I. Psychological AspectsTheodore R. Sarbin
II. Sociological AspectsRalph H. Turner
The concept of role came into psychology through the cross-disciplinary efforts of a number of social and clinical psychologists. The dramaturgical metaphor was first systematically exploited by George H. Mead and the University of Chicago sociologists in the 1920s. Its utility was demonstrated particularly in connection with the analysis of occupational and quasi-occupational categories. In the 1930s and 1940s several writers opened the way for social scientists to employ the concept in the analysis of interpersonal behavior. Among them were Moreno (1934), who pioneered the use of role playing as a tool of psychotherapy and role training; Sarbin (1943), who analyzed hypnotic behavior as role enactment; Cameron (1947), who demonstrated how the behavior of disordered persons could be understood as faulty role enactment, as failure in taking the role-of-the-other, and as ineptness in playing out one’s socially assigned roles; and Newcomb (1950), who used role as a central concept in his social psychology. Several writers have characterized role as the construct that bridges psychology and the other social sciences. It connotes not only overt actions and performances but also covert expectations held by an observer, or by a group of observers, such expectations serving as the basis for judging the propriety of the enactment. The concept of role is now widely used in the study of industrial and bureaucratic organizations, the family, small-group interaction, behavior modification, attitude change, behavior pathology, and recently, the analysis of psychological experiments where persons are assigned the role of experimental subject.
Definition of role
In general, the term “role” continues to be used to represent the behavior expected of the occupant of a given position or status. Thus, following the implication of the dramaturgical metaphor, an actor assigned to the position (or part) of Hamlet is expected to enact the role of Hamlet, the role being characterized by certain actions and qualities. A person who is assigned to the position of clergyman (or who elects to be placed in such a position in the social structure) is similarly expected to enact the role of clergyman characterized by certain typical actions and qualities. In this definition two features are emphasized: (1) expectations (i.e., beliefs, cognitions) held by certain persons in regard to what behaviors are appropriate for the occupant of a given position, and (2) enactments (i.e., conduct) of a person who is assigned to, or elects to enter, a given position.
Although the definition given above is satisfactory as a first approximation to the analysis of ongoing social behavior, more refined analysis has demanded the introduction of concepts that are more differentiated. In the first place, the behavior that serves as the dependent, or outcome, variable is seen as role enactment. That is, interest is focused on what the occupant of a given position does and says, such behaviors being noted by observers who may use behavior rating scales or free response verbalizations to make known their observations. Needless to say, the effectiveness, convincingness, validity, or propriety of such role enactments varies between persons and situations. To uncover the antecedents of this variation, social scientists have introduced additional theoretical conceptions and observational procedures. Among the variables that have been demonstrated to be antecedent to variation in effectiveness or propriety of role enactment are (1) validity of role expectations held by the actor, (2) accuracy of the actor in locating the other(s) (and reciprocally the self) in the proper role system, (3) sensitivity to situationally generated role demands, (4) available general and specific skills, (5) congruence of self and role, and (6) reinforcement properties of the audience. Each of the variables will be discussed below.
In order effectively to perform, i.e., to validate his occupancy of a social position, the actor must learn—either at first hand or vicariously—what performances are associated with what positions. In short, he must be acquainted with at least certain sectors of the social system; he must know what obligations, privileges, rights, and duties are the defining characteristics of each position he may be called on to occupy. Given such knowledge, he is in a better position to enact an assigned or ascribed role. The concept of role expectation is implicit in many practices in contemporary society. College students, for example, are frequently participants in orientation programs designed to provide them with the proper role expectations. Induction into the military or into bureaucratic organizations is similarly designed to acquaint inductees with their rights and privileges and with their duties and obligations. Failure to acquire expectations for recurring roles leads to enactments that are judged as inept, invalid, improper, antisocial, or illegal.
Role expectations may be viewed as actions or qualities expected of the occupant of a position. If viewed as actions, role expectations are codified as in a job description: the occupant is expected, for instance, to call the roll, open the windows, secure the doors. If viewed as qualities, role expectations are codified in adjectival terms; for example, the occupant is expected to be warm, friendly, outgoing, sincere, and cautious. This account of role expectations is necessarily oversimplified. Persons usually occupy multiple positions, some of which may be considered compatible, such as professor, father, and scientist, and some considered incompatible, such as high school teacher and bartender. The resolution of such incompatibilities has been studied in a number of contexts. These studies have provided us with a recognition of institutionalized rituals and personal adaptive techniques that serve to reduce the strain attendant upon occupying incompatible roles. In addition, the occupant of a position not only enacts role behavior that is congruent with the expectations held by others in the macrocosmic social system but also performs acts that make good his occupancy of concurrent positions in microcosmic personal role systems. Whereas the macrosystem points to the kinds of role behaviors appropriate to regularly occurring situations, miniature systems develop where the role expectations must be expressed in terms separate from the more obvious task expectations of the macrosystem. These may be denoted as idiosyncratic preferences, noncon-sensual expectations, and subtle or disguised expectations. Goffman (1961) has provided some enlightening illustrations of this notion under the rubric “encounters.” In a surgical team, for example, the chief surgeon may not only direct the flow of activities that are task oriented, thus meeting the expectations for surgeon, but also engage in behavior unrelated to the surgical task—such as making jokes—designed to maintain an optimal degree of tension in the performers.
The problems of assessing role expectations are no different from the problems of assessing any inferred variable. Expectations must be inferred from overt conduct, including self-reports, and from samples of behavior in unstructured test situations. An investigator may assess the role expectations held by a group of persons by observing their conduct in situations where an actor emits behavior in response to standard problems. Or he may make his inferences from replies to more direct questions, such as “What behavior should one typically expect from a person who is in the position of school superintendent?” Various devices have been created for more systematic investigation of role expectations, such as role inventories, designed to get at the actions expected of the occupant of a position, and adjective check lists and rating scales, designed to get at the qualities or personal dispositions expected of the occupant (Sarbin & Jones 1955).
