Frederic Charles Bartlett [Autobiography]
Frederic Charles Bartlett [Autobiography]
Citation: Bartlett, F.C. (1936). 'Frederic Charles Bartlett [autobiography]', in C. Murchison (ed.) A History of Psychology in Autobiography, Vol. III, Worcester, MA: Clark University Press (pp. 39-52).
The first book on psychology that I ever read was Stout's Manual. From this it may be gathered that my way into the science was by the philosophy of it. About the same time I read also the same author's Groundwork and his Analytic Psychology. The third of these books made a particularly great impression on me which remains to this day. I also read Ward's great "Article" in the Encyclopaedia Britannica, travelling once a week 18 miles to the nearest public library and bringing back copious notes with me.
Yet even in these early days not all my work was on the philosophical side. I did whatever physiology I could, but this was little and ineffective, for I had no available laboratory then. Soon after its publication I read Myers' Text-Book of Experimental Psychology. Much of this fascinated me, but again for lack of a laboratory I gained less than I should have done from my reading. With cardboard, pins, bits of string, and "elastic glue" I fixed up for myself such crude instruments as I could and tried a very prentice hand at experiment. At that time also my interest in political problems was extremely keen, and by the practical prosecution of this I was learning, though I hardly suspected it at the time, some social psychology in the field.
Three years after my introduction to psychology, and with a very predominant philosophical bias in my outlook upon life, I came up to St. John's College in the University of Cambridge. It had a good psychological record. It was the college of G. F. Stout, of William McDougall, and, in my own day, of W. H. R. Rivers. I went to it, in fact, largely because Rivers was there and my interests were turning strongly towards anthropology. However I decided to read moral science first, and that is how it was that I came into the Cambridge Psychological Laboratory. All students reading moral science then did-or were supposed to do-four hours of experimental work weekly in the psychological laboratory. From the first I did my full share, and later a good deal more than my full share, of this work. I started the laboratory courses with little liking and finished them with much. This change was due in the first place to C. S. Myers, and in the second place to Cyril Burt who was then helping Myers with the Cambridge classes. When Burt left for more im-[40] portant work I was asked to take his place and I accepted the offer eagerly. Even then it was in my plan to go for field work in anthropology after a few years' more intensive study of experimental method in the laboratory.
The War stopped that. My own war fate was happier than most, though it was not wholly in line with my desires. It gave me much work to do, in which whatever psychology I knew must be harnessed to the solution of some very practical problems. The War ended and Cambridge slowly struggled back to life again. Psychology, everybody found, had changed. It claimed to stand more by itself. Myers and Rivers had built for dreams that many had considered wide and wild. But their dreams were of the true prophetic order, and the development they designed has proceeded steadily ever since, though it is still far from its complete fulfillment.
* * *
Apparently a psychologist must have a descriptive label if people are to take much notice of him. He must be of this "school" or of that, and if his "school" is his own invention thrice blessed is he. There may be something ineffaceably controversial in the typical psychological temperament. This puts me in some difficulty, for, if I am to say what sort of a psychologist I am, I think I can say only that I am a Cambridge psychologist. The trouble about this is that Cambridge psychology of the laboratory type has never committed itself to any hard and fast and settled scheme of psychological explanation. I hope it never will.
To make the matter clear I must say how, what, for want of a better word, I shall continue to call laboratory psychology grew up in Cambridge.
In 1877 Dr. Venn and Dr. James Ward urged the need for founding a laboratory of psychophysics at Cambridge. But they fell foul of certain theologically minded mathematicians who were horrified that anybody should even think of measuring the human soul, and nothing was done. Fourteen years passed by. Other countries had by now definitely gained the lead in the new science of psychology. Ward was still anxious about the matter and in 1891 he succeeded in obtaining a grant from the University of £50 for psychophysical apparatus. If this had been the real start, however, it is difficult to see how experimental psychology could have developed in Cambridge as it has done. [41]
The decisive move came from the physiological side when, in 1893, Sir Michael Foster, then Professor of Physiology, invited Dr. W. H. R. Rivers to come to Cambridge in order to teach the physiology of the special senses. To Rivers, trained largely with Hering and a great friend of Sir Henry Head, the physiology of the special senses was inextricably bound up with their psychology. Four years later a University Lectureship was established in experimental psychology and the physiology of the senses. This was held by Dr. Rivers, and he was given a small room in the Physiological Department as his laboratory.
