RAWLS: Garfinkel's Conception of Time

This outstanding article by Anne Warfield Rawls (Bentley College), appearing in Time and Society, examines Harold Garfinkel's conception of time.   Garfinkel is best known as the "father" of ethnomethodology.   I apologize for how it transferred from the electronic journal to this format, in which a few initial sentences are omitted.  Enough great stuff though is here to make it worth examining!

ABSTRACT. Garfinkel articulates a significant conception of time – as situated and sequential – that works in tandem with his rendering of social order in terms of situated practices. However, because his treatment of the actor, action, group and time in situated terms differs significantly from more conventional theoretical approaches, critics have often mistakenly interpreted Garfinkel as focused on the individual, and indifferent to the significance of social structures, and their relations through time. What Garfinkel focuses on are practices, not individuals, and he argues that practices constitute the essential foundations of social structure. Given this view, the time dimension of practice is the significant time dimension for any study of communication and/or social order, which are both constituted in and through situations defined by mutual orientation toward practice. KEY WORDS • ethnomethodology • Garfinkel • practice • time • trust 

164 TIME & SOCIETY 14(2/3)
things could be different – or to make historical comparisons – let alone attend to the transformation from caste to class in relation to the economic and politi- cal changes in modern societies. Consequently, any theory that does not have what is considered to be an adequate conception of time is said to be conserva- tive – that is, focused only on what is, at a single point in time, and therefore unable to deal critically with questions of social justice and inequality, that is, with questions of what ought to be.

Given Garfinkel’s commitment to the study of situated practice, interaction that inhabits small bits of present time and local space, most scholars have assumed that Garfinkel has no adequate conception of time and therefore that he must have confused ought with is or, worse yet, have failed to consider the ought problem altogether.
Furthermore, because detailed descriptions of a single practice seem to have no relevance to other practices, in other times and locations, descriptions of such practices are thought to lack generalizability, the main criteria for signifi- cance employed by conventional social science. As a result, with regard to the so-called larger historical issues, Garfinkel’s approach has generally been dismissed – even attacked – as trivial (Coser, 1975).

This is a mistake. Not only does Garfinkel have a conception of time. But, that conception is relevant to the continually developing issues of justice and inequality that confront modernity in ways that more traditional notions of time have not kept pace with.1 A significant conception of time – one that might be referred to as the ‘heartbeat’ of a modern practice-based social order – stands at the very center of Garfinkel’s ethnomethodology. Conventional conceptions of time are about social order, but they are not of it in this deeply intrinsic sense.

Why then the perception that Garfinkel has no conception of time?

The confusion has arisen primarily because Garfinkel’s position involves transformations of several central ideas – including the conception of time – that make his treatment of these ideas hard to see from a conventional perspective. Because time is something different for Garfinkel, most scholars cannot see that he is addressing time at all. But, it also follows that in so far as his idea of time is completely different – and yet adequate to his subject – conventional criti- cisms of his position are irrelevant. They are irrelevant, rather than wrong, because Garfinkel’s transformations so completely change the field of argument that the criticisms simply lose their meaning when applied to the domain of social phenomena that he addresses.

Garfinkel is still talking about the essential issues – but the issues themselves have changed as society has changed, and the way in which Garfinkel takes them up is consequently unrecognizable to more traditional scholars. Garfinkel has insisted on allowing social practices themselves to determine what can be said about them: allowing empirical detail precedence over abstraction. New ways of talking about social orders are necessary. But, these changes in both
peer-00570994, version 1 - 1 Mar 2011

language and approach have been resisted by more conventional thinkers. It is a case of what Thomas Kuhn (1958) would have called ‘normal’ versus ‘revolu- tionary’ science. And, in keeping with classic cases of revolutionary science, Garfinkel’s influence has increased as resistance to his position has also increased. Even those who, in a friendly way, adapt something in the way of ‘methods’ from Garfinkel – not his point in any case – have generally felt the need to marry his approach to some conventional theoretical abstraction.

This is always a mistake. Garfinkel’s position is neither a method nor a theory. It is a complete reconceptualization of how the coherence of the social is made, coupled with an elaboration of the implications for studying it. It involves seeing things differently and thinking about them differently. But, the methods – referred to by ethnomethodology – are not research methods. They are the methods that ordinary persons use to create the recognizable order of the social world they live in – ethno (folk)-methods.

Garfinkel’s Transformations: Actor, Situation, Group and Time

In approaching the problem of intelligibility and the achievement of coherence in modern societies Garfinkel made four essential transformations that make his position incommensurate with existing perspectives on most points.2 These four transformations involve the actor, situation, group and time, and each trans- formation stands completely at odds with conventional social thought. Because these transformations are essential to his position, and work together in tandem, misunderstanding any one of them leads to a misunderstanding of the whole.3

With regard to the actor, Garfinkel turned his focus from whole biological persons, whose demographic characteristics follow them from place to place, toward what he referred to as the ‘actor.’ Although persons as such certainly exist, in Garfinkel’s sense, they do not act. In a situated encounter the only thing about a person that matters is their competence to participate and their commit- ment to the ongoing situated practices and situated identities. Therefore, in studying interaction, Garfinkel is concerned only with actors – performers of situated identities. Because it is actors who perform identities and produce the recognizable orders of situations, it is necessary to study the processes and problems involved for actors in producing and maintaining situated identities in particular local settings. To do otherwise, Garfinkel (1948/2005) argues, is to produce a series of paradoxes.4 Thus, the conventional focus on demographic characteristics, which preoccupies most social science, is, for Garfinkel, at best a distraction – and at worst, completely obscures the issues involved in the production of social order; as do the methodological issues of generalizability and reliability – tied to the collection of demographic data – which force researchers away from details and toward abstractions.5
peer-00570994, version 1 - 1 Mar 2011

166 TIME & SOCIETY 14(2/3)
Similarly, Garfinkel has no interest in locations as physical places with con- crete historical characteristics. It is not physical space that gives coherence to what we might still want to call the spacial characteristics of the places in which actors act. For Garfinkel, places are situations whose coherence is given by the practices which constitute them as situations of a particular sort. Practices have relevant histories that may be of interest. But, a place can be constituted by one practice at one time and a different practice at another. Practices define the situation and the identities that occur within it – not the location.

The same is true for physical objects, which Garfinkel (2002) refers to as ‘oriented objects’ and ‘phenomenal fields’. ‘Things’ are made to appear as objects of a sort only when oriented toward in specifiable ways in the context of a practice. In the context of a different practice, objects will be made to appear differently and will have different developing horizons.

Like actors and situations, social groups are not defined for Garfinkel by the demographic characteristics of members, their kinship relations, or the cultural practices and beliefs held by people across time, space and political con- nections. Groups, are defined by the practices that constitute their membership. That is, when a set of persons are mutually engaged in enacting a situated practice together in such a way that those practices define the relevant identities and interactional possibilities for all involved – constituting them as a recogniz- able and competent membership – they constitute a group.

Given these transformations of person, place and group, conventional con- ceptions of time – conceptions designed to compare the actions and beliefs of demographic persons and groups across physical spaces and cultural abstrac- tions – would be of no use. This is a very important point. If time, as con- ventionally conceived, arose as a way of making comparisons between demo- graphic groups, physical locations and demographic persons, time in this sense has no relevance for the study of the situated actors and situated practices which inhabit the foreground in the situated landscape of modern life. se elements of actors, situa- tions, interconnections between actors, and the situated order of practices, that are constitutive of mutually intelligible practice in contexts of situated inter- action. Only in so far as time is relevant to the production of order – in just these terms – will Garfinkel concern himself with time.

