Eric Edelman/

Collage by Eric Edelman

On September 1, 1953, William Scoville, a neurosurgeon at Hartford Hospital in Connecticut, operated on a twenty-seven-year-old man named Henry Gustav Molaison, who suffered from severe epilepsy. Scoville removed two pieces of tissue—the left and right sides of the hippocampus—from Molaison’s brain. The hippocampus, located near the center of the brain, forms a part of the limbic system that directs many bodily functions, and Scoville thought that epileptic seizures could be controlled by excising much of it. The result, however, as the journalist Philip Hilts wrote in Memory’s Ghost (1995), was that

from H.M.’s moment in surgery onward, every conversation for him was without predecessors, each face vague and new. Names no longer rose to the surface, neither histories nor endearing moments came anymore. Reassurances of welcome had to be sought every moment from every look in every pair of eyes.

H.M., as he came to be known in the medical literature (his real name was not disclosed until his death in 2008), could no longer remember anything he did. He could not remember what he had eaten for breakfast, lunch, or supper, nor could he find his way around the hospital. He failed to recognize hospital staff and physicians whom he had met only minutes earlier, remembering only Scoville, whom he had known since childhood. Every time he met a scientist from MIT who was studying him regularly, she had to introduce herself again. He could not even recognize himself in recent photos, thinking that the face in the image was some “old guy.” Yet he was able to carry on a conversation for as long as his attention was not diverted.

H.M.’s condition suggested that the hippocampus was essential for the conversion of short-term memories to long-term memories, and he became the most widely cited example in studies of the distinction between them. Eric Kandel, James Schwartz, and Thomas Jessell drew on his case in 2000:

Brain trauma in humans can produce particularly profound amnesia for events that occur within a few hours or, at most, days before the trauma. In such cases older memories remain relatively undisturbed…. Studies of memory retention and disruption of memory have supported a commonly used model of memory storage by stages. Input to the brain is processed into short-term working memory before it is transformed through one or more stages into a more permanent long-term store.1

Patient H.M., by Scoville’s grandson, Luke Dittrich, is a memoir of his grandfather and H.M. Much of the book describes, with justified quiet indignation, the failures of the neurosurgical procedures that were widely practiced by Scoville and other neurosurgeons in the past century.

The procedures that Dittrich describes have a long history. In the late nineteenth century, for example, Dr. Gottlieb Burckhardt, a Swiss psychiatrist, “performed the first modern neurosurgical attacks on mental illness.” Burckhardt had no experience or training as a neurosurgeon, but one of the first patients he selected for his experiments was a “‘disturbed, unapproachable, noisy, fighting’…fifty-one-year-old, ‘particularly vicious woman,’ who’d been institutionalized for sixteen years.” After five operations, over the course of which he removed eighteen grams of her brain, Burckhardt noted that his patient had become “more tractable.” As Dittrich writes, “Her previous intelligence, he added, ‘did not return.’” Burckhardt concluded that his patient “has changed from a dangerous and excited demented person to a quiet demented one.”

Psychosurgery became increasingly popular in the 1940s, and in 1949, Egas Moniz received the Nobel Prize for inventing the procedure called lobotomy, in which the neural connections to the prefrontal lobe are severed. Dittrich writes:

The Nobel Committee had endowed psychosurgery with a patina of nobility, demonstrating that future breakthroughs in the field might pay great professional, therapeutic, and scientific dividends. For ambitious tinkerers like my grandfather, the lure was irresistible.

He gives a fascinating portrait of Scoville, who sought professional advancement through his experimental operation on H.M., describing him as “a restless explorer in the operating room, never satisfied with existing techniques or methods, even the ones he had invented.” What emerges from Dittrich’s account is a profound sense of the ignorance, the arrogance, and the passion that drove his grandfather and other neurosurgeons to perform operations that often left their patients demented. They had a drive to innovate—to pursue new, untried, experimental procedures with unpredictable consequences—and were untroubled by their harmful outcomes.