It is axiomatic that in order to survive as a member of a society, a person must be able to locate himself accurately in the role structure. The simplest way to accomplish this is by seeking and finding answers to the question “Who am I?” Since roles are structured in reciprocal fashion, the answers can also be achieved through locating the position of the other by implicitly asking the question “Who are you?” The answers to the latter question are usually phrased in terms of role categories, such as man, teacher, friend, officer, secretary, and clown. In order to establish the position of the other, the actor must pay attention to the behaviors emitted by that other, scanning for cues that have reliability and validity. Physique and figure, hairstyle, facial adornment, and dress, among other things, are cues to which the actor may attend in order to locate the other, and reciprocally the self, in social space. The process of enculturation is heavily weighted with the learning of which emitted behaviors are signs or cues for which positions in the social structure. In ambiguous or partially structured situations, the actor may influence the role of the other by casting himself in one rather than another permitted role. His behavior, in short, serves as a potential constraint on the role enactment of the other, who must also locate himself in the social structure.
Under ordinary conditions the location of the other in social space occurs without much reflection, partly because much social behavior is ritualized, partly because the “badges of office” are more or less obvious. To locate the position of the other it is sometimes necessary to infer covert behavior on the part of the other, such as “feelings,” expectations, and attitudes. The more obvious cues and badges of office may not provide necessary information for correctly classifying the other. In this case the actor may search the ecology for additional cues, such as might be emitted in the form of posture, linguistic forms, facial expression, and gesture. Illustrative empirical research arising from this conception are two studies carried out in an effort to establish the relationship between accuracy of role perception on the basis of postural cues and indices of deviant or maladjusted role enactment. Sarbin and Hardyck (1955) constructed 43 stick figures in which all cues except postural ones were eliminated. These stick figures, black-and-white line drawings schematically depicting human postures, are presented one at a time to the subject, who selects, from among 5 choices, a word or phrase descriptive of the stimulus object. On standardizing groups, the frequencies of response to the 5 choices follow a J curve, the modal response containing at least 70 per cent of the distribution. When a subject’s protocol for the 43 stick figures shows a high proportion of modal responses, his performance is designated as conformant. Inmates of a state mental hospital who were classified as schizophrenic (whose deviant role enactments presumably were instrumental in establishing residence in the hospital) selected the modal response choice no more often than the low response choices.
A further study (Krasner et al. 1964) made use of a more refined scoring system. Those stick figures whose modal responses were terms indicating covert states, such as thinking, sadness, and anger, were scored apart from those stimuli whose modal responses were terms representing more overt conduct, such as running, sleeping, and standing at attention. Comparison of the performances of nonhospitalized control groups with hospital patients classified as schizophrenic showed significant differences in scores for the inference items but no significant differences for the non-inference items. This finding lends support to the proposition that the conduct of persons classified as exhibiting deviant role enactment (i.e., mental patients) is characterized by errors in locating others in the role system when they must depend on interpreting cues that represent covert states or dispositions.
Stated in an alternate form, the role perception variable considers the cue properties of the ecology, especially cues arising from the behavior of other persons. The accuracy with which a person notices such cues and draws conclusions regarding the role of the other is directly related to the accuracy with which he locates himself in the reciprocal or complementary role. The validity, propriety, and convincingness of his role enactment, then, depends on how accurately he perceives the role of the other. [SeePerception, articles onperson perceptionandsocial perception.]
Once a person correctly locates the position of the other on the basis of cues emitted by the other interactant, the possible choice of role behaviors is reduced from near-infinity to a small number. Further constraints on the choice of role behaviors are introduced when certain additional features of the situation are taken into account. These may be called role demands, that is, demands for a specific role enactment.
Silently operating to guide the actor in his choice of roles, these demands stem from cultural norms and may, in fact, dictate a role choice that is contradictory to requirements under conditions when the demands are not operative. Among such demands one may list norms for modesty, communication, control of aggression, cooperation, face-saving, public commitment, and prevention of embarrassment to others. Violation of such norms ordinarily carries heavy sanctions against the violator, hence their power to constrain the choice of role behavior.
Two series of experiments by Orne (1959) illustrate in a striking way the necessity of taking into account the role demand variable, or, to use his designation, the demand character of the environment. In his first series of studies he demonstrated that the role enactments of subjects properly motivated to simulate hypnotic subjects could not be differentiated from the role enactments of subjects who were instructed through traditional hypnotic procedures. By experimentally manipulating the role demands of the simulators, he approximated the demand character of the situation of the genuine hypnotic subjects. His simulators were instructed by an associate to pretend to be hypnotized and to try to deceive an expert in hypnotism. This study, together with a series of studies reported by Sarbin (1950), raises the possibility that hypnotic behavior is role enactment and that the manipulated demand for a specific role enactment is in large measure responsible for the outcome. One of the demands used by professional entertainers who perform hypnosis is that embarrassment to the other, in this case the hypnotist, be prevented. In a public setting the subject is in a position to embarrass the hypnotist by rejecting his instructions. However, for most persons the prevention of embarrassment is a powerful norm, and the informed hypnotist will capitalize on this fact.
Orne’s second series of studies (Orne & Scheibe 1964) is addressed to the social aspects of the psychological experiment. Generalizing from the effects of manipulating role demands in the hypnosis experiment, he hypothesized that the outcomes in psychological research might also be related to implicit or explicit demands. Such demands operate when a subject wittingly or unwittingly learns that he is in the experimental group or in the control group. The experiment was conducted in a laboratory setting that simulated sensory deprivation experiments, with an examination by a white-coated physician and the presence of a panic button in the experimental room, but without actual sensory deprivation. The experimental subjects reported many, if not all, of the bizarre and extreme phenomena attributed to actual sensory deprivation. Subjects who were in the control condition received the same experimental treatment (without the panic button) and reported none of the sensory deprivation phenomena. In brief, different role demands led to different outcomes. In this experiment the assignment to experimental or control group was made explicit. In the usual psychological experiment the subject is left free to speculate on whether he is in the experimental or control group. He will use whatever inputs are available to help define his role. The outcome of his performance in the role of experimental subject will be influenced by his responsiveness to the intended or unintended role demands. [SeeHypnosis; Persuasion; Suggestion.]
