Ethnomethodology and Breaching Experiments

Goffman showed how performances sustain a definition of the situation (2.4.2). Beneath performance there is a further layer: the tacit work by which a situation becomes recognisable as anything at all.

That is Harold Garfinkel's subject, and the discipline he founded has an ugly name and an extraordinary finding.

Garfinkel asked his students at UCLA to try a small experiment: engage an acquaintance in ordinary conversation and, without indicating anything unusual, insist that the person clarify the sense of their commonplace remarks.

Here is one of the exchanges his students recorded.

Subject: How are you? Experimenter: How am I in regard to what? My health, my finances, my school work, my peace of mind, my— Subject: (red in the face and suddenly out of control) Look! I was just trying to be polite. Frankly, I don't give a damn how you are.

And another:

Subject: I had a flat tyre. Experimenter: What do you mean, you had a flat tyre? Subject: (appeared momentarily stunned, then answered in a hostile way) What do you mean, "What do you mean?" A flat tyre is a flat tyre. That is what I meant. Nothing special. What a crazy question!

Now the reaction that matters. These are not people being teased about something trivial. Garfinkel's students reported subjects becoming anxious, bewildered, embarrassed and angry — far out of proportion to a slightly pedantic question. Several friendships were damaged. In another experiment, students were asked to spend an hour at home behaving as polite lodgers rather than family members: parents wept, siblings shouted, and one family concluded the student was having a breakdown.

Nobody had broken a rule of etiquette. Nobody had been rude. What had been withdrawn was something else entirely, and its withdrawal was intolerable.

The idea, in plain words

The name is literal and unlovely: ethno (members of a group) plus methodology (their procedures). Garfinkel coined it while studying jury deliberations and noticing that jurors were doing something remarkable — using an ordinary, everyday competence to produce recognisably legal decisions, without legal training.

Four core concepts

Indexicality. Every expression depends for its sense on the context of its utterance. "It's here." "He said no." "Later." "That one." Strictly, all language is indexical: a full specification of what an utterance means would require an endless account of context, and members never provide one. They rely instead on the et cetera clause — the tacit agreement to let the unstated remainder stand.

Notice what the breaching experiment did: it refused the et cetera clause. It demanded that the indexical be made explicit — which cannot be done, because each clarification is itself indexical and would require further clarification. The subject's rage was the rage of someone caught in an infinite regress they cannot name.

Reflexivity. In Garfinkel's specific sense (not the self-examination sense of 1.5.3): accounts of a situation are part of the situation they describe, and help constitute it. Saying "this is a serious meeting" does not merely describe the meeting; it makes it one. Descriptions and the described are not separable.

Accountability. Members conduct themselves so that their actions are observable-and-reportable — so that others can see what they are doing and, if asked, they can give a reasonable account of it. Social order, on this view, consists largely in the ongoing production of accountable conduct. Note how much of what you do is shaped by the anticipation of possibly having to explain it.

The documentary method of interpretation. Perhaps the most striking. Members treat each appearance as "the document of" an underlying pattern — while the pattern itself is known only through such appearances. Each new item is interpreted in the light of the presumed pattern, and simultaneously used to elaborate the pattern. It is circular, and it is how people make sense of everything from a friend's mood to a colleague's motives.

Garfinkel's most remarkable demonstration.

Students were told they were participating in a trial of a new approach to psychotherapeutic advice. Each was asked to describe a real personal problem to a "counsellor" in an adjacent room, then to ask questions that could be answered yes or no. The counsellor answered through an intercom.

The answers were predetermined by a table of random numbers.

What happened is the finding. The students did not conclude the answers were nonsense. They worked, sometimes very hard, to make them meaningful. A "no" that seemed to contradict an earlier "yes" was interpreted as the counsellor's way of getting them to think again, or as advice about a different aspect of the problem, or as reflecting information the counsellor had inferred. Inconsistencies were retrospectively reconciled. Students reported the advice as helpful.

Garfinkel's point is not that the students were foolish. It is that making sense is something people actively do, continuously, using the documentary method — and that the achievement is so robust it can produce coherent meaning out of random noise.

Two consequences follow, and both are large.

The order we experience in social life is not simply found. It is produced, locally, by the participants, in the course of the activity — Garfinkel's phrase is that social facts are "ongoing accomplishments."

