Theory & Science (2004)

ISSN: 1527-5558

The past, present and future of an alternate,asymmetric and incommensurable approach to sociology

Wes Sharrock
University of Manchester

The exchange between Alex Dennis and Michael Lynch involves three main elements – ethnomethodology, very largely the creation of Harold Gafinkel, Alfred Schutz’s (philosophical) account of science and social science, and, in a more subsidiary role, Sociology of Scientific Knowledge (SSK 1 ). The main issue of dispute is over whether Michael Lynch gives an accurate rendering of the extent to which ethnomethodology follows from - and adheres to - Schutz’s ideas. In the event that ethnomethodology is seen to deviate from Schutz’s ideas, the dispute continues over whether it is right to do so on the basis of arguments from SSK, whose research into the practice of science purportedly discredit Schutz’s ideas. Since this is a background introduction to this dispute, I have tried to keep the style as informal and the references as few as possible.

On the assumption that many readers of this journal will be largely unfamiliar with Schutz, Garfinkel, ethnomethodology and SSK, it is my role to provide a background account which will situate (rather than adjudicate) the dispute between Dennis and Lynch.

In my mind, there can be no question but that ethnomethodology originated in Harold Garfinkel’s adoption and transformation of Alfred Schutz’s ideas. This in no way diminishes Garfinkel’s ingenuity or achievement, for it was surely a remarkable achievement to see in Schutz’s work the possibility that Garfinkel identified, one that Schutz himself may have been incapable of perceiving. It was also an original and profound achievement to develop that possibility in the direction that Garfinkel took it. The debt to Schutz’s ideas in Garfinkel’s earlier writings is often plainly manifested by extensive direct quotations. Whether Garfinkel then ‘gave up’ on Schutz is a more difficult question to answer, not least because Garfinkel has been a rather reluctant publisher.

Garfinkel’s work on Schutz began during his doctoral studies at Harvard in the late 1940s and early 1950s. His first book appeared in 1967, to very considerable attention (a book that consists mainly of the relatively few papers that had been published up to then). His second book appeared only in 2002, when Garfinkel was into his eighties. Some papers were published in the intervening years, and some drafts have circulated in a samizdat fashion, but there was some disappointment that the recent book contained a very long editorial introduction and only a relatively small proportion of Garfinkel’s own writings. Add to this the fact the since the earliest period of his work Garfinkel has preferred to set out his ideas through extensive empirical examples, accompanied only by terse theoretical statements. AS a result, it can be difficult to discern whether he is using new words to restate an existing idea or is marking a significant change in position. Lynch is also right to say that Garfinkel maintains that scholarly faithfulness is not a priority of ethnomethodology, and that it is appropriate to raid thinkers for ideas that might be useful to empirical inquiry without undue respect for the context from which they are taken. It is also clear that in later writings Schutz’s name appears much less frequently (it appears only once in Garfinkel’s part of the new book, and then only in the course of an autobiographical remark). Nowadays, other names (such as Maurice Merleau-Ponty’s) appear with warm commendation. It is not easy to trace out the course or character of Garfinkel’s ideas since the 1960s, and there is, therefore, room for disagreement about it (as follows).

For those unfamiliar with the history of sociology over the last half-century, or with ethnomethodology in particular, there will be questions prior to those being argued between Dennis and Lynch. These may well ask why we should be interested in the question of Garfinkel’s relation to Schutz or the viability of Schutz’s ideas of science. The reader might well ask: but what are the ideas involved in these arguments? In an attempt to answer this question I will sketch Alfred Schutz’s philosophical, pre-sociological thought, and the place and character of his ideas about science and social science in that context. Schutz’s thought, like the thought of most significant sociological thinkers and social theorists, responds to the idea that science is a distinctively, even uniquely, rational mode of thought, which is why the discussion of science leads into a discussion of ‘the rationalities’. I will try to present Schutz’s thought on this topic in a way which makes the best and most favourable sense of it. I will then try to capture the transformation that Garfinkel worked on Schutz’s ideas, again focussing on the role of ‘the rationalities’ as a point of transition, explaining the critical significance of Garfinel’s move in the context of then contemporary American sociology, and describing, briefly, the set of sociological study policies that the transformation initially engendered. In doing so, I will indicate why Garfinkel, in his more recent terminology, characterises ethnomethodology as an alternative, asymmetrical, and incommensurable approach to sociology.

Schutz’S Treatment of Social Action

Schutz’s work was philosophical in character, and was mostly presented in essay form. 2 Some of those essays had an ‘applied’ character in that they talked about concrete examples (such as the experience of ‘a stranger’ or the relations between people ‘making music together’) but these were not based on empirical research. Schutz’s project originated in an attempt to clarify the philosophical presuppositions of Max Weber’s proposal that sociology be conceived as the study of social action. Weber was a most prominent turn-of-the-century figure in German social, historical and political thought, and is nowadays regarded – even revered – as one of the major, if not the major, precursons of sociology. His work is by no means out of date, and there are many who would call themselves ‘Weberian’ in their approach, or would at least affiliate themselves to ‘the social action’ tradition (one critic, Colin Campbell (1996) thinks that the social action tradition is nowadays dominant).

