“Of all the things I’ve lost, I miss my mind the most”
-Mark Twain
If you’ve never visited the Chicago Field Museum, I suggest checking it out if you are in town. The most complete T. rex skeleton ever found, “Sue”, awaits you in the entrance gallery. She looks like a tyrant, scaled, and crouching lizard. But how did paleontologists construct “Sue” for the display? Obviously no one knows for sure how dinosaurs looked 80 million years ago. When scientists uncover a bunch of randomly dispersed fossils, they may have a general idea of her appearance based on the 50 + T. rex specimens unearthed with certain common characteristics. But really they use their “best guess” to fill in all the details and reconstruct her.
Why are we mentioning how paleontologists puzzle together fossils in a blog about memory? Turns out, the processes may have similarities. But let’s step away from the fossil concept for a minute.
Most people have the utmost confidence in their mind and memory. Especially if a specific event was important or traumatic, they usually believe their memory is embedded like a photograph.
However, most people don’t remember nearly as good as they think. We often change the facts, adding fallacious details, without even realizing it.
In fact, we can even create completely fake memories. When participants are shown a photograph of a hot air balloon ride with themselves “photoshopped” in as a child, which was a completely fictitious event, 50% of people can be convinced the false memory was true. They can even recall specific details despite it never having happened. Photographs, especially, can be very distorting of our memory.
A few years ago we heard all over the news that an individual could accurately remember something emotional and traumatic. “I don’t have all the answers, and I don’t remember as much as I would like to”. But because certain details were, “Indelible in the hippocampus… the trauma-related experience is locked there…”. Most of us intuitively would agree with that… if an event was important enough to us, then how could we possibly forget?
However the problem with making life-altering memories “indelible”, has to do with how we analyze stimulating information and create a memory of such encounters. We know from research that our memories initially do not always encode details of events… of course this is different from trying to remember a single detail, like me just remembering the dress my wife wore at our rehearsal dinner. If tasked with just one thing, we’re pretty good. But when speaking about an overall experience, which is actually a series of occurrences, we tend to first remember the “general concept”, then later add in the minutiae. This minutia may be accurate, but sometimes becomes clouded by our perception of what “should” fit. For example, when participants of a research study are asked to remember an incident involving a man wearing a suit, most recall that the man was also wearing a tie, whether he was or not. Wearing a tie, “fits”, with wearing a suit. And the parts we “fit in” tend to be shaped by the information we already know or are familiar with. Turns out if someone is asking us to recall a memory, the way in which a question is posed also affects our recollection. When study participants are asked how fast two cars were moving in a car accident which they viewed in a video, and words like “smashed” are cued by the interviewers, people tend to recall the cars were traveling faster than they were. A week later the participants are also more likely to recall broken glass, even when none was visible in the video. Yet if interviewers cue using a word like the cars “bumped” each other, the participants recalled slower speeds. There are countless reasons why mistakes or embellishments intrude, ranging from what we believe is or wish were true, to what someone else told us, or what we want others to think. And whenever these flaws enter, they have long-term effects on future recollection. It’s a bit like mental, or memory, paleontology.
It is the mechanism in how a significant event is first processed as a “general” concept but then layered with subsequent details, that can lead to unreliable neural storage of memories. We simply are not computers… Memory is not a process of opening a file drawer and accessing information. It’s a process of reconstruction… just like how the Field Museum paleontologists reconstruct “Sue” from a general idea with bits and pieces of data. Maybe they get some of it right, but maybe they match together two fossils incorrectly or add in another dinosaur’s bone unintentionally.
