If I had to name the three historically most influential publications in the psychological study of memory, I would pick Ebbinghaus's “On Memory” in 1885, Bartlett's Remembering in 1932, and Miller's “Magical Number Seven” in 1956. Miller's paper inaugurated the modern period and the information-processing approach, whereas Ebbinghaus and Bartlett provided the foundations upon which the modern work has been built. How different they were!

Indeed, one can argue that Ebbinghaus and Bartlett are considered our founding fathers precisely because they were so different, each contributing to the study of memory his own distinct perspective. What made the differences in their perspectives important was that they occurred against a shared background of scientific values. Both were empiricists who put their faith in the study of memory through observation and experiment, rather than armchair analysis and speculation, as was the custom before their times. Where they differed was in the types of experiments they conducted and the kinds of observations they valued. Ebbinghaus focused on memory as an object of study; Bartlett was interested in the act of remembering. Although their approaches were contradictory, each made valuable and lasting contributions–a fact that attests to the enormous complexity of memory. There is more to memory than just remembering. Personal experience and general knowledge, reproduction and reconstruction are all aspects of “memory” but require their own methods of study and analysis.

There is no need to dwell on Ebbinghaus's contribution, except in contrast to Bartlett's.

On a brilliant afternoon in May 1913 the present Laboratory of Experimental Psychology in the University of Cambridge was formally opened. Dr C. S. Myers, the founder and, at that time, the Director of the Laboratory, invited me to take part in the Demonstrations which he had arranged. Accordingly, for several hours, I sat in a darkened room exposing geometrical forms, pictures and various optical illusions to the brief examination of a long string of visitors. That was the beginning of the present book. The interpretations by my observers of the figures which were placed before them were as various as they were attractive. It was clear that the course of normal visual perception may be determined by a very large number of different factors. It seemed probable that a carefully arranged experimental study might serve to disentangle many of these factors and to demonstrate their functions.

Encouraged by others, I set to work at once. But it very speedily became evident that an examination of normal perceptual process leads directly and inevitably to an investigation of related mental processes, and in particular to a study of imagery and of recall.

Long before this, Ebbinghaus had introduced the ‘exact methods’ of nonsense syllables into the Laboratory consideration of memory. As in duty bound, I followed his lead and worked for some time with nonsense material. The result was disappointment and a growing dissatisfaction.

INTRODUCTION

All of the methods so far described deal with factors influencing individual observers. They help to show what occurs when a person makes use of some new material which he meets, assimilating it and later reproducing it in his own characteristic manner. Already it is clear, however, that several of the factors influencing the individual observer are social in origin and character. For example, many of the transformations which took place as a result of the repeated reproduction of prose passages were directly due to the influence of social conventions and beliefs current in the group to which the individual subject belonged. In the actual remembering of daily life the importance of these social factors is greatly intensified. The form which a rumour, or a story, or a decorative design, finally assumes within a given social group is the work of many different successive social reactions. Elements of culture, or cultural complexes, pass from person to person within a group, or from group to group, and, eventually reaching a thoroughly conventionalised form, may take an established place in the general mass of culture possessed by a specific group. Whether we deal with an institution, a mode of conduct, a story, or an art-form, the conventionalised product varies from group to group, so that it may come to be the very characteristic we use when we wish most sharply to differentiate one social group from another. In this way, cultural characters which have a common origin may come to have apparently the most diverse forms.

The experiments now to be described were designed to study the effects of the combination of changes brought about by many different individuals.

SOME CHARACTERISTICS OF IMAGES

If the theory of remembering which has just been developed is justified, it is clear that images have fundamentally important parts to play in mental life. This hardly appears to be a popular view in modern psychology. It is not only the extreme behaviourist who is tempted to think that psychology can gain little advantage from a close study of images. In less radical circles also it is often held that, since images are generally vague, fleeting, and variable from person to person and from time to time, hardly any statement can be made about them which will not be immediately contradicted on very good authority. This view has arisen because most statements that have been made about images in traditional psychology concern their nature rather than their functions, what they are rather than what they make it possible to do; and many of the controversies that have raged about them have been concerned primarily with their epistemological status. Undoubtedly considerations about the nature of imagery, and its relation, on the one hand to a world of external objects, and on the other to immediate sensory perception, raise important, interesting and singularly intractable problems. It is not these with which we are now concerned, but simply with the question of what, exactly, images have to do in the general growth of mental life.