The idea of a ‘social action’ approach is, grossly, that of treating the way in which people in society act as being based in their own understandings. People respond to their circumstances, to the world around them, or to other people, on the basis of how they understand those circumstances. Those understandings will originate in the individual’s sociocultural setting. Though such an idea might seem the most transparent and reasonable conception, not everyone agrees, and this idea remains controversial.

Schutz (1899-1959) accepted Weber’s conception of sociology as the study of social action, but thought that Weber had failed to work through the philosophical presuppositions of that idea. Schutz undertook, in his first major work (1967, although his work on it began in the 1920s), to clarify those philosophical presuppositions. He adopted, as his means of clarification, the then newly-developed philosophy of phenomenology, as pioneered by his teacher, Edmund Husserl (1859-1938).

Schutz’s basic concerns remained fairly constant throughout his long career, and involved the elaboration of these philosophical presuppositions, forging (often critical) connections with recent or emerging tendencies in sociological thought.

The phenomenological approach, in application, undertakes to describe the fundamental forms of our experience of the world, and Schutz’s problematic can be thought of schematically as that of understanding how the objectivity of the social world is available in individual experience. For Schutz, the social world is experienced as ‘an objective reality’. (This is a designation that is important to many social scientists, though at the same time some of them want to see that as being threatened by, for example, Schutz’s approach. They fear that because Schutz concentrates only on experience, this portrays social reality as something merely ‘subjective’). Schutz seeks to describe the way in which the social world might be accessed by individuals from within society.

Schutz’s exposition locates a minimal point of departure: that our experience of the world (natural and social) takes place within a ‘here-and-now’, and it is fundamental to the world as I experience it, that it is arranged around me. (Thus, for example, I relate to events as ones that occurred before I was born or are happening to me now, or to people as those that I know and those that I do not, to places as ones within my current sight and reach, and those beyond the bounds of my present situation, and so forth.) There is an important sense in which the world is my world, but not in a way that contrasts with - or privileges me - relative to other people. The immediate surrounds of the succession of here-and-now situations which make up my life provide the primary site of my experience. But my experience is not restricted to whatever is currently and concretely present to me: I am aware that where I am now, and what I am currently experiencing, is a local manifestation of wider arrangements (as one is aware that one is ‘in Western Europe’ or ‘in the twenty first century’).

I experience and act in one or another, here and now, and my actions (premised on what I have acquired through experience) enable me to manipulate and affect my circumstances. But whilst I can change my world, I also recognise my own actions have proportionately small effect in shaping my circumstances. My actions, rather than being my creation, are overwhelmingly an inheritance. I am aware that – in respect of the social world – the prevailing mass of my present circumstances have been formed as a result of the actions of others, of those who are predecessors (and who may now be long dead), as well as of these who are contemporaries, still alive and active (and who are directly or indirectly affecting my situation). I am also aware that others are like myself in that they too undergo experience from their respective here-and-now, and I communicate with, and learn from, them. Thus, not only are my actual circumstances very much an inheritance, but so too are the understandings which I bring to bear on interpreting these. I have personally originated only a very small proportion of my understandings. Much of the (social) world I understand is second hand, through what I have learned from others. My own stock of understandings has been handed down to me from a ‘social distribution of knowledge’. Therefore, my experience is scarcely privatised: it is ubiquitously informed by understandings that I share with others, and it is experience of a world that is one and the same for me and others. It is a world that is known to us in common. It is a feature of our experience that we unreflectively assume (within limits) that the world appears to, and is known in, the same way by others. Indeed, this is not perhaps strong enough: as a matter of course we normally require it of others that their experience of things should reciprocate our own experience. Schutz talks of ‘reciprocity of perspectives’ and ‘interchangeability of standpoints’ to delineate this.

In this (so far) uncomplicated picture, my understandings are most relevantly comprised of ‘recipes’ that I can and do follow: standardised ways of doing various things which are practically effective, ones which I have also mostly learned from others. Thus, there is a ‘recipe’ to send a letter (e.g., put the letter in the envelope, buy a stamp, stick it on the envelope, stick the envelope in a post box.) Many of the understandings I have of what I (and others) should (and will) do comes through a stock of ‘typifications’ which, again, are mostly socially originated and shared. Thus, although I may have never met an individual golfer, I have understandings about what kinds of people golfers are, what kinds of things they do/should do, what I would have to do if I were to take up golfing, and so forth. Much of our relationship with ‘the wider society’ is conducted through such typifications - I entrust my money to bankers though I have might never met (and probably will never meet) the bank workers who process my account (but I know, for example, how many days to allow them to get the money registered in my account.)

