Knowing that Henry could recount the gist of experiences from his preoperative life, we wondered whether these memories would provide the main content for his postoperative dreams. Dreams are the product of our imagination, akin to mental imagery when we are awake. They are typically disjointed, weird, and fleeting—not narratives that make sense. Experiments with rats have shown that dreams have meaningful connections with our waking lives. We became curious about Henry’s dream content, given that he could not remember what he had done the day before.25
In 1970, I asked the CRC nurses to quiz Henry about his dreams when they woke him up in the morning. His answers tended to be similar from one day to the next, even though the nurse who woke him up changed from day to day. The same nurse might have elicited the same dream report on successive days. On May 20, he said he was running or being carried over hills; on May 22, he said he was driving a truck over hills with farmers, in pursuit of cattle rustlers; on May 23, he was in hills, but there were no trees; on May 26, he said he was in the country near the ocean—hilly—“like Louisiana with a sharp drop-off”; on May 27, he was racing over a hilly field with young men, age about twenty, to reach a spot so they could rest and sleep; and on June 6 he was walking through green hilly country—no trees.
To get a clearer understanding of Henry’s dreams, we designed an experiment to document what he dreamed about. Our goal in this 1977 study was to establish what dreaming was like without a functioning hippocampus and amygdala. We monitored his sleep patterns during the night using electroencephalography (EEG), a tool that records electrical activity produced by neurons firing in the brain. These recordings told us what stage of sleep Henry was in. Two students helped capture Henry’s dreams by waking him up during REM sleep and non-REM sleep, asking him whether he was dreaming, and if so what his dream was about. Like healthy participants, Henry reported dreaming during both sleep states.
Were Henry’s dream reports genuine or merely anecdotes that he created on the spot in order to oblige his interlocutors? I suspect the latter. Of course, if Henry’s dreams—like those of most people—were based on his own experience, they would have to draw on preoperative events, because he had no pool of recent memories to fuel them. Henry’s dream reports were highly realistic and lacked the disjointed, unreal quality of most dreams. His typical responses too closely resembled his waking accounts of events in his youth—watching cowboy movies, enjoying nature, and taking car trips along Jacob’s Ladder Trail and the Mohawk Trail in western Massachusetts. He had a small repertoire of factual stories related to preoperative experiences. Instead of describing real dreams, Henry was likely doing his best to be a cooperative participant, and what he came up with was the gist of memories from his dimly lit past.
Here is an example from a night when a student woke Henry at 4:45 a.m., during a period of REM sleep:
Student: Henry?—Henry?
Henry: Yeah?
Student: Were you dreaming?
Henry: I don’t know. Why?
Student: You don’t remember anything?
Henry: Well, in a way I do.
Student: What do you remember?
Henry: Well, I was trying to figure out—a house in the country, and I can’t make out just how it’s laid out. And—believe it or not—I was dreaming that I was a surgeon.
Student: You were?
Henry: Yeah—a brain surgeon. And—’cause that’s what I wanted to be—but I said no because I wore glasses. And I said, well, a little speck [of dirt] or something, and that person could go [die] if you were performing an operation.
Student: Uh huh.
Henry: This is about the thought that I had about being a doctor, a surgeon, a brain surgeon. And that was—the whole line—I mean when I thought of that particular line of surgery.
Student: And you said something about being in the country? How’s that?
Henry: In the country—performing operations or just being in the country too. Level area—I think of that—level area and wanted to know about that myself. Doubly, in a way, because I know that Dad had been brought up down south and it was flat down there. And I of course was brought up in Connecticut, and I milked a cow up in Canada. And—
Student: Did this all happen now?
Henry: No, this was reality.
Student: Oh, but then you weren’t dreaming?
Henry: I was dreaming about the—putting it all together.
Henry’s immediate report of “a house in the country” was possibly an actual dream, but here we run up against the problem of his limited span of memory. For his dream reports to be convincing, they would have had to occur within the span of his immediate memory, roughly thirty seconds. After that, the dream content would have evaporated, and the rambling conversation that followed would have tapped into his older stored knowledge.
I have no evidence to support the conclusion that Henry did not dream, but if he did, his dream experience must have been unlike that of healthy people. Some of the brain areas normally engaged during dreaming were replaced in Henry’s brain by fluid-filled spaces. For instance, the amygdala in healthy participants is very active during REM sleep, and the absence of this activation in Henry’s brain likely altered his sleep patterns and ability to dream. Also, he sometimes had nocturnal seizures whose aftermath left him out of sorts the next day, but despite our best efforts, we simply do not know the particulars of Henry’s nighttime events. In 1977, when we woke him to ask whether he was dreaming, he sometimes said, “Yes” and other times “No.” This pattern of responses indicates that he understood and evaluated the question, and that he did not come up with any old narrative on the spot just to satisfy the researcher. Still, the nature and quality of his dream content remain a puzzle.26
Henry’s impairment in autobiographical remembering, both preoperatively and postoperatively, limited his self-awareness. He delighted in telling us tales about his relatives and childhood experiences, but they lacked precise details. His unfolding autobiography was deficient in the rich array of sensory and emotional narratives that form the intricate tapestry of who we are. Without the ability to travel consciously back in time from one episode to another, he was trapped in the here and now. Given these limitations, it is appropriate to ask whether Henry had a sense of who he was. Was his self-awareness blunted by his amnesia?