The work thus begun developed considerably and psychology began to claim independent status. In 1901 a room was acquired in a dark and uncomfortable cottage and in 1903 the "department" was moved to a whole house, equally unsuitable from all points of view, but still a definite advance on the one-room plan. Meanwhile psychology had gained a powerful new recruit in Dr. C. S. Myers. He and Rivers had both joined in the Cambridge Expedition to the Torres Straits and the trip which took Rivers definitely over to anthropology brought Myers permanently within the ranks of the psychologists. This expedition did another thing. It put a social and ethnological stamp upon Cambridge psychology and this has perhaps done more than anything else to make Cambridge psychologists human as well as scientific.
Experimental psychology in Cambridge soon gave promise of breaking out of the bounds of its small house. Helped by the Moral Science Board and by the Cambridge University Association, Dr. C. S. Myers determined to build a laboratory worthy of the place and of times yet to come. The movement began about 1908. Dr. Myers was supported very generously by his family and his friends and with indefatigable energy he got the actual building in hand within some two years. A new physiological laboratory was being constructed about the same time, and, by good fortune and good will, the psychological department, though having an entirely independent administration, was housed in the same large building. Experimental psychology in Cambridge began in the physiological laboratory and it came back to its final home in the most intimate connection with the Physiological laboratory.
From this brief sketch it will be seen that laboratory psychology as it exists in Cambridge today is almost wholly the work of two [42] men: of the late W. H. R. Rivers and C. S. Myers, and particularly of Myers. Both were physiologists with wide biological training; both had a profound interest in anthropological and social research; both were medical doctors; both never allowed their love of exact method to override their study of human nature in or out of the laboratory.
It is now easier for me to say what Cambridge psychology is and what it is not. I would like to regard myself simply as a student of Rivers and Myers, carrying further perhaps the views which I have learned from them. They, together with their mutual friend, Sir Henry Head, have influenced my orientation in the subject more profoundly than any others, though I would not wish to underestimate the importance of the philosophical training that I had before I knew them.
I will now attempt to describe that orientation as definitely as I can. In doing so I shall drop the personal style and adopt a dogmatism of expression which I hope in no way represents an equal dogmatism of attitude.
* * *
Psychology is a biological science. This means that it definitely restricts its study to the conditions by which any type of animal and human response to stimuli and organization of stimuli is determined. It may, if it so prefers, further restrict itself to a study of those responses which are more complex than tropisms and the simpler reflexes. We agree to give up asking what a sensation, or an image, or an idea, or an emotion actually is and we ask under what conditions responses involving these occur. It may be said that we must know what they are before we can study their occurrence. The answer is that all such knowledge that is necessary can be given by the example type of definition.
There is no need that we should artificially restrict ourselves to the study of such conditions of response as can be observed by a second or a third person. Very often the most valuable information can be given in terms possible only to the person himself who responds. Always such information should be sought, and sometimes it may even be the only sort of information possible. When both modes of access are open they continually show such agreement as to warrant belief in first-hand observation as a sound method. For example, the description by the subject himself of the effects of adequate [43] stimulation of the pain sense organs-the slight latency, the "welling up" of pain, the swift adaptation-fits very well the study by physiological methods of the propagation of impulses following upon such stimulation.
Another general point which is of extreme importance is that, whenever stimulation sets up, either immediately or remotely, responses that can be only or best described by the subject's responding himself, these responses may themselves become important conditions determining subsequent reactions. This is by no means to be taken as axiomatic however. Either it must be demonstrated or the whole idea must be discarded.