The aforementioned objectivity of the social world resides (in significant part) in the availability of the considerable and elaborate stock of ‘social types’ and in the fact that I can use them to understand, relate to the activities that go on around me, and organise my own conduct. These types are common to other members of the same society, and I anticipate and require that their understandings and actions will be reciprocally informed by them. There is no guarantee that this will prove to be the case, but such understandings effectively get us through an awful lot of any day and a great part of our lives. These understandings have a ‘common sense’ character, very much taken for granted and inhabited unreflectively by members of the society. They make up the means by, and form in which, the society is known – the society is experienced as the world of everyday life, a matter of movement through familiar settings such as houses, streets, public places, service organisations, workplaces and so on – stopping at the pumps to buy fuel for the car, grocery shopping in the supermarket, making repairs to an office photocopier, working on a production line, keeping office files in order, ad infinitum. The content of ‘common sense’ is standardised to a given social setting, quite other understandings may play the same role in a different time or another place.

The notions of ‘the world of daily life’ and ‘common sense understandings’ are central in Garfinkel’s transformation of Schutz’s work into a deep source of methodological trouble for sociology. However, the uncomplicated story needs a little complication, and it is here that the topic of science and the notion of rationality come into play.

Science and Rationality

The topic of whether science is a superior form of thought – even uniquely so – is never far from the thoughts of sociologists, though the idea is sometimes viewed favourably, and at others as anathema. The topic is almost de rigeur for anyone setting out an approach to sociology, but Schutz has an additional reason for approaching it, which is as an unavoidable part of his project. This necessity arises from his aim of describing the form of our moment-to-moment experience. That experience has been characterised as an experience of the world of everyday life, for that is the fundamental form in which it comprises experience of social reality.

However, if we consider the flow of experience, we will recognise that this is not smoothly continuous – it has, rather, an episodic structure. For example, we are not always wide awake, and our experience can drift away from the immediate here-and-now and into daydreams, or we may fall fully asleep and find ourselves in the world of dreams. In daydreams, perhaps more so in dreams, we will experience the matter of those dreams as real, we will, in Schutz’s terminology, bestow, ‘the accent of reality’ upon these imaginings – we will have feelings of fright (even terror, or of affection or amusement) in response to events in our dreams, just as we would to comparable events in our waking lives. At the time, and for us, they are real. But we awake, and realise it was only a dream, that these things did not really happen.

There are other episodic features to our experience, as when we go to the theatre and watch a play. When we become engrossed in a play’s performance we forget that we are watching people dressed up and pretending to be certain characters. We respond to the events being enacted as if we were witnessing them, though when the performance is over we leave the theatre well aware that Caesar’s actual assassination took place many years ago, and not before our eyes that night. To conceptualise these episodic discontinuities in the flow of experience, Schutz used the expression ‘multiple realities’ to reflect the fact that whilst we are immersed in these diverse episodes we regard their inner occurrences as real. Amongst these ‘multiple realities’ can be included episodes of religious enactment or work in the sciences. The world of everyday life is one of these realities, but it does not stand on the same footing as the others, but is treated as ‘the paramount reality’. The movement into and out of these other realities involves discontinuous transitions – in the theatre, the lights go down, the play starts, and one is witnessing events in Ancient Rome but we do not move permanently into these worlds. The movement is from the world of daily life and back into it – we go from the street into the theatre and back onto the streets when the play is over. The world of daily life provides the standard of what really happened – that it was ‘only’ a dream - and there never was any monster chasing us, or the murder of Caesar was only a performance.

The ‘world of science’ shares these features of a self-contained reality, or, in Schutz’s formulation, ‘a finite province of meaning’. When someone works in a scientific laboratory they inhabit a world that is not accessible to them on the basis of everyday understandings, nor to their everyday modes of access. For example, the particle physicist inhabits a world of interactions between inconceivably small particles that cannot be seen with the naked eye (or even with a telescope), but can only be detected through use of hugely elaborate machineries. The working particle physicist acts in relation to that reality (his activities comprise manipulations of the behaviour of these miniscule particles that effect probings of the particle’s physical structure. Assuredly, these interacting particles have ‘the accent of reality’ whilst the physicist is at work. But the physicist (who cannot stay awake or at work interminably) enters the lab at the start of the working day, and exits at its end, thereby moving from and eventually returning to the everyday world. Driving home, the physicist’s environment is no longer that of the particles. Now, his/her world is that of the traffic and the numerous other drivers with, perhaps, their obstructive and irritating ways. Science too is a finite province of meaning, standing in the same formal relationship to everyday life as does absorption in a meaningful religious ceremonial or a captivating dramatic performance.

The discontinuity between such ‘self-contained’ provinces of meaning must be recognised as a relative one, and must not be over-emphasised. After all, as the assassins prepare to knife the Emperor, one does not stand up and shout ‘Look behind you!’ nor call the police. A schematic rendition of Schutz is being given, but one which will bear delicate elaboration.

Science as Self-Contained

Schutz talks about the way in which one inhabits these ‘provinces of meaning’: contrasting them in terms of the ‘attitudes’ that are appropriate to and dominant within them. The world of daily life is lived under the ‘natural attitude’, one in which the emphasis is on action, and where concerns are overwhelmingly practical. Things are getting done, and the interest is exhausted by being able to get them done, and in finding solutions to any problems that get in the way of doing what needs to be done. If the television won’t work, get the repairman!

