Podcast episode 37: Interview with Michael Lynch on conversation analysis and ethnomethodology

In this interview, we talk to Michael Lynch about the history of conversation analysis and its connections to ethnomethodology.

Download | Spotify | Apple Podcasts | Google Podcasts

References for Episode 37

Button, Graham, Michael Lynch and Wes Sharrock (2022) Ethnomethodology, Conversation Analysis and Constructive Analysis: On Formal Structures of Practical Action. London and New York: Routledge.

Fitzgerald, Richard (2024) “Drafting A Simplest Systematics for the Organization of Turn-Taking for Conversation,” Human Studies. https://doi.org/10.1007/s10746-023-09700-7

Garfinkel, Harold (2022) Studies of Work in the Sciences, M. Lynch, ed. London & New York: Routledge. https://doi.org/10.4324/9781003172611 (open access)

Lynch, Michael (1993) Scientific Practice and Ordinary Action. New York and Cambridge: Cambridge University Press.

Lynch, Michael and Oskar Lindwall, eds. (2024) Instructed and Instructive Actions: The Situated Production, Reproduction, and Subversion of Social Order. London and New York: Routledge.

Lynch, Michael and Douglas Macbeth (2016) “Introduction: The epistemics of Epistemics,” Discourse Studies 18(5): 493–499. See also the articles in the special issue.

Sacks, Harvey (1992) Lectures on Conversation, Vols. 1 & 2, Gail Jefferson, ed. Oxford: Blackwell.

Sacks, Harvey (1970) Aspects of Sequential Organization in Conversation.  Unpublished manuscript, U.C. Irvine.

Sacks, Harvey, Emmanuel A. Schegloff and Gail Jefferson (1974) “A simplest systematics for the organization of turn-taking for conversation”, Language 50(4), Part 1: 696–735. Available online

Transcript by Luca Dinu

JMc: Hi, I’m James McElvenny, and you’re listening to the History and Philosophy of the Language [00:14] Sciences podcast, online at hiphilangsci.net. There you can find links and references to [00:20] all the literature we discuss. Today we’re joined by Michael Lynch, who’s Professor Emeritus [00:26] of Science and Technology Studies at Cornell University. He’s going to talk to us about [00:32] conversation analysis and its links to ethnomethodology.

It’s probably fair to say that conversation [00:40] analysis, or CA, is a well-established subfield of linguistics today, which is concerned with [00:47] studying how interaction is achieved between speakers in an oral exchange. On a technical [00:54] level, conversation analysts typically proceed by making an audio or video recording of an [00:59] interaction and then transcribing it in a heavily marked up notation that conveys elements [01:06] of intonation, overlapping speech, gaze, and so on. Using these transcripts as empirical [01:12] evidence, the analysts then put forward theories about how the back-and-forth of conversation [01:18] is structured.

The seminal publication introducing conversation analysis was a 1974 article in [01:26] Language with the title, “A Simplest Systematics for the Analysis of Turn-Taking for Conversation”, [01:33] co-authored by Harvey Sacks, Emanuel Schegloff and Gail Jefferson. These three are widely considered [01:41] the founding figures of CA. But crucially, Sacks, Schegloff, and Jefferson had not been trained in [01:48] traditional linguistics programs. They were sociologists by academic upbringing. Moreover, [01:54] they were adherents of ethnomethodology, an approach to sociology pioneered by Harold Garfinkel. [02:02]

So the question arises as to how conversation analysis fits into linguistics and this broader [02:09] disciplinary constellation. Mike, can you illuminate this question a bit for us? [02:14] Where did conversation analysis come from, and how is it placed today? [02:19]

ML: OK, well, thank you, James, for the opportunity to speak to your podcast. To start, I’d like to [02:26] add that what you said about Sacks, Schegloff, and Jefferson also applies to me. I’m not trained as a [02:34] linguist, traditional or otherwise. My background is in sociology, but also like them, [02:39] I spent a lot of my career, particularly the last 25 years at Cornell, in interdisciplinary programs [02:46] of which sociology was a part. But my take on sociology through the field of ethnomethodology [02:53] is not normal sociology, as many people would tell you. I don’t want to go into that right now.