When people hear about Henry’s case, they frequently ask me, “What happened when Henry looked in the mirror?” If he could not remember anything since his late twenties, how did he adjust to seeing himself as a middle-aged and eventually an older gentleman? When Henry looked in the mirror, he never expressed shock or a lack of recognition; he was comfortable with the person he saw looking back at him. Once, a nurse asked him, “What do you think about how you look?” In his characteristic understated humor, Henry answered, “I’m not a boy.”
In the lab, we once showed Henry pictures of complex scenes, and he recognized them weeks later, based on a sense of familiarity and without explicitly recalling having seen them. Perhaps his own image did not surprise him for the same reason. Henry saw his face day in and day out for years. We know that the brain contains a region in the fusiform gyrus—a section of the temporal lobe that was preserved in Henry—specialized for processing faces. We also know that areas in the prefrontal cortex become active when people view their own face. These intact networks in Henry’s brain may have allowed him to perceive his own face as familiar, even as it changed, and to continuously update his mental self-image.
At the same time, Henry’s factual knowledge of his appearance and physical state was full of gaps. When we asked him his age or the current year, he often missed the mark by years or decades. He believed that he had dark brown hair even after he was partially gray, and described himself as “thin but heavy,” despite packing on pounds as he aged. He somehow reconciled memories of himself prior to the operation with his current appearance.
During the decades following Henry’s operation, his universe changed in
countless ways, but he was never shocked by these transformations. He unconsciously became familiar with new information in his environment as a result of repeated exposures day after day, which gave rise to slow learning over time—different from the way that healthy people learn about their world. During each encounter with his own face, with the people who took care of him, and with his environment, his brain automatically registered their features and integrated them into stored internal representations of objects and people. Otherwise, his gray hair would have surprised him, and he would have constantly wondered why he lived where he did, why the image on the television screen was in color, and what computers were. He somehow accepted these additions and innovations as they appeared in his life.
Being unable to establish new memories, Henry could not construct an autobiography as his life unfolded, and the narrative of his past was also sketchy. For many of us, our personal history is the most critical part of who we are, and we spend considerable time thinking about our past experiences and imagining how our stories will play out in the future. Our sense of self includes the story of our past and where we think we are going—our “to-do list.” We may imagine that we will advance in our careers, start families, or retire to better climates. In the short term, we have a plan for what to accomplish today, which friends we will see this week, and what we will do on our next vacation. Henry’s operation, in addition to depriving him of his declarative memory, prevented him from mentally traveling forward in time, in the short term or in the long term. He lacked the pieces to construct agendas for the next day, month, or year, and could not imagine future experiences. In 1992, when I asked him, “What do you think you’ll do tomorrow?” he answered, “Whatever is beneficial.”
Cognitive neuroscientists have called attention to the link between simulating future events and episodic retrieval. They identified a common brain circuit that is engaged in remembering the past and picturing the future. The process of imagining future events depends on medial temporal-lobe structures, the prefrontal cortex, and the posterior parietal cortex—the same areas critical for declarative memory. When we fantasize about our next vacation, we tap into long-term memory for details of past vacations and other knowledge. Remembering these past events and recombining them to create future scenarios requires the retrieval of information from long-term memory, and it is no surprise that amnesia interferes with this process. Constructing the future, like resurrecting the past, requires establishing functional connections between the hippocampus and areas in the frontal, cingulate, and parietal cortices. Without this network, Henry had no database to consult when asked what he would do the next day, week, or in the years to come. He could not imagine the future any more than he could remember the past.27
Eleven
Knowing Facts
The sharp contrast between Henry’s lack of episodic, autobiographical memory for the years before his operation and what seemed to be normal semantic memory for those same years raised the question of whether his episodic and semantic knowledge after his operation were equally affected. We had a wealth of evidence that his postoperative episodic memory was profoundly impaired—the hallmark of amnesia. But was Henry’s ability to acquire new semantic memories normal or deficient? To what extent could he learn and retain semantic information he encountered for the first time after his operation? We also wanted to explore how well the old semantic knowledge he had acquired before his operation fared during the years that followed. These questions drove numerous research projects in my lab.
Postoperatively, Henry had a normal attention span and could still talk, read, write, spell, and carry on a conversation, drawing on knowledge he had acquired preoperatively. He could retrieve the semantic information he absorbed before his 1953 operation because it was stored throughout his cortex, and he did not need his hippocampus to access it.