This orientation toward time as a situated feature of interaction does not ignore the ‘larger’ issues of inequality and power relations. It is in this inter- actional foreground that essential issues of power and exclusion play out in modern life. To the extent that mutually intelligible practice is necessary for the integrity of both self and interaction – and ‘trust’, in Garfinkel’s sense, is necessary for the production of mutually intelligible practices – situated action always has direct implications for justice and inequality.
Because practices require mutuality, they not only make egalitarian reciproc-
peer-00570994, version 1 - 1 Mar 2011

ity possible, but also work to exclude those who cannot display the qualities required for trust, creating new forms of exclusion and inequality that are mani- fest only in situated action and not in the structures and beliefs of the ‘larger’ social structure. These forms of exclusion elude formal means of control, including legislation and the formal measures of compliance that it depends on.

Therefore, focusing only on a combination of attitudes, beliefs and values, and structural inequalities, entirely misses these new, and increasingly preva- lent, forms of exclusion and inequality. This is one of the things that has made race and gender discrimination so elusive from a conventional research per- spective and enabled persons, governments, and corporations to continue practices of exclusion with impunity in a context of democratic law.6
The key to understanding both inequality and the possibility of egalitarian exchange, in any practice-based social context, lies in the relationship between situated orders and mutual intelligibility. Order and recognizability can only be achieved if actors share some set of expectations – or practices – that provide guidelines for how to use sounds and movements to create mutual understand- ing in particular contexts of use.

The underlying question is: what do actors need to do to make themselves understood by others – particularly others they know little about and have little in common with? How do social events come to be ordered – as an ongoing matter available for mutual inspection – in such a way that order is mutually available and constitutive of the coherence of the events? And, how is Time relevant to this endeavor – given the need to produce sounds and movements in ways that convey a recognizable meaning to others?

Time as a Constitutive Relationship between Parts of Situated Interaction

Garfinkel’s conception of time, then, will need to be relevant to the constitutive requirements of practices. How is something, that might be called time, relevant to an actor’s production of identities and situated practices recognizably for others? What does it mean to say that time is constitutive of order and meaning? The general perception has been that there is relatively little of interest about time in this regard. But, that is because most of the thinking about this issue has made use of conventional conceptions of time.

Just as Garfinkel’s position involves transformations of the conception of person, place, and demographic group, it also involves a transformation of the conception of Time. Rather than being a measure of something standing between different events, Garfinkel treats time as a relationship between the parts of a single event – a relationship that is constitutive of the organization, and hence the meaning, of those parts. Situated time as an organizing feature of interaction cannot be the time of groups, places, and demographic persons. It
peer-00570994, version 1 - 1 Mar 2011

168 TIME & SOCIETY 14(2/3)
cannot be a time that follows persons from place to place, or one that tracks their membership in demographic groups, or the transformation of demographic groups over time.
Time for Garfinkel is a feature of situated interactional practices – and the situated actors who inhabit and construct the coherence of these fragile and temporary spaces – and as such, time turns out to play an essential role. It is time as a constitutive relationship between the parts of a developing interaction, rather than time as a measure of a relationship to a completed act that occurred in some other place and time.
Most people, even when they insist that time is a social construction, never- theless think of it either in terms of an historical time line existing outside of and independently of situations, or as an inner duree experienced by each individual alone. Garfinkel’s transformations are designed to take into account the situated and socially produced character of all aspects of social orders – including time. As such he avoids both the positivism and the solipsism of the conventional view.
The coherences of interactional time are intrinsic and constitutive features of the situations in which they are produced, and which time is intrinsic to the production of. But this does not make them unique and ‘non-generalizable’. Rather, it means that they are available to all participating members as witness- able productions. Mutual intelligibility requires that these coherences must be available to participants – regardless of who they are, or what they believe – simply on the basis of their competent and attentive (trustworthy) participation in practices.7 Thus, constitutive features of time are distributed equally across all competent and trustworthy members of the situated practice in which they are produced.
This is one reason why generalizability is not a requirement for the study of sequential orders. The coherence of a practice does not have to appear anywhere else in the world in order to be a relevant organizational feature of this practice here and now. Constitutive features of practice may be unique, and nevertheless, be required for the achievement of recognizability in a given situated practice.
Modern societies are characterized by a situated form of interaction between strangers and mere acquaintances that requires persons to orient directly toward the production of meaning and order on the spot. It is not necessary to know anything about the person, their beliefs and values, or their demographic characteristics, to do this. In fact, this is what makes mutual intelligibility possi- ble in a democratic context. Furthermore, such interactions often involve the mastery of practices that integrate technology with human actors in rapidly changing ways. What persons bring with them in terms of personal characteris- tics must be set aside in order to participate successfully in the practice at hand.8
peer-00570994, version 1 - 1 Mar 2011

Situated Time and the Other-Oriented Character
The new modern form of character that results from the increasing importance of situatedness in modern highly differentiated societies has been referred to by David Riesman in The Lonely Crowd (1950) as ‘other oriented’, because actors need to focus on the situated others and the situated requirements of the inter- action, rather than on the principled requirements of traditions and beliefs. The positive consequence is that interactional practice becomes more adaptable, and in a densely populated multicultural society of strangers this form of adaptation makes communicative reciprocity between strangers possible. But Reisman was concerned that in the process of coming to focus more on the here and now of interactional practice, the modern character might have become shallow, because it is no longer focused on principles and beliefs.
This concern, which many others share with Reisman, associates morality with traditional beliefs and principles, the conceptions of time they generate, and the ability to make comparisons between them. This is ironic because those who take this position also argue that these same historical social forms con- stituted barriers to the achievement of justice and equality. It is because of a failure to consider this history that Garfinkel has been considered conservative. Yet, a modern character who does not operate on the basis of traditional principles is considered to be shallow, lacking moral focus.
Garfinkel is alone in making the adjustments to thinking about where morality lies that are required by changes in contemporary social forms. Somehow, in the face of a modern society that is increasingly based in situated practices, both the practices themselves and those sociologists who have studied them are considered to be less moral than the truly unequal and unjust social forms and the ideas and research practices they generated that these situated practices and characters have replaced.
This is backwards. Garfinkel’s transformation of the person into a situated actor, and location into situated action are designed to capture the uniquely con- stitutive quality of ongoing face-to-face interaction. He has, from the beginning, talked about trust and reciprocity – moral issues – as strongly implicated in the production of situated action and identity, in ways that must be respected by all participants in situations (Garfinkel, 1963; Rawls, 1989). Garfinkel’s trans- formation of time into situated interactional time shared by all competent participants in a situation: time used by actors to order interactional sequences, time standing at the heart of the coherence of practices is equally implicated in the issue of situated morality.
Far from lacking a conception of time, and being politically conservative as a result, Garfinkel’s position offers a conception of interactional time that requires a new way of orienting toward moral issues in modern societies. Society in its traditional forms may have all but disappeared as many have said
peer-00570994, version 1 - 1 Mar 2011

170 TIME & SOCIETY 14(2/3)
but people still make sense and connect with one another in public places, they still face inequality and exclusion and feel disconnection – they just do so in new forms. New ways of thinking about social orders and the moral issues inherent therein are required.
Sequentiality is the time dimension in and through which persons mobilize and organize the enactment of practices and the presentation of identities that comprise the order and meaning of modern society. Time, in this regard, is the sequential time of interactional face-to-face relations. As such, time is a consti- tutive feature of practices, and trust and reciprocity are implicated in its use. The order in which things are said and done – their placement before or after one another – is constitutive of how they mean.
As Garfinkel (1948/2005) says, a person may have many thoughts at once, but they cannot say them all at once. Faced with others with whom one wants to communicate – and with whom one has little in common – it is necessary to participate in orderly ways in recognizable practices. Some things have to be done and said first and others next. This placement of turns at practice first and next is one of the principal means by which people can indicate to one another what they mean. A second thing said or done can mark the interpretation of a first and so on. A thing said or done (or not done) after a question or statement can mean something very different from the same thing said before.
The position of action in a sequence is a part of what and how it ‘means’. This quality of time as a relationship between parts of interaction is not new – all face- to-face interactions have this character in all types of society, but the heavy reliance of modern societies on practices in which time – sequentiality – is the organizing principle with no fallback on shared beliefs, or biography, where cultural artifacts (already made) are not standing in the wings providing a shared ritual symbolic basis for understanding, is new. That words and gestures have different meanings according to their sequential placement must become a focus of attention in interaction. As Garfinkel (1948/2005: 106) says, ‘any sign can signify anything’, an early statement of the principle of indexicality, and partici- pants must attend to this possibility in constant and developing sequential details.