But [02:59] you asked about the background of Sacks, Schegloff, and Jefferson and where CA came from. I know less [03:06] about Schegloff’s and Jefferson’s background a little bit, but I know more about Sacks, [03:12] partly because I’ve been spending the last year and a half reading and rereading the two-volume [03:19] set of his lectures. A little bit about Schegloff. He wrote an MA thesis on the history of literary [03:26] criticism before pursuing a PhD in sociology at Berkeley at the same time that Sacks did. [03:32] Jefferson had an education and practical experience in dance choreography before she attended Sacks’s [03:40] lectures and switched into a PhD program with him at UC Irvine, and her father was a famous [03:48] radio psychiatrist.

Sacks had a law degree from Yale in 1958 and after that decided, [03:55] to the disappointment of his parents, not to pursue a law career. He was in the MIT, Cambridge, Harvard [04:05] area when he decided he wanted to go back to sociology and political science. He had studied [04:13] sociology as an undergraduate and he met Garfinkel and I believe also Goffman, who were on sabbatical [04:20] taking seminars from Talcott Parsons, a famous sociologist and Garfinkel’s mentor. And from there [04:29] he really hit it off with Garfinkel. Garfinkel encouraged him to go to the West Coast [04:34] and he pursued his PhD in sociology at Berkeley, where he did, for a time at least, work with [04:40] Goffman, although Goffman did not sign his PhD. And he stayed in touch with Garfinkel, was part [04:47] of groups that met, kind of forming the basis of ethnomethodology, which, to put a short gloss on [04:55] it, is the study of everyday actions as they are performed, at least preferentially in the case [05:03] of CA, using recordings of interaction naturally occurring (so-called) as a material for study. [05:12]

Sacks also was very widely read. I really recommend reading his lectures or at least some of them [05:19] because there – you can still get them online. They’re out of print, I believe. It makes clear [05:26] that he’s drawing from the history of oral languages, the cultures of ancient Greece and [05:33] Rome, the studies of Judeo and biblical culture and language. He also was apprised, to what depth [05:44] I don’t know, of ordinary language philosophy, Austin, Searle to some extent, but mainly Austin [05:51] and Wittgenstein as well. He didn’t mention it much in his writings or in his lectures, [05:58] but his sensibilities were definitely shaped by that background. And he also brings in themes [06:04] from law, which is not really obvious, but when you read the lectures and some of his unpublished [06:09] writings, you find that he has kind of a legal orientation to the organization, the rules, [06:17] the norms, procedures of ordinary conversation. There’s a bit of a legal background into what [06:23] he’s saying.

Now, you mentioned the 1974 paper on simplest systematics and turn-taking by [06:30] Sacks, Schegloff, and Jefferson. It’s often taken to be the beginning of conversation, [06:34] but his lectures, starting when he was a graduate student and living in Los Angeles and teaching [06:40] at UCLA, they start in 1964, so 10 years before that, and even the earlier lectures exhibit [06:49] themes about language, about many other things that show up in part in the turn-taking paper, [06:56] although compared to the ground he covers in the lectures, which of course are much more extensive, [07:03] he’s much broader, much more varied in his interests and his analyses than that paper [07:09] gives access to. So one of the things to keep in mind is that paper is treated as a foundation, [07:15] but when you go back to Sacks’s lectures, you see that there’s a lot missing from it, [07:20] and a kind of a restricted way of going about the study of conversation that it represents. [07:26]

JMc: You say that the 1974 paper is a bit restricted in terms of the ideas that Sacks had already [07:34] developed in his earlier lectures. Could you elaborate a bit and say in what ways it was [07:39] restricted, and do you think it’s because the paper appeared in Language and was being repackaged [07:45] for linguists? And if that’s the case, then what is the relationship or what was the relationship of [07:52] conversation analysis as Sacks conceived of it to other schools of linguistics at the time? [07:59]

ML: It was edited by a linguist at UCLA named William Bright [see Fitzgerald (2024)], and he did an unusual job of [08:09] doing the sole review and advice on the paper, mainly working with Schegloff, who was originally [08:17] not listed on a very early draft of the paper by Sacks and Jefferson. And then he was listed third, [08:25] and then second after his work with Bright, I guess. It was written in a different style than [08:33] some of Sacks’s earlier work, which also was very difficult to fathom, and lectures, [08:38] which are not so difficult to fathom, although very thought-provoking.

