Implications of Short-Term Memory for a General Theory of Memory 1
Total Page:16
File Type:pdf, Size:1020Kb
3"OURNAL OF VERBAL LEARNING AND VERBAL BEHAVIOR 2, 1-21 (1963) ADDRESS OF CHAIRMAN OF SECTION I (Psychology) Implications of Short-Term Memory for a General Theory of Memory 1 ARTHUR W. MELTON University o/Michigan, Ann Arbor, Michigan Memory has never enjoyed even a small The confluence of forces responsible for fraction of the interdisciplinary interest that this sanguine prediction about future progress has been expressed in symposia, discoveries, is reflected in this AAAS program on memory and methodological innovations during the (see other articles in this issue of this last five years. Therefore, it seems probable JOURNAL). Advances in biochemistry and that the next ten years will see major, perhaps neurophysiology are permitting the formula- even definitive, advances in our understanding tion and testing of meaningful theories about of the biochemistry, neurophysiology, and the palpable stuff that is the correlate of the psychology of memory, especially if these memory trace as an hypothetical construct disciplines communicate with one another (Deutsch, 1962; Gerard, 1963; Thomas, and seek a unified theory. My thesis is, of 1962). In this work there is heavy emphasis course, that psychological studies of human on the storage mechanism and its properties, short-term memory, and particularly the especially the consolidation process, and it further exploitation of new techniques for may be expected that findings here will investigating human short-term memory, will offer important guide lines for the refine- play an important role in these advances ment of the psychologist's construct once we toward a general theory of memory. Even are clear as to what our human performance now, some critical issues are being sharpened data say it should be. by such observations. Within psychology several developments have focused attention on memory. In the 1 This paper comprises, in substance, the author's first place, among learning theorists there is a Vice-Presidential Address to Section I (Psychology) of the American Association for the Advancement of revival of interest in the appropriate assump- Science, 1962. The author is particularly indebted tions to be made about the characteristics of to the Center /or Human Learning, University of the memory traces (engrams, associations, California, Berkeley, where a research appointment bonds, sHr's ) that are the products of ex- during the Fall semester of 1962-1963 gave the periences and repetitions of experiences. Thus, freedom from academic routine and the stimulating discussions that led to the repetition of the Hebb Estes (1960) has questioned the validity of experiment and also supported the preparation of the widespread assumption (e.g., Hull, 1943; this paper. Early exploratory studies on short-term Spence, 1955) that habit strength grows in- memory and the experiment on the recall of different crementally over repetitions, and has proposed sized verbal units were supported by Project MICHI- an all-or-none conception as an alternative. GAN under Department of the Army Contract DA- 36-039-SC-78801, administered by the United States More recently, he has examined (Estes, 1962) Army Signal Corps. Reproduction for any purpose of in detail the varieties of the incremental and the United States Government is permitted. all-or-none conceptions and the evidence 2 MELTON related to them. Already, some defenders of the spate of theorizing and research on the incremental concept (Jones, 1962; Keppel immediate and short-term memory during the and Underwood, 1962; Postman, 1963) have last five years. In 1958, and increasingly taken issue with Estes' conclusions, and it thereafter, the principal journals of human would appear that this fundamental question learning and performance have been flooded about memory will loom large in theory and with experimental investigations of human experiments for some time to come. At a short-term memory. This work has been somewhat different level, the revival of ex- characterized by strong theoretical interests, perimental and theoretical interest in the and sometimes strong statements, about the notion of perseveration or consolidation of nature of memory, the characteristics of the the memory trace (Glickman, 1961), and memory trace, and the relations between attempts to embody it in a general theory of short-term memory and the memory that learning (Hebb, 1949; Walker, 1958), have results from multiple repetitions. The contrast also focused attention on a theory of memory with the preceding thirty years is stril~ing. as a fundamental component of a theory of During those years most research on short- ~earning. term memory was concerned with the memory A second strong stimulus to research on span as a capacity variable, and no more. It memory from within psychology are several is always dangerous to be an historian about findings of the last few years that have forced the last five or ten years, but I venture to major revisions in the interference theory say that Broadbent's Perception and Com- of forgetting and consequently a renaissance