Since psychology is a biological science, its methods must be as rigorously exact as possible. Here we run into great difficulty. Physics and physiology, to both of which psychology on its experimental side is closely related historically, have their own ways of securing exactness of method. The former rests largely upon constancy of objective conditions. The latter, in order to secure instances suitable for the application of the same ideal, often resorts to the method of simplifying reactions by cutting out other responses than the one that is being experimentally studied. Naturally psychology has imitated these methods, but the results are often disastrous.
Constancy of objective determination is obviously consistent with variety of subjective attitude. Equally, since the human organism has grown up in a very varying environment, variation of objective circumstances is consistent with constancy of subjective orientation. Subjective attitudes or orientations are an important part of every response at the psychological level. It often becomes a matter of nice consideration whether the objective conditions or the orientation of the organism are the predominant determinants of the response. Neither can be neglected, but when the latter takes the lead it is folly for the psychologist to stick to the ideal of constant objective conditions merely because this is formulated and accepted by other scientists.
Again, the ideal of simplifying the response by cutting it off from others with which it is normally combined is all very well in its way, but it is a dangerous and slippery ideal in psychology. A physiologist who is interested in visual functions may isolate a single, or a few, appropriate nerve fibers with their connected sense organs, and observe how impulses are propagated when the sense organs are [44] stimulated. But if a psychologist who is concerned with relatively high-level responses like recall endeavors to isolate the response by simplifying the stimulus he has performed a very different operation. This has been the traditional method, as, for example, in the use of so-called meaningless material in the bulk of memory experiments. Though we simplify the stimulus, we leave the responding mechanism as complex as ever and we may merely force it to mobilize all its resources and to adopt unusual and often highly complicated modes of reaction.
The psychologist must refuse to be overawed by the demand for objective constancy and he must treat very carefully and critically the demand for simplification of response. Does it follow that the psychologist must give up every attempt to apply exact method, shut up his laboratory, and relapse into general speculation? This is not the answer that any biologically trained person is likely to give.
We come now to the positive ideas that underly the Cambridge approach to psychology, though no Cambridge psychologist would for a moment wish to maintain that they are his peculiar property.
When any person or animal enters an experimental situation he brings with him certain tendencies to specific response which, in combination with the objective, external conditions, will issue in conduct of one kind rather than of others. Some of these are very deep seated and practically common to all subjects, and we may call them instinctive. Others are almost equally deep seated, are extraordinarily persistent in a given subject even though the external conditions differ widely, and we may call these temperamental. Others again can be shown to come to birth in the course of special training, may often be set up by the play of experimentation itself, and these we may call learned or acquired tendencies. There is a class which demands very special study, for they run through practically all the members of a social group; they may appear as immediate, "simple" and inexplicable as anything can be, and they are also extremely persistent. These we may call socially determined tendencies.
Of whatever class these tendencies, attitudes, and orientations are, they are powerful determinants in all psychological response. The instinctive and the temperamental kind we have to try to pick out, and then make them starting points. We can no doubt raise questions about how they are built up. But if we do, though the prob- [45] lems are extremely interesting, their attempted solution takes us over into general biology and sets an issue at once in terms of heredity and inheritance. We can however take them as we find them and see how, when they are there, they work, and for this we can use both experiment and field observation. Of the other type we can legitimately raise questions both of origin and of results.
In all his work the psychologist must keep his eye upon these reaction tendencies which are not explicable by reference to immediate stimuli or situations, or organizations of these. If he does he will speedily come upon another problem of primary importance. Perhaps its nature may best be thrown up by a retort against a current psychological dogma. Everywhere we are being told that the psychologist must study "the whole organism." In actual fact he never does, and the view that utilizes this vague expression can be buttressed only by a relapse into an equally vague philosophy. What he does is to try to find out the order of importance of reactions, of reaction systems, and of partial factors in the objective situation. One subject observing structurally simple forms in the extreme periphery of the visual field will confidently follow the immediate visual or retinal pattern. Another will equally confidently follow some scheme of imagery, of the "primary memory" order or of some other. In each case there is a dominating partial response which sets the stream of his interpretation. Similarly in all other cases: the psychologist is just as concerned with partial responses as is the physiologist, though, by trying to study them in their normal or abnormal settings, he sets himself the task of working out a hierarchy of importance of partial responses in any instance with which he is immediately concerned.
