Learning from Harold Garfinkel’s Studies of Work in the Sciences

Introduction

In mid-October 1980, at the opening plenary session of a joint meeting of philosophers, sociologists and historians on the present state of social studies of science in Toronto (Canada), sociologist Harold Garfinkel gave an address on “The Work of a Discovering Science”. Garfinkel presented his analysis of an audio tape recorded in 1969 by astronomers working at night at Kitt Peak National Observatory in Arizona (USA) while searching for optical signals from a previously detected pulsar (thought to be a fast-rotating, radiation-emitting neutron star). This recording – intended by the scientists as a resource for subsequently describing their work in writing – documented the utterances of these astronomers while using a telescope and detector to search for, and find, periodic signals from the pulsar.

Garfinkel’s presentation – and its published version (Garfinkel et al., 1981) – perplexed many scholars. Drawing almost exclusively on an audio recording of a process that was obviously situated in the observatory’s dome as well as in astronomical research more broadly, Garfinkel described the pulsar as a “cultural object” that was discovered in the “first-time through” of the “night’s work”, noted this work’s “local historicity”, and invoked Gestalt notions (like finding “the animal in the foliage”) to describe these scientists’ achievement. He thereby offered a novel vocabulary for describing practical action and did so in a style of writing that was notoriously dense and difficult for new readers. Edited by Michael Lynch, Harold Garfinkel: Studies of Work in the Science (2022) presents writings of Garfinkel and transcribed recordings of his seminars that help much to make sense of Garfinkel’s studies of work in the sciences and other domains.

At the time of the Toronto conference Garfinkel was well-known as the founder of ethnomethodology, the study of the methodical ways in which people make sense. Ethnomethodological investigations are guided by the heuristic to “treat social facts as accomplishments”: “Where others might see ‘things’, ‘givens’ or ‘facts of life’, the ethnomethodologist sees (or attempts to see) process: the process through which the perceivedly stable features of socially organized environments are continually created and sustained” (Pollner, 1974, p. 27) as “artful achievements of communitarian practices” (Macbeth in Garfinkel, 2022, p. 69).

Ethnomethodology’s kinship with Durkheimian sociology should be clear from this formulation, but Garfinkel (1988) gave it a specific, and critically important, twist that is captured by the notion, and task, of “respecification”: describing how established notions (such as “observation” or “measurement”) are in fact achieved through concerted practical actions, so that these notions are merely “glosses” for the processes that constitute them.^[1] “Respecification” is a sort of ethnomethodological “interpretation.”^[2] Whereas Garfinkel and his students and colleagues had focused at first mostly on common sense methods and lay understandings, particularly in institutional contexts in law and psychiatry, in the 1970s and 1980s, Garfinkel (1986) moved to what he termed “ethnomethodological studies of work.” His concern was now to describe the methodical work of specific occupations and to identify “what is unique about particular kind of profession, i.e., to specify what makes driving a truck different from driving a taxi, which again is different from driving a train.” (Greiffenhagen & Sharrock, 2019, p. 255; italics in original).

It was in the course of these studies that Garfinkel turned to the study of scientific work. He might have been motivated to do so by the emerging field of science and technology studies, by the fact that his wife Arlene was a professional chemist, or – as Greiffenhagen & Sharrock (2019) speculate – by his frustration with the rejection of ethnomethodology by many of his colleagues in sociology. What certainly inspired him was his time as a fellow at the Center for Advanced Studies in the Behavioural Sciences at Stanford University (in 1975–76), where he met Harvard University historian of science Gerald Holton. Holton alerted Garfinkel to the “pulsar tape”, archived at the American Institute for Physics then in New York, and became a regular conversationalist on matters of scientific practice and the history of science. At this time, Garfinkel was supervising the PhD theses of Michael Lynch (who had done a laboratory study of neuroscience and was to transcribe the pulsar tape) and Eric Livingston (who studied mathematics).