The main topic of that [08:44] paper is turn-taking issues, that is one person speaking, coming to an end, another person, [08:52] or multiple persons, then vying for next turn, and so on and so forth. And that indeed is a [09:01] major theme in Sacks’s work and in Schegloff’s work, Jefferson’s, but there’s also a broader [09:08] conception of sequential analysis that’s in the lectures and also in unpublished manuscripts [09:15] that went through several drafts that Sacks wrote and which is yet to be published, but [09:22] presents, again, a somewhat different cast of the sequential analysis than you get in [09:29] this more exclusive interest in turn-taking and turn-transition [09:35] and the beginnings of turns in the 1974 paper. So there’s also lots of other themes about [09:44] phenomena that tie together utterances, not just at the beginnings and ends, but which [09:51] show topical continuity and coherence in a very interesting way in Sacks’s lectures. [09:59]

Sacks, in occasional remarks in his lectures and in a couple of papers where he talked about [10:05] understanding and organization of talk in a way that he sharply distinguished from [10:11] the orientation of linguistics, and the simplest aspect of this distinction that he emphasized was [10:18] that linguists treat the sentence as the basic unit and structural constituents of sentences [10:24] as embedded in sentences and as individually organized, cognitively or even neurologically, [10:34] to the extent that they could do that. He looks at sentences, parts of sentences, utterances, [10:41] in connection with those of other participants in conversation. And in some cases, you can get [10:48] a very different sense of not only the form, but also the meaning. He generally treats meaning [10:54] sideways in the sense that he doesn’t talk about it directly. He talks about in connection to [11:01] practices and structures of conversation, that you get a different sense of what’s being said [11:07] than when you take a sentence in isolation. Maybe linguists have caught on to this, but [11:14] at that time, and I think predominantly now, that orientation was distinctive of what Sacks and his [11:23] company were doing. [11:25]

JMc: I mean, in the mid-20th century in America, there were also schools of linguistic [11:31] anthropology, sort of ethnography of speaking, and so on, that looked at discourse and the use [11:38] of language in a particular cultural context. Do you think that Sacks would have felt that they [11:42] were still stuck with the sort of formal conception of a sentence as the basic unit of language? [11:48]

ML: He did know and addressed work by, you know, Gumperz and Hymes, and he actually participated [11:56] in a book they co-edited, and he knew of people in sociolinguistics, and Goffman himself was his [12:08] main contact at Berkeley when he was a student, and Schegloff was a student there too. [12:12] And there were differences. It seemed like a superficial difference, but for Sacks, [12:19] it’s very important, and I think there’s much to say about the difference, that his method of [12:27] working was usually, but not always, but usually he would try to record what he called naturally [12:36] occurring, or Garfinkel called, naturally organized everyday actions. So, bugging a phone or [12:45] recording – one of his favourite examples was some recordings he made behind a one-way mirror [12:53] of a group therapy session involving these mildly delinquent kids in Los Angeles at some point in [13:01] the probably early ’60s. And he goes to these tapes again and again, hears the same sequences, [13:07] discusses them again and again, often with a somewhat different framing in his lectures, [13:14] and finds in those recorded conversations, as he put it, things you would never imagine, [13:22] right, that people would say, and organizations of talk that you just don’t remember when, [13:28] you know, you think of a conversation you might have, or when you imagine an ideal typical [13:34] conversation.

And you find in Goffman and in social psychology, and in even some of the [13:44] more linguistically inclined sociolinguists, that they either still work on things like speech acts, [13:52] which are largely the actions of one person. They see the person as the organizational basis of, [14:00] and the person’s psychology or cognition, as the organizational basis of the structure of talk, [14:07] where moving the frame to sequences, and not just pairs of utterances, but more extended [14:15] connections and ties between one’s own and others’ utterances in an ongoing stream. [14:22] It’s not a stream of consciousness. It’s a stream of talk, which we’re recording, at least [14:28] Sacks, but it could be adequate to capture, not necessarily complete, but adequate for starting [14:34] a starting point.

