Denzin ASR Symbolic Interaction Ism and Ethnomethodology

Symbolic Interactionism and Ethnomethodology: A Proposed SynthesisAuthor(s): Norman K. DenzinSource: American Sociological Review, Vol. 34, No. 6 (Dec., 1969), pp. 922-934Published by: American Sociological AssociationStable URL: http://www.jstor.org/stable/2095982 .Accessed: 10/07/2011 01:47

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=asa. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access toAmerican Sociological Review.

http://www.jstor.org

SYMBOLIC INTERACTIONISM AND ETHNOMETHODOLOGY: A PROPOSED SYNTHESIS *

NORMAN K. DENZIN

University of California, Berkeley

The basic theoretical and methodological assumptions of symbolic interactionism and ethno- methodology are compared and points of synthesis are proposed. Similarities between the two orientations are noted, and these are seen to involve the problems of social organization, methodology, socialization, deviance, social control, face-to-face interaction, and the analysis of science as a social enterprise. It is suggested that these perspectives offer a much needed view of how individuals are shaped by and, in turn, create elements of social structure. Be- cause of their emphasis on the subjective side of social life, interactionism and ethnomethod- ology warrant serious consideration for their contributions to an alternative view of the individual and his social arrangements. Areas of empirical inquiry relevant to both points of view are stressed and a number of hypotheses are offered for future research. Such research, it is proposed, will shed light on what are now taken by many as irreconcilable diff erences between these perspectives.

THE development of a theoretical per- spective appropriate for the joint analy- sis of social psychological and socio-

logical problems has long concerned the sociologist. The methodology that would permit such an analysis has also remained an issue. Although various alternatives have been offered, ranging from the use of models taken from economics and psychology to structural-functionalism, none has proven completely satisfactory. My intent is to take two perspectives in contemporary so- ciology, one old and one relatively new, and to examine their potential for meeting the above issues. Specifically, I shall examine symbolic interactionism and ethnomethod- ology. Because both focus in some way on the individual, they provide a view of social organization that may be termed subjective and social psychological in nature. Analysis of the degree of convergence between the two should permit an expanded treatment of how individuals are linked to, shaped by, and in turn create social structure. These two perspectives are especially relevant to

the above problems because they also pro- pose special views of methodology.

THE PERSPECTIVES DEFINED

The ethnomethodology of Garfinkel (1967) and Cicourel (1968) proposes an analysis of the routine, taken-for-granted expectations that members of any social order regularly accept. Basic to this perspective is the at- tempt to sharply distinguish scientific from everyday atcivity. The problems of penetrat- ing everyday perspectives and giving them sociological explanations are repeatedly ad- dressed and the method of documentary anal- ysis is set forth as a preferred strategy. The abiding concern, however, is with the rela- tionship between everyday, taken-for-granted meanings, and the organization of these meanings into routine patterns of interaction.

Symbolic interactionism takes as a funda- mental concern the relationship between in- dividual conduct and forms of social organ- ization. This perspective asks how selves emerge out of social structure and social situations.

Both perspectives posit a link between the person and social structure that rests on the role of symbols and common meanings. To this extent they share a great deal in com- mon with the structural-functional perspec- tive. Locating the unit of analysis in the individual and interaction separates inter- actionism and ethnomethodology from other nointq of view.

* I am indebted to a number of colleagues and students for their critical reactions to earlier ver- sions of this essay, and especially to Herbert Blumer for his critical remarks. I am also grateful for the comments and criticisms of Howard S. Becker, Carl J. Couch, Harold Mark and the students in my 426 seminar on deviance at the University of Illinois.

What follows is my proposed synthesis of inter- actionism and ethnomethodology. This is not in- tended as an essay expressing widespread consensus as consensus is probably not Dossible at this time.

922

SYMBOLIC INTERACTIONISM 923

SYMBOLIC INTERACTIONISM

The interactionist assumes that human beings are capable of making their own thoughts and activities objects of analysis, that is, they can routinely, and even habitu- ally, manipulate symbols and orient their own actions towards other objects. A great deal of human conduct is of this routine na- ture. Once the meanings of objects have been agreed upon, conduct can flow along lines of custom, tradition, and ritual.

Because humans also possess the ability to self-consciously direct their own activities, the interaction process may be classified into those behaviors that are routinely organized and those that are actively constructed in a self-conscious and interpretative fashion (Blumer, 1966:537-538). Granted this as- sumption, a fundamental empirical question becomes the identification of the shifting modes of interpretation that characterize the interaction process. Clearly, interaction can- not be so grossly divided into either inter- pretative, or noninterpretative elements. There are many levels and shades of differ- ence between these two, and the extent to which action and objects move between these points remains to be identified.

Man's environment does not consist of ob- jects which carry intrinsic meaning. Social objects are "constructs, and not self-existing entities with intrinsic natures" (Blumer, 1966:539). Objects consist of any event that one can designate in a unitary fachion and around which one can organize action. The meaning of an object resides not in the object itself but in the definitions brought to it, and hence must be located in the interaction process.

These meanings typically derive from a group or organized interactional perspective. Human life is group life, and concerted ac- tion arises out of the ability of persons to be objects of both their own activity and other's. Joint actions, which represent the generic form of all interaction, rest on the ability of the human to grasp the direction of the acts of others (Blumer, 1966). For consensual lines of action to emerge there must exist a common community of symbols. Because the definitions of certain objects within a group's perspective are subject to continual negotia- tion, at least certain features of group life

are subject to negotiation and change. The basic object for all interaction is the self. Because the self carries a multitude of differ- ing interpretations, shifts in these definitions often give group life its changing character.