Cameron (1947), Gough (1948), and Sarbin (1950), among others, have presented arguments for the existence of a skill, aptitude, or competence that facilitates role enactment. Presumably acquired early in life through subtle learning experiences, this role-taking aptitude may be regarded as parallel to other aptitudes, such as mechanical, clerical, and drawing aptitudes. Authorities have not always been clear in their use of the term “role-taking aptitude”; in some contexts the referent is the ease with which a person enacts a number of roles, as in the case of the successful professional actor; in others the referent is the facility with which a person adopts and enacts a role on the basis of limited information. In the former referent the point of interest is role enactment as motoric (including gestural and verbal) behavior; in the latter it is cognitive and perceptual behavior.
Empirical studies have focused primarily on the cognitive component of this skill. With few exceptions, there has been no accurate measurement of the skill as a general cognitive competence. Most investigators, in fact, have drawn the conclusion that accuracy in predicting the behavior of others is a function of “stereotype accuracy”; i.e., the prediction is accurate insofar as the target person is regarded as an exemplar of a class of persons of known characteristics. If the subject knows the role system and the expectations that make up this system, he can extrapolate to any individual occupying a position in the system. The work of Gage (1953) is typical. Predictions of conduct were more accurate when based on general role expectations than on perceived idiosyncrasies of the target person. Although the experimental literature warns us to remain skeptical about a general role-taking aptitude, Cline and Richards (1960) appear to have demonstrated that such a disposition does exist. In a series of carefully controlled experiments some subjects made verifiable inferences about the social behavior of target persons after observing samples of behavior in sound motion pictures. The accuracy of these inferences could not be attributed to such conceptions as stereotype accuracy.
The motoric component has not been systematically studied except in the context of dramatic acting. Like the cognitive component, the motoric component of role-taking skill presupposes broad acquaintance with, and practice of, overt role behaviors associated with many positions. Although every person’s repertory of verbal and gestural actions is different from every other’s, a minimum of this skill is required for social interaction. Assessment of this variable can at present be achieved only through judgments made by observers. For example, Sarbin and Lim (1963) had their experimental subjects perform pantomime improvisations before a panel of specialists in dramatic arts who recorded on a rating scale their impressions of the convincingness of the performance. The purpose of the experiment was to determine whether relative responsiveness in the hypnosis situation could be attributed to the motoric component of the role-taking skill. The findings clearly showed that part of the variation in hypnotic behavior can be attributed to the motoric role-taking skill. In earlier writings the role-taking skill was likened to empathy, or the ability to take the part of the other. The term implied that the actor would take the perspective of the target person, sharing not only his attitudes and beliefs but also his feelings.
Critics of role theory have pointed to the confidence man as a person who is gifted in taking the role of the other but who obviously does not genuinely share the perspective of the victim. This observation has been used to question the utility of the concept of role-taking aptitude. The criticism may be outlined somewhat as follows: If the confidence man really takes the role of the other, would he not then see himself as the victim, feel the pain of loss or embarrassment, and reverse his behavior? This criticism confuses two dimensions in the cognitive process. The first is the cognitive component already discussed. On the basis of knowledge of people in general, acquaintance with the class of persons of which the intended victim is a member, and study of the intended victim himself, the confidence man draws inferences about the probable role behavior of the victim, given certain role demands that the confidence man may manipulate. The second dimension is an evaluative one and has no necessary connection with the first. The value that one places on his own conduct need bear no more relation to his skill in role taking than to any other skill.
The writings of George H. Mead contain observations and speculations about two conceptions that have become thoroughly entrenched in modern psychology, role and self. The extensions of role have already been discussed. For a more complete analysis of the variation in effectiveness of role enactment, the concept of self must be invoked. However, it does not stand alone but must be regarded as one of a set of conceptions—this set describable on a dimension of congruence or compatibility. The belief that a role is performed better if its requirements are consonant with one’s “natural inclinations” comes to us from folk psychology. By translating “natural inclinations” to self concept, or self characteristics, the folk belief can be translated into a scientifically testable hypothesis, namely, that the effectiveness or validity of role enactment is related to the degree of congruence of self and role.
In brief, the self may be defined as the residues of a human being’s transactions with objects and events, including other people. These residues are the referents for the symbol I (Sarbin 1954). To assess self structures, a large number of procedures have been devised, including autobiographies, speech and handwriting samples, responses to unstructured inkblots, endorsements of personality test items, and adjective check lists. In our linguistic system the most frequently used device for talking about self characteristics is the adjective. A person can describe his self as a catalogue of traits through the use of lists of adjectives, such as friendly, quiet, calm, vain, confident. The adjective check list, widely used in research settings, is based on this observation. Another widely used procedure in contemporary psychology is the personality inventory. This instrument is made up of incomplete I-statements requiring the respondent to indicate various predicates with himself as the subject, in this way expressing his preferences, attitudes, feelings, beliefs, values, and/or opinions. The dimensions used for organizing such sentences into scales are constructed for particular purposes by the investigator or distilled through analytic and statistical procedures. [SeePersonality measurement, article Onpersonality inventories; Self concept.]
Through the use of adjective check lists and inventories of I-statements one can systematically specify the qualitative requirements of a role (for example, a club treasurer must be honest and efficient) and compare such requirements with the self characteristics of persons who are assigned the role (for example, George sees himself as honest and efficient). When self and role characteristics are congruent, a convincing performance is expected, other things being equal. When self and role characteristics are incongruent or antagonistic, the enactment is likely to be invalid and unconvincing.
