It was remarkable that Henry could remember the plan of a house he had never seen before his operation. Walking from room to room, day after day for sixteen years allowed him to construct a mental map of the house over time. But his knowledge was more than a vague sense of what was where. For example, he could picture his house in his mind’s eye and tell me which way to turn to get from his bedroom to the bathroom, and where the front and back doors were located. His ability to recall the address in conjunction with the layout suggests that this home had become part of his world knowledge; this was information he should not have been able to learn.
Acquiring the floor plan of 63 Crescent Drive occurred without conscious awareness of the learning process, and with Henry’s attention focused on other things. Habit learning is also unconscious, but habits are noncognitive—automatic, involuntary, and inflexible. Henry’s spatial knowledge of his house was cognitive. He could use his spatial knowledge to picture the rooms in his house voluntarily, in relation to one another, and to describe consciously the route from point A to point B. This flexibility in navigating an internalized spatial map is markedly different from a habit.
Not until we saw the MRIs of Henry’s brain did we understand his remarkable capacity to draw the floor plan. By the 1990s, scientists had uncovered a network of brain regions, including the hippocampus and areas in the cortex, which are engaged when remembering the topography of spaces. Once we could see the precise structures that had been removed or spared in Henry’s brain, we discovered that some components of this brain network for processing information about space were still present. They included specific areas in the parietal, temporal, and occipital lobes—the somatosensory cortex, parietoinsular vestibular cortex, visual cortex, part of the posterior parietal cortex, inferotemporal cortex, and posterior cingulate/retrosplenial cortex.16
Clearly, enough tissue remained to form a memory of the house he had navigated countless times for years—a depth of exposure that we had not captured in our tests of Henry’s ability to learn. In navigating the rooms of his house every day, he learned via the same process of immersion that someone trying to learn a foreign language might engage. Simply by following his daily routine, he enriched his mental map in small increments day after day after day—a perfect example of learning by mere exposure.
This astonishing evidence of spatial knowledge, acquired slowly over time, raised the intriguing question of whether this ability would extend to a test of spatial orientation in the laboratory. Henry’s amnesia was not a detriment in this investigation because this spatial task did not rely on his long-term memory. We sought to discover whether, without a functioning hippocampus, Henry could create a mental cognitive map in a testing room.
During four of Henry’s visits to the CRC from 1977 to 1983, we assessed his spatial ability using a route-finding task. The test’s goal was to document his capacity to follow a route on a handheld map when walking from one landmark to another. Testing took place in a specially outfitted room at the CRC. Embedded in the tan wall-to-wall carpeting were nine red circles about six inches in diameter, in a three-by-three array. Henry held a large map that depicted the nine red circles on the floor as black dots. A path from dot to dot was drawn in heavy black lines, with a circle around the starting point and an arrowhead at the end. The test consisted of fifteen such maps. The letter N, designating north, was marked on each map, and a large red N was fixed on a wall of the room. Henry’s task was to walk from dot to dot along the path that corresponded to the one on the map. He was not allowed to turn the map, so as he walked, the map was not always in the same orientation as the room. North was always at the top of the map, but when he turned, the N on the wall could be to his left, his right, in front of him, or behind him. As a result, he had to make a series of mental translations from the coordinate system of the map to the directional correlates of the room. Henry walked patiently from dot to dot, but was usually unable to follow the path indicated on the map, and his performance did not improve with repeated testing using the same maps (see Fig. 6).
The network of brain circuits that enabled him to draw the floor plan could not support his performance on this laboratory test. He needed his hippocampi for map reading. Without these structures, Henry could not fathom the relation between the start and finish points on the CRC floor plan, and could not reconcile the changing position of his body with the static coordinates of the room.
Henry’s successful acquisition of a cognitive map of his house and his failure on the route-finding task seem contradictory. But the tasks themselves are fundamentally different. Through countless hours of practice, Henry slowly learned the geography of his house, without awareness and without consciously referring back to his declarative memory store. While the map test at the CRC was not a memory task, it did require Henry to form an instantaneous cognitive map, a task he could not perform without his hippocampi.
During the 1990s, as scientific knowledge grew with respect to how the brain processes complex thought, my colleagues and I continued to ask whether the preserved areas in Henry’s brain could support new learning about the physical world around him. In 1998, a young neuroscientist at the University of Arizona studied patients who had received small lesions to their right hippocampus or right parahippocampal cortex to alleviate their epilepsy. Patients with right hippocampal lesions were unimpaired on a spatial memory task, but those with damage to the right parahippocampal cortex exhibited severe impairment, suggesting that the parahippocampal cortex is vital for spatial memory. Henry’s right parahippocampal cortex was partially intact, so we wondered whether this part of his brain could support new place learning—the ability to navigate to a hidden target. To test this hypothesis, the Arizona researcher traveled to Boston to give Henry a simple spatial memory task, which he performed over nine test days during two visits to my lab in 1998.