Schutz builds his characterisation of these attitudes around the paradigm of the Cartesian radical sceptic, and, thus, in terms of the possibility for doubt. The radical sceptic is commanded to doubt everything that can possibly be doubted, which attitude is incompatible with action. Descartes sceptical teaching held that nothing is immune to doubt. A true sceptic would be endlessly busy doubting everything, and could never get around to doing anything.

The ‘natural attitude’ is a near polar contrast to the rigorous Cartesian sceptic. The natural attitude involves rejection of doubt for its own sake in favour of taking things at face value, and without question (one opens one’s mail without wondering if there might be a poisonous spider in the envelope). Doubt enters in daily life only where practically relevant, where the expectable routines and processes of daily life go wrong. Characteristically, doubt only enters so far as it is relevant to getting things put right again (practically). In between these two extremes – pure scepticism and the natural attitude - falls the world of science. Within this world there is greater value on doubt for its own sake, but much less license for it than there is in the sceptic’s domain. It is a feature of scientific work that doubts can be raised for their own sake. But whilst scientists typically raise doubts that certain things might not be as they presently appear, they do not undertake to doubt anything. They can only get on with their work if they restrict the possibility of doubting.

The word ‘attitude’ might suggest a strongly cognitivist or mentalist orientation on Schutz’s part. This might be a fair interpretation, given the Cartesian inheritance of phenomenology, but it might be asked whether there is anything indispensably cognitivist in what Schutz has to say. One can take ‘attitude’ as a label for a mental item, but one can equally well (and in my view, more usefully) understand it in the sense in which it occurs in the title of Angus Wilson’s novel Anglo-Saxon Attitudes. Here, attitude identifies a stance, posture or mode of address. Thus, in our daily lives we do not spend all our time mentally reserving the right to doubt, we just go at things in an unhesitating way.

Given that the hallmark of the natural attitude is suspension of the possibility of doubt, the fact that the scientific inquirer enjoys a greater latitude for doubt calls for its differentiation out from ‘the everyday’. The differentiation is not, and cannot be, an absolute one, for the reason that the suspension is only partial – the scientific inquirer remains, to a considerable extent, subject to the natural attitude – scientific doubt is very selective and focussed. The scientific researcher takes notes with a pen, on paper, at a desk, under a lamp, checking the time by the clock, etc. – it is what is being written with the pen on the paper that is incorporated within the world of science. Reducing it to a slogan, the scientist’s doubts are not ‘for their own sake’ but as a means to an end, namely the accumulation of knowledge. Action under the natural attitude can also involve the acquisition of knowledge. But, as pointed out, it is practical success, not the accumulation of knowledge, which is the objective. The everyday actor may seek to find out, but the search will be bounded by the need to, find out enough to overcome the practical problem in hand. Further, the scientist’s aim is not to accumulate knowledge as a matter of assembling an enormous catalogue of known items, established facts. Scientific inquiry is directed towards the accumulation of systematised knowledge - Schutz contrasts the ‘expert’ with ‘the man on the street’ (with the scientist included among experts). A main axis of contrast is with respect to the organised and systematic character of the individual’s stock of knowledge, with ‘the man in the street’ having a largely unstructured and ad hoc assemblage of knowledge. The contrast plays out in at least two other key ways. First, in terms of the nature of the problem to be solved by practical and scientific work respectively; and, second, the priority on closing the problem.

Somewhat simplistically put, everyday practical problems are the individual’s personal problems. I want something done, I’m trying to get it done, it starts to go wrong, I have to institute some inquiry to see if I can figure out why a usually reliable way of doing this has suddenly gone awry´┐Ż - this is only a problem because it obstructs my action and I want to dispose of it, otherwise I would never pay them any attention. By contrast, the scientific investigator is not usually investigating a problem because it is any kind of nuisance to him or her personally. Not knowing how to reconcile the large and the small scale in physics does not get in the way of public transport, government finance, police patrolling, weekend yachting or any other of anyone’s practical affairs, but it surely engages the attention of many scientists. Something is a problem in science not because of where it stands in the investigator’s practical priorities, but by virtue of its relation to the accumulated corpus of scientific understandings. The scientific investigator is not positioned in a personally defined situation, but in an impersonally identified one, relative to what is already known to (and currently unresolved) by the discipline’s dominant schemes. Thomas Kuhn’s hugely influential The Structure of Scientific Revolutions (1962) argues that the main business of the natural sciences is what he calls ‘normal science’, which does not involve seeking fundamental innovations but largely consists in working out the thorough details and upgrading the precision of the currently dominant scheme of ideas and associated forms of investigative practice. Kuhn’s account is congruent with Schutz’s to this extent: first, that most scientist’s working preoccupation is with refining an existing scheme (which means that their ‘problems’ are not personal ones, but ones formed relative to the scheme); second, that the concern to explore thoroughly, organise and integrate that scheme – which is what ‘normal science’ amounts to - testifies to Schutz’s claim that natural science inquiry prioritises precision of meaning and systematisation of knowledge.