Individual and Historical Time Versus Sequential Time

When time is treated as an individual experience, it is generally portrayed as having a primarily inner dimension in the flow of experience and is treated as unidirectional (from present to future, although it is available to reflection) and related to individual projects and goals. When time is treated as natural or historical, it is typically treated as external to individual experience and a characteristic property of social forms. In the case of external time, individuals are said to use abstract concepts to represent a historical time series to them-
peer-00570994, version 1 - 1 Mar 2011

selves. In both cases, the flow of time is represented as moving from past to future (even Schutz’s (1962) vivid present is oriented toward a future; it just never gets there).
For both individual and historical conceptions of time, the question of whether solipsism or positivism come into play is a problem. With regard to solipsism, if time is a feature of individual experience the problem is how people share it or communicate about it. In the case of positivism, if time is truly external, it is not possible for any individual to have an authentic experience of it. They must use conceptual representation, or belief, in place of the thing which they never actually experience.

Garfinkel (1948/2005) considers these various conceptions of time. All are characteristic of some social perspective, but, according to him all are ulti- mately derived from the constitutive sequential time of the vivid present:
Each type of social relationship has its particular type of time perspective, though each type is derived from the vivid present. There is the quasi-present in which the actor interprets the outcome of the other’s communicating actions without having participated in the ongoing process of the communicating acts [before taking a turn]. There is the peculiar time dimension in which the actor is connected with contemporaries he has never met, or with successors, or predecessors. There is the historical time in which the actor experiences the actual present as the outcome of past events. And so on. All of them originate in the intersecting of duree and cosmic time. (p. 182)
While both inner and external time originate in the vivid present, interactional time is different from the conceptions of both inner and external time. Its sequential constitutive properties – the commitment it requires – distinguish it from the others.

By recognizing the character of time as an organizing feature of mutual social interaction, Garfinkel avoids the problems of both solipsism and positivism. His argument has in common with Durkheim (1893/1933) the principle that, because an experience comes from a social event in which time was an ordering principle, the experience is both shared among all parties to the social event, and thus available to them all in the same way, and also experienced as an element of internal time flow during the event.
But, for Garfinkel, unlike Durkheim, sequential time is not just an outcome of interaction: something that already exists. It is an intrinsic ordering principle that creates a relationship between the parts of ongoing interaction moving forward. After the fact, accounts and narratives can be constructed that place the event into a larger context of order. This would be time in a historic sense. But, these accounts are not part of the original sequential order. They do not themselves constitute the ordered relationship between parts, or sequences, of interaction and, thus, can be both challenged and changed. Their ‘truth’ depends more on the perspective of the teller than on the order of the events themselves.
peer-00570994, version 1 - 1 Mar 2011

172 TIME & SOCIETY 14(2/3)
For Garfinkel this distinction between developing orders, and orders in retro- spect is an important one. Throughout his writing, Garfinkel distinguished between accounts as retrospective matters and prospective sequential orders.9 In the 1948 manuscript the distinction is made in the following way: there are, Garfinkel (1948/2005) says, ‘at least two types of signs’. The first is ‘the ready- made outcome of the other’s communicating acts, as, for example, a signpost’. These he calls ‘culture’. These signs already exist, do not require constitutive work to make them and can be oriented retrospectively. The second type of sign is one to which the actor attends as it is being produced, where that production is an integral part of the understanding, as for instance, he says,

two persons engrossed in conversation. In this latter type the signs are conveyed piece-meal, portion by portion and within a framework of space and time [sequen- tially]. While the one actor conveys his thought through this sequential order of actions, the interpreter follows with interpreting actions. (p. 181)

The first type, the ‘signpost’ or ‘culture’ is already made and does not require interactive collaboration. The interpretation of this type of sign is thus retro- spective and the interpretation of it is not interactive. The second is made over the course of a developing sequence and depends on the joint and ongoing inter- pretive efforts of parties to the conversation over its course.
Sequential time – as an order of ongoing events – is different from time as an abstraction. It is not a property of narrative accounts produced after the fact, but a feature of how events are ordered by actors to make them intelligible in the first place – going forward – by attending to the placement of sequences one after the other. In the 1948 manuscript Garfinkel describes what we might call sequence pairs:
A acts towards B as if the signs that B provides are not haphazardly given. When we say that A understands B we mean only this: that A detects an orderliness in these signs both with regard to sequence and meanings. The orderliness is assigned to B’s activities by A. The ‘validity’ of A’s conception of the signs generated by B are given in accordance with some regulative principle established for A when his return action evokes a counter action that somehow ‘fits’ A’s anticipations. (p. 184, emphasis in original)

Sequentiality involves a relationship between a first and a next where each has implications for the coherence of the other – in both directions – and each next can change the coherence of everything that went before as an ongoing and con- stantly developing matter. This forward–backward character of interactional exchange, reflecting back on itself as it moves forward – as a method of moving forward – is what Garfinkel meant by reflexivity.

Sequential time is not time used in the interpretation of existing or concluded events, but a property of those events as they move forward: an interrelationship between the parts. Furthermore, if all parties do not attend to the significance of
peer-00570994, version 1 - 1 Mar 2011

sequential placement in the same way, in the same event, immediate problems will result which will need to be attended to on the spot. The remedies for problems also make use of time as an ordering element of the remedy, or problem account. Thus, there is little room for what is conventionally meant by ambiguity, or solipsism, with regard to the sequential time of mutually intel- ligible practices.

Garfinkel recognized that in the flow of conversational events, as one thing follows another, the second event can change the meaning of, or give meaning to, the first. Because this is true, order – sequence or sequential time – offers a valuable resource for communication that all communicative practices make use of. Because these sequential orderings are constitutive of conversational practice, participants must attend to sequential orders of possibility, and social practices must be made to develop in such a way that complex elements of possible sequencing are incorporated into them moment by moment in expected and recognizable ways. These sequential orders don’t usually project into the future much farther than the next move, and the time dimension of interaction is, according to Garfinkel, experienced as what Schutz (1962) called a ‘vivid present’, rather than as a movement toward a future from a past.
In fact, sequentiality turns out to be one of the more important and constant features of interactional practice. This has been demonstrated most dramatically by conversational analysis. Building on Garfinkel’s initial insights and inspired by the pioneering work of Harvey Sacks, Emmanuel Schegloff, Gail Jefferson and Anita Pomerantz (1975), conversation analysis demonstrates the impor- tance of turns as a distribution system for ordering interaction.10 In their work, sequential time, conceptualized as a turn-taking-system, is demonstrated as an organizational feature of conversation.
Because sequential time not only occurs in public space, but the practices for ordering sequences must be oriented toward by all competent participants in any interaction, sequential time is experienced simultaneously, and in the same way, by all participants. Consequently, time, for Garfinkel is not only a property of individual inner experience, although it is also that. But, to the extent that time is an ordering principle of social practices – the time of interactional sequences – it is the same experience simultaneously for all participants. In this way Garfinkel’s conception avoids solipsism. Sequencing is a way of making – and distinguishing – moves: a public and shared dimension of time.