On the first trial, the researcher told Henry that there was a sensor hidden under a small carpet, but she did not show him where it was. The testing room was filled with objects—desks, chairs, shelves, and a door—which Henry could use to get his bearings in relation to the environment. On this learning task, Henry had to find the invisible sensor by chance, remember its location, and then find it again from memory. Before each trial, the researcher triggered the sensor by stepping on it while Henry looked the other way. She asked him to find the place under the carpet that would produce the sound when he stepped on it. Because the sound triggered by the sensor came from a distant speaker, Henry could not rely on the sound to find the sensor’s location. He was highly motivated in his search, even though he had to use his walker to navigate. On the first trial, he found the target. On fifty-four percent of subsequent trials, he walked directly toward the center of the carpet, and from there, on eighty percent of the trials, he took a direct path to the hidden sensor.17
Henry’s ability to locate the sensor was remarkable, given his severe amnesia and inability to explicitly recollect the testing episode or having heard the sound. His achievement on this task underscores the role of the parahippocampal cortex, of which three-quarters of an inch remained in his brain, in spatial memory. We know that he was not relying on his short-term or working memory to perform the task because more than sixty percent of his hits—finding the sensor—occurred a day after the first test session. His capacity to locate the sensor indicated that he was capable of limited long-term memory formation, and that structures beyond the hippocampus could support navigation.18
But because Henry was unable to consciously recollect any details of the testing episodes, we concluded that learning the sensor’s location was nondeclarative—learning that took place independently of the medial temporal lobe. Henry repeatedly found the sensor because of his implicit, nondeclarative knowledge of the target location. Whether his preserved parahippocampal cortex was solely responsible for his successful place learning was unclear because several other intact structures, like the striatum (located under the frontal lobes), could a
lso have mediated this learning.
Although it took decades to discover the detailed anatomy of Henry’s removal, one anatomical fact that we could hang our hats on from the beginning was that a large chunk of hippocampal tissue was missing on both sides of his brain. The maze experiments that Milner and I conducted with Henry helped establish the importance of the hippocampus for spatial learning. The later finding that he also performed poorly on the map-reading test, which did not require memory, indicated that his spatial deficit went beyond learning. Without his hippocampus, he could not process complex spatial information efficiently. He lacked the capacity to create a cognitive map in the usual sense. Other test results, however, highlighted an exception to the cognitive map theory and suggested a fractionation of spatial memory. Henry’s unexpected drawings of the house where he lived when neither hippocampus was in working condition means that other brain areas took over the job of encoding and storing that rich spatial information. An indication of a specific brain structure that Henry likely engaged when drawing his floor plan came from the spatial memory task on which he was able to find the sensor hidden under the rug. Previous work had shown that this task depended on the parahippocampal gyrus, part of which remained in Henry’s brain on both sides. So, on rare occasions, he somehow compensated for the devastating effect of his hippocampal damage by mobilizing preserved brain structures and networks.
A basic requirement for memory formation is intact perception. Henry passed this hurdle for sight, hearing, and touch, allowing my colleagues and me to test his learning and memory across these sensory modalities. His amnesia permeated all kinds of declarative memory without regard to the sensory portal that ushered in the to-be-remembered information. We consistently documented Henry’s disorder with a wide range of test stimuli—words, stories, faces, pictures, scenes, mazes, puzzles, and more. Observations of his everyday behavior supplemented the knowledge that we gleaned from extensive formal testing in the laboratory, giving us a full picture of his postsurgery life.
Six
“An Argument with Myself”
Henry seldom shared his introspections with anyone, so, for the most part, we had to infer his emotional life from observing his behavior. During our conversations with him, he seemed happy and content; he smiled often and rarely complained. You might imagine that if you were in his shoes, you would be habitually anxious, concerned that your behavior had been improper and fearful of what tomorrow would bring. But no one would describe Henry as a nervous or worried man. It is possible that his operation, which excised part of his emotional brain, protected him from the scary realities of his life. Still, he did have occasional dark episodes when he could become frustrated, sad, aggressive, or uneasy. Typically, these negative emotions would dissipate as soon as he was distracted.
At the time of Henry’s first visit to the CRC in 1966, his mother was in the Hartford Hospital after undergoing minor surgery. His father packed Henry’s clothes and brought him to Scoville’s office in Hartford, where Teuber had offered to pick him up and take him to Cambridge. Henry and his father had visited his mother in the hospital that morning, and by the time he met up with Teuber, Henry had only a vague sense that something was wrong with her. When Teuber asked who packed his bag for him, he answered, “Seems like it was my mother. But then that’s what I’m not sure about. If there is something wrong with my mother, then it could have been my father.” During the trip to Cambridge, Teuber repeatedly explained to Henry where his mother was and that she was fine, but Henry had a lingering feeling of uneasiness about his parents, wondering whether all was well. As he settled into his room at the CRC, his anxiety dissolved. We told him that he could call home, but he no longer knew why he would want to do so. The next afternoon, however, Henry told a nurse that he thought his mother was in the hospital or had heart trouble. He had regained some inkling overnight that his mother was ill.
