Sociology 333 – Introduction to Symbolic Interactionism (S.I.)

Arthur W. Frank

Having two theories before us—functionalism and interactionism—allows us to begin to ask: What is a theory? The answer will continue to shift throughout the course, but let me note a few components. I’m not sure it’s that useful to seek to "define" theory, but we can begin to delimit it.

Most theories have some version of:

Keep in mind also Alexander’s statement that for any theory, "the problems of action and order are not optional" (193b). Alexander reduces action problems to the dichotomy of rational and non-rational, and order problems to collectivist and individualist. He argues, and I agree, that

Thus the key difference to look for in these readings is how the authors understand society as "built up" by individuals through actions they choose, those choices guided by emotional, value-laden, and generally ‘inner’ considerations. Remember what Alexander also writes:

"Sociologists are sociologists because they believe there are patterns to society, structures somehow separate from the actors who compose it. Yet, while all sociologists believe such patterns exists [sic], they often disagree sharply about how such an order is actually produced." (193a)

Thus you should be asking yourselves, throughout the course:

Functionalists are most interested in patterns of mobilizing resources to achieve external goals, and complementary to that, sustaining internal integration. They believe these patterns are produced by actors who are socialized into a condition of "motivated compliance" with the functional imperatives of the system. As to Parsons’s own "research", many would argue that his failure to do much research was his greatest shortcoming. A general criticism of functionalist research is that it assumes on the basis of its theory what it then has to find when it goes out and looks—that’s unfair in some instances, and maybe not so unfair in others. Note that everyone has a criticism of everyone else, and functionalists criticize symbolic interactionist research for failing to develop a general theory that explains much of anything (symbolic interactionists then reply that what functionalists "explain" is a fictional "social system" they have dreamed up, and so the argument goes).

Symbolic interactionists are interested in patterns of interaction. They agree that actors themselves produce those patterns, but even within S.I. there is disagreement as to the degree of local power actors have to change patterns, when and how much. Thus within S.I. you can find more "collectivist" orientations and more "individualist" ones. My notes below on the readings are designed to help you see these differences.

Herbert Blumer coined the term symbolic interaction, though he only asserted it as the name for a theoretical orientation in the early 1960s. Blumer was a charismatic sociologist: a former pro football player (the precursor team to the Chicago Bears) who never went to seed. I’ve seldom encountered a man so hard to disagree with. His theory is quite simple—some would say ‘limited’ and others would say ‘parsimonious’—and is stated with wonderful clarity in his paper "Society as Symbolic Interaction." Remember: Blumer is not interested in elaborating a "theory" but in kick-starting a research program of direct observation, including ethnography and open-ended interviewing. What counts is to go into social settings and discover what the people there believe is the meaning of their actions.

Blumer’s philosophic anthropology is the most individualistic: "human beings interpret or ‘define’ each other’s actions instead of merely reacting….Their response…is based on the meaning which they attach to such actions" (232b). Blumer would admit that people do not, strictly speaking, make these meanings up on each occasion of interaction; meanings are based in shared symbol systems. But actors attach and recombine these meanings as they will.

Blumer then elaborates his three famous propositions (though Homans would claim they are not propositions, since they are not testable):

  1. "the human being has a self", by which Blumer means that humans "make indications" not only of the things (and people) around them, but of themselves; "the human being can be the object of his own actions" (233a). In Mead’s famous phrase that Blumer is elaborating, we can take our-selves as objects for ourselves. S.I. emphasizes self-consciousness: "conscious life…is a continual flow of self-indications." When Blumer writes that humans "make indications", he means that we single out objects (including other people and ourselves) and attach meaning to them; we constantly judge the object’s applicability to the situation we are in, making active judgments about how to use (or refuse) the object.
  2. Blumer’s second point shows that S.I. has its own continuum of rational and non-rational action. At the rational end, Blumer writes that the actor "has to note what he wants to do and how he is to do it" (233b). This sounds much like Parsons early action theory, but then Blumer presents a very different explanation of how the actor proceeds. "His action is built up step-by-step through a process of such self-indication." Blumer goes on to elaborate that S.I. is not a psychological reductionism. He makes one of his strongest individualist statements: "By virtue of indicating such things to himself, he places himself over against them and is able to act back against them, accepting them, rejecting them, or transforming them in accordance with how he defines or interprets them" (234a). These interpretations are not externally derived but. as Blumer writes (somewhat obscurely, on my reading): "The process of self-indication exists in its own right and must be accepted and studied as such." I’m honestly not sure quite what that means, but I believe it’s the core of Blumer’s thought. [Note: it’s almost a counter-claim to Durkheim’s principle that social facts exist "sui generis", in their own right.]
  3. We then get to the interactive part, which is what tips S.I. into the non-rational action category. Actors do not "build up" meanings on their own, but do so by "taking the role" of others; that is, by seeing situations from others’ perspectives. Meaning is a question of mutual alignment, not an individual construction. In this alignment actors seek not self-interest (Alexander’s defining quality of "rational" action) but community affiliation.

