views updated May 14 2018















It is almost pure tautology to say that human “social” phenomena are cases of the interaction between two or more human beings conceived as “persons,” “organisms,” “selves,” or “actors.” Hence, it may be thought that what is meant and implied by the concept of interaction and its theoretical context is the sheerest common sense. However, this is not so. This theoretical complex has had a long, complicated history, and the outlines of its place in modern social science, which is our concern here, have emerged only gradually.

On reflection this should not be too surprising. After all, science is not common sense, and its most basic theoretical ideas and frames of reference require development through complex intellectual processes which involve not only interpretations of observation but also theoretical and partly philosophical conceptualization. Perhaps this process has been particularly difficult in the case of human social action, because the subject matter is so close to immediate experience that isolating a scientifically usable scheme from the matrix of common sense is particularly difficult. In any case, for the limited purposes of this article it seems useful to begin by sketching the historical background of the problem.

Early history of the concept

From one point of view modern philosophy “got off on the wrong foot” for clarifying the nature of human interaction, while, from another viewpoint, it appears that indirectly this was fruitful over the long run. In any case, I think there is general agreement that of all the areas in which modern philosophy originated, the most relevant to the present discussion is the problem of knowledge.

The Cartesian schema

The focal early statement of this problem was Descartes’s Discourse on Method. This work is justly considered the basic philosophical charter of modern science, having posed the problem of the philosophical basis of empirical knowledge of the external world with the greatest clarity. This clarity, however, was bought at the price of assumptions and predilections which proved restrictive in two directions, both of which are central to our problem.

The first was the treatment of the significant “external world” as the physical world. This was natural, in that, among other things, the science of the time, which had recently made very striking advances, was overwhelmingly physical science. Consequently, the object of knowledge for Descartes was not conceived as “knowing” or “acting” (since physical objects do not act). Thus, his formulation blocked concern with the interplay between entities which are both subject and object at the same time and, hence, with the analytical distinction between these aspects of social actors.

The second was the treatment of the problem solely as one of knowledge. Presumably it was in order to facilitate this that Descartes treated his subject as a given—e.g., in the famous formula Cogito ergo sum—rather than analyzing it as a structured identity. In this respect the Cartesian analysis did not venture beyond asserting the existence of the “thinker” and the fact of his cognitive relation to objects in the external world.

Of course we can now say, almost at the level of common sense, that Descartes dealt with a limiting case of social action. First, he excluded the “inter” in our formula of interaction by assuming that there was no “action” on one side of the relation, i.e., that the object only came to be known and that “being known” was in no way a stimulus for the object to intervene in and possibly change the relation to the knower. Second, he excluded analysis of the complex nature of the “entity which” knows, which is part of the basic relational system of the subject-object relation. We would now hold that empirical cognition is an activity or “function” of persons, an understanding of which entails analyzing the structures and processes of personalities by virtue of which a variety of factors become so organized as to facilitate the “attainment of knowledge” as a goal output of personality systems. Furthermore, the recognition of complexity in the units on both sides of the relationship obviously entails complicating the conceptualization of the relational pattern between them. Here Descartes considered the relation merely one of the flow of “information” from object to subject, resulting in consequent “understanding,” or knowledge, with little specificity about how far and in what ways such understanding involved processes other than the simple input of information.

Differentiation of the Cartesian object

The Cartesian schema may be regarded as the primary reference point for a process of differentiation. Because it formulated a relational scheme, its differentiation necessarily cut across all of its components, involving both subject and object and the character of the relation between them. Since we are concerned with science, it is easier to understand the differentiation on the object side, which entailed the first step away from the more purely physicalist predilection of Descartes.

The human object came to be regarded not merely as a “knower” but also as a physical organism “behaving” (to use a later term) in an environment and actuated by “wants”—or, as Hobbes said, “passions”—which accounted for its action. This differentiation appeared even in Descartes’s own century, the seventeenth, notably in the writings of Hobbes and Locke. This development marked the beginning of utilitarianism, established the theoretical groundwork of both the discipline of economics and a major branch of psychology, and had important side effects upon law, political science, and sociology.

The utilitarian differentiation. The Cartesian pattern was maintained in the assumption that the individual’s wants were given. Even though these wants were also assumed to be plural, the problem of how, specifically, they were patterned and organized was not dealt with. However, the analytical concern was now no longer confined to knowing the external world but included “rationally” manipulating it through goal-directed activities. The early modern social scientist, then, is conceived of as an observer of objects who are at the same time actors striving to satisfy their wants through action. Moreover, only in a limiting case does the observer confine his observation to single individuals; generally he observes an interacting plurality. Economic exchange, through barter or the more elaborate market systems, became a prototype of such interaction, but the men of Hobbes’s state of nature, seeking to “destroy or subdue one another,” were also conceived of as interacting in this sense. Clearly, trying to satisfy wants or seeking to destroy others involves action in a sense not attributable, for example, to celestial bodies. Such wants or passions are easily distinguished from the activities—to use a term later made much of by Alfred Marshall—intended to implement them. Problems of the nature of the interaction systems generated by action conceived within this framework and of the conditions on which such systems can “function” become very important here. The nature and significance of “self-interest,” in the classic modern sense, and the basis of normative order in social systems become very problematic in this frame of reference.

Thus, the “utilitarian” frame of reference can be said to have emerged from the Cartesian problem-statement through the inclusion in the schema of a class of objects which are not physical and which interact in a sense in which neither knowers nor physical objects do. Although this conception emerges by differentiation on the object side of the Cartesian scheme, it implicitly raises the question about the position of the observer. Very clearly, knowledge of the situation, of the wants and activities of others in the interaction system, itself becomes a factor in want satisfaction. The utilitarian actor, considered as an observer, is a Cartesian “knower,” but he is more than that. Thus, introducing this additional element into the total scheme presents exceedingly important problems.

Idealistic differentiation of subject. As noted, Descartes left the structure of his subject unanalyzed: the “I” which thinks, and therefore exists, is given. Very broadly, the idealistic movement was an attempt to analyze the content of this given entity. Most crucially in Kant, it took the content of knowledge as its primary reference point. Contrary to the views of the British empiricists, knowledge was considered to be patterned and organized according to the Kantian schemata of intuition and the categories of understanding and not derivable from the “intrinsic” properties of the object world itself as they impinged on the subject in the form of “sense impresssions,” or in Locke’s term, “ideas.”

As for Descartes, the idealists’ reference here was the scientific understanding of the physical world. However, it had now been greatly “relativized”—in a special sense—because the major structure of empirical knowledge was attributed not only to the “nature” of the objects known but also to the “categories” in terms of which they are known. These categories could not be located in the objects of cognition, nor could they be treated basically as variant properties of the individual personalities of the knowers. In more modern terms, they constitute a cultural frame of reference which partly governs the whole action system so far as it is dependent on empirical knowledge. This raises a problem parallel to the utilitarian one concerning actors who are not only knowers.

It is fair to say that such a differentiation of the Cartesian subject, parallel to the utilitarian differentiation of the Cartesian object, was a principal consequence of Kant’s analysis. Thus “pure reason” concerned essentially the epistemological grounding of physical science, whereas “practical reason” regarded the other, especially noncognitive, concerns of human “actors.” Utilitarianism treated wants only as given and analyzed activities overwhelmingly by projecting a Cartesian rational knower into the role of the actor—hence the formula of rational self-interest. Kant took practical needs, which he considered predominantly moral, as essentially given, and he heavily discounted the possibilities of solving intellectually the underlying problems. Here he came close to the view that the moral imperative is existentially given.

Hegelianism. The Hegelian movement attempted to fuse the cultural component of Kant’s empirical epistemology, especially the “categories,” With the sphere of practical reason, thereby developing a unified idealistic metaphysics built about the key concept of the “objective spirit” (objectiver Geist). It then conceived of the whole of history as the “unfolding” of the world spirit, human action being essentially an “acting out” or implementation of the spirit’s “ideal” content.

Perhaps the primary disposition of the Cartesian phase of this broad intellectual development was to derive as much as possible from the inherent nature of objects; the conception of the “mind” as a tabula rasa which is only a recipient of sense impressions just carried this to an extreme. By contrast, idealism tended to attribute as much as possible to the creative activity of mind. On the one hand, this emphasized the importance in action of individual human agents as distinguished from the circumstances of their situations. But on the other hand, the problem of a cultural system, transcending (in the strict Kantian sense) the individual actor was necessarily extremely prominent.

Marxism. As the massive development of economics and, later, of psychology into firm disciplines led in the nineteenth century to the establishment of a strong intellectual tradition which positively institutionalized recognition of the wants activities differentiation within human objects of scientific observations, so the idealistic conception of an “unfolding” Geist could not satisfy for long. It required a parallel to the utilitarian differentiation between wants and activities. Such theoretical formulation emerged most clearly with the conceptions of Marx, who set a world of “material” factors over against the “ideal” factors of the Hegelian tradition. The famous aphorism about “setting Hegel on his head” makes clear, I think, that on broader grounds Marx intended to stay within the idealistic framework. The Marxian category of “material” was therefore in no way identical with that of “physical,” which had figured in Descartes and even Kant. It concerned, above all, those aspects of the human condition which are conditional to the attainment of human goals. The old primacy problem, which is inherent in the use of dichotomous conceptualization, came, in this case, to focus on whether primacy lay in the “ideal” realm or in the conditions necessary for their implementation. Marx’s materialism consisted essentially in his confronting the “Utopians” with the necessity to be “realistic” in taking account of such conditions. Importantly, his “material” system was not simply an “unfolding spirit” but a social system, in present terms an interaction system, however inadequately analyzed.

Modern developments

There are two limiting boundaries of the “action” aspect of the human condition: the biological, conceived of in terms of heredity and environment; and the cultural, conceived of as a symbolically defined system of order with normative primacy, to which men are obligated to conform on pain of this-worldly—or otherworldly—sanctions. Both major trends of social thought have been under pressure toward a reductionism grounded at one of these two boundaries. The utilitarian tradition has tended to be biologized, and the Kantian tradition to be “Hegelianized”—if one may take Hegelianism as the relative extreme of idealistic reductionism. At the same time, considerations similar to those that gave rise to the Marxian revolt against Hegelianism stimulated movements from both traditions that have brought conceptual definiteness and clarity into the middle ground between these two extremes.

Freud and personality theory

On the utilitarian side, the development of biological science in the latter half of the nineteenth century and the florescence of varieties of “social Darwinism” acutely posed the problem of how the essential components of human social action could be treated in a manner that would be realistic about the continuity between “human nature” and the organic world. Here the most important single figure was Freud, who, as a physician, started with the conception of man as organism but, as a psychiatrist, became primarily concerned with man’s behavior, not the internal state of his organs. Furthermore, the clinical method developed by Freud stressed the emotions and wishes and goals, rather than cognitive matters.

Freud started as an “instinctivist,” in the familiar hereditarist sense. Although, of course, he never abandoned emphasizing the importance of instinctual needs, he developed the concept of instinct itself from the more conventional idea of a hereditarily given pattern of behavior “triggered” by environmental stimuli to that of a highly generalized motivational system involving a complex relation between basic instinctual energy and the mechanisms of its goal specification and its control. The erotic complex, the focus of the “pleasure principle,” became a complex system which was by no means given and which was complexly integrated (or malintegrated, as the case might be) with the noninstinctual components of personality.

Relatively early, Freud gained the insight that the expression of instinctual need was regulated by the society’s moral standards—often, but in no simple sense always, in conflict with instinctual needs—and that these standards were introjected into the personality itself, becoming components of its structure. The final form of this conception crystallized about the famous idea of the superego. Later this basic mode of conceptualization was extended to the social environment, conceived of as an environment much in the Cartesian-Durkheimian sense. The famous “reality principle” came to focus on “object relations,” which for Freud meant relations to other persons, especially the parents, considered as agents of socialization. But these human objects were not only “adapted to” in the sense true for physical objects; they were also introjected—or, as we now usually say, internalized—to form part of the personality structure, particularly of the ego, in Freud’s sense.

Thus, Freud brought the distinctive properties of the social, as distinguished from the physical, environment of the action of the individual to the forefront of analytical consideration in two connections: that of normative standards and that of the more empirical aspects of the social object world. Indeed, it can be shown that even the id, the third of Freud’s primary structural subsystems of personality, is not purely “instinctual” but is organized about the “precipitates of lost objects,” especially those salient in the earliest phases of a person’s socialization experience (Parsons 1958).

Durkheim and the content of culture

Whereas Freud, from a biological starting point, arrived at the recognition of the distinctive properties of social systems, Durkheim began with the conviction that clear distinctions between social and personal systems were essential; in his famous phrase, society was a reality sui generis. In order to ground this, however, he had to escape the toils of utilitarianism. He chose to do this by harking back to the Cartesian frame of reference, including its cognitive primacy. Durkheim’s basic difference from Descartes was his insistence on exploring the distinctive category he called social facts, the facts of the social environment of the actor of reference. However, Durkheim also recognized a strong need to consider the actor as something more than a “thinker.” Furthermore, since the relevant environment was social, insofar as it included a plurality of individual actors they were all units of the same character as the original actor of reference and were conceived of as interacting with him.

From this start Durkheim came to converge with Freud at three essential points. The first was Durkheim’s primary starting position, his analysis of the distinctiveness of the social object world. If we combine this analysis with that of Freud, we can confidently speak of its distinctiveness vis-ávis not only the physical world in the narrower sense but also the organic world. The second was the idea that an essential aspect of the social environment is that it imposes normative requirements on the individual and sanctions him for compliance and noncompliance. The third was that the structure of this social environment, particularly its normative component, comes to be internalized in the personality of the individual. Otherwise, the moral authority of “society” as an agency of the control of the individual’s action—as an agency of constraint, in Durkheim’s sense—could not be understood. Although Durkheim did not develop a technical theory of the personality of the individual as a system, clearly the structure of his theoretical scheme articulates very directly with the type of personality theory Freud developed and even demands such articulation for theoretical closure.

Durkheim’s treatment of the normative components of social systems, however, went well beyond Freud, in a direction that brought him close to Kant. Since concern with the social system was primary for him (rather than residual, as for Freud), he was aware of the conception that normative components are part of a cultural system and in that sense transcend the individual. Durkheim developed this theme particularly well with his conception of “collective representations” and spelled it out in his analysis of religion in The Elementary Forms of the Religious Life (1912). He made it quite clear, as Freud did not, that systems of “representations”—no longer just a Cartesian mode of expression—were basically symbolic. In so doing, he took a most important step toward conceiving of the content of culture as consisting of codified symbolic systems and toward a general understanding of their articulation both with social systems and personalities.

Durkheim and Marx. Although this late development of Durkheim’s work brought him into direct contact with the idealistic movement, he was no more an idealist than Marx. It is interesting to compare them in this connection. Marx represents, of course, a particularly notable reaction against the Hegelian extreme, and indeed his conception of social systems is closer to Durkheim’s than has generally been believed. It can, however, be said that Marx was particularly ambiguous about the status of the normative components of social order. This ambiguity resulted from his scheme’s being incompletely differentiated in two main respects. First, the ideological-evaluative aspect was not clearly differentiated from the scientific aspect. For instance, with reference to “capitalism,” the moral condemnation of its normative structure tended to underrate its empirical importance in an analytical sense, suggesting that it merely cloaked exploitative interests. Second, Marxian thought shared with Hegelian and other postidealistic theories a commitment to historicism. This position denied the possibility, for the sociocultural field, of generalized analytical theory. Marxian theory is a theory of the development of a succession of dynamically linked particular socioeconomic systems, not an analytical theory of society in general. Marx does not use Ricardian economic theory as economic theory in the general sense but as the theory of capitalist process within one historical economic system.

Durkheim was not caught in these difficulties. He assumed, as a good Cartesian, that if social facts were facts, the general methodology of science, including general analytical theory, applied to them. And dealing with the problem of the “ideal” factors from a rather pre-Hegelian, Kantian viewpoint, he did not worry about the alleged dilemma of whether ideal or material factors determined human action “in the last analysis.”

Max Weber and the individual actor

If Durkheim avoided the ideal-material dilemma, Max Weber, starting from reference points within the German idealistic tradition, worked his way out of it, in a manner converging directly with Durkheim’s position. His crucial reference was the “motives of individuals” only through understanding (Verstehen) of the meanings of motives of actual and typical individuals could the motives be used to explain empirical courses of action—in relation, of course, to the conditions of the action situation. These motives of individuals were by no means the same as the wants of the utilitarians, precisely because, instead of being taken as given, their structure was a matter of major interest. Indeed, cultural meaning systems constituted a primary focus of Weber’s interest, as developed most clearly in his studies in the sociology of religion.

At the same time, Weber’s position is by no means a Hegelian idealism or a post-Hegelian “gestaltism” in the manner of Dilthey. In these latter cases the principal relation of a meaning complex (Sinnzusammenhang) to material reality is simply that of “unfolding.” Weber, however, made the crucial contribution of breaking down the rigid alternative of using either ideal or material systems as empirically closed, in that he developed a way of analyzing the complex interdependences between them. This is how Weber could become eminent both as a “sociologist of culture,” e.g., of religion, and as a sociologist of economic and political phenomena (particularly the latter), with his analysis of the social significance of law providing the most important link between them (cf. Parsons 1965). Thus, he managed to emphasize the complete reality of the “material” interests of persons and groups, while avoiding the faulty assumption of the Marxian model that these interests constitute a closed system which can be broken through only by a total revolutionary transformation. In regard to the individual actor, Weber, rather than having to arrive at a conception of internalization—as did the utilitarians, as well as Freud and Durkheim—quite naturally took it for granted, on the basis of his general theoretical position.

This orientation can be detailed only through the systematic analysis of interaction. Unit-by-unit social systems must be analyzed as engaged in detailed interchanges with each other, interchanges which constitute performances or sanctions according to which unit, the “sender” or the “receiver,” is the point of reference. Weber, more than any other figure emerging predominantly from the idealistic tradition, laid the groundwork and demonstrated the use of the more generalized schema toward which the movements of thought we are considering had been evolving. The relations of the authority of office to the use of power and of property to markets and exchange were primary focuses of the vast range of his empirical studies. Power and economic resources, of course, constitute particularly salient sanctions and resources in the moredifferentiated interaction systems, and Weber’s work may serve us as a kind of charter for analyzing them in social-system rather than purely economic or political terms.

A paradigm of social interaction

The broad outline of the present conceptions of interaction has emerged from the above movements. Its focus is a social system generated by and composed of the interaction of units which are “behaving organisms,” personalities, or various levels of collectivity. Acting units, however, are always involved in cultural systems, which express, symbolize, order, and control human orientations through patterned meaning systems consisting of both codes of meaning and specific combinations of symbols in particular contexts. At a minimum, an interaction system in this sense involves four analytically distinguishable aspects or components: (1) a set of “units” which interact with each other; (2) a set of rules or other “code” factors, the terms of which structure both the orientations of the units and the interaction itself; (3) an ordered or patterned system or process of the interaction itself; and (4) an environment in which the system operates and with which systematic interchanges take place. It can be seen that the various intellectual movements reviewed contributed one or another special emphasis or combination of components to this paradigm but that only at a late stage did anything like the complete paradigm emerge.

Pragmatism and the nature of the self

Before a fuller exposition of the paradigm is attempted, two further movements, which contributed less to its main outlines than those already reviewed but which have nevertheless been very influential, should receive brief attention. The first of these is primarily American and may be considered an aspect of the pragmatist movement. In a sense, James and Peirce cut through the structured rigidity of the European thought of their time to bring the whole self-object system into a new flux. Particularly in view of the “scientistic” trends in American thought, pragmatism raised questions about the self which were particularly important. James introduced a distinctively un-Cartesian pluralism into the concept: besides the I, which thought, there were an I, a me, and various other possible selves.

Symbolic interactionCooley and Mead. One branch of the pragmatist movement made a special contribution to social interaction theory, namely, that associated with Charles Horton Cooley and George Herbert Mead, which eventuated in a special kind of social psychology. It was Cooley who first took seriously the truly indeterminate character of the self as a structure independent of others. This led to the idea that the self developed in the process of interaction with others. As Mead said, to Cooley the “other” belonged in the same field as the self and was just as immediately given (Mead 1930).

