Henry Molaison was seven years old when he was hit by a bike. Following the head injury, he developed epilepsy that grew worse until, by the time he was 27, he could barely function. An average of one seizure and 10 losses of consciousness per week were too much to bear. As a last resort, his doctor decided to remove some parts of his brain, including a component known as the hippocampus.

The surgery was a success, and the frequency and intensity of his epileptic fits decreased. After the operation, Henry ‏(who died in 2008‏) remained the same intelligent and friendly person he had been. But something did change: He lost the ability to remember, and thus became one of the most fascinating subjects of neuropsychology research.

It would be incorrect to say that “patient H.M.,” as he was called by the researchers who examined him after he lost his memory entirely. His short-term memory, for instance, was excellent: He easily remembered every picture or long number they showed him, for a short period after looking at it. He was capable of holding conversations extending a few minutes. His long-term memory was also preserved; he very clearly remembered events that happened before the operation, such as incidents during World War II and childhood memories.

What H.M. lacked in a very profound way was the ability to turn short-term memory into long-term memory. He forgot everything shortly after it happened. Less than one hour after breakfast, he could not remember what he’d eaten, or even the fact that he had eaten at all. As long as his attention was focused on a certain subject – i.e., having a conversation – he was capable of absorbing new information. But the moment his attention was diverted − if someone interrupted him in the middle of that conversation, for example − he was incapable of returning to the previous subject.

For 30 years, once a month, H.M. would meet with one of the scientists who studied him. And every time she entered the room, she was compelled to introduce herself anew, as he did not recognize her. Nor did he recognize himself in the mirror, since he only remembered his own appearance as it had been prior to the operation. His identity was frozen at that point in time, and he could not remember any changes that took place in himself afterward.

The study of H.M. enabled researchers to draw several important conclusions about memory: First, that it is a skill that is separate from other cognitive skills: H.M.’s intelligence was not impaired at all. Second, that the transition from short-term memory to long-term memory takes place in the hippocampus. This is why H.M.’s ability to remember new things was impaired when this part of the brain was removed. Eventually, researchers realized that the hippocampus itself is not the permanent space for storage of long-term memory, since its removal did not affect his memory of events from the distant past.

For many years, the scientists believed that the defect in his long-term memory was absolute, and that he could not remember or learn anything new. But memory was found to be even more complex than they thought. The researchers showed H.M. a list of words and asked him to read it via a mirror. Normal people have a hard time reading a mirror image at first, but over time they learn and their reading speed improves. Learning is dependent, of course, upon memory, and so they were very surprised when it turned out that H.M. was also able to learn how to carry out that task.

Still, every time he came to the laboratory, he had no idea who the people standing around him were, or what the unusual task they were asking him to do was. Nevertheless, he continued to improve at the same rate as normal people. Even a year later, he could still remember how to read text in a mirror image.
In the 1980s, American researchers Neal Cohen and Larry Squire realized that this phenomenon was evidence of the existence of what they termed “implicit memory.” It was a breakthrough in memory research, and in the understanding that memory is not one mental ability, but a whole range of abilities.

‘Unable to learn’

Prof. Eli Vakil, the head of the Laboratory for Memory and Amnesia Research at Bar-Ilan University’s Multidisciplinary Brain Research Center, is very familiar with the case of H.M. As head of the Israel Defense Forces rehabilitation center for veterans who’ve suffered traumatic brain injuries, he has seen similar cases.

“In normal people,” says Vakil, “the two memory systems − explicit memory [conscious recollection] and implicit memory − work together, and it is difficult to distinguish between them. Nevertheless, we are aware of the difference between the two types of memory in everyday life. For instance, when your spouse asks you for your credit card’s PIN code, you will sometimes get confused and say you can’t remember. But if you go to an ATM and let your fingers do the walking, they will easily key in the number, so it turns out that we do, in fact, remember it. This is implicit memory.”