munication (1958), with its emphasis on of interest in it (Postman, 1961). First, there short-term memory as a major factor in human was the discovery by Underwood (1957) information-processing performance, played a that proactive inhibition had been grossly key role in this development. Fortunately, underestimated as a source of interference in many of the others who have made important forgetting. Then, the unlearning factor as a methodological and substantive contributions component of retroactive inhibition was given to this analysis of short-term memory have greater credibility by the findings of Barnes presented their most recent findings and and Underwood (1959). And finally, the thoughts in these Meetings on Memory, and joint consideration of the habit structure of they thus adequately document my assess- the individual prior to a new learning ex- ment of the vigor and importance of this perience, the compatibility or incompatibility recent development. Therefore I will refrain of the new learning with that structure, and from further documentation and analysis at the unlearning factor (among others) led to this point, since the impact of some of these the formulation of the interference theory of findings on our theory of memory is my forgetting in terms that made it applicable to main theme. all new learning (Melton, 1961; Postman, 1961; Underwood and Postman, 1960). Thus, THE DOMAIN OF A THEORY OF MEMORY this development focuses attention on the inter- A theory of memory is becoming important actions of memory traces during learning as for a number of different reasons, and some- well as their interactions at the time of how all of these reasons properly belong attempted retrieval or utilization in recogni- to a comprehensive theory of memory. Its tion, recall, or transfer. storage mechanism is the principal concern But perhaps the most vigorous force of biochemists and neurophysiologists; the directing attention within psychology to the morphology of its storage--whether as a need for a general theory of memory is multiplexed trace system with one trace per IMPLICATIONS OF SHORT-TERM MEMORY 3 repetition, or a single trace system subjected tive learning situation and any immediately to incremental changes in "strength" by subsequent trial. By convention among repetition--is becoming a principal concern psychologists, the change from Trial n to of learning theorists; its susceptibility to in- Trial n ~- 1 is referred to as a learning change hibition, interference, or confusion both at when the variable of interest is the ordinal the time of new trace formation and at the number of Trial n and not the temporal in- time of attempted trace retrieval or utilization terval between Trial n and Trial n ~ I, is the concern of forgetting and transfer and the change from Trial n to Trial n ~ 1 theorists; and the perhaps unique properties is referred to as a retention change when the of its manifestation in immediate and short- variable of interest is the interval, and the term retention is the principal concern of events during the interval, between Trial n psychologists interested in human informa- and Trial n ~ I. Learning and retention tion-processing performance. One knows in- observations generally imply that the char- tuitively that all of these different approaches acteristics of the task, situation, or to-be- emphasize valid questions or issues that must formed associations remain the same from be encompassed by a general theory of Trial n to Trial n ~ 1. When any of these memory, but nowhere--with perhaps the task or situation variables are deliberately exception of Gomulicki's (1953) historical- manipulated as independent variables between theoretical monograph on memory-trace theory Trial n and Trial n-Jr 1, the object of in- --will one find explicit systematic considera- vestigation is transfer of learning, i.e., the tion of these several different facets of the availability and utilization of the memorial problem of memory. products of Trial n in a "different" situation. Since my present, intention is to marshal Now, these operational definitions of learn- some data relevant to one of the main issues ing, retention, and transfer are completely in a general theory of memory--namely, the aseptic with respect to theory, and I think it question of whether single-repetition, short- is important to keep them so. In part, this term memory and multiple-repetition, long- is because it is useful to keep in mind the term memory are a dichotomy or points on fact that learning is never observed directly; a continuum--I feel compelled to discuss it is always an inference from an observed briefly what I believe to be the proper domain change in performance from Trial n to Trial of a theory of memory and to differentiate it n ~ 1. Furthermore--and this is the im- from a theory of learning. portant point for theory--the observed change After some exclusions that need not concern in performance is always a confounded re- us here, learning may be defined as the modi- flection of three theoretically separable events: fication of behavior as a fanction of ex- (i) the events on Trial n that result in some- perience.