Cooley, however, despite some arresting insights about the “looking-glass” features of the self, adhered to a semi-idealistic, subjectivist concept of “mind” which, though no longer individualistic in the Cartesian or the utilitarian sense, achieved only a truncated conception of interaction systems.

Mead, however, took the essential step of treating the individual as being both subject and object at the same time and in the same interpersonal system. Furthermore, he was far clearer than Cooley in showing that the personality of the individual emerges from the process we now call socialization precisely through the interplay between these subjective and objective aspects. This is not (as the idealists would have it) a process of the “unfolding” of the mind, individual or collective, but vitally includes the internalization of objects.

Mead also contributed a most important conception in his idea of the generalized other. Through symbolic interaction the individual learns to use and develop generalized codes that can interrelate a conception of the particular other with generalized categories and collectivities. This is the foundation, in the process of socialization, of the internalization of cultural, as well as social, systems, which in turn can come to be differentiated from each other. Mead took much longer steps than Cooley toward opening sociology and social psychology to the substitution of more technical research procedures for reliance on interpretative insight alone.

This was social psychology, in that it demonstrated and analyzed the intimacy of the relation between personality and social system by showing (in a way related to but different from Freud’s) that the personality cannot be understood independently of its articulation, including its genetic involvement, with social interaction. Moreover, particularly in Mead, who was a kind of symbolic behaviorist, there was an even fuller awareness than in Freud of the evolutionary continuity of phenomena between the human levels of action and those of organic life more generally. Mead also surpassed Freud in beginning the exploration of truly symbolic processes and in building a bridge between behavior theory and linguistics. There was certainly impressive convergence between this version of “symbolic-interaction” theory and the ideas of both Freud and Durkheim on the internalization of social objects.

The existentialist tradition

German, or, more broadly, continental European, thought has for a considerable period involved a strand of thinking which has recently gained prominence in analyzing interaction in a manner playing into a basic, if still relatively unclear, synthesis with the American tradition of social psychology. Perhaps it is most conveniently traced back to the Kantian phase in the development of idealism. Kant’s sphere of “practical reason” was specifically unstructured, in any sense comparable to that of either the phenomenal or the physical world. It was the world of will, individuality, and what may be considered socially unorganized meanings. It is perhaps particularly important here that this tradition experienced great difficulty in making the crucial discrimination between the biological-environmental and the cultural-cosmological reference poles of action systems. It has tended to merge the two in speaking of the “deeper” human needs of motivations.

In the early phase, perhaps Schelling was the interpreter of the Kantian tradition who veered furthest in this direction. Later, various more or less definitely “existentialist” orientations, but particularly those of Kierkegaard and Nietzsche, seem to be most prominent. A rough continuum among three distinct emphases evidently characterizes the more modern and more sociologically oriented phase.

Weber is at one end, self-consciously analyzing the interdependence of “intended meanings” and situation, interaction and sanctions, and being less concerned with the “fate” of subjective fantasies and hopes themselves than with the nature of the interaction systems generated by the complex modes of their implementation (and the failure thereof).

Perhaps Georg Simmel, who has had a very important impact on American social science, can be placed in the middle. Simmel attempted to confine “phenomenal” determinacy to the “forms” of interaction, and he devoted his immense intuitive talents to the interpretative understanding of the meanings individual actors and types of action injected into interaction, thereby creating, in a broad and rather loose sense, the determinate framework of such interaction. Significantly, Simmel’s influence came into American social science via the University of Chicago, which was also the home of Mead. Simmel’s “forms of social relationship” were not explanatory categories so much as a frame of reference for interpretive essays (see Naegele 1958).

The third movement, that furthest from Weber, is grounded philosophically in the phenomeno-logical tradition of Husserl and, in part, Heidegger and has affiliations with existentialism. Less immediately, it is also certainly linked with the post-Hegelian historicism that was Weber’s major critical foil. It focuses primarily on the most intimate experiences and feelings of the individual and from that vantage point mounts a relativizing criticism upon the more conventionalized and supposedly “superficial” structures of social life. At least certain elements in the movement stress individualism to the borders of a philosophical anarchy, running strongly counter to traditional sociological emphases on the grounding of basic order in social life.

Perhaps the most prominent writer in American sociology today strongly influenced by both a Meadian symbolic interactionism and a phenomenological viewpoint is Erving Goffman. Goffman’s most distinctive line of thought is a stress on the discrepancies between the self-image which the actor presents to others in the interactive process and his underlying private attitudes and preoccupations (Goffman 1956).

Empirical approaches

Since World War II there has been an important movement in the United States in favor of the empirical study of interaction, especially in small-group research. Three aspects of this may be mentioned. The so-called “group-dynamics” tradition was founded by the social psychologist Kurt Lewin (1939-1947; Lewin et al. 1944; Benne et al. 1950). It has been relatively eclectic, with an emphasis on the malleability of human goals through interaction. This environmentalism has been associated with a strong “action” orientation, i.e., toward changing behavior in desired directions through group participation experience.

The second movement has been the experimental and laboratory study of interaction in small groups by Bales and his associates (cf. Bales 1950). This group has concentrated on technical observational and analytical methods and theoretically has strongly emphasized the concept of social system at this level, as well as that of larger systems; in this respect it may be said to be in the theory-of-action tradition.

The third type of work is that of Homans, which took its departure from the study of informal organization in industry by Elton Mayo and his associates (cf. Homans 1950). In his latest work, Homans (1961) has, broadly in what above has been called a utilitarian framework, attempted to derive the main features of what he terms “elementary social behavior” from experimental psychology of Skinner’s type and from certain postulates of the theory of economic exchange.

Interaction and the social syste

The remainder of this article will attempt a systematic outline of both the components of an interaction system and some major aspects of the interaction process itself. I shall try to show that the principal emphases of the historical theories outlined above figure somewhere in a more generalized and theoretically comprehensive scheme. This more comprehensive scheme is a “theory of action,” in the sense long used by the author and also, in substance, by very many others, although their terminology may differ.

The concept of interaction is the first-order step beyond the action concept itself toward formulating the concept of social system. In speaking of action, we assume meaningful motivations and goal directedness. Motives, goals, and the like are expressed in, and hence must be interpreted as embodying, cultural-level symbolic form. There are infrastructures of all action systems which are not symbolically structured, but there is no system of action, in the present sense, which does not involve cultural symbolization—pre-eminently, of course, through language.

The concept of a dyadic interaction is convenient for clarifying certain fundamentals of interaction phenomena generally. But since it is a limiting case, general inferences from it should be made with care. This is true in the same sense that, although the unicellular organism is convenient for studying certain fundamentals of all organic life, alone it can scarcely provide adequate evidence for a theory of organic evolution.

The crucial reference points for analyzing interaction are two: (1) that each actor is both acting agent and object of orientation both to himself and to the others; and (2) that, as acting agent, he orients to himself and to others and, as object, has meaning to himself and to others, in all of the primary modes or aspects. The actor is knower and object of cognition, utilizer of instrumental means and himself a means, emotionally attached to others and an object of attachment, evaluator and object of evalution, interpreter of symbols and himself a symbol.

From these premises derives the fundamental proposition of the double contingency of interaction. Not only, as for isolated behaving units, animal or human, is a goal outcome contingent on successful cognition and manipulation of environmental objects by the actors, but since the most important objects involved in interaction act too, it is also contingent on their action or intervention in the course of events. The theory of games is perhaps the most sophisticated analysis of the implications of such double contingency. Of course, the contingency factor multiplies with each addition to the number of interacting units (for my own earlier formulations, see Parsons & Shils 1951).

Double contingency and more-complex contingencies have a crucial set of consequences. On the one hand, as analyzed pre-eminently by Hobbes, in an interaction system the possibilities of instability far exceed those to which isolated actors are exposed in relation to environments containing only nonactors, e.g., physical objects, as the significant objects. On the other hand, if the autonomy possessed by each acting unit relative to its environment is integrated with that of the others with which it interacts, the interaction system as a whole can gain vastly in autonomy, or freedom of action. Moreover, under certain conditions this enhanced autonomy of the system can be shared by the units within the system. In this situation a unit within the “organized” interaction system has greater freedom to act autonomously than does a unit which has the same capacities but which is isolated in relation to its physical environment, in the manner of Robinson Crusoe. This is the analytical basis of the “institutionalized individualism” which Durkheim so clearly demonstrated to be a product of the division of labor, in his sense.

Conditions of integration

The most important single condition of the integration of an interaction system is a shared basis of normative order. Because it must operate to control the disruptive potentialities (for the system of reference) of the autonomy of units, as well as to guide autonomous action into channels which, through mutual reinforcement, enhance the potential for autonomy of both the system as a whole and its member units, such a basis of order must be normative. It must guide action by establishing some distinctions between desirable and undesirable lines of action which can serve to stabilize interaction in these fundamental senses. Whether the stabilized system is “static” or “dynamically changing” in one or more of many senses is another issue. The theory of games can be said to have proved that a complex interaction system with no rules, but in which each unit is supposed only to be “rationally pursuing its self-interest,” cannot be stable in the above sense. This is a critical point for understanding the place of “rationality” in social behavior.

The concept of a shared basis of normative order is basically the same as that of a common culture or a “symbolic system.” The prototype of such an order is language. A language involves a code, consisting of the generalized norms which define “correct” speech or writing, as the basis for using symbols to formulate and transmit messages. Although there is considerable minor deviation, the massive fact is that all speakers of a language “observe” the norms of the code—“conform” to them, if you will—on penalty of not being understood.

Language, to be sure, is not a primary normative constituent of social systems in the sense true of the law in complex systems, but it is a primary normative constituent of distinct cultural systems. However, the point I wish to make here is that all culture is a matter of normative control, or the “guidance” of action. This is one sense in which the dyad is clearly a limiting case of interaction. However isolated a dyad may be in other respects, it can never generate the ramified common culture which makes meaningful and stable interaction possible. A dyad always presupposes a culture shared in a wider system. Furthermore, such a culture is always the product of a “historical” process long transcending the duration of a particular dyadic relationship.

As Durkheim made clear, for actors in interaction this common normative culture has a double significance. On the one hand, for each actor it constitutes a primary part of the situation or environment of his action. Its existence and the ways in which it guides the actions of system members are social facts of which the actor must take account. These facts include the probabilities of the imposition of sanctions contingent on action relative to the norms: rewards for conformity and negative sanctions for nonconformity. On the other hand, the normative culture becomes, in the paradigmatic cases, internalized in the personalities of individual actors—and institutionalized in collectivities—and thereby comes to control action, in part, by moral authority. To this extent conformity is voluntary, and hence internal sanctions come into play.

The phenomenon that cultural norms are internalized to personalities and institutionalized in collectivities is a case of the interpenetration of subsystems of action, in this case social system, cultural system, and personality. Since these subsystems are defined analytically, not concretely, it is understandable that the concrete boundary of any one subsystem may include within it spheres or zones which require an especially intimate integration with part of one or more other subsystems. Here the critical proposition is that institutionalized normative culture is an essential part of all stable systems of social interaction. Therefore, the social system and the culture must be integrated in specific ways in the area of their interpenetration.

The dyadic paradigm of interaction also constitutes a special limiting case in regard to the way in which an interaction system constitutes a collectivity. This point deserves special emphasis. Treatment of the dyad as the typical rather than the limiting case tends to perpetuate the utilitarian view of interaction and to underplay both solidarity and the role of normative culture in favor of the “wants” of individuals—or any other version of individual “interests.” Any given dyadic relation, as well as any given “individual,” should be seen in the context of a wider social system interpenetrating with a broader, shared culture.

Role pluralism and personality

Relative stability of a significant level of integration over time implies both a common normative culture and rather definite criteria of membership status. Members share with each other a level of solidarity not applying, in the relevant collectivity reference, to the relations between members and nonmembers. Solidarity involves some special quality and level of presumptive mutual trust and loyalty to the collective interest, on occasion involving the sacrifice of unit interests. In principle a collectivity is capable of “action in concert,” in the sense of taking collective action toward goals defined in social process as those of the collectivity and resisting centrifugal forces which might reduce the collective involvements of member actors to pure selfinterest. Indeed, the possibility of such action provides the primary basis for the boundaries between a social system and its environment, consisting of other social systems or other types of systems.

Dyadic interaction systems may constitute collectivities and be solidary in significant degree, but they are always subsystems of more-extensive social systems. One reason for this is the necessity of a common culture; thus, for example, the interaction possible without a common language is very limited indeed. A second reason involves the relation of the interaction system to the personalities of its members. A dyad, as a matter of empirical fact, never constitutes an independent society; a member of a dyad never interacts only with the particular role partner of that dyad. Hence, his whole personality, so far as it is engaged in social interaction, is never engaged only in a single dyadic interaction. Thus, although marriage is a particularly important dyadic relation, it is typical in all known societies that married couples have children and that the role of spouse is differentiated from that of parent; moreover, the nuclear family always constitutes a more inclusive collectivity in which each member plays plural roles.

The phenomenon of role pluralism is a central feature of all human societies, and this is more important the more highly differentiated the society. Therefore, the interactive spheres of different individuals, although overlapping and interpenetrating, are not identical. Any given individual participates in a considerable number of specific interaction systems, the more important and enduring of which are the stable collectivities in which he is a member. Thus, the unit of collectivity membership is not the individual in general but the person in role.

Two consequences follow from this. First, parallel to the interpenetration between social and cultural systems noted above, there must be interpenetration between social systems and personalities. Concretely, just as normative culture is internalized in personalities as well as institutionalized, there must be institutionalized expectations about the particular role in the particular collectivity which are also internalized in the personalities of incumbents. Typically, of course, internalized expectations of reciprocity shade in varying degrees into alienation and propensities for deviance.

Second, however, the specifications of normative culture to the different collectivities in which the individual participates and the expectations about behavior in the individual’s various roles must be integrated with each other at the level of the personality. One-to-one matching between the specific structures of particular personalities and the behavioral requirements of socially organized roles is precluded by, together, the pluralistic differentiation of subcollectivities in the social system and the plural role participations of individual persons. The sociological reason for this, which combines with genetic, psychological, and other kinds of reasons, is that no two persons have the same combination of role involvements—a circumstance greatly accentuated in societies where a substantial proportion of role involvements are nonascriptive —and, hence, role involvements are entered as a matter of, in some sense, voluntary choice. (Positing such a correspondence between the bases of social system and of personality integration has been a major fallacy in many theories of “culture and personality” and “national character.”)

Here we encounter again Durkheim’s analysis of the double relations between actor and normative culture. From the perspective of the social system the personalities of its participating members are at the same time, in different respects, both part of the social system, through interpenetration, and part of its environment. The zone of interpenetration is that of the expectations about role performance, since they are both institutionalized in the social system and internalized in individual personalities. Here it is particularly important that where the roles of role partners are differentiated, expectations are, not for identical, but for different yet complementary, performances. For instance, husband and wife, in their differentiated roles in the family, are generally expected to act, not alike, but differently, each having a distinct proper sex role. The differences between roles, as well as their common solidarity, are legitimized by the values shared between them.

Organism and environment

The personality of the individual, as an analytically defined action system, is one major parameter, linked to the living organism, the two being, in our terminology, ascribed to each other. They must nevertheless be distinguished analytically because the structure and mechanisms of the organism are physical, while those of the personality are psychocultural and learned. As with the personality and social system, there can be no one-to-one correspondence between the properties of an organism and the personality’s internalized content of normative culture and social role expectations. In certain contexts, this is very well known; thus, no expert contends that, in any but a “programmed” sense, there is specific anatomical or physiological structure distinguishing the speakers of a given language from those of another.

The organism is the link between action systems and the physical world. All concrete action is, in one aspect, the “behavior of organisms,” but only in one aspect. Thus, all linguistic communication involves the speech organs, the auditory apparatus, and the brain (or alternative mechanisms, as in the case of writing and reading). Since organisms are always located in particular places at given times, all social systems have their ecological aspect, i.e., there is location, movement, and distribution of organisms and activities in space. Clearly, an individual’s own body and the bodies of others are crucial objects of orientation to him in a wide variety of ways.

It seems to follow that the organism should be included in the physical environment of action systems, and hence of social systems. In the light of our traditions of thought, the physical environment is clearly the least problematic of the environments of interaction systems. However, the old difficulties over the sense in which the individual as a whole (including his organic aspects) should be included in the concept of social system can be resolved with the same logic that has been used in relating social systems to the cultural and psychological systems of action.

First, there is a category of objects which are only physical, whether they be “natural” objects or artifacts. These cannot and do not interact in human social systems, animals being a marginal case. In this sense, human organisms not only are physical but also interpenetrate with the other action systems. They are environmental objects and also, through interpenetration, parts of the action (and interaction) system.

This dual relation to interaction, however, does not apply equally to all aspects of the human organism. The concept “behavioral organism” designates the components of the organism for which interpenetration with personality, social system, and culture is most important. Some (for instance, H. A. Murray) have used the concept “vegetative organism” to designate aspects, such as most of the metabolic processes and mechanisms, that are minimally involved in action. One should not, however, assume that the line is empirically fixed; action phenomena may shift (through stress or psychotherapy, for instance) to involve rather directly organic processes that ordinarily are insulated from them.

Certainly, the involvement of the organism in interaction comprises all the modes of orientation and modalities of objects. The organism is perhaps particularly important as an instrumentality, but Freud’s concept of primary narcissism rightly considers the child’s love of his own body an authentic case of love. Similarly, there is an organic aspect interpenetrating with the nonorganic aspects in every subsystem of the orientation of actors. There has been considerable research, for example, on the organic “bases” of the emotions, starting with the well-known work of Cannon (1915).

Interaction as process

We may conclude with a brief outline of interaction as process. First, we presume that whatever the intermediate stages in the course of evolution from simple animal behavior to human social interaction, the latter is couched primarily at symboliccultural levels, although it certainly has various “subcultural” underpinnings. The action process, then, can be analyzed into two phases: what happens within the acting unit (e.g., a person in role or a collectivity) and what happens between such units. It seems to be generally acceptable terminology to refer to the former process as “decision” and to the latter as “communication.”

In decision processes, information communicated to the deciding unit (this is the interaction case, but environmental information may also be relevant) is “processed” in the light of the “dispositions,” goals, sentiments, etc., of the unit. An act is then performed, which typically consists of a communication to other units in the system. Whether the communication is verbal or not is an open question, as it may consist, for example, in a gesture of the sort Mead analyzed so clearly. This communication then becomes an input to the receiving units, including the promulgator, who may be, in terms of a stock phrase for this type of situation, “appalled by what he just said.”

Every output of communication involves crossing a boundary, as does its receipt as input. Its meaning must be interpreted and introduced into a combinatorial process, along with other inputs and with aspects of the internal structures and processes of the unit, whether a personality or a collectivity. This interpretive and combinatorial process constitutes “decision,” from which emerges new communicative output.

The output must also undergo a process that involves an indefinite number of stages before the communication reaches the target unit, units, or categories of units. In a variety of ways this process involves media of communication, which expose the communication to a variety of influences, such as modifications or distortions or maintenance of its “message” by special measures. Such influences are, of course, the outcome of decisions made by units through which the communication passes.

Generalized media of interaction

Of the many aspects of the communication process in interaction, one may be singled out for special comment, namely, the role of generalized media. I have already mentioned language a number of times as the prototypical medium. At the cultural level it is clearly the fundamental matrix of the whole system of media. Large-scale social systems, however, contain more specialized media (if you will, specialized “languages”), such as money, power, and influence (see Parsons 1963a; 1963b). Such media, like language, control behavior in the processes of interaction. They do so, however, by symbolic means, i.e., by presenting the actor, not with an intrinsically important object, such as a food object, but with a symbolic “representation” of such an object. Symbols can arouse the expectation that a meal will be served; hence they prepare the communication’s recipient for the experience of foodgratification and, within important limits, even substitute for the experience. The working of money in this regard is the best-understood example of a social system medium. It has, as the classical economists put it, no “value in use,” but only “value in exchange.” Possession of money symbolically concretizes expectations of access to gratifying objects of utility, but money is not itself such an object.

