Sociology 319

January 24-27, 2003


Harold Garfinkel and Ethnomethodology


The reading for this section is “Some essential features of common understandings,” from pp. 38-44 and p. 75 of Garfinkel (1967).


1. Introduction


The ethnomethodological approach studies social interaction through enacted conduct so that it is similar to the perspective of Mead and symbolic interactionists.  At the same time it differs somewhat from these perspectives in that it focusses on activities themselves and how they are conducted in particular contexts, rather than examining the meaning and interpretation in interaction.  Writers in the ethnomethodological perspective seem little concerned with the philosophic underpinnings of social interaction or in developing an all-encompassing theoretical structure for social interaction.  Rather, they analyze social interaction in particular situations and contexts, attempting to describe and understand the methods, procedures, and considerations involved in social interaction.  For the ethnomethodologist, social interactions should not be considered as rational or irrational, or subject to “error,” rather social interaction is a set of common sense procedures people use to deal with situations and contexts where there may be uncertainty and flexibility.


Heritage defines ethnomethodology as

the study of a particular subject matter: the body of common-sense knowledge and the range of procedures and considerations by means of which ordinary members of society make sense of, find their way about in, and act on the circumstances in which they find themselves.  (Heritage, p. 4).

This body of knowledge is “analytically primary to any theory of social action” and “consideration of these issues can be made an integral part of the theory of action” (Heritage, p. 5) and constitute significant research questions.  Garfinkel’s approach is “to detect some expectancies that lend commonplace scenes their familiar, life-as-usual character, and to relate these to the stable social structures of everyday activities” (Garfinkel, 1967, p. 37). 


In common with other interactionist approaches, ethnomethodologists look on these everyday interactions, and the practices involved in them, as having a regularity or stability so they form what sociologists term institutions and structures.  But these are always actively produced through interaction among social actors, but not necessarily in a conscious or meaningful manner.  Rather, social actors are accountable for their actions in the encounters that occur in particular situations and contexts.  This means that “social action in a context is an actively produced accomplishment” (Cohen, p. 90).  The ethnomethodological perspective emphasizes a social interaction that has a rationality of its own and the conduct is one of accomplishment – making sense out of situations and responding in an accountable manner.  That is, this social conduct not adherence to some rationality imposed by sociological study or norms and roles perceived to be common to all in society. 


Ethnomethodologists primarily examine particular encounters and attempt to uncover the common practices that social actors may assume or not consciously consider.  While they provide great insight into what these are and how work, it may be difficult to connect this approach to more comprehensive theoretical approaches to sociology.  More than the other interactionist perspectives, ethnomethodologists focus on methods of social action and may not attempt to understand social actors in the way that Weber, Parsons, or Mead did.  But the ethnomethodological perspective provides useful insight.  An example of how it can be applied to sex and gender inequality is provided in the article by West and Fenstermaker.


2. Garfinkel and ethnomethodology


Harold Garfinkel (1917 -  ) was born in Newark, New Jersey and, after studying economics, entered a doctoral program in Social Relations at Harvard University in 1946.  At the time, Talcott Parsons and other faculty members were attempting to develop a systematic theory and understanding of a general theory of social action.  The Structure of Social Action of Parsons had been published in 1937 and the action theory of Parsons forms the backdrop to Garfinkel’s approach.  In his doctoral dissertation, Garfinkel examined the different approaches to social action and “was in search of a theoretical framework which would directly catch at the procedures by which actors analyze their circumstances and carry out courses of action” (Heritage, p. 9).  In a study of juries, he found that that jurors “did not necessarily adhere to formal rules.  Instead, they brought their own methods to the proceedings, improvising upon the judge’s instructions by means of practices that they would use at home with their peers” (Cohen, p. 92).  While this might lead to doubting whether the verdicts are well founded, Garfinkel used this observation to being investigating other forms of social interaction and found similar procedures were used elsewhere. 


Garfinkel received his doctoral degree in 1952 and spent most of his academic career in the Department of Sociology at the University of California, Los Angeles (UCLA), where is listed as a Professor Emeritus on their web site.  Garfinkel coined the term ethnomethodology, meaning the methods used by people in accomplishing their daily lives.  His major work is Studies in Ethnomethodology, published in 1967, and the breaching experiments, for which he is noted, come from that work.  Many later ethnomethodologists were trained by Garfinkel although others are now associated with this sociological perspective.


