The term ”symbolic interactionism” was invented by Blumer (1937) to describe sociological and social psychological ideas he presented as emanating directly from Mead, especially but not exclusively in Mind, Self, and Society (1934). ”Symbolic interaction theory” is a term that is related to those ideas, though not necessarily in the specific forms presented by Blumer or Mead.
The fundamental character of symbolic interactionist ideas is suggested by the theoretical proposition that the self reflects society and organizes behavior and by related imagery that addresses the nature of society and the human being, the nature of human action and interaction, and the relationship between society and the person. That imagery begins with a vision of society as a web of communication: Society is interaction, the reciprocal influence of persons who, as they relate, take into account each other’s characteristics and actions, and interaction is communication. Interaction is ”symbolic,” that is, conducted in terms of the meanings persons develop in the course of their interdependent conduct. The environment of human action and interaction is symbolically defined: It is the environment as it is interpreted that is the context, shaper, and object of action and interaction. Persons act with reference to one another in terms of symbols developed through interaction and act through the communication of those symbols. Society is a label aggregating and summarizing such interaction. Society does not ”exist”; it is created and continuously re-created as persons interact. Social reality is a flow of events joining two or more persons. More than simply being implicated in the social process, society and the person derive from that process: They take on their meanings as those meanings emerge in and through social interaction.
Neither society nor the individual is ontologically prior to the other in this imagery; persons create society through their interaction, but it is society, a web of communication and interaction, that creates persons as social beings. Society and the individual presuppose each other; neither exists except in relation to the other. This conception of society implicitly incorporates a view of the human being as ”minded” and that ”mindedness” as potentially reflexive. That is, people can and sometimes do take themselves as the object of their own reflection, thus creating selves, and do this from the standpoint of the others with whom they interact. Selves are inherently social products, although they involve more than reflected appraisals of others in the immediate situation of interaction; in particular, selves involve persons as subjects responding to themselves as objects. Thinking takes place as an internal conversation that uses symbols that develop in the social process. Mind arises in both the evolutionary and individual senses in response to problems (interruptions in the flow of activities) and involves formulating and selecting from symbolically defined alternative courses of action to resolve those problems. Choice is an omnipresent reality in the human condition, and the content of choices is contained in the subjective experience of persons as that experience develops in and through the social process.
Following from this imagery is a view of human beings, both collectively and individually, as active and creative rather than simply responsive to environmental stimuli. Since the environment of human action and interaction is symbolic; because the symbols attaching to persons (including oneself), things, and ideas are the products of interaction and reflexivity and can be altered and manipulated in the course of that interaction; since thought can be used to anticipate the effectiveness of alternative courses of action in resolving problems; and because choice among alternatives is an integral feature of social conduct, one arrives at an image of social interaction as literally constructed, although not necessarily anew in each instance, in the course of interaction. One also arrives at an image that entails a degree of indeterminacy in human behavior in the sense that the course and outcome of social interaction cannot as a matter of principle (not uncertain knowledge) be completely predicted from conditions and factors existing before that interaction.
THE SYMBOLIC INTERACTIONIST FRAMEWORK
Labeling the ideas of symbolic interactionism a ”theory” is misleading. If one distinguishes between a systematic set of interrelated propositions about how a segment of the world is organized and functions and assumptions about and conceptualizations of the parts of that segment, symbolic interactionism has more the character of the latter than the former. That is, it is more a theoretical framework than a theory per se. While features of the framework appear to militate against attempts to formulate systematic theory by using it as a base and various proponents deny that possibility, a few sociologists have employed the framework in efforts to elaborate specific theories (e.g., Heise 1979; Stryker 1980, forthcoming; Stryker and Serpe 1982; Rosenberg 1984; Thoits 1983; MacKinnon 1994; Burke 1991). It is not possible to review such specific theories nor characterize the research that derives from that framework here. (For extensive references to classic literature and research literature before 1985, see Stryker and Statham . For more recent research, see one of the texts written from a symbolic interactionist perspective, such as Hewitt , or Symbolic Interaction, a journal sponsored by the Society for the Study of Symbolic Interaction and devoted to work emanating from the framework.)