When selves are consensually defined, sta- ble patterns of action will be observed. At the heart of group life lies a series of social selves that have been lodged in that struc- ture. Through the process of self-lodging, humans translate crucial features of their own identity into the selves, and into the memories and imaginations of relevant others. In this way Cooley's proposition that the other exists in "our imaginations of him" comes to life. By lodging the self in interac- tion, and in the selves of others, a reciprocal bond is created, and the firm foundations for future relationships are established. Self- lodging stands in distinction to what Goff- man (1959) has termed the process of pre- senting a self. It is certainly the case that selves have to be presented, but at some point in the cycle of recurrent interactions, the self moves from the presentational to the lodging phase. Central to self-lodging are variations on Cooley's notion of the looking glass self, which posited a three-fold process of presen- tation, identification and subjective interpre- tation. Cooley profoundly noted that upon the presentation of a self, the person took on a sense of pleasure or displeasure based on his interpretations of the other's reactions. This suggests that the self is a continually evaluated object-an object that rests on affective-emotional reactions and criticisms. It is not an object devastated through criti- cism, nor is it a calculated and planned object.

This view suggests that one motivational feature of human conduct is the self-lodging process. Humans return to those interac- tional quarters where the most basic features of their selves have been lodged. This ex- tends Foote's ( 1951 ) proposition that identi- fication of self and other provides the basis of motivation. The self-lodging hypothesis suggests that after the identification process has been negotiated, a recurrent pattern of identification arises which rests on the re- ciprocal definitions given the lodged selves. Once identification has been established, it ceases to become problematic. The respective selves now offer their own vocabularies and

924 AMERTCAN SOCIOLOGICAL REVIEW

grounds for identification. In this way moti- vation becomes an interactional process sub- ject to finely graded phases. The self and its interpretations are grounded in interaction, and as selves become lodged, the styles of, and reasons for, interaction become ritualized and subject to the firm rules of tradition and custom.

This conception suggests that human in- teraction may not be as rational and cogni- tively directed as some theories suggest. If persons return to those quarters and settings where crucial aspects of the self have been lodged, then their actions cease to be di- rected entirely on either the most rational or most effective grounds. Self-lodging may place a wedge between the most economical or rational selection of a goal. The wedge rests on the affective bond between the self and its relevant others, and this bond can justify actions grounded in custom, love, hate, jealousy, or profound respect.

Empirical indicators of self-lodging include variations on personal names, styles of speech, modes of dress, and the use of spe- cial gestures and body movements. Names (Hughes, 1952:130-144; Stone, 1962: 86- 118) provide convenient clues to the degree of knowledge and involvement each person has with the other. As naming moves from formal titles to variations on either the title, or the first name, greater self-lodging may be inferred. This is similarly the case with variations on speech, dress, and gesturing. Each person develops his own style which be- lies an attempt to lodge certain portions of the self in the interaction. Another indicator of self-lodging is the degree of involvement, often time- or topic-wise in conversation. To the extent that any encounter represents ver- bal communication, each person's involve- ment in the conversation can be charted by the frequency of utterances. In this sense dominating or contributing to a conversation becomes akin to lodging a portion of the self in that interaction. As an additional hypothe- sis, I suggest that persons judge interactions to be satisfactory or unsatisfactory in terms of their success at self-lodging. If valued por- tions of the self are not lodged, recognized, and reciprocated, a dissatisfaction concerning that encounter is likely to be sensed.

It must be noted that joint actions cannot be resolved solely into individual lines of

action. When persons come together for the purpose of accomplishing a task, exchanging a greeting, eating a meal, or making love, the observed behavior involves more than the intentions and meanings brought into the situation. The term "interaction" suggests a central feature of all joint actions-an emer- gent quality that may not have existed be- fore the parties came together. The fitting together of individual lines of action provides the basic feature of the joint action. Indi- viduals fit lines of action together by identi- fying the action they are going to engage in, and then by fitting these definitions around the other's definitions and interpretations. Hence, the joint action becomes more than the mere juggling of definitions-it is the fitting of disparate, conflicting, and often incomplete plans of action into a package of meanings that, at least for the moment of activity, provide the basis of interaction. This feature of the joint action suggests that interaction may have a variable career. Ob- servers can not photograph the beginning phases of an encounter and assume that agreements reached in that phase will remain unchanged until the end. The career of a joint action is contingent on the events that occur during its life time. While participants may initially agree on definitions, rules of conduct, and images of self, these definitions may be so vague as to permit conflicting points of view to later emerge to challenge the entire basis of joint action. On other oc- casions, joint actions may be interrupted as when a child is brought into a marriage. The entry of this new object demands renegotia- tion, relocation of selves, and perhaps even the adoption of an entirely different perspec- tive so that joint action can continue.

A fundamental concern of the interaction- ist has been with the conditions that give rise to new perspectives, new points of view and new lines of action. Any social world (Mead 1927:75-85) is sustained by the for- mation of a perspective that embodies the basic forms of thinking and acting common to that world. It was in the face of conflict, or the confrontation of divergent perspec- tives that Mead located change and creative activity. When an old perspective failed to provide answers, new points of view were seen as emerging. This simple point provided Mead (1917:53-83) with his basic doctrine

SYMBOLIC INTERACTIONISM 925

for the scientific method. Unlike everyday man, who could be satisfied with faulty, in- complete, and contradictory perspectives, the scientist deliberately sought out conditions, events, and activities that challenged previ- ously accepted interpretations. In the dialec- tical process of confrontation, analysis and resynthesis of perspectives, Mead lodged the basic features of science.

By placing emphasis on conflict and on situations which demand new interpretation, Mead anticipated subsequent statements concerning the positive functions of conflict (e.g. Coser, 1956). He also offered a view of social groups which stressed (Thrasher, 1927) the functions of conflict in the forma- tion of self-conscious group perspectives. Mead hypothesized that unless collectivities met with challenges, distinctive elements of a group structure were unlikely to occur.

