Symbolic Interactionism - Meaning Made in the Moment

Functionalism and conflict theory both start from above — systems, structures, institutions, groups — and work down to what people do.

Symbolic interactionism starts from the other end, and it makes a genuinely radical claim: that there is no society above the people in it, that structures exist only insofar as they are produced and reproduced in encounters, and that a sociology which cannot describe a conversation cannot describe anything.

Lessons 2.3.2, 2.3.3 and 2.4.2 introduced Cooley, Mead and Goffman as vocabulary. This lesson takes the school as a school: its premises, its methods, its achievements, and the objection it has never fully answered.

A queue that does not exist.

Twelve people are waiting for a bus. They are standing in a rough line on a pavement.

Now ask what makes this a queue.

There is no sign, no barrier, no marking on the ground and no official. Nobody has agreed anything. Most of them have not spoken.

And yet: everyone knows their position. Everyone knows who is in front of them and who is behind. If a thirteenth person arrives and stands at the front, something will happen — a glance, a small movement, a cough, possibly a word — and in the overwhelming majority of cases they will move.

The queue exists entirely in the shared understanding of the people in it, and it is maintained continuously by their behaviour. Remove them and there is nothing left. No structure persists. There is a pavement.

And yet the queue is not nothing. It constrains: you cannot simply walk to the front. It has a moral force: violating it produces indignation out of all proportion to the seconds involved. And it produces itself again the next morning, with different people, without anyone deciding.

Herbert Blumer's claim is that society is like the queue, all the way up. Institutions, organisations, classes and states are not containers that people move around inside. They are patterns of joint action, produced and reproduced by people interpreting each other's conduct, and they exist only for as long as that continues.

The intellectual problem is Blumer's objection to the sociology of his time, and it is a methodological one.

Mainstream sociology, in his account, treated people as media through which factors operate. Variables — class, education, gender, urbanisation — were said to produce outcomes, and the human being in between was a conduit. Stimulus in, response out.

Blumer's objection: this leaves out the only thing that actually connects the two.

Between the stimulus and the response, a human being interprets. They ask what this means, what kind of situation this is, who this person is to them, what is expected. They may interpret wrongly, or unusually, or in a way nobody predicted. And the response follows from the interpretation, not from the stimulus.

Therefore: a correlation between a variable and an outcome describes a regularity in how people are interpreting, without telling you what they are interpreting or how. It is Weber's incomprehensible statistic (4.3.1), stated as a methodological programme.

Blumer's positive claim: if action follows from interpretation, then the sociologist's job is to recover the interpretation — which requires getting close enough to the people concerned to find out what things mean to them.

The three premises

Blumer coined "symbolic interactionism" in 1937 and set out its foundations in Symbolic Interactionism: Perspective and Method (1969). Its three premises are worth having exactly.

One. "Human beings act toward things on the basis of the meanings that the things have for them."

Not on the basis of the things' physical properties, and not on the basis of what they mean to the observer. On the basis of what they mean to the actor. A chair is for sitting, for standing on, for blocking a door, for firewood, for throwing — depending on what it means here, now, to this person.

Two. "The meaning of such things is derived from, or arises out of, the social interaction that one has with one's fellows."

Meaning is not intrinsic to objects (which would be realism) and not attached by individual psychology (which would be idealism). It is produced between people. You learn what a thing means by seeing how others act towards it and how they respond to your action.

Three. "These meanings are handled in, and modified through, an interpretative process used by the person in dealing with the things he encounters."

Meanings are not applied automatically. The actor works on them — selects, checks, suspends, revises, rehearses — in an internal conversation. Blumer takes this directly from Mead's I and Me (2.3.3): the self is a process in which one indicates things to oneself and responds.

The inheritance

Pragmatism is the philosophical base, and it matters. William James, John Dewey, Charles Peirce and Mead all held that thought is an instrument for acting in the world rather than a mirror of it; that meaning lies in consequences; and that knowledge is tested in practice. Blumer's programme is pragmatism turned into a research method.