A study reported by Smelser (1961) illustrates the use of the variable of self-role congruence. Two samples of experimental subjects were selected, the members of one group having high scores on the trait “dominance,” derived from the California Psychological Inventory, the members of the second having low scores (see Gough 1957). The subjects, in pairs, were placed in a cooperative work situation and were systematically assigned roles calling for dominant or submissive enactments. The most productive pairs were those where dominant subjects had been assigned the dominant role and nondominant subjects the submissive role. The least productive pairs were those where the self characteristics and role characteristics had been reversed.
Employment of the self-role congruence variable has been especially useful in predicting effective enactment in the hypnosis role. Sarbin and Andersen (1964) have reported experiments whose results could be interpreted as showing that persons whose self characteristics are similar to the requirements of the role of hypnosis subject enact that role with greater convincingness. Their procedure involved the construction of an inventory of I-statements whose items reflected the kinds of role behaviors traditionally associated with hypnosis. Two of the dimensions of the role are (1) to become absorbed, to concentrate, and (2) to accept readily altered attentional and mood experiences. Illustrative of the /-statements written to tap corresponding self characteristics are the following statements: “I sometimes become so absorbed in a task that I am forgetful of other, less important aspects of the daily routine”; “I would not be afraid of anything that would temporarily change my awareness of my surroundings.” The study yielded significant correlations between scores on the self-assessment inventory and the degree of convincingness in the enactment of the role.
An alternate way of regarding this variable is available. Where the self becomes involved in the role, the enactment is more likely to be judged as valid, convincing, and proper. It has been suggested that forced compliance to a role whose requirements are incompatible with one’s self characteristics will not produce changes in the self. For example, attempts to indoctrinate prisoners of war in alien ideology generally fail because of lack of self-involvement in the role, even though the prisoners may give the appearance of involvement through the use of histrionic talents. [SeeBrainwashing.]
Reinforcement properties of the audience
Unlike many current attempts to understand social psychological phenomena, role theory addresses itself to continuities in conduct rather than to small samples of outcome behaviors. Social roles are enacted over time: the performances are protracted. To employ the dramaturgical metaphor further, roles are enacted before audiences—sometimes a one-person audience, as in psychoanalytic therapy; sometimes a small group audience, as in a work or discussion group; sometimes a large mass audience, as at political meetings, the theater, and professional football games; and sometimes a symbolic audience, as in a writer’s imagined picture of his reading audience. Audiences may serve two functions, providing cues helpful to the actor in locating his role and providing social reinforcements or sanctions. The plethora of studies on operant conditioning may be taken as a paradigm for the reinforcement properties of the audience (Krasner & Ullman 1965). In operant conditioning, selected responses of the subject are reinforced through social reinforcements, such as approval, praise, eye contact, friendly gesture, and nodding, or through more direct reinforcements, such as food and money. In the usual case of social interaction, audiences primarily provide social reinforcements. The effect of positive reinforcements, other things being equal, is to encourage the actor to continue his performance and to think that it is convincing, valid, and proper. Negative reinforcements, on the other hand, communicate to the actor that his role enactment is invalid and discourages the continuation of the role. [SeeCommunication, mass, especially the article onaudiences; Learning, articles oninstrumental learningandreinforcement.]
This article has attempted to identify current conceptions that may be organized under the general heading of role theory. The outcome variable of interest to psychologists whose work is guided by this orientation is role enactment, that is, the overt conduct of a person in his effort to validate the occupancy of the positions to which he is assigned in various macrostructures and microstructures. Role enactments are judged by audiences along various dimensions, in general reflecting the convincingness, propriety, effectiveness, or validity of the enactment. Because role enactments vary in these dimensions, investigators have attempted to isolate and identify the sources of variation. Six components of effective role enactment have been described and at least one empirical study introduced to illustrate the utility of each variable. The validity, effectiveness, or convincingness of role enactment depends on (1) a valid set of role expectations, (2) an accurate assessment of the role of the other (and reciprocally the self), (3) sensitivity to the role demands of the situation, (4)role-relevant cognitive and motoric aptitudes, (5)self characteristics congruent with the role, and (6) positive reinforcements for appropriate acts. The six variables are listed as if they were independent of one another. As yet they can be designated only as conceptually independent; further research will tell to what extent they are empirically independent—and under what conditions.
Theodore R. Sarbin
[Other relevant material may be found inPerception, articles onperson perceptionandsocial perception; Sociometry; Status, social; Stratification, social; Sympathy and empathy; and in the biography ofMead.]
Cameron, Norman A. 1947 The Psychology of Behavior Disorders: A Biosocial Interpretation. Boston: Houghton Mifflin.
Cline, Victor B.; and Richards, James M. Jr. 1960 Accuracy of Interpersonal Perception: A General Trait? Journal of Abnormal and Social Psychology 60:1–7.
Gage, Nathaniel L. 1953 Accuracy of Social Perception and Effectiveness in Interpersonal Relationships. Journal of Personality 22:128–141.
Goffman, Erving 1961 Encounters: Two Studies in the Sociology of Interaction. Indianapolis, Ind.: Bobbs-Merrill.
Gough, Harrison G. 1948 A Sociological Theory of Psychopathy. American Journal of Sociology 53:359–366.
Gough, Harrison G. 1957 The California Psychological Inventory: Manual. Palo Alto, Calif.: Consulting Psychologists Press.
Krasner, Leonard; and Ullman, Leonard P. 1965 Research in Behavior Modification: New Developments and Implications. New York: Holt.
Krasner, L.; Ullman, L.; and Weiss, R. L. 1964 Studies in Role Perception. Journal of General Psychology 71:367–371.
Moreno, Jacob L. (1934) 1953 Who Shall Survive? Foundations of Sociometry, Group Psychotherapy and Sociodrama. Rev. & enl. ed. Beacon, N.Y.: Beacon House.
Orne, Martin T. 1959 The Nature of Hypnosis: Artifact and Essence. Journal of Abnormal and Social Psychology 58:277–299.