Dittrich shows how H.M.’s case pointed the way to a better understanding of some of the more puzzling aspects of how our brains function and the nature of our conscious behavior. After surgery, he notes, H.M. was insensitive to pleasure and pain. When subjected to increasing levels of heat from a dolorimeter, which causes considerable pain in normal subjects, “Henry sat calmly,…even as his skin began to burn and turn red.” He lost “a capacity for desire”: “in the six decades between his operation and his death he never had a girlfriend, or a boyfriend, never had sex, never even masturbated.” H.M.’s insensitivity and his indifference to pleasure and pain seem critical to an understanding of his loss of memory. For all of our memories are subjective. Your memories are in relation to you, your friend’s memories are in relation to him or to her, and so on. The loss of pleasure and pain is a loss of subjectivity, of an ability to relate to objects, to persons, and to oneself—an ability H.M. lost when Dittrich’s grandfather removed his hippocampus.


Dittrich’s book concludes with an interview with Suzanne Corkin, a professor of psychology at MIT. For almost fifty years she studied H.M., and she and her mentor, Brenda Milner, wrote a number of important papers about the hippocampus’s function in establishing long-term memories. They showed that H.M. could no longer form memories of space or time or acquire general knowledge of the world, but he could learn new motor skills. Their work was the basis of the understanding of memory and hippocampal function since the 1960s. When Dittrich interviewed Corkin in 2015, he asked what she was going to do with her notes on H.M.:

Dittrich: Are you aiming to give his files to an archive?

Corkin: Not his files, but I’m giving his memorabilia to my department. And they will be on display on the third floor….

Dittrich: Right. And what’s going to happen to the files themselves?

She paused for several seconds.

Corkin: Shredded.

Dittrich: Shredded? Why would they be shredded?

Corkin: Nobody’s gonna look at them.

Dittrich: Really? I can’t imagine shredding the files of the most important research subject in history. Why did you do that?

Corkin: Well, you can’t just take one test on one day and draw conclusions about it.

Many readers will be shocked by the revelation that Corkin’s notes were shredded. (Whether they were remains a matter of controversy; in 2016 MIT responded to Dittrich with an open letter claiming that nothing was actually destroyed, and Dittrich then posted online a recording of his interview with Corkin telling him the material was gone.) Yet even had they survived, they would not have revealed much of the deeper significance of H.M.’s case, which has become evident only through new neurobiological research.

Recent studies of how the brain organizes space and regulates how one makes sense of one’s environment have shown that the hippocampus is concerned with much more than converting short-term memories into long-term memories. For example, H.M.’s sensations, thoughts, and perceptions after the operation had no continuity at all. “Every day is alone in itself,” Corkin quotes him as saying. Summarizing H.M.’s interview transcripts, Dittrich writes:

The most compelling moments were always the rare ones when Henry would try to explain what it was like to be him…. He never quite succeeded, since his amnesia wouldn’t let him hold on to the ideas long enough to get them out. He’d seem on the verge of a breakthrough, of a definitive statement, and then his train of thought would derail, and he’d start all over again.

These and other observations of scientists who studied H.M. are consistent with the more recent finding that, in the words of the neuroscientists Marc W. Howard and Howard Eichenbaum, “one of the functions of the hippocampus is to enable the learning of relationships between different stimuli experienced in the environment.” The work of Eichenbaum and others has begun to give us not only a new view of the function of the hippocampus, but a new understanding of the nature of memory. It is becoming increasingly clear that human and animal memory depend on the ability of the hippocampus to establish relations between an individual and his or her surroundings.

Laboratory of Comparative Human Cognition/UC San Diego

The Soviet neuropsychogist Alexander Luria, author of The Mind of a Mnemonist (1968),
with patients; 1960s

Studies by brain scientists including Eichenbaum and John O’Keefe have shown that the hippocampus is made up of cells with different kinds of functions. Most important are “place” cells, discovered by O’Keefe in research that won him the Nobel Prize, which respond to an animal’s location in space by causing electrical discharges called action potentials, creating mental maps of an animal’s environment. These maps are at various scales, like maps of an entire city as opposed to maps of individual streets. “Place cells,” wrote Howard and Eichenbaum in 2015, “are apparently not coding for a place per se but a spatial relationship relative to a landmark, or set of landmarks.”