George Herbert Mead (2.3.3) supplies the core: the self arises in interaction; role-taking — the capacity to see oneself from another's standpoint — is what makes coordinated action possible; the generalised other is the internalised community; and the I/Me dialectic makes the self a process rather than a thing. Mead published almost nothing on this; Mind, Self and Society (1934) was assembled from students' lecture notes after his death, which is worth knowing when reading it.

Charles Horton Cooley (2.3.2) supplies the looking-glass self — we see ourselves through what we imagine others see — and the concept of primary groups.

W. I. Thomas supplies the theorem that is the school's motto: "If men define situations as real, they are real in their consequences." (Co-authored with Dorothy Swaine Thomas, whose name was omitted from the attribution for decades — one more instance of 4.6.1's mechanism.)

And the Chicago School (5.1.1's rival tradition) supplies the empirical practice: Park's urban research programme, and a generation of ethnographic monographs.

The methodological programme

Blumer's methodological writing is as important as his theory, and it is combative.

His central demand: research must respect the nature of the empirical world and be shaped by it, rather than forcing the world into a preconceived scheme.

Two stages: exploration — flexible, open-ended immersion in the area, designed to develop a close and comprehensive acquaintance — followed by inspection, the intensive analytical examination of the elements identified.

And his attack on variable analysis is the sharpest part. Variables, he argues, are analytically detached from the interpretive process that connects them; they are frequently arbitrary in construction; and they can only capture what has been pre-specified, which means they cannot discover anything the researcher did not already suspect.

His famous distinction: most sociological concepts should be sensitising rather than definitive. A definitive concept specifies exact attributes and tells you what to look for. A sensitising concept "gives the user a general sense of reference and guidance in approaching empirical instances" — it suggests directions along which to look. Blumer's claim is that social life is too variable for definitive concepts to fit, and that pretending otherwise produces precision without accuracy.

What this looks like when done, and why the findings are of a kind other methods do not produce.

Howard Becker, Outsiders (1963) — how someone becomes a marijuana user. Becker interviewed users and found that the physiological effects of the drug are not automatically experienced as pleasurable. A novice must learn three things, in sequence: the technique for smoking that produces effects at all; to notice the effects and connect them to the drug; and to define those sensations as enjoyable rather than as alarming symptoms.

Each stage is learned from more experienced users, and failure at any stage ends the career. Becker's conclusion is that even an experience apparently as physical as being intoxicated is socially constructed — which no variable analysis of drug use would ever have discovered, because it requires knowing what the sensations mean to the person having them.

Erving Goffman, Asylums (1961) — a year of fieldwork at St Elizabeths Hospital in Washington, largely as an assistant to the athletic director, which gave him access to patients without a clinical role. His finding: much of what was read as symptomatology was an adaptation to the institution rather than an expression of illness. The mortification of the self on admission, the "moral career" of the patient, and the secondary adjustments by which inmates carve out unauthorised satisfactions — all of it visible only from inside, over time (2.3.7).

Anselm Strauss and colleagues, Psychiatric Ideologies and Institutions (1964) — the concept of the negotiated order. A hospital appears to run on rules and hierarchy. In practice, what actually happens is continuously negotiated between doctors, nurses, administrators, patients and families, and the negotiations are renewed daily. The formal structure is a resource in the negotiation, not a description of the outcome. This is the strongest direct challenge to the structural account, produced by watching.

And notice the common feature: in each case, the finding contradicts what the official account of the setting says is happening, and it could not have been obtained without prolonged close observation. That is the school's claim to a distinctive method, and it is a good one.

Go deeper

The two schools within the school

A real division, and worth knowing because it recurs in every methods argument.

The Chicago School (Blumer, Becker, Hughes, Strauss, Goffman loosely) — interpretive, ethnographic, sceptical of quantification, committed to fieldwork, and holding that meaning must be recovered through immersion.

The Iowa School (Manford Kuhn) — accepted the theory and rejected the methodology. Kuhn argued that interactionist concepts could and should be operationalised and measured, and developed the Twenty Statements Test ("Who am I?" answered twenty times) as a quantitative instrument for studying the self.

The dispute is the discipline's methodological split in miniature: whether the phenomena require interpretive methods or whether interpretive phenomena can be measured. Neither side won, and the argument recurs in every generation.

