This article is part four of an in-depth summary of the brilliant course:
Your Deceptive Mind: A Scientific Guide to Critical Thinking
In this article we look at why you can’t trust your memory:
- The limitations of memory
- Flashbulb memories
- Why confidence is not a good predictor of accuracy of memory
- Source amnesia
- Truth amnesia
- Thematic and detail memory
- Memory contamination
- The Rashomon effect
- False memories
- & more
Human memory is utterly flawed
Human memory is utterly flawed. Far from being a passive recording of events, our memories are constructed, filtered through our beliefs, and subjected to contamination and morphing over time. Memories can even be fused together or completely fabricated.
Limitations of Human Memory
There is a false assumption that all problems with memory are recall problems. That’s not true.
Some memories never form, in other words, we may experience something but never consolidate it from short-term into long-term memory.
Memories also degrade, two different memories can fuse together into one memory, or they morph and change over time so there is an actual loss of information. It’s not just difficult to access the memory, the memory is actually not there anymore.
A flashbulb memory is a type of long-term memory that we have for an unexpected emotional event—as opposed to the everyday mundane events of our lives.
Flashbulb memories tend to be vivid, long-lasting memories, and they are reinforced by the emotion of the event. In fact, a strong emotional experience, such as a traumatic event, strongly reinforces a memory. We tend to remember emotional events and traumatic events more easily.
The Challenger space shuttle explosion
A 1992 study by Neisser and Harsch questioned 106 students right after the Challenger space shuttle explosion about what they experienced when they heard that the Challenger had exploded as it was launching. They then re-questioned 44 of the 106 students 2 1/2 years later, and tested the students on the accuracy of their memory, about where they were, and what they were doing during the Challenger explosion.
25% of the students, scored 0/7 in terms of accuracy of details of that flashbulb memory.
50% of the students scored two or less.
Recall in general in the entire group was very bad. And it wasn’t an all or nothing problem. It wasn’t as if the students remembered 100% or 0% of the event, most students had some memory, but their memory for the details of the event was simply appalling.
Confidence is not a good predictor of the accuracy of memory
Interestingly, confidence in a memory is not a good predictor of the accuracy of a memory. We tend to naively assume that if we are very confident in a memory—if it feels vivid to us and can be easily recalled—it must therefore be accurate, but the research does not support this.
Memories of details tend to increase confidence, but having a vivid memory for detail does not necessarily predict accuracy. You may have a vivid memory that you feel very confidently about, but which is extremely inaccurate.
In relation to the disconnect between accuracy and confidence, there is something psychologists call source amnesia.
How often do you hear someone say, “I read somewhere” or “I heard somewhere” before relaying some fact or story.
Source amnesia is an example of the disconnect between accuracy and confidence. We have a particularly bad memory for the source of information, even when we can recall the information itself; our brains simply do not dedicate many memory resources to source information.
In our modern age where we’re dealing with multiple sources of information, and we have to access the reliability of those sources of information in order to say something meaningful about it, this source amnesia is a major problem. Knowing that you heard something somewhere is not that helpful if you have no way of assessing how reliable the source was because you don’t remember the source.
Similarly, truth amnesia involves remembering a claim much more easily than remembering whether that claim is a myth. In 2007 a study conducted by Ian Skurnik showed that as many as 27% of young adults misremembered a false statement as being true only three days after they were told it was false, and 40% of older adults misremembered a false statement as being true. They remembered that they’d heard the statement before, but they didn’t remember that it was false.
Truth status appears to fade faster than familiarity, as does source status. Therefore, familiarity leads to a truth bias, which gets reinforced with repetition. The effect of not remembering the truth status of a claim is worsened when the truth status is not revealed until the end. This has many implications for information campaigns that involve myth busting.
For example: If we say that it’s a myth that there is a causal link between vaccines and autism. What people remember is, “I remember there is some link between vaccines and autism, I don’t remember what the details were.” They don’t remember that it’s a myth, they don’t remember where they heard it from, they just have this familiarity with this association between vaccines and autism.
Thematic and detail memory
Psychologists also distinguish between thematic and detail memory.
Thematic memory is responsible for the gist of the overall emotional content of an event.
Detail memory is responsible for the details of an event.
We also tend to focus on and remember emotionally laden stimuli and ignore peripheral details. For example, witnesses will tend to remember that an alleged assailant was holding a gun and he was threatening. However, when the police ask them to remember such details as eye color and clothing, they may not remember those kinds of details at all.