A DEFINITION

In its modern aspects, social psychology had a brilliant beginning when Prof. William McDougall wrote his Introduction to Social Psychology, but since then it seems to have made very little advance. In one sense, McDougall's book, in spite of its outstanding originality and value, has had a bad influence. As everybody now knows, he based the whole of his treatment upon the view that instinctive activities are of primary importance in all social life. Every possible social reaction is considered to be ultimately explicable in terms of instinctive tendencies. This was a controversial subject when McDougall approached it, and it is so still. His treatment has tended to make psychological discussions of social problems a field of battle between instinct psychologies and psychologies of every other kind. Also his tentative catalogue of instincts has led to a large number of other catalogues of varying length and content, as if the mere listing of instincts itself sheds a great light upon social responses. In the hurly-burly of dispute a great amount has been said about instincts, but very much less about social conduct in any genuinely observed sense, except by anthropologists whose interests are apt to stop short at detailed description.

If any advance is to be made, the first thing that is necessary is a clear realisation by the social psychologist of the precise ‘lie’ of his problems. This may initially seem to be a fairly simple matter. We can define social psychology as the systematic study of the modifications of individual experience and response due directly to membership of a group.

INTRODUCTORY REMARKS

In several respects the notion of a collective unconscious is the counterpart, in social psychology, of the notion of memory traces in individual or general psychology. Both appear to assume that psychological material—images, symbols, ideas, formulae—are somehow individually preserved and stored up for use, either in the central nervous system of the individual, or somewhere in a persistent psychical structure which is the possession of a social group. The whole notion of collective unconscious is excessively difficult to discuss with any clarity, for it has been developed in an extremely confused fashion. Within its scope all sorts of psychological ideas, very different in their essential character, have been jumbled together.

Before I attempt discussion, it may be well to restate briefly the position which I have already reached.

When a number of people are organised into a social group, whether by appetite, instinct, fashion, interest, sentiment, or ideal, this group speedily develops certain characteristics peculiar to itself, which directly constrain the behaviour of its individual members. I have throughout treated these characteristics as the expression of active tendencies, for we have to consider them, not merely descriptively, as they are expressed in institutions, symbols, catch words, codes, and material culture, but also causally, as actual determining conditions of conduct and experience.

INTRODUCTION

When I was planning this investigation, in the early part of 1914, a question which was uppermost in my mind was the way in which conventional modes of representation and behaviour are developed within the social group and transmitted from group to group. Such conventionalisations are produced by a combination of innumerable small changes introduced by a large number of individuals; and it is not fantastic to suppose that there may be a parallel between them and the development, in the course of individual recall, of relatively fixed and stereotyped modes of representation or of reaction. There is no doubt that much human remembering is influenced directly and strongly by factors which are social in origin. The influence of these factors may be obscured by the ordinary laboratory methods of the study of memory, because of the exceedingly artificial character of the material which is used. But it seemed possible that suitably devised experiments could help to show how, by the accretion of many small changes, the individual human subject achieves reactions and modes of representation which are for him genuinely conventional. Perhaps, also, some light might be thrown upon the ways in which such changes, as they are passed about from person to person in a community, gradually come to develop conventionalisations which become current within a given social group. The way would thus be prepared for a more precise study of how the established conventions of one group fare, when they are introduced to another group possessing different conventions.

Among the social materials which illustrate this process of conventionalisation, none are more striking than the signs of written language.

