Michael Lynch's Remembrance of Harold Garfinkel

Social Studies of Science
Harold Garfinkel (29 October 1917 21 April 2011): A remembrance and reminder

Michael Lynch
Social Studies of Science 2011 41: 927 originally published online 29 September 2011 DOI: 10.1177/0306312711423434
The online version of this article can be found at:
Published by:
Additional services and information for Social Studies of Science can be found at: Email Alerts: http://sss.sagepub.com/cgi/alerts
Subscriptions: http://sss.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://sss.sagepub.com/content/41/6/927.refs.html
>> Version of Record - Nov 16, 2011 Proof - Sep 29, 2011
What is This?
Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
SSSXXX10.1177/0306312711423434LynchSocial Studies of Science

Harold Garfinkel (29 October 1917 – 21 April 2011): 

A remembrance and reminder

Michael Lynch

Social Studies of Science 41(6) 927–942 © The Author(s) 2011 Reprints and permission: sagepub. co.uk/journalsPermissions.nav DOI: 10.1177/0306312711423434 sss.sagepub.com
page2image11792 page2image12552
Department of Science & Technology Studies, Cornell University, Ithaca, NY, USA


This essay is a remembrance and also a reminder of Harold Garfinkel’s contributions to science studies. Garfinkel is best known as the founder of ethnomethodology, the sociological investigation of the production and coordination of ‘methods’ in non-scientific as well as scientific settings. In addition to studying the tacit organization of everyday activities, Garfinkel and his students also investigated practices in the natural and social sciences that elude formal methodological prescriptions and reports. Garfinkel’s work sometimes is acknowledged as a precursor to early ethnographies of scientific laboratories, but this essay argues that his conceptual and methodological innovations continue to have a pervasive, though often unacknowledged, place in science and technology studies and related fields.


ethnomethodology, Harold Garfinkel, science and technology studies, scientific methods, social action
Having just come out of a jungle, I can’t promise you that in leading you in to show you what I’ve found that I won’t lose the way for all of us. (Garfinkel, 2008: 101)

Harold Garfinkel died at his home in Los Angeles earlier this year, at the age of 93. Garfinkel was Professor Emeritus in Sociology at the University of California, Los Angeles (UCLA), where he had worked since 1954. He retired in 1987, but continued to teach occasional seminars for several years afterwards and maintained his scholarly activity until the very end. The publication of his writings, and of transcripts of his lec- tures and dialogues, is likely to continue for many years to come. Garfinkel was best

Corresponding author:

Michael Lynch, Department of Science & Technology Studies, Cornell University, 302 Rockefeller Hall, Ithaca, NY 14853, USA.
Email: MEL27@cornell.edu

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
928 Social Studies of Science 41(6)

known for having founded ethnomethodology, a distinctive approach to the social organization of practical actions. Although the main body of his work focused on a broad range of social phenomena, my concern in this essay is with his direct and indirect con- tributions to science and technology studies (STS). His contributions are far more exten- sive than is often recognized. Judging from submissions to this journal and the programs at recent Society for Social Studies of Science (4S) meetings, ethnomethodology is not much in evidence. This is not to say that it has vanished. This journal continues to publish occasional articles that exemplify, or at least draw upon, ethnomethodology and conversation analysis (a research program that developed from and has a continuing rela- tion to ethnomethodology). Ethnomethodology also is acknowledged, though less often than a few decades ago, for having made significant contributions to STS, largely in connection with laboratory studies and related ethnographic, discourse analytic, and reflexive approaches to research practices. Less often recognized is the extent to which conceptual and methodological innovations once associated with ethnomethodology have been folded seamlessly and, for the most part, invisibly into STS discourse and practice. Consequently, this essay is written both as a remembrance and a reminder of the significance of Garfinkel’s legacy for continuing work in science studies.

From accounting to accountability

Garfinkel grew up in New Jersey. His father ran a small business that sold furniture and housewares, and Harold began his undergraduate studies with a plan (if not a strong desire) to step into his father’s line of business. He attended Newark College – an ad hoc institution that served the swelling immigrant population in New Jersey and eventually evolved into Rutgers University. There, he had the good fortune to attend innovative courses taught by instructors who were students or junior faculty at Columbia University and other nearby institutions. Several of the students and instructors later went on to become notable social scientists. In an editor’s introduction to a compilation of Garfinkel’s writings, Anne Rawls (2002) recounts how Garfinkel enrolled in a course called Theory of Accounts, which covered the techniques of bookkeeping and account- ing. From that course he came ‘to understand that even in setting up an accounting sheet, he was theorizing the various categories into which the numbers would be placed’. Moreover, he realized, the constructed order of entries ‘was accountable to superiors and other agencies in a variety of complex ways’ (Rawls, 2002: 10). This account of Garfinkel’s experience with accounting was, no doubt, aided by retrospection in light of the concept of ‘accountability’ he developed much later. As a central theme in ethno- methodology, ‘accountability’ applies to accounts of all kinds – immediate communica- tive expressions as well as post facto bureaucratic records – and refers to how we present and organize our actions while anticipating their reception by unknown, as well as specifiable, ‘auditors’.

Inspired by his undergraduate experience, Garfinkel headed on a track that led him far from the family business. Prior to the Second World War, he successfully gained entry into one of the top programs in Sociology at the University of North Carolina, Chapel-Hill, where he received his Master’s degree in 1942. Early publications based on his thesis research on race relations in the American South demonstrated an acute sociological

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
Lynch 929

insight and, contrary to Garfinkel’s later reputation, an ability to write fluently in plain English. His first publication, a quasi-fictional account of a conflict that arose when an African-American woman refused to sit at the back of a bus when the vehicle crossed the (former) Mason-Dixon line en route from New Jersey to North Carolina, was included in a collection of best short stories of 1941 – more than a decade before the historic chal- lenges to the Jim Crow laws by Rosa Parks and the Freedom Riders (Garfinkel, 1941).