Sequentiality as an Order of Public Interpretation

The order in which things are said and done, their placement before and after one another is a large part of what they will come to mean: part of how they mean what they mean. Sequentiality is the time dimension within (through)
peer-00570994, version 1 - 1 Mar 2011

174 TIME & SOCIETY 14(2/3)
which persons mobilize and organize the enactment of practices and the presen- tation of self-identity that comprise the order and meaning of modern society.
Faced with others with whom one wants to communicate, it is necessary to participate in an orderly way in practices. Some things have to be said and done first and others next. This placement of turns at practice first and next turns out to be one of the principal means by which people can indicate to one another what they mean. Sequencing in interaction is one very important way of indi- cating what a particular action means. It can always been seen as related to what came before it. And what comes after it will be related to it. What something means at one moment – to the speaker and to others – can be changed com- pletely by what comes next. Jokes play on this possibility, but this is a constant feature of conversation as a whole, not only jokes.

Communication, according to Garfinkel (1948/2005), ‘is a temporal process’. Thoughts may occur all at once. But, they cannot be spoken all at once. ‘If one will reflect on his own experience,’ he says, they ‘will find that the succession of thought as it occurs “internally” undergoes a selective ordering process in which form it is presented “piece by piece” to the other fellow’ (p. 185). Conversation proceeds utterance by utterance.

Everything – what all speakers say and do – goes into making up what the communication will finally have meant. Garfinkel refers to the actor as a symbol treater and says that the symbol treater, ‘in treating these signs [of the other] generates further arrays of signs for treatment’ (p. 179). This, he says, results in a sequence built by both parties to the conversation that is organized by its temporal properties and is part of the immediate temporal experience of both: ‘the communicator does not experience only what he actually utters. A complicated mechanism of retentions and anticipations serve to connect one element of his speech with whatever preceded and what will follow until the unity of what he wants to convey has been grasped’ (p. 181).

With regard to such sequences, the timing and placement of the elements contribute to meaning. The way in which thoughts will be presented – the order, choice and placement of words – are all interpreted with regard to their relevance to the situation and time in which they are spoken and to the situated identity being accomplished.
Garfinkel (2002) uses a tutorial with a metronome to introduce students to the idea that time does not order practice – practice makes and orders time (see also Rawls, 2002). He distinguishes between ‘making the time we need’ and ‘clapping in time’ with the metronome. The point of this tutorial is that in order to make use of the metronome as an ordering device for clapping it is necessary to organize the clapping so that it covers the sound of the metronome. Although one might hear the metronome behind a piano softly playing, his point is that we have organized ourselves such that we would not have to hear it.
peer-00570994, version 1 - 1 Mar 2011

Sequences in Details

The elements of practice that comprise recognizable elements of ordering, placement, and sequencing, Garfinkel refers to in the 1948 manuscript as ‘tactics’ and ‘strategy’. Later he will refer to sequences in terms of their details only and not in terms of abstractions like these. These details constitute a sequential ordering of parts: a time element. For Garfinkel (1948/2005) ‘the tactics and the strategy arise from the fact that the agent cannot present his “stream of thought” at one instant’ (p. 185). Not only does communication require a temporal ordering, but the tactics and strategies involved in this temporal ordering vary from situated practice to situated practice and from situated identity to situated identity.
This emphasis on tactics, as elements in a situated temporal sequence by sequence mutually built and developing horizon of order, distinguishes Garfinkel’s position from the usual assumption that tactics and strategies arise in the individual mind (or as a feature of the developing trajectory of an indi- vidual’s actions) as plans for accomplishing an individual actor’s ‘projects’. Garfinkel maintains that mutual intelligibility – the possibility of communica- tion between people – requires a mutual engagement between people in a sequential production in which they must make use of tactics and strategies that are situated as methods for presenting their ‘thoughts’ in a form recognizable to others which, thus, conveys meaning to them.
This conversational/interactional form is inherently sequential. That is, the ordering features of talk – the placement of utterances – is a huge and essential tool that people use to render their ‘thoughts’ in a mutually intelligible form. And they never manage this without having the sequential back and forth character of interaction change what they mean. That is, what they will have meant in the end – even to themselves – will be what emerges from a collaborative sequential production, not what they thought they meant before the sequential series was produced.11
Each situation has different practices and situated identities that are relevant to it. The actor – as a ‘symbol treater’ who is enacting a particular situated identity – must produce speech that is recognized as appropriate to both the situated identity and the situation in which that situated identity enacts itself. According to Garfinkel (1948/2005), ‘communication refers to the process wherein the actor treats an array of signs . . . and in treating these signs gener- ates further arrays of signs for treatment’ (p. 179).

Treating the actor as a whole person pursuing a project obscures this process, because it obscures the fact that they must enact a different identity from situa- tion to situation and do so by enacting practices that are appropriate to the situation. In doing so, the whole-person – or demographic – approach, obscures the orderly and patterned ways in which actors use practices to produce their
peer-00570994, version 1 - 1 Mar 2011

176 TIME & SOCIETY 14(2/3)
identities and make sense in situated contexts of action. Focusing on the whole person results in looking for motives and projects rather than focusing on the constitutive features of sequential interaction.

Sequential Orders Versus Projects

One way of explaining the prospective ordering of interaction was offered by pragmatism: that individuals are pursuing projects. Schutz (1962) offers a more sociological version of this argument wherein what projects can be is con- strained by something like conceptual frames. Parsonian sociology employs a different form of the same argument in which individual behavior is explained as an orientation toward socially defined goals and projects to be pursued. For both of them the individual stands at the center, and action orients toward anticipated future goals. In spite of the fact that Schutz is considered micro and individual while Parsons is generally thought of as dealing with macro structures, they are the same in placing the individual at the center.
The underlying idea is that social orders result from an aggregate of indi- viduals pursuing their separate projects together. In order for this to work, how- ever, others would have to be able to ‘see’ what persons are trying to achieve. The various projects would also need to be coordinated. As long as people pursue projects the value of which are broadly shared within the society, it might be possible to achieve some social order in this way. In a traditional com- munity, for instance, in which people know each other well and share beliefs, customs and goals, others may be able to see in the actions of others something of the projects they pursue. But, even then, because much of interaction must be negotiated bit by bit, and projects and goals simply do not provide the means for doing that, a background of intelligibility – constructed through situated practice – would be necessary, even where projects were shared and visible.
If social order is thought to result from an aggregation of individual projects, time matters in a particular sort of way. Individuals come together with their projects. The projects themselves have a trajectory. Only if one person’s projects can be brought into conjunction with another’s can they become part of a joint reality. For Schutz (1945), they do not have the accent of reality, or are taken for granted, until this happens. So conceived, individual projects would begin in the mind. They would have a history in which they begin in anticipation without the accent of reality, then are brought into interactions and either receive the accent of reality, because they are successfully coordinated with others, or do not get the accent of reality because they fail in this coordina- tion.
Time would, in this sense, be the history of projects. It would follow projects from place to place. Projects themselves would have histories and their histories
peer-00570994, version 1 - 1 Mar 2011

would belong to the individual persons who anticipate and subsequently initiate them. Joint projects would be part of the history of groups.
Garfinkel’s time focus is on the more immediate matter of how sequences ever get enough coherence to become part of joint projects in the first place. This problem of intelligibility, and the role played by detailed orderings in it, is generally overlooked by conventional thinkers. Garfinkel notes that sequences establish relationships between the parts of interaction that make coherent orders of meaning possible.

