| 1 | Amnesia and Memory Research |
Daniel L. Schacter and Endel Tulving
University of Toronto
Recent years have witnessed an explosion of research in the analysis of memory and amnesia. For the student of normal memory, this has meant an endless stream of experiments and theory exploring countless phenomena of memory and the appearance of several journals devoted almost exclusively to this research. Although not quite so large in comparison, the ever-increasing flow of articles concerning human amnesia that have appeared in journals such as Cortex and Neuropsychologia over the past 15 years testifies to the growing vitality of this important sector of psychological research.
Unfortunately, and despite the increasing vigor of their respective fields, students of normal and abnormal memory have not enjoyed active communication with one another; consequently, the developments in the two fields over the past 15 or 20 years have proceeded largely in parallel. In the experimental psychology of intact memory, perhaps the sole point of meaningful contact with amnesia research has taken the form of the inclusion of clinical facts about the amnesic syndrome in theoretical discussions of the distinction between short-term and long-term memory (Atkinson & Shiffrin, 1968; Wickelgren, 1973, 1979). Beyond that, however, it would be difficult to find any systematic account of amnesic phenomena in the numerous recent theories of normal human memory, and utilization of amnesic data by mainstream experimentalists in attacking problems of normal memory is likewise rare. Indeed, several leading researchers have contended that amnesic phenomena currently lie outside the desirable boundaries of memory research. Murdock (1974), for instance, has cautioned that: âIt is difficult enough understanding the memory of normal college students; it will be time to consider the abnormal cases after we can cope with the normal cases [p. 5].â experimental psychologists, warning that: âextrapolations from pathological deficits to the structure of normal memory are of uncertain validity [p. 308].â
There are, however, some positive signs of change. For instance, one can point to the recent appearance of papers concerning various amnesic phenomena in âhard-coreâ memory journals: Several studies of Korsakoff patients have appeared in the Journal of Experimental Psychology (Cermak & Reale, 1978; Oscar-Berman, 1976), experiments concerning drug-induced amnesia have been published in the Journal of Verbal Learning and Verbal Behavior (Birnbaum, Parker, Hartley, & Noble, 1978; Eich, Weingartner, Stillman, & Gillen, 1975; Hartley, Birnbaum, & Parker, 1978), and a recent issue of Memory & Cognition contained an experimental study of Korsakoff amnesics (McDowell, 1979). Furthermore, the appearance of theories such as the one recently put forward by Wickelgren (1979) constitute a major advance in pulling together the insights of experimental psychology and the observations on amnesia.
It is only fair to point out that students of amnesia have harbored their own doubts concerning the usefulness of experimental psychology in the analysis of memory and its pathology. As long ago as 1901, Pierre Janet, one of the centuryâs most innovative clinical psychiatrists, sarcastically depicted psychologyâs role in the study of memory as follows: âThe psychologists in their descriptions admit of no other elementary phenomena of memory than conservation and reproduction. We think that they are wrong, and that disease decomposes and analyzes memory better than psychology [p. 102].â Despite similar misgivings expressed by other students of pathological memory from time to time, signs of hope are evident here too, as a number of contemporary amnesia researchers have sought to establish an alliance with experimental psychology. Expressed largely in the work of Warrington and Weiskrantz in England, and Cermak and his colleagues in Boston, the methods, findings, and theories of experimental psychology are beginning to find a home in the analysis of amnesia. Studies of amnesia that utilize the insights of experimental psychology have been undertaken by an increasingly diverse group of researchers in the past few years, and there is every reason to believe that this trend will continue in the future.
In short, there is reason to believe that we may be on the verge of a âgolden ageâ in which the interaction between the experimental psychology of normal memory and the investigation of amnesic deficits will be more thorough and meaningful than it has been in the past. Texts like the present one will hasten the arrival of this golden age, inasmuch as one of its important functions is to explore ways to remove the remaining barriers and problems preventing the integration and interaction of the two fields.