Garfinkel’s work took a more theoretical turn after he served in the war and began his doctoral studies with Talcott Parsons in the Social Relations Department at Harvard. Parsons and his students openly aspired to reinvent sociology by articulating a compre- hensive theory of social structure and social action. Although this is rarely acknowl- edged, Parsons’ conception of a science of society unified around a central theory had broad affinities with Thomas S. Kuhn’s (1970) notion of scientific paradigms in the natural sciences (see, especially, ‘Note on the concept of fact’, in Parsons, 1949 [1937]: 41–42). Though separated by a generation, both Parsons and Kuhn cultivated their understandings of science within the same intellectual and administrative milieu at Harvard. But, where Kuhn chronicled the emergence of paradigms in the physical sci- ences, Parsons set out deliberately to build a dominant theoretical framework for soci- ology. The venture was highly successful, as sociologists schooled in Parsons’ structural–functionalist framework staffed major departments and dominated North American Sociology (though without as much influence in Britain) until the late 1960s. In STS, this era is most often associated with Parsons’ former student Robert K. Merton, who allied himself with Paul Lazarsfeld to link functionalist theory with multi-variate analysis.

Although Garfinkel remained devoted to Parsons’ ideas until well into the 1960s, he was prominent among those whose actions and writings helped break the hold of structural–functionalism on American sociology. During his graduate studies, Garfinkel took up what became a life-long interest in phenomenology and existentialism, and in his dissertation he developed a theory in which routine actions that presume and antici- pate a normal, life-as-usual order of things, reflexively produce the resilient features of the social world that sociologists attribute to social structure. In an unpublished paper derived from the dissertation, Garfinkel (1960) contrasted Parsons’ ‘decisions’ on four ‘pre-theoretical problems’ with those of social phenomenologist Alfred Schutz, and in doing so, he articulated contrasting ontological and epistemological implications of Continental philosophy for contemporary social science decades before it became fashionable for social scientists to do so. In his dissertation, Garfinkel conceived of social structures as being realized in action–performatively–rather than being a tran- scendent set of forces impinging upon particular actions as though from outside. He attempted to integrate phenomenological insight with Parsons’ theory of social action, but after receiving his PhD and beginning his long career in sociology at UCLA, he eventually broke with Parsons’ theoretical framework and rejected the very idea of developing a theory that could encompass the varied and uncanny practices that consti- tute a society. He did not deny that our actions are always and already embedded in a mas- sively ordered society, but sought instead to probe the presumptive existence of social order with a series of idiosyncratic investigations.

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
930 Social Studies of Science 41(6)

A ‘company of bastards’

In the 1950s, Garfinkel coined the word ‘ethnomethodology’, a compound of ethno (people) and methodology (the systematic performance of scientific and non-scientific practices), and over the next decade he was able to recruit a group of highly talented students and colleagues (including Harvey Sacks, David Sudnow, Egon Bittner, Don Zimmerman, D. Lawrence Weider, and Aaron Cicourel), who set out to develop the research program adumbrated by that name. Ethnomethodology became both a name for the research program and for the substantive ‘things’ that ethnomethodologists studied. It was nominally inspired by the ‘ethnosciences’ (for example, ethnobotany, the anthro- pological study of native plant classifications and uses) and the ‘ethnography of lan- guage’ (the observational and analytical study of language use in specific cultural situations). However, the scope of ethnomethodology’s observational and analytical approach was exceptionally broad, as it included the circumstantial use of methods of all kinds. Language use, such as in conversational exchanges at a dinner table or delibera- tions among members of a jury, remained a crucially important research topic, but the methods studied also included embodied and technologically mediated practices, such as navigating through traffic situations and recognizing what is happening in specific social and technical environments. An important aspect of the methods and practices studied by ethnomethodologists is that they reflexively produce the very ‘social envi- ronments’ in which participants find themselves. Far from imagining, as is sometimes the case in cognitivist studies, that scientific methods (or, rather, idealized versions of such methods) provide analytical models or organizational templates for everyday activ- ities, Garfinkel proposed to come to terms with ‘native methodologies’ on their own ground. He and his students did not ignore the scientific or other practices in which methods and formal protocols are explicitly formulated, enforced, and criticized. Instead, they examined the day-to-day production of methods, just as they did with prac- tices in other settings. In addition to indicating that even the most banal activities were methodic productions, Garfinkel also wrote extensively about the practical accomplish- ment of professional work, such as coroners’ inquiries, jury deliberations, clinical record keeping, and social science research itself. Under his treatment, methods became ‘top- ics’ rather than (or in addition to) ‘resources’, and Garfinkel counseled ‘indifference’ to the epistemic privilege often assigned to scientific and other formal theories and meth- ods. The research ‘policy’ (as Garfinkel called it) of ethnomethodological indifference placed all methods and accounts of method on the same plane, ‘wherever and by whomever they are done’. When describing such methods, the ethnomethodologist was to refrain ‘from all judgments of their adequacy, value, importance, necessity, practical- ity, success, or consequentiality’ (Garfinkel and Sacks, 1986 [1970]: 163). Especially when applied to sociological methods and theories, such indifference provoked baffle- ment and fury from defenders of the discipline’s scientific aspirations.
The contentiousness surrounding ethnomethodology was at its strongest in the late 1960s and early 1970s, when some key publications (especially Garfinkel’s (1984 [1967]) seminal Studies in Ethnomethodology) helped to establish the field, and it became a subject of debate between its own proponents and those of established soci- ological approaches (Hill and Crittenden, 1968). At the time, many of Garfinkel’s

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
Lynch 931

students and followers presented themselves as radicals of an intellectual, if not political, variety. They courted controversy and got more than they bargained for, as they were denounced in review articles and plenary addresses by leading sociologists and anthro- pologists, such as James Coleman, Lewis Coser, and Ernest Gellner. Some of the denun- ciators likened ethnomethodologists to a cult of self-styled revolutionaries who hid the emptiness of their ideas behind a wall of jargon, and they accused them of overtly or covertly holding extreme epistemological positions ranging from positivism and behav- iorism to idealism and solipsism. Though Garfinkel’s supporters sometimes responded to such denunciations with bemused and defensive replies, they also took them as evidence that ethnomethodology had provoked the sociological ‘establishment’ to lash out in clumsy indignation at a partially understood challenge to intellectual orthodoxy. Years later, when it became common to refer to Garfinkel as the ‘founding father’ of ethnometh- odology, he would occasionally joke that the field was ‘a company of bastards’ (Garfinkel, 2007: 13), alluding to the reputation that his ‘company’ had acquired as a contentious lot, both in their relations to what they frequently called ‘conventional’ sociology and in their dealings with one another.

