The Group Fallacy in Relation to Social Science
Floyd Henry Allport
There is no such thing as a group mind; it is a misleading and harmful conception in every way, whether it is applied to crowd behavior, social conflict, revolutions, or the theory of the superorganic: such is the thesis of the author of this pamphlet.
(2) << INTENTIONALLY BLANK >>
(3) THE theory that a crowd possesses a mental life resulting purely from aggregation and superadded to the mental processes of its members seems to have perished at the hands of progress in social science. Its ghost, however, has been exceedingly difficult to lay: The convenient and picturesque manner of speaking in terms of groups as wholes has infiltrated much of our social thinking. The subtlety of this influence may be partly explained as follows. When we read that a certain army captured a city, or a certain football team defeated a rival team, the language, though not precise, is not misunderstood. It is clear that it is solely the individual soldiers or players who combined their efforts and accomplished the feat described. When we read, however, that the crowd becomes violent, emotional, or intolerant, or that it thinks in images or lacks reason, we are in danger of being misled into thinking that it is a crowd mind rather than the minds of individuals which is accountable for these phenomena. So long as the language is intended and accepted as purely descriptive and metaphorical no confusion exists. But the transition from description to explanation is in such cases very subtle, and not always recognized. The intangibility of the phenomena combines with the collective or abstract use of language to produce an error. This error is the attempt to explain social phenomena in terms of the group as a whole, whereas the true explanation is to be found only in its component parts, the individuals. Such an explanation is in itself false. We do not need a super-mind hypothesis to explain mob action, if we but take the trouble to study the individual in the mob and observe how he is responding to the stimuli afforded by the behavior of his fellows. This neglected field of study is being brought to the foreground by a modern social psychology whose data comprise the social behavior of the individual. The crowd mind theory is not only false; it retards in a special manner the discovery of the truth. Pointing toward the whole rather than the parts it withdraws attention from the latter and incites thought in precisely the wrong direction.
The influence of the social mind theory is as widespread as it is subtle. In the various guises it has assumed it has become amazingly protean. We find a counterpart of it in the social organism metaphor of Plato, as well as in the modern varieties evolved by Spencer, Espinas, and Münsterberg. In these alluring metaphors much space is given to developing a large conception which leads us nowhere; while the true
( 4) origin of social organization, the psychology of the individual, has been correspondingly neglected. We meet with the collective error again in the social systems of philosophical idealists. Here it plays freely into the hands of a metaphysics of the objective nature of mind. Professor Bosanquet, for example, argues for the existence of a "general will" which gives definition to individual wills. We meet with the fallacy again in theories of government and morals, where it takes the garb of national spirit, and absolution in conceptions of the State, the Law, and the Right. The abstracted social mind is put back again into the individual as the "collective" or "bee-man" of Sageret  as well as in the semi-mystical theories of Sidis  and Trotters .
A scholarly attempt to put the social mind hypothesis upon a tenable basis has recently been made by Professor McDougall.  In his view, which he calls the "Group Mind," the social reality is alleged to exist not in collective consciousness nor behavior, but in an organization or structure of social relationships which can be conceived to exist only in mental terms. A university, for example, is not comprised in its material aspects, nor even in its personnel. It is systematized relation of individuals and traditions, intangible, but real and mental, and carried along independently of particular individuals. Our answer to this is, of course, that, although not dependent upon particular individuals, this organized tradition is dependent upon some individuals. It exists in the attitudes and consciousness of these separate persons, just as it did in Professor McDougall's mind while he was developing the illustration. So far as we know the group mind has no other form of existence than this, namely in individuals; nor could we conceive of it exerting any effect upon the social order except through these agencies.
But perhaps the best way in which to deal with the fallacy under consideration is to expose its inadequacy when put to the test of explanation. Accepting at face value the social psychology of Le Bon, suppose we proceed, equipped with the concepts of crowd intolerance, emotionality, irrationality, and the like, to explain the actual mob phenomena of modern society. We should find that our terms merely describe, they do not explain. We are ascribing the actions of the mob to things which mobs generally do— which is mere tautology. There is at hand no means for explaining the differences in the behavior of different crowds, since all are emotional, irrational, and the like. Why should the excitability of one crowd express itself in whipping non-church-going farmers, that of another crowd in looting grocery stores, and that of still another in lynching negroes? These questions throw into relief the necessity of delving deeper for our notions of cause than terms which describe the
(5) crowd as a whole. We must seek our mechanisms of explanation in the individuals of whom the crowd is composed.
