfilled a flashbulb memory could not occur.

  They went on to speculate (without, it must be said, any reasonable scientific backing) that the reason we can have these kinds of memories is because of a unique biological mechanism that creates a permanent hard-wired record of the event in the brain. According to them, rememberers of flashbulb memories display extremely high confidence, often declaring the details with definitive statements – ‘I was definitely at home.’ ‘I remember it so clearly.’ Take these examples from the False Memory Archive, a project by the artist and Wellcome Trust Engagement Fellow A. R. Hopwood:10

  Sam: ‘I have a very vivid memory of watching the Challenger shuttle disaster in high school. I was standing in the Science/Media library in my school, and there was a TV on a tall metal rolling stand – the kind that was rolled into classrooms for presentations. Several of my friends were there. I can remember exactly how I was standing in the room, where I was, the angle I was seeing the screen from, the shock of others in the room.’

  Sue: ‘I remember my housemate coming home to our shared house, telling me that she heard on the radio that a plane had crashed into the World Trade Centre. I have a clear memory of that house and that she went to her bedroom, and I turned on the TV news.’

  Both those recollections are extremely vivid and detailed, and are told with a confidence and conviction that gives them the ring of truth. In 2014 Martin Day and Michael Ross at the University of Waterloo11 published a study that further explored rememberer confidence in flashbulb memories. Instead of a questionnaire, they conducted interviews. The first interviews took place shortly after the death of Michael Jackson, and they asked participants to recall how they found out about the death, where they were, their confidence in their memories, and whether they thought their memories would be durable. Eighteen months later, they asked the participants to once again recall the details of their whereabouts when they found out about the death, and asked them once again about their confidence in the memory.

  And what did they find? I’m afraid I have to pull the rug out from under you. The researchers found that these memories often lacked consistency, the descriptions of them changing between the two memory interviews, despite confidence in the accuracy of the memory remaining very high. This suggests that flashbulb memories may not be as permanent or accurate as Brown and Kulic originally proposed, and that people are overconfident in the accuracy of their recall for these kinds of events – which supports the idea covered in earlier chapters, that confidence does not necessarily imply accuracy. The original Brown and Kulic study had significant methodological problems, including having a small sample, relying only on self-report, assuming that participants’ reported memories were accurate, and making unfounded assumptions about how the brain works. Despite pitfalls, their research led to a large amount of further study of these flashbulb memories, resulting in an erroneous sentiment that memories for important historical events are safe from corruption. In reality, memories of witnessing cultural events may not be as strong and protected as we often assume. To demonstrate that, let’s revisit those accounts from Sam and Sue, which I hope you will agree were detailed and plausible. The problem is, I’m afraid they both had a little more to say:

  Sam: ‘The problem? The Challenger disaster happened two years after I graduated. When it occurred, I wasn’t in that high school, or that city – I was living in another part of the country. This memory is completely real to me, and yet I know that it’s false – it didn’t happen that way. I have no idea where I really was when, or if, I saw the Challenger disaster on TV – but I know for sure I wasn’t where I “remember” being.’

  Sue: ‘… we had stopped living in that particular house three years before 9/11. At the time of 9/11 we were sharing a house again, but in a different house. My memory has mixed up the death of Princess Diana and 9/11; same housemate, but wrong house.’

  This process of realising that a particular memory is inaccurate or impossible is called recollection rejection, a term coined by memory scientists Charles Brainerd at Cornell University and his colleagues in 2003.12 What is amazing about instances when we reject our memories in this way is that it does not necessarily mean that we will no longer have the memory, it simply means that our confidence that it happened diminishes greatly – or disappears altogether. More often than not, however, we are not confronted with contradictory evidence for our memories, an
d can come to accept accounts as our personal reality, even if they make no logical sense.

  Regarding trauma memories, there is also potential for the integration of grossly false information about events that we have, or think we have, experienced ourselves. Even our highly emotional memories can be totally false. How do I know this? Because it’s part of my job as a researcher to show that even our most vivid memories may be up to no good.