The purpose of this chapter is to confront some of these key issues and clarify problems surrounding them. The chapter is offered in a spirit of constructive criticism from two experimental psychologists who firmly believe that both the experimental psychology of memory and the analysis of amnesic deficits stand to benefit from increased understanding and utilization of each otherâs methods, data, and theories.
The chapter is structured as follows. First, several historical features of the relations between experimental psychology and amnesia research are considered, together with a brief analysis of the losses suffered by each because of its ignorance of the other. Second, some contemporary problems are outlined that have to be confronted before a further integration of the two fields of research can be effected. These problems include subject variability in studies of amnesia, logic of experimental inference, pitfalls in the application of experimental paradigms to amnesic populations, conceptual confusion in experimental study of memory, and difficulties encountered in the laboratory production of analogues of amnesia. In the third and final section of the chapter, a general framework for the study of memory and amnesia is briefly outlined.
HISTORICAL RELATIONS BETWEEN AMNESIA AND MEMORY RESEARCH
It may be indicative of the late 19th-century psychological Zeitgeist that Théodule Ribot published his classic treatise on amnesia, Diseases of Memory (1882), just three years before Hermann Ebbinghaus thrust his epoch-making experimental study of memory upon the scientific world in 1885. And it is interesting to observe that Ribot foresaw an emerging era in which phenomena of amnesia would be at the center of psychological attention:
The disorders and maladies of this faculty [memory], when classified and properly understood, are no longer to be regarded as a collection of amusing anecdotes of only passing interest. They will be found to be regulated by certain laws which constitute the very basis of memory, and from which its mechanism is easily laid bare [p. 10].
Alas, the central role that Ribot envisaged for amnesic phenomena in the new psychological science of memory never took shape: The lines of research initiated by his own work and the studies that followed in the tradition of Ebbinghaus occupied virtually nonoverlapping spheres in subsequent years.
The separation between amnesia and memory research becomes clear to anyone who studies the developments in the two fields in the late 19th and early 20th centuries. Investigations of intact memory in the experimental tradition initiated by Ebbinghaus were pursued largely in two countries, America and Germany. The experimental psychology of memory flourished in the 30 or so years following the publication of Ebbinghausâ book. Under the leadership of eminent researchers such as Georg MĂŒller, Ernst Meumann, and Mary Calkins, the young science ambitiously tackled numerous phenomena of memory (see Murray, 1976). But the proliferation of methods, data, and theory that characterized the experimental literature did not include a systematic interest in amnesic phenomena: The mainstream experimental journals were conspicuously free of articles dealing with memory pathology. There were exceptions, but these were exceedingly rare. For instance, in the first volume of the Psychological Review, we find a case study of amnesia resulting from gas poisoning reported by Dana (1894). In what was perhaps the most serious attempt by a mainstream experimentalist to grapple with the problems of amnesia, Burnham (1903) drew on available evidence on processes of organization and perseveration (MĂŒller & Pilzecker, 1900; MĂŒller & Schumann, 1893), to support his thesis that retrograde amnesia is attributable to what we now would call an encoding deficit: âThe memory is lost because it was never completely organized [p. 132].â And several years later, Hennig (1910) reported another pathological case study in the Zeitschrift fĂŒr Psychologie. These isolated instances, however, did not succeed in generating any sustained interest in amnesia on the part of experimental psychologists.
A reciprocal lack of interest in experimental psychology can be discerned when one examines the articles appearing in clinical journals that published research on amnesia. These papers were for the most part comprised of extensive case studies or descriptive analyses of small pathological populations (Coriat, 1907; Freund, 1889; Gregor & Roemer, 1907; Konrad, 1907; Korsakoff, 1889/ 1955; Kutner, 1906; Moll, 1915; Paul, 1889). The authors of these studies either did not know of, or did not care about, the experimental facts generated by mainstream memory researchers; the names of Ebbinghaus, MĂŒller, and Meumann that dominated the experimental literature are not to be found in most clinical reports. Again, the exceptions that can be uncovered were few. Krauss (1904) utilized recently developed tachistoscopic techniques for evaluating memory span in his case study of an amnesic patient. Buckley (1912), citing the experimental work of Kirkpatrick (1894) on modality effects, commented that data obtained from his patient contradicted the previous findings that visual memory exceeds auditory memory: âWords written for him were not remembered as easily as when they were spelled for him audibly [p. 436].â And in a rather thorough analysis, Pick (1915) applied theories of thought processes derived from the tenets of the WĂŒrzburg school to his analysis of Korsakoff patients and concluded that these patients exhibited a deficit in goal-directed thinking that could be comfortably explained by the WĂŒrzburg theory.