There are various other such media in human interaction. Freud’s “erotic pleasure” certainly constitutes one, as do the phenomena referred to by such terms as “affect” and “social acceptance,” and what W. I. Thomas called the “wishes” for response and recognition. The demonstration that such media are deeply needed by persons at a psychological level is excellent evidence of the phenomena of internalization discussed above and, more generally, of interpenetration.

Talcott Parsons

[Directly related are the entriesGroups, article onThe Study Of Groups; Integration; Norms; Systems Analysis. Other relevant material may be found inLanguage, article onLanguage and culture; Personality; Self Concept; Socialization; Utilitarianism; and in the biographies ofCannon; Cooley; Descartes; Durkheim; Freud; Hegel; Hobbes; Husserl; James; Kant; Lewin; Locke; Marshall; Marx; Mayo; Mead; Peirce; Simmel; Thomas; Weber, Max.]


Bales, Robert F. 1950 Interaction Process Analysis: A Method for the Study of Small Groups. Reading, Mass.: Addison-Wesley.

Benne, Kenneth D.; Bradford, Leland I.; and Lippitt, Ronald 1950 Social Action. New York: B’nai B’rith Anti-Defamation League.

Cannon, Walter B. (1915) 1953 Bodily Changes in Pain, Hunger, Fear and Rage: An Account of Recent Researches Into the Function of Emotional Excitement. 2d ed. Boston: Branford.

Cooley, Charles H. (1902) 1956 Human Nature and the Social Order. Rev. ed. In Charles H. Cooley, Two Major Works: Social Organization and Human Nature and the Social Order. Glencoe, III.: Free Press. → Each title reprinted with individual title page and pagination. Separate paperback editions were published in 1964 by Schocken.

Durkheim, Émile (1893) 1960 The Division of Labor in Society. Glencoe, III.: Fūree Press. → First published as De la division du travail social.

Durkheim, Émile (1895) 1958 The Rules of Sociological Method. 8th ed. Edited by George E. G. Catlin. Glencoe, III.: Free Press. → First published as Les règies de la méthode sociologique.

Durkheim, Emile (1912) 1954 The Elementary Forms of the Religious Life. London: Allen & Unwin; New York: Macmillan. → First published as Les formes e’le’mentaires de la vie religieuse, le systéme totemique en Australie.

Goffman, Erving (1956) 1959 The Presentation of Self in Everyday Life. Garden City, N.Y.: Doubleday.

Homans, George C. 1950 The Human Group. New York: Harcourt.

Homans, George C. 1961 Social Behavior: Its Elementary Forms. New York: Harcourt.

Lewin, Kurt (1939-1947) 1963 Field Theory in Social Science: Selected Theoretical Papers. Edited by Dorwin Cartwright. London: Tavistock.

Lewin, Kurt et al. 1944 Authority and Frustration. Iowa City: Univ. of Iowa Press.

Mead, George H. (1930) 1964 Cooley’s Contribution to American Social Thought. Pages 293-307 in George H. Mead, George Herbert Mead on Social Psychology. Rev. ed. Edited by Anselm Strauss. Univ. of Chicago Press.

Mead, George H. (1934) 1963 Mind, Self and Society From the Standpoint of a Social Behaviorist. Edited by Charles W. Morris. Univ. of Chicago Press. → Published posthumously.

Mead, George H. 1938 The Philosophy of the Act. Univ. of Chicago Press. → This volume consists almost entirely of papers unpublished during Mead’s lifetime.

Naegele, Kasper D. 1958 Attachment and Alienation: Complementary Aspects of the Work of Durkheim and Simmel. American Journal of Sociology 63:580-589.

Parsons, Talcott (1958) 1964 Social Structure and the Development of Personality: Freud’s Contribution to the Integration of Psychology and Sociology. Pages 78-111 in Talcott Parsons, Social Structure and Personality. New York: Free Press.

Parsons, Talcott 1963a On the Concept of Influence. Public Opinion Quarterly 27:37-62. → A comment by J. S. Coleman appears on pages 63-82; a communication by R. A. Bauer appears on pages 83-86; and a rejoinder by Talcott Parsons appears on pages 87-92.

Parsons, Talcott 1963b On the Concept of Political Power. American Philosophical Society, Proceedings 107:232-262.

Parsons, Talcott 1965 Evaluation and Objectivity in Social Science: An Interpretation of Weber’s Contribution. International Social Science Journal 17:46-63.

Parsons, Talcott; and Shils, Edward A. (editors) 1951 Toward a General Theory of Action. Cambridge, Mass.: Harvard Univ. Press. → A paperback edition was published in 1962 by Harper.

Simmel, Georg (1902-1917) 1950 The Sociology of Georg Simmel. Edited and translated by Kurt H. Wolff. Glencoe, III.: Free Press.

Simmel, Georg (1908) 1955 Conflict; The Web of Group Affiliations. Glencoe, III.: Free Press. → These essays appeared originally as “Der Streit” and “Die Kreuzung sozialer Kreise” in Georg Simmel’s Soziologie.

Simmel, GeorgGeorg Simmel, 1858-1918: A Collection of Essays. Edited by Kurt H. Wolff. Columbus: Ohio State Univ. Press, 1959.

Tiryakian, Edward A. 1962 Sociologism and Existentialism: Two Perspectives on the Individual and Society. Englewood Cliffs, N.J.: Prentice-Hall.

Weber, Max (1922) 1957 The Theory of Social and Economic Organization. Edited by Talcott Parsons. Glencoe, III.: Free Press. → First published as Part 1 of Wirtschaft und Gesellschaft.


The term “symbolic interaction” refers to the process by which individuals relate to their own minds or the minds of others; the process, that is, in which individuals take account of their own or their fellows’ motives, needs, desires, means and ends, knowledge, and the like. This process was first labeled symbolic interaction by Blumer (1937). Among sociologists it is often called social interaction.

The more distinctive problems of modern social psychology concern symbolic interaction—its rise, general characteristics, persistence, or dissolution. The more distinctive problems of modern sociology concern the occurrence, maintenance, or dissolution of special forms of symbolic interaction. Crowds, communities, friendships, economies, concerts, and cotillions are instances of such forms.

In the definition of “symbolic interaction,” the word “mind” is employed in the sense first clarified by C. S. Peirce, William James, John Dewey, and G. H. Mead (Morris 1932). It denotes instrumental activities that animals direct toward their environments. These instrumental activities, sometimes referred to as action or as psychological activities, relate the organism’s requirements to the conditions and resources in the environment that are relevant for meeting those requirements. What usually are considered to be the units or aspects of purely psychological events—for example, attitudes, beliefs, motivation, perception, thought, or choice—are here interpreted as ingredients or aspects of instrumental activity. In his capacity as a minded organism, the individual is called an “actor.” Groups or other collectivities can also be conceived of as actors, to the extent that they make decisions and relate to their own instrumental processes and to those of other collectivities.

One additional property distinguishes mind from other organic functions, such as respiration, ingestion, or excretion, that mediate between an animal and its outer environment. An activity is instrumental only if the probability of its appearance is affected by the relevance to the organism’s needs of that activity’s prior occurrence in similar situations. The term “learning” refers to changes in such probabilities. Situations are similar when they resemble one another in the requirements made by the organism and in the conditions afforded by the environment for meeting those requirements.

When individuals take account of one another’s minds, they observe, and adapt to, the existence of these instrumental processes as such. This means that they take into account something of the specifically instrumental character of one another’s behavior. It would be possible to perceive simply the physical patterns embodied in another man’s grimace or stride, but some perception of the place of those patterns in an instrumental process is involved in judging the grimace to be friendly or the stride to be hurried or determined.

The conception of symbolic interaction does not require that the individuals concerned reflect upon what they do. Mice and men learn to distinguish among the properties of their environment—among colors, sounds, and distances—and in this sense they are conscious, or aware, of those properties. They may not, however, be aware of their awareness; they may not know what they know. Similarly, it is conceivable, and probably common, for a man to be aware of his own mind or the minds of others without the existence of this knowledge becoming the object of his attention. Indeed, an individual often cannot say just why he has the impressions of others that he has or reaches the conclusions he does. It is proper to say that men aware of their own minds are “self-conscious,” but some other term, perhaps “reflective self-consciousness,” is needed to indicate the further step of knowing that one is self-conscious. That further step, or the even more complex stage of knowing that one knows what one knows, is a possible but not necessary development of symbolic interaction.

Emergence of the concept Symbolic interaction was first put forward as a distinct and important type of relationship in order to interpret some of men’s oldest observations. Since ancient times it has been thought that men’s behavior and experience differ from that of other animals, whether in degree or in kind. Specifically, men are often capable of rational activity—of identifying objectives in the environment and locating means by which those objectives may be attained. In the course of rational activity, men typically exercise control over their own behavior, exhibit reflective self-consciousness, and display an orientation to the past and future, as well as to the present. They experience themselves as deciding to act as they do. They employ some manner of signs or symbols, and they apply some normative standards in guiding their own behavior or that of others. They relate to the world in terms of its significance for their developing activities. In addition, sequences of human behavior are sometimes directed toward objectives which can be attained only in the fardistant future. On occasion, as in efforts toward achieving moral perfection, the objective may be conceived of as unobtainable yet men may feel obligated to approximate its attainment as best they can. Taken together, these several abilities especially conspicuous in man are said to constitute his “human nature.”

A major step in explaining human nature was taken in the seventeenth and eighteenth centuries, when several British philosophers, and especially John Locke and David Hume, proposed that it consisted, not of innate abilities, but of skills acquired in the course of men’s interacting with their fellows. Speculations about the exact process by which human nature emerged in interaction continued in Europe and the United States throughout the nineteenth century (Hughes 1958). It was, however, not until the two decades after World War I that a sociologist, Charles H. Cooley (1909), and a philosopher, George Herbert Mead (1934), provided a detailed and explicit account of the origins of human nature as aspects of the rise of symbolic interaction.

Mead and other writers who shared his outlook paid particular attention to views of behavior or of organized social life that precluded an interpretation of human nature’s origin in symbolic interaction. In Mead’s time instinctivism and behaviorism were the most conspicuous views of behavior that were obviously incompatible with Mead’s position. Instinctivism rooted human nature in biological inheritance. The behaviorism of Mead’s day sought, in physicalistic descriptions of the environment, properties which could be thought to determine conduct, quite apart from the relevance of those properties for the individual’s particular needs or his present lines of behavior. Mass, velocity, and extent are instances of these properties. These behavioristic accounts often accorded full ontological status only to such properties. Believing that behaviorism was inadequate to explain the rise of human nature, Mead and other writers sought to endow symbolic interaction with equal ontological status.

Writers who shared Mead’s position also objected to certain conceptions of organized social life, particularly the cultural determinism prevalent in the literature of their time. This determinism conceived of men as so completely immured within society, so thoroughly shaped as mere parts of society, that the role of novelty, fluidity, and change in social relations was completely obscured. By contrast, interpretations of social life as symbolic interaction conceived of actors as constantly establishing and re-establishing their mutual relations, modifying or abandoning them as the occasion demanded. Thus, social life was viewed as a process by which actors collectively solved problems, the nature and persistence of their solutions varying with the problems they defined.

Varieties of interaction

Symbolic or social interaction is conceived of as developing from prior relations that men have in their capacities as physical bodies, as organisms, and as actors. When people affect one another as physical bodies, we can speak of physical interaction. Two men hurrying around a corner from opposite directions, colliding, and falling to the ground provide a case in point. When individuals affect one another in their capacity as living organisms, we speak of ecological or biological interaction. In this capacity each individual, by using resources relevant for sustaining life and by other wise modifying those resources, affects the life of the others. When men acting as minded individuals, that is, as actors, behave toward their fellows as they might toward any other objects in the environment, we may speak of “behavioral” interaction. No requirement is then made that the actors take the minds of others into account.

Current notions and explanations of symbolic interaction depend on the assumption that men influence each other as physical bodies and that they cluster together and affect each other because of their status as living organisms. Explanations of symbolic interaction rest even more explicitly on the pre-existence of behavioral interaction, that is, on assumptions like the following: that minds and accompanying gestures already exist for men to observe; that patterned dependencies among actors exist ready to be discriminated; and that individuals come to identify the existence and significance of each other’s minds in the course of efforts to increase the amount and dependability of resources available to them. In short, both empirical observations and current theories require our treating behavioral interaction as a necessary condition for the rise and persistence of symbolic interaction.

We find symbolic interaction not only emerging over the long course of human evolution but also appearing, and then either persisting or dissolving, in men’s daily encounters. The principal occasions on which one may witness its ebb and flow are those when previously unrelated individuals establish contacts with one another’s minds or when existing patterns of symbolic interaction are subjected to strain. On such occasions appear events like those recorded in Table 1. The last three columns of Table 1 concern events which might be labeled symbolic interaction. The first column clearly refers to behavioral interaction.

The classification in Table 1 reflects the incompleteness and inaccuracy of our knowledge. Its categories are gross, recording little of the detailed understanding now available on each broad topic. However, the table does display key topics and suggests the essential unity among them.

The columns in Table 1 represent types of interaction presently known to occur, each having at least the characteristics that are minimal for behavioral interaction. Reading from left to right, we find that each column adds some characteristic to that minimum. In column IV we reach elements of what we are calling symbolic or social interaction. These elements are fully present in column V. Column VI treats a subclass of symbolic interaction. The type of relation represented in each column is assumed to be among the prerequisites for relations specified in columns further to the right. The reciprocal action described in any column can regress to types of interaction further to the left.

Row A in Table 1 records perceptions that individuals have of others and themselves. These perceptions concern characteristics of individuals as sources of influence. Row A1 and row A2 record coordinations required when actors relate to the conditions in row A. These coordinations involve some modification of each participant’s behavior. There are no reasonably standardized labels for rows A1 and A2 of the first three columns. This is so because the very nature of the processes is poorly understood and is the subject of controversy. Some investigators would be satisfied were “self-stimulation” inserted in row Al of column I and “interpersonal stimulation” in row A2 of the same column. Others would find this quite inappropriate. Concepts appropriate for columns II and III are even less certain.

On the other hand, there is considerable agreement

Table 1 - Some varieties of interaction
 Social stimulation (I)Circular reaction (II)Conversation of gestures (III)Interpretive interaction (IV)Symbolic interaction (V)Communication (VI)
A. Each actor acts toward himself and others as loci of:stimulicuesgesturessignsselvesidentity
1. Directed toward his own behavior, this action is: a. It results in some degree of:   insight self-control  
2. Directed toward others’ behavior, this action is: a. It results in some degree of:   empathy social influence  
B. Actors so engaged may yet have difficulties in coordinating their relations as:objectsactorscopartici pants in several types of actcoparticipants in a division of labor (i.e., a role system)coparticipants in several role systems 
1. The difficulties may be resolved if they pay attention to and routinize their use of the wider context provided by:cuesgesturessignsselvesidentity 

regarding the concepts recorded for rows A1 and A2 in column IV. Indeed, the same concepts are often employed to designate the contents of these rows in columns V and VI as well. Occasionally, more specific terms are used. Thus, “search for identity” might be inserted in row A1 of column VI, “role taking” in row A1 of column V, and “role playing” in row A2 of that column. However, firmer empirical knowledge about the processes involved must precede agreement on appropriate concepts and labels.

As row B indicates, difficulties may arise when actors try to coordinate their efforts. When that happens, actors have a greater chance of success if they know something of the conditions under which the behavior of their fellows occurs. Thus, they are likely to deal more effectively with one another as sources of stimuli if they are forewarned of those stimuli. Cues are stimuli which, an actor discovers, precede the appearance of other stimuli. Column by column, row B1 indicates progressively broader contexts of conditions, which, if perceived, make more dependable and effective an actor’s dealings with conditions specified in row A.

We now may notice that the figure’s successive columns are linked, in that the broader context developed in row B1 of any column (e.g., an understanding of cues in order to cope with related stimuli) becomes itself the focus of attention in the column immediately to the right. It is essential to offer a brief justification for saying that the conditions given in row B1 are a context for and an aid to coping with those given in row A.

The relation between cues and stimuli has already been mentioned. What of that between gestures and cues? A gesture is an overt behavior that occurs in the early stages of a particular act. It can serve, therefore, as a cue to later stages. It differs from other cues in being relevant to instrumental activity as the special kind of stimulus with which one seeks to cope. A sign is an overt behavior that occurs in the early stages of a class of acts. It can serve, therefore, as a cue to a mind, but it differs from other gestures by referring to qualities common to many acts instead of characteristics peculiar to any one of them. “Self” refers to signs representing classes of acts in which a given actor engages. These classes of acts are identified as his. Each of an actor’s selves provides a context for understanding and forecasting his more specific and immediate activities. “Actor’s identity” refers to signs representing the more persistent and general categories of action that he employs in all situations. Knowledge of his identity enables a more adequate interpretation and forecasting of the selves he will exhibit in particular situations.

Row B refers to events that are of great psychological and sociological importance but that are not well understood. In columns I to V of that row are specified relationships between participants in some joint action. The establishment and maintenance of each variety of joint action is a problem to participants. Our present information is too slight for any confident description of the problems appropriate to any single column. One can say with confidence only that the succession of problems from left to right across Table 1 should involve participants in increasingly complex relations with one another.

Social scientists still need to learn many things concerning each of these relationships. For example, little is known about the steps by which infants and young children come to identify each type of problem and to participate with others in its solution. Moreover, we have still to identify the steps that socialized actors must take to maintain each type of relationship once it is established. We have yet to discover more than the rudiments of the “grammar” of cues, gestures, signs, selves, and identities. We know but little of the means by which actors make valid interpretations of those phenomena. We presently are unable to specify in any detail the potentialities and limitations for collective undertakings of each variety of interaction.

Relations between types of interaction Although detailed knowledge of the different types of interaction has been slow to accumulate, many of the general processes by which one type of interaction changes into another have long been familiar to social scientists. Four of these processes—elementary collective behavior, socialization, institutionalization, and social control—can be at least partly described in terms of the relationships outlined in Table 1. Thus, elementary collective behavior is a change from the types of interaction in column V or VI to those in columns I to IV and back again. Socialization is a process in which an actor is trained to engage in any or all types of symbolic interaction or communication. Similarly, institutionalization takes place to some degree in any situation in which actors jointly define some relationships as legitimate and as necessary for their continued interaction. Finally, social control clearly involves change from one type of interaction to another, since it is a process in which actors encourage others to engage in, or prevent others from engaging in, some relationship because it meets or violates an institutionalized standard. Three of these four processes probably occur in most movements between columns. Thus, there probably are rudiments of socialization and of institutionalization in any move across these columns from left to right. Similarly, there are at least some rudiments of social control being exercised in all shifts, whether to the right or left. Collective behavior differs from the other three in referring to a sequence of changes in type of interaction from certain columns to others, in particular from column V or VI to columns I to IV and back again.

Sequences of stages in socialization like those proposed by Mead (1934), Freud and his followers (Fenichel 1945), or Newcomb (1961) distinguish successive steps in the movement from behavioral to symbolic interaction. Thus, Mead offers progressive discriminations in a cognitive sequence. From solo play to participation in simple games to relations with a generalized other, his sequence marks the steps toward interaction that is more fully symbolic. These steps progress from relations with a single other person to relations with several other persons and, finally, to adaptations to more-abstract social structures. In Freud’s work we find stages of growing dependence on others as minded individuals, beginning with a recognition of one’s dependence on their actions and ending with the commitment to support actively the social norms that facilitate, govern, and rationalize continued social interaction with them. In Newcomb’s studies appear the steps through which experienced social participants establish intimate contacts with one another, from (a) first impressions of one another’s overt behavior to (b) a classification of each other in terms of such common social roles as age, sex, and social class, thence to (c) a mutual reconnaissance of ideologies (e.g., those connected with politics or religion) that guide men’s relations as comembers of the larger society, and, finally, to (d) reciprocal investigations of more idiosyncratic attitudes and dispositions relevant to the subjects’ possible relations as friends and intimates.