The Rationalities

As mentioned, the thought that the scientific (or theoretical) point of view might be uniquely authoritative has had great attraction for many in the philosophy of science and in the social studies. This attraction has often been expressed in the idea that science is both the exemplar and general yardstick of rationality. Whether there is a general standard of rationality (or not) still occasions significant controversy, and doctrines of ‘relativism’ characteristically argue for a plurality of divergent, irreconcilable standards. The idea of rational action has often been understood to mean action in accord with the best scientific knowledge. When individuals act, they aim to achieve certain objectives, but can only be successful if the world around them can be manipulated in ways which will bring it into accord with the individual’s purposes. This will only happen if the individual knows how to manipulate the world, which requires that he or she knows what the world is really like. If science tells us what the world is really like, then it is only if the individual’s understandings conform to what science tells us that his or her actions can really succeed. Thus, science is the measure of rationality, and rational action is effective action.

Such a conception is, however, a stipulative, prescriptive one; it is a conception that the theorist contrives, usually as an exclusive standard. Schutz, however, proposes a descriptive one. The notion of rationality itself is not the creation of theorists, but is one that circulates within the society, where it is used in several different senses. Schutz does not deny that science has a distinctive claim to rationality, only that it is in terms of the features that have been identified above. As used amongst the members of society, rationality does not consist of a single unified conception. Rather, actions are considered rational if, inter alia, they are based on a strategy, they follow a rule of procedure that is generally recognised as appropriate, or if alternative courses of action are considered and their consequences analysed prior to the action. Garfinkel (1967: 271) gives, after Schutz, a tabulation of fourteen criteria of rationality, singling out four as distinctive to science. As Dennis rightly insists, Schutz’s argument is that ten of these criteria are common to both science and everyday reasoning (so that science, but not daily life, exhibits all fourteen of them). This certainly undermines the idea that there is any sharp and absolute distinction between science and everyday life with respect to rationality, meaning that science cannot be treated as a general arbiter of rationality.

The four criteria distinctive to science are:

These concisely summarise much of Schutz’s account of science, as given above.

Gartfinkel saw a further implication of Schutz’s conception, which was that the scientific standards of rationality cannot be recommended as standards for the evaluation of practical action. Ethnomethodology was initially notorious for various so-called ‘experiments’, which Garfinkel got his students to carry out. These were not really experiments in any strong sense, and were more properly called ‘demonstrations’. For example, Garfinkel’s students would be sent out to engage in conversations with other people, looking for any possible ambiguities in what the others said. When such ambiguities were identified, they were to be challenged with demands for clarification, which were to be persistently made in an attempt to achieve full clarity. Students were thus being asked to seek ‘semantic clarity and distinctness for its own sake’. Of course, any efforts to pursue this would soon be rejected, treated as a matter of perversity by those on whom they were inflicted. This simple demonstration clearly suggests the idea that seriously attempting to apply the scientific requirement in daily life would be disruptive of the conduct of practical affairs, rather than enhancing their effectiveness.

The Transition to Ethnomethodology

In this way, Garfinkel carried over a direct inheritance from Schutz’s work into ethnomethodology 3 through a series of similar demonstrations and the way of thinking he used them to illustrate.

Garfinkel did not, however, merely reiterate Schutz’s arguments. Rather, he took them in a quite different direction, one that meant withdrawal from further philosophical and theoretical argument to enable purportedly relentless occupation with empirical studies.

The idea that science is the general yardstick of rationality presupposes that there can be a general counter-position of ‘science’ with ‘common sense’. No one is going to reject the idea that science has sometimes overturned common sense conceptions (the same few examples – the Copernican revolution, Darwin etc. – are usually rolled out, and some, in the social studies, think that Marx and Freud have created comparable upheavals.) That science and common sense have sometimes conflicted does not, however, imply that the two must be systematically at odds, and Schutz’s arguments give reason for thinking they are not. Part of the reason for thinking that science must be the arbiter of rationality is the idea that admission to science’s body of knowledge is very strictly constrained. Supposedly, something is only admitted as a ‘scientific fact’ after it has gone through a stringent assessment, with experimental confirmation being the gold standard.

If this is so, then there is some reason for arguing that ‘common sense understandings’ cannot be seriously treated as expressing facts from a scientific point of view, for they have not satisfied those exceptional requirements. Thus, it is only scientific assessment which will establish whether any supposed fact, assumed by common sense is indeed a fact.

In society, these assumptions have led to recurrent derogation of common sense understandings. This played an important part in forming the frame of reference of (American) sociology’s methodology. Crudely, the supposition was that the assembly of a body of sociological knowledge would be achieved through strict compliance with the procedural rules of a science. Sociology could only know those facts that were established by the proper method, and hence could not depend upon any ‘common sense’ understandings since these were not, from the social science point of view, established facts. Hence the great preoccupation of post World War II sociology was with the securing of a method. Here Garfinkel discerned a serious difficulty, which he summarised as ‘the unsatisfied programmatic distinction between and substitutability of objective for indexical expressions’ (1967: 4).