We should fare no better if we were to depend upon the group mind theory for notions of cause. The "group mind" in the sense employed by its exponents is a static mind. It is a result, not a cause, of individual behavior. It offers no provision for explaining social change— change, that is, in the group mind itself. One might, for instance, ask how such a change could have been produced in the organized mental life of a group as the recent change from the tradition and practice of an alcoholic era to a régime of prohibition in our own country. For answers to such questions we must turn again to the responses of individuals to changing material and social environments and to the influence of leaders, inventors, and reformers. Such attempts as have been made to explain social movements in group terms have been purely on the descriptive plane. We may cite as an example Spencer's theory of a progressive, diversifying evolution of the social organism. Such metaphors are descriptive rather than revealing. They express; but they do not explain. They are soon passed by in the serious work of seeking causes.
The views which we have thus far examined are examples of what I have chosen to call the "group fallacy." This fallacy may be defined as the error of substituting the group as a whole as a principle of explanation in place of the individuals in the group. The word "group" is here used in the widest sense. Two forms of the fallacy may be distinguished. The first attempts its explanation in terms of psychology, assuming that it is possible to have a "group psychology" as distinct from the psychology of individuals. The second renounces psychology and relies upon some other form of group process for treatment of cause and effect. Both forms abolish the individual; and, it may be added, both therefore abolish the services of psychology as a possible helpmate of sociology.
Turning from explicit collective mind theories the task now lies before us of pointing out a less easily recognized but widely diffused influence of the fallacy under consideration. I shall present three examples from the more dynamic phases of sociological theory, namely social conflict and social change. These illustrations deal respectively with the psychoanalytic interpretation of group conflicts, the mechanism of revolutions, and the cultural approach to social causation.
1. The Group Fallacy in Social Conflict.— One of the most interesting varieties of the group fallacy is that which translates mental conflict within the individual into terms of dissociation within a hypothetical social mind. Although a number of writers have dealt with this pathological metaphor, its most elaborate development appears in a posthumous work of the late Dr. W. H. R. Rivers. The following points of fundamental resemblance are observed by him between the neurotic person and the abnormal social order. First, the cause of the disorder in each case lies deeper than the outer manifestations or symptoms. Social diagnosis, like the diagnosis of the psychoanalyst, must penetrate into the hidden forces of life. The prognosis, since the derangement is com-
( 6) -plex, is also uncertain in both cases. The repression of one portion of society by another is said to be closely analogous to repression and dissociation in the individual. In both cases the repressed element remains and causes trouble when the tension becomes too great. Rivers describes two forms of dissociative process. The first is unwitting: the individual merely turns his back upon that which is unpleasant. Just as we tend to let disagreeable experiences pass out of our attention, so one portion of the group (the upper class) preserves its peace of mind by ignoring the existence of poverty, disease, and kindred evils in the other portion. So far the metaphor is obvious. But if the "social mind" in which such dissociation takes place were to be taken literally there would result a ridiculous confusion. Shall we say for example that the upper classes who do the suppressing are conscious, while the lower classes, since they form the material which is banished from the social mind, are therefore unconscious.
Rivers' second type of dissociation is one in which the individual deliberately and wittingly forces painful experiences from his field of consciousness. The group analogy is that the more fortunate and powerful class of society "deliberately represses outward manifestations of the discontent [among the lower class] which social wrongs arouse." Freud himself invites confusion in this field by applying to the individual a term borrowed from social usage, namely, censorship. And at this point Rivers sounds a note of caution against literal interpretation which he himself does not heed. The persistence of the repressed element, he says, is common to both the neurotic individual and the abnormal group. Tensions are thus created breaking forth as hysterical behavior in the individual and catastrophic change in the social organism. Another type of outlet in the individual is furnished by the symbolism of dreams, which are elsewhere shown by Rivers to be analogous to the symbolism employed by the repressed faction of the social order. Thus a primitive tribe subjugated by a more powerful people preserved their religious ceremonies, but gave them a disguised character so as to conceal from the conquerors their true meaning. Hanging a hated person in effigy is another example of the social use of symbolism to release feeling without incurring punishment. The similarity of such mechanisms to the individual dream process is not to be questioned. We have no justification, however, for alleging dissociation and symbolism to be mechanisms of an "over-individual," or a "social mind." We have here a collection of individual inhibitions. A mental conflict exists between struggle responses against the oppressors on the one hand, and avoidance of punishment on the other. It is really a struggle between anger and fear. But this conflict and its release through symbolism lie, so far as the mechanism of explanation is concerned, wholly within the individual. To expand these mechanisms to the proportions of dissociation within a social mind is to destroy their significance.