Orne, Martin T.; and Scheibe, Karl E. 1964 The Contribution of Nondeprivation Factors in the Production of Sensory Deprivation Effects: The Psychology of the “Panic Button.” Journal of Abnormal and Social Psychology 68:3–12.
Sarbin, Theodore R. 1943 The Concept of Role-taking. Sociometry 6:273–285.
Sarbin, Theodore R. 1950 Contributions to Role-taking Theory: I. Hypnotic Behavior. Psychological Review 57:255–270.
Sarbin, Theodore R. 1954 Role Theory. Volume 1, pages 223–258 in Gardner Lindzey (editor), Handbook of Social Psychology. Cambridge, Mass.: Addison-Wesley.
Sarbin, Theodore R.; and Andersen, M. L. 1964 Base Rate Expectancies and Motoric Alterations in Hypnosis. International Journal of Clinical and Experimental Hypnosis 12:147–158.
Sarbin, Theodore R.; and Hardyck, Curtis D. 1955 Conformance in Role Perception as a Personality Variable. Journal of Consulting Psychology 19:109–111.
Sarbin, Theodore R.; and Jones, D. S. 1955 An Experimental Analysis of Role Behavior. Journal of Abnormal and Social Psychology 51:236–241.
Sarbin, Theodore R.; and Lim, Donald T. 1963 Some Evidence in Support of the Role-taking Hypothesis in Hypnosis. International Journal of Clinical and Experimental Hypnosis 11:98–103.
Smelser, William T. 1961 Dominance as a Factor in Achievement and Perception in Cooperative Problem Solving Interactions. Journal of Abnormal and Social Psychology 62:535–542.
The concept of role, borrowed from the stage, has been central in those sociological analyses which seek to link the functioning of the social order with the characteristics and behavior of the individuals who make it up. The concept has a relatively long history in American sociology and has attracted interest from some of its best minds. For instance, Park (1926, p. 137) noted that “everyone is always and everywhere, more or less consciously, playing a role…. It is in these roles that we know each other; it is in these roles that we know ourselves.” In this and most subsequent usages, the following elements appear in the definition of role: it provides a comprehensive pattern for behavior and attitudes; it constitutes a strategy for coping with a recurrent type of situation; it is socially identified, more or less clearly, as an entity; it is subject to being played recognizably by different individuals; and it supplies a major basis for identifying and placing persons in society.
George H. Mead’s observation (1934) that in mature social behavior the individual works out his own role by imaginatively taking the role of the other soon led to the idea that articulation between the roles played by partners in interaction determines whether interaction is harmonious and productive. The theme that personal and social disorganization arises from failure of the individual to choose between incompatible roles was developed by Mowrer (1935), and the problem of balancing rights and obligations in role articulation was explored by Kirkpatrick (1936). In the same year, Lin ton (1936) presented the first systematic statement identifying role as a segment of culture; he held that it consisted of behavioral prescriptions or norms bearing one-to-one correspondence with social status [SeeStatus, social; the biography ofLinton]. All these traditions have inspired descriptive studies, stressing personal and organizational strains from incomplete articulation of roles.
More recently such topics as the spontaneous emergence of similar informal roles in different discussion groups (Benne & Sheats 1948) and the hypothesis of role specialization in task and social-emotional behavior (Bales 1950) have been the subject of experimental study. Many organizational studies have demonstrated that lack of clarity and consensus in role conceptions is a factor in reducing organizational effectiveness and morale (see, for instance, Gross et al. 1958). One serious effort has been made to review and appraise this body of literature comprehensively (Rocheblave-Spenle 1962).
This variety of studies is an index of some major differences in usage. A few writers equate “role” with the actual behavior of an individual, although most distinguish between role as expected behavior or conceptions and role behavior as enactment. Furthermore, there is a continuing issue of generality, raised by Linton’s restriction of role to the prescriptions associated with a status (1936). Another influence is that of behaviorism, which led to an effort to think of a role as an inventory of specific behaviors or conditioned responses (Cottrell 1942), rather than as certain purposes, sentiments, and strategies in the context of which both routinized and improvised behavior are mixed. Authorities also differ on whether “role” should be said to consist only of the norms applicable to a position or should include not only norms but also expectations and conceptions with respect to behaviors that are only in the faintest degree normative. Finally, role is defined by some as essentially a part to be learned and played, whereas for others it is primarily a way of coping with an imputed other-role.
Emergence of roles
Despite definitional differences, it is possible to state, as a series of rough “postulates,” the tendencies that are usually assumed to constitute the normal role processes. Role theory will first be considered as it would apply to a self-contained, dynamic system of interacting roles; considerations of organizational and societal setting, as well as of individual personality, will be discussed later.
Role differentiation and accretion
In any interactive situation, behavior, sentiments, and motives tend to be differentiated into units, which can be called roles; once roles are differentiated, elements of behavior, sentiment, and motives that appear in the same situation tend to be assigned to the existing roles. This postulate suggests that one should study degrees of differentiation, as indicated by clarity of role definition and degree of role overlap. The hypothesis has been widely accepted that individual adjustment to role and group effectiveness is positively related to clarity and negatively related to overlap, although excessive clarity may be overly restrictive to the individual and rigidifying to the organization, and some overlap may facilitate empathic understanding and contribute to interpersonal bonds. The postulate also suggests that attention should be given to what is grouped and what is separated into different roles. Internal consistency of roles is thought to facilitate individual adjustment and group efficiency; internal contradiction is held to be explained with the help of such concepts as role set (discussed below), culture contact and assimilation, imposition of role structure by authority, and rigidification through formalization and tradition. Feasibility of enactment by a single individual has been the most common principle invoked to explain the peculiar way in which the content of one role can be separated from that of another; additional principles, such as those arising from the functional requirements of the group, are noted below.