An additional feature of the joint action must be noted. I suggest that all forms of interaction rest on some combination of the following rules: (1) civil-legal codes that exist to protect the owners of objects, as well as the objects; these would include laws pro- hibiting violence to the self, or property, and are commonly upheld by specific authorities; (2) rules of etiquette, which exist to sustain the ceremonious occasions of interaction (Goffman, 1959, 1963); these rules find their way into etiquette books and cover problems of dress, introduction, leave-taking rituals, etc.; and (3) those rules which display the distinctive nature of enduring social relation- ships; these are relational rules of conduct which often derive from civil-legal, or polite codes, yet they serve to redefine and often make irrelevant other prescriptions. Such would be the case for rules sanctioning loud or profane language among co-workers, open states of undress among the married, and white collar theft. Relational rules define how the self is to be presented, and display the forms that self-lodging is to take. Agree- ment to use nicknames, to swear on occasion, to steal behind an employer's back, or to ignore certain clothing rules indicates the ways in which selves have moved beyond the presentational phase into various degrees of reciprocal lodging. A complete interactional study would include all three categories of rules and its central concern would be the extent to which persons actively construct

their own meanings for taken-for-granted standards of conduct. The distinctive nature of many social relationships is represented in their relational rules which make problem- atic events other collectivities assume, and take for granted rules that other's argue over.

The Methodological Assumptions of Interactionism

Because human interaction involves be- havior of both the covert and overt variety, and because the meanings attached to ob- jects often change during an encounter, the interactionist endeavors to relate covert symbolic behavior with overt patterns of interaction. This additionally demands a concern for the unfolding meaning objects assume during an interactional sequence. The usual strategy is to work from overt behaviors (Mead, 1934:1-8) back to the meaning attached to those behaviors and objects. This feature of interactionist meth- odology suggests that behavioral analyses alone (see for example Webb, et al., 1966) are insufficient to establish valid explana- tions of human conduct. Similarly, an analy- sis of the meanings, or definitions held by a set of persons and carried into a real or proposed interaction will not supply the needed link between those symbols and interaction. Thus, our first methodological principle asserts that covert and overt forms of conduct must be examined before an investigation is complete. Because this prin- ciple suggests that meaning shifts during interaction, a basic problem for research is the identification of interpretational phases. Studies must be conducted which determine at which point during an encounter objects cease to be negotiated.

A second principle focuses on the self as an object and a process. The investigator is directed to examine behavior from the perspective of those being studied, and he must indicate the shifting meanings and statuses assigned the self. At certain times the self ceases to be a negotiated object, assumes an agreed upon meaning, and inter- action then turns to other concerns. This may be observed in many ritual encounters where the basic activity lies above the self, or in the interaction process. Social games, routine work, and even participation in a

926 AMERICAN SOCIOLOGICAL REVIEW

religious ceremony represent such occasions. By making the self a central object of study, analysis can quickly establish what is taken for granted and what is problematic for the respective interactants. A commitment to this principle permits the researcher to es- cape the fallacy of objectivism which is the substitution of the scientist's perspective for those studied.

Taking the role of the acting other leads to a third principle. The researcher must link his subjects' symbols and meanings to the social circles and relationships that furnish those perspectives. Unless meanings are linked to larger social perspectives, analysis remains largely psychological. This suggests a two-step process for any study; meanings at both the individual and interactional levels must be examined.

A fourth methodological principle directs researchers to consider the "situated aspects" of human conduct. If behavior occurs within social situations and if the meaning attached to those situations influences subsequent be- havior, then the situation becomes a dimen- sion of analysis. Four components of the situation may be distinguished: the inter- actants as objects, the concrete setting, the meanings brought into the situation and the time taken for the interaction. Variations in behavior can arise from definitions given the respective selves, the other objects that constitute the situation (e.g. furniture, light- ing), the meanings and definitions for action that are held before interaction occurs, and the temporal sequencing of action.

The situation as an intrusive variable can- not be ignored. The entry of alien others, the failure of mechanical equipment (Gross and Stone, 1963), or shifts in levels of mu- tual involvement, all relate to interaction as a situated process. In this way concrete situations become both places for inter- action and objects of negotiation. It is im- possible to separate the two from situational analysis.

Because the interaction process is charac- terized by both stability and change, a fifth principle demands that research strategies be capable of reflecting both aspects of group life. Research methods can be judged by their ability to yield both kinds of informa- tion. Because of the interaction between the observer and his environment, we mention

parenthetically that the act of making ob- servations becomes symbolic and subject to personal bias and even ideological prefer- ence.

For the interactionist the preferred con- cepts are sensitizing. This does not mean that operationalization is avoided-it merely sug- gests that the point of operationalization is delayed until the situated meaning of con- cepts are discovered. At this point standard methods of observation can be employed. An- other feature of this process is the use of multiple methods of observation. Commonly termed triangulation, (Webb, et. al., 1966) this directs the researcher to utilize different tools in the observational process. This strat- egy assumes that no single method can ad- equately treat all the problems of discovery and verification. Each method has restric- tions, and if several different methods are combined in the same study, the restrictions of one are often the strength of another.

The triangulation process assumes the fol- lowing elements: a series of common data bases; a reliable sampling model that recog- nizes interaction; a series of empirical indi- cators for each data base; a series of hypotheses; and a continual reciprocation between data and hypotheses.

A final methodological principle relates to theory. Formal theory (Simmel, 1950) is a common goal of interactionist research. Al- though historically, or situationally, specific propositions are recognized, propositions with the greatest universal relevance are sought. This assumes that human affairs, wherever they occur, rest on the same interactional processes. Formal theory, in this sense, ex- tends Merton's view of middle-range theory to a position that calls for soundly grounded empirical propositions of an all-inclusive, universal nature. (See Glaser and Strauss, 1967). Properly speaking, the interactionist has not achieved this goal; interactionism re- mains a perspective or conceptual framework and is not a theory in the strict meaning of the term.

