Psychology at Sweet Briar

The Case Study of H.M.
Kristi Chatham '00

      In an attempt to learn more about the brain, case studies are often used to provide a detailed examination of a patient's condition resulting from brain damage. Researching by this method has contributed a great deal to our understanding of the brain's structures and the mechanisms by which they function. In particular, the in-depth study of a patient referred to as H.M. has contributed to the development and acceptance of many theories in the areas of cognition and neuropsychology. Specifically, a detailed investigation of his condition has significantly expanded our knowledge about memory and language.

      At the age of 27, H.M. underwent a surgical procedure called a bilateral medial temporal resection that was rendered as a last resort to alleviate him of debilitating epileptic seizures. Bilateral temporal lobe resection involves ''destroying bilaterally the anterior two-thirds of the hippocampus and hippocampal gyrus, as well as the uncus and amygdala, but sparing the lateral neocortex'' (Milner, 1970, p. 34). This procedure was successful in abating the seizures; however, it left him with some damaging side effects. H.M. now suffers from severe memory impairment; a condition referred to as anterograde amnesia. Although he is able to remember events early in his life, prior to the surgery, he is unable to remember events occurring after the surgery (Milner, 1970).

      Soon after the procedure, researchers began evaluating H.M.'s condition and observed that he could remember new information as long as his attention was focused on it. Yet, as soon as his attention shifted, he would forget. For example, H.M. was able to remember the number 584 for over 15 minutes. However, when his attention was directed away from the number, to a new topic, he not only forgot the number, but also forgot he had been given a number. The ability to remember the numbers for this extended period of time has been explained by continuous verbal rehearsal. The use of continuous verbal rehearsal in retaining new information has been supported by research examining H.M.'s inability to remember nonverbal stimuli. When given simple tasks to perform, such as determine if two nonverbal stimuli are similar or different over a delay period of about 60 seconds, H.M.'s performance was remarkably lower than that of normal subjects (Milner, 1970).

      Another study was done to examine H.M.'s ability to learn. During a maze learning study, H.M. was asked to learn the correct path through a 28-point stylus maze. This task proved to be very difficult for him. After repeated trials, he did not exhibit any signs of advancement. However, when given a shorter maze, H.M.'s performance considerably improved. Not only was he able to remember the maze without error, but learning curves showed that he retained what he learned. Although he did not explicitly remember learning the maze, two years later his performance at this task revealed that he had maintained some memory of it. It has also been suggested that the significant difference in H.M.'s performance on the long versus the short maze is the result of his inability to verbally rehearse past his span of immediate memory. When the maze exceeded his immediate memory, H.M. completely failed at remembering the maze (Milner, 1970).

      Evaluating H.M's performance on tasks such as these has enabled researchers to answer many questions about two controversial theories, stage-of-processing and distributed-memory theories, which have been debated for some time. Both of these theories have been used to examine the relationship between memory and language; however, more recent research on H.M. shows stronger support for the distributed-memory theories.

      According to the stage-of processing theories, "language and memory are separate units and processes that are organized into four sequentially ordered stages or processing modules: comprehension, storage, retrieval, and production" (MacKay, Burke, & Stewart, 1998, p 29). Some aspects of H.M.'s condition, involving language and memory support these theories. In particular, separate modules for 'long term and short term' and 'implicit and explicit' memories have been defined to explain his condition. An example of this would be in the maze learning study, H.M. did not remember learning the maze; however, his greatly improved performance showed that he actually had. This finding supports the stages-of'-processing theories for the existence of different modules for implicit and explicit memories. One benefit of the stages-of-processing theories is that subdividing the stages of memory easily creates new modules, which can be used to explain a variety of conditions (MacKay, Burke, & Stewart, 1998).

      According to the distributed-memory theories, memory and language depend on connections between 'nodes' or neural units. One theory, in particular, that has been focused on to explain H.M.'s language disabilities is the Node Structure theory (NST). This theory suggests that there are three processes involved during language production and comprehension: "activation processes; the strengthening of existing or already established connections; and the formation of new connections between nodes" (MacKay, Burke, & Stewart, 1998, p.33).

      The mechanism used to form new connections is extremely important in understanding the underlying causes of H.M.'s condition. According to the Node Structure theory, his inability to remember new information is believed to be caused by his incapacity to form new connections between nodes. New connections between nodes are formed by 'binding nodes' located within the hippocampus. This theory has led researchers to suggest that the surgical removal of part of H.M.'s hippocampus damaged the 'binding nodes', necessary to form new connections and new memories. However, his ability to remember or comprehend, a familiar word or a memory from youth, would not require input from the hippocampus. The recall of a familiar word would be activated through a process known as engrainment. As a result, the connections between the already formed nodes would be strengthened (MacKay, Stewart, & Burke, 1998).