It gives you a very different insight. It’s not just that, you know, he’s being [14:38] empiricist, always wanting stuff recorded from the ground. He also used newspaper articles and [14:46] snippets from the Bible and all sorts of stuff. But his main resource was recorded conversation [14:52] that he could play again and again and again. And another aspect of it was he could, [14:56] with transcript, which he didn’t treat as the primary ground, the recording was the primary [15:02] ground as, you know, an adequate record of what people were doing. Assuming they spoke a language [15:09] you spoke and you had enough insight into who they were, what they were talking about and so forth, [15:14] that you could find recognizable structures that required no special skill, no special [15:24] knowledge to recognize and to try to stay with that rather than try to override it with [15:29] an overly technical understanding. That those materials he saw to be a source of insight, [15:36] not just material from which to derive inductive inferences. [15:41]

JMc: So what does structure mean to an ethnomethodologist, and specifically to Sacks? [15:49]

ML: That’s a very good question. Sacks had a love for machine metaphors. He talks about machinery [15:57] of conversation, the turn-taking machine. Occasionally in his lectures, he acknowledges [16:05] that when he’s talking about machinery, he’s talking about rules, or you could even say [16:09] maxims, or, you know, regularities, even, that occur, but he just loved to talk about machinery. [16:18] And he also loved to invest agency in the machinery, rather than in people’s intentions, [16:27] motives, cognitive organization, right? So it was kind of a gestalt shift from the speaker [16:37] to the speaking in concert with others as the, not ultimate origin necessarily, but as [16:45] an organizational basis for what people are doing, saying, orienting to, and so forth. [16:51] It’s not that he emptied the person. Gail Jefferson once made a joke about, [16:58] “Sacks was somebody who treated people in the same way that you would treat algae.” [17:04] He has a line in his lectures that is really funny where he says he’s got nothing against [17:10] anthropomorphizing humans any more than when physicists anthropomorphize their data. [17:17] He’s got nothing against it, but nothing particularly in favour of it. So there’s this [17:23] kind of strange indifference that he expresses, but it leads to a very unique insight. [17:31]

JMc: But at the end of the day, Sacks still talks in terms of rules, maxims, or structures and so on, [17:39] because isn’t it a sort of, would it be reasonable to say that one of the core ideas of [17:45] ethnomethodology is that the ethnomethodologist seeks to discover organization sort of from the [17:51] perspective of the participants in a particular situation? [17:56]

ML: Yes, and I think that Sacks held to that. And the perspective of the participants didn’t require [18:05] some sort of magical trip of mind reading. But in his case, not necessarily Garfinkel’s, [18:13] in his case, he used the overt recording materials, the surface, [18:22] as the organization that the members were paying attention to insofar as they would hear what the [18:31] other is saying and react often without hesitation in a way that showed an understanding, or in some [18:39] cases a misunderstanding, of what the other said, and that would be then dealt with downstream in [18:45] the conversation. And so he was treating the surface materials, which sounds very shallow, [18:51] but in this he had some backing by the likes of Wittgenstein. And the skepticism about having [19:01] to always delve into interpretation, reading between the lines and that kind of thing, was [19:08] not his procedure. And he had a deep basis for that in both Wittgenstein, Garfinkel, and to some [19:18] extent Goffman.

And so there is this orientation in the analysis to, “What are the parties doing?” [19:27] And it’s very important to know that the term “conversation analysis,” which Sacks didn’t use, [19:34] actually, at least not in his lectures, he talked about the analysis of conversation, [19:41] and he and many of his colleagues for a while talked about conversational analysis, A-L, [19:47] “conversation” with “al” at the end. And it got conventional to talk about CA or conversation [19:55] analysis, and everybody went along with that. But the idea was that the analysis is being done [20:02] on the ground floor by the person’s talking. It’s not something where you take data, you code it, [20:09] you do experiments to try to eliminate the lack of comparability from one occasion to another. [20:18] And for him, the problem was to address how it is that parties hearing what they hear, [20:24] knowing what they know, can continue in the way they continue in a conversation. [20:30] And how do they respond to what another says? Now, it may be they misinterpret it, or it may be [20:36] that they interpret differently than the speaker meant, and the speaker doesn’t indicate that [20:43] that’s the case. I mean, there’s a lot of things that can happen, but the orientation analytically [20:49] was to try to recover, as Garfinkel would call it, what persons were doing. So that the rules, [20:56] say the rules of turn-taking or the facts of it, as they talk about the turn-taking paper, [21:02] that one speaker speaks at a time, transitions occur without gap or overlap, as both a description [21:09] and in some sense, a basis for normative organization, that these are not strict [21:16] inviolable rules. They are procedures that also have noticeable, regular features that you could [21:24] call structures in the way conversation is organized. And Sacks tried to then delve into [21:30] that to try to answer the question, how do members do it, given that they’re flying by the seats of [21:37] their pants with very limited time constraints on understanding and response, especially in a [21:43] situation where there’s competitive talk, that there’s no timeout. And so how do they do that [21:49] is his big question, and how do they reconcile things like that speaker change recurs in [21:56] conversation, that is, you know, one speaker speaks, another does, etc., etc., [22:00] that with the idea that they can do it without gap or overlap, how do they do that? [22:05] And he had a lot to say about that. I can’t summarize it in a few words, but that was the problem. [22:12]