Before we consider some of the consequences that followed from the mutual isolation of amnesia and memory research, we should point out that a brief period of reciprocal interest did seem to take shape starting around 1930. For instance, we find three studies of amnesia appearing within a year of each other in the experimental journal Archiv fĂŒr die Gesamte Psychologie. Ranschburg (1930), a well-known experimental psychologist, applied mainstream theory to disorders of attention and memory; Krauss (1930) attempted to conceptualize amnesia in terms of Kurt Lewinâs recently formulated dynamic theories; and Störring (1931) informed experimental psychologists of a dramatic case of pure amnesia that in several respects resembles the famous modern case of H.M. Similarly, three papers on amnesia appeared in close temporal proximity to each other in various American psychology journals. Both Ray (1937) and Sears (1936) were impressed by the apparent similarities between clinically documented retrograde amnesia and the laboratory phenomenon of retroactive inhibition. Sears cautiously concluded that the resemblances between the two phenomena were âsuggestive [p. 237],â whereas Ray argued more boldly that: âthese phenomena are manifestations of the same underlying mechanism [p. 341].â And Conkey (1938) used a variety of newly developed psychological tests to explore memory and other intellectual deficits following closed-head injury.
The interest in amnesia demonstrated by the aforementioned psychologists in the 1930s was matched by a new willingness on the part of clinicians to utilize conceptual tools provided by experimental psychology. BĂŒrger (1927), for instance, offered an interpretation of memory pathology cast in terms of the newly popular gestalt theory. Syz (1937) provided an interesting discussion of concussion amnesia that drew heavily on experimental work on interference and on what appeared to be a new and iconoclastic theory of memory proposed by Bartlett (1932). Gillespie (1937) attempted an even more detailed account of amnesia within the framework of Bartlettâs theory and explored the possibility that defective operation of schemata and voluntary recall contributed to amnesic pathology.
The manifestation of mutual interest between the two groups of students of memory in the last 10 years or so before the second World War was not exactly overwhelming, and the reasons for the apparent beginning of a rapprochement are not clearâneither are the reasons for its apparently abrupt termination.1 But it does seem appropriate to suggest that the seeds of several intriguing research possibilities were sown during this period.
For the most part, as we have seen, there existed a gulf that separated the experimental psychology of memory and the clinical study of amnesia in the years following the contributions of Ebbinghaus and Ribot. What were its consequences? We believe that both sides missed out on potentially key insights from the other. On the clinical side, the parade of case studies that dominated the journals remained free of standardized experimental procedures and explored only a few restricted areas of memory function in amnesic patients. It is precisely in these areas that the experimental psychology of memory could have aided students of amnesia most: Mainstream memory researchers had developed an arsenal of research techniques and generated a vast body of data that could have increased substantially the scope of amnesia research and enriched its analytical precision. Early students of amnesia paid scant attention to these possibilities, and their research remained correspondingly narrow.
But the experimentalists also paid a price for ignoring the data generated by clinical observations of amnesic deficit. As we have argued elsewhere (Schacter, Eich, & Tulving, 1978), amidst the impressive amount of experimental evidence reported by post-Ebbinghausian researchers, one important area of memory function remained virtually unexamined: the process of retrieval and the psychological conditions that affect its success or failure. Memory research at the turn of the century was powerfully influenced by the classical doctrine of associationism, which regarded one variable as particularly crucial to the outcome of memory performance: the strength o...