Though wonderfully concise in identifying the difficulty, the point, in plainer English, was that research sociologists were attempting to do something that many significant philosophers - Husserl and Bertrand Russell amongst them – had recognised to be deeply problematical. They were – because their methods had to process data consisting of natural language materials - unavoidably involved in trying to capture the meaning of natural language through the application of stringent rules. In principle, the requirement was for formal rules to provide a way of systematically and exhaustively eliminating ‘indexical’ features from the use of language. This was, though, precisely the thing that the most powerful and ingenious logical systems had failed to accomplish, which meant that sociologists could only get on with their work if they were willing to be satisfied with much less than a thoroughly principled solution to this problem. Sociologists would have to settle for practical ways of coordinating the meaning of ordinary language (pervaded by ubiquitous indexical expressions) with the representation of that in their formal theoretical schemes.

In other words, sociological researchers continued to make extensive and unexamined use of indexical expressions (and of the common sense understandings in terms of which the sense of indexical expressions is defined). Thus, sociologists were not involved in progressive elimination of indexical expressions, but in a theoretically and methodologically undisciplined reliance on them. Such sociologists could no doubt suppose that they were continuing to rely on such expressions as only a temporary measure, but that did not remove or eliminate the problem. In a context where there are high ambitions for the use of formal systems of thought, the problem of ‘indexicals’ is not a trivial one, and it is possibly an unsolvable one.

The aim of substituting objective for indexical expressions was to institute the thorough purging of common sense understandings from the sociological corpus. But if Garfinkel’s argument was right, then this substitution remained only programmatic, something promised but far from realised, and certainly presently unsatisfiable. This meant that ‘common sense understandings’ continued to play an inappropriate and largely unacknowledged role in the context of dominant forms of sociological thinking and research. Common sense was officially condemned, and to the extent that common sense was thought about at all (rather than made unreflective use of), it was only as a methodological nuisance.

Schutz’s argument had been that common sense understandings make up social life. The implication is clear: common sense ought, therefore, to be a topic of sociological study, rather than solely an object of invidious comparison. The point was not to ‘defend’ common sense against science, but to bring it within the purview of sociological reflection and investigation. Garfinkel’s unprecedented move was to institute investigations into the ways in which common sense featured in the conduct of social affairs, to investigate how ‘common sense sociological reasoning’ is done (for it is important to recognise that if ‘sociological reasoning’ is a matter of reasoning about ‘how the society works’, then all members of society - not just professional sociologists - engage in sociological reasoning).

One of the great exemplars of sociological methodology was (and, to an extent remains) Emile Durkheim’s Suicide. The study was meant to demonstrate the necessity of sociology by showing that variations in suicide rates were due to social factors, proving sociology’s necessity by showing that the true causes of (variations in) suicide (rates) could only be identified through the kind of theoretical and methodological apparatus now being envisaged by Garfinkel’s contemporaries. It is no coincidence, then, that Garfinkel made an investigation of the ways in which suicidal deaths were actually investigated. The investigations were in service of the coroner’s office, not studies made by sociologists, and Garfinkel sought to show that the work of determining how someone died (whether suicidally or not) was pervaded with ‘sociological’ reasoning, with the investigating professional’s common sense understandings of how people come to die, of the kind of social circumstances that provoke suicide attempts, the standardised ways in which people act in preparation for suicidal acts,and the like.

The title ‘ethnomethodology’ identified the project of capturing the practical ways in which society’s members implement their common sense understandings in courses of sociological reasoning that are embedded in the conduct of their everyday affairs – reasoning about the ‘known’ social circumstances under which people die, kill themselves, are murdered is integral to official investigations of cause of death. The medium of Garfinkel’s innovative programme was therefore meant to focus extensively on the ways in which people talk in the course of their activities, the examination of ‘indexical expressions’ in their home environments, with the aim of understanding how discourse conducted on the basis of common sense understandings (and by means of natural language) actually ‘works’. Thus, the project could be summarised as studying the rational properties of indexical expressions.

Such a project was at odds with the major sociological projects in place, and remains so today (though the critical force of ethnomethodology’s implications contributed to the removal of the quest for formal theory from its then central place in sociology). Noting the role of common sense and natural language in the organisation of society’s affairs was not to be understood as a corrective observation that could be accommodated by simply adjusting the way in which sociology was generally done. That is why an independent enterprise – ethnomethodology – was needed. The title captures the nature of the project – ‘ethno’ is commonly used as a prefix to identify fields which study the indigenous practices of a community, and ‘methodology’ nominates the study as that of indigenous methodological practices, the multifarious ways in people establish ‘matters of fact’, the ways in which – as part and parcel of their organised affairs, as part and parcel of organising their affairs - they find things out, satisfy themselves that they have detected how ‘social reality’ is.