ETHNOMETHODOLOGY

The ethnomethodologist directs attention to the question of how a social order is possi- ble. For Garfinkel (1967) the answer merges a Durkheimian concern for large collective

SYMBOLIC INTERACTIONISM 92 7

representations with an interactionist concep- tion of the rules, norms, and meanings that members of any social order daily take for granted.

These rules, which any bonafide member of a social order is aware of, include the fol- lowing assumptions: (1) interaction flows in a temporal sequence and statements in any encounter cannot be understood without ref- erence to the actual flow of events; (2) per- sons in any situation will talk about many things that are only tacitly recognized, if at all; (3) normal background affairs and conditions in any situation are taken for granted and typically go unchallenged during an encounter; (4) once a situation is defined, this definition holds for the duration of the encounter; (5) any object present in the situation is what it is presented as being; (6) the meanings given an object on one occasion will hold for future occasions, sug- gesting that definitions of one set of inter- actants will be the same as those any other person or persons would develop were they in the same situation; (7) interactants identify and attach meaning to objects by the use of standard terms, symbols and labels; (8) while persons base their definitions of situations on their own bi- ographies and past experiences, any discrep- ancies that would arise in an encounter because of variations in biography or ex- perience are held in abeyance. In short, sit- uations are defined through the process of interaction; therefore, persons will often feel a degree of conflict between their public and private definitions.

A basic interest of the ethnomethodologist has been the penetration of normal situations of interaction to uncover these taken for granted rules. This is typically phrased in terms of how one could disrupt normal social events so that any person's conception of the normal, real, and the ordinary would be challenged. In Garfinkel's studies (1967:54) the common strategy has been to design quasi-experimental field studies in which three conditions are created. First, the situa- tion is structured so that the subject studied could not interpret it as a game, an experi- ment, a deception, or a play. Second, the sub- ject is given insufficient time to reconstruct the situation in his own terms. Third, he is given no aid in forging new definitions.

At several points Garfinkel reports experi- ments which meet the above three conditions. On one occasion students were asked to play as boarders in their own homes; on another, they were told to overpay and underpay for objects purchased in a store; in one experi- ment medical students were given discrepant information regarding an application for medical school; and in another study stu- dents were told to violate the usual rules of tick-tack-toe.

In all of these studies, which Garfinkel in- sists are only exploratory and illustrative, it was found that persons who act as "every- day experimenters" find it difficult to chal- lenge the routine rules of interaction. Feel- ings of distrust, hostility, anger, frustration, and persecution were reported by his student experimenters. The focus of interaction was soon lost when the "experimenter" attitude was assumed, and for all practical purposes the students were unable to carry on normal interaction. Garfinkel explains this inability with the concept of trust, which he defines as one's assumption that all others he en- counters will share the same expectations and definitions of the situation and that the other person will act on the basis of these assump- tions, even in problematic situations.

This concept suggests that when one or more interactants are forced to distrust the other, the normal background features of the situation suddenly become problematic, and the organization of joint action soon col- lapses. These experiments represent small- scale studies concerning the basis of collec- tive behavior. They also offer data on the interpretational phases of encounters.

Another broad concern of the ethno- methodologist has been with the routine productions of persons in social organiza- tions. The basic hypothesis guiding these studies, which have ranged from analyses of mental health clinics, to hospitals, police departments, juvenile courts, and suicide prevention centers, is that members of any social organization develop a special per- spective for handling their clients. It is argued that the perspectives of any given organization will be sufficiently different from any similar organization to make com- parisons between such agencies problematic. Ethnomethodological studies have sug- gested that (1) organizations perpetuate

928 AMERICAN SOCIOLOGICAL REVIEW

themselves through time by generating fictitious records; (2) comparable organiza- tions differ in the meanings they assign to the same events (e.g. birth, death, mental illness, cured, etc.); (3) the production of organizational records is basically an inter- actional process based on rumor, gossip, over- heard conversations, discrepant information and biographically imperfect bookkeeping. Cicourel (1968), for example, noted that agencies created to process juvenile delin- quency routinely produced delinquents by piecing together long series of conversations between the predelinquent, his parents, the arresting officer, the counselor, and the judge. The sum total of these conversations, trans- lated into official reports, represented the organizational documentation that a delin- quent act had or had not occurred. And (4) in piecing together these organizational re- ports it was found that members routinely relied on open-ended categories to classify cases. What Garfinkel (1967: 73-75) calls the "et cetera clause" refers to this tendency of persons to fit events into a pattern that complements their on-going action.

It is important to note that these studies amplify the research of interactionists on the labeling process. Becker (1963) has sug- gested that deviance does not reside in social acts, but must be traced to definitions that arise during interaction. Cicourel's research suggests that deviance may be as much or- ganizational as interactional in nature and must be related to the working perspectives of members of social control agencies. Gar- finkel's studies propose that disruptions of everyday perspectives can create feelings of distrust which become translated into deviant labels.

In several senses the ethnomethodologist has merged the interactionist's concern with deviance with the study of social organiza- tion and social relationships. Their studies cross-cut the analysis of civil-legal, polite interactional and relational rules.

These studies suggest an additional point of convergence. Traditionally, interaction- ists have employed a conception of the or- ganization that stresses the interpretations given to positions in a division of labor work, noting that the key to an organization lies in its informal structure. This has led to the

proposition (Hughes, 1956) that beside every task division of labor rests a moral hierarchy of positions which dictate how per- sons are to relate. A typical interactionist study of social organization begins with the formal structure and then details how moral and ideological variations transform that structure into a going concern of social re- lationships (e.g., Strauss, et al., 1964). This complements the ethnomethodologist's strat- egy but suggests that the formal structure cannot be ignored.