There is considerable evidence that the activities of hippocampal neurons also help establish our relationships to many other types of environmental and internal stimuli, such as sounds, odors, pain, pleasure, and fear. Howard and Eichenbaum proposed that “the spatial map in the hippocampus is a special case of a more general function in representing relationships…including both spatial and non-spatial [stimuli].” In each case, the neurons are able to convey a relationship to our consciousness. The hippocampus also organizes temporal stimuli (including when an event took place) and sequential stimuli (indicating the order of a series of events). The hippocampus receives and integrates many other varieties of information to create multisensory relations, which is what memory is all about.2

But in what sense are relationships of this kind involved in remembering other sorts of information that apparently have nothing to do with specific events or our environment, such as random lists of words and numbers? Consider, for example, Alexander Luria’s description in his book The Mind of a Mnemonist (1968) of a patient, S, who could

recall tables of numbers written on a blackboard. S. would study the material on the board, close his eyes, open them again for a moment…and…reproduce one series from the board.

How is this ability to recall random words and numbers, even years later, related to what scientists have recently suggested is the function of the hippocampus, which is apparently essential to our capacity to remember? Luria describes how the mnemonist remembers. He never recalls arbitrary lists of words or numbers without first establishing a setting—a relation—in which he heard the lists:

Experiments indicated that [the mnemonist] had no difficulty reproducing any lengthy series of words whatever, even though these had originally been presented to him a week, a month, or a year, or even many years earlier…. During these test sessions S. would sit with his eyes closed, pause, then comment:… You were sitting at the table and I in the rocking chair… You were wearing a gray suit and you looked at me like this… Now, then, I can see you saying…

In other words, the mnemonist accesses (i.e., recalls) what appear to be imprinted words only by recalling the setting in which they were first “imprinted” in his memory. Once he recalls that setting, S. has a technique that allows him to memorize arbitrary lists of numbers, words, or both. The mnemonist, Luria notes, when given a long series of words to memorize, would

find some way of distributing these images of his in a mental row or sequence. Most often (and this habit persisted throughout his life), he would “distribute” them along some roadway or street he visualized in his mind. Sometimes this was a street in his home town, which would also include the yard attached to the house he had lived in as a child and which he recalled vividly. On the other hand, he might also select a street in Moscow. Frequently he would take a mental walk along that street…and slowly make his way down, “distributing” his images [evoked by the words] at houses, gates, and store windows.

There is no example in Luria’s book suggesting that the mnemonist can recall without establishing a setting. We would suggest that all recollections depend on a setting that the individual may or may not be aware of.

This mnemonic technique has been known since the ancient Greeks. Cicero tells us that an aristocrat named Scopas was giving a banquet, at which the poet Simonides chanted a poem in honor of his host that included “a passage in praise of Castor and Pollux.”3 Subsequently a note was brought to Simonides that two young men were waiting for him outside, but when he went to greet them he did not find them. Meanwhile the banquet hall collapsed during his absence, killing all of the guests. The corpses were badly mangled and could not be identified. Simonides remembered the place where each of the guests was sitting and was therefore able to identify them.

Simonides is generally known as the inventor of the art of memory. Most remarkable is that the art he invented operates not unlike the way the hippocampus creates human and animal memory by means of cells that map location in space, or create temporal markers, or encode sequences of events.

Essential to the brain’s creation of memories is that all of our memories are subjective—they are created from the point of view of the individual who is remembering. We have a sense of self because we have a preexisting sense of our body that contains that self. The basis of our subjectivity is our “body image,” a coherent, highly dynamic (it is constantly changing with our movements), three-dimensional representation of the body in the brain. This body image is an abstraction the brain creates from our movements and from the sensory responses elicited by those movements—using one’s left hand to pick up an apple, for example. “The coherence of consciousness through time and space is again related to the experience of the body by way of the body image,” John Searle wrote in these pages in 1995. “Without memory there is no coherent consciousness.”4

Since our subjectivity depends on our body image, if our body image is altered for neurological reasons, so too are our recollections. After he badly injured his leg on a mountain in Norway, Oliver Sacks described what is known as the “alien limb” phenomenon in his book A Leg to Stand On (1984):

The leg had vanished, taking its “place” with it. Thus there seemed no possibility of recovering it…. Could memory help, where looking forward could not? No! The leg had vanished, taking its “past” away with it! I could no longer remember having a leg. I could no longer remember how I had ever walked and climbed.

Since the nineteenth century it has been known that the brain creates “maps” of the body in the cortex. There is a cortical map of sensations (a sensory map) and a cortical map of movement (a motor map). In the sensory cortical map (also known as the sensory homunculus), the region in the brain that is activated, for example, by touching the hand, fingers, and arm—the cortical area that “represents” the sensations created by a cotton swab moved from the tip of the fingers to the arm—is adjacent to the representation of the face.