The great criticisms

It cannot handle structure or power. The most serious charge, made from every direction. If everything is negotiated in the encounter, what accounts for the fact that some parties consistently prevail? Gouldner attacked Goffman for producing a sociology of the "cocktail party" that ignored the institutional and economic order. Marxist critics noted that a framework which cannot see class, capital or the state is not equipped for most of what matters.

The reply available: interactionists do study power — Becker on labelling and the moral entrepreneur, Goffman on total institutions and stigma, the negotiated order literature on hierarchy. But the reply is partial, and the framework is genuinely better at describing how domination is enacted in an encounter than at explaining why the parties arrived with such unequal resources.

It is astructural and ahistorical. A negotiated order is negotiated within limits nobody at the table set. The interactionist can describe the negotiation and struggles to explain the limits.

It cannot generalise. Rich, deep accounts of specific settings, with unclear scope. What does one hospital tell you about hospitals? Interactionists reply that the aim is analytic rather than statistical generalisation — the identification of processes that recur — but the question of how far a process travels remains open in each case.

And the astructural bias may be built into the method. If you study by immersion in settings, you will see what happens in settings. The distribution of resources across settings, the historical processes that produced them, and the institutional arrangements that connect them are all invisible from inside any one of them.

What is not a good criticism, and is frequently made:

"It's just describing everyday life." The findings above — that intoxication is learned, that psychiatric symptomatology is partly institutional adaptation, that hospital order is negotiated daily — are none of them obvious, and all of them contradict the official accounts of their settings.

"It denies that structures exist." Blumer's position is that structures exist as patterns of joint action that must be continuously produced. That is a claim about the mode of existence of structures, not a denial of them — and it is closer to Simmel (4.4.1) and to Giddens's structuration (5.11) than to a denial.

"It's unscientific." The demand that concepts fit the phenomena rather than the reverse, and that research proceed by exploration before inspection, is a methodological position with arguments behind it. It can be wrong; it is not unserious.

The offspring

The school's productivity is best measured by what came out of it.

Labelling theory (10.2) — Becker, Lemert, Kitsuse. Deviance is not a quality of an act but a consequence of the application of rules and sanctions by others. "The deviant is one to whom the label has successfully been applied." This transformed criminology.

Dramaturgy (2.4.2, and 5.3.2) — Goffman's whole programme.

Grounded theory — Glaser and Strauss's The Discovery of Grounded Theory (1967), the most influential qualitative methodology of the last sixty years, is a direct product of this tradition: theory generated from data through constant comparison, rather than tested against it.

Social constructionism (2.4.1) — Berger and Luckmann synthesised this tradition with Schutz and Durkheim; and Spector and Kitsuse's constructionist account of social problems asks not what a problem is but how a condition comes to be defined as one.

Ethnomethodology (2.4.3) — Garfinkel's radicalisation: not what meanings people hold, but the methods by which they produce the sense of an orderly world, moment to moment.

Conversation analysis — Sacks, Schegloff and Jefferson's development, which turned the study of interaction into a precise empirical science of turn-taking, repair and sequence organisation, and which is arguably the most methodologically rigorous thing to come out of qualitative sociology.

The sociology of emotions — Hochschild's emotional labour (4.1.4) and the feeling rules literature.

And the sociology of work, medicine, education and deviance are all substantially interactionist in their qualitative traditions.

"Symbolic interactionism is a micro theory." Its claim is that the macro consists of patterned joint action, which is a claim about all levels rather than a restriction to one. Whether that claim works is the live question.

"Blumer just restated Mead." Mead published almost nothing on this and Mind, Self and Society is posthumous lecture notes. Blumer's three premises, the concept of joint action, the methodological programme and the attack on variable analysis are his.

"The Thomas theorem means perception is reality." It says definitions have consequences — that acting on a false belief produces real effects. The bank run (5.1.2) is the illustration.

"Interactionists ignore power." They study its enactment in encounters — labelling, stigma, total institutions. The fair charge is that they cannot explain why the parties arrive so unequally equipped.

"Sensitising concepts are vague concepts." Blumer's claim is that they should suggest directions of enquiry rather than specify attributes in advance, because the phenomena vary too much. It is a position, not an evasion.