JMc: Just to sort of summarize the picture of how CA came into existence, do you think it’d be [22:17] fair to say that Sacks was the great theoretician and Jefferson provided the sort of technical [22:25] apparatus required through her transcription system? [22:29]

ML: Well, I think you have to also mention Schegloff, since he was the major figure in the period of time after ’75 and until he stopped [22:37] working in 2012 or ’10. Jefferson struggled to maintain a career. She never thought of herself [22:44] as a sociologist. I’m not sure what she thought of herself as. She was a conversation analyst. [22:50] And she spent the last roughly 20, 25 years of her career living in the Netherlands as an independent [22:59] scholar, occasionally employed, but mainly working on her own stuff. I was told, I haven’t seen it, [23:07] I’d love to see it actually. She transcribed the Watergate hearings [correction: Watergate tapes recorded in Nixon’s White House office], or at least a good part of [23:13] them. And I don’t know what’s happened to that transcript because she died in, I think it was [23:18] 2007. And I don’t know what’s happened to those records, but she kind of faded out of the scene [23:25] pretty early on, and Schegloff was the major character.

And Schegloff and Sacks obviously [23:32] worked closely together. I think Schegloff had a somewhat different, more structured, [23:38] more disciplined orientation than Sacks, which was probably good for maintaining CA as a [23:45] quasi-discipline, sub-discipline, whatever you want to call it. But Sacks was not just a [23:51] theoretician. He was widely read, very creative. During his lifetime, people called him a genius. [23:58] I went to Irvine, somebody told me, “This guy’s a genius.” Not that… That’s not necessarily the reason I went [24:03] there, but… And it’s sort of like, yeah, he was a genius, but I don’t believe in the concept. [24:10] He did more than just theorize. I think, again, if you read the lectures, you get a sense of [24:16] the various things he did. It wasn’t always the same from beginning to end. And there’s [24:21] different threads of his analysis that have been picked up, particularly what he called [24:26] membership category analysis, which has an attraction for some people. [24:33] So he was involved in the production of it. I think, though, he was, in his own words, [24:39] sort of the methodologist of ethnomethodology, and Schegloff worked differently, and Sacks kind of [24:46] went along with that in some of the stuff they collaborated with.

To break it down into, yeah, [24:52] there was Jefferson’s transcription system, which, yeah, she developed and deserved credit for it. [24:59] But more than that, she deserved a lot of credit for some of the analyses she did, [25:03] which are brilliant. She was really an amazing character. And Schegloff is also a very formidable [25:11] intellect. And so all three of them had their own shape in what they did, and it didn’t break [25:17] down in terms of theory and technical aspects of it. It was much more varied for all three of them. [25:24]

JMc: And what’s Garfinkel’s relationship to conversation analysis? [25:31]

ML: Yeah, inconsistent. Informally, he was very disappointed with the direction that CA had [25:38] taken, but at the same time, particularly in public statements to other sociologists, [25:45] he would really defend CA, and he would say, and I think he meant this, [25:49] that it was the crown jewel of ethnomethodology. It was the most developed, most technically [25:55] developed, most procedurally developed area of ethnomethodology. But it also diverged from [26:02] ethnomethodology. And I think people who currently come into CA, particularly from other fields other [26:09] than sociology, just don’t see much connection with Garfinkel. He’s treated as kind of a woolly [26:16] predecessor who spoke incomprehensibly and was besotted with phenomenology, etc., etc. [26:27] And certainly there are differences. Yet you can find in Sacks’s work and also Schegloff’s and some [26:34] of Jefferson’s that they were doing ethnomethodology at the same time they were [26:39] also developing CA as an independent field with its own interdisciplinary links, [26:48] not just to linguistics, but to communication studies, to psychology to some extent, [26:55] anthropology. You know, language, nobody owns language, ordinary language particularly, and [27:02] so it shows up in odd places. [27:06]

JMc: Great. Well, thank you very much for answering those questions.