In any interactive situation the meaning of individual actions for ego (the actor) and for any alters (partners in interaction) is assigned on the basis of the imputed role. This postulate is inspired by the common observation that identical actions mean totally different things when viewed as expressions of different roles.
A tendency (never completely realized) for the social character of a role to shift toward correspondence with patterns of behavior is actually seen to take place within the role context. For example, changes in the wife and mother role have been attributed to removal of many traditional domestic activities from the home, despite inertial factors that caused a lag in the adaptation.
Every role tends to acquire an evaluation, both in terms of rank and in terms of its socially favorable or detrimental character. The perceived utility of the role, the power resident in the role, and association of the role with other valued roles are among hypothesized determinants of role evaluation.
Role as interactive framework
The establishment and persistence of interaction tend to depend on the emergence and identification of ego and alter roles. This postulate extends the meaningfulness postulate to interaction. Identifications are tentative and are based on cues, so the character of interaction in any initial encounter, or under grossly changed conditions, is tentative and changeable.
Each role tends to form as a comprehensive way of coping with one or more relevant alter roles. The actual character of any given role involves a constant strain toward equilibrium between actual behavior, evaluative adjustments between roles, and the tension of maintaining each as a viable role in relation to the other. For example, in any role relationship of superordination-subordination, such as parent-child, teacher-student, master-servant, policeman-criminal, the subordinate role will incorporate mechanisms that afford some latitude for action.
There is a tendency toward consensus regarding the content of roles in interaction. Because of the adaptation that takes place as roles form, some consensus regarding what enactors of each role are to do seems inevitable, although the degree of consensus required for a viable interaction system is not great. Role dis-sensus has generally been thought an obstacle to harmonious interaction (as in the family), although the evidence is not entirely consistent.
There is a tendency for stabilized roles to be assigned the character of legitimate expectations, implying that deviation from expectation is a breach of rules or violation of trust. Willard Waller’s observations (1938) about the sense of injustice that arises in married life when one partner deviates from what has come to be the customary pattern of behavior suggest the following hypothesis: the degree to which ego ascribes legitimacy to his expectations for alter’s role varies directly with the degree to which the selection and enactment of his role depend on alter’s role and role behavior. The evaluation of the role also appears to make a difference, suggesting this principle: to the extent to which alter’s role is positively valued with respect to rank and favorableness, ego will translate his anticipations regarding alter’s role into legitimate expectations. Thus an unfavorably evaluated role, such as criminal or indigent, evokes legitimate expectations of continuance in the role only to the extent of justifying adaptive behavior by relevant others, such as surveillance by police and anticipatory defensive behavior.
Role and actor
Once stabilized, the role structure tends to persist, regardless of changes in the actors. This postulate takes two specific forms. First, when an actor leaves the group and is replaced by another, there is a tendency to allocate to the new member the role played by the one who leaves. Second, if one actor changes roles, there is a tendency for another actor to make a compensatory change of roles in order to maintain the original role structure. The latter principle has been referred to as role appropriation by Perry, Silber, and Bloch (1956), who noted that in some families, when the parent became disorganized and assumed a childlike role of dependence in a disaster situation, a child suddenly blossomed into responsibility and helped to supply family leadership.
There is a tendency for a given individual to be identified with a given role, and a complementary tendency for an individual to adopt a given role, for the duration of the interaction. The postulate of role allocation, in combination with the meaningfulness postulate, suggests that an actor’s behavior is without clear meaning until he settles on a particular role in interaction. In combination with the postulate of legitimate expectation, role allocation suggests that an individual whose behavior cannot be fitted into an identifiable role will be regarded as violating the terms of interaction.
A role allocation has taken place when relevant alters interact with ego on the basis of the same role as he is playing. Ego’s act of attempting to determine the role that alter will play has been dubbed “altercasting” by Weinstein and Deutschberger (1963). The alter’s use of a relatively unambiguous response to confer the right to play the chosen role is sometimes designated as validation.
The bases for role allocation can be inferred from the postulate of evaluation in conjunction with the postulate of role complementarity (see above). Since roles are evaluated, there is likely to be competition for the preferred roles, and the principles of any competitive situation will be at work. Because of complementarity of roles, it is essential that an individual be able to play the role with sufficient adequacy; demonstrated adequacy or inadequacy therefore affects the willingness of others to allocate to him a particular role.
To the extent to which ego’s role is an adaptation to alter’s role, it incorporates some conception of alter’s role. If the actor is guided chiefly by his conception of alter’s role (role taking), his behavior has an element of improvisation about it; he simply follows the course of action that seems reasonable for coping with alter under the circumstances. A further implication of the role-taking postulate is that the individual learns roles in pairs or sets, not singly.
Role behavior tends to be judged as “adequate” or “inadequate” by comparison with some definite conception of the role in question. When combined with the role allocation postulate, this observation leads to the following generalization: to the degree to which a role is favorably evaluated, judgments of role adequacy tend to be translated into evaluations of the person playing the role.
The degree of role adequacy legitimately expected of ego tends to be a function of the degree of role adequacy attributed to alter. Conversely, the degree to which ego can legitimately claim the privileges of his role tends to be a function of his degree of role adequacy. With this postulate, which has been elaborated by Gould-ner (1960), we call attention to two facts, namely, that role behavior does not normally correspond precisely with role conceptions, and that legitimate expectations are adjusted in the course of interaction.
The organizational setting supplies both direction and constraint to the working of the processes already outlined, thus bringing them into more complex interrelationships.
Organizational goal dominance
To the extent to which roles are incorporated into an organizational setting, organizational goals tend to become the crucial criteria for role differentiation, evaluation, complementarity, legitimacy of expectation, consensus, allocation, and judgments of adequacy. To the extent to which the organization has well-defined goals, the most salient role differentiation will be according to different kinds of tasks that together accomplish the goal. The expectations assigned greatest legitimacy and held with most consensus will tend, under these circumstances, to be those with most obvious task relevance; judgments of role adequacy will therefore weigh task adequacy more heavily than other aspects of role adequacy.