"It has been superseded." Grounded theory, labelling theory, conversation analysis, constructionism and most qualitative sociology descend from it. It stopped being a separate school by being absorbed.

Why it matters

It supplies the discipline's account of how structure is actually made. Functionalism and conflict theory both tell you what structures do; neither tells you how a structure gets from one day to the next. The interactionist answer — that it is produced continuously by people interpreting each other, and therefore has to be brought off each time — is the best available, and it is what makes change explicable at the level where change actually happens.

It underwrites qualitative method. The argument that action follows from interpretation, and that interpretation must therefore be recovered, is the justification for ethnography, in-depth interviewing and observation. Every qualitative study rests on it, whether or not it cites Blumer.

It produces findings other methods cannot. Learning to be intoxicated; the institutional production of symptoms; the negotiated order of a hospital. Each contradicts the official account of its setting and each required someone to be there.

And it names the school's own limit honestly. A framework built by immersion in settings sees what happens in settings, and is weakest on the distribution of resources across them. That limit is real, and the next school in Part 5 — and Bourdieu in particular (5.10) — is substantially an attempt to fix it.

Blumer's three premises: people act toward things on the basis of the meanings those things have for them; meanings arise out of social interaction; and meanings are handled and modified through an interpretive process.

The objection to mainstream sociology: it treats people as media through which variables operate, omitting the interpretation that actually connects stimulus to response. A correlation describes a regularity in interpreting without telling you what is being interpreted.

Joint action must be constructed anew each time. A recurring pattern looks like a structure; it is an achievement that conceals the work of producing it. Institutions persist because people keep producing them — and can therefore fail or change in ways structural accounts cannot predict.

The inheritance: pragmatism (thought as an instrument for acting); Mead (role-taking, generalised other, I/Me); Cooley (looking-glass self); the Thomas theorem — "if men define situations as real, they are real in their consequences," co-authored with Dorothy Swaine Thomas, whose name was long omitted.

The method: exploration then inspection; concepts should be sensitising rather than definitive; and research must be shaped by the empirical world rather than forcing it into a scheme.

The findings: Becker on learning to experience intoxication as pleasurable; Goffman's Asylums on symptomatology as institutional adaptation; Strauss's negotiated order. Each contradicts the official account of its setting.

The criticisms: it cannot explain why parties arrive at the encounter so unequally equipped; it is astructural and ahistorical; and its scope of generalisation is unclear. The astructural bias may be built into the method — immersion in settings shows you settings.

The offspring: labelling theory, dramaturgy, grounded theory, social constructionism, ethnomethodology, conversation analysis, the sociology of emotions.

Symbolic interactionism — the perspective that social life consists of people acting on meanings produced and modified in interaction.

Joint action — the collective form built by fitting together individual lines of action; must be constructed anew each time.

Role-taking — seeing oneself and the situation from another's standpoint.

Thomas theorem — "If men define situations as real, they are real in their consequences."

Sensitising concept — a concept suggesting directions of enquiry rather than specifying attributes in advance.

Definitive concept — a concept specifying exact attributes and indicators.

Exploration and inspection — Blumer's two research stages.

Negotiated order — Strauss's finding that institutional order is continuously renegotiated rather than fixed by rules.

Labelling — the process by which deviance is produced through the application of rules and sanctions.

Grounded theory — Glaser and Strauss's method of generating theory from data through constant comparison.

Analytic generalisation — extending findings by identifying recurrent processes rather than by statistical inference.

One — watch a joint action being produced. Stand where people queue, order, or pass each other in a corridor. Identify the moment-by-moment signals by which the pattern is achieved. Nobody is following a written rule.

Two — find a learned sensation. Coffee, spirits, chilli, exercise pain, a musical genre. Ask how you learned to experience it as pleasurable, and from whom. Becker's finding generalises further than drugs.

Three — apply the Thomas theorem. Find a case where a false belief produced real consequences, and trace how the outcome was afterwards cited as proof of the belief.

Four — test a sensitising concept. Take "stigma," "career" or "negotiated order" into a setting you know. Note that it directs your attention without telling you what to find. That is Blumer's claim, tested.

Five — look for the official account and the actual one. In any organisation, compare what the handbook says happens with what happens. The gap is where interactionist research lives.