These are the problems of the psychologist: to show the functions of reaction tendencies in combination with the external exciting stimuli, in many cases to show how such tendencies grow up and in all cases how they operate, and to determine the order of importance or dominance of partial reactions and reaction systems in whatever setting they may occur. With these firmly in mind he will use whatever methods he can: psychophysical, statistical, and any others. As all his problems are ultimately problems of conditions and of determination he is throughout scientific, and whenever he judges that the response he wants to study can be produced to order he will use experiment. [46]
Let us now reverse the picture, so to speak. All that has been said so far seems to imply that psychology is through and through functional. So it is. But it appears to be a peculiarity of psychological functions-or at least notably characteristic of them-that the material with which they deal may influence their course of expression; by "material" is here meant psychological material, not mere objective or external conditions. What constitutes psychological material? Sensations, percepts, images, ideas, thoughts, some at least of whatever is present in volition and emotion, and no doubt a good many other things for which the inventive psychologist can find names. Sensations have been roughly manhandled by recent psychologists and we are told that they are a kind of psychological Mrs. Harris- "there ain't no such person." But a touch, a pain, a light-flash does seem different from perceptions mediated by the same sensorial reactions and in a perfectly understandable sense more primary than these. However that may be, the real point is that this sort of psychological material, once it has occurred, makes a difference to subsequent reactive processes. The presence of a visual scheme may make all the difference between a series of muscular reactions in a darkroom well adapted to secure some practical end and one blundering and ill adapted to the same end. Thus, however "functional" a psychologist may be, he dare not shut his eyes to problems of psychological material. Precisely as a subject has reaction tendencies ready to come into play in every experimental situation, so also he has organized psychological material ready to influence his reactions and to set the course of his interpretations of whatever external conditions may be present.
This psychological material is what an introspectively minded subject will describe if he is asked to say what a given response appears to him to deal with, or, since there is reason to agree that he cannot always arrive at such description, it is of that order of thing. A subject will say that he is perceiving or responding to a tonal field having a certain order of arrangement or certain properties; or he is perceiving a visual pattern possessing boundaries, color, location, and standing against some visual background. In remembering he may say that he is dealing with visual images having temporal and other characters, or with vocally arranged material; or it may be that we can demonstrate in some indirect way that he is utilizing a scheme built upon materials that belong to his past life and so arranged that [47] he can construct from it whatever is needed in the interests of the specific tendency to recall. He is not able to describe this scheme as something that he can find by introspection, but it has a character theoretically the same as other things, e.g., images, sensory patterns, ideas, and so on, that he is able to find in this way. The introspective method gives us psychological material and with it no psychologist can dispense.
The reaction tendencies spoken of already are not in the same position. No direct inspection ever really reveals them or their character. They are hypotheses necessary to account for the varied grouping of psychological material and themselves subject to influence by specific instances and groupings of such material that have already been achieved. They characterize or describe how an individual deals with any psychological material with which he is confronted. Sometimes they are hypotheses made by the experimenter on the basis of the way in which objective facts are selected or grouped by his subjects. For example, two people observe the same landscape. One picks out and describes all the rocks and another all the trees and flowers. The onlooker, finding this markedly different grouping of like materials, infers that the observations are directed by the different tendencies underlying different interests. Sometimes they are hypotheses made by the subject on the basis of how psychological material appears to him. For example, in an experiment on peripheral visual perception with tachistoscopic exposure, a subject finds that one visual pattern, in a whole series of successive exposures, stands out without challenge from others. He can describe this more or less rapidly, without finding that a number of alternative formulations threaten to break in, and he says: "I am confident." Anyway, the tendencies are not of such a nature that they can be set up and directly observed by the subject, whereas psychological material is of that nature. The tendencies are necessary hypotheses which the psychologist must use in order to account For diversities and for diversities of grouping of psychological material.
