Online with Garfinkel


Jan Blommaert

The Durkheim and the Internet project (DAI in what follows) being completed, I now move on towards a more radical exercise: using some of Harold Garfinkel’s central intuitions as a lead into forms of online analysis. This exercise, I should underscore, builds onto DAI and does not replace or qualify it – it extends it. For a summary of DAI, see Blommaert (2018).

This extension is warranted, I believe, because of one methodological outcome of the project: the “four lines of sociolinguistic methodology” that I designed as a way to investigate new forms of collectivities in online-offline contexts. Here they are:

  1. Patterns of communication necessarily involve meaningful social relationships as prerequisite, conduit and outcome;
  2. Such relationships will always, similarly, involve identities and categorizations, interactionally established;
  3. Thus, when observing patterns of communication, we are observing the very essence of sociation and “groupness” – regardless of how we call the “groups”.
  4. And specific patterns of interaction shape specific forms of “groups.

I added the following reflection to these four lines:

“Groups, then, are not collections of human beings but patterned sets of communicative behaviors and the relationships with which they are dialectically related. Whenever we see such ordered forms of communicative behavior, there is an assumption of active and evolving groupness – sociation – but the analytical issue is not the nature of the group (or the label we need to choose for it) but the specific social relationships observable through and in communication. All other aspects of sociation can be related to this. So if one needs the definition of a group: a group is a communicatively organized and ratified set of social relationships.”

This analytical point pushed me to a re-examination of Garfinkel’s work, notably Garfinkel (2002). I shall not follow Garfinkel in any canonical way, however. The nature of the exercise I undertake here would prevent it, and the fact that Garfinkel’s incredible methodological idiosyncrasy makes much of his book barely readable further supports that decision. Fortunately, Anne Rawls does a great service to Garfinkel in her introductory essay to the book (Rawls 2002 and other contextualizing essays, 1987, 1989). And finally, I reject several of Garfinkel’s assumptions and principles. But there remains much that can be profitably reformulated and redeployed as well. Let me summarize these reformulated elements.

Garfinkel’s intuitions

Let me start by listing what I see as productive fundamental intuitions in Garfinkel’s work. The connecctions with the “four lines” above should be clear at once.

  1. Garfinkel focuses on social order as a locally accomplished social fact. In this, he is entirely empirical, in the sense that he rejects any conceptual a prioris and prioritizes the actual, observable social actions as a site of “structure” and “theory”. That naturally implies that Garfinkel rejects the old binaries of “micro vs macro” or “structure vs agency”, as well as an ethos of scientific practice in which conceptual and theoretical “implementation” is sought.
  2. Rather than to take (predefined and “known”) individuals and groups as a starting point in his analysis, he takes situated actions as the point of departure; the people acting within such situations are merely the “local staff, its local production cohort” (Garfinkel 2002: 247). And in line with G.H. Mead, action is interaction.
  3. Actions can be shown to have “autochthonous order properties”, i.e. “empirically observable properties of the congregational work of producing social facts” (id. 245). Rawls (ibid, FN) further clarifies: “Congregational refers not only to to the idea that these social facts are made collaboratively by a group, but that the population cohort has its cohort or congregation by virtue of being engaged in doing just this thing”.
  4. In other words: groups are made by the actions they are involved in, and Garfinkel emphasizes “situations that provide for the appearances of individuals” (Rawls 2002: 46).
  5. Such involvement is predicated on the recognizability of actions and their properties of order. Social actions occur as formats, the characteristic features of which are recognizable to others and, thus, intelligible as action x, y or z. Garfinkel’s example of a queue (2002, chapter 8) is telling: it is the queue itself that organizes the behavior of people as a queue. The queue has a set of “properties of order without which the phenomenon ceases to be recognizable as what it is” (Rawls 2002: 45).
  6. This aspect of formatting is reflexive: there is no “external” or explicitly stated rule for action, but its execution “must work and be seen to work by others” (Rawls 2002: 41). Thus, rules become reflexively apparent after their implementation in social action. It’s when a queue has been formed that people can tell you that there is a queue, and that it starts thère, not here. Social actions “have a [normative] coherence when one is finished with them that they did not have at the outset” (ibid).
  7. Recognizability and reflexivity as features of social action involve and presuppose at least two things: (a) that no social action is “individual” in any sense of the term but always interactional; (2) that the formats of social action need to be learned, acquired.

It is clear that Grafinkel attributes a sui generis character to situated social action and its forms of order: its characteristics cannot be reduced to individuals’ intentions and interests, nor to external (“institutional”) constraints. In fact, the sui generis character of situated social action is an echo of Durkheim’s qualification of “social facts” as having a sui generis quality – the very foundation of Durkheim’s sociology. And just like Durkheim’s statement, Garfinkel’s is easily (and widely) misconstrued. So we must be precise here. The sui generis character of situated social action involves – contra methodological individualism – that individual social beings are constrained in their choices of action; people rather “enter into” the ordeliness of situated social action, as soon as such an order is recognizable, and attribute intelligibility to their own actions in that way. Their actions become meaningful to others by entering into the orderly procedures that make such actions recognizable as specific actions.