Harold Garfinkel: Studies of Work in the Science documents Garfinkel’s “turn to science.” Edited by Lynch it consists of two parts. Part I presents an unpublished manuscript, originally intended as a chapter for an edited volume but retracted after an argument with its editors. It has a long title: “Respecifying the Natural Sciences as Discovering Sciences of Practical Action, I & II: Doing So Ethnographically by Administering a Schedule of Contingencies in Discussions with Laboratory Scientists and by Hanging around Their Laboratories.” Following Lynch I shall refer to this manuscript as “Respecifying.” Garfinkel intended to include it in a book with the provisional title Working Out Durkheim’s Aphorism, Book Two: The Lebenswelt Origins of the Sciences.^[3] Part II is the edited transcript of five seminars that Garfinkel gave in Spring and Summer 1980 in preparation for his plenary address in Toronto. These recordings became available through the Garfinkel archive in Newburyport, Massachusetts (USA), and the Garfinkel Archive Project at the Universität Siegen, organized and led by Garfinkel’s literary executioner, Anne Warfield Rawls.

Michael Lynch’s presence in this volume could perhaps be described as polyphonic. As one of the most influential ethnomethodologists of the second generation, Lynch is not only an editor who is deeply knowledgeable about the contents and the persons involved. He was also listed by Garfinkel as a co-author of “Respecifying”,^[4] and he features as a discussant in the recorded seminars. Furthermore, Lynch is a co-author of the “pulsar paper” (Garfinkel et al., 1981). Lynch is wary of editing Garfinkel’s prose to make it easier to understand as doing so entails dangers of interpretation. He opts instead for editing Garfinkel’s text only minimally and lets Garfinkel do the explaining himself in the seminars that make up Part II of the book.

What Lynch can get away without is a nagging challenge for me in this text: speaking for myself in reviewing a manuscript and seminars that are fragmentary and difficult to summarize. I embark on this task as an astronomer-turned-anthropologist who did an ethnographic study of astronomers. For me the pulsar paper remains one of the most intriguing and enigmatic texts on practical action. It has helped me to see my ethnography in a new light (Hoeppe, 2014), but it is relevant far beyond studies of astronomy and other sciences.

My aim in this essay is to identify some highlights of what we can learn from the book for studies of work in science and elsewhere. I shall first walk through the book and identify a series of key topics. Then I take a critical look at some of them and ponder the limits of Garfinkel’s approach.

“Respecifying”

Garfinkel begins “Respecifying” with making two consequential moves. First, he sets out with dismissing the unity of sciences, positing instead a disciplinary, practical and material diversity of science that was properly acknowledged by philosophers and sociologists only later (Dupré, 1993; Knorr-Cetina, 1999). Secondly, he seeks answers to the question: “What is the work of a discovering science?” Doing so sets his concerns apart from those of the Sociology of Scientific Knowledge (Bloor, 1976) and constructionism (Latour & Woolgar, 1979; Knorr-Cetina, 1981) which gained prominence at the time. The latter had already considered laboratory work in a certain detail, but they did so by highlighting contingencies in laboratory work and the “manufactured” nature of scientific facts – often with the ethnographer looking on in ironic bemusement (cf. Francis & Hester, 2004). Garfinkel, by contrast, seeks to understand the “hard sciences” as “discovering sciences of practical action”, and calls for considering them from the inside. As Lynch explains in the introduction, for Garfinkel laboratory work oriented to making scientific discoveries is also about “investigating as well as using practical actions” as these scientists are “reflexively discovering a local organization of practical actions as well as what those practical actions disclose, stumble upon, negate, or prove” (p. 10; italics in original). Each natural science is a “science of practical action” – in contrast to sociology, which Garfinkel calls a “talking science” (p. 156). This distinction is manifested starkly in the ability of laboratory sciences to “lose the phenomenon” (pp. 23, 49, 90, 158).^[5]

Garfinkel seeks empirical access to the sciences of practical action in lieu of doing ethnography. He begins with an innovative approach to talk with laboratory scientists, eliciting their stories by prompting them with “unmotivated observables” (p. 42) – formulations of practical action extracted from earlier conversations, such as “Losing the phenomenon”, “Wasting time”, “Making an experiment work”, or “An issue can get settled” (p. 23) – and asking his conversationalists what it is that he is thereby talking about. He calls the “unmotivated observables” “coat hangers” on which these conversationalists can hang their stories in response. This yields a sort of grammar of the “structures of practical action” (p. 42) in laboratory research.