That’s the core of S.I., and the rest of Blumer’s essay is thinly veiled refutation of Parsons and elaboration of what S.I. has to say about organizations and their reality. If Blumer thinks that "social structures" exist only in the minds of sociologists who spend too much time in their studies and too little doing direct observation, he does acknowledge the reality of organizations. Note how Blumer seeks to combine his own version of collectivism ("the organization of human society is the framework inside of which social action takes place") but, predominantly, individualism (this framework "is not the determinant of that action" but is itself "the product of the activity of acting units", that is, people (237b). As he says below, structural features set conditions within which people act, "but they do not determine their action."

The S.I. research program, then, is to understand what sorts of frameworks various parts of social life are, but more important, how actors jointly define meanings and construct lines of action within those settings. As I wrote above, the preferred method is ethnographic, and I want to emphasize that S.I. took a large hit when rules of "informed consent" of research subjects (promulgated in the early 70s and becoming increasingly restrictive even as I write) made it impossible to do true "participant observation" research, in which those being observed were unaware that one of the group members was actually a sociologist. The current S.I. interest in "auto-ethnography"—using one’s own experience as a source of research data—is one of a continuing series of attempts to recover what was lost.

I’ll say a good deal less about the other three readings. Each takes up the problem Alexander suggests, above, of how members acquire the patterns of their action. Each is more "collectivist" than Blumer, though whatever "framework" is described, it never determines the action that takes place within it.

Anselm Strauss is best known today as one of the developers of the qualitative method of "grounded theory." The reading is from the early Strauss, when he was trying to differentiate his theoretical position from Mead’s work, of which Strauss was one (of many) editors. In his discussion of the relative importance of personal history in people’s lives (or, their relationships with their ancestors), he is elaborating what is "symbolic" about meaning, and well as opening up a personal, historical dimension to interaction: we also interact with those whom we experience coming before us, creating possibilities and expectations that frame our lives.

Strauss begins discussing question of group membership—who’s in the group and who’s not—because these memberships furnish us with many of our meanings (in Blumer’s terms, they direct our ‘indications’). He then turns to history, writing: "A man must be viewed as embedded in a temporal matrix not simply of his own making [note the collectivist direction], but which is peculiarly and subtly related to something of his own making—his conception of the past as it impinges on himself" (229a). In this typical S.I. turn of phrase, the collectivist side is acknowledge but then subsumed within an individualist framework. We do not "make" our respective pasts, but we do form our own conceptions of how that past will "impinge" on us.

Strauss then takes his argument to the group or nation-state level, discussing how nationalist movements often have to create a past that is consistent with their definition of the present. A number of later studies describe this process as it takes place in Northern Ireland, Yugoslavia, and parts of Africa. Strauss ends with another typical S.I. formulation, "Each generation perceives the past in new terms, and rewrites its own history" (230b). Of course that rewriting takes place within a certain framework (both of content and of what counts as "history"), but the emphasis is on people creating meaning.

Turning to the reading from Goffman, I’m not sure why the editor chose a particularly dense section, except that it shows Goffman at his most collectivist: the emphasis is on the game as a structure (a container, as Goffman sometimes says) of interaction, and the reader has to keep in mind that "players" are still choosing both how to play and whether to play. Goffman would be unhappy to find himself in a "S.I." section of a text; he was often at pains to declare his distance from symbolic interaction. In one famous passage he wrote that while W.I. Thomas’s famous theorem (‘If men define situations as real, they are real in their consequences.’) is true as it is written, it obscures the more important reality that persons do not "define" their situations to any significant degree; most situations are already understood, culturally, as occasions for this and not that. Thomas is right, we do define our situations; but Goffman emphasizes, we do not make up these definitions for ourselves. Thus Goffman is at the collectivist end of the S.I. continuum, if he is a "symbolic" interactionist at all (he is clearly an interactionist, but interested in frameworks, not symbols).

