William S. Verplanck
Sometimes, problems need to be restated. Here is one attempt.
1. There was an incident.
2. There were individuals who took part in it.
3. The incident is over and done with; records made at the time are incomplete. The incident can no longer be examined or reexamined, nor “rerun,” yet present and future actions depend on exactly what happened, so far as we can tell. A decision must be made now. What we do now will be determined by what was done, said, and seen then, and we must have an acceptable statement of that now, to choose what to do next.
4. The incident must be “reconstructed,” so that it can be described now, as it might have been described as it was occurring, by an all-observant observer-recorder. What was said then must be reinstated now, as it might be, had it been taped.
5. “Reconstruction” must be done using, on the one hand, such physical records as may have been made at the time (the “physical evidence”) and on the other, the behavioral records, the recall of persons who participated in the event.
If the events that must be reconstructed had been planned, that is, if their occurrence had been predicted, people would likely have prepared themselves, by choosing and using a method of recording, as by taking movies or tapes of baby’s first birthday party, or of the bride as she looks throwing the bouquet, or of the swearing-in. Indeed planned events most often require the development of an “official” version, which may or may not correspond with the events as they take place. Secretaries take minutes, court recorders produce verified court records, and heads of state sign red-taped treaties. But the eventual reconstruction of events with such official “legal” records will nevertheless depend on the recall of participants. History, biography, autobiography, and journalism abound with instances of recall of “the real story” of such-and-such. Behavioral reconstructions of the event often serve to control further events, despite the written or other official record, which can be, and sometimes are, thoughtfully falsified by omissions, by additions, by ordinal inversion. Written records can be challenged, the tape can be erased, or cut and spliced.
Regrettably, psychologists have left the investigation of these phenomena to novelists, playwrights, diplomats, journalists and the like and ignored their implications.
6. Until time-travel becomes feasible, and there are perfect observers and recorders, there will be no means of establishing the “truth,” the veridicality of a reconstructed event. In day-to-day living, as in science, veridicality will be judged as is any theoretical account: by the degree to which the account achieves agreement or at least acquiescence among observers. In our legal system, two or more such accounts are presented, and a jury must choose among them.
7. Given this it is ironic that the models for such reconstruction-by-recall that psychologists employ in their theorizing are drawn not from the facts of how we behave, but from the various and sundry technological methods that cultures have developed for making, and reserving, a variety and multitude of physical records, which do not share the properties of recall.
Psychological theory proposes that we make some kind of record, correct or not, in our heads, our neurons, or short or long-term storage, as the events occur, and that this record is preserved, essentially intact through time, except insofar as it progressively deteriorates, like so many canned goods on the grocer’s shelf, or as it may be tampered with. Then, at the time of recall, these theories propose that the question is primarily one of finding the records in the files, or library, and of taking them out for inspection and use. Theories differ on what gets stored, words, phrases, pictures, or books–on how they are catalogued (“structure”), and on whether they may be stored, rearranged, or recataloged, even edited while in the files and as they are retrieved, zipped out to the circulation desk (“organization”).
The implausibility of such accounts is perhaps best attested by the fact that civilization has progressed hand in hand with the development of methods for preserving, transmitting, and retrieving physical and physically reproducible records of events–from pictures and sculptures through writing, printing, libraries, museums, films, tapes, and so on. People found they could not depend on recall; they developed ways to make and to save records. Once these records could be made, stored, and consulted at later dates, complex and extensive societies could emerge. Records do not work like “memories.” Memories are not stored; records are stored. Records are perhaps the least plausible of metaphors for understanding remembering.
8. This seems to be the pattern for psychological theorizing. Our theorists use as models for behavior just those things, processes, and concepts that people have developed to do the things that people cannot themselves do. When we have a machine (in the broad sense) that will do these things very effectively indeed, then we say, that’s how we (our “minds,” our “brains,” our “spirits,” our “memories”) work. It should be unnecessary to point out that the wheel and the sail do not serve as useful models for walking and running or for swimming. Even less are their principles useful for predicting when and where we will go, although they may determine how we get there.
As we develop new technologies, we flatter ourselves for the wrong reason. We flatter ourselves because we think we work like that new gadget–not because we have developed something that words better than we do, better because it works differently, on different principles.
Our machines do not work like us, nor we like our machines, except insofar as machines have been developed to work with us, to interface with us, and because we choose what they are to do.
A parallel in nonsense is given by “cognitive psychologists” who study human thinking as if we were all preliterate. They not merely ignore, they exclude from the study of thinking the papers and pencils, the formulas, maps, pictures, diagrams, and notes that are integral to “higher thought processes.”
