Henry’s performance reflected his personal semantic knowledge—where he came from, his family history, and his own past. He had a general sense of identity. But, his deficient autobiographical knowledge—personal, unique episodes—meant that his self-awareness was significantly limited.
Memory researchers have long wondered why retrograde amnesia often seems to unfurl backward in time from the onset of anterograde amnesia, so that the distant past is most vivid and recent years fade away. One theory, the Standard Model of Memory Consolidation, posits that memories are consolidated—become fixed—over a long period that could last from months to decades. Psychologists Georg Elias Müller and Alfons Pilzecker at the University of Göttingen first proposed this theory in 1900. In the mid 1990s, neuroscientist Larry Squire and his colleagues at the University of California, San Diego, adopted the Standard Model as the centerpiece of their thinking about retrograde amnesia. According to this theory, the brain needs the hippocampal system during the early stages of consolidation to store and recover memories, but over time this system becomes unnecessary as areas in the frontal, temporal, parietal, and occipital lobes take over the responsibility for maintaining all memories over the long term. Once the memories are solidly stored, they no longer rely on the hippocampal system to be accessed again. In short, the hippocampal network plays only a temporary role for all forms of memory. According to this model, more recent memories are lost in cases of amnesia and dementia because these newer memories have not been fully consolidated and still depend on the hippocampal system.16
The flaw in the Standard Model is that it assumes that all memories are processed in the same way, whether they consist of general knowledge of the world (semantic) or personal experiences (episodic). Henry’s recollections pointed to an important distinction between the two types of memory: he could remember facts he had learned before his operation, but when asked to recount what happened in his personal life at specific moments in time, he struggled. Declarative memories are not all processed in the same way. Through studying Henry, we learned that the ability to store and retrieve autobiographical memories that are detailed and vivid always depends on the hippocampal system, whereas remembering facts and general information does not.17
A better model for understanding Henry’s retrograde amnesia emerged in the late 1990s, when neuroscientists Lynn Nadel and Morris Moscovitch introduced the Multiple Trace Theory of Memory Consolidation. Although Henry’s case did not contribute to their original thinking, our findings with Henry provided strong support for their views. This theory draws on Tulving’s proposal that we process facts differently from unique experiences. Like the Standard Model of Consolidation, the Multiple Trace Theory recognizes that semantic memories—world knowledge—can eventually become independent of the hippocampal system because they do not require us to remember the context in which we learned them or to connect and relate pieces of stored information. For instance, we do not remember that we learned the year of Columbus’s first voyage to the Americas while sitting in the back row of a second-grade classroom; we simply remember 1492. Remembering how we celebrated our twenty-first birthday, however, requires accessing the specifics of when, where, and what happened. According to the Multiple Trace Theory, it is possible to retrieve facts—like 1492—without the hippocampal system, but it is not possible to retrieve any unique experiences—like the birthday celebration—unless the hippocampal circuits can communicate with the cortical circuits. Evidence from Henry’s case favors the Multiple Trace Theory over the Standard Model of Consolidation because it indicates that traces of autobiographical events forever rely on the hippocampal system for retention and retrieval.18
The Multiple Trace Theory offers an alternate explanation for why people with amnesia are more likely to remember early experiences than later ones. According to this theory, neural processes in the hippocampus provide pointers, or an index, to all the distant sites in the cortex where memories of our experiences are stored. Think of this process in terms of a visit to a local library, where we look up a subject such as “birds of the Caribbean” in the card catalog, and then peruse the shelves to find the books. Each time we activate a memory trace by reminiscing about it or attaching a new piece of information to it, we create a new pointer, a new entry in the card catalog, to that memory. So, our memory of an exciting phone call with a job offer may be linked to multiple pointers within our hippocampal system through a web of associations from the times we have relived the moment or told other people about it. In this model, earlier memories have had a chance to knit themselves more firmly into the brain by accumulating these pointers over time. Retrograde amnesia hits newer memories harder because they are linked to fewer of these anchors and are, therefore, more vulnerable to being wiped away.
The controversy between the Standard Model of Consolidation and the Multiple Trace Theory focuses on autobiographical memory. Unlike semantic or general memory, autobiographical memory is episodic and rich in detail, and specifically includes what researchers call experience-near details—the specific sounds, sights, tastes, smells, thoughts, and emotions that accompany a unique event. The Standard Model assumes that autobiographical memories depend on a functioning hippocampus only for a limited time; after that, they become independent of the hippocampus and are stored in the cortex. This theory would predict, therefore, that Henry’s autobiographical memory for his preoperative years was intact.
In contrast, the Multiple Trace Theory posits that retrieval of autobiographical episodes always requires hippocampal engagement. According to this view, the hippocampus is a memory index or pointer to the cortical areas that store the sensory and emotional qualities that make up episodic events. If the Multiple Trace Theory is correct, then Henry’s ability to recollect autobiographical events in his early life would be impaired.
