Reading guide to: Heritage, J 'Ethnomethodology', in Giddens, A and Turner, J (eds) (1987) Social Theory Today,Cambridge: Polity Press.
Garfinkel's work has always been controversial, partly because his own writing is compressed, and it does not foreground 'classical sociological reference points' (224). It is really best understood as aimed at classical sociological concerns, including the nature of social action, intersubjectivity, and the social constitution of knowledge. Garfinkel was able to effectively criticise the dominant work of Parsons, and move towards an abandonment of a concern with motivations as explanations of social action. Instead, the interest was to be directed to 'a procedural approach', activities by which actors makes sense of every day action. Because this is closely connected to the ways in which they explain, or 'account' for what has happened, there is a need to look at socially available knowledge as a necessary part of social action.
Garfinkel criticises Parsons.
Parsons had gradually developed a whole system to explain the connections between action, organisations, and the wider society. The assumptions made were that human beings act positively to realise their goals, but that they also need to achieve some social regulation of these actions in order to avoid chaos or the predominance of 'force and fraud' (227). Parsons stressed the role of the central value system to co-ordinate such activity, and suggested that the effects of this value system could be found at a number of levels [see file]. Interestingly, those critics that stressed the role of social conflict also shared this view of the central importance of social motivations. Garfinkel argued that a major omission in this kind of sociology is a consideration of the actor's own knowledge of social interactions. Parsons assumed that scientific rational action was some kind of norm: he assumed this partly in order to defend sociology as the scientific study of action. Fully subjective action, which did not follow scientific rational procedures, could be discounted. Heritage also suggests that the actual motivations and knowledge of the actor were also suspected -- it helped people obey social constraints if they did not fully understand the principles on which they acted: 'moral values are to be an effective prophylactic against... chaos... [for full conscious understanding of social principles] could give rise to Machiavellian calculation which... would undermine the moral constitution of society and to leave social order dependent on unstable coalitions of interest' (229). [ Seems to me this is precisely where ethnometh leads...]. For both of these reasons, actors were left as having a very limited understanding of social interaction -- they were '"judgemental dopes"' for Garfinkel.
Garfinkel reads Schutz
Schutz insisted that social knowledge was an integral part of social action, and had investigated the common-sense categories of mundane actions. It was essential that social scientists investigate these rather than impose some theoretical model instead. Heritage summarises Schutz's main insights (230 -- 232):
(1) Ordinary actors take for granted the naturalness and objectivity of the social world, which they construct using mechanisms of 'subjective time' [see file].
(2) Ordinary actors use their social knowledge to build ideal types of others, of varying degrees of explicitness.
(3) Ordinary actors understand each other through the 'general thesis of the reciprocity of perspectives', which assumes that experiences of others are 'identical for all practical purposes' [Heritage, quoting Schutz, page 230].
(4) Commonsense knowledge is not organised systematically or consistently, and thus action does not correspond to the models in scientific rationality.
Indeed, Heritage argues, the complexities of social life could not take place if action did so correspond. This leads Garfinkel to abandon the attempt to model action on a scientific basis. The way is open for more concrete study of 'practical commonsense reasoning in mundane situations of action' (231).
Garfinkel develops his programme
Analysts need to study such mundane actions, without privileging sociological categories. They are to do so with 'ethnomethodological indifference', which is 'studying the systematic properties that practical reasoning and practical action while refraining from judgements which have the effect of endorsing or undermining them' (231). [A naive commitment of value freedom, for me, which hopes to avoid the effects of the analyst's values by simply striving to ignore them].
Garfinkel's Studies in Ethnomethodology offered two main innovations. The 'breaching experiments' were designed to see what would happen if the general notion of reciprocity of perspectives were questioned. Secondly, a number of experiments were designed to show how ordinary actors actually did try to understand and explain each other. These two approaches enabled research on the emerging new focus for ethnomethodology -- the 'cognitive problem of order' as an element of social action.