3. Common understandings


The reading for this section contains an analysis of common or implicit understandings associated with a simple conversation (pp. 1-2), Garfinkel’s comments on some implications of this (pp. 2-3), seven cases of breaching experiments (pp. 4-6), and a short conclusion. 


The initial conversation reported is apparently understood by husband and wife in that the conversation is continued and developmental – proceeding through time, issues, and events.  From this, Garfinkel draws a number of conclusions (pp. 2-3).  Note that these did not appear to impede the flow of the conversation.


·        Several matters were not mentioned or were left unsaid.  These stand as background for the conversation, demonstrating the importance of context and how meaning is not inherent in particular words or symbols but is connected to context.

·        Some (non-time) matters were understood in connection to the time sequence of events – parking meter and store visited.  This demonstrates the importance of considering time, sequence, and flow in social interaction – these appear to be absent from some approaches to social action and interaction.

·        Understanding work was done by both partners to the conversation.  Interaction sociologists emphasize the work and accomplishment associated with daily activities, and this example demonstrates this.  While the mental energy may not have been all that great in this case, the conversation did require work and commitment.

·        Biography and history, as well as prior, current, and future interaction all were involved in the conversation.  In this example, these appear to have worked adequately, so that the conversation was understood by both parties, continued, and proceeded in a particular direction.

·        Waiting was also a crucial aspect of the conversation.  Each party appears to have waited for something and was willing to listen to what the other party said. 

·        In the middle paragraph on p. 2, Garfinkel argues that, as a result of these considerations, a new time parameter is introduced – not clock time but a time “constitutive of  ‘the matter talked about.’”   That is, there is a sequence and timing of events and understandings, but these are not in strictly clock time, but may be sequenced relative to the understandings of the parties, in terms of both “process and product.”  For example, “your loafers need new heels badly” must have been a prior observation (from before all these events occurred) and yet occurs at the end of the trip.  Examples of this sort demonstrate a more sophisticated human understanding of time than as a linear, continuous, directional measure which moves forward in particular units.  Also note the later reference to the “memory drum” (p. 3) where the common understanding has greater capability and creativity than do computer-type memories.


Garfinkel then outlines some implications for sociological analysis of interaction and, in particular, conversations.  Some of the following points should be connected with Garfinkel’s critique of explanations of social action arguing that it is guided by conscious, utilitarian, rational considerations (see end of 1st complet paragraph on p. 3).


·        Transcripts or audits of conversations may be very incomplete and, possibly, misleading (item 1, bottom of p. 2). 

·        Words and expressions do not necessarily have the same meaning throughout a conversation.  At other places, Garfinkel notes that expressions are indexical, that is, expressions “which depend for their sense upon the circumstances of their production, of who said them, when, where, in relation to what and so forth” (Sharrock and Anderson, p. 42).

·        In the conversation, references to events are vague and not restricted to “a clearly restricted set” as might be required by utilitarian theoretical approaches – even though the events themselves must have been clear-cut, in involving individuals in ordinary time and experience.  Garfinkel further notes that events are open to “internal relationships, relationships to other events” (bottom of p.2) and connections to other possible objects, events, relationships, and possibilities.  That is, there is a fairly open set of options and outcomes, not so much a restricted set as utilitarian models of choice might imply.

·        Just as time is not linear (last bullet of section above), so the expressions move ahead and back in the conversation.  There is a progressive realization of the whole situation, in both prospective and retrospective senses.  (item 3, p. 3).

·        As noted in item 4, p. 3, all expressions are context dependent – again a demonstration of the indexical aspect of expressions.


In the first complete paragraph on p. 3, Garfinkel draws out two results of this analysis – these appear to be critiques of other approaches to social interaction.  He first implies that humans do not treat the events or recount them in a “temporally constituted character” or as “precoded entries on a memory drum.”  Although probably not intended as a critique of information treated in computer fashion, the implication is that human minds and interaction work differently than do computers.  While Garfinkel does not specifically address the issue of consciousness and mind here, implications concerning its complexity, ability, organization, and working are implied.  Second, the last part of this paragraph seems to be a direct criticism of utilitarianism and rationality (as in the approach of Parsons) as a means of explaining social interaction.  In the next paragraph, he comments “persons refus to permit each other to understand .. in this way” – that is, individuals entering into social interaction conduct affairs in other ways.  Instead of the well-defined ends and means, within well understood constraints, there is a “vagueness” and “unnoticed” aspects that are socially acceptable (sanctioned), and it may be as much this as anything that produces “reasonable, understandable, plain talk” (p. 3).  At the same time, these properties of discourse are required so people can “conduct their common conversational affairs without interference” (p. 3). 