Many of Garfinkel critics complained of their difficulties with understanding what he was saying. His most prominent work – Studies in Ethnomethodology (Garfinkel, 1984 [1967]) – is not a book I would recommend for readers interested in gaining a casual acquaintance with ethnomethodology, but for those willing to struggle with it, it is one of those rare academic books that can yield remarkably different insights with each re- reading. In person, and in his writings, Garfinkel resisted his interlocutors’ efforts to integrate his ideas with familiar sociological concepts and established ‘schools’, or to account for ethnomethodology in terms of set-piece distinctions between ‘macro’ and ‘micro’ levels, ‘structure’ and ‘action’ theories, ‘subjective’ and ‘objective’ orientations, and ‘quantitative’ and ‘qualitative’ methods. Taken uncharitably, such resistance could be dismissed as a strategy for exaggerating novelty, but it is worth keeping in mind that in the 1960s, existential phenomenology and Wittgenstein’s later writings (which Garfinkel also drew upon) were unknown to all but a very few North American social scientists. Many of Garfinkel’s readers and interlocutors had no inkling of where he was coming from. He preferred to address them as ‘novices’, which did not endear him to the leading sociologists he included in that category. Garfinkel also was very tough on his students, ever vigilant to detect (what he construed as) careerist efforts to appropriate his teachings in order to repackage them as more palatable academic commodities.


Contrary to prevailing efforts to develop abstract theories that derived individual actions from larger social structures, Garfinkel delved more deeply into the minutia of social life. By doing so, he did not aim to reduce actions to psychological or neurological causes; instead, he attempted to pursue communicative actions all the way down to their constituent details. The details never ceased being actions: the utterances that made up a conversational exchange or the embodied actions that composed queues and traffic jams retained their logical status as socially organized practices rather than individual behaviors that traced back to motives, unconscious habits, or brains.

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
932 Social Studies of Science 41(6)

Starting with his dissertation research and for many years afterwards, Garfinkel developed a distinctive style of investigation in which ‘trouble’ was a central method- ological theme. Trouble could be deliberately induced to ‘explode’ orderly scenes or observed in studies of disrupted events and difficult lives, such as the struggles of an intersexed person ‘Agnes’ to convince the UCLA clinic that she had always been a ‘natural’ woman in need of corrective surgery (Garfinkel, 1984 [1967]: Ch. 5). Trouble was Garfinkel’s Archimedean lever for revealing tacit, taken-for-granted practices that perform what one simply is (a ‘normal’ man or woman, a spouse, a stranger or close intimate, a competent nurse). For Garfinkel, ‘social facts’ were ‘accomplishments’, though they never ceased being ‘things’. He borrowed the idea of ‘perspective by incongruity’ from literary theorist Kenneth Burke, but his methodological use of trouble was original, and later proved influential for STS efforts to use practical difficulty and discord as leverage for revealing taken-for-granted practices that perform what a ‘fact’ or ‘thing’ is. He became famous for his exercises – experiments, of sorts, designed as ‘aids to a sluggish imagination’ – often performed by (or on) his students. Some of them were akin to practical jokes, as they aimed to disrupt taken for granted household routines, conversational protocols, and the delicate balance between disclosure and concealment in ordinary interaction. One particularly notable exercise involved a series of simulated exchanges between a hidden ‘counselor’ and a series of unwitting students who responded to an advertisement for free counseling sessions. The students were instructed to present a series of ‘yes–no’ questions to an unseen ‘counselor’, who would respond in oracular fashion with a simple ‘yes’ or ‘no’ to each question (Garfinkel, 1984 [1967]: 79–94). Unbeknown to the students, the counselor’s responses were generated randomly, and yet the students – though often showing consternation – were able to make coherent sense of the series of questions and answers. As Lucy Suchman (1987: 65) observes, the flexible logic that allowed the students to find coherence in the counselor’s random replies was akin to what Joseph Weizenbaum (1976: 26) described in his reflections on the surprising success of his early natural language program Eliza for eliciting sustained engagement in ‘counseling’ sessions, even from interlocutors who knew fully well that they were confronted with computer-generated text and not a psychotherapist.

Trouble of a kind that became familiar – and methodologically useful – for STS also had a revealing role for elucidating one of Garfinkel’s best-known themes: ‘indexical- ity’. This theme also had a place in artificial intelligence research, as well as in the longer history of logicians’ efforts to regiment natural language for constructive– analytical purposes. In an analysis of early machine translation programs, Bar-Hillel (1954) identified ‘indexical expressions’ as a source of trouble. Pronouns and other deictic expressions are prototypical indexicals, as particular words such as ‘me’, ‘over there’, ‘later’, or ‘home’ index what speakers presume their auditors can discern from the immediate context of use. The early machine translation programs could only work with dictionary equivalents, without recourse to worldly contexts of use, and so it was necessary to use competent human assistance to repair both the input and output texts to make the former intelligible to the machine and the latter intelligible to a natural language user. As Garfinkel observed, particular words classified as indexicals are but the tip of the iceberg, as virtually all natural language-use relies upon context for intel- ligibility (and ‘context’ itself is notoriously open-ended). But, far from concluding that

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
Lynch 933

natural language was a disorderly mess, he and his former-student/colleague Harvey Sacks delved into the peculiar way in which indexicals and other ‘glossing practices’ were central to the production of concerted actions; and, further, that their use was irreparable and without need for repair in everyday usage.