"Fun in Games", from which this reading is taken, is one of two long essays that make up Goffman’s book Encounters. The other essay is more famous for its depiction of what Goffman called role distance. Here his work seems more complementary to Blumer: the "role" is the socially constructed framework, but people work to demonstrate to themselves and others that they are ‘more than’ the role in which society places them. That essay is about people playing with—and against—the expectations that adhere to their roles. The section you’re reading is about the structure of interaction, and represents an interest in games that Goffman would develop in his later book, Strategic Interaction, which is perhaps his most "rational" and "collectivist" theoretical version, and not regarded as one of his best books. So to be blunt, this is not the Goffman reading I would have chosen for inclusion in the book. But Goffman is always worth reading.

The paper begins with the idea of "engrossment" (239b) that Goffman has developed in his earlier book, Relations in Public. There he shows that a rule of interaction requires each person to present him or herself as "engrossed" in the present interaction; to seem to have one’s mind elsewhere is damaging to the interaction and embarrassing to the others involved (lapses question the validity of others as interactants worth being with). Note that interaction for Goffman is always a collaborative performance, in which each has a responsibility to the others to sustain the situation and, in particular, to avoid anyone’s embarrassment. Thus Goffman has some harsh things to say about cheating in games.

By the end of p. 240 the topic seems to have drifted to something else, not clearly announced on my reading [note on writing style: always tell the reader what your point is, especially if you’ve changed directions]. His point was not clear to me until 243b, when he writes of "the kind of veil that will be drawn over his feelings while in communication with another." We have a basic problem in interaction: how much of ourselves and our feelings do we let show? We all need some kind of veil, but if it is too thick, others will find us less than engrossing and interaction will falter; if it is too thin, inappropriate feelings may "flood out" as Goffman frequently worries (he is consistently worried about scenes of "flooding out"; recall here Alexander’s argument about presuppositions).

The argument resolves itself in an assertion of the collectivism in interactionism: "but in this case the disguise is socially standardized; it is applied by the individual but not tailored by himself to his own particular needs" (243b). Again, we define situations but we do not make up the definitions ourselves.

Goffman’s subtle back-and-forth shifts over the boundary of individualism/collectivism, and his commitment to working out the problem of order at the level of interaction, is most clearly seen at the top of 244 when he writes of "feeling that is generated in the institutional situation at large but disguise what is being expressed sufficiently to ensure the orderliness of the particular encounter." In other words, we have and show feelings (non-rational), but we do so within occasions structured by institutions. And any show of feeling has to be moderated so as to keep the encounter orderly (again, his fear of "flooding out" or excessive display that would say too much about actual attitudes toward the situation and others in it).

Stanford Lyman’s paper is more an attempt at theoretical synthesis than Blumer might have thought advisable. Lyman begins with the need for habits. These are necessary but, necessarily, dull in their repetitive nature (that’s what a habit is, repetitive). The question you should be asking is, what’s Lyman trying to explain? It’s not immediately clear. In part he’s playing a theory game, trying to concoct his own stew by mixing a bit of this and that. On at the top of 249 does he offer some kind of definition of habits, as "praxes", or ways of getting things done. Habits are the routines through which social life is done, or, the practices through which life is practiced. He proceeds to quote some interesting statements, but as I read I’m still not getting what he thinks his contribution is. I think I finally got it on 250a when he describes "crises" that interrupt the habitual flow. Something happens and suddenly our habit is no longer adequate as a practice: it doesn’t get the job done, and we have to think—innovate—how to do what we’re doing (and maybe even try to remember what it was this habit was supposed to do). On 250b we get back to Goffman’s interest in embarrassments as crises of the habitual in interaction: embarrassment requires repair. {I imagine Goffman replying to Lyman that even crises have their habitual responses.}

Again the theory issue is the collective and the individual. Habits are collective, but they get disrupted, and the responses to this disruption are individual in the sense of Mead’s "I" that injects some unpredictability into action (the "me" acting only according to the expectations of others). As I understand Lyman, the crisis calls out some "I" response. I like, or think I like, Lyman’s metaphor of society as a multiplex theatre without walls "to separate its autonomous scenarios". The issue here the problem of how—out of deference to our fellow interactants—we show ourselves to be engrossed in one scenario even as another may be bursting through, demanding our attention.

I’ve spent more time talking about the collectivist/individualist dimension (order) than the rational/non-rational dimension (action). Note that all four authors place the emphasis in action on the actor’s responsibility to sustain the interaction. Even in games, each plays *with* the other players, not against them. When someone wins, that is an outcome required by the structure of the game, a culmination to the mutual engrossment of players and spectators. "Winning" is not seeking a personal reward so much as it is sustaining an interactional framework that requires someone win (see Goffman’s example in note 7, p. 245). Next week, in the world of George Homans, that emphasis will shift.