Henry’s reports of preoperative autobiographical episodes were sparse; I am aware of only two. He related the first to Brenda Milner in the 1950s when he described a significant event that occurred when he was ten. “I can remember the first cigarette I ever smoked. It was a Chesterfield; I took it from my father’s cigarettes. I took one mouthful, and did I cough! You should have heard me.” For decades, this was the only autobiographical memory Henry conveyed to us.19
It was not until 2002 that we learned about a second autobiographical experience from Henry’s past. At that point, we were systematically examining his autobiographical memory using a newly developed structured interview designed to elicit personal details. Sarah Steinvorth, a postdoctoral fellow in my lab, devoted several sessions to Henry’s interviews, during which she asked him to describe an event from each of five life periods: childhood, teenage years, early adulthood, middle age, and the year prior to testing. She then asked him for as many details as possible about that event. If he had trouble thinking of an event, she would help by suggesting typical life events, such as a wedding or moving to a new house. The key to this study was patience and persistence. Steinvorth might have spent half an hour suggesting possible events to elicit a memory before moving on to another time period. She also discouraged Henry from choosing an event he had recounted repeatedly in the past. (We all have stories like these—events that we have retold so many times that we eventually narrate them blandly, without vividly reliving the wealth of sensory experiences.) As she questioned Henry, Steinvorth talked him through the different time periods of his life, and he struggled to produce any details. When she asked him for an event in his childhood, for instance, he mentioned falling in love with a girl whose father was a police captain, but was unable to describe a specific episode—something that occurred at a particular time and place—associated with that experience.20
Then one day, Henry delighted Steinvorth with an astonishing narrative.
“Can you think about one specific event, something that lasted several hours, from your early childhood up to age eleven?” she persisted. “Can you come up with something like that?”
“No, I can
’t,” Henry said.
“Would you like to go to another time period, and see if you can come up with anything from that time period?”
“It would be better, in a way,” he agreed.
“Okay. Let’s try that. So maybe you can think of something—can you think about one specific event you were personally involved in that happened when you were between eleven and eighteen years old?”
He was silent for a while, so she repeated the question, but he still did not respond.
“Henry, are you tired? Would you like to have a break? Or are you just—”
“I’m trying to think.”
“Okay, I’m sorry. I didn’t want to interrupt you.”
“I think of the first plane flight.”
“Say that again.”
“The first plane flight.”
“Your first plane flight?”
“Yeah.”
“Tell me about it.”
Henry proceeded to describe, in great detail, the experience of a half-hour “sky ride” in a single-engine airplane when he was thirteen. He accurately described the layout of the Ryan aircraft, its instruments, steering column, and spinning propeller. He and the pilot sat side by side, and at one point, the pilot allowed him to take the controls; he recalled having to stretch his legs to reach the foot pedals. As Steinvorth questioned him, he remembered that it was a cloudy day in June, and as they flew around Hartford, he could see the buildings along Main Street. As they approached the airport to land, they passed over a cove where boats were docked. For the second time since his operation, a memory from the past was replete with experience—near details, the excitement of a specific event, complete with landmarks, colors, and sounds.
I was amazed when I read the transcript of Steinvorth’s interview; I had no recollection of having heard this story before, but as it turns out, I was mistaken. In the course of writing this book, I reread an interview we conducted with Henry in 1977 while we were putting electrodes on his scalp for an all-night sleep study. He was chatting casually with the researcher when he related a similar narrative:
And Brainard Field. Well, I know that too. Both before they had the transports in there, and when there were just private planes there. Why I remember is because in ’39, that’s where I went up. When I flew. Yeah. In an airplane. And I’m always glad I handled it too. That was something right there. ’Cause I know my mother and my dad were both afraid of airplanes. And when I went up, it was just before I graduated. It was because I was going to graduate that I was able to go up—about two and a half bucks. ’Cause the guy who took me up then was from—private pilot from Rockville—and he worked for the [unintelligible]. I got a little extra ride.
When I rediscovered this additional piece of evidence, obtained twenty-five years earlier, I was convinced that Henry’s memory of the plane ride was genuinely autobiographical, and not a confabulation.
During the 2002 interview, Steinvorth asked Henry if he could remember any other specific events from his past. Although he came up with some promising attempts, none approached the vividness of the plane ride. He talked about taking a train trip with his mother when he was about seven, but simply recited facts, not unique episodes. They boarded the train in Hartford, switched trains in New York, and rode the second train all the way to Florida. He recalled sleeping on the top bunk while his mother took the lower one, and eating meals on the train. He also remembered visiting Canada with his parents while still in grammar school. He talked about milking a cow during that visit, and when Steinvorth quizzed him about the details, he could offer some information—sitting on a stool, being inside a barn with about twenty other cows, having to pull one teat and then the other. But again, despite her nudging, he could not paint the picture in a way that distinguished it as a genuine reliving of the experience rather than as a general description of milking a cow. It did not qualify as an autobiographical memory.