A simple example of an interactional device for making sense with indexical expres- sions is an interactional practice that Garfinkel dubbed ‘Rose’s gloss’ (he attributed it to Edward Rose, who later disavowed any credit for it). Garfinkel and Sacks (1986 [1970]: 183–184) describe it as a playful trick that Rose perpetrates on an unsuspecting host who picks him up at an airport during a visit to a city he had never seen before. While riding as a passenger in the host’s vehicle, Rose looks out the window and turns his head, appar- ently tracking something in the passing scene. He then remarks, ‘It certainly has changed’, eliciting from his host a response such as ‘It was ten years before they rebuilt the block after the fire’ (p. 184). (Incidentally, I have noticed during conversations with an elderly relative living through a long course of senile dementia that she recurrently uses variants of ‘Rose’s gloss’ to initiate conversational sequences.) The point of the story is not that Rose confabulates an account (or, rather, that he elicits an unwitting confabulation from his host), but that, as in the case of the phony counselor exercise, the host is able to find and build a usable coherence by working with the scenic and temporal circumstances of the ‘gloss’. Exposing such a practice in ‘thin air’ suggests the possibility that it thrives, if less visibly, in ‘thick air’ as well.

Studies of work

A line of research on the professions and sciences that Garfinkel began in the 1970s, and which he and some of his students sustained in subsequent decades, partly paralleled devel- opments in social studies of science at that time. Like early laboratory studies (Latour and Woolgar, 1979; Knorr, 1981), this research was ethnographic and focused on discourse and practical actions in the workplace. The various studies Garfinkel and his students conducted encompassed more than laboratory or scientific work. They construed ‘work’ very broadly to encompass practices and practical reasoning accomplished both within and beyond places of work (see Garfinkel, 1986). In parallel with Bloor’s (1976) criticism of the sociology of knowledge for excluding the technical contents of science and mathematics, in lectures and seminars in the early 1970s, Garfinkel identified a more general ‘gap’ in sociol- ogy’s treatment of occupational and other practices, a gap to which he gave the name ‘the missing what’. By this he meant that sociological studies of work, religious activities, and many other specialized practices elaborated upon the background ‘social’ characteristics of participants–their places in class systems and divisions of labor, their conflicts and values, their attitudes, and so forth–but they did not deem it relevant to address just what practitio- ners did together when accomplishing particular occupational or organized activities. A whole series of topics appeared to be open to investigation: how musicians played music together; how dancers coordinated their gestures; how surgical teams performed opera- tions; how mathematicians demonstrated proofs to one another; and, of course, how labora- tory scientists conducted experiments, recorded their findings, and analyzed results.

Garfinkel did not accuse sociologists of disregarding ‘the missing what’ out of lazi- ness or ignorance. He recognized that they were rarely in a good position to explicate the
Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
934 Social Studies of Science 41(6)
technical activities in other fields, which often have their own methodologies and pedagogies. However, as his earlier studies of sociological research practices revealed, formal methodologies produce their own ‘missing whats’: ‘gaps’ between explicitly prescribed action and the actual moment-to-moment production of constituent research tasks such as administering interviews, coding questionnaire and interview responses, and writing-up results. Such gaps can be sources of problems (for example, when a novice tries to master a practice), but this does not imply that competent practitioners are unconscious of how they do their work, or even that they do not ‘know’ what they are doing until ethnographers reveal it to them. As noted earlier, troubles with formal instruc- tions, plans, directions, maps, and so forth are revealing, and Garfinkel (2002: Ch. 6) explicitly designed a program for explicating troubles observed in the course of ‘instructed actions’. In line with that program, Suchman’s (1987) Plans and Situated Actions concisely presented and deftly elucidated how inexperienced users of complex ‘intelligent’ technologies routinely encountered trouble when attempting to follow instruc- tions on how to use them, and her research provided a major initiative for developing computer-supported cooperative work (CSCW), a research program of interest to compa- nies as well as academic researchers in the communication and information sciences.

Garfinkel’s program seemed to require a student to master the practices in other fields in order to do ethnomethodological research on them. Indeed, he gave a name for this requirement: ‘the unique adequacy requirement of methods’ (Garfinkel, 2002: 175–176). Simply understood, this meant that a student needed to be able to perform the practice; that is, to acquire competency akin to what Collins and Evans (2002) much later called ‘contributory expertise’, as a condition for explicating the performative organization of the constituent practices. He proposed further that successful ethnomethodological stud- ies would produce ‘hybrids’ with the disciplines studied. Needless to say, this was a demanding requirement, and there were some casualties among his students who got caught between their studies and the practices studied (others of us evaded the require- ment and settled for a verbal or conceptual understanding in the interest of getting on with our lives).

Some of the studies of work that Garfinkel and his students undertook were on mathematics and the sciences (Garfinkel et al., 1981; Livingston, 1986; Lynch, 1985). These studies paralleled and partly converged with the early ethnographies of science (Collins and Pinch, 1982; Knorr, 1981; Latour and Woolgar, 1979; Traweek, 1988; it should be noted that, despite the staggered publication dates, the research in each of these cases began in the 1970s). There were at least two key differences, however.

One distinctive feature, which was most evident in Livingston’s (1986, 1999) studies of the ‘lived work’ of mathematical practice, was that ethnomethodological studies of work were not limited to observations of other peoples’ workplace activities. Ethnomethodologists also aimed to understand practices as they were revealed in talks and publications. Livingston’s demonstrations of mathematicians’ work take as their sub- ject the materials for various proofs. In his texts, he works through the proofs step-by-step while specifically calling the reader’s attention to the practices necessary to ‘follow’ along. These include idiosyncratic ways of drawing connections and using notations, as well as standard ways of presenting figures and equations. To the extent that it ‘works’ with a particular reader, Livingston’s (1986) text becomes a place in which practical

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011

Lynch 935
mathematics is performed and analyzed. In this case, the unfolding proof on a page of text is the workplace (also see Bjelic and Lynch, 1994).