Garfinkel joins Goffman here, and Rawls attributes the same sui generis character to Goffman’s notion of interaction order: “the interacion order has an existence independent of either structures or individuals” (Rawls 1987: 139). This point, too, has often been overlooked, and Goffman’s concept of self, consequently, has often been misrepresented as strategically performed identity, central to his social theory. In actual fact, Goffman’s self is “a dramaturgical effect arising diffusely from a scene that is presented” (Goffman 1959: 253, in Rawls 1987: 139; italics added). So it is not just “performed” but interactionally ratified – morally sanctioned – by others: “both in its capacity as performer and performed, the self ultimately depends upon interaction” (ibid.). Such interactions require a scene – an orderly and recognizable situation that makes the dramaturgical effect (the particular, enacted and ratified self) an intelligible outcome of social action. In Rawls’ (ibid) terms:

“The self is therefore not the ontological starting point for a theory of social order. For Goffman it is an end product, the existence of which depends upon a presentation order which is the primary constraint of situations of co-presence”.

This presentation order is replete with reciprocally exchanged moral expectations – “involvement obligations” – providing a degree of security in social encounters (cf. Rawls 1987: 140). There is slightly more space for empathy and anxiety in Goffman’s view of order than there is in Garfinkel’s, and Goffman’s “ground rules of interaction” are moral ones (id.: 142). Goffman’s insistence on the ritual character of interaction (often seen as an insistence on communicative routine) is in actual fact an insistence on the maintenance of a moral order in social action. And this is done in view of the interaction order itself (sui generis), “and not directed toward the reproduction of social structure at all” (id. 145).

Rawls here brings Goffman and Garfinkel together once again: both rejected “micro vs macro” and “agency vs structure” distinctions, since for both, whatever we understand by “structure” should be empirically observable in the orderly features of actual, situated social action; the former actually coincides and identifies with the latter. And in both, the self is an outcome, a product, an effect of the orderliness of situated social action – which, consequently, should be attended to in full detail. In most work, situated social action would be seen as a building block or a reflection of “larger” social-structural phenomena (power, class, gender, race, etc.). What we have here is a radical refusal to treat situated social action as “just” the small stuff that relates to bigger stuff. Instead, we get a view in which the big things are right there, in and through situated social action – which is, consequently, a big thing. Social order in any form is interactional.


Garfinkel’s radicalism is certainly appealing because it refutes most of mainstream social theory, with a particular vehemence reserved for deductive theory-internal analysis, concepts-as-realities and simplistic interpretations of “micro vs macro” and “agency vs structure”. Aspects of this refutation are compelling and inescapable, while others are potentially fertile as a heuristic, and still others are probably nonsense. Thus, I will adopt the elements I sketched above and add two important qualifications to them.

  1. I maintain the theoretical framework designed in DAI, with its emphasis on complexity, mobility, scalarity and polycentricity. The “social order” and its “autochthonous order properties” that Garfinkel was after (and Goffman’s “interacion order” and its “involvement obligations”) are, consequently, made more precise and accurate when we see them as ordered indexicalities occurring in social arenas that are by definition polynomic, dynamic and flexible.
  2. Garfinkel’s view of situated social action as necessarily recognizable presupposes a mutually assumed sharedness of expectations (which he confirms), and of resources. The latter remains unaddressed, while it is precisely the sociolinguistic dimension of DAI. While situated social action may be a form of order sui generis, the stuff that enters into such actions isn’t: it is conditioned historically and assumes its concrete shape in interactions in the form of entextualizations, the nature and valuations of which need to be learned and acquired. So here is the second qualification to Garfinkel’s intuitions: we need to add to them an awareness of the concrete historical conditions enabling certain forms of action to assume certain kinds of order not others. This, I underscore, does not mean that we need to revert to an older vocabulary of institutionalization, routinization or even “macro” or “structural” aspects of action. What we need to do is to see situated social action as historically conditioned (and we can take some cues here from Bourdieu, for instance). This, I believe, is crucial if one wishes to maintain the claim about the sui generis character of the orderliness of such situated social action.

The historical conditions for action include infrastructural conditions as well. I underscore this because we intend to go online with Garfinkel – entering into a world not just of queues in front of the Starbucks counter at LAX, but of virality, memes and social media profiles. And a world not just of presenting and presented selves but of selfies – new technologically mediated modes of self-presentation for which Garfinkel, Goffman and others provides necessary, but insufficient, analytical frames. Such infrastructures have changed the “order” of social actions, and we must take them on board.


Blommaert, Jan (2018) Durkheim and the Internet: Sociolinguistics and the Sociological Imagination. London: Bloomsbury.

Garfinkel, Harold (2002) Ethnomethodology’s Program: Working Out Durkheim’s Aphorism (ed. Anne Warfield Rawls). Lanham: Rowman & Littlefield.

Goffman, Erving (1959) The Presentation of Self in Everyday Life. New York: Doubleday Anchor.

Rawls, Anne Warfield (1987) The Interaction order sui generis: Goffman’s contribution to social theory. Sociological Theory 5/2: 136-149.

—– (1989) Simmel, Parsons and the interaction order. Sociological Theory 7/1: 124-129.

—– (2002) Editor’s introduction. In Garfinkel (2002): 1-64.





Author: jmeblommaert

Taalkundig antropoloog-sociolinguist, hoogleraar Taal, Cultuur en Globalisering aan Tilburg University. Politiek publicist.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s