Example 1 -- breaching experiments
The idea here is to experiment with ordinary social interactions in order to highlight the processes that are at work in rendering them 'normal'. Ordinary situations are to be disrupted. [Note that Garfinkel was well aware that to make social interactions 'senseless' was to produce 'anxiety, shame, guilt and indignation' in the innocents who were not in on the experiment -- Garfinkel's words cited in Heritage page 233]:
(1) A game can be easily disrupted by having one of the players deliberately break the rules -- a game of tick tack toe (noughts and crosses) was disrupted when an experimenter 'erased the mark [of their opponent], moved it to another cell, and then made their own move while avoiding any indication that something unusual was being done' (234). After 250 trials [!] 'over 75 per cent actively objected to [such action] or demanded an explanation of it'. In other words, 'The experiment showed decisively that the discrepant behaviours motivated immediate attempt at normalization' (234).
(2) Garfinkel also realised that this could be extended to other social actions, resulting in the same 'kind of bewilderment, anger and vigorous attempts to restore the situation' (234). Thus he asked students to demand clarification of everyday actions -- e.g. to demand what people meant by sentences such as 'I had a flat tyre'.
(3) Garfinkel apparently realised the almost limitless possibility of such disruptions in producing 'really nasty surprises... [as in] standing very, very close to a person while otherwise maintaining an innocuous conversation... [or] saying "hello" at the termination of a conversation'. It seems that he inferred that 'all actions... may have a constitutive structure, and that perhaps it is the threat to the normative order of events, as such that is the critical variable in evoking indignation' [original punctuation, taken from Studies in Ethnomethodology, quoted in Heritage, page 235].
Heritage believes that these experiments tell us something important about action itself, and justify the emphasis on procedures. Apparently, analysing the production of action enables the the subjective elements to be analysed 'from a social science perspective', since they 'lie behind actions' and are 'available to the actors by virtue of a combination of contextual knowledge and their tacit grasp of the procedural structure of their own activities' (235).
[There is just so much to criticise here. The claim that subjective intentions can be best studied through concrete action is a classic behaviourist one, which assumes that behaviours are somehow transparent indicators of something more mysterious lying behind them -- this is so in practice only by employing unstated interpretive procedures, of course, which ethnomethodologists cannot discuss because they are being 'indifferent'. This might be seen as a 'scientific' approach only by making some implicit contrast with other ways of grasping underlying motivations. There seems to be a classic positivistic intention in pursuing this new approach. I also find it very hard to warm to the juvenile 'breaching experiments' which treat the innocent subjects involved as mere raw materials -- worse than treating them as judgemental dopes in my view -- and seem to take delight in the power relations that produced indignation and anger in them, and quiet triumph in those in the know. So much for 'indifference' again -- it is the indifference of the biologist for the laboratory rat or of the police interrogator for the rights of the naive suspect].
Example 2 -- establishing social order
Actions take place within a context, and knowledge of this context is an important 'resource through which they are understood' (235). Thus, in one example in Garfinkel, the conversations between husbands and wives make no sense unless the context is grasped, 'a background of matters that were assumed to be known in common' (236). Another experiment to illustrate this involved students pretending to be boarders in their own homes, suspending these background understandings -- once more this produced 'quarrelling, bickering and hostile motivations', in Garfinkel's own words (Heritage: 236). Studies of a psychiatric clinic revealed that contextual knowledge was needed in order to operate procedures, especially when coding ambiguous cases.
Such contexts have been 'imputed', Heritage tells us [by ethnomethodologists, presumably?], but one example in Garfinkel apparently shows how contexts are invoked by ordinary actors. This is the case study of 'the documentary method of interpretation', [on which there is a separate file on this website -- Heritage discusses it on pages 237 - 8]. Heritage shows that 'an indefinitely large range of matters'if are involved in such documentation, and that the understandings of actors themselves 'were provisional, "loose" and subject to revision' (238), that is not at all governed by tight rules [and so must the ethnomethodologists' understanding be similarly provisional? ] Instead, actors apply rules using 'ad hoc devices such as "unless", "etc" and "let it pass"' (238). [Is it as ad hoc is this? Are there no rules, conventions, habits or social purposes 'lying behind' these apparently ad hoc devices? Are not ethnomethodologists attempting to understand these ad-hoc devices in terms of some underlying principles?].
