The Memory Illusion

by Dr Julia Shaw

  Let’s start with the memory processes involved. Unless you are currently looking into a mirror, your perception of what you look like is a type of memory. It is a memory not only of when you last looked in the mirror earlier today, but of all other times you looked in a mirror or looked at photos of yourself. This means you almost certainly have a composite image of, say, your face in mind when you think of yourself. The problem is that this patchwork memory of what you look like never stood a chance because it can never actually exist in reality. You cannot look today like you did every day until today. Ageing alone makes that impossible, never mind everyday blemishes and style changes. This helps us to understand why, in response to certain photos, we say, ‘That’s a bad picture of me!’ Often what makes it a bad picture is simply that it is at odds with what we think we look like, at odds with our memory of ourselves.

  In 2008 psychology researchers Nicholas Epley at the University of Chicago and Erin Whitchurch at the University of Virginia18 published results from a series of studies on how good we are at identifying ourselves, in an article called ‘Mirror, mirror on the wall’. They took photographs of their participants and digitally altered them, producing both more and less attractive versions of the original photo by morphing it to fit a standardised highly attractive or highly unattractive face. The photos were morphed to varying degrees, to give a continuum of faces.

  Two to four weeks later the researchers presented the various versions, including the original photo, to the participants and asked them to select their unmodified photo out of the series. Most participants selected images that had been morphed with the attractive face to be between 10 and 40 per cent more attractive than the original photo. Participants picked the original, unedited, photo of themselves less than 25 per cent of the time for all conditions. Both men and women clearly thought they were significantly better looking than they actually were, and systematically picked enhanced versions of themselves.

  What about identifying others? When participants were asked to identify the faces of friends out of a similar array of modified photos, the researchers found that participants showed a similar bias here as with their own faces – it seems we think that our friends are also more beautiful than they actually are. However, the same is not true for a stranger’s face – according to this study we seem pretty good at correctly identifying people we have known for only a short period of time. On average, participants picked photos of themselves that were 13 per cent more attractive than their unedited pictures, picked photos of their friends that were 10 per cent more attractive than their unedited photos, and for strangers picked photos that were 2.3 per cent more attractive than actuality.

  We could chalk this bias up to the idea that we generally think we are above average, as discussed in Chapter 6. Or perhaps we could say that because we know ourselves and our friends well, we see an inner beauty reflected outwards. But there is also a third alternative: you have distorted your self-perception, and your perceptions of those closest to you, over time.

  This actually has less to do with automatic memory processes and more to do with vanity. We all choose the best photos of ourselves and those closest to us to present to others, paying particular care with photos that we put online or into formal documents. This is precisely the problem – by only including the best photos, with our most dolled-up faces, in photogenic situations, we are setting ourselves up to be bad at recognising ourselves on regular days.

  An article published in 2015,19 spearheaded by psychological scientist David White at the University of New South Wales and supported by the Australian passport office, looked at how good we are compared to strangers at knowing what we look like. In this study the first set of participants were asked to download ten photos of themselves from Facebook, and to rank how much each photo looked like them from best likeness to worst likeness. They were then asked to take a one-minute webcam video of their face and two additional still photos.

  These were then used in a face-matching task by a second set of participants who did not know the first set. They were asked to compare the Facebook photos with the webcam video taken during the study and to also rank them on likeness, saying that the photos and videos were either very similar or not similar at all. The strangers picked different ‘best likeness’ pictures than the participants did themselves. The question became: Who is better at knowing what we look like, strangers or ourselves?

  This was addressed by looking at a third set of participants who were asked to match the Facebook photos that had been rated as having a ‘good likeness’ by the people in the photos themselves, and those picked by the participants involved in part two. This third set of participants was more accurate at matching a person’s Facebook photos with the two stills taken in the study when the Facebook photos had been selected by a stranger; photos selected by strangers led to a 7 per cent increase in accuracy in matching photos of the same person. In other words, it turned out that the second set of participants were better at identifying what the first set of participants looked like than the first set of participants were themselves. This might be taken to indicate that strangers know what we look like better than we do ourselves. According to White’s team, ‘It seems counterintuitive that strangers who saw the photo of someone’s face for less than a minute were more reliable at judging likeness. However, although we live with our own face day to day, it appears that knowledge of one’s own appearance comes at a cost. Existing memory representations interfere with our ability to choose images that are good representations or faithfully depict our current appearance.’20

  Research like this seems to indicate that we come to believe that we actually look like the person we present ourselves to be on Facebook and other platforms – we come to internalise our own online facade.

