Archives of Memory

A Soldier Recalls World War II

Alice M. Hoffman, Howard S. Hoffman

• Publisher: The University Press of Kentucky
• Year: 2021
• ISBN:
  9780813187426
• Topic: History
• Subtopic: Military & Maritime History
• Index: History

"Tell me about the war"—these words launched a ten-year project in oral history by a husband-and-wife team. Howard Hoffman fought in World War II from Cassino to the Elbe as a mortar crewman and a forward observer. His war experiences are of intrinsic interest to readers who seek a foot soldier's view of those historic events. But the principal purpose of this study was to explore the bounds of memory, to gauge its accuracy and its stability over time, and to determine the effects of various efforts to enhance it.

Alice Hoffman, a historian, initiated the study because she recognized the critical role of memory in gathering oral history; Howard Hoffman, the subject, is an experimental psychologist. Alice's tape-recorded interviews with her husband over a period of ten years are the basic material of the study, which compares the events as recounted in the first phase of the interviews with later accounts of the same experiences and with the written records of his company as well as the memories of fellow soldiers and the evidence of photographs and other documents.

This engrossing story of World War II breaks new ground for practitioners of oral history. The Hoffmans' findings indicate that a subset of human memory exists that is so permanent and resistant to change that it can properly be labeled "archival". In addition to describing some of the circumstances under which archival memories are formed, the Hoffmans describe the conditions that were found to influence their storage and retrieval.

1A Psychological Overview of Memory

Psychologists have been investigating memory for at least 100 years. It has usually been defined as the capacity or faculty of retaining and retrieving impressions from the past by means of recalling them or by recognizing them when some aspect of the impression is presented. This definition is generally accepted. A debate arises when we ask how the process works. Where and how are the impressions stored? Which impressions are stored—all of them or just some, and if just some, how are they selected for storage? What strategies are used for retrieval from memory, and what factors may impinge upon retrieval? Are these impressions susceptible to change and/or decay over time? The investigation of such issues has resulted in debate and the creation of opposing camps of theorists.

In fact, two of the earliest pioneers in the study of memory, Hermann Ebbinghaus (1885) and F.C. Bartlett (1932), took different approaches to the task of attempting to understand memory processes which have not as yet been brought into one organic conception. It is possible when studying most of the contemporary scientific literature on memory to assign a study either to the Ebbinghaus or to the Bartlett tradition.¹

It was primarily German professors who began to study the processes of the mind experimentally. Their own training was in areas where the laboratory method was well established, such as physics or physiology. Hermann Ebbinghaus, who had been so trained, observed that the experiments being conducted by most of his colleagues were restricted to the analysis of sense perception. He wanted to go a step further and use laboratory methods to look at “the workings of the mind and to submit to an experimental and quantifiable treatment the manifestation of memory.”²

He recognized the inherent difficulties in the task he had set himself. “How,” he asked, “can we keep constant the bewildering mass of causal conditions which, insofar as they are of mental nature, almost completely elude our control, and which, moreover, are subject to endless and incessant change?”³

Ebbinghaus attempted to provide appropriately controlled conditions by inventing the nonsense syllable, which consisted of a vowel surrounded by two consonants that did not form a three-letter word. He made up 2,300 such syllables and then proceeded to make lists of varying lengths and measure the amount of time and the number of trials which he needed before he could recite the list without error. He also attempted to determine the number of trials required to relearn a list after the passage of time had caused him to forget it. He was the principal subject of his own experiments, and while he realized that his tests would be primarily of individual significance, he hoped nevertheless that his experiments would provide a generally valid set of relationships. In conception and aim, then, our own study belongs to the Ebbinghaus tradition.

It must be observed that Ebbinghaus did exceedingly well—so much so that at a centennial symposium held at the twenty-sixth annual meeting of the Psychonomic Society in Boston in 1985 a group of memory theorists honored his contribution and agreed that his empirical techniques and conceptual framework are still a useful guide for research in the field. In the 100 years since Ebbinghaus made his studies, psychologists have been preoccupied with testing his propositions and with analyzing and extending their applicability to a wide variety of learning tasks. Ebbinghaus found, for instance, that attributed meanings cause the syllables to be learned more quickly. He discovered a savings effect, namely, that when material must be relearned, it can be mastered more quickly than when the first attempt to learn it was made. The implication is that, while the material is not available for recall, some memory trace must facilitate relearning the material.

Ebbinghaus found that learning is more efficient when sessions are distributed than when the same amount of time is spent in one session. And he plotted a forgetting curve which demonstrated that information is lost from memory rapidly at first but that the rate of loss levels out over time. This curve has proven valid for many types of learned material over a wide range of circumstances.