A second distinctive feature is signaled by Garfinkel’s dictum that each science is a distinctive ‘science of practical action’ (Garfinkel et al., 1981: 142, n.). In contrast to the analogy of scientists with ‘natives’ studied by anthropologists (of the classic variety) who purport to be able to see and describe what the ‘natives’ take for granted, Garfinkel was proposing that the ‘natives’ were masters of their own methodologies. The task for his students, though, was not to recite formal tenets of scientific methodology (there would be no point in doing so), but to examine the relevant methodic actions; actions that speak more directly than what textbooks and philosophers say abstractly about methodology. The idea was that such actions would show that, and how, doing science not only would produce natural order, but that at the same time it would also produce and analyze relevant features of ‘social’ order. This contrasted with a general constructionist position that everything in a laboratory, without exception, is an artifact, and that what scientists describe as natural facts are manufactured products derived from an artificial environment (Knorr-Cetina, 1983: 118ff.). Instead, the ethnomethodological idea was that scientists produce a distinctive ‘archeology’ when they ‘discover’ research artifacts and attribute them to particular origins within a complex of instruments, disciplined (or undisciplined) practices, and ontological presumptions (Lynch, 1985). Often, this ‘archaeology’ is negative: by analogy, the scientists discard the potshards and preserve the dirt for analysis. The archeology in question is nothing other than the natural science at work, as natural phenomena are related to artifacts in the manner of a gestalt figure and background. The relation is co-constitutive, reflexive, and not necessarily a ‘regress’ of a vicious kind.
Starting in the late 1970s, Garfinkel had some contact with Bloor, Latour, Woolgar, Collins, and other rising ‘stars’ in science studies, but for the most part he remained aloof from science studies (this did not hold for all of his students, however). He also never embraced the word ‘construction’ or any of the variants of constructionism that became emblematic of social studies of science. In part, this was because he reserved the term ‘constructive analysis’ for reference to the sociological practice of constructing models and indicators that stood proxy for the concrete ‘society’ that was always beyond methodological reach. He preferred the term ‘production’ to ‘construction’, per- haps because it implicated action and performance, and was less easily confused with a common idiom in the sciences to discredit findings as (often unwitting) products of artifice.

The idea that methods should be treated as topics rather than resources, and the explicit ‘policy’ of ‘indifference’ towards the special epistemic status often attributed to professional scientific methods resonated with, and perhaps informed, David Bloor’s (1976) more familiar postulates of ‘symmetry’, ‘impartiality’, and ‘reflexivity’. The parallels are especially obvious between ‘ethnomethological indifference’ and Bloor’s ‘impartiality’: the idea that the sociology of scientific knowledge (SSK) ‘would be impartial with respect to truth or falsity, rationality or irrationality, [and] success or fail- ure’ of the systems of knowledge studied, and that ‘[b]oth sides of these dichotomies will require explanation’ (Bloor, 1976: 7). There was a key difference, however. Unlike Bloor, Garfinkel never proposed to give causal explanations of the practices studied, let

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
936 Social Studies of Science 41(6)

alone causal sociological explanations. Instead, he included the social science disciplines and their explanatory efforts among the practices that ethnomethodologists studied. Ethnomethodological indifference was directed, first and foremost, toward the methods and explanatory apparatus of sociology. Lay and professional social explanations, social concepts, survey methods, interview techniques, and so on, were topics, not resources, for ethnomethodology. Although professional sociologists are committed to particular concepts (race, class, gender, norms, values, motives, and so on), and often are highly skilled with using particular analytic methods, variants of these concepts and methods are ubiquitous and, more importantly, constitutive, in the organization and administra- tion of contemporary life.
For their part, Bloor, Collins, and other proponents of SSK could never accept Garfinkel’s indifference to sociology’s epistemic status, as it threatened to undermine their reliance upon sociological concepts and methods as explanatory resources – reliable means through which to discern and demonstrate the social conditioning of scientific knowledge. In contrast, Bruno Latour and Michel Callon explicitly took on board ethnomethodology’s displacement of sociology into the world it studied when they began to formulate actor-network theory (they did not, of course, adopt ethno- methodology wholesale). (For a recent acknowledgment of the debt to Garfinkel, see Latour (2005: 54); for ethnomethodological critiques of constructionism, see Coulter (1989), Sharrock and Anderson (1991), Lynch (1993), and Macbeth (1996).)

The legacy

One not insignificant aspect of Garfinkel’s legacy is that his word is now officially embedded in English language, though dictionary entries on ‘ethnomethodology’ are not very helpful because they tend to define it as a ‘method’. Ethnomethodology is not a ‘method’ as such, as it includes an open variety of research practices that aim to come to terms with the production of ‘methods’ of all kinds. The Wikipedia (n.d.) entry is slightly better: ‘the study of the everyday methods people employ for the production of social order’. However, that version also is too limited, especially for readers of this journal, because a significant amount of research in ethnomethodology examines novel and controversial practices, as well as methods that are ‘everyday’ only for people who are at home with highly specialized activities. The fact that ethnomethodology remains poorly defined and not well understood after so many years may indicate that the project is unfinished.

Garfinkel outlived several of his students and colleagues, each of whom made dis- tinctive contributions to ethnomethodology and conversation analysis: Melvin Pollner (1940–2007); David Sudnow (1937–2007); D. Lawrence Wieder (1938–2007); and Gail Jefferson (1938–2008). Another, Egon Bittner (1921–2011), outlived Garfinkel by 2 weeks. Their deaths are a reminder that ethnomethodology itself has aged consider- ably in past half-century. Pollner, Sudnow, Jefferson, Wieder and Bittner began their academic careers in the 1960s, at a time when Harold Garfinkel and Harvey Sacks were developing ethnomethodology and conversation analysis. Sacks died in a traffic accident in 1975, at the age of 40, leaving behind an impressive collection of writings, data archives, and recorded lectures. The lectures were published in a two-volume set by Blackwell in

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
Lynch 937

1992, and they remain an invaluable resource even for readers who have no particular interest in ethnomethodology or conversation analysis (Sacks, 1992). Jefferson, Pollner, Sudnow, Wieder and Bittner did not just apply ethnomethodology and conversation anal- ysis to specific instances and settings. In distinctive ways, they developed memorable cases and themes, and devised methodological resources that have proved, and should still prove, to be sources of insight and methodic practice in STS, among other fields.