As per the subtitle of “Respecifying”, “Hanging around Their Laboratories” is another route to the sciences of practical action. In his circle only Lynch (1985) had done so with his “analytical ethnography” (p. 27) of a neuroscience laboratory. Although Lynch participated in a minor part of laboratory work he arguably did not fulfill what Garfinkel termed the “unique adequacy requirement of methods” – the analysts’ competence of performing the practices they study.^[6] By contrast, Garfinkel praises Lynch’s fellow PhD student Eric Livingston for having achieved unique adequacy in mathematics. Livingston, Garfinkel claims, could be taken seriously by mathematicians (p. 28). Livingston had studied for a Master’s degree in mathematics to conduct an ethnomethodological study of proving in mathematics (Livingston, 1986).^[7]

Troubled by his own lack of unique adequacy in science, Garfinkel returns to an extended discussions of the “coat hangers” and their praxiological import. The latter comes about through these formulations’ embeddedness in specific courses of laboratory action and their detailed interactional unfolding. Whereas Garfinkel seeks to specify the conditions for unique adequacy (p. 49), his own study of balls rolling down an inclined place to replicate Galileo’s famous experiment is not included in “Respecifying” (but see chapter 9 of Garfinkel’s 2002 book Ethnomethodology’s Program). A particular, “tendentious” (p. 71), understanding of detail – qualified by Garfinkel as detail* (with the asterisk marking that the common meaning of the word is only a placeholder for what remains to be specified in context) – becomes salient for the description of practical action. It is not Garfinkel who conveys it most clearly in “Respecifying”, but Douglas Macbeth, who contributes a fascinating description not of science but of playing pick-up basketball gracefully (Appendix 2). Garfinkel endorses Macbeth’s description (in Appendix 3) and concludes “Respecifying” with a summary of its programs and policies (in Appendix 4).

The 1980 Seminars

The 1980 seminars precede the version of “Respecifying” presented in part I by eight years. Tape-recorded by Garfinkel, who also arranged for their transcription, they are easier to make sense of than his dense writing. Stylistically less arcane, the seminars provide useful background on the pulsar tape and its interpretation, as well as to his understanding of studying practical action.

Garfinkel highlights topics from Gestalt psychology and the 1969 postscript of Thomas Kuhn’s Structure of Scientific Revolutions (1970). Gestalt switches feature prominently in Kuhn’s account of the incommensurability of paradigms before and after a scientific revolution. But as suggested by the explicit reference to the “Lebenswelt Origins of the Sciences” in the subtitle of “Respecifying”, Garfinkel is also influenced by Husserlian phenomenology. Edmund Husserl’s former student Aron Gurwitsch (1901-1973), who had considered Gestalt themes phenomenologically, had a major influence on Garfinkel. Whereas Gestalt switches are often considered as perceptual phenomena, Garfinkel arguably follows Kuhn under the influence of Gurwitsch and describes the astronomers’ discovery of optical signals from the pulsar as “the animal in the foliage” – an “achievement of practical action” (pp. 120, 132). As Garfinkel et al. (1981, p. 132) note, “[t]he 'foliage' is the local historicity of their embodied shop practices. The 'animal' is that local historicity done, recognized, and understood as a competent methodic procedure.” The “transposition” (Lynch & Eisenmann, 2022) from perception to practical action is highly instructive for making sense of Garfinkel’s later work more generally.