Altering Memory Details
What’s much worse than not having a good memory for details is the fact that we actually alter the details.
Again, memories are not passive, they’re constructed. When you remember a memory, you’re not just experiencing this stored memory, you’re actually reconstructing the memory as you’re recalling it. You’re updating the memory, you’re bringing it up to date with everything that you’ve learnt since the last time you remembered or thought of that memory.
We construct a narrative, a story in our heads about what happened to us in the past, and that narrative has emotions and themes attached to it, and we alter details in order to be in line with our thematic narrative. Therefore, details are not only inaccurate, but they’re also biased thematically. We invent details, and we change details, in order to make a story more emotionally punchy.
An excellent example of this is Jean Hill a famous witness to the JFK assassination. She was standing very close to the vehicle when the shots were fired. Her report is recorded on the day of what she saw, as well as photographic and video evidence for corroboration.
Yet her story slowly morphed over time to fit the narrative theme that she had developed of a conspiracy that there was more than one shooter.
Initially her reports were very simple, she said that she was standing next to the car when the shots went out, she saw what happened to the president, but otherwise she didn’t really see a thing.
But over the years she started to embellish her story and add details that maybe she saw something over at the grassy knoll, and then years later she saw a person over at the grassy knoll, and then eventually she even recalled running after a person who was trying to flee the scene from the grassy knoll. These details were all added bit by bit over years, each time she was interviewed.
She remembered the gist of the event, and she remembered the narrative that was developed that there was a conspiracy of multiple shooters, and then the details morphed over the years documented on multiple interviews to fit that narrative, completely disconnected from the reality of what she actually saw as documented immediately after the event.
In this way, memories of past events become contaminated.
Memory contamination occurs when we incorporate details that we are exposed to after an event into the memory of the event itself.
Researcher Elizabeth Loftus and her collaborators found that people incorporate misleading details from questions or other accounts into their own visual memory. These are false memories that are constructed by someone asking a leading question, such as, “How fast were you going when you slammed into that car?”—implying that you were speeding.
For example: If you show subjects a film of something happening, and then just start asking them leading questions, “What was the woman in the video wearing?” (when there was no woman in the video) they will remember seeing a woman, and they will start inventing details, they will tell you what she was wearing. They will start to remember it vividly. These are false memories that are easily constructed just by asking a slightly leading question.
There is also a need to conform to what we think we know about the event we’re trying to recall. Witnesses, therefore, tend to contaminate each other’s accounts, bringing them in line with each other. This is not deliberate deception on the part of the witnesses, this is just how our memory works. In order to keep our memories consistent we update the details so they all fit together, and if we have to alter or make up details of a memory in order to do that, our brains will happily gloss over any inconsistencies, fill in the gaps in our narratives, so in the end we have a consistent narrative, even though it maybe completely inaccurate.
If you have three or four people who witness a UFO, a light in the sky, they may all give a very consistent account, and that consistent account maybe interpreted as evidence that they experienced something real, however, if they talked with each other about what they saw they have contaminated each others memory, and now they will incorporate the details that the other person suggested to them into their own memory. And they may have vivid detailed memories of seeing things, that were simply the result of the communication they had among themselves.
Additionally, we can even make up details entirely. We tend to invent details to fill in the gaps in order to create a consistent story and reinforce the emotional themes. This relates to the fact that memory is a construction—not a passive recording—we are constructing memories, and our brains do this in many trivial ways.
For example, when we blink, we miss a tiny bit of visual information, and our brains stitch together the visual information so that we have one continuous stream. We do this cognitively as well, we stitch together different bits of things we perceive about an event in the process of confabulation.
For example: There is one famous study in which there is a staged theft in a public area where there are a couple of confederates (subjects who are participating in a psychological study) that are carrying out the scenario. A person simply says, “Hey, somebody stole my radio!” Police officers who are also confederates come in and they take eyewitness testimony. People who are there will describe the radio, they will describe the suspicious person who probably took it, when in fact, there never was a radio, no one ever came in and took it, the entire scenario was made up.
But since the witnesses were told there was a radio, they were told it was stolen, and they’re being asked to recall what happened, they invent all the details they need to in order to make the narrative make sense.
Forced confabulation occurs when leading questions are asked about a film that was viewed, for example.
In studies of forced confabulation, subjects can remember seeing a nonexistent scene that was never in the film that they were asked about. Again, we make up these details to fill in our internal narrative, and they become indistinguishable from “real memories”. A memory of an event we made up, or confabulated, feels like a memory of something we actually experienced, and we don’t have a way of telling the difference.