“Among amnesiac head injuries, the division between the two systems is clear. I had one patient, an amnesiac fellow who was an injured IDF vet, for whom I tried to find employment. I contacted the owner of a printing press who agreed to employ individuals undergoing rehabilitation, and this fellow began a trial period. A few days later, I called to ask how it was going, and the employer said he was trying to teach the guy to operate a huge paper-cutting machine, but in vain. I asked him not to despair, but he called me back two weeks later to say that he was giving up because ‘This guy is just unable to learn.’

“I went to the printing press and asked the amnesiac patient if he knew what he was supposed to do. He said that he did not know. In an attempt to encourage him, I asked him to describe, stage by stage, the required actions, but he just shrugged his shoulders and said he could not remember. The employer looked at me helplessly. Out of some unclear impulse, I picked up a bundle of sheets of paper, thrust it into the fellow’s hands, and said: Go ahead, cut it. Without hesitation, the guy went up to the machine, set the bundle down, raised something here, pulled and pushed there, pushed some buttons two at a time − precisely what needed to be done − and completed the task.

“It was incredible. He remembered what he’d been taught perfectly, but didn’t know that he remembered. He lost the explicit memory, the one of whose existence or absence we are aware, but he had not lost the implicit memory.”
So it would be wrong to say about someone that “he has a good memory,” because there are many types of memory?

Vakil: “Human memory is much more complex than we once thought. People have always tried to define this ability, which in large part determines our identity and makes us the people we are. After all, who are we if not for the memories we have accumulated and the things we’ve learned?”

Over the generation, people have developed models for describing memory, which have changed in accordance with the technology of the period, explains Vakil. For example, Plato referred to memory as a wax tablet on which experiences are inscribed. In the Middle Ages, thinkers viewed memory as a sort of library − an image that combines both the notion of inscribing and the notion of storing the experiences. Based on this model, memories are cataloged in file cabinets or drawers, so that when we want information, we scan the cabinets and pull out the appropriate material.

In the 19th century, memory was described as a telephone switchboard that activates lines that have previously been laid down by experience. Later on, people would refer to memory as a recording device or camera. Today, in the computer age, people are accustomed to comparing memory to a supercomputer that efficiently absorbs, processes, stores and retrieves information.

But the developments in brain research over the past few decades have taught us that memory is, in fact, none of the above, Vakil says: “Human memory is dynamic and variable; it is formed and is erased, it’s explicit and implicit. If we’re going to compare it to a computer, than we should describe it as a network of computers, each of which has a different specialty, and these computers are linked together and operate in coordination.”

What would be an example of a memory network?

“Take, for example, our conversation. It is an episode, in the course of which your brain creates what we call ‘episodic memory.’ This memory is composed of everything connected to the episode: contents of the conversation, the time and place in which it is conducted, my face as I am speaking. All of this information is inscribed in your brain as memories that are linked to one another in a way that I can liken to a bunch of balloons. The content of the conversation is one balloon, recognizing my face is another balloon, the place, the time and the background − each one has its own balloon. All of the balloons are bunched together in a ring, which is held by the balloon seller. Thus, when he pulls the string of one balloon, he moves other balloons, too, and that is like the act of remembrance: I activate one memory and it pulls and arouses another memory.

“What is interesting is the relationship between the unconscious memories and the conscious ones. In our episode, perhaps you are intending to remember only the content of the conversation that we are having, because that is what is important to you, but your brain is also absorbing and remembering a myriad of other things that belong to the episode: the red dress you are wearing, my striped shirt, the picture hanging on the wall behind me, and maybe also thoughts that are popping into your head at the same time as the conversation between us is happening. So the conversation is only one memory among many in this episode.

“This creates fascinating possibilities of remembrance. If I asked you tomorrow what picture was hanging on the wall in my office, it is likely that you would say you can’t remember. But if I showed you the picture, you might possibly more easily remember our conversation. Because I moved one balloon, I also jiggled the others, and even the jiggling of an unconscious memory can influence and arouse the primary conscious memory. The memory is context-dependent, even if the context is unconscious.”