      Recently there have been numerous observations within H.M.'s case that contradict the previous belief that he could complete the other stages of processing involving comprehension, retrieval, and production. For example, after successfully completing a set of 'nonverbal' task, he was unable to describe what he was doing. This left some researchers to suggest that he also may have a language-related deficit (MacKay, Burke, & Stewart, 1998).

      After re-evaluating his case, it appears that H.M. does in fact suffer from a language production and comprehension impairment. These inferences were made after re-examining three studies done between 1970-1973. In Study 1, H.M.'s ability to describe two meanings of an ambiguous sentence was examined. In study 2, the coherence, usage of proper grammar, and comprehension of H.M.'s descriptions of ambiguous statements were examined and compared to normal subjects to evaluate his ability to communicate. In study 3, H.M.'s speech about events happening prior to the surgery were compared to those of normal subjects. The results from these studies indicate that H.M. does suffer from language deficits (MacKay, Stewart, & Burke, 1998).

      These studies support the belief that H.M. has difficulty comprehending language. Indications of his deficit include: problems with explaining the first meaning, inability to repeat single meanings, free association of his own output, impossible interpretations, unusual pronoun use, blending of two concepts, failure to follow the experimenters requests, and misreadings. According to MacKay, Stewart, and Burke (1998) many of these observations can not be explained solely by his memory deficit or by the stages-of-processing theories. However, the distributed-memory theories have been able to explain H.M.'s language deficit.

      Another important point to make about H.M.'s ability to detect ambiguity, in studies 1, 2, and 3 (described above), is that this task requires two connections to be made between the same pair of nodes. This process is even more difficult for the damaged hippocampus to accomplish. In order to explain H.M.'s difficulties MacKay, Stewart, and Burke (1998) have proposed the 'inference hypothesis'. According to this hypothesis, the newly formed weak connections between nodes enable him from describing the multiple meanings of the word at the same time. This occurs because the act of describing the first meaning interferes with recalling the second meaning of the word (MacKay, Stewart, & Burke, 1998).

      Some have rejected the belief that H.M. suffers from a comprehension and language deficit. Therefore, in an attempt to explain the evidence supporting this notion, it has been suggested that damage to his limbic system, which is responsible for motivation, might have effected his performance on the tasks. Maybe H.M. did not follow the experimenter's requests because he was not interested or motivated to do so. However, throughout the experiment H.M. expressed frustrations about his inability to complete the tasks. This would not have been true if he was merely not motivated (MacKay, Stewart, & Burke, 1998).

      Research has also been done to determine if H.M. suffers from visual perception deficits. In an attempt to examine his visual perceptive abilities 'hidden figure' studies has been done. In this study, H.M. was given the task of identifying hidden figures and target figures. The hidden figures were defined as familiar figures placed within different context, backgrounds, and interrupting lines. Target figures, on the other hand, were characterized as familiar figures without interrupting lines. This results of this study showed that H.M.'s ability to identify the 'hidden' and 'target' figures is also impaired. This finding was extremely surprising because past studies showed that his performance on perceptual tasks was superior to others (MacKay, Stewart, & Burke, 1998).

      Some have suggested that his inability to identify the hidden figures may be the result of short-term memory overload. However, as with deficits in language, others have proposed the 'binding problem' to explain H.M.'s difficulties in visual perception of hidden figures. According to this explanation, identifying familiar figures and forms does not require the formation of new connections between nodes. However, identifying and distinguishing the overlapping lines in the hidden figures would require the formation of new connections between nodes. Again, as with the formation of new connections between nodes in language, damage to the hippocampal system is believed to be responsible for H.M.'s inability to form new connections that would aide in accurate visual perception (MacKay, Stewart, & Burke, 1998).

      Studies have shown that other patients with damage to the left hemisphere of the brain, compared to the right hemisphere, experience a greater inability to identify ambiguous meanings. Other patients with damage to the left hippocampal system in particular have also shown comprehension deficits. This observation has led researchers to suggest that H.M.'s condition is also a result of damage to this area, rejecting the previous belief that the condition may have resulted from damage caused by his epileptic seizures (MacKay, Stewart, & Burke, 1998).

      Although there are still many unanswered questions, each day we are learning more about the relationship between language and memory. Through our understanding of H.M.'s condition, the fields of cognition and neuropsychology have made great advancements. Research in this area is extremely challenging and the ability to study patients with brain damage is one of our most valuable resources for solving the many mysteries of the brain.