Many of these developmental schemes are guided by the image of a complete or perfected social situation. In such perfected situations, we are told, the full range of skills that distinguish symbolic interaction are employed. Cooley’s vision of the primary group (1909) belongs here, as does the psychoanalysts’ picture of mature love (Abraham 1925; Fromm 1947). In each case, experience with the perfected social situation is advanced as the necessary condition for training and reinforcing individuals in symbolic interaction. Each writer reminds us that fully symbolic interaction is rarely achieved, always tends to revert to behavioral or biological interaction, and must be reachieved over and over again. Indeed, it is the effort to achieve some essential stability of symbolic interaction that forces men to act in support of social relations—to distinguish and legitimize their joint actions, in contrast with activities that support only their personal interests. This is the origin of social systems.

Guy E. Swanson

[Directly related are the entriesCommunication; Groups, article onThe study of groups; Learning theory; Semantics and semiotics. Other rele-vant material may be found inCollective behavior; Language; Social control; Social institutions; Socialization; and in the biographies ofCooley; Dewey; Freud; Hume; James; Locke; Mead; Peirce.]


Abraham, Karl (1925) 1953 Character-formation on the Genital Level of Libido-development. Pages 407-417 in Karl Abraham, Selected Papers. Volume 1: Selected Papers on Psychoanalysis. New York: Basic Books. → First published in German in Volume 7 of the International Journal of Psycho-analysis.

Blumer, Herbert 1937 Social Psychology. Pages 144-198 in Emerson P. Schmidt (editor), Man and Society. New York: Prentice-Hall.

Cooley, Charles H. (1909) 1956 Social Organization: A Study of the Larger Mind. In Charles H. Cooley, Two Major Works: Social Organization and Human Nature and the Social Order. Glencoe, III.: Free Press. → Each title reprinted with individual title page and pagination. Separate paperback editions were published in 1964 by Schocken.

Fenichel, Otto 1945 The Psychoanalytic Theory of Neurosis. New York: Norton. → See especially pages 463-540.

Fromm, Erich 1947 Man For Himself: An Inquiry Into the Psychology of Ethics. New York: Holt.

Hughes, H. Stuart 1958 Consciousness and Society: The Reorientation of European Social Thought, 1890—1930. New York: Knopf.

Mead, George H. (1934) 1963 Mind, Self and Society From the Standpoint of a Social Behaviorist. Edited by Charles W. Morris. Univ. of Chicago Press. → Published posthumously.

Morris, Charles W. 1932 Six Theories of Mind. Univ. of Chicago Press.

Newcomb, Theodore M. 1961 The Acquaintance Process. New York: Holt.


Dramatism is a method of analysis and a corresponding critique of terminology designed to show that the most direct route to the study of human relations and human motives is via a methodical inquiry into cycles or clusters of terms and their functions.

The dramatistic approach is implicit in the key term “act.” “Act” is thus a terministic center from which many related considerations can be shown to “radiate,” as though it were a “god-term” from which a whole universe of terms is derived. The dramatistic study of language comes to a focus in a philosophy of language (and of “symbolicity” in genera]); the latter provides the basis for a general conception of man and of human relations. The present article will consider primarily the dramatistic concern with the resources, limitations, and paradoxes of terminology, particularly in connection with the imputing of motives.

The dramatistic approach to action

Dramatism centers in observations of this sort: for there to be an act, there must be an agent. Similarly, there must be a scene in which the agent acts. To act in a scene, the agent must employ some means, or agency. And it can be called an act in the full sense of the term only if it involves a purpose (that is, if a support happens to give way and one falls, such motion on the agent’s part is not an act, but an accident). These five terms (act, scene, agent, agency, purpose) have been labeled the dramatistic pentad; the aim of calling attention to them in this way is to show how the functions which they designate operate in the imputing of motives (Burke [1945-1950] 1962, Introduction). The pattern is incipiently a hexad when viewed in connection with the different but complementary analysis of attitude (as an ambiguous term for incipient action) undertaken by George Herbert Mead (1938) and by I. A. Richards (1959).

Later we shall consider the question whether the key terms of dramatism are literal or metaphorical. In the meantime, other important things about the terms themselves should be noted.

Obviously, for instance, the concept of scene can be widened or narrowed (conceived of in terms of varying “scope” or circumference). Thus, an agent’s behavior (“act”) might be thought of as taking place against a polytheistic background; or the over-all scene may be thought of as grounded in one god; or the circumference of the situation can be narrowed to naturalistic limits, as in Darwinism; or it can be localized in such terms as “Western civilization,” “Elizabethanism,” “capitalism,” “D day,” “10 Downing Street,” “on this train ride,” and so on, endlessly. Any change of the circumference in terms of which an act is viewed implies a corresponding change in one’s view of the quality of the act’s motivation. Such a loose yet compelling correspondence between act and scene is called a “scene-act ratio” (Burke [1945-1950] 1962, pp. 1-7).

All the terms are capable of similar relationships. A “purpose-agency ratio,” for instance, would concern the logic of “means selecting,” the relation of means to ends (as the Supreme Court might decide that an emergency measure is constitutional because it was taken in an emergency situation). An “agent-act ratio” would reflect the correspondence between a man’s character and the character of his behavior (as, in a drama, the principles of formal consistency require that each member of the dramatis personae act in character, though such correspondences in art can have a perfection not often found in life). In actual practice, such ratios are used sometimes to explain an act and sometimes to justify it (ibid., pp. 15-20). Such correlations are not strict, but analogical. Thus, by “scene-act ratio” is meant a proposition such as: Though agent and act are necessarily different in many of their attributes, some notable element of one is implicitly or analogously present in the other.

David Hume’s An Inquiry Concerning Human Understanding (first published in 1748) throws a serviceable light upon the dramatistic “ratios.” His treatise begins with the observation that “moral philosophy, or the science of human nature, may be treated after two different manners.” One of these “considers man chiefly as born for action.” The other would “consider man in the light of a reasonable rather than an active being, and endeavor to form his understanding more than cultivate his manners” ([1748] 1952, p. 451). Here, in essence, is the distinction between a dramatistic approach in terms of action and an approach in terms of knowledge. For, as a “reasonable being,” Hume says, man “receives from science” his proper food and nourishment. But man “is a sociable, no less than a reasonable being…. Man is also an active being; and from that disposition, as well as from the various necessities of human life, must submit to business and occupation” (ibid., p. 452).

Insofar as men’s actions are to be interpreted in terms of the circumstances in which they are acting, their behavior would fall under the heading of a “scene-act ratio.” But insofar as their acts reveal their different characters, their behavior would fall under the heading of an “agent-act ratio.” For instance, in a time of great crisis, such as a shipwreck, the conduct of all persons involved in that crisis could be expected to manifest in some way the motivating influence of the crisis. Yet, within such a “scene-act ratio” there would be a range of “agent-act ratios,” insofar as one man was “proved” to be cowardly, another bold, another resourceful, and so on.

Talcott Parsons, in one of his earlier works, has analytically unfolded, for sociological purposes, much the same set of terministic functions that is here being called dramatistic (owing to their nature as implied in the idea of an “act”). Thus, in dealing with “the unit of action systems,” Parsons writes:

An “act” involves logically the following: (1) It implies an agent, an “actor.” (2) For purposes of definition the act must have an “end,” a future state of affairs toward which the process of action is oriented. (3) It must be initiated in a “situation” of which the trends of development differ in one or more important respects from the state of affairs to which the action is oriented, the end. This situation is in turn analyz-able into two elements: those over which the actor has no control, that is which he cannot alter, or prevent from being altered, in conformity with his end, and those over which he has such control. The former may be termed the “conditions” of action, the latter the “means.” Finally (4) there is inherent in the conception of this unit, in its analytical uses, a certain mode of relationship between these elements. That is, in the choice of alternative means to the end, in so far as the situation allows alternatives, there is a “normative orientation” of actions. (1937, p. 44)

Aristotle, from whom Aquinas got his definition of God as “pure act,” gives us much the same lineup when enumerating the circumstances about which we may be ignorant, with corresponding inability to act voluntarily:

A man may be ignorant, then, of who he is, what he is doing, what or whom he is acting on, and sometimes also what (e.g. what instrument) he is doing it with, and to what end (e.g. he may think his act will conduce to some one’s safety), and how he is doing it (e.g. whether gently or violently). (Nichomachean Ethics lllla5)

This pattern became fixed in the medieval questions : quis ( agent), quid ( act), ubi (scene defined as place), quibus auxiliis (agency), cur (purpose), quo modo (manner, “attitude”), quando (scene defined temporally).

The nature of symbolic action

Within the practically limitless range of scenes (or motivating situations) in terms of which human action can be defined and studied, there is one over-all dramatistic distinction as regards the widening or narrowing of circumference. This is the distinction between “action” and “sheer motion.” “Action,” is a term for the kind of behavior possible to a typically symbol-using animal (such as man) in contrast with the extrasymbolic or non-symbolic operations of nature.

Whatever terministic paradoxes we may encounter en route (and the dramatistic view of terminology leads one to expect them on the grounds that language is primarily a species of action, or expression of attitudes, rather than an instrument of definition), there is the self-evident distinction between symbol and symbolized (in the sense that the word “tree” is categorically distinguishable from the thing tree). Whatever may be the ultimate confusions that result from man’s intrinsic involvement with “symbolicity” as a necessary part of his nature, one can at least begin with this sufficiently clear distinction between a “thing” and its name.

The distinction is generalized in dramatism as one between “sheer motion” and “action.” It involves an empirical shift of circumference in the sense that although man’s ability to speak depends upon the existence of speechless nature, the existence of speechless nature does not depend upon man’s ability to speak. The relation between these two distinct terministic realms can be summed up in three propositions:

(1)There can be no action without motion—that is, even the “symbolic action” of pure thought requires corresponding motions of the brain.

(2)There can be motion without action. (For instance, the motions of the tides, of sunlight, of growth and decay.)

(3)Action is not reducible to terms of motion. For instance, the “essence” or “meaning” of a sentence is not reducible to its sheer physical existence as sounds in the air or marks on the page, although material motions of some sort are necessary for the production, transmission, and reception of the sentence. As has been said by Talcott Parsons:

Certainly the situation of action includes parts of what is called in common-sense terms the physical environment and the biological organism …these elements of the situation of action are capable of analysis in terms of the physical and biological sciences, and the phenomena in question are subject to analysis in terms of the units in use in those sciences. Thus a bridge may, with perfect truth, be said to consist of atoms of iron, a small amount of carbon, etc., and their constituent electrons, protons, neutrons and the like. Must the student of action, then, become a physicist, chemist, biologist in order to understand his subject? In a sense this is true, but for purposes of the theory of action it is not necessary or desirable to carry such analyses as far as science in general is capable of doing. A limit is set by the frame of reference with which the student of action is working. That is, he is interested in phenomena with an aspect not reducible to action terms only in so far as they impinge on the schema of action in a relevant way—in the role of conditions or means…. For the purposes of the theory of action the smallest conceivable concrete unit is the unit act, and while it is in turn analyzable into the elements to which reference has been made—end, means, conditions and guiding norms—further analysis of the phenomena of which these are in turn aspects is relevant to the theory of action only in so far as the units arrived at can be referred to as constituting such elements of a unit act or a system of them. (1937, pp. 47-48)

Is dramatism merely metaphorical?

Although such prototypically dramatistic usages as “all the world’s a stage” are clearly metaphors, the situation looks quite otherwise when approached from another point of view. For instance, a physical scientist’s relation to the materials involved in the study of motion differs in quality from his relation to his colleagues. He would never think of “petitioning” the objects of his experiment or “arguing with them,” as he would with persons whom he asks to collaborate with him or to judge the results of his experiment. Implicit in these two relations is the distinction between the sheer motion of things and the actions of persons.

In this sense, man is defined literally as an animal characterized by his special aptitude for “symbolic action,” which is itself a literal term. And from there on, drama is employed, not as a metaphor but as a fixed form that helps us discover what the implications of the terms “act” and “person” really are. Once we choose a generalized term for what people do, it is certainly as literal to say that “people act” as it is to say that they “but move like mere things.”

Dramatism and the social system

Strictly speaking, then, dramatism is a theory of terminology. In this respect a nomenclature could be called dramatistic only if it were specifically designed to talk, at one remove, about the cycle of terms implicit in the idea of an act. But in a wider sense any study of human relations in terms of “action” could to that extent be called dramatistic. A major difficulty in delimiting the field of reference derives from the fact that common-sense vocabularies of motives are spontaneously personalistic, hence innately given to drama-laden terms. And the turn from the naĩve to the speculative is marked by such “action words” as tao, karma, dike, hodos, islam (to designate a submissive attitude), all of which are clearly dramatistic when contrasted with the terminological ideals proper to the natural sciences (Burke [1945-1950] 1962, p. 15).

The dramatistic nature of the Bible is proclaimed in the verb (bara) of the opening sentence that designates God’s creative act; and the series of fiats that follows identifies such action with the principle of symbolicity (“the Word”). Both Plato’s philosophy of the Good as ultimate motive and Aristotle’s potentiality-actuality pair would obviously belong here, as would the strategic accountancy of active and passive in Spinoza’s Ethics (Burke [1945-1950] 1962, pp. 146-152). The modern sociological concern with “values” as motives does not differ in principle from Aristotle’s list of persuasive “topics” in his Rhetoric, One need not look very closely at Lucretius’ atomism to discern the personality in those willful particles. Contemporary theories of role-taking would obviously fall within this looser usage, as indicated on its face by the term itself. Rhetorical studies of political exhortation meet the same test, as do typical news reports of people’s actions, predicaments, and expressions. Most historiography would be similarly classed, insofar as its modes of systematization and generalization can be called a scientifically documented species of storytelling. And humanistic criticism (of either ethical or aesthetic sorts) usually embodies, in the broad sense, a dramatistic attitude toward questions of personality. Shifts in the locus and scope of a terminology’s circumference allow for countless subdivisions, ranging from words like “transaction,” “exchange,” “competition,” and “cooperation,” or the maneuvers studied in the obviously dramalike situations of game theories, down to the endless individual verbs designed to narrate specifically what some one person did, or said, or thought at some one time. Thus Duncan (1962) has explicitly applied a dramatistic nomenclature to hierarchy and the sociology of comedy. Similarly, Goffman (1956) has characterized his study of “impression management” as “dramaturgical."

Does dramatism have a scientific use

If the dramatistic nature of terms for human motives is made obvious in Burke’s pentad (act, scene, agent, agency, purpose), is this element radically eliminated if we but introduce a synonym for each of those terms? Have we, for instance, effectively dodged the dramatistic “logic” if instead of “act” we say “response,” instead of “scene” we say “situation” or “stimulus,” instead of “agent” we say “subject” or “the specimen under observation in this case,” instead of “agency” we say “implementation,” and instead of “purpose” we use some term like “target"? Or to what extent has reduction wholly taken place when the dramatistic grammar of “active,” “passive,” and “reflexive” gets for its analogues, in the realm of sheer motion, “effectors,” “receptors” (output, input), and “feedback,” respectively? Might we have here but a truncated terminology of action, rather than a terminology intrinsically nondramatistic? Such issues are not resolved by a dramatistic perspective; but they are systematically brought up for consideration.

A dramatistic analysis of nomenclature can make clear the paradoxical ways in which even systematically generated “theories of action” can culminate in kinds of observation best described by analogy with mechanistic models. The resultant of many disparate acts cannot itself be considered an act in the same purposive sense that characterizes each one of such acts (just as the movement of the stock market in its totality is not “personal” in the sense of the myriad decisions made by each of the variously minded traders). Thus, a systematic analysis of interactions among a society of agents whose individual acts variously reinforce and counter one another may best be carried out in terms of concepts of “equilibrium” and “disequilibrium” borrowed from the terminology of mechanics.

In this regard it should also be noted that although equilibrium theories are usually interpreted as intrinsically adapted only to an upholding of the status quo, according to the dramatistic perspective this need not be the case. A work such as Albert Mathiez’s The French Revolution (1922-1927) could be viewed as the expression of an anima naturaliter dramatistica in that it traces step by step an ironic development whereby a succession of unintentionally wrong moves led to unwanted results. If one viewed this whole disorderly sequence as itself a species of order, then each of the stages in its advance could be interpreted as if “designed” to stabilize, in constantly changing circumstances, the underlying pattern of conditions favorable to the eventual outcome (namely, the kind of equilibrium that could be maintained only by a series of progressive developments leading into, through, and beyond the Terror).

Though a drama is a mode of symbolic action so designed that an audience might be induced to “act symbolically” in sympathy with it, insofar as the drama serves this function it may be studied as a “perfect mechanism” composed of parts moving in mutual adjustment to one another like clockwork. The paradox is not unlike that which happened in metaphysics when a mystical view of the world as a manifestation of God’s purposes prepared the way for mechanistic views, since the perfect representation of such a “design” seemed to be a machine in perfect order.

This brings up the further consideration that mechanical models might best be analyzed, not as downright antidramatistic, but as fragments of the dramatistic. For whatever humanist critics might say about the “dehumanizing” effects of the machine, it is a characteristically human invention, conceived by the perfecting of some human aptitudes and the elimination of others (thus in effect being not inhuman, but man’s powerful “caricature” of himself—a kind of mighty homun-culus).

If, on the other hand, it is held that a dramatistic nomenclature is to be avoided in any form as categorically inappropriate to a science of social relations, then a systematic study of symbolic action could at least be of use in helping to reveal any hitherto undetected traces of dramatistic thinking that might still survive. For otherwise the old Adam of human symbolicity, whereby man still persists in thinking of himself as a personal agent capable of acting, may lurk in a symbol system undetected (a tendency revealed in the fact that the distinction between “action” and “sheer motion” so readily gets lost, as with a term like kinesis in Aristotle or the shift between the mechanistic connotations of “equilibrium” and the histrionic connotations of “equilibrist”). Similarly, since pragmatist terminologies lay great stress upon “agencies” (means) and since all machines have a kind of built-in purpose, any nomenclature conceived along the lines of pragmatist instrumentalism offers a halfway house between teleology and sheer aimless motion.

At one point dramatism as a critique of terminology is necessarily at odds with dramatism as applied for specifically scientific purposes. This has been made clear in an article by Wrong (1961), who charges that although “modern sociology after all originated as a protest against the partial views of man contained in such doctrines as utilitarianism, classical economics, social Darwinism, and vulgar Marxism,” it risks contributing to “the creation of yet another reified abstraction in socialized man, the status-seeker of our contemporary sociologists” (p. 190). He grants that “such an image of man is …valuable for limited purposes,” but only “so long as it is not taken for the whole truth” (p. 190). He offers various corrections, among them a stress upon “role-playing,” and upon “forces in man that are resistant to socialization,” such as certain “biological” and “psychological” factors—even though some sociologists might promptly see “the specter of ’ biological determinism’” (p. 191) and others might complain that already there is “too much ’ psychologism’ in contemporary sociology” (p. 192).

Viewed from the standpoint of dramatism as a critique of terminology, Wrong’s article suggests two notable problems. Insofar as any science has a nomenclature especially adapted to its particular field of study, the extension of its special terms to provide a definition of man in general would necessarily oversociologize, overbiologize, overpsychologize, or overphysicize, etc., its subject; or the definition would have to be corrected by the addition of elements from other specialized nomenclatures (thereby producing a kind of amalgam that would lie outside the strict methodic confines of any specialized scientific discipline). A dramatistic view of this situation suggests that an over-all definition of man would be not strictly “scientific,” but philosophical.