  Memory hacking

  I am a memory hacker. I get people to believe things that never happened.

  Usually when I tell people that this is part of my job they ask ‘But why?’ The answer is that I believe that through generating memories in a lab-based experiment we may be able to begin to figure out how such memory illusions operate. We cannot hope to prevent such errors from happening until we fully understand why they occur.

  I’ll explain what I actually do. It’s not hypnosis, or torture, or anything like that – it’s simple social psychology. I apply what has been learned from decades of research, and basically do the reverse of the careful, non-leading questioning which I supervise when working with the police. I deliberately provide what I think are the ideal conditions for the development of memory illusions. Let me take you through the process, one step at a time.

  Step 1: I source adult participants for an ‘emotional memory study’, and get them to provide contact information for some informants – people who know them really well, such as their parents.

  Step 2: I contact the informants and ask them to describe emotional experiences that the potential participant may be able to recall from a particular time in their lives, between the ages of 11 and 14. At this point I also harvest information about who the potential participant was best friends with at the time and where they lived.

  Step 3: I filter the participants so as to include only people who have not actually experienced any of the emotional events I am planning to implant, but who have experienced at least one other emotional event. I invite these people to participate in the study.

  Step 4: The participants come in for the study. They are under the impression that it is a study on emotional memory, but have no idea that the actual premise includes the possible implanting of false memories. I get their consent to participate in an emotional memory study. The deception I use, in that I don’t disclose that one of the memories they may recall will be false, is of course in compliance with the permissions granted to me by my university research ethics board. I begin to ask the participant, in a structured way, about their memory of a true emotional event which I have learned about from the informants. It might be being bullied at school, fainting on vacation or any other emotional event. As a result of this I gain credibility as someone who knows about emotional memories that they have experienced.

  Step 5: I introduce a false event, telling the participants they did something that I know they did not actually do. The event in my most recent experiment, published in 2015 with Professor Stephen Porter and conducted at the University of British Columbia,13 involved telling participants that they had committed a crime with police contact – assault, assault with a weapon, or theft – or had experienced another emotional event – an animal attack, a bodily injury, or losing a large sum of money and getting into trouble with their parents. The script we used for the police contact scenario was as follows:

  Okay [participant’s name], thank you for telling me about the first event. You did a great job. The other event your parent[s] reported happening was an incident where you were in contact with the police. I will be asking you about this next. On the questionnaire, your parents stated that when you were about [age] years old, you [false event]. It happened in [place], in the fall and you were with [friend or relative] when it happened.

  No other information was divulged about the supposed event.

  Step 6: After being told that this supposedly happened in the first interview, participants initially correctly say something along the lines of ‘I don’t remember this’, so I offer to help them out. I offer to do a visualisation exercise with them. In this exercise I get the participants to close their eyes and picture what the event would have been like. Little do they know that I am doing this to get them to access their imagination instead of their memory. After they do this visualisation exercise during the first interview, they normally don’t have many details. At this point I send them home with the instruction that they are not allowed to talk to anyone about the study, that they should try to visualise the memory at home, and that they should come back to see me in a week’s time.

  Step 7: The participants return to my lab a week later, and I ask them again to take me through their recollections of the true memory. Then I ask them about the false event. At this point, many participants begin to ‘remember’ and report details of it. ‘Leaves were falling. Blue skies. I stole a CD. I punched a girl for teasing me. The policeman had brown hair.’ I encourage them and say they are right on track – positive reinforcement. I also repeat the visualisation exercise, to get them to imagine more details that they can mistake for memory details. I send them home again, tell them to try to get more details, and to come back for a third time, a week later, where I repeat the process – true memory recollection followed by false memory recollection with visualisation – for a final time.

  Step 8: After three interviews, I harvest the full-blown false memories. At this point many of my participants are divulging a tremendous number of details about an event that never happened, talking about them with confidence. It’s like memory magic.