Memory is processed in the hippocampus and limbic system. It is postulated, and incorrectly generalized, that strong and coordinated stimulation of the neural circuits in the hippocampus, either with repetitively practiced memorization or the importance of a particular memory, leads to strengthened connections between those neurons. This is called, “long term potentiation” (LTP). It is felt that LTP leads to structural changes in the hippocampus, and once a memory undergoes this “synaptic consolidation”, it is permanent or “indelible” like a file on a computer. This storage sequence can occur with studying for a test and remembering facts for educational purposes. However, this learning pathway involves the brain seeing nearly the exact same information, over and over again, to boost the memory. The brain does a good job of imbedding this information into the mind forever, or at least for a while. In contrast, the pathway may be different in remembering a traumatic or significant event. Again, a life-altering memory is processed initially as a generality, and then small, and sometimes changing, details creep in later. The subsequent minutia can cause the hippocampal neurons fire in uncorrelated ways, leading to a reduction in the strength of the same synapses that LTP strengthens. This is termed, “long term depression” (LTD). The significant memory quickly becomes falsely incorporated in our brains with countless errors. Then later, the act of retrieving such memory puts those synapses into a labile state, from which they must re-stabilize to persist. Without “reconsolidating” the memory, or long-lasting LTP, the information becomes more mixed up. If the content of the original memory is updated with new information at the time of retrieval, memory distortion could occur. To further complicate, if multiple different memories now exist of an event (the original memory plus new versions formed while retelling time and time again), retrieving inadvertently and unknowingly mixes up the versions. And if we don’t think about that specific memory for a time period, other unrelated but similar memories intrude into the original file. If someone else recalls their memory of the same event, their version contaminates ours. This all results in inaccurate memories of very significant life events, despite the utmost confidence that we’re remembering like it was yesterday. Most of us can relate to this when we sit with friends or family, nostalgically telling stories from years back, but we all seem to remember the same thing a bit differently. It’s not that we’re intentionally lying, we now just have a totally incorrect version of the memory in our hippocampus… and we trust our hippocampus. As George Costanza said in “The Beard” episode of Seinfeld, “Jerry, just remember… It’s not a lie… if you believe it…”
Some evidence suggests that emotional arousal from certain events also leads to activation of the basolateral amygdala, with the release of cortisol, modulating memory storage. But in both animals and humans with high states of arousal, memory encoding may be enhanced or impaired depending on the individual’s stress response. There is a cortisol response in retrieval as well. In humans, high levels of cortisol during retrieval may improve the retrieval for the central gist of the memory, but impair the details. This may explain why interviewing people immediately following an arousing event when stress levels are high, often produces vague and disjointed memory. But a 2nd interview a few hours later may produce a more coherent story with additional details. However, how can we know when our cortisol levels have baselined enough to not interfere with memory retrieval, yet the remaining memory is not contaminated with false minutiae?
We won’t even delve into the psychological factors which may influence memory through a “self-protective” mode.
One problem with the human hippocampus is that it’s evolutionarily older than some other parts of our brain. It has half the number of layers compared to regions (neocortex) that make humans complex. The structure of the hippocampus is similar across various mammalian species, ranging from the rodent through primates such as humans. Do we expect a rodent to have a photographic memory??? Of course not… that’s why they’re so easily caught in a trap. A rodent just needs some basic and generalized memory… and so it’s unfortunate that this structure in humans has evolved at a snail’s pace.
In the courtroom, the fallacy of memory is becoming apparent. It has been reported that mistaken eyewitness identifications in criminal cases may occur more than 75% of the time when DNA later exonerates the convicted individual. Even more concerning, research specifically examining eyewitness testimony, or the memory of traumatic events has shown weak or negative correlations between a person’s confidence and the true accuracy of their memory. We are a very confident society, but often wrong.