Legitimate role definers
To the extent to which roles are incorporated into an organizational setting, the rights to define the legitimate character of roles, to set the evaluations on roles, to allocate roles, and to judge role adequacy tend to be lodged in particular roles.
To the extent to which roles are incorporated into an organizational setting, role differentiation tends to link roles to statuses in the organization.
To the extent to which roles are incorporated into an organizational setting, each tends to develop as a pattern of adaptation to multiple alter roles. The teacher role must incorporate tenable adaptations to pupils, parents, other teachers, and principals, as well as to less salient alters. Discrepant expectations may arise from relations with each of the alters; Merton (1957) describes several mechanisms that are employed to minimize the conflict in the role set.
To the extent to which roles are incorporated into an organizational setting, their persistence is intensified through tradition and formalization. A strain then develops between formalized or traditional role definitions and the informal role structure.
Because society is built on accommodation among many organizations, it introduces multiple organizational referents for roles and multiple roles for the actor.
Economy of roles
Similar roles in different contexts tend to become merged, so as to be identified as a single role recurring in different relationships. As a corollary, roles in situations of limited generality and social significance tend to be shaped in accordance with the pattern of roles in situations of greater generality and social significance. One index of this tendency is that we often speak of male and female roles, of heroic and unheroic roles, of judge and defendant, when seeking meaning and order in quite simple interaction.
To the extent to which roles are referred to a societal context, differentiation tends to link roles to social values. Status is not thereby made less important as a point of anchorage, but social values take the place of organizational goals as major anchorages for roles. Values thus supply one of the major bases for the simplification of an otherwise unmanageable profusion of recognized roles in society; even in more limited contexts, as we have seen in the discussion of role economy, roles for interaction tend to be fitted into a framework of social values.
The individual in society tends to be assigned and to assume roles that are consistent with one another. The role conflict which arises from failure of this process has been subject to considerable study. Mechanisms such as organizational compartmentalization facilitate the tendency; there are also various recognized transitionl cues, as when a functionary who wishes to pass from one role to another speaks of “putting on a different hat.” Perhaps the most specific mechanism for achieving consistency is the tendency for roles that have the greatest generality, such as age, sex, and occupation, to serve as qualifying criteria for the allocation of other roles.
Role and person
When we turn to the consequences of role process for the individual as a sociological entity, we look at the actor in relation to a role-in-situation, and we observe the person in the setting of society. For the actor the primary concern is coping with strain; for the person the dynamics hinge on the management of the several roles in his repertoire.
The actor tends to act so as to alleviate role strain arising from role contradiction, role conflict, and role inadequacy and to heighten the gratifications of high role adequacy. Role strain comes largely from failure of the many processes already outlined to function adequately, so as to leave unclear, incomplete, and contradictory elements in a role. This may come about through failure of role cues, gross lack of role consensus, and so forth. Little has been noted that is distinctive about adjustments to role strain, as compared to ways of coping with any other type of stress.
The individual in society tends to adopt a repertoire of role relationships as a framework for his own behavior and as a perspective for the interpretation of the behavior of others. It is important to remember that socialization is partly a matter of learning somewhat culturally standardized role conceptions and partly a matter of acquiring a viable image of the social world by means of the tentative and adaptive processes that make up role enactment.
The individual tends to form a self-conception by selective identification of certain roles as more characteristically himself than other roles in his repertoire. The self-conception, from a sociological point of view, is not a set of traits but an organization of roles that both articulates the person with the society and incorporates role evaluations. On the basis of the roles included, the individual is said to develop a sense of personal prestige, which is likely to be reflected in his bearing, his self-assurance, and other aspects of his interpersonal relations. On the basis of his perceptions of role adequacy in these most “ego-involved” roles (Sherif & Cantril 1947), the individual develops a self-esteem.
Adaptivity of self-conception
The self-conception tends to stress those roles which supply the basis for effective adaptation to relevant alters. Thus, while there is no simple or uniform self-conception held by all persons defined by society as “criminal,” the established criminal must somehow take account, in his self-conception, of the fact that his criminal identification is recurrently the most salient consideration shaping the behavior of others toward him in a variety of situations.
To the extent to which the person must play roles that contradict his self-conception, these roles will be assigned “role distance,” together with mechanisms for demonstrating the lack of personal involvement that the person feels when playing these roles (Goffman 1961).
Ralph H. Turner
[Directly related are the entriesIdentity, psychosocial; Interaction; Norms; Status, social. Other relevant material may be found inPersonality measurement, article onsituational tests; Social Psychology; Socialization; and in the biographies ofLinton; Mead; Park; Waller.]
Bales, Robert F. 1950 Interaction Process Analysis: A Method for the Study of Small Groups. Reading, Mass.: Addison-Wesley.
Banton, Michael P. 1965 Roles: An Introduction to the Study of Social Relations. London: Tavistock.
Benne, Kenneth D.; and Sheats, Paul 1948 Functional Roles of Group Members. Journal of Social Issues 4, no. 2:41–49.
Biddle, Bruce J.; and Thomas, Edwin J. (editors) 1966 Role Theory: Concepts and Research. New York: Wiley.
Cottrell, Leonard S. Jr. 1942 The Adjustment of the Individual to His Age and Sex Roles. American Sociological Review 7:617–620.
Goffman, Ervtng 1961 Encounters: Two Studies in the Sociology of Interaction. Indianapolis, Ind.: Bobbs-Merrill.
Goode, William J. 1960 A Theory of Role Strain. American Sociological Review 25:483–496.
Gouldner, Alvin W. 1960 The Norm of Reciprocity: A Preliminary Statement. American Sociological Review 25:161–178.
Gross, Neal; Mason, Ward S.; and Mceachern, Alexander W. 1958 Explorations in Role Analysis: Studies of the School Superintendency Role. New York: Wiley.
Kirkpatrick, Clifford 1936 The Measurement of Ethical Inconsistency in Marriage. Ethics 46:444–460.