Perhaps the most important claim of the ethnomethodologist is the statement that the productions of the sociologist are similar to those in everyday life. This echoes the con- cern of Mead and others for distinguishing scientific from everyday activities.

The ethnomethodologist's argument in- volves the following points. First, all sociolo- gists are (or should be) concerned with de- picting the taken-for-granted affairs of actors in any social order. The sociologist will find that he is forced to make decisions regarding the relationship between his concepts and his observations. In making these decisions, he will note that unclassifiable instances ap- pear, that coding schemes become too nar- row, that statistical tests are inappropriate, or that observations bear little, if any, re- lationship to central concepts and hypoth- eses. In the process of deciding when an observation fits or does not fit a conceptual category Garfinkel (1967:78-79) suggests the sociologist make use (even if uncon- sciously) of the documentary method of analysis. In applying this method it will be found that any instance of classifying an observation rests on the earlier discussed as- sumptions of daily interaction. That is, events will be placed in a temporal sequence, certain statements will be ignored, and com- mon vocabularies will be assumed. If the method of data collection rests on inter- views, it is argued that the researcher must give attention to the interaction that occurs between himself and the respondent. In this context Garfinkel (1967) and Cicourel (1967) suggest that while it is commonly assumed that interviewers and respondents achieve a "rapport" during the interview, this hypothesis is problematic. Data gath- ered via interviews and questionnaires are

SYMBOLIC INTERACTIONISM 929

viewed by the ethnomethodologist as col- laborative products created by the sociolo- gist and his subject. To understand such productions demands a knowledge of the routine meanings held by subjects.

At the heart of the ethnomethodologist's position lie a series of assumptions taken from Schutz (1963). For Schutz there ex- isted an impenetrable barrier between the scientific and the everyday conceptions of reality. The fundamental conflict between scientific and everyday conceptions arose, Schutz argued, (1963:325-326) from the fact that the scientist could never fully, com- pletely, and accurately enter into the "here and now" world of everyday man. The an- swer was to construct rational models of action that actors in everyday life would never fully live up to. Schutz's (1963:340- 341) everyday actor becomes a rational con- struct manipulated by the sociologist.

The adoption of these assumptions can be seen in recent statements by Garfinkel and Cicourel. At several points Garfinkel (1967: 262-283) elaborates the differences between scientific and everyday rationalities. On other occasions Garfinkel and Cicourel de- liberately construct models of everyday man which rest on a social game perspective. This model is one attempt to construct a series of rationalities that would permit the analysis of everyday action.

CRITICISM OF BOTH PERSPECTIVES

Before offering a synthesis of these per- spectives, a review of their problematic fea- tures is necessary. Interactionism has been criticized on methodological and theoretical grounds.

Theoretically, critics have challenged the metaphorical basis of the dramaturgical view which they suggest is alien to the interaction process because it gives man an unattractive motivational commitment-that is, to ever win support for a presented self. On other grounds it has been suggested that interac- tionists are so vague on the self as a concept that firm empirical observations cannot be gathered. Indeed, interactionists are vague themselves on the causal status of this con- cept, and Kuhn (1964:61-84) has identified some nine different variants within the per- spective on this dimension.

This criticism reflects a general dissatisfac- tion with the perspective because it offers too few concrete hypotheses. Compounded with this issue is the nature of interactionist re- search. This problem was raised earlier and possible strategies for improving such re- search were suggested.

Some have suggested that interactionism offers an excellent view of social relationships and social groups but fails to adequately treat larger forms of social organization. In my judgment this criticism is largely mis- placed because it ignores the long line of research on the sociology of work and organ- izational settings stimulated by Hughes. Sim- ilarly, the study of collective behavior within interactionism offers a perspective for the analysis of mass society (e.g., Blumer, 1957; Klapp, 1964; Couch, 1968).

Ethnomethodology has been criticized for its phenomenological bias which on occasion has restricted its practitioners from treating the relationship between individual defini- tions and larger social units (Coleman, 1968: 128). Others have suggested that the per- spective does not suitably treat the role of the self in interaction and some have noted that there is no clear demonstration of how taken-for-granted assumptions operate in daily interactions (e.g., Swanson, 1968: 122-124). The failure of the ethnomethodol- ogist to precisely indicate the nature of the documentary method has also raised criti- cisms (Coleman, 1968; Hill and Crittendon, 1968).

Collectively I would criticize both perspec- tives for their failure to clearly indicate the source of meanings and definitions. In addi- tion, they as yet offer no firm strategies for measuring the interaction process. In defense of each, I would suggest that they offer a view of human conduct that recognizes the complex role of interaction in shaping activ- ity. They point to important distinctions between the interpretative and noninterpreta- tive elements of conduct, and Garfinkel's studies offer experimental strategies for fu- ture research. The persistent treatment of scientific and everyday activity within both frameworks offers a fresh perspective on the sociology of knowledge, and this is especially the case with Garfinkel's analysis of scien- tific rationalities.

930 AMERICAN SOCIOLOGICAL REVIEW

A VIEW OF SOCIAL ORGANIZATION AND THE INTERACTION PROCESS:

A PROPOSED SYNTHESIS

If face-to-face interaction is characterized by shifting modalities of interpretation, then a major point of convergence between ethno- methodology and interactionism is the treat- ment of the meanings given to social objects.

The Interpretational Process

This suggests several hypotheses concern- ing the movement of objects from interpre- tative to non-interpretative roles. The first suggests that any event challenging normal interpretation creates pressures to bring that event into the flow of interaction. When an object is taken out of its non-interpretative status and held up for consideration, frus- tration and groping will be observed at a rate proportionate to the importance assigned the object. Thus some objects can be quickly settled upon if they occupy a relatively low position in the interaction. This would be the case with breakdowns in mechanical equip- ment at sociable gatherings; they can either be replaced with other objects, or ignored.