Since it is demonstrable that psychological material affects the subsequent operation of reaction tendencies, while reaction tendencies account for the grouping and differentiation of psychological material every bit of psychological observation must be a cooperative effort in which both observer and subject must play their parts. When an observer experiments upon or observes himself, he gains a part of his [48] relevant information directly, as a subject does, and a part by hypothesis, as an outside observer must.
Just the same sort of problems arise about psychological material as about reaction tendencies. From the psychologist's point of view it looks very much as if some organization of such material is as ready made as are instinctive and temperamental tendencies, as liable from the first to influence the course and character of specific reactions. Such, for instance, may be those simple spatial relational references we come to call "above," "below," "right," "left"; such may be the sensory basis of the "light-shade," "sound-silence" differentiations, and such may be some part of the material which is bound up with temperamental reactions. The great part of organized psychological material is, however, built up in the course of experience. We must say how it is built and how, when it is built, it operates.
Undoubtedly, when we turn our investigation upon any instance of psychological material, as it is dealt with by an actual reaction tendency or group of tendencies, we find it both complex in structure and organized. Consequently it may seem necessary to say that in every case this material must be understood to play its part in an indivisible manner-"as a whole." This phrase is just as applicable and as inapplicable as the parallel phrase: "the total organism." For instance, visually presented material will be dealt with by one subject predominantly as colored and by another predominantly as formed. The one does not utterly ignore the form nor the other the color, but different components are dominant in the two cases. It is sometimes said that psychologists must utterly eschew the analytic attitude towards psychological material. This cannot mean that we must refuse to analyze any instance of such material into theoretically simple elements, and regard these as discrete in character and unaffected by any combinations into which they may enter, for hardly anybody ever has done this, and hundreds of psychologists, of all sorts of views, have warned us against doing this. Perhaps it means that we must avoid all theoretical analysis, sticking only to such forms as are found in the actual reactions themselves. Thus a visual pattern may be reacted to as a balance of light and shade with the brighter parts predominating in the responses made. Then the light sensory material and the darker sensory material may be treated as genuine components of the visual pattern content, and in [49] so treating them we literally adopt an analytic attitude, though our analysis is derived from a study of the reaction itself and not from any theoretical division of the whole psychological material into minimum parts. If this is what the contention that we must treat psychological material in the mass or "as a whole" means, then "as a whole" is a strikingly misleading expression to use. Just as we have to study the order of importance of partial reactions and reaction systems in their setting, so we have to try to establish the order of dominance of partial components in their setting, if we are studying psychological material.
Clearly this leaves room for a structural as well as a functional psychology. The view developed here is less interested in questions of structure, but traditionally they have been uppermost in psychological investigation. Moreover, there seems no good reason why a structuralist, if he merely wants to describe completed psychological structure in simple terms, should not push analysis much beyond those distinctions which are made in the course of actual reactions, and so make use of "mental elements," "laws of association," and all the usual stock in trade of traditional psychology. Having done this he must not try to give his results a genetic significance, and he must, of course, recognize that combinations or organizations of "mental elements" may have different properties from the elements themselves.
This rapid and inadequate development of a point of view inevitably leaves many loose ends lying about and raises many points that require far more drastic discussion, but it carries a further important implication. It is merely what it pretends to be: a point of view, an indication of how to use experiment and field observation in psychology, and an expression of a belief that psychology can grow only by the patient use of these methods. It is not a scheme of explanation. Just where the psychologist could best have imitated other scientists he has gone his own way all too often. A physiologist, studying peripheral nerves, does not go on to dogmatize about the operations of the central nervous system, or about respiratory or circulatory functions. He looks for working explanations ad hoc and not for simple explanatory principles covering all functions of the organism. The psychologist far too often works on perception, or imaging, or emotions, or instinctive reactions and then tires of the effort involved in first-hand observation and gets out of his one bit of real work the clue to everything in mental life. Complete systems [50] and schemes of psychological explanation are the biggest stumbling-block to progress in psychology.