In the work of some (e.g., Blumer 1969), symbolic interaction ”theory” is intended to be a general sociological frame that is applicable to the intellectual problems of sociology as a discipline from the most micro to the most macro levels. In the work of others (e.g., Stryker 1980), it is a frame restricted in utility to issues of social psychology. The first position does not seem defensible, because any framework brings into special focus particular variables and leaves unattended—at least relatively—other variables and because the symbolic interaction framework highlights interaction, social actors related through interaction, and subjective variables ”internal” to those actors. It thus neglects features of the sociological landscape relating to large-scale social systems—the state, the economy, the ”world system,” demographic variables, and so forth—and does not easily pose sociological questions involving interrelationships among those features of large-scale social systems. This neglect has led to criticism of the symbolic interactionist framework as lacking the social structural concepts needed for the analysis of power and consequently as an ideological apology for the status quo (see Meltzer et al. 1975; Stryker 1980; Reynolds 1990). Although many (e.g., Maines 1977) deny the validity of this criticism, pointing to work by Hall (1972) and others, the criticism may be justified if the claim is that symbolic interactionism is a general sociological framework; it is not valid if the more restricted claim for its utility is made. However, there remains concern about the adequacy of the framework for problems of a distinctively sociological social psychology that centers on the reciprocal relationships of social units and social persons. There also is concern about whether the framework admits of and provides readily for the articulation of sociological and social psychology concepts. These concerns arise from the ways in which social structural concepts enter, or fail to enter, the symbolic interactionist frame (for presentations of an avowedly social structural version of the framework, see Stryker 1980, forthcoming). Whatever the intended coverage—from all of sociology to a limited social psychology—the framework traditionally has been conceived as knowing no cultural boundaries; that view, however, has been questioned (Hewitt 1990).
Implicated in the description of symbolic interactionist imagery provided above are many of the central concepts of the framework. The meaning of ”meaning” is fundamental. By definition, social acts involve at least two persons taking each other into account in satisfying impulses or resolving problems. Since social acts occur over time, gestures—parts of an act that indicate that other parts are still to come—can appear. Vocal sounds, physical movements, bodily expressions, clothing, and so forth, can serve as gestures. When they do, they have meaning: Their meaning lies in the behavior that follows their appearance. Gestures that have the same meaning (implying the same future behavior) to those who make them and those who perceive them are significant symbols.
Things, ideas, and relationships among things and ideas can all be symbolized and enter the experience of human beings as objects; objects whose meanings are anchored in and emerge from social interaction constitute social reality. Although meanings are unlikely to be identical among participants, communication and social interaction presuppose significant symbols that allow meanings to be ”sufficiently” shared. Because significant symbols anticipate future behavior, they entail plans of action: They organize behavior with reference to what they symbolize. In the context of the ongoing social process, meanings must be at least tentatively assigned to features of the interactive situations in which persons find themselves; without the assignment of meanings, behavior in those situations is likely to be disorganized or random. The situation must be symbolized, as must its constituent parts; it must be defined or interpreted, and the products of that symbolization process are definitions of the situation. Those definitions focus attention on what is pertinent (satisfying impulses or resolving problems) in an interactive setting and permit a preliminary organization of actions appropriate to the setting. Tentative definitions are tested and may be reformulated through ongoing experience.
From the point of view of the actors involved, the most important aspects of a situation requiring definition are who or what they are in the situation and who or what the others with whom they interact are. Defining the others in the situation typically is accomplished by locating them as members of a socially recognized category of actors, one (or more) of the kinds of persons it is possible to be in a society (e.g., male or female, young or old, employed or unemployed). Doing this provides cues to or predictors of their behavior and permits the organization of one’s own behavior with reference to them. When others are recognized as instantiations of a social category, behaviors are expected of them and actions that are premised on those expectations can be organized and directed toward them. Through this process, the introduction of early definitions of the situation can produce, although not inevitably, behavior that validates the definitions. This is an insight that underlies the notion of altercasting (Weinstein and Deutschberger 1963) and appears in the development of expectation states theory (Berger et al. 1974). When such behavior becomes routinized and organized, it also can serve to reproduce the existing social structure.