THE PROCESS OF CONVENTIONALISATION

The greatest stimulus to social change probably always comes from outside the strict limits of the changing group. This is the primary psychological reason why large and important organisations, such as an army, have groups of attachés, or liaison officers, with alien friendly organisations; why commercial undertakings or other associations often form selected groups among themselves with more intimate interrelations than they possess with similar bodies outside of the special grouping, and why, in the world as we know it, there have grown up any number of elaborate methods of facilitating contact between social groups. Yet when a technique, a custom, or an institution is adopted into one group from another, by whatsoever means, the selective conservation of the recipient group always works it into a pattern which is distinctive of itself. It is this process of the development of characteristic patterns within the group, into which all alien material that is retained must be fitted, that we are now to study. The emphasis passes, for the moment, away from psychology, in the strict sense, towards sociology. It is not with emotions, images, ideas, individual attitudes, that we are concerned, but with objective changes of culture. The general name which, as I have already stated, I propose to use to cover the whole of the processes involved is conventionalisation. The problem is: Here is an element of culture coming into this group from another. What are the main principles of the changes it must undergo before it finally settles down to an accepted form in its new social setting? The range of this question is immense.

REQUIREMENTS OF A THEORY OF SOCIAL RECALL

The pathway to an understanding of the general mechanism of remembering is strewn with rejected theories; but concerning social memory proper there has been much less speculation. For the most part the arguments used have turned upon analogy. A more or less elaborate likeness has been drawn between the social group and the human individual, and on the basis of this, whatever is attributed to the latter has been ascribed to the former. This is certainly unsatisfactory. What is required is a direct study of social facts, and conclusions should be founded upon these facts alone. Speculations based upon analogy are bound to appear incomplete and unconvincing.

There are, it is true, a number of views regarding the social origin of memory. Prof. Pierre Janet has, for example, written upon this topic in his most persuasive and attractive manner. Human beings, he believes, confronted by a difficult world, have been forced to invent various special ways of dealing with adverse circumstances. The most important of these all concern the utilisation of absent objects. Many of the most significant biological advances, such as the development of distance receptor organs, are directly connected with this, and memory, which is something utterly different from conservation, is a triumphant solution of the problem. Now according to Janet the need for dealing with absent objects becomes acute only in a social situation; and this is why the development, or invention, of memory is inextricably bound up with some form of overt conduct and especially with the use of language.

THE METHOD OF APPROACH

The most persistent problems of recall all concern the ways in which past experiences and past reactions are utilised when anything is remembered. From a general point of view it looks as if the simplest explanation available is to suppose that when any specific event occurs some trace, or some group of traces, is made and stored up in the organism or in the mind. Later, an immediate stimulus re-excites the trace, or group of traces, and, provided a further assumption is made to the effect that the trace somehow carries with it a temporal sign, the re-excitement appears to be equivalent to recall. There is, of course, no direct evidence for such traces, but the assumption at first sight seems to be a very simple one, and so it has commonly been made.

Yet there are obvious difficulties. The traces are generally supposed to be of individual and specific events. Hence, every normal individual must carry about with him an incalculable number of individual traces. Since these are all stored in a single organism, they are in fact bound to be related one to another, and this gives to recall its inevitably associative character; but all the time each trace retains its essential individuality, and remembering, in the ideal case, is simple re-excitation, or pure reproduction.

Now we have seen that a study of the actual facts of perceiving and recognising suggests strongly that, in all relatively simple cases of determination by past experiences and reactions, the past operates as an organised mass rather than as a group of elements each of which retains its specific character.

ITS BEGINNINGS

No doubt there is a sense in which it is true to say that experiment in psychology is “at least as old as Aristotle”; but certainly it can claim no great age as a method of systematic exploration of human reactions. This is a matter of some significance; for it means that, before experiment was systematically applied in psychology, the experimental method had already a long history of development in other realms, upon which the early experimenters in psychology built both their aims and their methods. All the pioneers in experimental psychology were trained either in physics or in physiology. Their influence, both for good and for ill, still remains stamped upon the accepted methods of the psychological laboratory. Further, as was natural at the time, they were often men of a strong philosophic bent. Gustav Theodor Fechner, who is usually regarded as the founder of experimental psychology, was, in fact, concerned mainly to establish a panpsychic view of the Universe. The tendency to use psychological experiment chiefly as a buttress to some all-embracing philosophical theory is one that has clung to experimental psychology ever since its earliest days, and has provided the critics of this branch of science with many of their most potent modes of attack.