9. We must remember that remembering occurs at a given locus in time and space. The word refers to a kind of behaving, and interacting of the individual with what is around at the time that person remembers–with the environment including other individuals who are there, or who will read what we write down, as well as with those inferred past events. Remembering will always show itself (as behaviors do) as dependent upon (“a function of”) that environment as well as on the individual (with a unique reactional biography) who is recalling. Remembering will also depend on not only one previous behavior-segment–the person’s behavior participating in the event being reconstructed, but aggregates of behavior-segments in that biography, especially those relating in any way to the incident and to those that have occurred before any occasion of recall. Until instances of remembering, of sharing the past with those present, are investigated as such, we will not learn much about their processes. If we insist on bridging the months and years, with a “record,” and remain bound by Newtonian metaphysics, we will also remain limited to trivialities, designed for explaining away basic characteristics of remembering as “distortions” that occur at the time of recall.
10. Forgetting? We had better remember that forgetting is a word that can be used only when we are talking about those very specific things we taught the individual to do, or observed that individual doing and that we looked for again after a time. To use the term we must have a full set of data, taken at the first instance, as Ebbinghaus painstakingly tried to do when he set the pattern for “scientific” thinking and research on remembering. This pattern has limited research on remembering: That is, the complement of what is forgotten in the set of what we know was learned, what we observed in the first instance. We can’t study remembering by studying forgetting. We don’t have the full deck of cards.
The pattern demands “rigorous experimental controls,” over both first instances, and the later occasion for control. It applies barely at all to what we do every day.
These controls obscure the importance of events excluded by the arbitrary presumptions of the theorist-experimenters. The method asks the participant to recall in one or another way the experimenter’s myopic version of the original event– the pictures, stories, number, little scenes, paired-associates, or nonsense syllables–but not that the participant thought the whole thing was stupid, or had to go to the bathroom, or was intensely bored by the whole thing, but needed the points to pass the course.
The notion of “experimental control” as we have explained it ensures that setting factors (and the behavior-sets they govern) are ignored, controlled out of existence and hence to be neglected in the theory. They should be center stage. We have been obsessed by “intentional learning.”
11. A non-problem: What about lying? May not the participant lie when recalling? How are we to know that a subject is not “lying” when recalling so-and-so, but all other evidence leads us to believe–makes it certain–that that individual was not present?
Perhaps there was no lying. The language recognizes what psychologists do not: a social behavior, a matter of people dealing with other people; that we deceive ourselves, as well as others; that we may lie to ourselves, just as we may love ourselves, remind ourselves, and so on. We interact with ourselves no differently than we do with others. We recall events that did not happen. We misremember.
But that is another story, to be dealt with elsewhere.
12. Only an interbehavioral, an operational, a naturalistic approach can help us climb out of the pitfall we have dug for ourselves by assuming that some unique and identifiable thing must persist, be stored, through time–an engram, a trace, a hologram, or whatever.
13. Such an approach would note that “time” as it shows itself in remembering shares few properties with “physical time.” Attempts to develop theories of remembering that incorporate the physicist’s concept of covert Cartesianism. We can use physical time in identifying landmarks, and quantifying the distances between them, but no more.
14. In sum, what we observe, what happens as we recall, tells us that remembering is not the retrieval of stored records. It is a social interaction, controlled in some part by past events as recorded.
And what is accepted as an accurate account of a past event must be recognized for what it is, a set of inferences from the social behaviors of the individuals who took part in it, and from the records.
Since the drafting of this paper, a relevant publication has been brought to the writer’s attention: Roediger, H.L., III. Memory metaphors. Cognitive Psychology, 1980, 8, 231-246. This scholarly review presents in diverting detail the full array of the metaphors, some of which we have alluded to. It not merely explores all the blind alleys, it even identifies the central goal box, the “ultimate metaphor for the human mind.” This will be recognizable, Roediger tells us, as an “unobservable construct…supported by converging operations,”–the unseeable top of a sort of Eiffel Tower, unseeable because, one must presume, the tower disappears up into a haze of truth. The author does not seem to be concerned with why he (and the “Science of Psychology”) should choose to get to such a goal by so convoluted a path, by so many false entries, reversals, backtrackings. He accepts the maze and the goal as givens, and never wonders whether one needed to enter it in the first place.
This paper, in contrast, points out not solely that all these metaphors for memory should have been recognized as the blind alleys they are but also that the whole experimental-theoretical construction, while suitable for diversion, is one that precludes our liking outside it, outside the enchanted maze. Outside, there are people remembering some things, misremembering and forgetting others, noticing, ignoring, rejecting.
The only metaphors that will work will be other behaviors, other interactions of the individual and the things, events, and people around.