In 1932, Frederick C. Bartlett, professor of psychology at Cambridge University, published a book entitled Remembering. He was critical of the use of nonsense syllables and unrelated word lists to study memory. Bartlett believed that it is impossible to ensure that all experimental subjects begin the effort to learn and recall material unaffected by previous experience with the material to be remembered. He observed, as had Ebbinghaus, that the human subject makes innumerable associations in the effort to remember even nonsense syllables. But he carried his observation to the conclusion that the memory process is in and of itself inextricably bound up with the development of structural organization, or what he termed “schemata,” and he suggested that a study divorced from the history of the subject and from the natural tendency to seek meaning and association has a doubtful claim to be focusing upon memory at all.

He asked his subjects to remember pictures and stories. In one study he presented subjects with a series of five postcards, each with a picture of a man in one branch or another of the British armed forces. Since the study was being made during World War I, his subjects were quite familiar with the fighting services and were generally interested in them. Each of his twenty subjects was given ten seconds to examine each card and was then asked to describe the cards and to answer a series of questions about them, for example, “Which man has a mustache?” or “Which man is smoking a pipe?” After a period of two weeks or so, each subject was again asked to describe the cards and to answer questions about them. Bartlett tested for accuracy in the sequence in which the cards had been presented. He found that his subjects made frequent errors in this regard, although none was wrong as to which card had been presented first. This finding is of particular interest, for as will be seen there is a considerable primacy effect in Howard’s recollection of his World War II experience. That is, he recalled the first time he did or experienced something more vividly and more accurately than he recalled similar subsequent events.

Bartlett also found that his subjects could be divided into two groups: those who relied primarily on visual images and those who relied upon language cues in attempting to recall the cards. Those who relied upon the image were more sure of the correctness of their response, but in fact there was no difference in the number of errors made by the two groups. Bartlett remarked that, in strategies for recall, “seeing is believing.”⁴

Elizabeth Loftus in her work on memory reports the same phenomenon with respect to the confidence of visual imagers in their memories. Like Bartlett, however, she found that they were not more accurate in their ability to recall than those who processed their experiences verbally.⁵

In another set of experiments, Bartlett presented a story to his subjects that was derived from an Eskimo folktale. It concerned two young men who went to hunt seals. When his subjects repeated the story, Bartlett found that they showed a clear tendency to shorten it, to reduce it to its gist, and to reproduce that essence. He also found that subjects tried to make the story accommodate their own viewpoint and experience. Thus “something black came out of his mouth” became “foamed at the mouth” in many of the reproductions. And canoes frequently became boats. He further found that, once the story had been rationalized to fit constructions satisfying or understandable to the subject, subsequent reproductions remained stable and were subject to very little, if any, change.⁶

Since Bartlett’s publication of his findings, there have been relatively few studies of the accuracy of report over time. His contribution did not receive the attention that it deserved until quite recently. Most studies of memory have been concerned with very short retention periods. In many cases, in fact, they have involved retention periods of less than a minute.

There are a number of reasons for this emphasis. First, psychologists have suggested that there are two stages of memory, short-term memory and long-term memory.⁷ Short-term memory is information stored for only a few seconds, for example, a telephone number, which may be stored for no more than the time required to dial it.

Material retained for any appreciable amount of time beyond a minute belongs to the latter classification, and the basic processes involved are perhaps not effectively different whether one is examining retention for a day or a week or years. Therefore, the experimental analysis of the basic process involved is facilitated if shorter time intervals are used. Second, individual differences are smaller in memory over short periods. Information retained over long periods of time is affected by a host of potentially confounding variables, such as perceived importance, interest, comprehensibility, knowledge, personality, attitudes, temperament, prejudices, and so forth. Obviously at some point all these individual factors will have to be taken into account and controlled for, but experimental psychologists have deemed it wise to attempt to understand the underlying mechanisms first and to look at ways in which people or animals perform similarly with the expectation that such a strategy is more likely to uncover these basic mechanisms. Third, longitudinal studies are inherently difficult because over time the original group of subjects may die, become ill, or otherwise be unavailable to the researcher. Fourth, it is unfortunately true that longitudinal studies require a great deal of time between their conception and their completion, and in the academic environment of publish or perish few psychologists can afford a lengthy delay.

While studies on memory over long periods of time are relatively rare, some do exist. In the early twentieth century, there were a number of studies on the accuracy of report over time. These studies corroborate Bartlett’s observation that the reports show considerable inaccuracy and that the inaccuracies tend to be in the direction of bringing the report of the event into congruence with expectation or with accepted norms and standards. Moreover, a more recent study by H. Kay suggests that memory is very much influenced by its original encoding process or by the storage process itself. In other words, the assimilation of the material greatly affects its retrieval. Kay asked subjects to listen while he read two short passages. They were then asked to write a verbatim account of these passages. When they had done so he read a correct version to the subjects. A week later they were asked to reproduce the passages. This process was continued for seven weeks. In each case, subjects remembered their own reproductions far more accurately than the correct version even though at each session they were presented with the correct version.⁸

Obviously the fact that they had to write out their own version may have produced more powerful and persistent learning of their own version. It would be interesting to replicate the study with tape-recorded responses from the subjects rather than written responses to see what effect the change might have on the results. It is perhaps noteworthy, however, that, when Michael Howe replicated the study and obtained substantially the same results, he found that, according to subjects’ reports of their subjective experience, they knew they were reproducing their own version and not the correct version but that they were able to recall only their own version. It might also be interesting to see whether subjects could correctly identify their own version and the correct version when both were presented. In this way their recognition could be tested as well as their ability to recall the material.