A second hypothesis suggests that the fun- damental objects for any interaction in- volve those that must be negotiated over. In short, taken-for-granted objects will not account for the complete variance in be- havior. Those objects which are accorded explicit interpretative status will significantly determine the flow of events. On the other hand, taken-for-granted objects cannot be ignored, and I would hypothesize that ob- jects in this class receive earliest attention in any encounter. Once their meaning can be taken-for-granted, they cease to operate as problematic elements.

The nature of the interaction process is such that a complete a priori classification of objects cannot be given. Earlier I sug- gested that the self represents the most sig- nificant object for interpretation. To this should be added the meanings brought into the situation and perhaps the situation as well; although the situation, in a concrete sense, is likely to contain the greatest pro- portion of non-problematic objects.

The problem of meaning still remains vague in both perspectives. As a point of empirical inquiry, meaning can be treated as an element of the covert symbolic act, and by

self-reports measured in terms of the ex- pectations for action that are brought into the interactions (e.g., McHugh, 1968). Fol- lowing Garfinkel, such expectations would include assumptions concerning who was go- ing to be present, the length of time to be spent, the types of selves one was going to present, the degree of knowledge held about the occasion, and the types of objects that were going to be encountered. Once the inter- action begins, overt activities could be linked to the shifts in meaning that the participants were constructing as they interacted.

In this way interaction could be measured by the frequency of joined actions. The emergent effect of interaction would be rep- resented by the frequency of disrupted plans of action; that is, how frequently partici- pants had to alter plans of action brought into the encounter. Interactions could then be examined in terms of their emergent qual- ities. Those that flowed basically along non- interpretative lines would be judged less emergent, and so on.

Deviance, the Labeling Process and Agents of Social Control

Because interactionism and ethnomethod- ology have focused on the deviance and label- ing process, an additional series of hypotheses can be offered. If Garfinkel's conception of trust is redefined to specify those situations where two or more actors assume that the other will abide by decisions mutually agreed upon, then violations of trust become viola- tions of these agreements. Examples would include entrusting another with a dark secret about the self, withholding salient informa- tion from outsiders, or simply continuing to interact along consensual lines. Breakdowns in joint action could be partially traced to breakdowns in the trust-taking attitude and would be vividly displayed in the betrayal process that characterizes interactions be- tween normals and persons defined as men- tally ill (e.g., Sampson, Messinger and Towne (1962:88-96).

A first hypothesis emerges: continued vio- lations of trust create strains in the relation- ship which culminate in attributions of devi- ance directed toward the trust-violator. This suggests that only certain breakdowns in con- sensual interaction will produce a deviant label. Violations of the relational order would

SYMBOLIC INTERACTIONISM 931

lead to a greater proportion of deviant labels than would violations of polite-interactional and civil-legal codes. Thus a failure to abide by rules of deference and demeanor, refusals to act in terms of an agreed upon division of labor, or leaking of crucial information to outsiders would represent significant con- cerns around which trust would have to be sustained.

If the labeling process is raised to the or- ganizational level, the readiness of members within social control agencies to validate a deviant label would vary by their perception of the degree of trust-violation attributed to the potential deviant. That is, the extent to which the member of a social relationship can validate a claim of trust-violation in- creases the readiness of members of social control agencies to accept and process the potential deviant. This hypothesis must be conditioned by the fact that social control agencies continually monitor their deviant populations in terms of the ability of those populations to meet perceived organizational needs. As Cicourel (1968) indicates, this monitoring process serves collective organ- izational needs as well as the concerns of individual members. When these concerns converge with the attributions of deviance among members of social relationships, greater rates of labeling would be expected, suggesting that members of social control agencies may turn away valid instances of deviance simply because they do not meet the needs of the organization. On the other hand, deviants may be created by such agen- cies when valid attributions of deviance are not given within the social relationship. This would be especially so when members of the social relationship are in conflict with the social control agency. Current examples in- clude police monitoring of ghetto areas, and police enforcement of drug abuse laws among college youth. In these instances the labeling process becomes a political and ideological issue with members of social control agencies responding more to political demands than to socially validated instances of deviance.

Social Relationships, Socialization and the Languages of Interaction

Returning the discussion to social relation- ships and trust, breakdowns in consensual action often arise because persons hold con-

flicting definitions of the salient objects in their environment. If a wife persists in di- recting interaction around objects which a husband regards as taken-for-granted, con- sensus may soon collapse and joint action becomes problematic. A sense of dissatisfac- tion with the interactional partner may be created which soon eventuates in reciprocal alienation; this in turn produces a situation of deviance attribution. The outcome may be divorce.

Another area of mutual concern is the lan- guage of interaction. It appears that at least two languages, one silent, and one vocal, characterize the interaction process. On the silent level, rules regarding body spacing, gesturing, the control of body noises, and aromas, and the ordering of words can be observed (Sommer, 1968). These often re- main unstated, and represent the "back- ground expectancy set" of the interactants.

On the vocal level, prescriptions concern- ing proper address, naming, tone of voice, and choice of vocabulary are observable. These represent the overt aspects of interac- tion and their expression displays the salient features of the silent dialogue (Goffman, 1963).

If interaction involves both languages, then ethnomethodology and interactionism provide a perspective for analyzing the con- tingencies of face-to-face encounters. An im- portant line of investigation becomes the problem of socialization. How are persons taught rules that are seldom vocalized? How are sanctions brought to bear upon perceived violations of the silent language? At what point in the socialization process are children assumed to be responsible for their silent behaviors?