Similarly, the dramatistic concept of a scene-act ratio aims to admonish against an overly positivistic view of descriptive terms, or “empirical data,” as regards an account of the conditions that men are thought to confront at a given time in history. For insofar as such a grammatical function does figure in our thoughts about motives and purpose, in the choice and scope of the terms that are used for characterizing a given situation dramatism would discern implicit corresponding attitudes and programs of action. If the principle of the scene-act ratio always figures in some form, it follows that one could not possibly select descriptive terms in which policies of some sort are not more or less clearly inherent. In the selection of terms for describing a scene, one automatically prescribes the range of acts that will seem reasonable, implicit, or necessary in that situation.

Dramatistic analyses of order

Following a lead from Bergson (1907, especially chapter 4), dramatism is devoted to a stress upon the all-importance of the negative as a specifically linguistic invention. But whereas Bergson’s fertile chapter on “the idea of nothing” centers in the propositional negative (“It is not”), the dramatistic emphasis focuses attention upon the “moralistic” or “hortatory” negative (“Thou shalt not”). Burke (1961, pp. 183-196) has applied this principle of negativity to a cycle of terms implicit in the idea of “order,” in keeping with the fact that “order,” being a polar term, implies a corresponding idea of “disorder,” while these terms in turn involve ideas of “obedience” or “disobedience” to the “authority” implicit in “order” (with further terministic radiations, such as the attitude of “humility” that leads to the act of obedience or the attitude of “pride” that leads to the act of disobedience, these in turn involving ideas of guidance or temptation, reward or punishment, and so on).

On the side of order, or control, there are the variants of faith and reason (faith to the extent that one accepts a given command, proscription, or statement as authoritative; reason to the extent that one’s acceptance is contingent upon such proofs as are established by a methodic weighing of doubts and rebuttals). On the side of disorder there are the temptations of the senses and the imagination. The senses can function as temptations to the extent that the prescribed order does not wholly gratify our impulses (whether they are natural or a by-product of the very order that requires their control). Similarly, the imagination falls on the side of disorder insofar as it encourages interests inimical to the given order, though it is serviceable to order if used as a deterrent by picturing the risks of disorder—or, in other words, if it is kept “under the control of reason.”

Midway between the two slopes of order and disorder (technically the realm where one can say yes or no to a thou-shalt-not) there is an area of indeterminacy often called the will. Ontologically, action is treated as a function of the will. But logo-logically the situation is reversed: the idea of the will is viewed as derivable from the idea of an act.

From ideas of the will there follow in turn ideas of grace, or an intrinsic ability to make proper choices (though such an aptitude can be impaired by various factors), and sacrifice (insofar as any choices involve the “mortification” of some desires). The dramatistic perspective thus rounds out the pattern in accordance with the notion that insofar as a given order involves sacrifices of some sort, the sacrificial principle is intrinsic to the nature of order. Hence, since substitution is a prime resource available to symbol systems, the sacrificial principle comes to ultimate fulfillment in vicarious sacrifice, which is variously rationalized, and can be viewed accordingly as a way to some kind of ultimate rewards.

By tracing and analyzing such terms, a dramatistic analysis shows how the negativistic principle of guilt implicit in the nature of order combines with the principles of thoroughness (or “perfection”) and substitution that are characteristic of symbol systems in such a way that the sacrificial principle of victimage (the “scapegoat”) is intrinsic to human congregation. The intricate line of exposition might be summed up thus: If order, then guilt; if guilt, then need for redemption; but any such “payment” is victimage. Or: If action, then drama; if drama, then conflict; if conflict, then victimage.

Adapting theology (“words about God”) to secular, empirical purposes (“words about words”), dramatistic analysis stresses the perennial vitality of the scapegoat principle, explaining why it fits so disastrously well into the “logologic” of man’s symbolic resources. It aims to show why, just as the two primary and sometimes conflicting functions of religion (solace and control) worked together in the doctrines of Christianity, we should expect to find their analogues in any society. Dramatism, as so conceived, asks not how the sacrificial motives revealed in the institutions of magic and religion might be eliminated in a scientific culture, but what new forms they take (Burke [1945-1950] 1962, pp. 406-408).

This view of vicarious victimage extends the range of those manifestations far beyond the areas ordinarily so labeled. Besides extreme instances like Hitlerite genocide, or the symbolic “cleansings” sought in wars, uprisings, and heated political campaigns, victimage would include psy-chogenic illness, social exclusiveness (the malaise of the “hierarchal psychosis”), “beatnik” art, rabid partisanship in sports, the excessive pollution of air and streams, the “bulldozer mentality” that rips into natural conditions without qualms, the many enterprises that keep men busy destroying in the name of progress or profit the ecological balance on which, in the last analysis, our eventual well-being depends, and so on.

The strongly terministic, or logological, emphasis of dramatism would view the scapegoat principle not primarily as a survival from earlier eras, but as a device natural to language here and now. Aristotle, in the third book of his Rhetoric (chapter 10), particularly stresses the stylistic importance of antithesis as a means of persuasion (as when a policy is recommended in terms of what it is against). In this spirit dramatism would look upon the scapegoat (or the principle of vicarious victimage) as but a special case of antithesis, combined with another major resource of symbol systems, namely, substitution.

In the polemics of politics, the use of the scapegoat to establish identification in terms of an enemy shared in common is also said to have the notable rhetorical advantage that the candidate who presents himself as a spokesman for “us” can prod his audience to consider local ills primarily in terms of alien figures viewed as the outstanding causes of those ills. In accord with this emphasis, when analyzing the rhetorical tactics of Mein Kampf, Burke (1922-1961) lays particular stress upon Hitler’s use of such deflections to provide a “noneconomic interpretation of economic ills.”

While recognizing the amenities of property and holding that “mine-ownness” or “our-ownness” in some form or other is an inevitable aspect of human congregation, dramatistic analysis also contends that property in any form sets the conditions for conflict (and hence culminates in some sort of victimage). It is pointed out that the recent great advances in the development of technological power require a corresponding extension in the realm of negativity (the “thou-shalt-nots” of control). Thus, the strikingly “positive” nature of such resources (as described in terms of “sheer motion”) is viewed dramatistically as deceptive; for they may seem too simply like “promises,” whereas in being powers they are properties, and all properties are problems, since powers are bones of contention (Burke 1960).

A dramatistic view of human motives thus culminates in the ironic admonition that perversions of the sacrificial principle (purgation by scapegoat, congregation by segregation) are the constant temptation of human societies, whose orders are built by a kind of animal exceptionally adept in the ways of symbolic action (Burke [1941] 1957, pp. 87-113).

Kenneth Burke

[See also Ethics, article onethical systems and social structures; Historiography, article Onthe rhetoric of history; Literature; Religion; Role; Semantics and semiotics; Systems analysis, article onsocial systems; and the biographies of Aristotle; Hume; Mead.]


Benne, Kenneth D. 1964 From Polarization to Paradox. Pages 216-247 in Leland P. Bradford, Jack R. Gibb, and Kenneth D. Benne (editors), T-Group Theory and Laboratory Method: Innovation in Re-education. New York: Wiley.

Bergson, Henri (1907) 1944 Creative Evolution. New York: Modern Library. → First published in French.

Burke, Kenneth (1922-1961)1964 Perspectives by Incongruity and Terms for Order. Edited by Stanley Edgar Hyman. Bloomington: Indiana Univ. Press. → Two representative collections of readings from Burke’s works. Each collection is also available separately in paperback from the same publisher.

Burke, Kenneth (1937) 1959 Attitudes Toward History. 2d ed., rev. Los Altos, Calif.: Hermes.

Burke, Kenneth (1941) 1957 The Philosophy of Literary Form: Studies in Symbolic Action. Rev. ed., abridged by the author. New York: Vintage. → The Louisiana State University Press reprinted the unabridged edition in 1967.

Burke, Kenneth (1945-1950) 1962 A Grammar of Motives and A Rhetoric of Motives. Cleveland: World.

Burke, Kenneth 1955 Linguistic Approach to Problems of Education. Pages 259-303 in National Society for the Study of Education, Committee on Modern Philosophies and Education, Modern Philosophies and Education. Edited by Nelson B. Henry. National Society for the Study of Education Yearbook 54, Part 1. Univ. of Chicago Press.

Burke, Kenneth 1960 Motion, Action, Words. Teachers College Record 62:244-249.

Burke, Kenneth 1961 The Rhetoric of Religion: Studies in Logology. Boston: Beacon.

Burke, Kenneth 1966 Language as Symbolic Action: Essays on Life, Literature, and Method. Berkeley: Univ. of California Press.

Duncan, Hugh D. 1962 Communication and Social Order. Totowa, N.J.: Bedminster Press.

Goffman, Erving (1956) 1959 The Presentation of Self in Everyday Life. Garden City, NY.: Doubleday.

Hume, David (1748) 1952 An Inquiry Concerning Human Understanding. Pages 451-509 in Great Books of the Western World. Volume 35: Locke, Berkeley, Hume. Chicago: Benton.

Mathiez, Albert (1922-1927) 1962 The French Revolution. New York: Russell. → First published in French in three volumes. A paperback edition was published in 1964 by Grosset and Dunlap.

Mead, George herbert 1938 The Philosophy of the Act. Univ. of Chicago Press. → Consists almost entirely of unpublished papers which Mead left at his death in 1931.

Parsons, Talcott 1937 The Structure of Social Action: A Study in Social Theory With Special Reference to a Group of Recent European Writers. New York: McGraw-Hill.

Richards, Ivor A. (1959) 1961 Principles of Literary Criticism. New York: Harcourt.

Rueckert, William H. 1963 Kenneth Burke and the Drama of Human Relations. Minneapolis: Univ. of Minnesota Press.

Wrong, Dennis H. 1961 The Oversocialized Conception of Man in Modern Sociology. American Sociological Review 26:183-193.


Most gratifications of human beings have their source in the actions of other human beings. To experience excitement in sexual pleasure or contentment in love, to enjoy intellectual stimulation or relaxing diversion, to achieve professional recognition or a happy family life, to satisfy the lust for power or the need for acceptance—to attain any of these ends requires that one induce others to behave in certain ways. The fact that many rewards men seek can be obtained only in social interaction is what underlies the conceptualization of interaction as social exchange.

Basic assumption

The basic assumptions of the theory of social exchange are that men enter into new social associations because they expect doing so to be rewarding and that they continue relations with old associates and expand their interaction with them because they actually find doing so to be rewarding. Associating with another person may be intrinsically rewarding, as in love and in sociability, or it may bring rewards that are extrinsic to the association itself, such as advice from a colleague and help from a neighbor.

In either case, the desire to satisfy some want is assumed to underlie the association. As Simmel (1908, p. 6) put it: “Social association refers to the widely varying forms that are generated as the diverse interests of individuals prompt them to develop social units in which they realize these interests, be they sensual or ideal, lasting or fleeting, conscious or unconscious, causally impelling or ideologically inducing.” To be sure, not all needs or interests are satisfied directly in social interaction, as hunger illustrates, and not all social interaction is primarily governed by an interest in rewards, since irrational forces and moral values also influence it. But many aspects of social life do reflect an interest in profiting from social interaction, and these are the focus of the theory of social exchange. Far from being confined to strictly rational conduct oriented toward material gain, however, the theory is intended to encompass all striving for rewarding social experiences, including the desire to further humanitarian ideals or spiritual values as well as the pursuit of personal advantage and emotional satisfaction.

The conception of social interaction as an exchange process follows logically from the assumption that men seek to obtain rewards in their social associations. If a man is attracted to others because he expects associating with them to be rewarding to himself, he will wish to associate with them in order to realize the anticipated rewards. Likewise, for them to engage in social interaction with him, they must also have an interest in doing so. But their interest in associating with him depends, according to the assumption, on their expectation that interacting with him will be rewarding to them. To implement his desire to associate with them, therefore, he must demonstrate to them that associating with him would benefit them. In brief, to reap the rewards expected from attractive potential associates, a man must impress them as a desirable associate by implicitly conveying the promise that social interaction with him will be rewarding for them too.

A person who derives benefits from associates is under obligation to reciprocate by supplying benefits to them in turn. People often go out of their way to do favors not only for friends but also for mere acquaintances and even for strangers, and they thereby create social obligations. The individual who fails to discharge his obligations and reciprocate in some form for benefits received robs others of incentives to continue to befriend him. Besides, such an individual is likely to be accused of ingratitude. This very accusation indicates that reciprocation for favors freely given is expected, and it serves as a social sanction to discourage men from forgetting their obligations. Gratitude, as Simmel ([1902-1917] 1950, p. 387) noted, “establishes the bond of interaction, of the reciprocity of service and return service, even when they are not guaranteed by external coercion.”

When obligations for benefits received are discharged by providing benefits in return, both parties profit from the association, and their exchange of rewarding experiences fortifies the social bond between them. A man who helps others earns their gratitude and appreciation, and he puts them into his debt, which promises to bring him further rewards in the future. These advantageous consequences of doing favors are undoubtedly a major reason why men frequently go to great trouble to help others and enjoy doing so. Giving is, indeed, more blessed than receiving, for having social credit is preferable to being socially indebted. To be sure, there are men who selflessly work for others without thought of reward and even without expecting gratitude, but these are virtually saints, and saints are rare. Other men also act unselfishly sometimes, but they require a more direct incentive for doing so, if it is only the social acknowledgment that they are unselfish. Such social approval is, of course, a very significant reward men seek in social interaction.

Defining social exchang

Exchange is not restricted to economic markets: social exchange is ubiquitous. Neighbors exchange help with chores; discussants, ideas; children, toys; friends, social support; politicians, concessions. The novice must meet the demands of the group to find acceptance in it. Colleagues exchange advice, and if the superior competence of one prevents the rest from reciprocating in kind for his advice, they discharge their obligation by paying their respect to his abilities and thus raising his status. Even the lover whose only apparent concern is to please his girl seeks to win her affection in return for his devotion. Groups and organized collectivities, too, are engaged in social exchange. For example, the medical profession receives exclusive license to practice medicine in return for assuming the obligation to meet the health needs of the community, or a political party makes concessions in its program to an interest group in return for support at the election booth.

Homans (1961, p. 13) developed the first systematic theory that focuses on social behavior “as an exchange of activity, tangible or intangible, and more or less rewarding or costly, between at least two persons.” Of special concern to Homans are the psychological processes that motivate men to engage in exchange, and the psychological reductionism of the theory has been criticized by other sociologists. Homans was, however, by no means the first to call attention to social exchange. Anthropologists had earlier discussed the significance and pervasiveness of the exchange of gifts and services in simpler societies. For instance, Mauss (1925) had presented a general analysis of gift exchange in such societies. But anthropologists were not the first to observe this phenomenon either.

Given the ubiquity of social exchange, it is perhaps not surprising that social philosophers have discussed it ever since antiquity. Aristotle’s Nicomachean Ethics (1162a34-1163a24) deals extensively with social exchange, which he distinguishes from economic exchange by saying that it “is not based on stated terms, but the gift or other service is given as to a friend, although the giver expects to receive an equivalent or greater return, as though it had not been a free gift but a loan.” Many writers of the intervening centuries, such as La Rochefoucauld (1664), Mandeville (1714), and Adam Smith (1759), have been intrigued by the exchange nexus observable in much of social life. More recently, a conception of exchange is implicit in Whyte’s discussion of the obligations of a gang leader (1943); it is explicit in Blau’s analysis of consultation in a group of government officials (1955); and it is an underlying element in Thibaut and Kelley’s theory of dyads and triads (1959).

The pervasiveness of social exchange makes it tempting to explore the fruitfulness of the concept by trying to apply it to all social conduct. But the concept of exchange loses its distinctive meaning and becomes tautological if all behavior in interpersonal relations is subsumed under it. Although much of social conduct is oriented toward expected returns from others—indeed, more than we usually think—not all of it is.

The concept of exchange can be delimited with the aid of some illustrations of why a man gives money to others. First, he may do so because they stand in front of him with guns in a holdup. While this could be viewed as an exchange of his money for his life, it seems preferable to exclude the results of physical coercion from the definition of the term “exchange.” Second, a man may donate money to charity because his conscience demands that he help the poor and without expecting gratitude in any form from them. While this could be viewed as an exchange of his money for the internal approval of his superego, here again it seems preferable to exclude conformity with internalized norms from what is meant by the term “exchange.” Third, an uncontrollable impulse may compel a man to squander his money; such behavior motivated by irrational drives does not entail any exchange either. Finally, a man may give alms to beggars because he enjoys receiving their expressions of deferential gratitude and discontinue giving them money if they fail to react with such expressions. This last case illustrates social exchange, whereas the others delineate the boundaries of the concept. In brief, the concept of exchange refers to voluntary social actions that are contingent on rewarding reactions from others and that cease when these expected reactions are not forthcoming.

Social and economic exchang

The very term “social exchange” is designed to indicate that social interaction outside the economic sphere has important similarities with economic transactions. Above all, the expectation that benefits rendered will yield returns characterizes not only economic transactions but also social ones in which gifts and services appear to be freely bestowed. Moreover, the economic principle of eventually declining marginal utility applies to social exchange as well. Advice from an expert colleague is worth much to a man who needs help with a problem, but once the problem has been clarified, additional counsel is no longer as valuable. No matter how much two friends enjoy one another’s company, after a certain amount of time together they will become less eager to continue their association. The more a man concentrates on obtaining a given social reward rather than others, the more the significance of the alternatives forgone will impinge upon his consciousness, making this reward relatively less significant. All of these examples manifest the marginal principle in social life.

There are, however, also important differences between social and strictly economic exchange. The most basic difference is that the obligations incurred in social transactions are not clearly specified in advance. In economic transactions the exact obligations of both parties are simultaneously agreed upon: a given product is sold for a certain price. Both commodities may change hands at the time the agreement is reached, or a contract is made that stipulates precisely the obligations either party has to discharge in the future. In social exchange, by contrast, one party supplies benefits to another, and although there is a general expectation of reciprocation, the exact nature of the return is left unspecified. Indeed, it must remain unspecified, since any attempt to specify it in advance destroys the social meaning of the transaction by transforming it into a merely economic one. Doing a favor has an entirely different social significance from making a bargain. If a man does a service for another and then indicates what the return for this service should be, he reveals that he does not want to consider the service a favor but prefers to make it part of a bargain; he thereby insists on keeping the relationship businesslike and refuses to enter a more sociable association. If the recipient immediately states what return he will make, he reveals the same disinclination to enter a sociable relation.

Social exchange, therefore, entails supplying benefits that create diffuse future obligations. The nature of the return is invariably not stipulated in advance, cannot be bargained about, and must be left to the discretion of the one who makes it. Thus, if a person gives a dinner party, he expects his guests to reciprocate in the future. But he can hardly bargain with them about the kind of party to which they should invite him, though he expects them not simply to ask him for a quick lunch if he has given a formal dinner for them. Generally, a man expects some expressions of gratitude and appreciation for favors he has done for others, but he can neither bargain with them over how to reciprocate nor force them to reciprocate at all. Any attempt to assure repayment for one’s generosity discloses that it was really not generosity in the first place. The distinctive significance of social obligations requires that they remain unspecific, and the fact that social, as distinguished from economic, commodities have no exact price facilitates meeting this requirement.

Since the recipient is the one who decides when and how to reciprocate for a favor, or whether to reciprocate at all, social exchange requires trusting others, whereas the immediate transfer of goods or the formal contract that can be enforced obviates such trust in economic exchange. Typically, however, social exchange relations evolve in a slow process, starting with minor transactions in which little trust is required because little risk is involved and in which both partners can prove their trustworthiness, enabling them to expand their relation and engage in major transactions. Thus, the process of social exchange leads to the trust required for it in a self-generating fashion. Indeed, creating trust seems to be a major function of social exchange, and special mechanisms exist that prolong the period of being under obligation and thereby strengthen bonds of indebtedness and trust. In the ceremonial gift exchange of the kula among the Trobriand Islanders, for example, returns for gifts received at one expedition can be made only at the next, many months later, and hasty reciprocation is generally condemned (Malinowski 1922, pp. 210-211 in 1961 edition). In our society, similarly, it is considered improper to reciprocate for a gift or return an invitation too quickly. The condemnation of posthaste reciprocation stimulates the growth of trust by constraining exchange partners to remain under obligation to each other for extended periods.