Instructed Action Versus Projects

Garfinkel’s (2002) idea that instructions have praxeological validity is instruc- tive in contrasting his position with Schutz. Instructions indicate a trajectory for action. But, unlike projects, they belong to situations and events, not persons. Instructions have sequential properties. Instructions, in Garfinkel’s sense, must also be distinguished from rules. Rules, as traditionally conceived, may be said to exist in advance of being followed. This introduces a time dimension with regard to rules that is not the same as Garfinkel’s. Unlike a project that is intended before the fact, or rules that are construed as a set, one cannot know what instructions mean until they are in the process of being followed. They exist only over a sequential time series. They are instructable, but there is both more and less in the instructing than the instructions. The actions taken in following the instructions are part of the coherence of the instructions.
For Garfinkel, sitting with a set of instructions in front of you for assembling a chair, one cannot ‘see’ how to follow the instructions. It is the following, over a sequential time series, that makes the instructions coherent. Certainly the page of instructions is experienced as real. But, the anticipated following has no coherence because one cannot ‘see’ how to follow the instructions. They are not a description of the process. One looks at the page and tries to see how to follow the instructions. One sometimes turns the page this way and that trying to see how it ‘fits’ the pile of parts one has on the floor. One sometimes moves the parts themselves around.
Usually there is a first step that one can follow, and then with some more ‘trying to see’ a second step. Often after one or two steps, several more steps will appear to be clear. The way of seeing what to do begins to be clear. But, even when this happens it can turn out at some next step that something has been done wrong at an earlier stage. Then one has to go back, retracing steps, until one finds the mistake that can now be seen. Then the following steps – already taken – rearrange themselves and are seen differently even though one might say they had been ‘real’ and have been taken concretely and ‘seen’.
Instructed action is not a process of bringing an anticipation to reality. In fact,
peer-00570994, version 1 - 1 Mar 2011

178 TIME & SOCIETY 14(2/3)
it is the reverse; a process of bringing reality to an anticipation. That is, antici- pation is vague, or abstract, until one begins to work out how things might go together. Like a conversation – one starts with something that is constantly changing and becoming something else.
It is a very concrete process of enacting a series of steps that will come to complete a course of action. The steps cannot be imagined until they are being done. Of course the person has the intention, or project, of completing or build- ing the item. But, that project is completely irrelevant to the actual doing of the work that brings the item into being. The project cannot be used to build the item. There is no sense in which the project becomes more real as the item is built, either. The project was a real intention in the first place. In that sense the project, as a project, was always complete. But, the item is not complete until the sequence of its building can be clearly seen. And, the intention as an out- come that can be imputed to the builder does not exist until the sequential series is complete. Intentions, in this sense, come at the end, not at the beginning of things.
In fact, the completed item is sometimes so different from what the person intended that they are disappointed and realize that, while they have success- fully produced the required sequence of moves, they have not realized what they had thought of as their ‘intended’ project. In other words, the ‘intention’ in advance and the intention as an outcome are not always relevant to one another. Without accomplishing their intention, a person may nevertheless have followed a series of steps for assembling an object, and both that object and that process have been completed successfully. It is not intentions or projects that give action meaning.
The difference between Garfinkel and those who see interaction as ordered by projects, is that the latter are talking about bringing an anticipated mental project to reality through a series of steps. The idea is that at the beginning the project is only a thought and has no reality. At the end it exists in an inter- subjective world and its reality is acknowledged by others. This is a time sequence over the course of which, they argue, epistemological validity is achieved as intentions are actualized.
Garfinkel, as we saw earlier, points out that this does not happen. What he argues is quite different. People may indeed bring anticipated projects to inter- action, but these are not what concern him. In order to bring any projects to fruition, Garfinkel realized that people would have to have some methods for doing so. The important thing about an interaction is not that a project that was only imagined has at some point become real for the participants. The important thing is the medium through which something mutually intelligible has been achieved: the member’ s methods for producing mutually recognizable sequences.
This medium transforms both the anticipated project and the outcome, and
peer-00570994, version 1 - 1 Mar 2011

time plays an important role in the transformation. One may bring a project to an encounter. But, the ways that things must get done there will determine how that project can come out. The person will have to come to understand their project in a developing way that allows it to be constructed out of those methods available as sequences unfold.
Even more importantly, there are dynamics, not only of methods, but of reciprocity with regard to others, that are critical to any situated interaction. These dynamics will exert an influence on every project. Not only will indi- vidual projects be changed over the course of the interaction, but individuals will change their own understandings of what they wanted in the first place. These initial understandings may even fade entirely into the background as actors get caught up in the general interactional exchange, sequence by sequence, which takes all their attention as they are engaged in it over its course. They cannot attend to imagined projects while also being fully absorbed in managing the necessary interactional reciprocities over their sequential course.

Trust, Sequences and Motivation

It is the elements of communicative sequencing, by which Garfinkel means the recognizably produced features of situated identity and practice, that effect the plan of presentation in any given situation.
In considering the sequential aspects of situated practice Garfinkel (1948/ 2005) early on, in the 1948 manuscript, made a distinction between strategy – which involved a situated orientation to the whole of the ongoing interaction – and its sequential properties over its course, and what he called ‘sub-strategies,’ which operated across short sequences of turns, or what he called an extended series. ‘If strategy be considered to designate the communicative “plan” as it appears to the “auditor” [hearer]’, he says, ‘then tactics refer to the sub- strategies which operate in extended series’ (p. 184).
While Schutz (1945) had focused on ‘projects,’ and Parsons (1937) had focused on broader life goals and values, Garfinkel began by situating strategies and tactics in particular locations of interaction within which people have involvement obligations because of the need for reciprocity – that is, all parties to the interaction must orient toward the same practices in recognizably the same ways.
The importance of interactional detail cannot be overemphasized. Garfinkel is proposing ‘sub-strategies’ that operate sequentially, and distinguishes those from plans that operate as valued (even situated) goals, or situated courses of action more generally. There is something essentially ‘in its own right’ about the sequences of action that he calls sub-strategies. Unlike ‘projects’ more conventionally conceived, they take their own coherence as an end, and their
peer-00570994, version 1 - 1 Mar 2011

180 TIME & SOCIETY 14(2/3)
sequential properties are intrinsic to their coherence. They may have an objec- tive in the situation beyond themselves, but they don’t accomplish anything if they themselves are not coherent – so the order of sequences demands a great deal of attention in its own right.

The order of sequences is in this sense an example of something rather pure – ends in themselves – and, as Garfinkel (1963) will argue later, in the ‘trust’ paper, they demand a rather pure reciprocity on their own behalf. They have coherence as sequences, and they make demands on reciprocity and attentive- ness, just in their own right, as orderly sequences of action, regardless of intended outcomes.

As such, they offer a very different answer to the motivation question. Any per- son with the intention of being understood must attend to the careful building of such sequences. Thus, attention to sequential orders always supercedes goals and objectives as a motivation in the immediate time frame of ongoing interaction.
The essence of mutual intelligibility, for Garfinkel, consists in these situated sequential orders. Actors are constrained to build meaningful sequences of action in just the ways that are appropriate to the situations they find themselves in, with regard to just those identities they are trying to present. Changes in sequential orders will result in changes in identity. Thus, not only is this situated order real, and can be observed in details, but its order is necessary and conse- quential. Any meaningful action – in order to be meaningful – must display such an order, and recognizably.