While some interactionists disdain the term, expectations attached to social categories—again, the kinds of persons it is possible to be in a society—are roles. Situations frequently allow one to locate others in multiple categories and open the possibility that conflicting expectations will come into play; in this circumstance, no clear means of organizing responses may be available. Defining oneself in a situation also involves locating oneself in socially recognized categories; to respond reflexively to oneself by classifying, naming, and defining who and what one is is to have a self. The self, conceived in this manner, involves viewing oneself as an object. The meaning of self, like that of any object, derives from interaction: To have a self is to view oneself from the standpoint of those with whom one interacts. The self, like any significant symbol, provides a plan of action. By definition, that plan implicates the expected responses of others.
People learn, at least provisionally, what they can expect from others through role taking, a process of anticipating the responses of the others with whom one interacts. In effect, one puts oneself in the place of those people to see the world as they do, using prior experience with them, knowledge of the social categories in which they are located, and symbolic cues available in interaction. On such bases, tentative definitions of others’ attitudes are formulated and then validated or reshaped in interaction. Role taking permits one to anticipate the consequences of one’s own and others’ plans of action, monitor the results of those plans as they are carried out behaviorally, and sustain or redirect one’s behavior on the basis of the monitoring. Because roles often lack consistency and concreteness while actors must organize their behavior as if roles were unequivocal, interaction is also a matter of role making: creating and modifying roles by devising performances in response to roles imputed to others (Turner 1962).
Many social acts take place within organized systems of action; consequently, both role taking and role making can occur with reference to a generalized other, that is, a differentiated but interrelated set of others (Mead’s example involves baseball players anticipating the responses of other members of their team and those of their opponents). Not all others’ perspectives are equally relevant to an actor; the concept of significant other indicates that some persons will be given greater weight when perspectives differ or are incompatible. It is implied here that meanings are not likely to be universally shared or shared in detail; if they are not, accuracy in role taking and difficulty in role making also will vary. It also is implied that smooth and cooperative interpersonal relations do not necessarily follow from accurate role taking: Conflict may result from or be sharpened by such accuracy.
The symbolic interactionist ideas reviewed here have a history. Many issue directly from Mead. Mead’s ideas are part of a tradition of philosophical thought with roots in the Scottish moral philosophers Adam Smith, David Hume, Adam Ferguson, and Francis Hutcheson and, more proximately, in the American pragmatists Charles Peirce, William James, and John Dewey. They also contain important admixtures of evolutionary and dialectic premises. Mead’s thought overlaps considerably with that of a number of sociologists who also wrote in the first decades of the twentieth century, in particular Charles Horton Cooley and William Isaac Thomas. Cooley’s axiom that society and the person are two sides of the same coin (the coin, he added, is communication) and that of Thomas asserting that if humans define situations as real, the situations are real in their consequences, capture much of the essence of symbolic interactionism. A host of sociologists connect that past with the present; among others, Burgess, Blumer, Waller, Sutherland, Hughes, Shibutani, Kuhn, Cottrell, Hill, Lemert, Lindesmith, Mills, Miyamoto, and Stone are linked to a more contemporary set of persons that includes Goffman, Lofland, Becker, Lopata, Strauss, Geer, Weinstein, Farberman, Couch, Denzin, Bart, Maines, Reynolds, Turner, Daniels, Scheff, Wiseman, Heise, Stryker, Burke, Heiss, Fine, Hochschild, Weigert, McCall, Snow, and Hewitt. (For reviews of the history and literature of symbolic interactionism, see Stryker and Statham 1985; Meltzer et al. 1975; Reynolds 1990; Lewis and Smith 1980.) The presence of these researchers in a common listing does not indicate their adherence to a common credo; there may beas much conceptual difference as similarity among them.