Howe concluded that “when we speak of distortions in memory it would often be more accurate to speak of distortions in the material as it enters the retention system; once the material is stored, it is unlikely that much further distortion occurs.”⁹

Such findings have led many psychologists to conclude with D.B. Bromley that “short-term memory declines with age, whereas long-term memory is relatively unaffected except by disuse.”¹⁰ There is ample evidence that short-term memory declines with age; the ability to retain a telephone number long enough to dial the number correctly, for example, is a skill which deteriorates considerably with increasing years. The evidence that long-term memories are affected, however, is less clear-cut. It is, in fact, unclear whether memory is, or is not, a unitary process. There may or may not be a variety of kinds of memory. Furthermore, it is not clear exactly what memory is. Is it a chemical change in the brain? Is it the result of changes in neuronal pathways? Are memories laid down in a specific geographical location, or are they diffused in some way throughout the brain? These questions are currently under intense investigation. While some progress has been made, psychologists seem to agree that it is not yet possible to say with certainty where or how memories are stored in the brain.

Karl Lashley spent a lifetime looking for the specific location of memories. He systematically removed parts of the brain of rats that had learned a maze. No matter where he cut, however, memory survived, and finally, he gave up the search.

In 1951 Wilder Penfield, a Canadian brain surgeon, reported a study which generated considerable interest both in the scientific community and in the popular press. Brain surgery is usually performed under local anesthesia because the brain itself is not susceptible to pain. As a result the patient is conscious throughout the operation. Penfield reported that he could elicit specific memories in human patients who were undergoing brain surgery. If he touched a particular spot in the temporal lobes of the brain, patients would report a particular recollection. Moreover, it seemed to run off in time rather like a tape recording would.

Penfield’s work gave credence to the notion that the brain functions rather like a tape recorder and that not only specific memories can be evoked but also the associated perceptions, which have also been recorded. Moreover, Penfield believed that all experience is preserved and that any experience is available for replay if it can be appropriately evoked or retrieved.¹¹ There are however a number of difficulties with Penfield’s results. The most basic problem involves the difficulty of knowing whether the responses being evoked by the surgeon’s probe are in fact actual memories of real events or rather some kind of hallucination. Furthermore, no other neurosurgeon has reported similar findings.¹²

The suggestion that the memory is rather like a storehouse of audio and video tapes available for replay, however, has been an attractive one. It fits in with studies showing that memories can more easily be retrieved if some part of the experience can be brought to consciousness. For example, some studies show that, if a student can arrange to take a test in the same room in which he or she studied, the amount of material available for recall will increase substantially. In other words, the environment itself serves to cue or set in motion the retrieval process.¹³

In 1980 Duncan Godden and Alan Baddeley suggested that this environmental effect would serve to enhance recall but not recognition.¹⁴ They distinguished between intrinsic context and extrinsic context. They defined intrinsic context as those aspects of a stimulus which are inevitably processed when the stimulus is perceived, such as the voice in which an item is spoken or the typeface of a word on the page. Extrinsic contexts are irrelevant to the perceived item, such as the color of the walls in the room. They found that recall is enhanced by both intrinsic and extrinsic contexts but that recognition is enhanced only by intrinsic contexts. In other words, if you know that you will be given a multiple choice or true/false test, it would not enhance your performance to study in the room where you will take the test, but if you are to take an essay-type exam, then studying in the room where you will take the test may be advantageous.

Much of psychoanalysis is based on the assumption that all memory is retained somewhere in the recesses of the mind and that by application of categorical or contextual cues (that is, pieces of information or word associations), the patient can be led to retrieve memory that has been suppressed and has hence been unavailable for recall.

Sigmund Freud suggested that forgetting is due, in large part, to...

Table of contents

• Cover
• Half title
• Title
• Copyright
• Contents
• List of Maps
• Foreword by Forrest C. Pogue
• Foreword by Charles T. Morrissey
• Introduction
• Biographical Note on Howard Hoffman
• 1. A Psychological Overview of Memory
• 2. Stateside Experiences: First Recall Document
• 3. Stateside Experiences: Analysis of Memories
• 4. The Italian Campaign: First Recall Document
• 5. The Italian Campaign: Analysis of Memories
• 6. From Southern France to the Elbe: First Recall Document
• 7. From Southern France to the Elbe: Analysis of Memories
• 8. War’s End and Return to Civilian Life: First Recall Document
• 9. War’s End and Return to Civilian Life: Analysis of Memories
• Conclusion
• Appendix 1. The Italian Campaign: Second Recall Document
• Appendix 2. Daily Log, Company C, Third Chemical Mortar Battalion
• Appendix 3. Combat Units Supported by Hoffman’s Company
• Notes
• Acknowledgments
• Index