* * *
I will drop this dogmatic habit of writing, which I do not like, and resort to a more personal style. A brief exposition of a general point of view is almost sure to be misleading in some places and to satisfy nobody less than its author. Nevertheless I hope what I have written will make it dear that I believe firmly that psychology, regarded as a science, has a great future before it. Whether on its theoretical side, as enabling us to understand the diverse workings of the mental life, or on its applied side, as enabling us to direct and control those workings in the world as we find it, I think that its greatest achievement lies yet ahead.
Perhaps this also is due to the influence of Dr. C. S. Myers, but I am sometimes tempted to think that in a few years it may be found that those who are applying psychology to industry, to medicine, to the organization and training of a nation's fighting services, and to legal procedure have done more than any others to the advance of psychology as a science. The exigencies and nature of the demands in these fields mean that an investigator must hold himself aloof from hard and fast theory, while at the same time he must define his immediate problems clearly. Whether this will be so or not, the method of work in the applied field is, in these respects, the one that has to be adopted by any psychologist who believes his study to be a special branch of biological science.
* * *
Is there any royal, best road into psychology? I cannot think there is. Even when it is treated scientifically, and I should be sorry to imply that I think this is the only way in which it can be treated, psychology is a very wide subject, in which there is room for the exploitation of the most varied interests. All I can say on this point is that if I personally had an entirely free hand in the direction of my own training from the beginning I would make it different in certain ways from what it was. Assuming I had reached the stage at which specialization can profitably be begun, I would first have a good practical training in physiology and biology and also in the elements of physical science. Contemporaneously I would read some philosophy and more logic. The positive value of [51] this would, I think, lie in a training in laboratory technique on the one hand and in the technique of thinking on the other. As it is, people who have the first very often lack the second, and people who have the second almost always lack the first. The student whose training is wholly in the field of natural science seems often to lack flexibility when he comes to psychology and to be overmuch in awe of "exact method." The student whose training is wholly in philosophy and logic is generally too prone to rapid speculative generalization and inadequately reactive to the complexity of actual fact. If I had to choose either the one or the other, I should unhesitatingly adopt the scientific training and discard the philosophic and logical, for it is quite certain that scientific advance is never merely a matter of applied logic. But I can never see why the two should not be combined. The negative value of the plan I propose is that this kind of training would do something to prevent that elaboration of physical and physiological speculation which is common among psychologists, and would help to produce a proper skeptical attitude toward universal systems of fixed psychological explanation. After two or three years of this sort of study I think I should be ready to plunge into psychology proper. I would insist upon having laboratory work, and a good deal of it, from the first. A thorough training in psychophysical methods would come early, and so would practice in the use of statistics. Elaborate theory of statistics I would leave out altogether, but I would learn how and when to use statistics as a tool. Outstanding psychological theories I would have presented as history, of value to promote orientation in the learner, but not to be regarded in any single case as dogmatic truth. All the teaching in psychology from the beginning I would have as informal, as conversational, as full of perfectly free discussion as it possibly could be.
The laboratory should be a place to work in, not to live in. I am a firm believer in the great value of dumping a student into a social group that he is not familiar with, for I think that this stimulates observation of genuine human life reactions more than anything else. Further what I learned in social contacts at first hand I would try to bring into the laboratory and test under more exact conditions.
* * *
I will finish with the sort of remark that ought to be perfectly obvious. A psychologist who thinks that his work is done, that all [52] that is now needed is the application of a final scheme to new instances, is dead. Psychology will go on and leave him lamenting. Like the reactions it studies, psychology is living and oriented forward: there can be no end to its achievements.
Copyright Note:
This article is included with permission from Clark University, as Clark University Press is no longer in operation.