At that time, it was unclear what would account for the recovery of this memory; we speculated that Henry was simply less tired than the day before. Since then, however, numerous experiments in animals and humans have demonstrated that sleep sometimes improves memory consolidation. During sleep, memories may be reactivated and replayed, making them stronger and less susceptible to disruption. Different kinds of memory are enhanced by different stages of sleep, which in turn recruit different brain structures. For example, conscious, declarative memory performance benefits from deep (slow-wave) sleep, while unconscious, nondeclarative memory is enhanced by light (rapid eye movement, REM) sleep. Researchers have also found that REM sleep improves memory for emotional (especially negative) information more than non-emotional. According to the CRC nurse, he “slept fairly well” during his first night, and activation in preserved brain areas, including emotion circuits, may have strengthened a fragmentary memory of his mother’s illness so that it surfaced the next day.1
The hospitalization of Henry’s mother had dual representation in his brain, consisting of a factual element and an emotional element. He quickly lost the factual content—Mother is in the hospital having minor surgery—but the vague emotional content—something is wrong—lingered for days. Without a functioning hippocampal circuit, Henry could not preserve the facts about his mother’s hospital stay in his long-term memory; but a larger network of brain areas, the limbic system and its connections, helped maintain his anxiety. The emotional component of the situation had privileged access and processing that helped establish an affective memory trace. Limbic is an anatomical term that means margin, referring in this case to a continuous band of cortical and subcortical structures nearest the boundary of the cortical covering. In 1877, this ring of cortex was originally believed to be concerned with the sense of smell, but a new proposal in 1937 described it as an anatomical basis for emotional behavior (see Fig. 7).2
In the 1937 version of the circuit, information traveled from one brain area to another in a loop—the hippocampal formation to the mammillary bodies of the hypothalamus to the anterior thalamus to the cingulate cortex to the parahippocampal gyrus and back to the hippocampal formation. In 1952, another researcher added the amygdala to the circuitry. The hippocampus is no longer believed to mediate emotions, whereas the amygdala is seen as the hub of emotional responses. This complex structure receives information from all the senses and from areas that process feelings of well-being and distress. The amygdala also sends information back to many of the same areas, creating vast networks specialized for emotional perception, expression, and memory. A mounting number of studies reject the idea that different emotions map one to one onto specific brain areas, recognizing instead that each emotional response evolves from a collaboration among many brain areas. By this view, the brain stores individual emotional experiences, for all kinds of emotion, by recruiting a wide range of brain structures in the limbic system and beyond—networks that support basic cognitive operations, both emotional and non-emotional. Henry’s brain was occasionally able to create such networks.3
The removal of Henry’s amygdala and hippocampus caused a malfunction in his basic limbic circuitry, so it was reasonable to expect that his ability to process emotions might be altered. But we learned from our earliest studies of Henry that he could experience a range of emotions. During his first visit to MIT in 1966, a CRC nurse woke Henry up each day at 4:00 a.m. to take his vital signs. On these occasions, she chatted with him briefly and then made careful notes in his chart. On eight of the sixteen nights of his stay, he inquired where his parents were, and whether they were OK. With parts of his distributed limbic system still functioning, he could feel anxious about his parents—and was doomed to re-experience this emotion.
Memory can be a burden: it forces us to revisit unpleasant events from the past. But without his memory, Henry could never properly mourn or process the losses that are an inevitable part of life. He did not remember that a favorite uncle had died in 1950. According to his mother, he became distraught each time he heard the news. As this emotion gradually fa
ded with the passage of time, he would occasionally ask when his uncle would visit again.
In 1966, the same year as his first visit to the CRC, Henry suffered a tremendous loss. In December, when Henry was forty, his father died of emphysema at Saint Francis Hospital in Hartford. Mrs. Molaison told me that Henry was quite depressed after her husband’s death but did not consciously grasp that his father was gone unless someone reminded him. She told us that at one point, Henry became angry and rushed out of the house when he discovered that some of his prized guns were missing, claimed by an uncle after his father’s death. When the uncle learned that Henry was upset, he returned the guns, and Henry’s anger subsided. He was disturbed by the disappearance of the gun collection because it was a focal point of his world. The guns had been on display in his room every day since his youth, so naturally their absence was noticeable and distressing. They also represented an emotional bond with his father and were prized possessions in their own right.
For at least four years, Henry was unable to articulate the fact that his father had died. Seven months after Mr. Molaison’s death, Mrs. Molaison asked us not to tell Henry that his father was dead because he might react as if hearing the sad news for the very first time. My inclination was the same: I did not question him explicitly about his father because I knew it would upset him. But in August 1968, when I was testing him, he talked about his father in the past tense, so over time, his brain may have absorbed the painful fact into unconscious memory traces that stored it. Still, he would have moments of uncertainty. Without a functioning hippocampus and amygdala, Henry did not form long-term emotional memories; instead, he used what he had at his disposal—numerous interconnected cortical areas that stored preoperative memories of his father, his home life growing up, and the concept of death. Over time, he gradually connected the dots and understood on some level that his father was gone for good.
Permanent Present Tense Page 12