In addition to the creation of wonderful books like Marcel Proust’s “In Search of Lost Time,” are there any applications to this finding?

“The fact that memory is context-dependent is relevant, for example, to police interrogations of witnesses. A person happens to be in a certain place and becomes a witness to a crime. His testimony is important, but the memory is blurred. Studies show that in order to refresh his memory, it is worthwhile interrogating him at the crime scene, not at the police station, so that the context will help him to recall. There are places where they make sure to bring the witness to the crime scene at the same time of day that the crime occurred, in order to recreate as closely as possible the contexts that will arouse the desired memory.

“This is also something that can help us in everyday life. I’m sitting in my chair, intending to call someone. Suddenly there’s a knock at the door, and by the time I’ve gotten up to open it, spoken with the visitor and said good-bye to him − I’ve forgotten what it was that I had wanted to do. This happens a lot to older people. The recommendation is to go back and physically sit in the chair, in order that the context − the spatial conception, the body position, the things I saw out of the corner of my eye when I was sitting there − will help me to raise up from memory the thing that I wanted to do.”

Sometimes it’s semantics

The information of explicit memory is stored in various regions of the cerebral cortex, explains Vakil. Every region that is responsible for processing a certain type of information stores it, as well. For instance, the area responsible for vision both processes and stores the visual information, and the hearing region does the same for auditory information. The link between them, the ring that binds them, is found in the hippocampus.

Conversely, implicit memory is found in “lower” regions, beneath the cerebral cortex, which are responsible for relatively simple cognitive functions. Implicit memory usually endures head injuries much better than explicit memory, but among individuals with Parkinson’s disease, for example, implicit memory is more likely to be impaired, whereas the explicit memory is preserved.

You spoke about distinguishing between explicit memory and implicit memory, between long-term memory and short-term memory, and you talked about episodic memory. Which memory “complements” episodic memory?

“Episodic memory derives from a one-time event, from a particular episode. Conversely, there is conceptual memory, which is not dependent on any particular event. If I asked you, for instance, who the first prime minister was, you would retrieve the answer from your memory not because you had contact of any sort with Ben-Gurion. This memory, of information, is called ‘semantic memory.’

“But evidently semantic memory also starts out as episodic memory: One day, your kindergarten teacher Shoshana showed you a picture in kindergarten of a man with very little hair and said: this is David Ben-Gurion, our first prime minister. This formed in you an episodic memory. You went home and told your mother, and she said that Ben-Gurion knows how to stand on his head. Over time, you were exposed to more such episodes, and the memory became semantic memory − abstract memory that is not context-dependent.”

Which reinforces the Aristotelian, as opposed to the Platonic, approach to the terms.

“Semantic imprints are indeed formed by events that recur in different variations. Sometimes when you want to examine the memory of an adult person, you ask him, for example, to describe the wedding that he attended the previous day. He says it was great, that they ate and drank and had a good time. Does that mean he has a good memory? Not exactly. We meant to examine his memory of the wedding episode, but possibly he is reporting on a semantic memory − that is, a memory that he has constructed from the many weddings he has attended. After all, at most weddings you eat and dance and have a good time.”

You could say that he is lying. Not consciously, perhaps, but nevertheless he is reporting on a memory that is not true.

“Indeed, semantic memory is liable to create a problem of bogus memories. The global leader of this field of research is [an American] scientist named Elizabeth Loftus, who engages in the validity of eyewitness accounts. After all, what is an eyewitness account? It is a type of memory test, and it turns out that memory, at times, is deceptive when tested.”

“This was one of the tests: Loftus went with two unknown actors to a movie theater, where there was a long line of people in front of the ticket booth. The first actor was older and meticulously dressed, with shined shoes, a suit and tie. The other actor was young and sloppily dressed, wearing torn jeans and worn-out shoes. At one point, Loftus sent the young actor to join the line. He stood at the end, and as time passed the line continued to form behind him.
“In the next stage, she sent the older actor, who did not stand at the end of the line but pushed ahead and cut in front of the younger actor. The young actor turned to him and said: Sir, there is a line, to which the older one answered: Excuse me, but I was here first. The younger one: You cut ahead of me. The older one: You’ve got nerve.