However, Kuhn’s influence is profound as well, demonstrated by Garfinkel et al. (1981, p. 142, fn. 28) describing how the “pulsar is attached to nature”, with reference to Kuhn’s 1969 postscript. Arguably there is also a kinship to Latour & Woolgar’s (1979, p. 177) constructionist notion of “splitting and inversion.” Yet where Latour & Woolgar describe scientific practice as a work of ordering, Garfinkel goes further and deeper: he regards the “production” of the “finding” of the pulsar “as nothing less than the production of social order” (p. 123) – echoing an insight articulated already in the preface of Studies in Ethnomethodology (Garfinkel, 1967, vii). Epistemic and social orders are intertwined in scientific work.

Latour & Woolgar’s Laboratory Life (1979) is, famously, concerned with inscriptions and writing. Garfinkel addresses writing only late in the seminars, when – again inspired by Gurwitsch ­– he remarks how it is through producing a written account that these astronomers “render” their discovery as an example of “Galilean science” (p. 146). It is a hallmark of Galilean science, Garfinkel claims, that “the work is always accompanied by the research reports, the research papers, the published accounts, and the publishable talk of that work – that is, the publicly available and the publicly responsible talk of that work” (italics in original), ascribed to the voice of a “transcendental analyst” (p. 159). This voice hides what is, in the work of scientists, always and unavoidably a “Lebenswelt pair” (p. 28; italics in original) of lived, practical action on the one hand and its textual rendering on the other.

How is the study of the “sciences of practical action” a study of the natural sciences?

After reading the book I agree with Lynch that Garfinkel provides “a distinctive perspective on the natural and social sciences that remains highly original and pertinent to research on science, social science, and everyday life today” (p. 1). Both parts of the book could inspire a wide range of empirical studies of practical action, of the sciences and other occupations, including Garfinkel’s (and Macbeth’s) notes on coat hangers, detail*, Gestalt switches and the praxeological import of Kuhn’s Structure. Garfinkel’s “distinctive perspective” notwithstanding, readers will formulate their own list of questions about the book. The questions that occupied me most are these: How is the study of the “sciences of practical action” a study of the sciences, and of scientific discovery? What does it include, and what does it leave out?

I second the bewilderment with Garfinkel et al.’s (1981) pulsar paper, first expressed by Collins (1983) and recently echoed by Knorr-Cetina (2019), how one can possibly reduce a discovery to certain audio-recorded exchanges without thereby missing essential context. Minimally, the distinction between a discovery (as a historically contextualized achievement) and discovery work (as a course of action) remains in place. But even if one acknowledges this and attends to discovery work as a course of action, I remain puzzled by Garfinkel’s assessment in the seminar that “[w]hat we’re saying is that the optical pulsar is the achievement of a practice, that (includes) itself (as) a practice” (p. 118).

Garfinkel jokes that “this sounds like an operational definition with pepper and salt on its tail” (p. 118). This may be so, but I am rather puzzled by what one may call Garfinkel’s operational use of language.^[8] On the face of it, Garfinkel’s claim that a physical object (the pulsar) could possibly be the result of the practices of its discoverers will be absurd for scientists (and scientific realists). Is this constructionism gone wild? Saying so misses the point inasmuch as any perception or (scientific) observation is procedural. Even “seeing something” is not the result of a simple “looking” (Coulter & Parsons, 1991). Garfinkel’s use of the words “optical pulsar” must be read alongside with his notion of the “independent Galilean pulsar” that is posited by textual practices.