From ethnomethodology’s point of view, professional sociologists have no special or distinctive status compared to (other) members of society. Much of professional sociology’s reasoning itself relies upon ‘common sense understandings’ carried over from the sociologist’s own membership in the society. It is by presupposing or explicitly appealing to common sense understandings that they are able to make claims about ‘how the society works’ and about ‘what actually happened on this occasion’, and ‘what that person really meant by what he said’ that they are able to make assertions that are intelligible to (and plausible to) each other. For most sociologists, common sense understandings serve as a means of talking about other matters, a basis on which they can discuss, for example, whether there is inherent conflict between managers and workers on the shop floor, whether marriage is still a viable institution, what people’s increasing inclination to mark their body signifies and so forth. ‘Managers’, ‘workers’, ‘shop floor’, ‘marriage’ ‘instiutution’, and ‘body markings’ are all common sense typifications, originating in the natural language before being taken up by sociology. Far from being expunged from sociology’s discourse, they are readily used to define its topics. It is not ethnomethodology’s role to disparage those topics or to deny the validity of the discussions about them. Rather, the crucial point is that one cannot make such extensive use of common sense understandings without requiring that they remain the unexamined medium of one’s discourse. One cannot accompany the use of a mode of expression with a thorough reflection on how one is organising and using that mode of expression – an attempt at the latter would disrupt the former. Ethnomethodology’s imperative (to examine the organisation of common sense understandings and natural language use as its own topics) dislocates the possibility of making sociological studies of the usual kind.

Ethnomethodology’s earliest works were mainly devoted to inspecting the most commonplace and unremarkable of occurrences with a view to showing that ‘common sense sociological reasoning’ was an identifiable topic and exploring how it might be most effectively talked about. One of the most productive outgrowths of that early period was the exercise of ‘conversation analysis’ (Sacks, Schegloff, Jefferson, 1974), now an independent and flourishing enterprise. Conversation analysis (CA) sought a systematic and disciplined examination of the organisation of commonplace conversations, of the ways in which the most ordinary and unremarkable exchanges of talk are organised, to identify the indigenous methods whereby speakers of a language conduct their conversations. CA developed a very systematic method of examining talk which engendered a coherent and comprehensive account of the way parties to conversation do their talking, the account being centred on the conversationalist’s practical procedures for producing and distributing turns at talk.

Relatively Recent Developments

Garfinkel’s insistence that ethnomethodology make studies - rather than argue theory - has consistently involved an antipathy toward the use of any preconceived method. From his point of view, a preconceived method is likely to obscure the features of whatever is being studied, rather than reveal it. Though CA seemed like the first and finest fulfilment of Garfinkel’s project, more recently developments have become contentious (and Mike Lynch has been at the forefront of an argument from ethnomethodology and against CA - or, more technically, against ‘foundationalist CA’ - cf. Lynch (2000)). I think that part of this case can be summarised as follows: CA has now elaborated not exactly a preconceived method, but certainly one which is not brought to bear upon the phenomena, running similar risks to those associated with preconceived methods. Garfinkel’s ethnomethodology is meant to resist, not embrace, crystallisation into a fixed method.

Since the early studies, the main novel development in ethnomethodology has been the development of a programme of ‘studies of work’. The election of ‘work’ as a topic is partly a means of explicating the difference between ethnomethodology and social science more generally. Sociology features a huge number of studies that are, or could be, called studies of work. This topic is available to sociology because innumerable members of the society (in a society like ours) go to work each day, and devote several hours of that day to getting on which whatever work is theirs. One can search the standard literature on the sociology of work assiduously without finding much in the way of descriptions and analyses of the work that people actually do. One can find all sorts of other topics discussed that are associated with the workplace - what kinds of friendships people at work form, what kind of satisfactions they find in their work, whether and how far they are obedient to management imperatives inter alia – but one will not find much actual study of what people are doing when they are assembling television sets out of components, directing air traffic across UK air space, preparing and recording a TV advertisement, improvising jazz piano playing, etc.

Sociology (characteristically in Garfinkel’s view) relies on the fact that people at work organise their affairs so that the work gets done, but is simply satisfied to know that, somehow, a production line is up and running, somehow communications between pilots and air controllers achieve a safe and efficient distribution of craft within the most crowded airspace, somehow a film crew assemble the materials for, undertake the preparations of, and film and edit something suitable for broadcast, or that the pianist plays in a way which intricately integrates with other members of the group. And so on. If one is interested in how social order is made to happen, how people put their affairs together so that the world of everyday life turns out to be, as it extensively does, much the same today as it was yesterday and will be tomorrow, then one cannot really be satisfied that somehow these things get done. The fact that there is little if any study of what is included under that ‘somehow’ seems to represent an enormous absence from any attempt to understand how society works.

Ethnomethodology has, however, every reason to look into those matters, to open up the topic of what goes on at the work site, attempting to specify for exemplary cases, how – through their understandings – they arrange to carry through whatever work it is that they do. The thought is not, of course, that this problem arises only in respect to studies of work which are singled out only as an illustrative case.

The ‘studies of work’ emphasis has perhaps intensified an emphasis upon the specific and local character of organisation. Another reason why sociology generally cannot capture the things that ethnomethodology seeks out is because it is primarily motivated by the desire to generalise. From ethnomethodology’s point of view, it is apt to exhibit what Wittgenstein condemned as a disdain for the particular case. Certainly, the objective of generalisation requires extrication from detail, the analytical discounting of many specific, localised and detailed features. From the point of view of understanding how work gets done at the work site, however, it is an unavoidable feature of their work that doers must attend to and manage the specific, localised and detailed. Taking care of all practically necessary details to the extent required to finish off the job is, in key respects, what doing the job means. Thus, emphasising the fact that the nature of work can only be accessed through close examination of the details of work’s practice (since the details necessary to that practice cannot possibly be anticipated or imagined) is at the fore of many of Garfinkel’s recent pronouncements. This emphasis has been very influential upon the field of Computer Supported Co-operative Work (CSCW) and has informed recent thinking about the design of large-scale computer systems (programming and designing systems cannot be effective without attention to detail).