  Better together

  Our socially constructed memory, biased as it may be, is not all smoke and mirrors. Most research in cognitive psychology shows that recalling memories collaboratively is usually disruptive and inhibits accuracy. However, Celia Harris21 and her team from Australia wanted to change this. In 2011, they published a fascinating article that set out to challenge the existing research methodologies, which often focus on strangers. They wanted to focus instead on people who knew each other very well, and how they recalled shared memories – both of personal and impersonal events.

  In the first study of its kind, the researchers interviewed 12 couples who had been married for 26 to 60 years to see how they remembered collaboratively. They asked them to come in for two sessions, two weeks apart. In each interview the participants were given a list of random words to remember. They were also asked some personal details, including names of people they knew. In session one each spouse was asked to recall the impersonal list of random words and the personal list of names on their own. In session two the couple was interviewed together and asked to do the same thing. It was found that some couples displayed collaborative inhibition, meaning that they actually reduced the quality and amount of information that they were each able to remember, while others showed collaborative facilitation, meaning that they helped their husband or wife remember more. Whether partners helped or hindered their spouse’s memory depended on how they remembered together.

  The researchers found that there were both memory-diminishing factors, which were indicated by a lack of cohesion between the partners, and memory-enhancing factors, which were the result of an interactive style of recall. For example, the couples that helped each other remember in an interactive way might have a conversation something like this:

  Participant 1: ‘What was his name again? Ed something …’

  Participant 2: ‘Yes, Ed Sherman.’

  Participant 1: ‘So, Ed Sherman was with us at the dinner.’

  Participant 2: ‘The dinner to celebrate Nancy’s birthday.’

  Participant 1: ‘Yes, with the giant cake that tasted like cardboard.’

  In doing so they would fill in each other’s memory gaps and work as
a team to rebuild the story.

  Wanting to go beyond recalling lists of names is work spearheaded by my close friend Annelies Vredeveldt at the Vrije University of Amsterdam.22 While I cannot say that I am impartial here, I think her research is absolutely fantastic. In one particular study, from 2015, she recruited couples when they were leaving a theatre, having just seen a performance of Bossen, a play which includes a three-minute scene in which one of the characters murders his father and then rapes his twin sister. It was the recall of this scene that Annelies and her team decided to focus on. The participants they recruited had no idea that they were going to be asked to participate in a memory study later, so they could not have prepared for this eventuality in any way or watched the play with it in mind.

  The couples, who had known each other for an average of 31 years, were interviewed about their memories of the play, first alone and then together. It was found that working together did not necessarily help the participants to remember more about the target violent scene, but they did seem to make fewer errors when working together than when they were interviewed individually. When they were interviewed alone, each spouse made on average 14.6 errors, while when they were interviewed together, each made on average 10 errors. Annelies calls this effect, when we report fewer details that are inaccurate, ‘error-pruning’. It may be due to our being more cautious with what we share in company, being less likely to report details we are unsure about. Corroborating Celia Harris’s findings from earlier, Annelies also found that certain strategies were better for helping spouses remember: acknowledging what the other person was contributing, repeating and rephrasing their insights, and elaborating on each other’s statements.

  According to Annelies and her team, ‘Taken together, our findings suggest that, under certain circumstances, discussion between witnesses is not such a bad idea after all.’ This is good news because most of our memories seem to be in some way collective. According to Elin Skagerberg and Daniel Wright,23 a whopping 88 per cent of real-world eyewitnesses have co-witnesses, many having more than three others watching an event, and over half discussed the event with at least one of the others who were there. This is much like other events in our lives, where we have at least one friend by our side with whom we immediately discuss what happened.

  So, what do we do with these results? We know from previous chapters that remembering together can have highly problematic repercussions. In situations where memories are shared with others, we can steal them, distort them, or create entirely new complex false memories. Researchers like Vredeveldt agree with such problems of plasticity, but they also argue that under certain circumstances the likelihood of our generating false memories and making errors may not be as dire as in others. In particular, when we know someone really well or use supportive and collaborative memory retrieval strategies we may be at a lower risk for misremembering. However, in practice, the implications of this research are unknown. We haven’t studied the benefits of remembering together in nearly enough detail or depth. When working with strangers it’s not clear whether we will be able to reproduce these favourable conditions or whether the effects of co-witnessing events and reminiscing about them will always be problematic and lead to memory distortions.

  For now, the research strongly suggests putting your virgin memories on paper before they have the chance to be tainted by social processes. After recording a memory in a place that can be accessed later, then you can go ahead and share it with others. Just beware that at this point your friends and family can just as easily help or hinder how your memory develops.

  A world of witnesses

  As soon as we share our lives on our social media sites, we find ourselves involving an almost infinite number of people to be the co-witnesses of our lives. This has irrevocable implications for our memories, for better or worse.