Egon Bittner worked as a research assistant for Garfinkel when he was a student at UCLA in the late 1950s and early 1960s. He is listed as a collaborator for a chapter in Garfinkel’s Studies in Ethnomethodology (1984 [1967]: Ch. 6, ‘“Good” organizational reasons for “bad” clinic records’). That chapter provides a highly original analysis of the ‘deficiencies’ of organizational records for sociological purposes. Going back at least as far as Emile Durkheim, sociologists have expressed skepticism about the use of official records as sociological data. The problem with such records is that they tend to present the organization (or the particular agents who compiled them) in a favorable light. Garfinkel and Bittner turned that deficiency into a phenomenon, and elucidated how case files were compiled to anticipate possible inquiries about the motives and adequacy of the clinical work they ‘described’. Bittner went on to a distinguished career in sociology at Brandeis University, and was President of the Society for the Study of Social Problems in 1981–82. He is best known for his extended ethnographic study of the day-to-day organization of police work (see, for example, Bittner’s (1967) essay on the pragmatic use police powers of arrest for ‘keeping the peace’). In addition to producing outstanding ethnographic studies, Bittner (1965) wrote a brilliant essay on ‘the concept of organiza- tion’, which demonstrated how the concept of organization is itself used by members as a constitutive feature of how they order their daily affairs.

Gail Jefferson was studying dance as an undergraduate student at UCLA in the 1960s when she enrolled in one of Harvey Sacks’ courses. Jefferson was so taken with Sacks’ lectures that she began to record and transcribe them. Sacks’ lectures themselves used tape-recordings of phone conversations, meetings, and group therapy sessions, and Jefferson later developed a distinctive transcription system for annotating (one might say, choreographing) silences, overlapping utterances, and features of stress and prosody that elude standard lexicography. This system, and the type of close analysis of interac- tional organization that it facilitated, became an established fixture of conversation anal- ysis. Jefferson sometimes spoke of herself as a ‘technician’, and she professed to have little background or interest in theory, but her work is rich with understated theoretical significance. Many of her publications are co-authored with Sacks and Emanuel Schegloff, but her own work on the interactional organization of intimacy and ‘troubles talk’ (Jefferson, 1988), and her reflections on transcription (Jefferson, 1983) not only express an acute and incisive attention to detail, they show deep insight into how such details compose and express linkages between the fleeting minutia of personal conduct and what Garfinkel, after Durkheim, called the production of the ‘immortal society’. Jefferson’s work is rarely discussed, or even mentioned, in STS research, but her tran- scription system – and, more importantly, the attention to detail her system embodied – has provided many of us with a means with which, as Foucault once put it in a very different context, ‘to push a little farther back the foamy line of language, to make it encroach upon that sandy region that is still open to the clarity of perception but is already no longer so to everyday speech’ (Foucault, 1989: 208).
Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011

938 Social Studies of Science 41(6)

Melvin Pollner is best known for his work on ‘mundane reason’. His 1970 dissertation, later revised for publication by Cambridge University Press was a lucid introduction to ethnomethodology that also developed a distinctive approach to that field (Pollner, 1987). Pollner took some rather unpromising empirical material – notes from the steady- stream of minor cases observed in traffic court – and developed a unique empirical investigation of the ‘problem of reality’. Assuming the perspective of an anthropologist witnessing an exotic ritual, he delved into the presuppositions of the ‘mundane reason- ing’ deployed when, for example, a motorist accused of speeding blames the offense on a faulty speedometer, or when a judge invokes differential perspectives and interests to resolve a ‘reality disjuncture’ between a policeman’s and motorist’s accounts of the speed of the vehicle in question. Pollner’s insightful analysis had, and continues to have, clear applicability for analyses of disputes and other exchanges among scientists in which ontological questions are negotiated in relation to matters of reputation, credibility, interest, technical mediation, and perspective. Pollner also endeavored to apply his insights about reality ascription to the methodic practices and observational claims made by ethnographers, and his insights fed directly into debates about reflexivity in science studies (for example, Woolgar, 1988). Pollner, together with one of his teachers, Don Zimmerman, also co-authored an early essay that articulated the ethnomethodological injunction to treat methodic practices as topics rather than (or as well as) resources for sociology (Zimmerman and Pollner, 1970) – an injunction that later was deployed as a fundamental argumentative and analytical resource in STS.

David Sudnow was a fellow graduate student with Sacks and Emanuel Schegloff at Berkeley in the 1960s, and he also was a faculty member with Sacks in the School of Social Sciences at UC, Irvine, in the 1970s. Though he was a brilliant and highly original analyst of practical actions, unlike Sacks he did not develop a sustained line of academic work. Instead, he wrote on a series of, seemingly disconnected, topics. His early ethnog- raphy, Passing On, remains one of the finest ethnographies of daily routines in a medical setting (Sudnow, 1967). He also wrote insightful essays on the informal logic of plea- bargaining in criminal courts (Sudnow, 1965), and on the temporal parsing of visual scenes through photography (Sudnow, 1972). Sudnow left Irvine in the early 1970s, and for the remainder of his life he held occasional, short-term academic appointments while developing a method for teaching music. Sudnow was an accomplished jazz musician. His phenomenological study of playing improvisational jazz on the piano (Sudnow, 1978 [2001]) was written well before ‘the body’ became a favorite theme for social scientists, but it remains a rare exemplar of how to describe embodied (and not just bodily) conduct. His ethnographic and phenomenological explications of the temporality of embodied actions and the ‘logical’ organization of improvised conduct have great potential for social studies of science, but remain largely unexploited.
D. Lawrence Wieder is best known for his book Language and Social Reality: The Case of Telling the Convict Code (Wieder, 1974). As the title suggests, the study exam- ines a much-chronicled ‘code’ among convicts, but instead of simply codifying the rules (such as, ‘Do your own time’, and ‘Don’t snitch’ on a fellow inmate), Wieder closely examines the circumstances in daily prison life in which ‘the convict code’ becomes salient – as something to recite, invoke, or ‘tell’. His study provides a distinctive picture of how moral codes function in daily activities. Rather than treating them as the