COMMONALITIES AND VARIATIONS
Thus, no single version of symbolic interaction theory satisfies all who find its core ideas appealing and useful in conducting research and analyses. There appear to be three fundamental premises of a symbolic interactionist perspective that are shared by those who acknowledge their intellectual roots in this tradition of sociological thought (Stryker 1988). The first holds that an adequate account of human social behavior must incorporate the perspective of participants in interaction and cannot rest entirely on the perspective of the observer. The second is that the self, that is, persons’ reflexive responses to themselves, links the larger social organization or structure to the social interaction of those persons. The third asserts that processes of social interaction are prior to both self and social organization, both of which derive and emerge from social interaction.
Each of these premises leaves open issues of considerable importance with respect to the content, methods, and objectives of interactionist analyses on which symbolic interactionists can differ. Some sociologists for whom the three core premises serve as a starting point believe that social life is so fluid that it can be described only in process terms, that concepts purportedly describing social structures or social organization belie the reality of social life. Relatedly, some believe that actors’ definitions, which theoretically are central and powerful as generators of lines of action, are reformulated continuously in immediate situations of interaction, making it impossible to use preexistent concepts to analyze social life (Blumer 1969). Others accept the ”reality” of social structural phenomena, viewing the social structure as relatively stable patternings of social interaction that operate as significant constraints on actors’ definitions. Social structure is thought to make for sufficient continuity in definitions to allow the use of concepts derived from past analyses of social interaction in the analysis of present and future interaction (Stryker 1980). The first premise hides, in the term “account,” the important difference between those who seem to believe that given the constructed character of social behavior, only an ”after the fact” understanding of past events is possible (Weigert 1981) and those who believe that sociology can build testable predictive explanations of social behavior (Kuhn 1964). Similarly, some argue that the perspective of a sociological observer of human social behavior is likely to distort accounts of that behavior andso must be abjured in seeking to capture the perspectives of those who live the behavior that is observed (Denzin 1970), and others argue directly or by implication that the requirement that accounts incorporate the perspective of the actors whose behavior is observed dictates only that actors’ definitions be included in developed explanations, not that they constitute those explanations (Burke 1991). The first group tends to argue that the best, if not the only ”legitimate,” methods are naturalistic, primarily observational (Becker and Geer 1957); the second group tends to be catholic with respect to methods, refusing to rule out categorically any of the full range of possible social science methods and techniques (Heise 1979).
With respect to the second premise, interactionists differ in the degree to which they assign an independent ”causal” role to the self as the link between social organization or structure and social behavior. For many, self can and does serve as an independent source of that behavior (McCall and Simmons 1978). For others, social organization or structure (as the residue of prior interaction) builds selves in its image, thus making the self essentially a conduit through which these structures shape behavior, not an independent source of that behavior (Goffman 1959). Similarly, there is variation among symbolic interactionists in the degree to which the self is seen as the source of creativity and novelty in social life, the degree to which creativity and novelty in social life are seen as probable as opposed to simply possible (occurring only under a specific and limited set of social circumstances), and the degree to which social life is constructed anew rather than ”merely” reconstructed in the image of prior patterns (Turner 1962; Hewitt 1997; Stryker and Statham 1985).
The third premise is interpreted by some as denying that social organization and selves have sufficient constancy to permit generalized conceptualization or the development of useful a priori theory on the basis of any investigation that can carry over reasonably to any new investigation (Glaser and Strauss 1967). In the view of others, this premise does not deny that there is in social life a reasonable constancy that implies a sufficient constancy in both selves and social organization to permit the elaboration of useful theories employing general concepts that potentially are applicable to wide instances of social behaviors (Heise 1986). Some emphasize the behavioristic elements in their intellectual heritage from Mead, concentrating on how concerted lines of social action are constructed (Couch et al. 1986; McPhail and Wohlstein 1986), while others adopt a stance that attends primarily to the phenomenological worlds of the actors (or interactors) they study (Denzin 1984).