  For better, in a very basic sense, remembering life events through social media is going to enhance memories for those particular events. In the scientific literature this is sometimes known as ‘retrieval practice’,24 which means that simply recalling information enhances our memory of it. Studies of the effect have shown that simply recalling something can lead to better retention of information than studying the same information for the same amount of time. This line of research suggests that ten minutes of reminiscing may be better for your memory than ten minutes of studying.

  Social media also gives us an unprecedented ability to access corroborating evidence for our memories. By Instagramming our food we document where we had lunch and what we ate. By tweeting our opinions we can go back to see if and how we have changed our attitudes over time. By adding friends on Facebook we can see when we first met someone and how our relationship with that person has evolved. We have an astonishing amount of personal data that allows us to track, and confirm, many of our memories. In the case of false memories, this can of course be tremendously useful. With the world as our witness, if we ever get into trouble we can default back to the internet for proof of what happened.

  But, from trying to divide our attention more, to having the potential for misinformation to come from virtually anyone, to putting less effort into remembering facts because we can just Google them later, there is also a far more problematic side to social media memory. Having intrusive social media prompts and notifications constantly reminding you of certain events and thrusting more and more information at you also has the potential to severely distort your reality.

  This is in part related to the retrieval-induced forgetting effect we discussed in Chapter 3. Every time we remember something, the network of cells that make up that memory becomes active, and that network has the potential to change and lose the details about which we do not directly reminisce. For example, say you are reminded on Facebook of a vacation. The prompt may be a single photo of the event with a caption. As you remember the particular moment in which the photo was taken, it is possible, even likely, that you are forgetting related and unmentioned information of other things that happened during the day.

  Of course, it’s not just social media that can have this memory-altering effect. Rehashing memories in any situation has the potential to distort them. What is different about social media is that the prompts are being selected from your online persona so they already represent a distorted, social-media appropriate, version of your life. This amounts to a double-distortion – distorting the memory in your brain with a previously distorted memory from your online persona.

  By having social media dictate which experiences count as the most meaningful in our lives, it is potentially culling the memories that are considered less shareable. Simultaneously it is reinforcing the memories collectively chosen as the most likeable, potentially making some memories seem more meaningful and memorable than they originally were. Both of these are problematic processes that can distort our personal reality.

  How do you know whether you are recalling your experienced reality or your online, crafted, reality? You probably can’t tell the difference, as social processes of remembering become magnified and have the potential to infiltrate in ways that were previously not possible. Social media and our ability to connect with others is introducing a fascinating new set of challenges and benefits that memory researchers are only just beginning to explore. It’s a brave new world, and we can all look forward to exciting developments in how we remember together.

  9. TOOKY PULLED MY PANTS DOWN

  Satan, sex, and science

  Why we can falsely remember traumatic events

  SOMETIMES FALSE MEMORIES can turn into real horrors.

  When we find ourselves at the intersections of the concepts we have discussed throughout this book, we can end up in a perfect storm. A storm of misconceived assumptions about how memory works, about faulty memory retrieval techniques, and overconfidence. We can come to spin amazingly complex and fantastical tales of horror that have tremendous implications for our loved ones, for ourselves, and for justice.

  To show
you what I mean, let me tell you about a case that I consider highly problematic. This story has been recalled many ways by many people, but here is a version that I have pieced together and believe represents the basic outline of what most people would probably agree happened, much of which is supported by a detailed account written in 1995 by investigative journalist Charles Sennott.1

  The case begins in the spring of 1984. We are in Malden, Massachusetts. A four-and-a-half-year-old boy named Murray Caissie lives here. According to his mother, Murray has been wetting the bed for quite a while, but recently it has been getting worse. That summer he begins to use the same infantile babbling as his 16-month-old brother and acts up more often. He is caught in sexually suggestive play with his little cousin. His mother is getting increasingly concerned about his behaviour.

  One evening Murray is having trouble going to pee, and begins to cry. His mother begins to wonder whether maybe his troubled behaviour is a result of something bad that happened to him. This worry begins to escalate sharply, and she apparently wonders whether her son may have even been the victim of sexual abuse. This concern may stem from the knowledge that her brother was abused when he was a young boy.

  Murray’s mother asks her brother to talk to Murray. Her brother tells Murray the story of the time he was molested when he went away to camp as a child. He says to Murray that he should tell him if anything like that ever happened, if anyone ever did anything to undress him or make him do things that he did not want to do. Murray thinks about this, and then he tells his uncle about being taken to a room on his own where Tooky ‘pulled my pants down’. Tooky is the nickname of a man called Gerald Amirault who works at the Fells Acres Day Care Center, which Murray attends. Amirault had once been asked by Murray’s teacher to change the boy’s clothing after he had wet himself at the Day Care Center several months earlier.