We were particularly interested in Henry’s language capacities because we wanted to know whether he needed his medial temporal-lobe structures to maintain his old semantic memories—knowledge about the world that he had squirreled away before 1953. A key component of semantic memory is lexical memory—stored information about words, including their meanings and forms (singular versus plural). The overarching issue was whether established lexical memories are preserved in amnesia. We designed our experiments to address three questions: Do medial temporal-lobe lesions compromise the capacity to use already-learned (preoperative) lexical information? Do these lesions affect grammatical processing? As time passes, does long-term lexical memory decay?1
When Henry spoke conversationally, his speech content was measured and deliberate. My colleagues and I wondered whether the characteristics of his speech processing were the same as in other people. In 1970, a graduate student in our lab proposed a test that would challenge Henry’s speech-processing mechanisms with linguistic ambiguities, sentences that have more than one meaning. Generally, there are three types of linguistic ambiguity—lexical, surface structure, and deep structure.
Lexical: When a strike was called it surprised everyone.
Surface structure: A moving van out of control is dangerous.
Deep structure: Visiting relatives can be a bore.
The student created sixty-five ambiguous sentences, including multiple sentences for each type of ambiguity, as well as twenty-five unambiguous sentences with one meaning (for instance, Jim bought a parka at the ski shop.)2
In the experiment, the student read the sentences aloud, and a prompt card in front of Henry asked, Does the sentence have one meaning or two? Normal participants could mentally restructure and correctly interpret the ambiguous sentences. Although Henry did not detect the presence of two meanings as often as our control subjects, when he did, he grasped all three kinds of linguistic ambiguity, including deep-structure ambiguity—as in Racing cars can be dangerous.
This study illustrated Henry’s normal capacity to retain, over several seconds, the various elements of a sentence and their relation to one another. He detected ambiguities less frequently than control participants because his short-term processing capacities became overloaded and his long-term memory failed to activate. Short-term memory can store a small amount of information temporarily, and although Henry’s brain could do that, its limited capacity was insufficient to disambiguate some of the sentences.
Around the same time, a psychologist at the University of California, Los Angeles, reported a similar set of experiments. Working independently when he was a graduate student in our department at MIT, he created thirty-two ambiguous sentences and read them aloud to Henry, giving him two instructions: “Find the two meanings of the sentence as fast as possible,” and “Say ‘yes’ and give the two meanings in the order you saw them.” This researcher imposed a ninety-second time limit, so if Henry did not find two meanings during that window, the trial was counted as an error. Henry detected the two meanings in sentences with lexical and surface ambiguities on more than eighty percent of the trials, but when given sentences with deep-structure ambiguities, he scored zero. Contradicting our previous results, the California researcher contended that structures removed from Henry’s brain—“the hippocampal system”—played a central role in language comprehension.3
This conclusion seemed incorrect to my team. We knew Henry to be conversational and improvisatory in talking with everyone he met. We also understood from decades of research dating back to the mid 1800s that language expression and comprehension were not localized to the hippocampal system. Many researchers, including myself, argued that language is mediated by multiple cortical circuits, primarily in the left hemisphere in most individuals, and not by the hippocampus or parahippocampal gyrus.
I decided to test the UCLA researcher’s findings by administering his sentences to Henry myself, with important changes in procedure. I asked him to read the sentences aloud, but also to reread any sentence from which he omitted a word; I called his attention to omissions by putting my finger under words he had overlooked. Then,
I asked him to say “yes” when he saw two meanings in the sentence and to describe them in the order in which he noticed them. Because Henry generally did everything slowly, I gave him unlimited time to interpret the sentences, instead of cutting him off after ninety seconds as the UCLA researcher had done. My results showed that Henry could indeed detect deep-structure ambiguities, provided he read the entire sentence without omissions, and provided he went at his own pace. Indeed, the fact that Henry could converse effectively—as he did every day with lab members and the CRC staff—attested to his ability to grasp the underlying meaning in sentences.
We continued to study Henry’s linguistic abilities until nearly the end of his life. We did not believe, as the UCLA researcher did, that his bilateral medial temporal-lobe lesions impaired his appreciation of linguistic ambiguity or any other speech-processing capacities. To make our case definitively, in 2001 a graduate student and a postdoctoral fellow in my lab gave Henry a host of tasks to assess his store of word knowledge and his ability to use grammatical rules. After examining his performance on nineteen tests, we concluded that his trouble detecting ambiguities did not result from a fundamental deficit in word knowledge or grammar.4
Henry could readily name objects depicted in color pictures and line drawings. For instance, the researchers showed him cards with a picture and a word. Half the time, the word matched the item in the picture, and the other half, the word and the picture were unrelated. Henry could distinguish between the matching and non-matching cards with only a slight deficit. Likewise, he performed well on tests of basic grammar. For instance, he could give the plural forms of nouns and the past-tense forms of verbs, and could turn an adjective into a noun (“The man is stupid. In fact, his ____________ is noticeable.”) Henry scored as well as control participants when we asked him to listen to a sentence and determine whether it was grammatically wrong.5
Permanent Present Tense Page 28