The Performance of Speaker and Hearer

Garfinkel (1948/2005) refers to this developing horizon of communicative utterances as a ‘performance’ in which selves, confronting one another in a face-to-face situation, construct both the intelligibility of the situation and the identity of their respective selves within it:
Like a working action it is a performance for it embodies an intention to realiza- tion. The listener experiences the occurrences of the other’s action as events occurring in outer time and space, while at the same time he experiences his interpretive actions as a series of retentions and anticipations happening in his inner time and connected by the intention to understand the other’s ‘message’ as a meaningful unit. (p. 181)
While the interaction goes on, the listener and hearer share an orientation toward space and time in which their inner and outer experiences intersect in what Garfinkel calls a ‘common vivid present. Both can say later, “We experi- enced this occurrence together”’ (p. 181). This is the common present of sequential exchange.
peer-00570994, version 1 - 1 Mar 2011

There is, however, an important difference between the positions of speaking and listening. The speakers hear and see themselves while speaking and monitor what they are saying to see whether it has conveyed what they meant. ‘All of these experiences belong to the communicator’s inner time duree’. The listener on the other hand has a different job. They have to grasp the point, see what type of action the communicator is making, and ready themselves for a response that displays whatever understanding they have made of the utterance. The job of the hearer – the observer – is to make what they can of what the speaker says. Their orientation toward the interaction is therefore different.
The positions of speaker and hearer change continually over the course of interaction – each has different obligations and each involves a different aware- ness of self and other. But, both are mutually engaged, and as Garfinkel says: ‘Both vivid presents occur simultaneously. A new time dimension is therefore established, that of a common vivid present’ (p. 181).

The sequence of turns – the exchange of positions of speaker and hearer back and forth – creates a common shared historical sequence, a time frame, which the two (or more) have shared in a very special way. They built it together out of their collaborative work and it would not exist except for that collaborative work, and does not exist unless that work is successful.

Reflexive, or Sequential, Interpretation

Each person understands the signs produced in these sequences in their own way – but because they are built in a progression, as sequences, each of which confirms (or does not confirm) the anticipations of the prior speaker, there is a mutual orientation to orderliness that keeps differences in the interpretation of symbols from being a problem. Each can see what the other takes their utterance to have meant. And they can revise their own understanding as they go along.
Ordinary language philosophers (see in particular Paul Grice, 1957), have argued that there is no way of understanding whether the person you are speak- ing to has understood you. Garfinkel describes an order of conversational sequencing in which there is a constant confirmation of understanding as par- ticipants go forward. This sequential order does not require that all parties understand everything in just the same way. But, each next utterance, while it may mean somewhat different things to different parties to the conversation, works only by establishing a mutual understanding of the last. This provides a mechanism for the constant mutual display of understanding in which the next speakers continually confirm or disconfirm the expectations of prior speakers, and actors are not left in doubt as to whether others have understood them.

The way Garfinkel handles interpretation sequentially avoids the whole problem of how two people get the same idea. They don’t need to. The speaker
peer-00570994, version 1 - 1 Mar 2011

182 TIME & SOCIETY 14(2/3)

says something. Then it is up to the hearer/observer to make what they can of it. They construct an action that responds to what they have made of what the other did. The other can tell a good bit from this about how they have understood what they did. Sometimes they even learn something they did not know about what they did. You might make what you thought of as a compliment: ‘You look good today’ and the person says ‘Oh, so I don’t usually?’ This is not a misunderstanding on the part of the hearer. It is new information about how what one said could be heard. It could also be a joke and it might be a problem. But, only further utterances can tell if this is so. That is, if the first speaker doesn’t take it as a joke it isn’t, unless and until, the second speaker says it is and has that confirmed by the first, and so on.
The essential point is that meaning emerges over a sequential course of utter- ances. It is the job of the observer to make what they can of what someone says. The speaker is busy building a practice in a recognizable and appropriate style. They are deeply embedded in the natural attitude and cannot think about what they are doing. The observer can. Then they switch places and the former speaker becomes the observer and makes what they can of what the prior speaker said. This is their turn to reflect on the sequence. They display their understanding to one another sequentially and in this way ‘confirm’ for one another the understandings they have reached, whether mutual or not.

When Garfinkel worries about the adequacy of scientific observation (in social science), the worry is as follows: the serious problem with scientific observation is that it stops after the first part. The scientific observer interprets, but because they are not part of the ongoing activity they have no opportunity turn by turn to check out their understandings. If they stop the action to ask – or ask afterwards – neither of those moves has the same confirming character of a sequential exchange. Asking treats something as an accomplishment – as just a thing – that is in fact a complex sequential order, whose meaning is its order – as accomplished over its course – and not an account given after the fact. People do not consciously know and cannot tell about these orders. They are taken for granted. In fact, as Garfinkel’s (1967) research shows, people get angry if others try to confirm understandings in this way, by asking. And it is an infinite regress anyway because shared understanding is required for every utterance involved in the clarification. Confirming meaning can only be done by building – not clarifying – sequentially.

This is the real difference between ethnomethodology and other forms of field method. The ethnomethodologist must insert themselves into the situation and get their information in the way that ordinary actors do, sequence by sequence. Interpretations must be confirmed on the spot – and not after the fact – not by asking, but by doing, sequence by sequence.

Ethnomethodologists know that they cannot ask actors what they were doing or what it meant. People cannot say how they order sequences: the process is
peer-00570994, version 1 - 1 Mar 2011


taken for granted, and must be discovered as an order of affairs, not as an account. Ethnomethodologists, as a consequence, do not focus on typifications that they think actors are using, or on projects that actors may be orienting toward. For an ethnomethodologist not to be fully and completely embedded in the situation, not just for part of the time, but while they collect and analyze their data, is a problem. The conventional objectivity that researchers value, in fact, prevents understanding. Every interpretation by a researcher can only receive confirmation by making a move in a developing sequence that gets a response, which requires being embedded in the ongoing sequential order and even then over and over again.


While a conception of time is a necessary element of any social theory, it was only in the nineteenth century that time itself became a focus of analysis. Marx (1992) recognized that analyzing contradictions in the production of social inequality involved relations over time. Durkheim (1893/1933) recognized that particular configurations of time had a social element, and that time was consequently not experienced in the same way, was not even the same thing in different societies.

In the twentieth century, social thinkers focused on the development of the western notion of time, contrasted modern and traditional time, and pointed out that the clock time we have come to live by developed in tandem with factory life: that is, with western capitalism. Similarly, the idea of punishment in units of time in prisons developed at the same time that payment for labor by units of time appeared in the factory.

Pragmatism and phenomenology both involved a focus on time as an element of the organization of perception. Existentialism analyzed the relationship being and time. Norbert Elias (1987/1992) and Karl Mannheim (1952) outlined sociologies of knowledge that included time as a central figure. Time is also involved in the idea of rational choice as social action oriented toward norms and goals.

Garfinkel’s refusal to address this increasingly important focus on time in conventional terms has created the impression that he takes time for granted. He does not. His treatment of time as sequential is informed by consideration of pragmatism and phenomenology, as well as the more conventional structural perspectives of Parsons and Marx.
Even so, Garfinkel’s focus on sequential time will still be a concern to some. They will feel that it is impossible to consider the ‘important’ questions, which they feel are larger than the situation, from the standpoint of sequential time. It seems ironic, however, that on the conventional view, an individual’s orienta-
peer-00570994, version 1 - 1 Mar 2011

184 TIME & SOCIETY 14(2/3)
tion toward projects in terms of norms and values is considered to be an indi- cation of something called the ‘larger social structure’ and is therefore worthy of study, while the sequential structures which require mutuality and thus make collectivity possible are considered to be small and individual.

If we accept the need for attention to sequential orders (or ‘methods’) for rendering bits of interaction recognizable we must also recognize a need to rethink the conceptions of the actor, group, place and time. The actor must be the one in just this place, attending to just this interactional matter, engaged in attending to the production of just this sequential detail. The group is that con- stituted by just these actors, just so occupied, in sequentially producing just this practice, just here, just now. Furthermore, action is not only embedded in sequences of time – but its coherence consists largely in its sequential order. This is the social order that historical time assumes but does not examine.