Parenthetically, we are reminded of a contemporary instance of social conflict and evasion to which the collective viewpoint might be amusingly applied. There is said to be an unwritten law of censorship adhered to by managers of low class theatricals. It is this: that any
(7) stage joke, no matter how salacious its meaning, may be allowed to pass if it has also a different meaning intelligible to those who are too pure minded to comprehend the other. This arrangement is indeed convenient, for it permits the clergyman and the roué to sit side by side in the front row, each enjoying the performance from his own angle, while the tranquility of the social mind remains serenely undisturbed!
That the pathological form of the group fallacy leads in precisely the wrong direction is evident upon closer analysis of the relation between social conflict and mental conflict. A significant fact taught us through psychoanalysis is that one horn of the dilemma present in mental conflict is usually social in character. It consists of a system of socialized habits inculcated in the individual through stimulation by others, and striving in opposition to the unmodified egoistic drives. We deny ourselves immediate cravings because to satisfy them would infringe upon the needs and desires of others. Were it not for this denial, an overt or actual conflict would result between ourselves and other members of society. To avoid such social conflict the socialized reactions inhibit the unsocialized, and between them engender in the individual a mental conflict. Sometimes it is fear of the social environment rather than socialized habits which represents the social force in the conflict. Thus in Rivers' example of veiled ceremonials the lower class, not daring to risk overt social combat with their masters, develop in themselves a mental conflict between hate and fear which finds its release in some disguised manner. On the other hand, when the members of the upper class shut out of their consciousness the miseries of the lower, this behavior holds sway only in the absence of overt conflict. When the masses rise in revolt the scene of the conflict is at once shifted from within the mind of the aristocrat to the field of outward combat between groups. At every turn therefore social and mental conflict are inversely related in their occurrence. The more the conflict lies within the individual, the less it lies within the group, and vice versa. Instead therefore of using the mechanism of individual neurosis to explain conflict in terms of the group as a whole, we must conclude that that mechanism is precisely the one which cannot be used in that manner. Mental conflict is surely an important concept for understanding social causation; but the interpretation must always be through a collectivity of individual conflicts, and never as a phenomenon of the group as such.
2. The Group Fallacy in the Theory of Revolution.— I wish now to call your attention to a theory which represents more nearly than the one just discussed a type of the group fallacy common in sociological writing. This is the formula for revolutions developed by Professor Ellwood. At the outset, however, let me state that Professor Ellwood has always been a staunch opponent of the group fallacy, and has done no little service for the recognition of the individual in social science. Even in the theory which I cite there are portions indicating clearly that the individual has not been overlooked.
With these provisions in mind, let me attempt to state Professor EIlwood's view. The basis of social revolutions, according to this theory, is the rise of immobility of the institutions of society. Through short-sightedness or selfish interest the persons in power in these institutions block the normal processes of social change. Governments become despotic and exert too rigid a censorship. Religion and education likewise may become ultra-conservative. Reactionary and intolerant public sentiment may foster an immobility of the social order. Sooner or later, as the conditions of life change, the forces opposing these inflexible institutions accumulate until the old habits are overwhelmed. To quote the author:
The breakdown of the old habit [in another place the term "social habit" is used] may be sudden, and the society, being unused to the process of readjustment and perhaps largely lacking in social machinery therefor, is unable for a greater or less length of time to reconstruct its habits and institutions. There ensues . . . a period of confusion . . . in which competing classes strive for the mastery. If the breakdown . . . concerns habits and institutions which affect the system of social control, we have the disorders which essentially characterize a social revolution.