“A fight broke out, just as we ourselves have seen happen. At this point, Loftus approached the people standing in the line and asked them who was there first. Seventy percent of the people said that the older man was standing in the line first, and that the younger man had cut in. When they were asked if they were certain, they swore that that was the case, and when they were asked if they would be willing to testify under oath in court, they answered in the affirmative.”

Hitting a wall

Our memories are dynamic, constantly being built and destroyed, says Vakil. Loftus, he says, compares memory to a wall that repeatedly collapses, whose bricks we gather up so as to rebuild it. Sometimes we are missing a brick, and then we hide the fact and build a slightly different wall. Often, when the episodic memory is vague, we complete it on the basis of the semantic memory we have, which is a stereotypic memory.

In the case of the experiment with the two men, people were not certain of what happened, and so they activated their semantic memory, according to which disheveled young people tend to cut in line and disturb the peace much more than do older people. “Our memory is definitely not a recording device that documents things as they are,” Vakil notes.

Of course, this can be a problem for anyone who seeks justice in the courts.
“When I lecture before judges , they are indeed very disturbed by this. After all, if a credible person comes along, one whom the judge has no reason to suspect of harboring any bias or intent to lie, then he believes him. The problem was that the people in line did not admit they did not clearly remember who was standing there first. They did not say, ‘It seems to us; it seems reasonable to us; considering the circumstances’ ... They thought they remembered because they were activating their semantic memory.”

“With the development of the technology of DNA testing as a means of identifying criminals with certainty, there have been many retrials at which it has been proven that many people who were once found guilty of a crime are in fact innocent. It was found that more than 50 percent of them were convicted on the basis of eyewitness testimony. Stereotypes have developed in the course of evolution as semantic memories, and they are critical to our survival, as they enable us to quickly make what might be life-saving decisions. If in the middle of the night on a nearly empty subway train we encounter a gaunt, feeble person − as opposed to a large man with rippling muscles − it’s obvious we will automatically be afraid of the latter, even though the former could turn out to be dangerous and the latter might be a nonviolent poet.

“The semantic memory we have constructed out of a myriad of episodic memories in which both feeble people and large, muscular men have taken part, taught us that as a rule, it is preferable to avoid an encounter with the latter. However, sometimes these semantic memories are at odds with our episodic memory.”

Let’s go back to everyday life. In order to remember a book or lecture, do you have to want to remember it, or is it better to just “let it flow”?

“Generally speaking, it is customary to say that when there is a desire to remember, we do it better than when the learning is random. But some researchers say it is not the desire that determines, but rather the level of processing that we do with the new information. The fact that you want to remember something new causes you to process it more, in other words, to understand its significance and to link it to information that already exists in your mind; and the more that the information is processed, the more it is remembered. This is easy to see when subjects are asked to remember a list of words. If we ask one group of people to find a rhyme for each word, and members of the other group to consider its meaning, we will see that the second group will remember many more words from the list. Attributing meaning to a word is a deep processing of the information, and it is therefore remembered better.

“There is a good chance that if the book or lecture provoked a certain emotion in you, you will remember them better. All of us remember, for instance, where we were when Rabin was assassinated, because we remember things that happen to us when we are in a high state of arousal. This happens because the hippocampus, which is, as we said, responsible for the ability to transfer memory into long term, is connected to the amygdala, which is part of our emotional brain. This attribute of memory also developed as a critical survival skill. If I went hunting in the savanna and a tiger attacked me and I barely managed to escape, it is good if the memory of this bitter place is inscribed in me, so that next time, I will avoid getting too close to it.