Henry’s intense memories of his first cigarette and his airplane ride stood in marked contrast to his indistinct memories for the rest of his preoperative life. His ability to vividly recall these two extraordinary episodes is grounded in the powerful emotions he felt during both autobiographical experiences. Steinvorth asked Henry to rate his plane ride from one to six in terms of how much his emotional state changed during the flight, how personally important it was to him at the time of the flight, and how important it was to him at the time of testing. Henry rated his emotional state during the flight as six—a “tremendous emotional change.” He rated the personal importance of the experience at the time of the flight, in retrospect, as five and at the time of testing as six. The memory traces of this unique experience stuck in Henry’s brain because its salience and emotional significance strongly activated brain areas that support vivid encoding and storage of emotional information—his hippocampus, prefrontal cortex, and amygdala. This activation would have occurred not only at the time of the actual ride but also every time he told his friends about it afterward. Over time, the thrilling memory became more and more robust, a rich representation available for retrieval decades later.21
In contrast to his difficulty recounting specific episodes from his past, Henry consistently performed much better on tests of his remote semantic memory, general facts about the world. In one such test, Steinvorth asked him to focus on the public event itself rather than on his personal experience when he learned about the event—in other words, the semantic, not the episodic, knowledge. She gave him a list of cues and asked him to recall a specific public event, such as a major crime or the wedding of a celebrity, during different time periods in his life. Henry could recall public events from each of the preamnesic time periods. For instance, he chose to talk about “a big accident,” and proceeded to describe the 1937 Hindenburg disaster in some detail. His ability to summon this type of general information lent further evidence to the idea that autobiographical and semantic memory are stored and retrieved in different ways. His intact memories of public events provided compelling support for the conclusion that his autobiographical memory deficit could not be attributed to a general failure in retrieving, recalling, or describing a detailed narrative structure.22
Scientists continue to debate the merits of the competing Standard Model of Consolidation and the Multiple Trace Theory. Our own results with Henry are consistent with the Multiple Trace Theory: Henry’s ability to remember semantic information from his early, preoperative life remained strong, but his damaged hippocampus prevented him from remembering almost all autobiographical events. The two memories he could elicit—the first cigarette and the plane ride—were startling exceptions that revealed two outstanding moments in his life.
In an attempt to explain the nature of Henry’s amnesia using the Standard Model, Squire raised two issues. First, he suggested that Henry’s inability to retrieve preoperative autobiographical memories might have resulted from age-related disease, basing this suggestion on abnormalities in Henry’s 2002–2004 brain scans. We can rule out this explanation, however, because we have evidence of this autobiographical memory vacuum in my 1992 interview with Henry, in which he not could retrieve an episodic memory related to his mother or his father. At that time, his brain did not show any age-related abnormalities.23
Squire’s second explanation for Henry’s lack of preoperative autobiographical memories was that Henry may have had autobiographical memories soon after his operation, but these memory traces withered away over time. If this reasoning is correct, then the same logic should apply to his memories of public events. They also should have dissipated over the years, but they did not. Henry’s normal performance on the Public Events Interview showed that he could vividly recall public events from the very same years when his autobiographical memories were missing. His semantic-memory performance was spared, but his episodic, autobiographical memory was impaired. Based on our experiments, I stand by the view, in agreement with the Multiple Trace Theory, that we need a functioning hippocampus to re-experience un
ique moments in our past, regardless of how long ago they were acquired. An increasing number of studies also support this theory.24
A critical question in weighing the Standard Model of Consolidation against the Multiple Trace Theory is whether amnesia affects episodic, autobiographical memory and semantic memory differently. The two theories make clear predictions, and Henry’s case exposed the difference between them. His test results, supporting the Multiple Trace Theory, taught us that the network of brain areas that supports the retrieval of remote autobiographical information is distinct from the network that sustains the recovery of remote semantic information. The former is impaired in amnesia, while the latter is not. Medial temporal-lobe structures are engaged in the initial encoding, storage, and retrieval of both kinds of memories. Then, during the process of consolidation, semantic memories become permanently established in the cortex, while episodic, autobiographical-memory traces continue to depend on medial temporal-lobe structures indefinitely. Thus, as far as we know, the removal of this tissue from Henry’s brain left him with only two autobiographical memories.
In the late 1970s we did not know how important sleep is for memory consolidation, and we did not understand its central role in neural plasticity. At that time, we knew little about the neural basis of dreams, and studies of the cognitive neuroscience of dreaming did not exist. What we did know was that there is a relation between eye movements and different stages of sleep, and between different stages of sleep and dreaming. Armed with this basic knowledge, we set out to examine the effect of Henry’s massive medial temporal-lobe damage on his dreams. We were tantalized by the Freudian possibility that his dream reports would expose stored secrets that we had been unable to elicit through his conscious recollections. Would we get a glimpse of his unconscious wishes?
Permanent Present Tense Page 27