One of the best ways to study the accuracy of event-related memory involves “flashbulb” events. After all, it’s difficult to study a bunch of individual people remembering personal events and make generalizable conclusions. There are key events in our world, like the assassination of JFK, the Challenger explosion, or the September 11, 2001 attacks, which sear themselves into our memories. Most people know (or think they know) exactly where they were and what they were doing when they first learned of these events. We have raw footage available of the events for further reference. Researchers can study people in these scenarios right after they occur, and then follow them for years to see how their memory holds up. For 9/11, a survey was sent to about 3,000 individuals 1 week after the attack, asking them their recollections which were recorded in writing. The 2nd survey, asking the same questions, was recorded 11 months later, and participants offered consistent answers only ~63% of the time. The 3rd survey, 3 years after the attack, showed only 57% consistency. This was stable at 10 years. Incorrectly remembered facts included not just the number of planes, crash sites, or order of events, but even basic details about where the participants were located or what they were doing at the time of the attack. After about 2-3 years the inconsistency of the memories stopped changing, but unfortunately, participants began just repeating the same faulty memories, in exquisite detail, with great confidence. This suggests that individuals stop “filling in” missing details with guesses at a certain point, and retain/ recount factually false information thereafter with the true belief that they’re remembering accurately. And when shown their own original written statements, which you’d think would reset their recollection… they adamantly stood by their current memory, stating what they previously wrote was wrong.
These studies reflect what we know about the Ebbinghaus “forgetting curve”. Even after one week, research suggests a drastic drop in memory retention.
Some argue that emotionally charged “flashbulb” memories are distinct from other traumatic memories because for many they involve learning about an event rather than directly experiencing that event. That being said, new research may suggest that people can manifest many of the symptoms of PTSD following even indirect exposure to trauma. And of course, the 9/11 data surveyed mainly New Yorkers, who may not have been in the immediate vicinity of the Towers, but directly were affected in some fashion.
Research involving memory retention of significant events is very interesting, as it affects our interpretation of everyday media stories. It affected how most of us swayed our opinion, one way or the other, in the “Me Too” movement, Clarence Thomas/ Brett Kavanaugh hearings, or even in the firing of Brian Williams a few years back. We presume someone is flagrantly lying, and dismiss any possibility that this simply displays the underperformance of the normal human hippocampus. A sadistically comical component about Brian Williams was that when he incorrectly recounted the 2003 helicopter scandal, the actual pilots of those Chinook’s also erroneously recalled incorrect dates but without disparagement. He was fired and chastised, but they surely weren’t. Jon Stewart, on “The Daily Show”, satirized this was, “infotainment confusion syndrome”. “It occurs when the celebrity cortex gets its wires crossed with the medulla anchor data”. Of course, it’s really just the mere limitations of a 3-layered allocortex comprising our hippocampus. At just 18-20 weeks of gestation, the fetal hippocampus already resembles the adult version. The cerebral cortex (neocortex), on the other hand, continues to develop even after birth. So how can we expect a structure formed so early in embryogenesis to function as flawlessly as our cerebral cortex which needs more than a full pregnancy to mature?
We also see the pitfalls of memory retention in medicine… specifically in discussing diagnoses and procedures with surgical patients. I would not equate the trauma of a neurosurgical discussion to 9/11 or assault, but surgery can still be considered a very significant and traumatizing life event for some people. Memory for medical information is often poor and inaccurate, especially when the patient is anxious. They tend to focus on diagnosis-related information and fail to register instructions on treatment. Patients with lower education recall about 38% of items freely in the clinic setting, while patients with higher education tend to recall about 65%. This is inherently problematic when we decide how much should be explained to the patient and how much they can comprehend. Physicians, and especially neurosurgeons, are criticized for failing to speak in “layman’s” terms. Clinicians could poorly be perceived if patients feel inadequately informed about risks of procedures or reasonably expected outcomes. Yet we know that the greater the amount of information presented, the lower the proportion will correctly be remembered. Of interest, studies involving consent for surgery show that patients tend to recall benefits more than complications. This may be minimized if the information is provided verbally, with visual tools, and in writing if possible.
To better understand the memory recollection of fellow human beings, we all just need to be a little more aware of the limited capabilities of the hippocampus. In many ways, our hippocampus is evolutionary “as old as” the T. rex. And while paleontologists having been displaying “Sue” since the late 1990s based on their “best guess” of how the individual fossils fit together, newer T. rex research in 2004 suggests the primitive tyrannosaurus may have been covered in feathers and stood taller… There may have been a completely different look to the most famous predator of all time, just as our memory of significant events may sometimes be wholly different than the truth.