A counterpart of the alien limb is the “phantom” limb—a limb perceived by an amputee who feels as if an arm or leg that was lost in surgery is still attached to the body. The phantom limb might be extremely painful. When points remote from the amputation line are touched, such as the amputee’s face, he or she paradoxically feels a phantom limb. Remarkably, memories related to the original limb may be linked to the phantom limb. The subject may even perceive that the phantom limb is wearing a wedding ring or jewelry; when the weather turns humid, the phantom limb may experience arthritic pain. The patient’s phantom limb is not only a recollection of the lost arm or leg, but one that includes the patient’s experiences related to that limb.

Or take the case of a man whose memories are transformed when he becomes blind, as the theologian John Hull describes in his book Touching the Rock (1990). Hull became increasingly blind between the ages of twenty and forty. When he lost his sight, he noted, “the proportion of people with no faces increased…. I have fairly clear pictures of many people whom I have not met again during these three years, but the pictures of the people I meet every day are becoming blurred. Why should this be?” Hull answers his own question:

In the case of people I meet every day my relationship has continued beyond loss of sight, so my thoughts about these people are full of the latest developments in our relationships. These have partly converted the portrait, which has thus become less important. In the case of somebody I know quite well but have not seen for several years, nothing has happened to take the place of the portrait, and when I think of those people, it is the portrait which comes to mind.

Hull goes on to say that he was deeply distressed that he was losing the visual portraits of his wife and children. Hull’s memories (as is true of all of our memories) were continuously being “updated.” He could still visualize people he had known before he became blind and had not been in contact with since. But now that he was living in a world without any new images, his memories of people with whom he was regularly in touch were being updated into a nonvisual form—the sounds of their voices and the sensations of touching their hands and faces. When one becomes blind, the continuity of visual memory is lost.

When memories are first formed, they are “short-term” and unstable. But with time, the physical representation of the memory in the brain formed by the synaptic junctions between neurons becomes more stable. This process is called consolidation. The stabilized memories then become “long-term” memories. H.M.’s brain was unable to create long-term memories. Recent neurophysiological studies have shown that even long-term memories are very dynamic and that each time the brain tries to activate a “memory trace”—the physical representation of the memory in the brain, also called the “engram”—the nature of that trace changes. In other words, memories are altered every time the brain recalls them. This alteration of an existing memory is called reconsolidation. Because the memory trace changes, you can never remember the same thing twice in exactly the same way.

The process of reconsolidation, scientists have shown, changes the memory—that is, the way the memory is represented at the synaptic junction is altered. The recognition of the malleability of memory is nothing new. What is new is the observation that the connections between neurons that many scientists believe have a central part in generating memories change whenever the brain seeks to recover the information they represent. These changes may be the reason we can generalize. Over time, some memories are assimilated into categorizations or generalizations. When we recall taking the subway, we do not necessarily recall each trip separately but rather taking the subway in general; and such recollection may include an image of the subway. The brain simplifies our understanding of our environment and our relationship to it.

Memory may appear to be a reproduction of images, sounds, and even thoughts that can be stored in the brain in a manner analogous to the way information can be stored on a CD, but it is becoming increasingly evident that this is too limited an understanding. Rather, as Eichenbaum, O’Keefe, and others have shown, memory is the establishment by the hippocampus of complex relations among a variety of sensory stimuli from the point of view of the individual who is remembering. Thus when Scoville removed H.M.’s hippocampus, H.M. lost more than an ability to convert short-term memories to long-term memories; he lost the ability to establish such relations.

Yet scientists still don’t understand the ways that changes in the synaptic junctions between neurons, or changes in the neurons themselves, are related to our memories, thoughts, and actions. Indeed, neurobiology has yet to define the physical nature of the long-lasting changes in neuronal connections that are created as memories are formed. Even a simple memory must involve vast numbers of such changes. Advanced techniques for imaging brain activity, such as fMRI, reveal which brain regions are activated when a memory is recalled, but the resolution is far too low to study individual neurons, let alone individual synapses. As Luke Dittrich has so aptly shown, much of what we know about memory today still comes from studying the irreparable harm done to H.M.