Social benefits are also less detachable from the source that supplies them than are economic commodities. At one extreme is the diffuse social support derived in a love relationship, the significance of which depends entirely on the person who supplies it. At the other extreme are such economic goods as shares in a corporation or money, the value of which is completely independent of the supplier. Most social benefits are intermediate between these extremes, having a value that is extrinsic to the exchange relations in which they are supplied but having this value modified by the significance of these relations. A man who consults a colleague is interested in good advice, whatever its source, but his personal relation with the consultant makes it more or less easy for him to ask for assistance and to understand the advice he receives. (Although in the economic sphere the services of the friendly corner grocer may be preferable to those of the impersonal supermarket, such personal relations generally encroach less on economic than on social exchange.)

Economic exchange may be considered a special case of the general phenomenon of exchange, with social exchange being the excluded residual category. When goods and services are given a price in terms of a single medium of exchange, economic transactions are institutionalized. Their price defines the value of commodities independent of any particular exchange relations, making this value separable from that of other benefits accruing in these relations, and it permits exact specification of the obligations incurred in economic transactions. Economic institutions, such as the impersonal market, are designed to exclude other considerations than price from exchange decisions. Many social benefits have no price, either because they are never traded on an economic market, as is the case with social support, or because they are not so traded in this instance, as exemplified by the advice from a friend in contrast with that from a professional consultant. These are the benefits that enter into social exchange, which means that their supply is not contingent on stipulated returns, though there is a general expectation of reciprocation. The fact that the return is at the discretion of the one who makes it gives social exchange its fundamental significance for developing bonds of trust and friendship, and mechanisms such as the social norms prohibiting bargaining and hasty reciprocation tend to protect this discretion. The most important benefits involved in social exchange, furthermore, do not have any material value on which a price can be put at all, as exemplified by social approval and respect.

Exchange and powe

A paradox of social exchange is that it serves not only to establish bonds of friendship between peers but also to create status differences between men. The kula exchange described by Malinowski (1922, p. 92 in 1961 edition), for instance, “provides every man …with a few friends near at hand, and with some friendly allies in the far-away, dangerous, foreign districts.” The potlatch of the Kwakiutl, on the other hand, is a system of giving away valuables in which “status in associations and clans, and rank of every kind, are determined by the war of property,” as Mauss ([1925] 1954, p. 35) noted. An important function of gift exchange in simple societies is, in the words of Levi-Strauss (1957, p. 85), “to surpass a rival in generosity, to crush him if possible under future obligations which it is hoped he cannot meet, thus taking from him privileges, titles, rank, authority, and prestige.” In modern society, too, supplying benefits to others serves sometimes as an expression of friendship for them and at other times as a means for establishing superiority over them.

A person who gives others valuable gifts or renders them important services makes an implicit claim to superior status by obligating them. A benefactor is not a peer but a superior on whom others depend. If they return benefits that adequately discharge their obligations, they deny his claim to superiority; and if their returns are excessive, they make a counterclaim to superiority over him. Continuing mutual exchange strengthens bonds between equals. But if they fail to reciprocate with benefits that are as important to him as his are to them, they validate his claim to superior status. In simple societies the resulting differentiation of status seems to be rooted in the institutionalized significance of one-sided benefactions, while in modern societies it is typically due to unilateral dependence on a supplier of needs.

The recurrent unilateral supply of important benefits is a basic source of power. A man with resources at his disposal that enable him to meet other men’s needs can attain power over them provided that four conditions are met, as suggested in a somewhat different formulation by Emerson (1962). First, they must not have resources that the benefactor needs, otherwise they can obtain what they want from him in direct exchange. Second, they must not be able to obtain the benefits he has to offer from an alternative source, which would make them independent of him. Third, they must be unable or unwilling to take what they want from him by force. Fourth, they must not undergo a change in values that enables them to do without the benefits they originally needed. If these four conditions are met, they have no choice but to comply with his wishes and submit to his power in order to obtain the needed benefits. The four alternatives to compliance are assumed to be exhaustive; in their absence, the supply of important services inevitably generates power.

Under specifiable conditions, then, exchange processes give rise to a differentiation of power. A man who commands services others cannot do without, who is independent of any services at their command, and whose services they can neither obtain elsewhere nor take from him by force can attain power over them by making the satisfaction of their needs contingent on their compliance with his directives. By acceding to his wishes, they reciprocate for the benefits he supplies. The exchange balance is restored as unilateral services are compensated by an imbalance of power. The man who recurrently supplies needed services to others makes them dependent on and obligated to him, and their accumulated obligations constrain them to comply with his wishes lest he cease to supply further services. Their indebtedness to him takes the form of a pool of willing compliance on which he can draw at his discretion whenever it is to his interest to impose his will upon them.

The compliance of men with another’s wishes and his consequent power, with which they repay him for services received, may appear to be no different from other social rewards that enter into exchange transactions. Yet there is a basic distinction between the differentiation of power and mutual social exchange, just as there is a basic distinction between social and economic exchange. The distinguishing criterion lies in the answer to the question, Who has discretion over the repayment? In economic exchange, neither party can exercise discretion over making the return, since the exact conditions of repayment are specified when the initial transaction takes place. In mutual social exchange, the nature and timing of the return are decided on by the one who makes it, that is, the recipient of the original benefit. In power relations, on the other hand, the return is made on the demand of the one to whom it is owed, that is, the supplier of the original benefit. Accumulated obligations and unilateral dependence transfer the power of discretion over the return from the debtor to the creditor and transform an exchange relation between peers into a power relation between superior and subordinate.

Secondary exchang

The study of complex social structures must take into account the social forces that emerge in them and that are not observable in face-to-face interaction. To be sure, the concept of exchange itself refers to emergent properties of social relations that cannot be reduced to the psychological processes that motivate individual behavior. Exchange theory is concerned with the interaction processes that emerge as individuals seek rewards in social relations, whatever the psychological forces that lead each to want certain rewards. The differentiation of power in a collectivity gives rise to still other social processes in the complex structure, and these may be conceived of as constituting a secondary exchange that becomes superimposed upon the primary one characteristic of interpersonal relations.

Power makes it possible to enforce demands, and these demands are judged by those subject to the power in terms of social norms of fairness. The fair exchange of power by a ruler or ruling group elicits social approval, whereas unfair demands that are experienced as exploitative or oppressive evoke social disapproval. Thus, a secondary exchange—of fairness in the exercise of power in return for social approval by subordinates —emerges as power becomes differentiated in a collectivity. The social forces set into motion by this secondary exchange lead to legitimation and organization, on the one hand, and to opposition and change, on the other.

Collective approval of power legitimates that power. If men profit from the way they are governed by those in power and consider the demands made on them to be fully justified by the advantages the leadership provides, common feelings of loyalty are likely to develop as they communicate to each other their appreciation of the leadership. Their joint obligations to the leadership tend to find expression in social norms that make compliance with its directives mandatory. The collectivity of subordinates repays those in power for the benefits derived from their leadership by enforcing the leaders’ directives as part of the enforcement of the collectivity’s own social norms, that is, by legitimating the leadership’s authority. For the distinctive characteristic of legitimate authority is that a superior’s commands are obeyed, not because of his sanctioning power but because of the normative pressure exercised among the subordinates themselves, particularly once these normative constraints have become institutionalized. Authority, in turn, promotes organization.

Collective disapproval of power engenders opposition. Men who share the feeling of being exploited and oppressed by the excessive demands of those in power are inclined to communicate their grievances to each other. A wish to retaliate by striking down the oppressors is often kindled in these discussions, in which men receive social support for their aggressive feelings. An opposition ideology may be adopted that further justifies and reinforces the hostility against existing powers. It is out of such shared discontent that opposition movements develop: for example, that men band together to organize a union against their employer or a radical party against their government. Such opposition is an important catalyst of basic social change.

A prime determinant of social conduct is the institutionalized system of values in a society: the distinctive values that define the in-group’s identity, the common standards of morality and of achievement, the values that legitimate governing authority and organization, and the ideologies that sometimes foster opposition to those in power. Guided by these values, men often set aside immediate self-interest and considerations of exchange; for instance, the professional’s standards may require him to help clients, disregarding the return he receives from them.

However, social values and norms largely set broad limits on conduct without prescribing it in detail. Within these limits, men are free to pursue their interest in social rewards, and considerations of exchange do apply. While social norms prohibit lying and cheating in order to get advice from another, they permit inducing him to give advice by expressing genuine respect or by any other means not specifically proscribed. Both common values and exchange principles influence social conduct, and neither may be neglected in studying it. Of particular importance in the analysis of social life is the modifying influence social values have on the rewards in which men are interested. Patriotic or opposition ideals often inspire men to make great material sacrifices, for these values make furthering the common cause more rewarding for them than material gain.

Exchange theory is most directly concerned with face-to-face relations, and thus it must be complemented by other theoretical principles that focus on complex structures with institutionalized values. However, even in the study of complex structures, exchange theory has something to contribute.

Peter M. Blau

[Directly related are the entriesConflict; Cooperation; Duty; Exchange and display; Friendship. Other relevant material may be found inCoalitions; Groups; Power; Sociometry; and in the biographies ofMaussandSimmel.]


Blau, Peter M. (1955) 1963 The Dynamics of Bureaucracy: A Study of Interpersonal Relations in Two Government Agencies. Rev. ed. Univ. of Chicago Press.

Blau, Peter M. 1964 Exchange and Power in Social Life. New York: Wiley.

Boulding, Kenneth E. 1962 Conflict and Defense: A General Theory. A publication of the Center for Research in Conflict Resolution at the University of Michigan. New York: Harper.

Emerson, Richard M. 1962 Power-Dependence Relations. American Sociological Review 27:31-41.

Gouldner, Alvin W. 1960 The Norm of Reciprocity: A Preliminary Statement. American Sociological Review 25:161-178.

Homans, George C. 1958 Social Behavior as Exchange. American Journal of Sociology 63:597-606.

Homans, George C. 1961 Social Behavior: Its Elementary Forms. New York: Harcourt.

La rochefoucauld, FranÇois (1664) 1940 The Maxims. Oxford Univ. Press. → First published as Réflexions … et maximes morales.

LÉvi-Strauss, Claude (1957) 1964 The Principle of Reciprocity. Pages 84-94 in Lewis A. Coser and Bernard Rosenberg (editors), Sociological Theory. New York: Macmillan.

Malinowski, Bronislaw (1922) 1960 Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea. London School of Economics and Political Science Studies, No. 65. London: Routledge; New York: Dutton. → A paperback edition was published in 1961 by Dutton.

Mandeville, Bernard (1714) 1957 The Fable of the Bees: Or Private Vices, Publick Benefits. Edited by F. B. Kaye. Oxford: Clarendon. → A reprint of the 1924 edition. The introduction and notes include a life, a bibliography, a critical and historical evaluation, and an annoted bibliography of secondary works.

Mauss, Marcel (1925) 1954 The Gift: Forms and Functions of Exchange in Archaic Societies. Glencoe, III.: Free Press. → First published as Essai sur le don: Forme et raison de l’échange dans les sociétés archaïques.

Simmel, Georg (1902-1917) 1950 The Sociology of Georg Simmel. Edited and translated by Kurt H. Wolff. Glencoe, III.: Free Press.

Simmel, Georg (1908) 1958 Soziologie. Berlin: Duncker & Humblot. → The translation of the extract in the text was provided by Peter M. Blau.

Smith, Adam (1759) 1892 The Theory of Moral Sentiments. London: Bell.

thibaut, John W.; and keixey, Harold H. 1959 The Social Psychology of Groups. New York: Wiley.

whyte, William F. (1943) 1961 Street Corner Society: The Social Structure of an Italian Slum. 2d ed., enl. Univ. of Chicago Press.


The functioning of modern society depends a great deal on the success of human encounters. Summit meetings, labor-management negotiations, biracial conferences, indeed the success of virtually every area of public life ultimately rests on the effectiveness of people working together. Private life is no less subject to this requirement. Marital crises, parent-child misunderstandings, and the appallingly high proportion of homicides committed by close relatives of the victim—all of these attest to the excruciating difficulty of achieving satisfying and productive relations between people.

Why does a civilization that copes so successfully with the physical world make so little progress toward understanding the factors that lead to successful human relationships? This inability has always been at the center of religious and philosophical discussions. Recently, social scientists have applied their experimental methods to this problem. Psychotherapists, especially Harry Stack Sullivan, have given increased attention to the problem through the recognition of the centrality of interpersonal relations in the development of personality. Some of the new trends in psychotherapy concentrate on the small group and the family as the therapeutic unit. These techniques are based on a recognition of the importance of interactional factors and on the assumption that an individual’s personality can be understood better by seeing him in interaction with others. [See Mental disorders, treatment of, article ongroup psychotherapy; and the biography of Sullivan.] In this essay the basic concern is with the relations between behavior at the social or group level and the personalities of the participants in this social interaction.

The problem

Before exploring the empirical work that has been done in the field, it will be useful to develop a more precise statement of the problem. Intuitively, it seems that it should be possible to predict, at least to some degree, the nature of social interaction from a knowledge of participating personalities. Experience suggests that two highly combative, unyielding people, for example, seem not to fare as well as a more flexible pair and that the effect of two personalities on each other is greater in the case of people who must live close together for a long period of time and whose decisions affect each other, such as men in a submerged submarine or persons in a space capsule going to the moon, than it is for two people who associate little and have stereotyped, structured relations with each other, such as a bus driver and a passenger or a theater ticket seller and a customer.

The outcome of an interaction requires not only a knowledge of the properties of the elements (personalities) but also of the laws that relate them (laws of human interaction). To understand social interaction more fully, consideration must also be given to influences other than personality. Factors such as social role and cultural norms affect social interaction. There is considerable variability, too, in the consistency of individual behavior. There are times when the same behaviors are elicited by virtually any other human being. Some severe neurotics, for example, will feel anxious or suspicious in the presence of any other person. These feelings exist regardless of roles, personalities, or setting. This type of reaction is found more often in pathological personalities, but it is not uncommon in the general population.

Sometimes identical behaviors are elicited by the external individual characteristics of others, such as their social role or ethnic group membership. For example, a person’s reaction may be the same to all authority figures, or all women, or all short people, or all unknown people, or all Negroes, or all popular people. Some external property of the observed individual causes the reaction regardless of the behavior of that individual. There also are consistent reactions toward social groupings, such as crowds, cocktail parties, or a group composed only of old men. The personality of the individual members of the social group does not influence the response.

Often people respond in the same way toward all persons who are domineering, or very affectionate, or withdrawn, or indecisive, or ingratiating, regardless of the remainder of the personality. These reactions are sometimes combined with the above category (“subgroups”) to elicit, for example, a reaction to indecisive women specifically or to aggressive Jews as opposed to aggressive people in general. This category differs from the above in that the person is reacting to personality attributes rather than to external characteristics.

Sometimes, too, people react to other people as unique individuals. This is a reality-oriented reaction and is associated with mental health. The reaction is to a person as he really is, recognizing both his similarities to and differences from other people and groups.

The role of personality factor

These considerations of social interaction make clear that the personalities of the group members account for only a part of the behavior observed in social interaction but a very significant part. There is evidence that as a social interaction increases in time, personality factors play a more important role in determining the course of the relationship. For a complete understanding of social interaction, situational and personality factors must be taken account of simultaneously. Predictions based only on situational factors seem doomed to a limited success and usually result in abstract statements, such as “people have a general tendency to conform to group norms.” Specification of the applicability of this generalization requires further work with situational factors, such as the types of decisions to which people conform; individual factors, such as the status differences among the conformers; and personality traits, such as rebellion and the need to be accepted.

Similarly, attempts to explain social interaction in terms of personality factors only also appear inadequate. The limited success of attempting to predict the outcome of a marriage based on the complementarity of needs attests to the necessity for considering situational factors as well as personality factors in order to understand social interaction.

Thus, the precise statement of the problem of this essay is the following:

Which characteristics of social interaction are predictableand to what degree and under what circumstancesfrom a knowledge of the personalities of the interacting individuals and of the laws governing the interaction of personalities?

Approaches. Investigators of the relation between personality and social interaction have used several approaches. At times they have studied the personality of one individual; at other times two or more; and, at still other times, the personalities of all of the members of a group. Sometimes, too, the entire group is studied as if it were a single, individual personality.

Combinations of personalities

The predictability of social phenomena from a knowledge of the individual personalities is directly tested by establishing a group and predicting the subsequent interaction of its members.

Means and variabilit

Approaches to studying the results of group composition may be distinguished by the various assumptions that are made about interaction by the investigators. Some of the earliest work assumed that social interaction was predictable from mean scores on a trait or from the variability of that trait among the group members. Results of the following type were obtained:

High group means on the personality characteristics of adventuresomeness, vigor, dominance, purposefulness, orderliness, willed application, and freedom from anxiety are associated with high performance on tasks requiring vigorous coordinated action and with a preference for such tasks over discussion.

High variances on personality measures of tough- versus tender-mindedness, “bohemian aggressiveness,” and paranoid suspiciousness are associated with the group’s dislike for a task requiring the resolving of opinion and attitude differences, with the group’s slowness in ranking attitude preferences, and with a feeling by its members that other members hinder progress (Cattell & Stice 1954).

Similarities and difference

Another set of experiments was based on the assumption that social interaction is predictable from a knowledge of the similarities of, or differences between, personality traits of the interactors.

Defense mechanisms. Differences in defensemechanism preference of two persons in interaction were used as a basis for investigation (Cohen 1956). Three pairs of subjects were used: one pair manifested projection (the tendency to attribute to someone else the impulses which the subject finds unacceptable in himself), another pair had similar defenses other than projection, and a third pair had defenses that were dissimilar. The members of each pair were asked to read some short stories which would arouse their common area of disturbance, to make individual judgments about the motives of the actors in the stories, and to arrive at a common decision about the motives through discussion.

The pairs of subjects who used projection perceived their own interaction as more hostile than did pairs of subjects who utilized other defenses. This severe negative effect was obtained only when projectors were paired with each other. When projectors interacted with persons using other defenses, their perceptions of hostility were no different from those of any other pairs having dissimilar defenses. The interaction that was perceived as next most hostile was that of two regressors, followed by that of two persons using reaction formation, while the least hostility was felt by two interacting users of denial or repression. [See Defense mechanisms.]

Authoritarianism. Another experimenter explored similarity or differences in authoritarianism of group members as a determinant of interaction. It was hypothesized that leaders emerging in groups made up entirely of authoritarian members would, for example, be generally less effective than would leaders emerging from groups made up entirely of equalitarian members. The hypotheses were supported by the data, although not always to a statistically significant degree (Haythorn 1953).

Interpersonal needs. In another investigation, similarity of interpersonal needs, that is, needs that can be satisfied only through the attainment of a satisfactory relation with other people, was used as the basis for composing ten-person groups for purposes of learning about human relations. Groups were composed along the dimensions of the Fundamental Interpersonal Relations Orientation (FIRO) theory of interpersonal behavior (Schutz 1958). The groups were divided into those with (a) strong attempts to control others—high expressed control, (b) strong desire to be controlled by others—high wanting to be controlled, (c) strong desire to receive affection—high wanted affection, and (d) strong preference for initiating behavior in areas such as control and affection. The scores of each group were randomly distributed on the other uncontrolled scales. After one week the group members could accurately identify their own characteristic from observing themselves interacting. It was easiest for them to identify a positive common group trait, such as expressing control, and most difficult to identify the common group trait when it was a covert desire, such as wanting affection. The type of interaction exhibited by each group was unique and quite consistent with the prediction based on their composition. The highly controlling group ejected their leader in the third meeting, while the group that was high in wanting to be controlled became highly dependent on their leader. The group high in initiating behavior spent their time talking about authority, competition, and feelings of adequacy and proceeded to compete with each other and with the leader. The group high in wanting affection spent much of their time discussing the problems of expressing feelings and crying. Thus, personality data obtained from a paper-and-pencil test did predict several aspects of group behavior.