“Post-traumatic response is, of course, an unsuccessful mutation of this linkage. For a person involved in an accident who has experienced severe emotional trauma, the terror may at times be imprinted in our memory so much so that it will not yield. The mistake that is sometimes made in cases of shell shock, and this happened a great deal in the Yom Kippur War, is to evacuate these victims to the rear. The sudden detachment from the severe emotional experience leaves the memory imprinted. The appropriate treatment is to move the victims away from the front line, but leave them within the war zone, and even keep them in uniform, so as to give them the feeling they are still part of what’s going on. In this way, the experience is not fixed on the isolated emotional event, but is diluted by other emotions, sounds and sights.”

And age also affects the memory, of course.

“Implicit, episodic memory reaches its peak in the third decade of life, between 20 and 30. All of the stories about the fantastic memory of children are rubbish. Children easily acquire language, that is a conceptual process, but memory of abstract things, in which content that requires a high level of processing is juxtaposed, develops at a much older age.

“But with age, this memory weakens, and some people even develop a dementia, such as Alzheimer’s, that causes severe memory loss. [Researchers in] one of today’s hottest fields try to predict why some people develop dementia and others not, and why sometimes at age 90 and sometimes at age 60.

“One of the most famous studies in this field was conducted on several hundred nuns in the United States, who bequeathed their bodies to science. In their lifetimes, the nuns had a similar lifestyle in terms of daily regimen, nutrition, hours of sleep and inclination toward spirituality. After their deaths, when their brains were examined, it turned out that among them was a group of nuns whose brains were afflicted with Alzheimer’s. Their brains demonstrated various attributes of the disease − for example, unique proteins that produce sediments in the brain − but strangely enough, while they were alive, there was no behavioral manifestation of this brain disease. It turned out that these particular nuns were the most educated and cognitively-developed members of the entire group. Numerous other studies conducted since then confirm these findings and show a significant correlation between years of education and development of the disease.”

So what will people without that much education do? Are they sentenced to forget?

“A key concept in this context is ‘cognitive reserves.’ It would seem that the diploma on the wall is less important than the cognitive activity to which the education perhaps testifies. Cognitive activity affects memory because it helps develop extended neural networks and therefore produces complex representations of reality. This does not prevent the biochemical process of protein sedimentation that is typical of Alzheimer’s from taking place in the brain, but active people have adequate cognitive reserves, which will ‘absorb the blow’ in one place and compensate for it through activity in other places.

“I would compare it to two cities, one of which has a good road network, and another that has only one main road and then a few minor ones. There is a big difference between the cities when a traffic jam occurs on one of the roads: In the networked city, alternate routes will easily be found, whereas in the other city, all of the traffic will come to a halt.”

Is there such a thing as “excess memory” − people who have too much memory or who cannot forget?

“Yes. Our natural mechanism is actually to forget, and this, too, is an attribute of human memory that distinguishes it from a computer. Our brain is built to produce memory and then to erase and change it, so that we will not be excessively connected to memories. There is a survival benefit to this attribute of memory, as well. If we did not forget experiences such as childbirth or other harsh events, we would have a hard time functioning. That is what happens to people with post-trauma syndrome.”

“However, there are people who have an absolute memory. In other words, they remember everything. And I am not talking about what we call ‘idiot savants,’ people who, aside from having a phenomenal memory, suffer serious defects in their other cognitive faculties. It turns out that there are normal people who live and function as we do, but who, whenever they are presented with a date from the past, know not only what day of the week it was and what the weather was that day, but exactly what they were doing that day, what they were wearing, the conversations they had, etc.

“On the face of it, this is great, right? What could be bad about it? But almost none of these people is able to maintain a couple relationship. It may be that in order to be in a relationship you have to know how to forget, not to win every argument. One of these women said, for example, that with every argument that she’s ever had with a partner, she always remembered everything that was said, and this proved to be destructive. Another woman related that bitter memories are especially hard for her, because when she recalls one of them, she does so to the full depth of the emotions that originally accompanied it. Such that forgetfulness can evidently be a blessing, as well.

“Someone once told me that forgetfulness encourages creativity, because you have to create things from the fragments of memory. That is interesting, and it is apparently true, as well. I once knew a mathematician who told me the same thing. He forgets the things that he’s learned, and that enables him to blaze new trails.”