Heritage points out that the constant striving for order in the use of documentary methods seems at odds with the impatience and annoyance of those subjected to the breaching experiments, who 'responded almost immediately with outrage and hostility' (239). Apparently, these differences depend on how the experimenters' behaviour were interpreted -- but at least this shows the continuing effort to make sense: 'the ... predominantly hostile reactions... [arose from a view]... of the experimenters' behaviour as motivated by presently undisclosed -- though probably disagreeable -- intentions' (239). What this shows is the centrality of 'procedural trust'-- even when actors encounter a breakdown of understandings, they still try to interpret such a breakdown. In this way, interpretive procedures are always involved, in any action whatsoever -- 'There is thus no uncategorisable action' (240). [And ethnomethodology becomes queen of the sciences].
Society as rule governed
The number of theorists have seen action as a matter of following moral rules, as did Parsons [and including Winch -- see file]. These rules have to be internalised through socialisation. But no place is left for the reasoning of actors, as we have seen, or the ways in which they use their knowledge to interpret, or 'make accountable' the actions of themselves and others. Once you allow for these ordinary understandings, however:
(1) The context of action ceases to be set by external factors, and is instead interpreted by actors themselves, as part of an ongoing process: '"Action" and "context" are mutually elaborative and mutually determinative elements in a simultaneous equation that actors are continually solving and re-solving to determine the nature of the events in which they are placed' (242).
(2) It follows that a small number of central values can never be simply operated by actors. Instead there is 'an indefinitely large number of differentiable situations of action' [and thus no limit to the ways in which these situations can be described, Heritage tells us in a note (242)].No simple set of values can govern or determine this wide variety of contexts and actions. The participants themselves must decide in which rules to apply, according to the situations in which they find themselves. Only the actors themselves, and not an outside observer, can grasp this process of application [although ethnomethodologists claim to be able to describe sequences involving application].
Instead of normative expectations determining action, they tend rather to emerge from action, as part of a process or sequence. Actors come to a situation with any implicit or tacit grasp of how to behave, but modify their behaviour as the action proceeds -- in this way, any social setting can be viewed as 'self-organising with respect to the intelligible character of its own appearances' (Heritage, page 244, quoting Garfinkel). This is what the breaching experiments reveal, of course -- most actors are able to see that normative expectations had been breached, and are able to interpret these breaches by reference to another set of normative expectations [note the apparent emphasis on functional re-establishment of equilibrium]. [This is one problem with ethnomethodology for Bourdieu -- the norms seem to have to be renegotiated and recognised every time actors meet. Bourdieu thinks that they will have been selected and sedimented in the unconscious habitus long before] .
It follows that the norms and values are better seen as resources rather than as binding commitments, used to understand action, make it intelligible or accountable. This introduces both a reflexive and a calculative element to social action. [Heritage again drifts in the direction of functional order here in arguing that these are likely to underpin the inhibition of deviancy and motivate 'compliant conduct' (248). I would like to see him discuss the emergence of calculative reason as corrosive of social trust, the growth of cynicism and cheating which he hinted at earlier. He thinks this unlikely, but never explains why -- this may be the price paid for ignoring issues of structure and power].
Garfinkel is not good at explaining why norms persist, especially as they are no longer seen as heavily institutionalised, as in Parsons. Heritage suggests that the frequent need to make actions accountable, especially deviant ones, helps preserve norms cognitively [but it is still not clear where they live, especially if we do not allow for an unconscious]. There seems to be a constant need to interpret, negotiate and account for both factual and moral matters. This constant 'accommodative work' seems to imply a constant rediscovery, with no real memory of past events to go on --'[the context of action] is always recognised, as Garfinkel elsewhere puts it, for "another first time"' (247).