The conception offered above is drawn essentially in terms of the group as a whole. The social order becomes rigid and inelastic until demolished by convulsive social change. The dynamics involved are purely in terms of needed change, resistance to change, and final overthrow leading to temporary chaos,— all these as phenomena of the group itself. So far as quoted above we find no attempt to portray the causes underlying this struggle for change, such, for example, as the blocking of instinctive or prepotent responses involved in the economic and family life of the members of society. These of course are phenomena of the individual. The group-wide aspect of social inelasticity, though descriptively true, is not the causal factor. Not lack of change itself, but thwarting of individual responses, produces struggle. There have, in fact, been revolutions precipitated by changes in institutions, rather than by immobility. According to Martin revolution is not a sign of the unyielding character of social controls, but of their very weakening, a condition which gives the restless proletarian a chance to assert himself. Social inelasticity is therefore a descriptive concept; it does not penetrate to the level of explanation.
A further trace of the group fallacy lurks in the phrase "social habits." A sudden change in the government destroys the habits of society and brings confusion, until habits of response toward the new régime are learned. Strictly speaking there are, of course, no "social habits"; but only a collection of habits of individuals. Shall we then say that post-revolutionary disorder is due to lack of habituation of the individuals to tile. new type of government? Such lack of habituation
(9) no doubt exists (again we have a descriptive truth) ; but the cause of the confusion lies rather in the unrestrained following by individuals of whatever factions serve their own interests, and in the new oppressions and revolts caused by the temporary accession of a tyrannous dictatorship. Again causation is discovered in terms of the individual. A real danger lurks in the prefixing of the term social, not merely to the word habit, but to any term denoting a psychological process.
In fairness to Professor Ellwood let me now quote a portion of his account which shows that he was by no means wholly insensible to the basic individual factors, and in which such factors are recognized although the language is not wholly free from the confusion just described. His account continues:
[The party of revolt] is composed in general of those individuals whom the changed conditions of social life have most affected, in other words, of those individuals on whom the old social habits set least easily, and whose interest therefore lies in another adjustment. (Italics by the present writer).
And, again, in a footnote:
The motivation (cause) of revolt in large masses of men is always lack of adaptation.
The criticism which I have ventured is therefore largely a question of emphasis. No fault can be found with such writing as an interesting, and even valuable, piece of description. My only complaint is that it tends (unconsciously to the writer) to usurp the prerogative of explanation. As such it leads us away from the discovery of true causation which lies, not in groups, but in persons. Much writing, I fear, both in sociology and social psychology, partakes of this error in method.
3. The Group Fallacy in the Theory of the Super-Organic.— Turning now from these mental forms of the group fallacy, we may examine one which renounces psychological explanation. This is the super-organic hypothesis developed principally by Professor Kroeber. This theory is concerned with the social as such; that is, with the super-organic. A study of individuals gives us a knowledge only of individuals; it can never reveal to us the reality of organized society in itself. The data of the super-organic consist of culture in the broadest sense, a body of "super-organic products" "carried along from individual to individual and from group to group independent of the nature of these individuals and groups." Professor Kroeber insists that we must study the laws of development and change in these data alone. He affirms that both analysis (description) and determination of process (explanation) are possible wholly within the plane of the super-organic. Culture, in other words, is explicable in terms of itself. It is true that he suggests a psychology for the study of super-organic processes. He demands, however, "not an individual or `psychic' psychology, but a social or `superpsychic' psychology, in short, sociology" (as he conceives it).
From the standpoint of the psychologist a psychology such as that just specified has no existence. I do not believe it ever will exist. For psychology is a study of the individual; to extend its principles to larger units is, as we have seen, to confuse their meaning. To extend them to entities which are not even organic is to make them wholly unintelligible. We do not consider a reasoned argument to be identical with the thinking behavior which produced it. It is the product of a psychological process, not the process itself. In the same way culture, being essentially recorded thought, is a product not a process. Whatever the terminology employed, Professor Kroeber has effectually turned his back upon psychology as aid in sociological explanation.
The freedom of the super-organic from dependence upon organic and individual sciences Professor Kroeber bases upon a classification of sciences into four planes or levels. The data of these sciences are, respectively, the inorganic, the province of physics and chemistry; the vital organic, represented by physiology; the mental organic, constituting the field of psychology; and the super-organic, which are to be treated by sociology. The biologist is said to accept life and inquire into its forms and processes as such. Organic life may, it is true, be reduced to the elementary basis of physics and chemistry. But that is not the main task of the biologist; since, if it were, biology would not differ from physics or chemistry. Similarly, psychology should accept the mental as such and analyze the processes within this field, ignoring the physiological elements to which mental phenomena are reducible. Finally, sociology treats of the organized products of mental life, seeking in these alone its material of analysis and explanatory process.