Compatibility and complementarity

More complex assumptions about interaction underlie another study aimed at predicting the behavior of pairs that were composed on the basis of compatibility and complementarity.

Defense mechanisms. Building upon the study of defense preferences reported above, predictions were made, based on clinical knowledge, about the result of each combination of defense mechanisms —denial, isolation, projection, regression, turningagainst-self, and intellectualizing. For example, for two isolators it was reasoned that each of the isolators wants to handle feelings by intellectualizing or rationalizing and expects that the other will deal with an anxiety situation in the same way so that threatening feelings need not be considered. Since each agrees with this means of handling the problem, their interaction would be expected to be positive (Waxler 1960).

It was found, as predicted, that pairs of two regressors were least satisfied, while the regressor—isolator pairs liked each other and were most satisfied. The satisfaction and liking ratings were not related to an individual’s defense mechanism but, instead, were related to the combination of such mechanisms in the pair.

Interpersonal needs. Another approach centers on the concept of compatibility of interpersonal needs (Schutz 1958).

The types of compatibility proposed may be understood best by considering the diagonals of the diagram shown in Figure 1. The horizontal axis represents the behavior a person expresses toward others; the vertical axis, the behavior the persons wants from others. This schema is used separately for each interpersonal area proposed in the theory—inclusion (need for human contact), control (need for influence), and affection (need for love). The high-interchange quadrant represents interactions of those who prefer a great deal

of exchange of the “commodity” (contact, influence, love) relevant to the interpersonal area. The low-interchange quadrant represents those people who neither initiate nor want to receive these commodities. In order to be compatible, two people should be similar with respect to the interchange variable. Compatibility based on similarity along this diagonal is called interchange compatibility.

The other diagonal extends from a point representing those who desire only to initiate or originate behavior to a point representing those who wish only to receive it. For example, for control the lower right quadrant represents someone who likes to give orders and not take them, while the upper left represents someone who does not like to give orders but who likes to be told what to do. On this diagonal, two people should complement each other, that is, be equidistant from the center in opposite directions, in order to be compatible. Compatibility based on complementarity along this diagonal is called originator compatibility.

Conflict arises when there is disagreement regarding who shall originate relations and who shall receive them. For each need area (inclusion, control, affection) there are two types of conflict: between two originators—competitive originator incompatibility—and between two receivers—apathetic originator incompatibility.

In the inclusion area, competitive conflict is between two persons, each of whom wants to “select his own company.” Each wants to join only the activities he wishes but not to have others join him. The apathetic conflict is between two persons, both of whom want to be included but neither of whom will act to join the other.

In the control area the competitive conflict is between two persons, each of whom wants to be dominant and to run the activities but does not want to be told what to do. The apathetic conflict in this area is between two submissive people, each of whom wants to be told what to do but neither of whom will take the initiative for doing it. This situation arises with a boss who cannot make decisions and an employee with no initiative.

In the affection area the competitive conflict is between two persons who desire to originate close relations but not to receive them. One example of this would be the Don Juan, for whom pursuit is an end in itself and reciprocation is threatening. The apathetic conflict is between two people who want to be liked but do not want to initiate warm feelings. An example would be two co-workers secretly fond of each other but neither ever initiating a personal relation.

Interchange compatibility highlights a different aspect of the relation. Whereas the situations covered by originator compatibility refer to need satisfactions by a specific person, interchange compatibility refers to the general context or “atmosphere” in which a relationship exists. On the interchange axis, two individuals’scores should be similar for maximum compatibility.

In the area of inclusion, high interchange means high interaction with others; low interchange, to being separated from other people and being alone. Conflict arises because one person likes to be with people and the other person prefers to be alone.

In the area of control, high interchange means a preference to be both controlled by others and to control their actions; low interchange, to neither influencing nor being influenced by others. Conflict arises because one person wants a highly structured relation and the other person wants a free and spontaneous one.

In the area of affection, high interchange means reciprocal preferences for close personal relations with people; low interchange, a reciprocal preference for maintaining affectional distance. Conflict arises when one person wants an atmosphere of personal warmth and closeness and the other person prefers a more impersonal task-oriented atmosphere.

These indices of compatibility have been applied to several experimental situations, including selection of roommates and sociometric selection of naval personnel and task groups. Results seem to be fairly consistent: use of compatibility scores increases the predictability from a chance level of 50 per cent to about 70 per cent (Schutz & Gross 1959).

By use of this theory of compatibility, success was achieved in predicting the ability of an experimenter to condition a subject by considering the interpersonal compatibility of the two. When subject and experimenter were incompatible the conditioning proceeded more slowly than when subject and experimenter were compatible. Then, in an ingenious experimental maneuver, each experimenter asked his subject to continue with the conditioning procedure while the experimenter left the room. The subjects from the incompatible pairs immediately improved their performances to the point where they equaled the compatible subjects, thus confirming the hypothesis that it was the presence of the experimenter that led to the depressed rate of conditioning (Sapolsky 1960).

Another set of investigations was based on Winch’s hypothesis (Winch 1958), which asserts that marital success is a function of the husband’s and wife’s need complementarity, particularly in the areas of dominance-submission and nurturance-succorance. Unfortunately, to date the results of this method of investigation are not encouraging.

Larger groups. Two attempts have been made to expand the compatibility concept to groups of more than two members. An experiment using groups of college students illustrates one method of arranging five-membered groups to be compatible or incompatible. Four groups of college students equated for intelligence were composed according to a pattern of interchange compatibility based on high-affection scores. Four other groups were established on the basis of interchange compatibility reflecting low affection and four followed an incompatible pattern.

High-affection compatible groups contained one member who was predicted on the basis of his test scores to be a “focal person,” another who was predicted to be a “main supporting member,” and three members who were predicted to be less intelligent and less assertive. All of the subjects in these groups had a liking for close, intimate relationships. The subjects in the compatible groups with low-affection scores were selected to play similar roles, except that they were all persons who liked to keep others at a distance. In the incompatible groups there were two focal person-follower pairs, one pair had high-affection scores, and the other had low-affection scores; there was one neutral person. It was predicted that the two subgroups would clash and that, therefore, energy which would ordinarily be devoted to productive work would be diverted into interpersonal problem solving; hence, these groups would be less productive.

Each of the 12 groups met 14 times in a laboratory over a period of six weeks, during which time each group was given a sequence of several types of tasks to perform.

The difference in productivity between the “compatibles” and the “incompatibles” was highly significant. Both types of compatible groups clearly outperformed the incompatibles, and there was virtually no difference in productivity between the high-affection compatibles and the low-affection compatibles (Schutz 1958).

Organizational structure. Another attempt to devise a theory for compatibility of groups larger than two arose out of the desire to assess compatibility within an organization. Although no rigorous empirical work has yet been done to test this method, it has already revealed the necessity of considering factors other than personality.

To measure the compatibility of people working together within a complex structure, by summing the degree of compatibility of each dyad within the organization, is an unsatisfactory method of assessing the compatibility of a large group since:

(1)The over-all compatibility of an organization is often more dependent on some dyads than on others. For example, how well the boss gets along with a key department head is often far more pivotal for organizational compatibility than is the compatibility of two subordinates doing a routine job.

(2)Compatibility between people who must work together closely to produce something is more important than compatibility between two whose jobs do not require their interaction.

These two considerations suggest a method for extending the concept of compatibility to organizations. The organization must first be analyzed in terms of the acual working groups, that is, a chart should be drawn indicating who must actually work with whom in order to accomplish a certain task. Within these groupings the importance of each dyadic relationship must be assessed.

The best way to determine the functional pattern of the organization is to do it subjectively, that is, to have as many people as possible who are acquainted with the organization make such an analysis. A consensus of informed judgments is usually possible.

Likewise, the best means of weighting the importance of any given dyad to the functioning of the organization is to obtain a subjective judgment, that is, to get a consensus of all knowledgeable people. It may be possible in some organizations to approximate the weighting by making it proportional to the dyad’s status level within the organization; the higher up in the hierarchy, the more important to the organization’s functioning is the dyadic compatibility, but it would be inaccurate to assume this for all situations. Experience, so far, indicates only that the number of distinguishably different levels of importance are usually small, of the order of three or four within an organization.

After the functional groups and weightings are determined, all dyadic compatibilities between relevant pairs are computed and weighted. The organizational compatibility can be represented by the sum of these weighted compatibilities divided by the sum of the weights—in other words, the mean compatibility over the organization.

This method is very new but gives promise of extending the notion of compatibility to groups larger than two.

Personality as internalized interaction

One of the most interesting theoretical developments in the area of personality and social interaction concerns the use of social interaction as a model for theorizing about personality. This position reverses the usual analogy that a larger social unit is like an individual organism and asserts that the individual can be treated as a small group of interacting people.

This assertion is not an attempt merely to establish an analogy, but it is a contention that the same laws hold for both the individual and the group levels and that the substitution of corresponding variables from these levels will result in correspondingly valid hypotheses. A parallel can be made between a person and a group, since it is often assumed that an individual has a group “within” him. In the course of an infant’s development many initially ambiguous figures in his environment are gradually brought into focus and differentiated. These figures are then introjected to various degrees and exert a differential influence on the behavior of the developing child and, later, on the behavior of the adult. [See Developmental Psychology; Identity, psychosocial; Perception, article onperceptual development.]

The individual, therefore, may be conceptualized as a group in which he is struggling to become the leader. This group is composed of all those people whom he has incorporated into his own ego—his introjections. Just as some leaders seem to be dominated by one group member, so may an individual be influenced by a particular introjection; just as external forces influence group behavior, so do an individual’s external personal relations affect the interaction of his introjections; just as a group at times acts as if it were torn by dissension, so does an individual’s behavior at times reflect internal conflict; just as groups vary in cohesiveness, so do individuals vary in their integration; just as groups become immobilized and unproductive, so do people.

An individual arriving at a decision may be regarded as symbolically working out the interaction of the group within himself to reach a decision. The group equivalent of individual behavior is a group decision or a topic that a group chooses to discuss, as opposed to idiosyncratic topics. Group decision behavior is analogous to ego functioning.

Berne (1961) states that the individual personality consists of parts called the Parent, the Adult, and the Child. Fairbairn (1952) emphasizes the importance of internalized objects as a conceptual framework for understanding individual personality. Schutz (1961a) evolves his point of view through the vehicle of psychoanalytic ego psychology, especially as developed by Hartmann (1939). Hartmann discusses ego development in terms of three sets of factors: inherited ego characteristics, influences of instinctual drives, and influences of outer reality. Schutz transforms this into corresponding terms for small groups, and Bennis (1964), in a review of organization theory, expresses organization problems in much the same terms.

There seems to be general agreement that each social unit must deal with an environment, or outer reality, with its own internal emotional forces or interpersonal needs and with those factors involved in task behavior, such as intellect and other abilities. The following integration is suggested.

Outer realit

It is necessary for each social unit to establish and maintain sufficient contact and interaction with outside groups and individuals to avoid isolation of the unit, but it is not necessary to engage in so much contact that the unit loses its identity.

It is also necessary to establish and maintain sufficient control over outer reality so that the group can function satisfactorily without outside interference, and yet it is not necessary to assume so much control that the group is forced to undertake more responsibility than it can handle.

There is, furthermore, a need to establish and maintain sufficient closeness and intimacy with outside reality so that the group can feel the pleasures of friendship and affection, and yet there is no need to experience so much intimacy with outside reality that the actions of the group become distorted and detrimental to group objectives.

Interpersonal need

The group’s existence is greatly dependent upon everyone’s feeling that he is part of the group. The desire for inclusion becomes a motivation for efficiency in activities, such as notifying members of meetings. However, group members must also be allowed to maintain some degree of distance from other group members and some degree of individuality.

It is necessary for members to influence other members to some extent in order to make decisions and to establish behavior patterns leading to a restriction of the amount of control some members have over others. The institutional procedures of majority rule and consenus are often used to balance control relations.

People must relate to each other with sufficient warmth and closeness so that group processes may occur. If there is not enough freedom to express feelings among members, then productivity suffers because of the tie-up of energy in the suppression of hostile impulses. However, excessive intimacy and closeness may have the effect of detracting from the main purposes of the group and of personalizing task issues to an undesirable extent. Hence, it is necessary to balance the degree of closeness in groups.

Task behavio

In the area of the task behavior the precise functions of the group are not as clear, but they do include at least the following: (a) establishment and clarification of the hierarchy of group goals and values; (b) recognition and integration of the various modes of problem solving existing within the group; and (c) development and utilization of the full potential of each group member.

Examples and extension

These descriptions of group functions are directly transferable to individual personalities and organizations. Specification of these areas of concern originated in ego psychology and, through the use of the analogy to small groups, generated the above theoretical structure. The value of the approach is also felt in the other direction—suggestions for personality theory derived from the analogy to small-group theory. Two illustrations will develop this point.

In Redl’s (1942) analysis of group leadership he distinguishes ten types of “central persons,” that is, a person around whom a group forms as a result of his eliciting common feelings from group members. These types suggest ten ways in which individual personalities are integrated, that is, ten kinds of ego integration. For example, one of the ways groups are formed, according to Redl, is through the use of the central person as an object of love. Another mode of group formation uses the central person as the object of aggression. Perhaps these two processes occurring within one individual may be distinguishable by strong feelings of self-love on the one hand and of self-hate on the other. For example, a person in whom the self-hate self concept exists at an unconscious level is analogous to a group organized around a central person on the basis of the members’ commonly shared feelings of aggression toward that person.

Another type of group formation centers around a common-conflict solver, a person who acts in such a way as to resolve enervating conflicts in other group members. Where there is a great deal of conflict within an individual, he may select an introjection that is not a source of conflict in order to orient his own ego, thereby allowing him to solve this problem. One might wonder, however, about the stability of such an individual; for groups this formation is a very volatile one, the leadership pattern being dependent on the immediate conflict only. Possibly, a useful classification of pathological conditions may be made on the basis of inadequate types of ego integration.

This parallel leads to lines of research in which the group, conceptualized as one individual, becomes the basic tool. Attempts can be made to reproduce in the group various kinds of ego-integration patterns found in individuals and to examine their consequences. Since it is ordinarily easier to see phenomena in a group than within an individual, experimental personality research can be expedited with the use of the small group.

Another application of this point of view has been made by certain psychotherapists, although their theoretical basis may be somewhat different. The gestalt therapists and those in sociodrama make use of the multifaceted personality. For example, the therapist may say (in an appropriate situation), “Imagine someone inside your head saying to you, ’ You are a mature adult,’ and someone else inside your head saying, ’ You are an immature child.’ Make them have a discussion or argument with each other until one wins. Act out each part.” In this way the therapist is helping the patient to understand the nature of a conflict. He is doing it by utilizing the concept that a personality is made up of introjected people. Often the patient will visualize his father, or himself at an earlier time, or mother, or boss, or some other significant figure in his past as one of the people in his head. The internal conflict he is presently experiencing can then be understood as a struggle between parts of “others” that he has internalized.

There is ample promise that progress is being made toward predicting social interaction from a knowledge of the personality of the interactors and the laws of human interaction. The notion that the laws of group behavior and the laws governing the individual personality are similar or perhaps even the same is an exciting one and seems likely to lead to some fruitful investigations. In short, the apparently ineffable relations that exist between men are slowly being identified and measured.

William C. Schutz

[Other relevant material may be found in Conflict, article on Social Aspects; Groups; Identity, psychosocial; Organizations; Self concept; Socialization; and the biographies of Cooleyand Mead.)


Bennis W. 1964 The Decline of Bureaucracy and Organizations of the Future. Unpublished manuscript. → Paper presented at the annual meeting of the American Psychological Association.

Bekne, Eric 1961 Transactional Analysis in Psychotherapy. New York: Grove.

Cattell, Raymond; and Stice, Glen F. 1954 Four Formulae for Selecting Leaders on the Basis of Personality. Human Relations 7:493-507.

Cohen, Arthur R. 1956 Experimental Effects of Ego Defense Preference on Interpersonal Relations. Journal of Abnormal and Social Psychology 52:19-27.

Fairbairn, William R. (1952) 1954 An Object-Relations Theory of the Personality. New York: Basic Books. → First published as Psychoanalytic Studies of the Personality.

Hare, Alexander P. 1962 Handbook of Small Group Research. New York: Free Press.

Hartmann, Heinz (1939) 1958 Ego Psychology and the Problem of Adaptation. Translated by David Rapaport. New York: International Universities Press. → First published as Ich-Psychologie und Anpassungs-problem.

Haythorn, William 1953 The Influence of Individual Members on the Characteristics of Small Groups. Journal of Abnormal and Social Psychology 48:276-284.

Kerckhoff, Alan; and Davis, Keith 1962 Value Consensus and Need Complementarity in Mate Selection. American Sociological Review 27:295-303.

Perls, Frederick; Hefferline, Ralph; and Goodman, Paul (1951) 1962 Gestalt Therapy. New York: Julian.

Redl, Fritz 1942 Group Emotion and Leadership. Psychiatry 5:573-596.

Sapolsky, Allan 1960 Effect of Interpersonal Relationships Upon Verbal Conditioning. Journal of Abnormal and Social Psychology 60:241-246.

Schutz, William C. 1958 FIRO: A Three-dimensional Theory of Interpersonal Behavior. New York: Holt.

Schutz, William C. 1961a The Ego, FIRO Theory and the Leader as Completer. Pages 48-65 in Luigi Petrullo and Bernard Bass (editors), Leadership and Interpersonal Behavior. New York: Holt.

Schutz, William C. 1958 FIRO: A Three-dimensional Journal of Abnormal and Social Psychology 62:275-281.

Schutz, William C. 1964 Organizational Compatibility. Unpublished manuscript.

Schutz, William C; and Gross, Eugen F. 1959 The FIRO Theory of Interpersonal Behavior: Empirical Tests and Applications to Business Administration. Pages 161-172 in California, University of, Los Angeles, Western Data Processing Center, Contributions to Scientific Research in Management. Berkeley: Univ. of California Press.

Waxler,N. 1960 Defense Mechanisms and Interpersonal Behavior. Ph.D. dissertation, Harvard Univ.

Winch, Robert F. 1958 Mate-selection: A Study of Complementary Needs. New York: Harper.


Interaction process analysis is an observational method for the study of the social and emotional behavior of individuals in small groups—their approach to problem solving, their roles and status structure, and changes in these over time. The method is sometimes treated as a type of content nalysis, because, like content analysis, it is employed to estimate the relative strength of various underlying determinants of overt behavior. It is distinguished from content analysis, however, in that the observer abstracts from the content, in the ordinary sense of “what is talked about,” and focuses attention instead upon the form of the behavior and the changing patterns of action and reaction among individuals by which the content is communicated.

The term “interaction process analysis” was introduced into the literature as the title of a book by the present author (Bales 1950). It was presented as a generic designation for a number of similar methods. Some of these methods were then being used in experimental studies of .groups, particularly studies by researchers in group dynamics; others were being used in studies of counseling and therapy, particularly research in nondirective therapy, and in studies of classroom groups. Earlier methods for the study of child play and for the study of discussion groups had appeared as much as twenty and thirty years previously. Theoretically, therefore, there is justification for using “interaction process analysis” as a generic term, but in practice it has been closely associated with the particular method and set of categories advanced by the present author.

The method to be described here was the first to be self-consciously developed as a general-purpose descriptive and diagnostic procedure designed to produce theoretically relevant measures for all sorts of small groups, thus encouraging the development of empirical norms. The method has been used by a number of different investigators, and norms as well as a number of useful empirical generalizations have resulted. Table 1 shows the categories with information on the distribution of rates in a sample of 21 widely different types of situations (representing many times that number of groups) observed by different investigators.