And the finding generalises uncomfortably. Horoscopes, therapeutic interpretations, ambiguous feedback, a boss's cryptic remark, a divine sign, an algorithmic recommendation: wherever people expect meaning, they will construct it, and the coherence they experience is their own work.

Against the cultural dope

Garfinkel had a target, and naming it clarifies the whole project.

He objected that mainstream sociology — Parsons above all (5.1) — treated people as what he called cultural dopes: figures who act out the stable features of a society by following internalised norms, without judgement, without skill, and without any account of how a rule gets applied to a situation.

His objection is sharp and correct. A rule does not apply itself. Knowing "be polite" tells you nothing about what to do in this room, with this person, at this moment. Between any rule and any action lies an enormous amount of interpretive work — assessing what kind of situation this is, what the rule would look like here, and what would count as having followed it. That work is skilled, it is done continuously, and Parsons' framework had no place for it.

So ethnomethodology is not a rejection of social order. It is an insistence that order is achieved rather than obeyed, and that the achieving is a competence worth studying in its own right.

This is why Garfinkel belongs in 1.2.1's six answers to the problem of order, and why he sits alongside Goffman as the discipline's most serious student of the micro level.

Go deeper

Agnes: gender as an ongoing accomplishment

One of Garfinkel's most influential studies, and it must be handled with care.

"Agnes" was the pseudonym of a person Garfinkel studied in the late 1950s at UCLA — assigned male at birth, living as a woman, seeking surgery. Garfinkel's interest was not clinical. It was that Agnes had to achieve something consciously and effortfully that others accomplish without noticing: being recognisably, unremarkably a woman in every ordinary situation.

Because she had to work at it, she could describe it. Her account made visible an enormous body of practice — how to sit, how to speak about a past, how to handle a question about school, what to know and what not to know, how to manage a situation where a wrong assumption might be exposed — that is entirely invisible to those who acquired it in childhood.

Garfinkel's conclusion: gender is not a status one has but a continuous practical accomplishment, produced and re-produced in every interaction, and normally invisible precisely because it is so thoroughly routine. That insight runs directly into West and Zimmerman's "Doing Gender" (5.7, 8.6), one of the most cited papers in the discipline.

Two honest notes. Agnes later disclosed that part of her account to the researchers had been untrue — which Garfinkel reported, and which is itself a finding about accounts. And the study belongs to its period: the framing, the clinical gatekeeping and the language reflect the late 1950s, and contemporary scholarship on trans lives has moved a long way beyond it. Read it for what it demonstrates about the achievement of a taken-for-granted status, and read it with awareness of its context.

Studies of work, and where the tradition went

Garfinkel's later programme turned to studies of work: the actual, moment-by-moment competences through which practitioners produce their disciplines. Laboratory scientists producing a finding; air traffic controllers producing safe separation; jurors producing a verdict; mathematicians producing a proof; call centre workers producing a service encounter.

The insistence is always the same: describe what practitioners actually do, in detail, rather than what the official account of the occupation says. This has proved genuinely useful — in the design of technology and workplaces, in safety-critical industries, and in the sociology of science (5.14), where laboratory studies owe a substantial debt.

The tradition's most productive offspring is conversation analysis (2.4.4), founded by Garfinkel's student Harvey Sacks, which took the same commitment — order at all points, evidence in the data — and applied it to talk with formidable rigour.

The criticisms

It brackets power and structure. The standard objection: ethnomethodology describes how order is produced and declines to ask whose order, or why the participants occupy the positions they do. A conversation between a landlord and a tenant is analysed as a conversation. Practitioners answer that context should be shown to be relevant in the data rather than assumed from outside — a methodological discipline with real integrity and real costs. Read alongside 2.2.9 and 2.1.12.

It can look trivial. Enormous care spent on greetings and pauses. The reply is that this is where social order is actually manufactured, and that the appearance of triviality is exactly what a taken-for-granted background produces.

The writing is famously difficult. Garfinkel's prose is genuinely hard, sometimes gratuitously. 1.4.6's warning about jargon applies, and it applies to a tradition whose findings deserve better sentences.

And the breaching experiments would not now be approved. Deliberately distressing people without consent, damaging friendships and alarming families, fails every contemporary ethical standard (7.2). They were a demonstration, not a method, and Garfinkel largely abandoned them. Do not replicate them on anyone who has not agreed.