All four of Garfinkel’s transformations (actor, place, group and time) are designed to take into account the situated character of social orders (Rawls, 1987, 2005). Formal institutional orders involve accounts, which work retro- spectively (Mills, 1940; Rawls, 1987). But, interaction must be built first time through each time it is done, or as Garfinkel says ‘each next first time’. Participants must be able to ‘see’ what it is that others are doing and saying before they can interpret what they have done and produce either sequential interpretations, or retrospective accounts. That requires a prospective ordering of events that cannot be accomplished by accounts or formal orders. Of course the need – or potential need – to produce an account later can also become a feature of the prospective ordering of situated practices.12

Not only does Garfinkel develop a conception of time that is central to his position, but his conception is crucial to understanding modern societies. The idea of historical time that informs the positions from which Garfinkel is challenged: the idea of time as history, is a product of the development of modernity that does not question the reliance of the coherence of modernity on sequences of time: the question Garfinkel takes up. As such, this conventional conception of time obscures what it claims to represent.

Typically, conceptions of time involve the inner time flow of individual expe- rience, the time of natural events (or external time), to what extent either or both are influenced by social relations (or the reverse), and the relationship between the two. For all the substantial differences between them, this is as much true of Pragmatism and Phenomenology as it is of Parsons and Weber. And, while Durkheim and Marx consistently treated time as both a social construction and as an element of social construction – they still placed their emphasis on the history of praxis instead of on time as an internal dynamic of practices.

By contrast with this tradition, Garfinkel recognizes that time, as an ordering principle of interaction, is constitutive of a public and shared dimension of experience: he recognizes that time, in its sequential character, while public, is
peer-00570994, version 1 - 1 Mar 2011

nevertheless also experienced both internally and simultaneously by all partici- pants who are successfully engaged in the same interaction.

Norbert Elias (1987/1992), one of the more important scholars of time, argued that time is a social convention which developed to allow people to com- pare things that do not happen at the same time or place. One of his main points, which he shares with Garfinkel, is that time as a social convention reproduces the structure of the society and its state of development, not a natural phenome- non. That we have thought of time as an objective measure and not as a social construction has introduced an element of positivism into most social, scientific and historical studies.
But, this is not what is important about the contrast between Garfinkel and Elias. For Garfinkel, time is also a social construction. Where their positions differ is that for Elias the focus was on the development of a social convention that could be used in the abstract to make comparisons across populations and events. He argued that this development was critical to the development of modern societies. Elias did not consider the possibility that time was itself an ordering principle of social action in situ.

Elias even suggests that the idea of time as an in-situ phenomenon is a problem, arguing that when people only use, or conceptualize, time in this way their societies have not developed. It is the development of the abstract idea of time that signals the development of a society for Elias. This is a very western- centric view, one the one hand, and also, ironically, one that ignores significant developments toward a practice-based interaction order in modern western societies. Quite a paradox.

In this regard, Elias’s position differs from Garfinkel’s in the same way that Parsons’s does. In studying social order, Parsons was not interested in particular instances of social order. He didn’t expect them to reveal anything interesting. What he wanted were abstractions that could be applied to broad collections of social phenomena. In fact, this is true of almost all social science. The search has been for abstractions, and time has been treated as just another such abstrac- tion.
Instead of seeking abstractions, Garfinkel has searched for order in details. One of the most persistent criticisms of ethnomethodology, in fact, has been that it does not get beyond details. What are all the details for? Why can’t ‘these people’ make up some abstractions?

As soon as a conventional field researcher begins to see some patterns, they make up abstractions: typifications. Ethnomethodologists respond that these abstractions entirely lose the social orders they claim to represent. But, con- ventional thinkers don’t question their faith in abstractions, and it is a faith. Meanwhile, ethnomethodologists continue to be ridiculed for their interest in details. Garfinkel’s point is that order and intelligibility are constituted in and through the details and so also the solutions to the problems of order, justice and
peer-00570994, version 1 - 1 Mar 2011

186 TIME & SOCIETY 14(2/3)
inequality lie in those same details. If abstractions entirely lose the details, any social science that resorts to abstractions must, by definition, fail.

This focus on abstractions extends to the treatment of time. The insistence that time as an abstraction is superior to situated time, while related to other criticisms of ethnomethodology, is largely responsible for to the criticisms of Garfinkel for not focusing on history. According to Elias (1987/1992), an ability to compare – to develop measures of comparison – is critical to social development. He con- siders that a society that has not developed such abstractions is not developed as a society. While this may or may not be the case (and is a strikingly western-centric view, many non-western languages not having past or present tenses), it is essentially irrelevant to a consideration of studies of interaction.

For Garfinkel there is a parallel between focusing on time as a way of com- paring aspects of different groups and the focus of conventional social science on the demographic characteristics of groups. In both cases, Garfinkel argues, an understanding of how coherent social phenomena are produced is being sacrificed so that we can produce a fictively objective measurement across groups and situations. And, as Elias himself contends, the measures are not objective – the abstractions are social constructions and at best relative.

Whether there is any truth to Elias’s claim that societies cannot develop until they develop the idea of time as history, time in both traditional and modern societies constitutes an essential aspect of the ordering process in ways that a focus on historical time does not capture – and in fact obscures. So, here is a question for Elias: can societies develop until they have developed an idea of time as a sequential ordering principle of mutually intelligible practices?
In so far as interaction in either traditional or modern societies is ordered, time, as a sequential phenomenon, must constitute an element of its ordering, whether or not people have developed an abstract conception of time.
Because time as an abstraction is itself a social construction and results from social settings, as an abstract conception, time cannot be used to understand what goes on in the processes that produce it. But, time as an ordering feature of interaction is still very much relevant and can be found concretely in situ.

Elias’s main point was that time itself has not been studied. He was certainly right about that at the time – and in spite of increased interest over the past few decades, time is still a neglected area of study. But, it is essential to note, in this regard, that neither type of time has been studied. It is all well and good to point out that time is a social convention that reflects the structure of the society and that we need to study it as such. But, it is also true that time as an ordering principle of interaction is a social convention and as such it should also be studied. If we are interested in how time reflects the social structure of a society, we should also be interested in how that social structure is ordered in its own right and in an ongoing fashion, especially if that ordering also makes use of time.
peer-00570994, version 1 - 1 Mar 2011