In a survey of the usage of the method, Bales and Hare (1965) describe the setting, type of group, and other details, and report the total summary profile for each of 21 different studies. The profiles of acts initiated are reported in percentage rates in each category, usually on large numbers of acts. The 21 studies give indications of the ways in which the rates in each category are likely to vary over many situations and investigators. However, the results reported in Table 1 should be used as norms only with caution, since the sample may not represent the appropriate range of variability for many diagnostic purposes. Table 1 represents variability due to different observers, situations, organizational settings, sizes of group, tasks, gross

Table 1 - List of categories used in interaction process analyses, showing distribution of rates obtained in each category in a sample of 21 studies by different investigators*
Category of act initiatedMedian roteRange of lowest 7 ratesRange of middle 7 ratesRange of highest 7 rates
*The rates for a given category across the total sample of 21 studies are rank-ordered and divided into the lowest seven, the middle seven, and the highest seven. Cutting points are established so that the range shown contains all seven rates, but no gap is left between thirds of the distribution. The labels “Low,” “Medium,” and “High” are added for use in connection with Table 2.
Source: Bales & Hare 1965.
1. Shows solidarity2.30.1-1.92.0-3.33.4-9.5
2. Shows tension release6.20.5-5.35.4-7.47.5-30.4
3. Agrees9.50.9-7.98.0-13.613.7-21.5
4. Gives suggestion3.70.0-2.93.0-7.07.1-25.2
5. Gives opinion22.83.5-19.819.9-27.227.3-33.4
6. Gives Orientation29.013.6-21.021.1-33.733.8-56.9
7. Asks for Orientation5.11.8-3.94.0-7.27.3-12.9
8. Asks for opinion2.20.5-1.92.0-3.94.0-9.3
9. Asks for suggestion0.40.0-0.20.3-0.91.0-2.2
10. Disagrees4.40.6-3.03.1-5.35.4-14.3
11. Shows tension2.20.2-1.71.8-3.03.1-15.5
12. Shows antagonism1.60.0-0.50.6-2.42.5-9.2

differences between groups because of age, sex, general physical condition of members (such as whether or not intoxicated), and the like.


The observer studies the list of categories and definitions until he is thoroughly familiar with them, not only singly, but as an ordered scheme. Prior to the interaction he assigns and memorizes an identification number for each of the participants. In observing he keeps his eyes on the group as much as possible. He divides the ongoing behavior, nonverbal as well as verbal, into separate acts, each of which is recorded by entering the identification number of the person speaking, followed by the identification number of the person spoken to, under the category which best describes the act. The criterion as to how much behavior constitutes an act is pragmatic—enough to allow the observer to make a classification. A single act is essentially equivalent to a single simple sentence. Tone of voice, facial expression, bodily movement, and cues of all kinds, nonverbal as well as verbal, are used in making the classification and in determining to whom the act is directed. The group as a whole is recognized as a recipient of communication, as well as specific other individuals. But the group as a whole is not recognized as an actor. Acts performed in unison (notably laughing) are recorded by a single score but later credited back to individual actors. The observer scores continuously, usually at a rate of ten to twenty scores per minute. Comparable methods differ from each other not only as to the list of categories but also on all these points of procedure. The issues have

been discussed in several standard works on methods (see Research Methods in Social Relations 1951; Festinger & Katz 1953; Lindzey 1954).

Interaction process analysis is designed for on-the-spot concurrent recording of the behavior, but it may be applied to a sound recording or a written transcript of interaction. The method was developed and is most easily applied with the aid of a laboratory observation room, connected to the group meeting room by one-way mirrors, with sound monitoring, tape recording, and a paper-moving device called an interaction recorder to enable the observer to keep scores in time sequence. None of these aids is strictly necessary, however. Hare (1957) has applied the method to observation of the leadership behavior of boys on a playground; he used nothing more elaborate than a pencil and a pad of scoring forms. For a field study Strodtbeck (1951) used a pickup truck with a portable sound recorder. Husband-and-wife couples sat in the front seat and talked. The sound recordings were translated, where necessary, by local assistants, Strodtbeck added notations of sighs, laughs, and the like, and the interaction scoring was done by a third person from the written transcript. In an experimental study of jury deliberations Strodtbeck and Mann (1956) noted a few words of content as well as who-to-whom and nonverbal signs at the time of interaction, and the scoring was done later from these notes plus the sound recording. Some researchers may prefer a two-step method like this for the sake of higher reliability or to ease the training problem, but many successful studies have been made using on-the-spot scoring. Learning the method may require one or two months. Through training, correlation between two observers of rates in a given category over a series of sessions may be as high as .90 for some pairs of observers on the categories of higher frequency, but the same observers may find it difficult to obtain correlations higher than .60 for the categories of lower frequency.

Frame of reference

In order to present the correlates of measures to be obtained from interaction process analysis it is helpful to make reference to a three-dimensional “property space” (for which see Barton 1955) that has turned up repeatedly in factor analyses of interpersonal behavior, perceptions, and ratings persons make of each other in small groups. The space was first recognized as rather general by Carter (1954) and Couch (1960), who factor-analyzed data of their own and reviewed a series of other factor analytic studies. The work of Borgatta, Cottrell, and Meyer (1956) and Borgatta (1960) has been particularly notable in carrying on this inquiry. Hare (1962) has presented a review of the relevant studies, to which should be added the work of Leary (1957), who constructed an interpersonal diagnostic system around two of the dimensions. It has not yet been shown that the different property or factor spaces cited are all indicators of the same fundamental evaluative tendencies of persons as they view other persons and actions, but the assumption is plausible. A factor analysis by Couch (1960), which employed an exceptionally large number of measures from a study conducted with the present author, turned up as the three most important factors the same three found most important in previous studies. Couch’s factor loadings form the factual basis for the description and interpretation employed here. The types of variables included overt social interaction (the present categories), personality tests, interpersonal perceptions and ratings, value statements, and observer’s ratings.

Dimensions of social evaluation

Let us suppose that there are at least three fundamental dimensions of social evaluation involved as one person views another in a group setting. The first is concerned with the degree of power, dominance, ascendance, or individual prominence of the person as perceived by the evaluator. In ordinary language this dimension is often referred to as if it were vertical in physical space: a person is said to move “upward” in the group as his power, dominance, or prominence increases and “downward” as his power decreases.

The second dimension is concerned with the pleasant or unpleasant quality of feeling aroused by the person. If the feeling aroused in the evaluator is pleasant, one of acceptance and liking, the evaluation may be called “positive.” If it is unpleasant, one of rejection and disliking, it may be called “negative.” For purposes of visualization one may think of this dimension as horizontal, with positive to the right-hand side and negative to the left-hand side.

The third dimension is concerned with the value of the other person in the performance of group tasks and the achievement of group goals—that is, those tasks and goals given by acceptance of whatever authority the group recognizes. A spatial metaphor is often used to refer to this dimension: the direction toward achievement of group goals is said to be “forward”; the direction away from achievement is said to be “backward.” The person who moves forward may be said to conform to the values given by acceptance of the authority that is effective within the group, while the person who moves backward may be said to deviate.

The spatial metaphors should not be taken too seriously—they are suggested here as an aid to visualization and memory. It may be, of course, that the spatial metaphors continue to be used because they provide an effective, though primitive, way of remembering a complicated set of relationships. It is not easy to find a better model for a three-dimensional set of relationships than the physical space that surrounds us. If the reader wishes to think in terms of the theory, he is urged to adopt the physical space model. He may then think of the directions in relation to his head as the center of the space.

According to this theory, then, persons judge or evaluate each other according to their significance in relation to power, affection, and contribution to group tasks. The dimensions of social evaluation must be understood broadly as correlated clusters of more concrete attitudes of persons toward others, not as single well-defined attitudes. But so far as they can be thought of as variables, the factor studies suggest that we ought to think of the dimensions as unrelated to each other—as uncorrelated (“orthogonal” in the language of factor analysis). If the variables were actually uncorrelated, individuals would be found about equally distributed in all positions described by all combinations of the directions. Thus, the factor analytic findings imply that the possession of a position of power by a person in a small group should not lead us to expect that the possessor will be either positively or negatively evaluated (liked or disliked), or that he will be either conforming or deviant with regard

Table 2 - Key for interpretation of high and low rates on the interaction profilea
interaction categoryDirectional component indicated if the rate of initiating is:Directional components indicated if the rate of receiving is:
a. High and low rates of interaction in all categories of interaction initiated and received by an individual may be combined to predict the attitudes other members may have toward that individual, that is, where they place him in the three-dimensional evaluative space. The combination is performed by addition of the directional components.
b. A directional component is an indicator of a direction of movement in the three-dimensional evaluative space. The dimensions and directions are:
1. Shows solidarityLow = DFHigh = UBLow = NFHigh = PB
2. Shows tension releaseLow = UHigh = DLow = DPFHigh = UNB
3. AgreesLow = NBHigh = PFLow = UBHigh = DF
4. Gives suggestionLow = 0High = 0Low = DNHigh = UP
5. Gives opinionLow = DPBHigh = UNFLow = DNBHigh = UPF
6. Gives orientationLow = UHigh = DLow ~ NHigh = P
7. Asks for orientationLow = DNHigh = UPLow = UPFHigh = DNB
8. Asks for opinionLow = NHigh = PLow = UPHigh = DN
9. Asks for suggestionLow = 0High = 0Low - SHigh = F
10. DisagreesLow = PHigh = NLow = DPBHigh = UNF
11. Shows tensionLow = UFHigh = DBLow = PBHigh = NF
12. Shows antagonismLow = DPHigh = UNLow = DPBHigh = UNF

to the task-oriented common values of the group. We should not expect that the task-oriented conformist is necessarily well liked, nor that the socially-emotionally oriented deviant is necessarily disliked. Knowing the person’s position on any one dimension we should not expect to be able to predict his position on either of the other two dimensions.

The conception of three independent dimensions or scales along which any person or any act within the group may be given a position by the social evaluation of group members is, if truly sound and typical, a useful basis for classifying not only positions within the group but also directions of motivational movement, types of values, types of acts, types of roles—in fact anything which may be considered an object of social-psychological evaluation. Properly understood, such a dimensional space is an appropriate starting point for a classification of the functional problems of small groups (the a priori base from which the interaction categories of the present system were originally derived). The classes of functional problems as originally presented (Bales 1950) were four in number: adaptive, integrative, instrumental, and expressive. Unfortunately, their relations to each other were not very clear. The empirical results of factor analysis now suggest that the adaptive direction may be identified as the downward direction of yielding power or submitting. Its opposite is upward or dominating. The integrative direction may be identified as the positive direction. Its opposite is disintegra-tive or negative. Finally, the instrumental and expressive directions turn out to be the two opposite directions of the same dimension, here called forward and backward.

Thus, to move upward rather than downward in the social evaluation of another person or the group means to acquire the connotation in the minds of others of overcoming by possession of power rather than of adapting because of lack of power. To move positively rather than negatively means to acquire the connotation of promoting feelings of social acceptance and integration rather than feelings of antagonism and rejection. To move forward rather than backward means to acquire the connotation of contributing instrumentally to the achievement of group goals rather than of encouraging deviance by the expression of fantasies, tensions, or feelings in such a way as to hamper achievement.

Status and leadership

Is “higher social status” the primitive meaning of “upward”? Probably what most sociologists mean by “higher status” is not simply “upward” but a combination of “upward,” “positive,” and “forward,” in the present dimensional system. High status of an individual in the small group, as in the large, is probably best thought of as an additive combination of high evaluation on several components, not necessarily correlated with each other. In the present case, the components are uncorrelated with, each other over a population of many individuals in many groups. Leadership, like high status, is a complex direction. Apparently what most group members mean when they rate a member high on leadership is that he combines the components of “upward,” “positive,” and “forward” movement. The traits which make for each of the components are uncorrelated over a population of many individuals in many groups. Hence, “great men” who combine them are rare. They must be simultaneously high on “activity,” “likeability,” and “taskability,” to use the language of an earlier formulation (Bales 1958).

The interaction profile

An interaction profile is an array of the rates of activity in each category. The profile of acts initiated by an individual is obtained by calculating the percentage of his total acts falling in each of the categories. The profile of acts received by an individual may be obtained as well as the profile of acts initiated. Individuals may be compared to each other or to general norms (the norms given in Table 1 are not completely appropriate for the examination of individual profiles, and especially not for acts received, but they are the best available). Comparison with these norms, or with some more appropriate set, can be translated into judgments of the individual as high, medium, or low on each category. These judgments in turn can be used as diagnostic indicators for placement of the individual on the three dimensions described above. Table 2 shows the placements most consistent with empirical evidence so far.

The interaction matrix

An interaction matrix is a tabulation of the number of acts addressed by each individual in a group to each other individual, and to the group as a whole, with appropriate subtotals and totals, as shown in Table 3. The symbol Σ1 is used to indicate the subtotal of acts addressed to specific individuals by a given actor. The symbol ΣO is used to indicate the subtotal of acts he addresses to the group as a whole rather than to specific individuals. The symbol ΣR is used to indicate the subtotal of acts he receives from all others. Finally, the term “total participation” is used to designate the sum of these three subtotals—that is, the total number of acts initiated plus the total number of acts received by that individual. The internal cells of the matrix show the number of acts each individual addresses specifically to each other individual.

The matrix is most easily examined if ordered according to rank of the individuals on total participation, as in Table 3. In most groups the amount initiated by a given actor to a specific other person tends to be of the same order of magnitude as the amount the other addresses to the actor. This can be seen in Table 3. But more exactly, it is generally true that the lower of the two participators addresses a little more to the higher participator than the higher participator addresses to him. It is generally true that persons in uncontested higher power positions address a relatively large part of their total participation to the group as a whole (Σ O) rather than to specific individuals. For the highest participator in many groups the amount addressed to the group as a whole (Σ O) exceeds the sum addressed to all individuals specifically (£1). This is the case in Table 3. For the lowest participators, on the contrary, the amount adddressde

Table 3 - Aggregate who-to-whom matrix for 18 sessions of six-man groups (individuals ordered according to rank on total participation*)
Individuals ranked for each matrix before aggregation of matrices.
Source: Adapted from Bales 1953. Copyright 1953, The Free Press, a Corporation.
Receiver Initiator123456Σ1Σ0Total participation Σ1 + Σ0+ΣR
1 1,2389615454453173,5065,66114,370
21,748 4433101751022,7781,2116,302
31,371415 305125692,2857424,971
4952310282 83491,6766763,660
566222414483 281,1414432,456
64701261146544 8193731,757
ΣR5,2032,3131,944 1,308872565 12,2059,10633,516

to the group as a whole tends to fall short of the amount addressed to specific individuals, as Table 3 also shows. When, for a given individual, the amount addressed to the group as a whole exceeds the amount addressed to specific individuals, it may be inferred that he is at least trying to move “upward.” Conversely, when the amount he addresses to specific individuals exceeds the amount he addresses to the group as a whole, it may be inferred that he is tending to move “downward.” The ratio of the two amounts can be used to give a more refined measure.

An act addressed to the group as a whole may be taken as a bid for power; an act received in return which does not require an answer or an argument may be taken as an agreement or vote of confidence. The actual power position of a given person in the group may be roughly estimated, then, by counting the number of other persons who will regularly respond to his bids for power with agreement, thus leaving him free to continue further bids for power by speaking to the group as a whole, instead of arguing with each of them specifically. The person who is successful in this way will build up a high total participation. For most individuals there is no better indicator of relative power or position in the upward-downward evaluation of the group than his total participation compared to that of others.

Phase movements

Under certain conditions of organization, and when there is a clear-cut but complex task of group decision to be arrived at within a given session, groups tend to move through an orbit of directional movement over the course of the session (Bales & Strodtbeck 1951). They may start with preliminary pleasantries that average toward the positive side. As they begin with the task proper they move downward with high rates of giving orientation, then forward and upward as rates of orientation fall and rates of giving opinion and suggestion rise. The movement tends toward the negative side as giving opinion rises accompanied by disagreement. (Actually, rates of agreement as well as disagreement tend to rise, but on the average the negative movement is at first stronger.) Some groups, especially those which begin with low status consensus, may hang in this phase indefinitely. Others, particularly if their status problems are not severe, pass through the crisis, and begin to swing back toward the positive side as agreement is reached. The average movement swings upward and backward as the group begins a period of joking and laughing. As the laughter grows and spreads, the movement swings backward, downward, and positive, completing the orbit at about the same place as the starting position, on the positive side.

Somewhat similar cycles can be described for longer time spans for groups (Heinicke & Bales 1953). It seems probable that a similar orbital model could be useful for examining successful as compared to unsuccessful episodes in child socialization, as well as other social processes which have a basic phasic character. This remains to be seen. It also seems likely that the conceptualization of evaluative space as a theory will have many applications in which actual observation of interaction may play only a minor part. It should be noted that estimates of position in the space may be made from many bases other than interaction.

Robert F. Bales

[Directly related are the entries Groups, article onrole structure; Leadership. Other relevant material may be found in Sociometry.]


Bales, Robert F. (1950) 1951 Interaction Process Analysis: A Method for the Study of Small Groups. Reading, Mass.: Addison-Wesley.

Bales, Robert F. 1953 The Equilibrium Problem in Small Groups. Pages 111-161 in Talcott Parsons, Robert F. Bales, and Edward A. Shils, Working Papers in the Theory of Action. Glencoe, III.: Free Press.

Bales, Robert F. 1958 Task Roles and Social Roles in Problem-solving Groups. Pages 437-447 in Society for the Psychological Study of Social Issues, Readings in Social Psychology. 3d ed. Edited by Eleanor E. Maccoby, Theodore M. Newcomb, and Eugene L. Hartley. New York: Holt.

Bales, Robert F.; and Hare, A. Paul 1965 Diagnostic Use of the Interaction Profile. Journal of Social Psychology 67:239-258.

Bales, Robert F.; and Strodtbeck, Fred L. 1951 Phases in Group Problem-solving. Journal of Abnormal and Social Psychology 46:485-495.

Barton, Allen H. 1955 The Concept of Property Space in Social Research. Pages 40-54 in Paul F. Lazarsfeld and Morris Rosenberg (editors), The Language of Social Research. Glencoe, III.: Free Press.

Borgatta, Edgar F. 1960 The Stability of Interpersonal Judgments in Independent Situations. Journal of Abnormal and Social Psychology 60:188-194.

Borgatta, Edgar F.; Cottrell, Leonard S.; and Meyer, Henry J. 1956 On the Dimensions of Group Behavior. Sociometry 19:223-240.

Carter, Launor F. 1954 Evaluating the Performance of Individuals as Members of Small Groups. Personnel Psychology 7:477-484.

Couch, Arthur S. 1960 Psychological Determinants of Interpersonal Behavior. Ph.D. dissertation, Harvard Univ.

Festinger, Leon; and Katz, Daniel (editors) (1953) 1965 Research Methods in the Behavioral Sciences. New York: Holt.

Hare, A. Paul 1957 Situational Differences in Leader Behavior. Journal of Abnormal and Social Psychology 55:132-135.

Hare, A. Paul 1962 Handbook of Small Group Research. New York: Free Press.

Heinicke, Christopher; and Bales, Robert F. 1953 Developmental Trends in the Structure of Small Groups. Sociometry 16:7-38.

Leary, Timothy 1957 Interpersonal Diagnosis of Personality: A Functional Theory and Methodology for Personality Evaluation. New York: Ronald Press.

Lindzey, Gardner (editor) (1954) 1959 Handbook of Social Psychology. 2 vols. Cambridge, Mass.: Addison-Wesley. → Volume 1: Theory and Method. Volume 2: Special Fields and Applications.

Research Methods in Social Relations. Rev. ed. By Claire Selltiz et al. (1951) 1964 New York: Holt.

Strodtbeck, Fred L. 1951 Husband-Wife Interaction Over Revealed Differences. American Sociological Review 16:468-473.

Strodtbeck, Fred L.; and Mann, Richard D. 1956 Sex Role Differentiation in Jury Deliberations. Sociometry 19:3-11.

interaction (social)

views updated Jun 11 2018


views updated May 29 2018

interaction The term used when an operator interacts with a computer by means of a set of input devices to achieve a desired effect. See also acknowledgment, prompt, echoing, feedback.

interaction (statistical)

views updated May 09 2018

interaction (statistical) See STATISTICAL INTERACTION.

About this article


All Sources -
Updated Aug 08 2016 About content Print Topic