The ‘studies of work’ orientation has included a concern with the work of the sciences and of mathematics. In this area, Lynch has made a leading contribution with a meticulous study of laboratory work, and, subsequently, a now extensive collection of related studies. The interest in science and mathematics is in how scientific and mathematical work is done – ‘at the worksite’ stands for ‘in the laboratory’, amongst other things. The interest in recovering the (historical and contemporary) specifics of (some) scientific and mathematical works led to an involvement on Lynch’s part with another, largely independently originating, exercise called ‘social studies of science’ or ‘the sociology of scientific knowledge’ (SSK). This exercise – much influenced initially by Thomas Kuhn, who later vigorously dissociated himself from it – was itself concerned that the philosophy of science, but showed little knowledge of how science was done by working scientists. The preconceptions of philosophy of science were to be put right by empirical evidence, and therefore a natural focus of SSK’s attention was on showing that scientific – mainly laboratory – practice did not conform to the picture of scientific method as portrayed in either philosophy of science texts or in the methods’ sections of scientific reports. Though the investigation of scientific and mathematical practice might be a matter of common cause, the interests of SSK and ethnomethodology are also significantly divergent. Lynch has been a dissident within Social Studies of Science, often in direct dispute with other participants (who are often in dispute amongst themselves). 4 In his criticism of CA and in his methodological examination of Scientific Practice and Ordinary Action, Lynch has sought to make critical use of what he thinks can be learned from the study of practical scientific work. It is in relation to the significance to these results (especially to the evaluation of Schutz) that the disagreement above breaks out.

For my part, I do not think that Schutz’s account of science and its place in relation to the world of daily life is – if I have reproduced it correctly – significantly affected by work in SSK. I am not aware of studies that show that semantic clarity and distinctness are not held at a premium in the natural sciences (the social sciences are a whole other matter) or that scientific work is not oriented to, and structured in terms of, an accumulated disciplinary corpus of currently accepted understandings. But this is not the place, and there is certainly not the space, to elaborate on the nature of SSK or to argue about the relation between what it claims to show and what might actually be licensed by its studies.

I hope the latter part of this paper explains why ethnomethodology claims to be an asymmetric, alternate and incommensurable sociology. After all, its central preoccupation with the study of practical sociological reasoning in context is incongruous with the major projects of contemporary sociology (wherein ethnomethodology continues a rather marginalised existence).


Campbell, Colin (1996) The Myth of Social Action, Cambridge, Cambridge UP

Garfinkel, Harold (1967) Studies in Ethnomethodology, New Jersey, Prentice-Hall, 1967

----------------------(editor) Ethnomethodological Studies of Work, London, Routledge

---------------------- (2002) Ethnomethodology’s Program (edited by Ann Rawls), Lanham MD, Rowman and Littlefield

Golinski, Jan (1998) Making Natural Knowledge, Cambridge, Cambridge University Press.

Kuhn, Thomas (1962) The Structure of Scientific Revolutions, Chicago, Chicago University Press

Lynch, Michael (1985) Art and Artifact in a Scientific Laboratory, London, Routledge

--------------------(1996) Scientific Practice and Ordinary Action, Cambridge, Cambridge University Press

--------------------(2000) ‘The ethnomethodological foundations of conversation analysis’, Text, 20 (4): 517 – 32

Sacks, Havey, Emmanuel Schegloff and Gail Jefferson, (1974) Á simplest systematics for the organization of turn-taking in ordinary conversation’, Language, 50 (4) :696-735

Pickering, Andrew (editor) 1992, Science as Practice and Culture, Chicago, University of Chicago

Schutz, Alfred (1962) Collected Papers Volume 1 (edited by Maurice Natonson, The Hague, Martinus Nijhoiff

----------------- (1964) Collected Papers Volumes 2 (edited by Arvid Broderson), The Hague, Martinus Nijhoiff

------------------(1967) The Phenomenology of the Social World (translated by George Walsh and Frederic Lenhert), Evanston, Northwestern University Press, (first published Vienna, 1932)

Zammito, John H. (2004) A Nice Derangement of Epistemes, Chicago: University of Chicago Press


1. For review of some of the ideas, topics and figures in this field see Pickering (1992), Zammito (2004) and Golinski (1998).

2. The subsequent expression draws mainly on Schutz’s ‘Common-sense and scientific interpretations of human action’ (1962: 3-47); ‘The social world and the theory of social action’ (1964: 3-62); and ‘The problem of rationality in the social world’ (1964: 64-90).

3. This section, up to ‘Relatively recent developments’, draws on Garfinkel (1967: 1-18)

4. See Pickering (1994) for Lynch’s debate with David Bloor, one of the pioneers of the contemporary sociology of scientific knowledge, along with other arguments between other leading figures in the field.