  You may not think this would work on you, but the statistics suggest otherwise. In this particular study we found that 70 per cent or more of participants in both the criminal and other emotional conditions developed full false memories. I defined full false memories by a number of criteria, including the participant at least reporting ten details related to the event, and saying at the debriefing that they believed that the event actually happened. Many of the accounts given were richly detailed. Here an excerpt from one participant’s transcript; we’ll call her A:

  I remember being so shocked when the cops came. That was bad. That was bad.

  What was so bad? Here is an earlier part of her account:

  That’s what set me off, [she] was calling me a slut … Because I was a virgin … I know she wasn’t close to me. And I think we were like, we were like, following her, and taunting her … And I got pissed off. And it wasn’t a huge rock. It wasn’t … but I did find it, it was a decent sized rock and I threw it at her head. I threw it at her … I ended up going home. S took off. And then, I remember being in the house, I think we were eating dinner, and uh, and then we … the doorbell rang and my mom went and answered the door, and then I remember mom yelling for me to come to the door, so I went, and there’s two cops standing there.

  You might still not be convinced. You may think only vulnerable people would fall for this. Certainly, if we use poor or deliberately misleading interview techniques on individuals who are highly compliant, very young, sick, or low functioning, we might expect these kinds of results. In this study, however, I specifically selected participants who were not vulnerable in any of those ways. They were ordinary university students.

  That means that the most compelling explanation for the results we got is that even people like these are susceptible to social pressures and bad memory retrieval techniques that helped them to imagine things, and helped them to mistake those imagined things for real experiences. This is all in line with ample research conducted by other teams, who have successfully implanted other kinds of emotional false memories such as spilling a bowl of punch on the bride’s parents at a wedding, published by Ira Hyman and his colleagues at Western Washington University in 1995,14 or of being attacked by a vicious animal, as generated by Stephen Porter and his colleagues at the University of British Columbia in 1999.15

  In two follow-up studies I conducted16 I showed videos from my false-memory-of-crime study to new participants. Thes
e new participants did not know that some of the videos they were asked to watch involved false memories. I had each participant view a recording of the same person recalling first a true memory and then a false one. In both studies, participants were no better at correctly guessing the true and false memory than if they had picked randomly. The evidence seems to be that these memories feel real to rememberers, and they therefore look real to others – they can become part of the rememberer’s personal past, whether they actually happened or not.

  In the wild

  But perhaps you’re still not convinced. Perhaps you think fake memories generated in a lab situation don’t prove anything about our experiences in the real world. If that’s your opinion then you are not alone – which is why false memories have also been studied ‘in the wild’. Rather than creating scenarios themselves, researchers can piggyback on real-life events that they know already cause the kinds of conditions they require – being highly negative and emotional, generating a level of stress that we should never seek to generate in a lab environment.

  An example of this kind of highly stressful situation is navy training. Let’s imagine the scenario. You are a 26-year-old member of the US Navy, and as part of your survival training you’ve been put in a simulated prisoner-of-war situation. You have just finished a four-day evasion exercise in the wilderness. You are tired, you are sore, you are hungry. And now, much to your dismay, you find yourself held captive in a prisoner-of-war camp. Technically you know this is a mock scenario, but the types of stresses you are going to be subjected to are directly modelled on the experiences of actual prisoners of war.

  For the interrogation you are locked into a room with someone you have never met, one-on-one. You undergo a vicious half-hour where you are subjected to physical abuses intended to make you talk, including slaps to the face, abdominal punches, being slammed into the wall, and being placed in stress positions. You are required to stare directly into your interrogator’s eyes for much of the ordeal. His face is not covered. You can clearly see him at almost all times. Then he places you in isolation. During this prisoner-of-war exercise you are placed under acute stress for a total of 72 hours. This is, by most people’s standards, a potentially traumatic experience.