These questions are cross-cut by a particu- lar image of socialization. Contrary to some theories which regard socialization as a dis- continuous learning process into well-defined roles, ethnomethodology and symbolic inter- actionism suggest that socialization is never- ending, and often involves more of what is not said than what is stated (e.g., Olesen and Whittaker, 1968; and Clausen 1968: 130-181). Socialization thus represents an ubiquitous feature of all interactions-the apprehension of another's perspective so that joint action can occur. Indeed, if in- teraction is regarded as a potentially emer- gent event, socialization is one aspect of

932 AMERICAN SOCIOLOGICAL REVIEW

the role-taking process. If success at joint action is measured by the ability of persons to fit lines of action together, then this suc- cess represents the quality of the ongoing socialization process.

The Study of Social Organization

Because both perspectives provide a sim- ilar view of the interaction and socialization process, they offer a powerful strategy for organizational analysis. The following points can be indicated. First, organizations become territories of interaction that are focused around complex spoken and unspoken lan- guages. These languages represent the salient organizational concerns, often work-related, and they offer prescriptions for action that frequently run counter to formally stated organizational goals. Each organizational role position can be seen as having its own special language, both silent and vocal. This language will communicate special socializa- tion strategies which are daily tested and reaffirmed through interaction. These lan- guages will tend to cluster within, and in- deed give focus to, special social orders that exist hand-in-hand with other social orders. The organization is then conceived in terms of competing perspectives and social orders, each of which rests on its own language. The sum of these languages represents the or- ganization's collective perspective. The sum of the social or moral orders becomes the behavioral representation of the organization.

In this way organizations are broken down into interactional units, each of which offer special ways of thinking and acting. It be- comes difficult to speak of one organization, or one organizational perspective. An organ- ization represents a multitude of shifting and competing languages and social orders: a social order that is held together, if at all, by a few very salient symbols such as univer- sity X, or mental hospital Y. The name of the organization perhaps represents the only salient symbol all participants would agree upon.

Methodology and the Study of Scientific Conduct

Both perspectives offer a series of hypothe- ses relevant to the analysis of scientific con-

duct and the development of sociological methodology. The scientist is judged by his ability to challenge accepted perspectives- that is by his ability to be self-consciously interpretative. Developments in science re- flect growth stimulated by this challenging stance. In this way Mead partially antici- pated Thomas Kuhn's notion of the para- digm model of scientific development. Of equal importance was Mead's ability to sepa- rate the rationalities of scientific conduct from the perspectives of everyday life, a point more fully elaborated by Garfinkel in his analysis of the forms of rationality (also see Blumer, 193 1). A hypothesis which runs slightly counter to the Mertonian and Par- sonian image of scientific behavior emerges (Kaplan, 1964:855-857). Scientific conduct is so imbued with elements from everyday life that unless the scientist self-consciously directs his activity in terms of the norm of discovery, his behavior is unfavorably judged. The norms of the scientific institu- tion parallel those of other enterprises, most notably art, theology, and philosophy (Swan- son, 1968:123). Consequently, Merton's (1957) four norms of universalism, commu- nism, disinterestedness and organized skepti- cism are not unique to science.

In this way Mead and Garfinkel have opened the way for a more open discourse of the value-free problem. It is now impossible to view science as other than a value-laden enterprise-a position similarly reached by Gouldner (1962) and Becker (1967). It can be no other way if the assumption that sci- ence as a human enterprise is granted. Con- sequently, the scientific norms of rationality which include (1) official neutrality toward the meaning of objects; (2) an irrelevancy for the real world; (3) an indifference to chronological time; (4) perfect communica- tion; (5) standards of publicity, remain norms that are imperfectly realized (Swan- son, 1968: 123). This is the thrust of Gar- finkel's critique of modern sociology. The sociologist's belief in a perfect system of ra- tionalities has led him further away from the world of social events. The sociologist has pursued his normative system at the expense of concrete behavioral analyses of face-to- face interaction. By forgetting that he is re- sponsive to social demands, the sociologist has overlooked the irrational elements of his

SYMBOLIC INTERACTIONISM 933

own conduct-most notably his inability to make sound observations, to reliably code documents, to conduct face-to-face interviews and so on.

The thrust of ethnomethodology and inter- actionism becomes clear-the sociologist can never ignore the interactional features of his own conduct. Interactionism and ethnometh- odology offer some recipes for corrective action. They may be stated as hypotheses.

If the scientific observer is subject to in- teractional demands, and hence less than perfect as a recorder of social events, then multiple observers and multiple methods, which overcome one another's restrictive bi- ases, become the most valid and reliable strategies of observation. This suggests that any observation based on the triangulation principle will yield data that are more reli- able and valid than an investigation that is not so based.

Because the scientist brings unique inter- pretations to bear upon his own conduct, a major source of variance in any investigation becomes the nature of these interpretations. Specifically, this predicts that two experi- menters, for example, will produce different findings to the extent that they conceive their role differently (see Friedman, 1967). This would also hold for interviewers, coders, and unobtrusive observers.

Third, the interaction process between an observer and a subject must be examined for its effect on the data. Whenever interaction is emergent, observational encounters become noncomparable events for purposes of collec- tive analysis. If an encounter proceeds along taken-for granted lines, which is a measure of the degree of nonemergence, similar ob- servational encounters could be pooled for collective analysis. This proposition directs investigators to record carefully the nature of the interaction process with a special eye to events that they judge to be unique within each encounter.

Additional hypotheses could be offered, but space restricts their elaboration. The basic point is that both interactionism and ethnomethodology direct scientific interest in the scientific process itself.

In concluding this proposed synthesis, I would offer as a final point the convergence of ethnomethodology's interest in the study of face-to-face interaction with the methodo-

logical principles of symbolic interactionism discussed earlier. The documentary method, as a strategy of pointing to empirical in- stances of theoretical concepts, can be easily merged with the use of analytic induction, sensitizing concepts, the method of role tak- ing, and the strategies of linking individual perspectives with larger social units.