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
Lynch 939
structural functionalists did, as abstract governing principles, he treats them as discursive resources and repertoires that are used strategically and artfully by savvy convicts in interactions with one another, with prison authorities and, not incidentally, with ethnog- raphers. Later, Wieder conducted research on the management of audience perspective by stage magicians (he was himself an accredited magician), and on the organization of human–animal communication. The latter study (Wieder, 1980) is rarely cited in the STS literature on the subject, but provides an illuminating case study of routine interactions between chimps and ‘chimpers’ (technicians who handle the animals). Wieder suggests how those communications are crucial for setting up experiments, but are not admissible within the ‘methodogenic ontology’ of ‘behavioristic operationalism’ – the predominant frame in which experiments are designed. Behavioristic operationalism turns the inter- subjectivity of the face-to-face encounter between chimp and chimper into ‘unobserv- able’ and ‘subjective’ epiphenomena. This early study is rich with suggestive analytic themes that pertain to recent STS interests in human–animal relations.
Pollner (1991) once wrote that ethnomethodology had settled down and moved to the ‘suburbs’. His metaphor suggested that ethnomethodology had lost its sharp, critical edge and had become comfortably settled at the margins of academic respectability. With this recent rash of deaths, it is now tempting to suggest that ethnomethodology is moving from the suburbs to the graveyard, but while there is reason to be concerned about the fate of the field, it would be premature to forecast a gloomy end. As an indication of the glacial pace of change in fields like sociology, ethnomethodology was only recently established as a section of the American Sociological Association, but it also continues to crop up in unexpected quarters, and has a substantial presence in the field of CSCW (see, for example, the essay on ‘technomethodology’ by Button and Dourish, 1998). As an indication that it is far from finished, ethnomethodology remains one of the least-well- understood orientations (sometimes misnamed as ‘paradigms’) in the social sciences today. As Garfinkel (2002: 91) pointed out, the question ‘What is ethnomethodology?’ continues to be asked, despite decades of research and numerous attempts to answer it in a straightforward or not-so-straightforward way. At the very least, continued pursuit of the question indicates that interest persists and novel answers may be forthcoming.
Garfinkel was a remarkable and volatile character. It would not be inaccurate to describe him as a temperamental, self-fashioning genius, though that posture was far from constant in his dealings with students and colleagues. There were many contradic- tions in his life and work – many of which appear to have been deliberate. Though he disavowed ‘ironic’ social science – by which he meant descriptions that presume to know other and better than the ‘actors’ studied about the sources of their own actions – his writ- ing is infused with a more familiar kind of irony: an undertone that suggests a not-so-well concealed joke on the world. Despite teaching of the ubiquitous and inescapable every- dayness of social existence, and despite denying a transcendent ‘place’ for a privileged few who presume self-reflexive perspicacity, Garfinkel often seemed to be speaking for posterity. Despite his incisive demonstrations of the contingency of social existence, he tried to control his legacy. And, yet, he often was very generous with time and credit, and he repeatedly emphasized that the legacy he wanted was not a monument to him as a great theorist, but a continuation of the unfinished business of ‘the company’. At this point, it seems safe to say that he does and will continue to have a legacy, albeit one that he may never have imagined.

Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
940 Social Studies of Science 41(6)
Garfinkel is survived by his wife Arlene, his daughter Leah, and son Mark, as well as by many former students.


An abbreviated and edited version of this obituary was published in The Guardian (UK), 13 July 2011 (available at www.guardian.co.uk/education/2011/jul/13/harold-garfinkel-obituary (accessed 20 July 2011)). I would like to thank the editors of The Guardian for permission to use portions of that obituary in this essay.


Bar-Hillel Y (1954) Indexical expressions. Mind 63: 359–379.
Bittner E (1965) The concept of organization.
Social Research 32: 239–255.
Bittner E (1967) The police on skid row: A study of peace keeping.
American Sociological Review
32: 699–715.
Bjelic D and Lynch M (1994) Goethe’s ‘Protestant Reformation’ as a textual demonstration:

Comment on Jackson. Social Studies of Science 24(4): 703–724.
Bloor D (1976)
Knowledge and Social Imagery. London: Routledge & Kegan Paul.
Button G and Dourish P (1998) On ‘technomethodology’: Foundational relationships between

ethnomethodology and system design. Human–Computer Interaction 13(4): 395–432. Collins H M and Evans R (2002) The third wave of science studies: Studies of expertise and
experience. Social Studies of Science 32(2): 235–296.
Collins HM and Pinch T (1982)
Frames of Meaning: The Social Construction of Extraordinary
Science. London: Routledge & Kegan Paul.
Coulter J (1989)
Mind in Action. Oxford: Polity Press.
Foucault M (1989)
The Birth of the Clinic. London: Routledge.
Garfinkel H (1941) Color trouble, a short story. In: O’Brien EJ (ed.)
The Best Short Stories of
1941: The Yearbook of the American Short Story. Boston: Houghton Mifflin.
Garfinkel H (1960) A comparison of decisions on four pre-theoretical problems by Talcott Parsons and Alfred Schutz. Unpublished paper, Department of Sociology, University of California,