As a criticism of this argument I shall endeavor to show that the separation of sciences into these distinct strata is apparent rather than real. The sciences really overlap one another, and possess certain fields in common between adjacent higher and lower planes. This relationship may be stated precisely as follows: The phenomena studied by any science are approachable from two different viewpoints. The first is that of description, the second is explanation. A complete program for any science embodies both these forms of approach. Now the essential fact is that in the hierarchy of sciences the field of description of one science becomes the field of explanation for the science immediately above it. Not all of the descriptive material of the lower science is used by the higher; but only that which is relevant to the explanation of the data studied by the higher science. Let us illustrate by tracing through this hierarchy a single example of importance in the psychological level, namely the reflex arc concept.
First, the physiologist notes that when a stimulus, such as a pin prick, is applied to a sensory nerve ending a certain muscle contraction follows. He notes also certain properties of this event, such as latent time, refractory phase, and inhibition of other reflexes by this one. These are descriptive aspects of that physiological unit called the reflex arc. By the aid of the microscope he is also able to describe such minute features as the synapse. But the physiologist must not be content with mere description; he must explain. In order to do this he must borrow
(11) certain principles from the lower sciences, physics and chemistry. Thus to account for neural transmission and the action of the synapse he employs the laws of electro-chemical change, polarization and combustion. The existence of an intermediate science, such as organic chemistry, proves how closely the organic is dependent upon the inorganic for its causal principles. Description is thus carried in physiological terms, explanation in physico-chemical terms. But we call this science physiology.
The psychologist in turn is attracted by the field of human behavior. He observes the higher integrations of response, such as emotions, habits, and thought, their speed of operation, and ability to inhibit or reinforce one another. He is interested not so much in the reflex as a detached physiological unit as he is in what response (involving usually a pattern of reflexes) is linked up through synaptic functions with a particular stimulus. The realm of phenomena described by the psychologist thus transcends in scope and complexity that of its lower constituent science, physiology. But how about explanation? It will be seen that for principles of causation in the study of behavior we must descend directly to the reflex arc level, and accept as explanatory its conditions and characteristics as described by the physiologist. Instincts and emotions are conceived as reflex patterns, involving more or less innate coordination of synapses. Learning and thought involve selection among reflexes and fixation,— a process explained by change of resistance at the synapse. Nerve transmission, altered resistance, and correlation at the synapse are therefore conceptions which belong in two sciences. For the physiologist these words are descriptive of things which universally occur. To explain them he must descend to principles formulated as descriptive laws in physics and chemistry. For the psychologist these neurological conceptions are explanatory. Since they are of universal occurrence he leaves their deeper explanation to the physiologist, and applies them directly as explanations of the higher phenomena which psychology describes.
Turning now to the sociologist, we find that the data which he describes reach the highest point of breadth and complexity. They embrace collections of individuals in organized societies, the products of such organizations, and the changes which they undergo. This is indeed a vast field for descriptive analysis. Yet for explanation sociology is in its turn dependent upon the descriptive formula of the science just below it, namely psychology. Just as psychology has to seek its causation within the units (reflex arcs) of which its material, individual behavior, is composed; so sociology must find its explanatory principles in the units (individuals) of which society is composed. The formulations of individual behavior which psychology explains at a deeper level sociology accepts and uses as tools for explanation. Thus social continuity can be understood only through the concept of learning; social control demands the knowledge of instinctive mechanisms through which government and other institutions coerce individual behavior; while progress and cultural change rest largely upon invention, which is in turn the thought process of individuals.
To recapitulate: the sociologist describes social or collective phenomena and explains them in terms of individual behavior; the psychologist describes behavior and explains it in terms of reflex mechanisms; the physiologist describes the reflex mechanism and explains it in terms of physical and chemical changes.