Clearly, these possibilities for important variations in symbolic interactionist thought are not independent of one another: Those who subscribe to a view emphasizing the fluidity of social life and the moment-to-moment situated character of definitions also are likely to emphasize the degree to which social order continuously emerges from fluid process, the self organizes social behavior in an unconstrained fashion, and creativity and novelty characterize human behavior. They also tend to insist that the point of view of the observer contaminates reasonable accounts of social interaction, that there is little utility in an analysis of conceptualizations and theory emanating from earlier analyses, and that understanding, not explanation, is the point of sociological efforts.
The set of views presented in the preceding paragraph identifies symbolic interactionism for many of its most passionate adherents and perhaps for a majority of its critics. Those approaching their work from symbolic interactionism so defined tended to present what they did in both conceptual and methodological opposition to available alternatives in sociology. For example, Blumer (1969) devoted much of his career to championing direct and participant observation aimed at accessing the interpretations of those whose ongoing interaction sociologists sought to understand as opposed to both statistical and structural analyses, whose categories, data, and mathematical manipulations seemed to him devoid of actors’ meanings. Critics of symbolic interactionism attacked it and its adherents for being nonscientific and asociological. To circumscribe symbolic interactionism in the manner of these adherents and critics belies the diversity in views on key issues represented in the work of those who use the framework.
Interest in the symbolic interactionist framework within sociology has fluctuated. That interest was great from 1920 to 1950, reflecting in part the dominance of the University of Chicago in producing sociologists as well as the institutional structure of sociology. Through the 1950s and into the 1970s, interest waned, first as the structural func-tionalism of Parsons and Merton gained ascendance intellectually and Harvard and Columbia became institutionally dominant and later as Marxist and structuralist emphases on macro social processes swept the field. Symbolic interactionism, when not decried as reactionary or asociological, became the loyal opposition (Mullins 1973). Indeed, Mullins predicted that it would disappear as a viable sociological framework.
More recent events contradict that prediction: Symbolic interaction’ has had a remarkable revitalization in the past three decades (Stryker 1987), and there has been a corresponding resurgence of interest in the framework. The revitalization and resurgent interest reflect various sources. One is an emerging realization among sociologists with a structural orientation that their theories could benefit from the sociologically sophisticated theory of the social actor and action that symbolic interactionism can provide and the related increasing interest in linking micro to macro social processes. A second lies in a series of changing emphases in the work of contemporary symbolic interactionists. Although much recent work in a symbolic interactionist framework reflects traditional conceptual, theoretical, and methodological themes, on the conceptual level, newer work tends to adopt a ”multiple selves” perspective, drawing on William James (Stryker 1989; McCall and Simmons 1978) rather than viewing the self as singular or unitary. Theoretically, there is greater attention to emotion, to affective dimensions of social life (Hochschild 1979; Thoits 1989; MacKinnon 1994; Ervin and Stryker, forthcoming), correcting for a ”cognitive bias” in the framework; there is also greater appreciation for structural facilitators of and constraints on interaction and on self processes. While not yet prominent in contemporary interactionism, the groundwork has been laid (e.g., in Stryker and Statham 1985) for the reintroduction of the concept of habit, which was central in the writings of John Dewey and other forerunners of interactionism, in recognition that social life is not invariablyreflexive and minded. Current symbolic interactionism is methodologically eclectic and tends to be more rigorous than it was in the past, whether the methods are ethnographic (Corsaro 1985) or involve structural equation modeling (Serpe 1987). Also contributing to the revitalization of symbolic interactionism is the attention to its ideas, often unacknowledged but sometimes recognized, paid by a psychological social psychology that is predominately cognitive in its orientation. For cognitive social psychology, concepts are mental or subjective structures formed through experience, and these structures affect recognizing, attending, storage, recall, and utilization of information impinging on the person; of prime significance among concepts functioning in these ways are self-concepts. The link thus forged between cognitive social psychology and symbolic interactionism is mutually advantageous. Symbolic interactionism benefits from the ”legitimacy” implicit in the attention given to its ideas and from the expanded pool of researchers focusing on those ideas; cognitive social psychology can benefit from understanding that cognitions are rooted in social structures and processes.