This is perhaps why Garfinkel does not describe the “optical pulsar” as a physical object, but as the “achievement of a practice”. It is through his use of operational terms that Garfinkel makes the record of these astronomers’ work (as a process) amenable to ethnomethodological study. But in doing so he distorts the words of the scientists whose practices he ostensibly seeks to describe from within. “Optical pulsar” is not a members’ term. Starting with the title of their discovery article (Cocke et al., 1969) these astronomers never write of an “optical pulsar” but keep writing of “optical signals” from the (already discovered) pulsar. Their work is intended to refine knowledge of an object that had been discovered at other wavelengths (the pulsar had been previously discovered at radio wavelengths). The optical observations were not intended to discover a new object, but to refine and extend observations of one that had been discovered previously – through quite independent courses of practical action. There are multiple practices, employing different technologies. Arguably it is through them that the pulsar gets “attached to nature” and is secured as a discovery. Garfinkel’s account in the seminars (as well as Garfinkel et al., 1981) misrepresents what these scientists did and what they understood to have done. It ignores what philosophers call scientific robustness (cf. Hoeppe, 2020). It is through multiple kinds of observation – and by presumably detaching an epistemic claim from a singular course of action – that discoveries are commonly secured in the sciences (Hoeppe, 2019), a point that is also ignored by constructionist accounts like Latour & Woolgar (1979). This ignorance is clearly recognized by some natural scientists familiar with constructionist accounts.^[9]

Garfinkel’s grammatical collection of the “structures of practical action” that includes “Losing the phenomenon”, “Making an experiment work”, and “Finding the animal in the foliage” certainly captures something that is unavoidably part of skilled practical action in laboratory experimentation (and in data analyses; see Hoeppe, 2014). But much of this would be just as relevant to, say, a craft workshop. So is Garfinkel’s collection specifically about the natural sciences? Is that even his concern?

When Bruno Latour (1986, p. 1) probed into “what is specific to our modern scientific culture” and sought “the most economical explanation”, he found it in documentary practices: uses of “immutable mobiles” – inscriptions that are flat, mobile, combinable and superimposable. As we have seen, Garfinkel paid much attention to the “missing what” – a Lebenswelt pair’s element of lived, practical action. But if Latour is right, Garfinkel misses much of what is essential about science – something that is likewise not captured by ethnomethodological accounts of the natural sciences like Garfinkel’s (2002) experiments with Galileo’s inclined plane. Conversely, of course, it is Latour who misses the embodied work that Garfinkel describes.

Garfinkel acknowledges that “the report, that is the announcement of the discovered pulsar, is present from the beginning as a relevant task in presence to what’s going on in the lab” (p. 188). This, he argues, is characteristic of “Galilean science.” The ethnomethodological pulsar paper (Garfinkel et al., 1981) includes the astronomical pulsar paper (Cocke et al., 1969) as an appendix. However, with the pulsar tape being a resource for the scientists’ writing, Garfinkel does not address this production of Galilean science. The work of Dorothy Smith can be a guide to its workings, as she points out that there is a “movement between the locally historic and the textually mediated discourse” in science and beyond (1984, p. 65; see also Hoeppe, 2014, 2020). Her attention to the “locally historic” is inspired by the pulsar paper. It is through such movements – which can themselves be respecified, or “interpreted” ethnomethodologically – that epistemic claims are detached from singular courses of action.

Outlook

The continuing relevance of Garfinkel’s studies is predicated by the nature of social life. That practical actions are ubiquitous in (and make up) all walks of social life, including the natural sciences, cannot be wished away. For this reason alone it is regrettable that, following their “first wave” in the 1970s and 1980s, ethnographic studies of laboratory work, including ethnomethodological ones, remain rare. Harold Garfinkel: Studies of Work in the Science offers valuable resources that might inspire new work, on the sciences and any other domain of practical action. Yet it can also help recognizing and formulating the limits of the later Garfinkel’s program. Greiffenhagen & Sharrock (2019) have recently drawn attention to the limits of Garfinkel’s notion of “hybrid studies”, something also touched upon in this book. The challenge of studying scientific practice by adhering to Garfinkel’s programmatic restriction of focusing on embodied Lebenswelt phenomena, but doing so without missing their essential context (in something like Smith’s “textually mediated discourse”), is perhaps another.