  1. The claim that Garfinkel is not interested in questions of justice and inequality has always been a mystery to me in any case. His first two publications (‘Color Trouble’ (Garfinkel, 1941) and ‘A Research Note on Inter- and Intra-Racial Homicides’ (Garfinkel, 1949) focused on issues of racial injustice – this at a time when most white scholars were busy ignoring race altogether. Somehow, when the politics of the 1960s defined interactional studies as conservative the label operated as a stereo- type, like any prejudice, and as a consequence the ‘truth’, as in the content of Garfinkel’s work, was not consulted.
  2. Durkheim (1893/1933) is a partial exception in so far as he also focused on practices and recognized their increasing significance in modern life. But, in other respects, in so far as Durkheim retained a conventional notion of person and group, his position continued to rely on abstractions. Because his argument was so revolutionary it had some inconsistencies (Rawls, 2004). If his position is used to frame Garfinkel’s, the abstractions make a nonsense of the thing. But, if Garfinkel is used to explain what Durkheim might have been working toward there is an interesting connection between the two positions.
  3. In writing this article, I have relied heavily on an early manuscript that was finally published in 2005 as Seeing Sociologically, but which was originally completed by Garfinkel in 1948. I have done so, not because Garfinkel places a greater emphasis on time in this manuscript. In fact, the ideas about sequential organization developed there only become more important in Garfinkel’s later work. I have referred so frequently to the early manuscript for two reasons. First, this is where Garfinkel explained his ideas about time, rather than just using or developing them empiri- cally; and second, because the manuscript, since it remained unpublished for almost 60 years, is largely unknown and deserves to be written about.
  4. Persons appear to have multiple and conflicting identities and often to stand at a distance from their own social roles, thus causing psychological distress. Garfinkel argues that his position solves the problem of seeing an individual as divided between multiple roles. If we take seriously the time dimension of interaction then we can see that at each moment in time, or to be more precise, with each move that a person makes interactionally (all moves being ordered in a time dimension sequen- tially), they are enacting only one identity. A person does not enact two identities at once. They may at one moment enact one identity and at the next moment another. But each will be enacted through practices appropriate to that identity. Seeing a person as divided between multiple identities happens when you treat an actor as a whole person. If we treat the actor as only the identity being enacted at any given moment, the problem of competing roles, and of role distance disappears. There is only one identity in one bit of situated time.
  5. Generalizability and reliability require abstracting from situated orders such that their details are lost. If there is anything important going on in situated interaction, it would be lost to any method that generalized over time and space. Why should action only be significant in so far as it remains constant across bits of time and space? If the parts of action lost through abstraction are the parts that made it valuable and meaningful in the first place, then of what value is generalizability if it loses them? And why would we think it was valuable in the first place? We have a
peer-00570994, version 1 - 1 Mar 2011

188 TIME & SOCIETY 14(2/3)
meaningful encounter. But, as scientists, do we feel there is nothing of value in it
until we have reduced it to unrecognizable abstractions?
  1. In addition to focusing on demographics and not on the relevant features of inter-
    actional exclusion, methods that rely on institutional reifications (statistics) will almost always reproduce the institutional fictions that the data was originally pro- duced to maintain.
  2. Competenceisaveryimportantcriteriainthisregard.Incompetentparticipationwill result in a lack of mutual intelligibility. Competence is assured by the mutuality of sequences. This is called praxeological validity in Garfinkel’s (2002; see also Rawls, 2002) terms.
  3. When personal characteristics cannot be put aside, problems arise. Garfinkel (2002: Chapter 9) gives some nice examples of scientists who have been trained in one scientific lab and then move to another. The practices of labs differ. My favorite story involves how the killing of a rat differs from one scientific lab to another. Theoretical, abstract, or so-called ‘generalizeable knowledge’, the kind you get from a textbook, can often get in the way in science labs where a shared practice is the foundation for coordinated work. Even ‘standardized’ bench practices vary from lab to lab. There are also implications for the study of disabilities which can prevent people from producing practices recognizably. Derek Coates and I have done some interesting research on blindness in this regard, with the finding that not being able to see the unfolding of sequentiality sometimes makes it impossible to participate in producing the orders required by mutual intelligibility (Coates and Rawls [forth- coming]).
  4. It has been my argument that one of the distinct qualities of more formal practices is that they eliminate this step-by-step character of practices and thereby build in a great deal of potential for ambiguity. There is also a power differential, however, which is even more important. The step-by-step building of practices requires a high degree of reciprocity and equality. When practices formalize, they no longer require this and thus by moving away from step-wise practices some people can come to wield more power than others.
  5. See especially what has come to be known as ‘the turntaking paper’ by Harvey Sacks, Emmanuel Schegloff and Gail Jefferson (1974). See also Sacks (1996).
  6. Weber (1921/1968) says something like this about how a person finds out what they meant when they see how others take it – how they react – a sequential placement in a time series. However, he does not mention time in this regard (see p. 8).
  7. Garfinkel (1967) gives a good example of this in ‘Good Reasons for Bad Clinic Records’. There he argues that the incompleteness of the intake files is oriented toward their eventual use to produce organizational accounts. Because they are oriented toward this use, information that would seem to be missing from a treat- ment or a research perspective is missing from them. Their production in its course is oriented toward their subsequent accountable uses.
Coates, Derek and Rawls, Anne Warfield (forthcoming) ‘Trust, Trouble and Lies: Language Games of Blindness’. Paradigm Press.
peer-00570994, version 1 - 1 Mar 2011

Coser, Lewis (1975) ‘Two Methods in Search of a Substance’, American Sociological Review 40: 691–700.
Durkheim, Emile (1893/1933) The Division of Labor in Society. New York: Free Press. Elias, Norbert (1987/1992) Time: An Essay. Oxford: Blackwell.
Garfinkel, Harold (1941)
‘Color Trouble’. The Best Short Stories of 1941 and the
Yearbook of the American Short Story. Boston: Houghton Mifflin.
Garfinkel, Harold (1948/2005)
Seeing Sociologically. Boulder, CO: Paradigm Press. Garfinkel, Harold (1949) ‘Research Note on Inter- and Intra-racial Homicides’, Social
Forces 27: 369–81.
Garfinkel, Harold (1963) ‘A Conception of and Experiments with “Trust” as a Condition

of Stable Concerted Actions’, in O. J. Harvey (ed.) Motivation and Social Interaction.
New York: The Ronald Press.
Garfinkel, Harold (1967)
Studies in Ethnomethodology. Englewood-Cliffs, NJ: Prentice-
Garfinkel, Harold (2002)
Ethnomethodology’s Program: Working Out Durkheim’s
Aphorism. Boulder, CO: Rowman and Littlefield.
Grice, Paul (1957) ‘Meaning’,
Philosophical Review 66: 377–88.
Kuhn, Thomas (1958)
The Structure of Scientific Revolutions. Chicago: University of
Chicago Press.
Mannheim, Karl (1952)
Essays on the Sociology of Knowledge. New York: Oxford
University Press.
Marx, Karl (1867/1992)
Capital. London: Penguin.
Mills, C. Wright (1940) ‘Situated Action and Vocabularies of Motive’,
Sociological Review.
Parsons, Talcott (1937) The Structure of Social Action. New York: Free Press. Pomerantz, Anita (1975) ‘Second Assessments’. Unpublished Dissertation, University of
California, Los Angeles.
Rawls, Anne Warfield (1987) ‘The Interaction Order
Sui Generis: Goffman’s
Contribution to Social Theory’, Sociological Theory 5(2): 136–49.
Rawls, Anne Warfield (1989) ‘Language, Self and Social Order: A Re-Evaluation of

Goffman and Sacks’, Human Studies 12(1): 147–72.
Rawls, Anne Warfield (2002)
‘Introduction’ to Ethnomethodology’s Program: Working
Out Durkheim’s Aphorism, by Harold Garfinkel. Boulder, CO: Rowman and
Rawls, Anne Warfield (2004)
Epistemology and Practice: Durkheim’s The Elementary
Forms of Religious Life. Cambridge: Cambridge University Press.
Rawls, Anne Warfield (2005) ‘Introduction’, in H. Garfinkel
Seeing Sociologically.
Cambridge, MA: Harvard University Press.
Riesman, David (1950)
The Lonely Crowd. Cambridge, MA: Harvard University Press. Sacks, Harvey (1992) Lectures (Volumes I and II). Oxford: Blackwell.
Sacks, Harvey, Schegloff, Emmanuel and Jefferson, Gail (1974) ‘A Simplest System-

atics for the Organization of Turn-taking in Conversation’, Language 50: 696–735. Schutz, Alfred (1945) ‘Multiple Realities’ in The Collected Papers I: The Problem of
Social Reality. The Hague: Martinus Nijhoff.
Schutz, Alfred (1962)
The Collected Papers I: The Problem of Social Reality. The
Hague: Martinus Nijhoff.
Weber, Max (1921/1968)
Economy and Society. Berkeley: University of California