Los Angeles.
Garfinkel H (1984 [1967])
Studies in Ethnomethodology. Oxford: Polity Press. (Englewood Cliffs,
NJ: Prentice Hall, 1967).
Garfinkel H (ed.) (1986)
Ethnomethodological Studies of Work. London: Routledge & Kegan Paul. Garfinkel H (2002) Ethnomethodology’s Program: Working Out Durkheim’s Aphorism. Lanham,
MD: Rowman & Littlefield.
Garfinkel H (2007) Four relations between literatures of the social science movement and their

specific ethnomethodological alternates. In: Hester S and Francis D (eds) Orders of Ordinary
Action: Respecifying Sociological Knowledge. Aldershot: Ashgate, 13–29.
Garfinkel H (2008)
Toward a Sociological Theory of Information. Boulder, CO and London: Para-
digm Publishers.
Garfinkel H, Lynch M and Livingston E (1981) The work of a discovering science construed

with materials from the optically discovered pulsar. Philosophy of the Social Sciences 11(2):
Garfinkel H and Sacks H (1986 [1970]) On formal structures of practical actions. In: Garfinkel

H (ed.) Ethnomethodological Studies of Work. London: Routledge & Kegan Paul, 158–189. (Originally in: McKinney JC and Tiryakian EA (eds) Theoretical Sociology: Perspectives and Development. New York: Appleton-Century-Crofts, 337–366.)
Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
Lynch 941
Hill R J and Crittenden K S (eds) (1968) Proceedings of the Purdue Symposium on Ethnomethod- ology. Purdue, IN: Institute for the Study of Social Change, Department of Sociology, Purdue University.
Jefferson G (1983) Issues in the transcription of naturally-occurring talk: Caricature versus capturing pronunciational particulars. Tilburg Papers in Language and Literature, No. 34: 1–12. Available at: www.liso.ucsb.edu/Jefferson/Caricature.pdf (accessed 7 July 2011).
Jefferson G (1988) On the sequential organization of troubles talk in ordinary conversation. Social Problems 35: 418–441.
Knorr K (1981) The Manufacture of Knowledge: An Essay on the Constructivist and Contextual Nature of Science. Oxford: Pergamon.
Knorr-Cetina K (1983) The ethnographic study of scientific work: Towards a constructivist sociology of science. In: Knorr-Cetina K and Mulkay M (eds) Science Observed. London: Sage Publications, 115–140.
Kuhn T (1970) The Structure of Scientific Revolutions, 2nd edn. Chicago: University of Chicago Press.
Latour B and Woolgar S (1979) Laboratory Life: The Social Construction of Scientific Facts. London: Sage Publications.
Latour B (2005) Reassembling the Social: An Introduction to Actor-Network Theory. Oxford: Oxford University Press.
Livingston E (1986) The Ethnomethodological Foundations of Mathematics. London: Routledge & Kegan Paul.
Livingston E (1999) Cultures of proving. Social Studies of Science 29(6): 867–888.
Lynch M (1985)
Art and Artifact in Laboratory Work: A Study of Shop Talk and Shop Work in a
Research Laboratory. London: Routledge & Kegan Paul.
Lynch M (1993)
Scientific Practice and Ordinary Action: Ethnomethodology and Social Studies
of Science. New York and Cambridge: Cambridge University Press.
Macbeth D (1996) The discovery of situated worlds: Analytic commitments, or moral orders?

Human Studies 19(3): 267–287.
Parsons T (1949 [1937])
The Structure of Social Action, Vol. 1: Marshall, Pareto, Durkheim, 2nd
edition. London and New York: Macmillan.
Pollner M (1987)
Mundane Reason: Reality in Everyday and Sociological Discourse. Cambridge,
UK: Cambridge University Press.
Pollner M (1991) Left of ethnomethodology: The rise and decline of radical reflexivity.
Sociological Review 56: 370–80.
Rawls A (2002) Editor’s introduction to: Garfinkel H,
Ethnomethodology’s Program: Working
Out Durkheim’s Aphorism. Lanham, MD: Rowman & Littlefield, 1–64.
Sacks H (1992)
Lectures on Conversation, 2 vols. Oxford: Blackwell.
Sharrock W and Anderson B (1991) Epistemology: Professional scepticism. In: Button G (ed.)

Ethnomethodology and the Human Sciences. Cambridge: Cambridge University Press, 51–76. Suchman L (1987) Plans and Situated Actions. Cambridge: Cambridge University Press. Sudnow D (1965) Normal crimes: Sociological features of a penal code in a public defender’s
office. Social Problems 12: 255–276.
Sudnow D (1967)
Passing On: The Social Organization of Dying. Englewood Cliffs, NJ: Prentice-
Sudnow D (1972) Temporal parameters of interpersonal observation. In: Sudnow D (ed.)
in Social Interaction. New York: Free Press, 259–279.
Sudnow D (1978 [2001])
Ways of the Hand: The Organization of Improvised Conduct. London:
Routledge & Kegan Paul. Revised edn: Cambridge, MA: MIT Press, 2001.
Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011
942 Social Studies of Science 41(6)
Traweek S (1988) Beamtimes and Lifetimes: The World of High Energy Physics. Cambridge, MA: Harvard University Press.
Weizenbaum J (1976) Computer Power and Human Reason: From Judgment to Calculation. San Francisco: W.H. Freeman.
Wieder DL (1974) Language and Social Reality: The Case of Telling the Convict Code. The Hague: Mouton.
Wieder DL (1980) Behavioristic operationalism and the life-world: Chimpanzees and chimpanzee researchers in face-to-face interaction. Sociological Inquiry 50(3/4): 75–103.
Wikipedia (n.d.) Ethnomethodology. Available at: http://en.wikipedia.org/wiki/Ethnomethodol- ogy (accessed 20 July 2011).
Woolgar S (ed.) (1988) Knowledge and Reflexivity: New Frontiers in the Sociology of Knowledge. London: Sage Publications.
Zimmerman DH and Pollner M (1970) The everyday world as a phenomenon. In: Douglas JD (ed.) Understanding Everyday Life: Toward the Reconstruction of Sociological Knowledge. Chicago: Aldine, 80–103.

Biographical note

Michael Lynch was a student of Garfinkel’s in the 1970s and collaborated with him in the early 1980s. Lynch’s book Scientific Practice and Ordinary Action (Cambridge University Press, 1993) discusses the relevance of ethnomethodology for social studies of science. He is current editor of Social Studies of Science.
Downloaded from sss.sagepub.com at CORNELL UNIV on December 30, 2011