Similarly, Garfinkel's use of the quasi- experiment in natural field settings can be- come a model for more rigorous studies of face-to-face encounters. This is especially so if recent findings on experimenter effect and subject perceptions are incorporated into the experimental design (e.g., Friedman, 1967).

REFERENCES

Becker, Howard S. 1963 Outsiders: Studies in the Sociology of De-

viance. New York: The Free Press. 1967 "Whose side are we on." Social Problems

14 (Winter):239-248. Blumer, Herbert.

1931 "Science without concepts." American Jour- nal of Sociology 36 (January):515-531.

1957 "Collective Behavior." Pp. 127-158 in Joseph B. Gittler (ed.), Review of Sociol- ogy. New York: John Wiley and Sons, Inc.

1966 "Sociological implications of the thought of George Herbert Mead." American Journal of Sociology 71 (March) :535-544.

Cicourel, Aaron V. 1967 "Fertility, family planning and the social

organization of family life: Some methodo- logical issues." The Journal of Social Issues 23 (October):57-81.

1968 The Social Organization of Juvenile Jus- tice. New York: John Wiley and Sons, Inc.

Clausen, John A., (ed.). 1968 Socialization and Society. Boston: Little,

Brown. Coleman, James S.

1968 "Review symposium of Harold Garfinkel, studies in ethnomethodology." American Sociological Review 33 (February) :126- 130.

Coser, Lewis A. 1956 The Social Functions of Conflict. Glencoe,

Illinois: The Free Press. Couch, Carl J.

1968 "Collective behavior: An examination of some stereotypes." Social Problems 15 (Winter) :310-322.

Foote, Nelson N. 1951 "Identification as a basis for a theory of

motivation." American Sociological Re- view 16 (February):14-21.

Friedman, Neil, J. 1967 The Social Nature of Psychological Re-

search. New York: Basic Books.

934 AMERICAN SOCIOLOGICAL REVIEW

Garfinkel, Harold. 1967 Studies in ethnomethodology. Englewood

Cliffs, N. J.: Prentice-Hall. Glaser, Barney G. and Anselm L. Strauss.

1967 The Discovery of Grounded Theory. Chi- cago: Aldine.

Goffman, Erving. 1959 The Presentation of Self in Everyday Life.

Garden City, N. Y.: Doubleday. 1963 Behavior in Public Places. New York: The

Free Press. Gouldner, Alvin W.

1962 "Anti-minotaur: The myth of a value- free sociology." Social Problems 9 (Winter, 1962):199-213.

Gross, Edward and Gregory P. Stone. 1963 "Embarrassment and the analysis of role

requirements." American Journal of Sociol- ogy 70 (July):1-15.

Hill, Richard J. and Kathleen Stones Crittenden (eds.).

1968 Proceedings of the Purdue Symposium on Ethnomethodology, Lafayette, Ind.: Pur- due Research Foundation.

Hughes, Everett, C. 1956 Men and Their Work. Glencoe, Ill.: The

Free Press. Hughes, Everett C. and Helen MacGill Hughes.

1952 Where Peoples Meet. Glencoe, Ill.: The Free Press.

Kaplan, Norman. 1964 "Sociology of science." Pp. 852-881 in

Robert E. L. Faris (ed.), Handbook of Modern Sociology. Chicago: Rand Mc- Nally.

Klapp, Orrin E. 1964 Symbolic Leaders. Chicago: Aldine.

Kuhn, Manford H. 1964 "Major trends in symbolic interaction the-

ory in the past twenty-five years." Socio- logical Quarterly 5 (Winter):61-84.

McHugh, Peter. 1968 Defining the Situation. Indianapolis, Ind.:

Bobbs-Merrill. Mead, George Herbert.

1917 "Scientific method and individual thinker." Pp. 176-227 in John Dewey (ed.), Crea- tive Intelligence: Essays in the Pragmatic Attitude. New York: Henry Holt and Co.

1927 "The objective reality of perspectives." Pp. 75-85 in Edgar Sheffield Brightman

(ed.), Proceedings of the Sixth Interna- tional Conference of Philosophy. New York: Longman, Green and Co.

1934 Mind, Self and Society, Chicago: Univer- sity of Chicago Press.

Merton, Robert K. 1957 Social Theory and Social Structure. Glen-

coe, Ill.: The Free Press. Olesen, Virginia and Elvi Whittaker.

1968 The Silent Dialogue. San Francisco: Jos- sey-Bass.

Sampson, Harold, Sheldon L. Messinger and Robert D. Towne.

1962 "Family processes and becoming a mental patient." American Journal of Sociology 62 (July):88-96.

Schutz, Alfred. 1963 "Common-sense and scientific interpreta-

tion of human action." Pp. 302-346 in Maurice Natanson (ed.), Philosophy of the Social Sciences. New York: Random House.

Simmel, Georg. 1950 The Sociology of Georg Simmel. Kurt H.

Wolff (ed. and Tr.), Glencoe, Illinois: The Free Press.

Sommer, Robert 1968 Personal Space. Englewood Cliffs, N. J.:

Prentice-Hall Stone, Gregory P.

1962 "Appearance and the self." Pp. 86-118 in Arnold M. Rose (ed.), Human Behavior and Social Processes. Boston: Houghton Mifflin.

Strauss, Anselm L., Leonard Schatzman, D. Ehrlich, R. Bucher, and M. Sabshin.

1964 Psychiatric Ideologies and Institutions. New York: The Free Press.

Swanson, Guy E. 1968 "Review symposium of Harold Garfinkel,

studies in ethnomethodology." American Sociological Review 33 (February) :122- 124.

Thrasher, Frederick M. 1927 The Gang. Chicago: University of Chicago

Press. Webb, Eugene J., Donald T. Campbell, R. D.

Schwartz, and L. Sechrest. 1966 Unobtrusive Measures: Nonreactive Re-

search in the Social Sciences. Chicago: Rand McNally.