Fechner took his degree in medicine, but turned soon to physics and mathematics. In developing his methods he built upon the work of E. H. Weber, the physiologist, and had the close co-operation of A. W. Volkmann, who was Professor of Comparative Anatomy at Leipzig.

INTRODUCTION AND METHOD

In one sense the whole of the subsequent experiments to be discussed in this book will deal with processes of imaging. For imaging consists essentially in the utilisation of experiences which are no longer fully presented to perceptual sensory organs, and such utilisation is a part of all remembering processes. But before discussing the exact ways in which imaging enters into remembering, it is necessary (a) to get as definite an idea as possible of precisely what should be included in imaging, and (b) to determine at least some of the most important of the general conditions and types of imagery.

The possibility of using ink-blots in the study of imagination was suggested by G. V. Dearborn, and used by him in a brief “Study of Imagination” Some other experimenters had already made use of this type of material, before I carried out the experiments which I am about to discuss, and since then similar ink-blots have been used by a number of investigators for various purposes. I prepared a series of blots, some of which are reproduced in the accompanying plates. Thirty-six blots were used in all. They were variously shaded or coloured, mounted on ordinary post-cards, and laid face downwards in front of the subjects. The instructions were: “Here are a number of ink-blots. They represent nothing in particular, but might recall almost anything. See what you can make of them, as you sometimes find shapes for clouds, or see faces in a fire”. The subjects turned over the cards for themselves.

INTRODUCTION

The main data of my experimental studies have now been set forth and it is time to consider whether they have any consistent theoretical significance. As everybody knows, there is no lack of theories of memory. Biologists, philosophers and—though, perhaps, to a smaller extent—psychologists have put forward numerous speculations. Biologists, on the whole, have treated memory mainly as a repetitive function and have speedily become immersed in theories of how specific ‘traces’ maybe made and somehow re-excited. Philosophers have naturally tried to find how what is recalled is related to a ‘real’ world and have discussed the validity of the information that is given by recall. Most of the more careful psychological work, especially that which has an experimental basis, has been concerned with special problems in the general field; for example, with the normal course of learning and forgetting; with the influence of special conditions such as position in a series, intensity of stimulation and the like; with classifications of the typical kinds of association and with a study of ‘association strengths’. Psychopathologists have considered prodigies, whether of remembering or of forgetting, and have given detailed attention to the question of the affective basis of recall. The classical theories have often been adequately summarised and I shall make no attempt to describe them here. More recent general psychological theories are still in a fluid state and no brief summary could do them justice. I shall therefore proceed at once to consider how my own results may lead to theoretical conclusions regarding the characteristics and functions of recall.

INTRODUCTION

Not only stories, descriptions of events and arguments, but various types of picture material—art forms, decorative patterns, graphic representations of common objects and the like—frequently pass rapidly from person to person and from group to group. In the course of this transmission the material often suffers considerable change. It is therefore interesting, both from the point of view of the remembering of picture material and from that of its bearing upon the development of conventional representations, to see how far such a course of change can be experimentally produced. It is well known that a person may perfectly remember what he cannot reproduce well by drawing. But it has constantly appeared that when any complex form is presented for visual observation certain features of it will be dominant. These are pretty sure to appear again in any reproduction that may be effected; and, whether they are depicted accurately or not, it seems probable that it is these which set the direction of change in the course of serial reproduction.

Accordingly I collected a number of very simple picture representations and submitted them to a course of repeated and serial reproduction under exactly the same conditions as have been described for the prose passages. As before, I shall give here only a very small selection of the many results that have been collected. This selection will sufficiently illustrate a few of the outstanding characteristics of the reproductions. The principles are found to be broadly the same, both in repeated and in serial reproduction. I shall therefore draw my illustrations from the chains of reproduction only.