Whereas Professor Kroeber represents description and explanation upon the same plane in the hierarchy of sciences, the argument just advanced would prove that the explanatory portion of each science is derived from the descriptive portion of the science just beneath it. Overlapping, rather than separation, of the planes of the sciences seems therefore to be the rule. A sociology as a science of pure super-organics aims to step entirely out of the hierarchy of scientific method. It does not overlap, but is pushed up, detached from its moorings, and cut off from its explanatory support in psychology. Consequently a purely hypothetical set of social dynamics is projected by Professor Kroeber as awaiting future discovery, any true understanding of which has thus far been revealed neither in theory nor in practice. If the foregoing analysis is sound, the theory of super-organic causation fails by the very standard upon which its claim for existence was based.
Anthropologists may perhaps take exception to the remark that no causal principles can be discovered in the purely social order, and may refer to the laws of cultural dynamics, recently established by ethnology. In fairness to them this claim must be examined. It is true that laws of culture growth and change have been worked out. These are well summarized by Professor Ogburn under such topics as curve of cumulative growth, diversification, diffusion, independent discovery, cultural inertia, survivals and cultural lag. It is true also that these concepts are stated wholly in terms of culture itself. It is my contention, however, that they are fundamentally descriptive rather than explanatory. The dynamics which they involve cannot be truly understood without
(13) recourse to the psychology of the individual. Thus Professor Ogburn freely employs psychological factors in explaining the mechanism of cultural inertia and cultural lag.
A rich field, but little explored, lies ahead for the elucidation of cultural development in psychological terms. Though ethnologists may well lose patience with the poverty of attempts thus far made, the fact that there are great possibilities still remains. A few illustrations may here be in place.
First, as to the origin and progress of culture. Invention, a term lightly used by ethnologists, must be explained in psychological rather than cultural terms. The need or prepotent drive behind inventive behavior exists only in individuals. In manner of procedure invention is but a variation of trial and error learning, shortened by the implicit random movements of thinking. The culture base short-cuts individual inventive behavior; but it is absurd to say that it can take the place of such behavior. We find learning of this character extending through vast reaches of time from the earliest neolithic culture down to the modern machine shop.
As another example we may cite the universal culture pattern described by Dr. Wissler. The origin of this universal pattern is inscrutable until we conceive its various parts as means for the adaptation of the prepotent, or instinctive, needs of man. Thus the innate responses of struggle and defense are behind the development of such culture products as government and warfare. The hunger drive gives rise to property, scientific knowledge, and material culture in tools, though these of course serve other needs as well. Family and social systems are evolved largely as satisfactions of sexual interests. The speech complex, through the control it gives over others, serves all the prepotent needs.
A third phase of culture explanation lies in the problem of continuity. In its most essential nature culture is not a group of super-organic products. It is distinctly organic and lies within the individual. Social causation lies not in the tool but in the socially inculcated habits of constructing and using the tool. This theme has been immortalized in Mark Twain's story of the Connecticut Yankee. Culture upon a descriptive plane may be studied at large. This is the task of the anthropologist. In an explanatory sense, however, it must be sought within the individual. Explanation of this sort is a part of the program of the sociologist. The theory of the super-organic is a well meant but futile attempt to transplant the historical and descriptive method of ethnology into the field of social causation.
At every turn we are thus led back to the behavior of the individual as the source of explanation of social facts. Does this prove that the social as such is to be banished from consideration? By no means. The fact that we decline to use the group, its products, or its changes as principles of explanation, does not lessen our interest in these phenomena. They are, after all, the chief objective of the sociologist. We have not, as Professor Kroeber charges, denied reality to the social order because
( 14) we do not explain that order in terms of itself. When we turn to the individual for causation we but follow the rule of the other sciences in explaining the complex in terms of the simple, the whole in terms of its parts.
I have previously referred to the fact that a science does not employ all the descriptive elements of the science just below it for its own purpose of explanation. Neuro-muscular physiology for example suited the phenomena of psychology better than the physiology of circulation or digestion. In the same way sociology may choose a certain phase of psychology for its explanatory groundwork. This phase is that part of the behavior of the individual which stimulates other individuals or is a response to such stimulation from, others; in other words, the social behavior of the individual. Such indeed would be my definition of social psychology; and I would suggest that its thorough and scientific formulation would provide sociologists with a set of principles of the highest value for explaining the phenomena of a purely social order. The work of sociology, therefore, would be to describe social aggregates and social change in terms of the group, but to explain these phenomena in terms of the social psychology of the individual.