"Ethnomethodology is about breaking rules." The breaches were a demonstration device. The subject is the methods people use to produce recognisable order.

"Reflexivity here means self-examination." In Garfinkel's usage it means that accounts are part of what they describe. Different concept, same word — check which is meant.

"It denies that structure exists." It insists structure is accomplished rather than merely obeyed. That is a claim about mechanism, not about existence.

"The counselling experiment shows people are gullible." It shows sense-making is active work, robust enough to produce coherence from randomness. That is a competence, not a defect — and it applies to you.

"Agnes proves gender is a choice." It shows gender is a continuous practical accomplishment, normally invisible because routine. That is a claim about production, not about choice.

"You should try the breaching experiments." Not on people who have not consented. They caused real distress and would not pass any ethics review today.

Why it matters

It identifies the layer beneath everything else in Part 2. Norms, roles, institutions and performances all presuppose participants who can recognise what kind of situation they are in. That recognition is skilled work, and Garfinkel is the only theorist who made it the object.

It shows order being manufactured in real time. 1.2.1 asked how millions of strangers cooperate. Part of the answer is that they are, continuously, doing interpretive work to make each other's conduct recognisable — and the work is invisible until it fails.

It explains why you can make sense of almost anything, and warns you accordingly. The documentary method is why ambiguous feedback, cryptic remarks and randomly generated advice all yield meaning: the coherence is partly your own production.

And it delivers the finest available demonstration of 1.1.2's habit. Seeing the strange in the familiar is difficult because the familiar is a background you are using rather than looking at. Garfinkel found a way to make it visible — by taking it away — and the violence of the reaction is the measure of how much of your world it holds up.

Ethnomethodology studies the methods ordinary members use to produce and recognise social life as orderly and accountable. Garfinkel's shift: from what rules do people follow to what do they do to make situations recognisable.

Breaching experiments withdrew the unstated common ground of interaction and produced anxiety, bewilderment and anger far out of proportion to the offence — showing that the background is constitutive, not merely convenient.

Four concepts: indexicality (sense depends on context; the et cetera clause lets the remainder stand), reflexivity (accounts are part of what they describe), accountability (conduct is produced to be observable-and-reportable), and the documentary method of interpretation (appearances read as documents of an underlying pattern known only through appearances).

The counselling experiment — random yes/no answers experienced as meaningful advice — shows sense-making as active, robust accomplishment.

Against the cultural dope: a rule does not apply itself; order is achieved, not obeyed.

Agnes demonstrated gender as a continuous practical accomplishment, feeding directly into "doing gender" — and should be read with awareness of its period and of Agnes's later disclosure.

Criticisms: brackets power and structure, appears trivial, is written impenetrably, and used methods that would not now be approved.

Ethnomethodology — the study of members' methods for producing and recognising orderly social life.

Breaching experiment — a demonstration in which taken-for-granted expectations are deliberately disrupted.

Indexicality — the dependence of an expression's sense on its context of use.

Et cetera clause — the tacit agreement to let unstated background remain unstated.

Reflexivity (Garfinkel's sense) — the constitutive relationship between accounts and the situations they describe.

Accountability — the production of conduct as observable-and-reportable to others.

Documentary method of interpretation — treating appearances as evidence of an underlying pattern that is itself known only through those appearances.

Ongoing accomplishment — social order understood as continuously produced rather than pre-existing.

Cultural dope — Garfinkel's term for the passive rule-follower assumed by mainstream theory.

Studies of work — detailed description of the practical competences through which occupations are actually performed.

Doing gender — the production of gender as a continuous interactional accomplishment (West and Zimmerman, 5.7).

One — find the et cetera clause. Take one sentence someone said to you today and try to specify everything a stranger would need to know for it to mean what it meant. Notice that you cannot finish.

Two — spot the documentary method. Recall a time you decided what someone "really meant." List the items you treated as evidence, and note that you selected them using the pattern they were supposed to establish.

Three — a consented breach. Ask a willing friend to let you conduct a small breach — respond to "how are you" with a genuine two-minute answer, or ask for clarification of one commonplace remark. Debrief immediately. Note what you felt as well as what they felt.

Four — an accomplishment you can see. Identify something you have had to do consciously that others do without noticing — speaking a second language, navigating a class you were not raised in, managing a chronic illness in public. Write out the practices involved. That is Agnes's contribution, applied to yourself.