Accountability in practice
[Despite the difficulties mentioned above, curiously ethnomethodology advocates empirical study]. Ordinary descriptive accounts are not adequate to uncover practices of accountability. Nor can 'the analysis of verbal accounts of action... in any way substitute for the analysis of action itself' (248). This means that non-verbal actions can be studied, such as queuing up. Happily, activities like this appear to possess an 'inherent or "incarnate" intelligibility' (249). However, at the same time, Garfinkel apparently believes that that did is not possible to develop some neutral description outside ordinary understandings. An active interpretation is required instead, realising that explanations are never independent of the contexts in which they are employed -- they are 'indexical' (249). [I'm still not sure whether this applies to sociological descriptions as well].
Heritage tells us that the term indexical means that expressions 'require contextual knowledge in order to recover their referents' (249) Garfinkel thinks that all expressions must be covered by this link to context -- to tell the difference between a description and 'an irony, a joke or a metaphor' (249). Radical indexicality like this is a useful resource for ordinary sense making. Indeed, indexical expressions are actually embodied in interaction, contributing to the setting, and used in 'a range of ways to manage ordinary settings of activity' (250). Language therefore becomes far more than a transparent descriptive recording, but an activity in its own right. Heritage observes that 'this treatment generates more problems than it resolves. This is entirely to the good' (250), partly because it permits empirical study [and a limitless research programme, and it permits useful criticism of rival approaches].
One theme began by examining the processes of typification, as understood in Schutz. Thus underlying assumptions and presuppositions of various professional frameworks were examined, especially in the field of deviance or in more general 'people processing' organisations. Unlike labelling theory, the technical formal aspects of categorisation were of most interest. Thus Sudnow studied the practices of lawyers and their typifications of offenders and offences, and how they affected plea-bargaining (251 - 2). Other studies showed how bureaucratic rules were actually applied in a number of contextual ways, that this might often be a source of 'discretionary power' [an interesting departure from technicality and indifference?], and that all rules required active interpretation even where apparently clear cut (252 - 3).
Garfinkel's own study of a psychiatric clinic's records discovered that only very vague records were kept, enabling a kind of strategic interpretation to justify the power of a therapeutic regime. 'The absence of detail... served as a defensive resource by ensuring that the records could only be competently read by entitled personnel' (254). [Garfinkel on the verge of discovering power here?].
The use of official statistics are clearly open to criticism on similar grounds, since records are themselves the result of active interpretation and the priorities of the collecting agencies. Cicourel showed this with the construction of records of juvenile delinquency and the association with various background factors such as broken homes (254). Police records too were actively constructed. In this way, 'as circular process' connected assumptions and events. Studies of suicide statistics make the same point -- that knowledge of background factors involved in suicides feed forward into coroners' decisions to classify some acts as suicides but not others (255).
Heritage suggests that these studies make excellent critical points, but that there are more positive applications too: processes of normalization are widespread in organisations and enable them to do their work. This offers new possibilities to understand the social constitution of knowledge in institutions themselves. [It also reminds me of Foucault's insistence that power/ knowledge has a positive role in permitting organisations to function].
This is considered as a particularly successful aspect of ethnomethodology. It is 'resolutely empirical' (256), focusing on actual concrete actions and sequences of them. [So the full empiricist claims are clear here?].
Schegloff and Sacks are credited with developing the approach, which seems to be based on the assumption that the order found in conversations reflects some deeper order among members of society. There was also a methodological interest to see if ordinary conversations could be formally described. There is a tradition of looking at ordinary ['naturally occurring'] conversations rather than institutionalised forms, partly because ordinary forms are seen as 'primary', a major vehicle of socialisation and a 'benchmark' for more formal kinds (257).
Thus the research appears 'naturalistic', specific, and 'uncontaminated by interventions from the researcher' [Heritage operationalises this value freedom in a note on page 258 -- avoiding manipulation of behaviour, pre-coded schedules, deliberate interviews and so on. This is still oblivious to the effects of interests, or, indeed, the processes of making sense describes as applying to ordinary people]. Great detail is to be sought [how much? -- see the file on Sacks]. The idea is to uncover social competencies which sustain interaction, to understand any 'institutionalised procedures' [rules], and to reveal the effects of context.