Linton, Ralph 1936 The Study of Man: An Introduction. New York: Appleton.
Mead, George H. 1934 Mind, Self and Society From the Standpoint of a Social Behaviorist. Edited by Charles W. Morris. Univ. of Chicago Press.
Merton, Robert K. 1957 The Role-set: Problems in Sociological Theory. British Journal of Sociology 8:106–120.
Moment, David; and Zaleznik, Abraham 1963 Role Development and Interpersonal Competence: An Experimental Study of Role Performance in Problem-solvingGroups. Boston: Harvard Univ., Graduate School of Business Administration, Division of Research.
Mowrer, Harriet 1935 Personality Adjustment and Domestic Discord. New York: American Book.
neiman, Lionel J.; and Hughes, James W. 1951 The Problem of the Concept of Role: A Re-survey of the Literature. Social Forces 30:141–149.
Park, Robert E. 1926 Behind Our Masks. Survey (New York) 56:135–139.
Parsons, Talcott 1951 The Social System. Glencoe, 111.: Free Press.
Perry, Stewart E.; Silber, Earle; and Bloch, Donald A. 1956 The Child and His Family in Disaster: A Study of the 1953 Vicksburg Tornado. National Research Council, Committee on Disaster Studies, Disaster Study No. 5. Washington: National Academy of Sciences-National Research Council.
Rocheblave-SpenlÉ, Anne Marie 1962 La notion de role en psychologie sociale: Étude historico-critique. Paris: Presses Universitaires de France.
Rommetveit, Ragnar (1953) 1955 Social Norms and Roles: Explorations in the Psychology of Enduring Social Pressures With Empirical Contributions From Inquiries Into Religious Attitudes and Sex Roles of Adolescents From Some Districts in Western Norway. Oslo: Akademisk Forlag; Minneapolis: Univ. of Minnesota Press.
Sherif, Muzafer; and Cantril, Hadley 1947 The Psychology of Ego-involvements, Social Attitudes and Identifications. New York: Wiley; London: Chapman Hall.
Slater, Philip E. 1955 Role Differentiation in Small Groups. American Sociological Review 20:300–310.
Strodtbeck, Fred L.; and Mann, Richard D. 1956 Sex Role Differentiation in Jury Deliberations. Sociometry 19:3–11.
Wwaller, Willard (1938) 1951 The Family: A Dynamic Interpretation. Revised by Reuben Hill. New York: Dryden.
Weinstein, Eugene A.; and Deutschberger, Paul 1963 Some Dimensions of Altercasting. Sociometry 26:454— 466.
Znaniecki, Florian 1965 Social Relations and Social Roles: The Unfinished Systematic Sociology. San Francisco: Chandler. -” Published posthumously.
There are two rather different approaches within role theory. One develops the social anthropology of Ralph Linton and gives a structural account of roles situated within the social system. Here, roles become institutionalized clusters of normative rights and obligations: Talcott Parsons's celebrated account of the sick role is a good example. An alternative approach is more social-psychological in tenor and focuses upon the active processes involved in making, taking, and playing at roles: it is part of the traditions of symbolic interactionism and dramaturgy, the latter of which analyses social life through the metaphor of drama and the theatre.
The structural account of roles locates a status in society, such as that of a teacher, and then tries to describe the standard bundle of rights and duties associated with an ideal type of this position. These expectations, which are socially based, constitute the role. Any given person will possess a number of statuses (for example mother, teacher, golf-captain), and these constitute a status set, with each status harbouring its own role. Every role brings a number of different partners, each with their own set of expectations, so that a teacher (for example) may have students, colleagues, heads, governors, and parents as role partners, each of whom makes somewhat different expectations upon his or her behaviour. The sum total of the expectations of these partners is the role-set. When these expectations are in disagreement, as is frequently the case, sociologists talk of role conflict and role strain. In the Parsonsian system of social theory, these role patterns are defined through the so-called pattern variables, or choices between pairs of alternative norms. This theory is a useful heuristic device for mapping the organization of societies through normative patterns, but it does have a tendency to oversimplify normative expectations by assuming too much consensus in society, and by a reification of the social system. A sophisticated version will be found in Ralf Dahrendorf's Homo Sociologicus (1968), at one time controversial, now unwarrantably neglected.
The contrasting social psychological view is focused much more upon the dynamic aspects of working at roles: it examines the interactions in which people come to play their roles rather than describing the place of these roles in the social structure. Here, the emphasis is on the ways in which people come to take the role of the other (role-taking), construct their own roles (role-making), anticipate the responses of others to their roles (altercasting), and finally play at their particular role (role-playing). In some versions of this theory (for example that propounded by Erving Goffman), attention is given to the ways in which roles are performed: sometimes people may embrace their parts fully (role embracement) and play out the details of their role in cherished detail. At other times they may perform their parts with tongue-in-cheek (role distance)—showing that they are much more than the simple role they play. Or they may play roles cynically in order to manage the outcomes of the situation (impression management). In all of these accounts the concern is with the dynamics of working at roles, where roles are not fixed expectations, but emergent outcomes. Perhaps the most useful accounts of this approach to role theory are to be found in Goffman's The Presentation of Self in Everyday Life (1959) and Encounters (1961).
Role theory is certainly not the prerogative of sociologists. The notion of analysing social life through the theatrical metaphor is obvious in Greek theatre, in Shakespeare's declaration that ‘All the world's a stage’, and in modern notions of ‘theotocracy’ such as are introduced in Stanford Lyman and Marvin Scott's The Drama of Social Reality (1975). Louis A. Zurcher's Social Roles (1983) is still a good introduction to the field.
role / rōl/ • n. an actor's part in a play, movie, etc.: Dietrich's role as a wife in war-torn Paris. ∎ the function assumed or part played by a person or thing in a particular situation: she greeted us all in her various roles of mother, friend, and daughter religion plays a vital role in society.