The Rashomon effect
This is sometimes referred to as the Rashomon effect, which is based on a 1950 Akira Kurosawa film about a rape and murder in which the same story was told from four different accounts, and it was a very artistic depiction of how different accounts of the same event can be. Everyone has their own perspective, filter, and memory, and therefore when they compare stories after the event, there may be striking differences in how each person constructs the same event that they all experienced at the same time.
You’ve probably experienced this many times yourself. Imagine a time when you’ve had a very heated discussion with other people over an hour or so, and then afterwards when things have calmed down and you’re comparing notes about who said what, you may find that you and the others have very different accounts about what just happened. You may feel like you needed a camera rolling when the conversation was taking place because what you’re claiming now is completely not what happened. But you have to realize that your memories are as biased, constructed and flawed as everyone else’s in the room, and that’s why none of them fit together. You all have different perspectives and different filters.
Inventing false memories and experiences
The ability to confabulate can lead to what are called false memories. In 1988, Ellen Bass and Laura Davis wrote a book called The Courage to Heal in which they promoted the idea that there was an epidemic of what they called “repressed memory syndrome”. They believed that many people, mostly women, were abused as children, often sexually abused, and even the victims of bizzare, ritualistic, satanic torture. They felt that the memories of these highly traumatic events were repressed and later manifested as adults as anxiety, depression, eating disorders and other problems.
What Bass and Davis created however, was an epidemic of what is now known as false memory syndrome. Therapists who followed their prescription encouraged clients to remember details of abuse, they said, the first thing you have to do is recognise that your eating disorder (or whatever) is a manifestation of childhood abuse, they started with that premise, they said, I want you to imagine yourself being abused, what was happening, imagine the details, who was doing it etc. they essentially came up with a very good prescription for manufacturing false memories which led to many thousands of people manufacturing memories of intense abuse as children and believing that they were real. In fact, there are cases of people who were sentenced to prison for committing abuses when the only evidence against them was from false memories.
Budd Hopkins is another example of how easy it is to innocently or naively manufacture false memories. He was a painter who became an amateur UFO abduction investigator. His research technique was to use hypnosis to interview subjects who suspected that they may have been abducted by aliens.
Usually what would happen is someone would read a book or watch a TV show about alien abductions and then think, “Hmmm, I wonder if I was abducted, I had some unusual experience, or I saw a UFO once”, and some of these people may have found themselves in the hands of someone like Budd Hopkins, and used hypnosis in order to recover what he believed was their partially repressed or erased memory of their alien abduction. Under hypnosis which is essentially a condition in which people are highly suggestible, he would ask a series of very leading questions. He would encourage them to imagine themselves being abducted. It’s therefore not surprising that a standard abduction scenario would emerge out of those hypnosis sessions, and then his clients would come to really “remember” being abducted by aliens.
Children are particularly susceptible to suggestibility and creating false memory.
False experiences are very easy to generate
In addition to false memories, false experiences are very easy to generate. In one study, participants read an advertisement for a new but false brand of popcorn that vividly described both the taste and feel of the popcorn. A week later, participants were asked whether they were actually given the popcorn to taste or whether it was just described to them, and a certain percentage of the people who were never exposed to the popcorn remembered having eaten it. In other words, they inserted themselves into the memory of an event they only read about.
This is a caution for any profession that must solicit a history from another person, such as police officers, physicians and therapists.
People in these professions need to remember the incredible tendency for suggestibility and creating false memories. That’s why it’s important not to ask leading questions, but open-ended questions and let patients fill in the details themselves.
Furthermore, not only do we need to avoid encouraging people to invent memories, but we also need to make sure there’s always some external, objective verification. Courts have been moving in this direction and using more caution about the validity of eyewitness testimony and recognition of the needs to validate any testimony with some kind of objective evidence.
Memories are highly constructed. Every time you remember something you are recreating the memory, updating it and changing it’s details. Memories change over time, they fade over time, we lose important details like whether or not they’re correct, and we can fuse one memory to another, take one detail from one memory and insert it into another, they also morph to fit our internal narrative.
The bottom line is that it’s naive to implicitly trust our own memories. This maybe unsettling, but it’s an important step to recognise that we need to be realistic and humble about the limitations and flaws of human memory, it’s an important step along the journey towards true critical thinking.