There are two major dimensions: (a) papers on 'particular lexical formulations and referring expressions' (258 - 9), and (b) the 'sequential organisation of interaction', such as 'conversational turn taking, and the problems of entry into and exit from conversations (259) [see file for some examples] . The latter are particularly commended in displaying 'new standards and rigour and comprehensiveness in the study of social interaction... [and in introducing]... further analytical concepts for the study of interaction' (259).
These new concepts include the 'adjacency pair', which indicates that certain activities are commonly paired, so that production of the first one implies or expects production of the second. This helps understand common patterns and expectations in interaction, and begins to point to a basic competency in conversation, the ability to recognise such adjacency. There is also a possible 'primary mechanism' involved, since replying with the second part implies 'the public understanding of the prior utterance' (259), and thus could indicate some initial shared understanding. This helps justify Garfinkel's view that order and understanding are somehow implicit or inherent. Adjacency pairing has been applied to a wide range of behaviour since.
More recent developments include interaction in institutional settings, such as classrooms, courtrooms and so on. These are more 'piecemeal', since institutions narrow the possible range of options. This shows the operation of conventions again. The range and type of participation is also limited . These conventions themselves as sometimes discussed and justified in interesting ways (see pages 260 - 1). Studies of turn-taking systems have also flourished, and they take specific forms in different institutions, and indeed might explain the operations of these institutions. [And there is reference to the unexplained concept of power again on page 261]. Even these small piecemeal studies are cumulative, though, demonstrating the power of the ethnomethodological viewpoint.
Studies of work
This is more recent, and tends to focus on the activities of scientists and mathematicians, although others are not ruled out. The studies focus on 'specific competencies' required in an occupation, and offer a detailed analysis instead of the more common 'external' analyses, which tend to focus on abstract formal descriptions of group functioning, or whatever. Heritage expects that these studies will be critical of official statements by practitioners, and will require detailed empirical study instead. They might begin by focusing on what practitioners themselves define as their relevant work activities and competencies, and how they actually perform, how they display characteristic 'logical and reasoned' occupational action (263). A range of techniques including ethnographic analysis, textual analyses and the analysis of conversations will be required to grasp this multi-faceted occupational action.
Garfinkel suggests that research must always display an 'unique adequacy requirement' (Heritage, 264), relating to the highly specific nature of the occupation, much of which might be opaque to conventional sociological study. This means that any researcher should be competent practitioners themselves [an abandonment of theoretical indifference is implied here as well? Researchers will go native?]. Heritage believes that a neutral form of description will maintain 'ethnomethodological indifference', presenting scientists and their activities 'without celebration, irony or relativisation and without transforming their technical features or downgrading their achievements' (264). Studies like this also bring even scientists within sociological investigation.
This section ends with Heritage making some strangely positivised claims for ethnomethodological study: they focus on 'specific, discoverable material practices', offering 'detailed descriptions... which, like observations in the natural sciences, can be reproduced, checked, evaluated and form the basis for naturalistic study and conjecture' (265).
Garfinkel has transformed the theory of action by studying practical reasoning instead of imposing some scientific rational model as a standard. Context becomes important. Description supplants questions of ultimate validity or effectiveness' (265). Research has headed towards naturalistic study. Because Garfinkel is interested in the public accountability of action, it helps focus on understandings which are 'socially shared and used'. It is the procedures of action which bridge 'the gap between cognition and action, both practically for the actors and theoretically for the social scientist' (266). [We still lack any description of how sociologists or ethnomethodologists actually do this themselves, however -- why no detailed studies of the practical, occupational work of ethnomethodologists?].
These procedures are fundamental to social action, social organisation and into subjective understanding. It is this that has spawned so many detailed study since [which implies a purely rational commitment to ethnomethodology?]. The project turns away from 'prematurely theorised conceptions of the social world and towards the empirical phenomena of social activity in all their richness and diversity' (266). This indicates 'Garfinkel's continuing stand against "all attempts, no matter how thoughtful, to specify an examinable practice by detailing a generality"' (266). [This reminds me of the curious and internally contgradictory 'aversion against universal' in post-modernism, ably criticised by Honneth].