These comments lead to the seven cases of breaching experiments. The purpose of the breaching experiments is to uncover the tacitly accepted practices and procedures by attempting to violate them.  They demonstrate the existence and “centrality of a cognitive order by withdrawing it momentarily” (Cohen, p. 90).  If you conduct such experiments yourself, be careful how you do this and be prepared for the consequences of social disapproval.  Friendships could be destroyed if these are more than temporary or quick experiments.


In the concluding remarks on p. 6, Garfinkel adopts what I consider to be a democratic and participatory approach to social interaction and analysis of social interaction, and one that demonstrates both human attention to patterns and regularity and to improvisation and creativity.  He comments that such issues do not belong just to philosophers or sociologists.  Rather, ordinary people in everyday interaction with others, have a common sense knowledge of interaction and structures.  They use a variety of methods to deal with social interaction, institutions, and structures, and enter into activities in a common sense but sociological manner.


4. Cohen on Garfinkel


For the ethnomethodologist, “social action in a context is an actively produced accomplishment” (Cohen, p. 90).  While the sociologist may have difficulty in constructing a theory of social action, individuals in social interaction construct common-sense understandings through their production of social action.  While ordinary people in situations may not understand the tacit procedures they adopt, they are generally good at developing and using these procedures.  That is, there is a “cognitive order” of “normal practices [that] produce intelligible features of social organization, an order in the events that actors take for granted in their everyday lives” (Cohen, p. 90).  When this cognitive order breaks down, there can be “a profoundly disturbing type of anomie in which nothing appears to make any sense at all” (Cohen, p. 90). 


Also note how social action is never in isolation for Garfinkel, as it might for Weber or Parsons.  While the latter note orientation toward others is required for social action, their model of social action still begins with an individual.  Garfinkel’s situations and contexts always involve two or more social actors in a social interaction.  Further, “any symbol or gesture considered apart from interaction is always ambiguous until specified as meaningful by being introduced into a specific, locally constructed context” (Cohen, p. 91). 


Cohen concludes the section on Garfinkel by discussing conversational analysis (CA) – “the notion that talk has generalized forms that influence the content of interaction, leaving all questions of meaning aside” (Cohen, p. 91).  This form of analysis takes ordinary conversation as a form of social interaction that requires work and energy, has tacit procedures, and results in accomplishments of particular forms of social interaction that become expected and repeated.  These are often interpreted as roles or norms.  For CA, Cohen notes there are some patterns and strategies – for example, insertion of a question will lead to some response to the question, or no response.  In any case, the question becomes an important interjection and one that other actors must address.  “CA thus embeds the cognitive problem of order as deeply as possible in the practices of daily life” (Cohen, p. 92). 


Cohen also notes some criticisms of CA, saying that it may result in emphasizing form over content and neglect other social practices.  Another problem is the equality implicit in the conversations provided by Garfinkel.  In bureaucratic situations or in situations with inequality among participants, conversations often take quite a different form and direction.  At the same time, CA might help to understand some of these situations – example of these are provided in the next section.


5. Application of ethnomethodology to sex and gender inequality

a. Accomplishment of gender

In the article, “Power, Inequality, and the Accomplishment of Gender: An Ethnomethodological View,” Candace West and Sarah Fenstermaker use an ethnomethodological perspective to examine how people accomplish gender in everyday life.  They provide examples of conversational analysis and participation in other daily tasks to demonstrate that analysis of individual characteristics or gender roles “obscures the work involved in producing gender in everyday activities.”  Rather, “gender is an accomplishment: an emergent feature of social situations that is both an outcome of and a rationale for the most fundamental division of society” (WF, p. 151).  In doing this, they demonstrate the power and inequality associated with sex and gender interaction and comment on the possibilities for social change in forms of such interaction. 

WF define ethnomethodology as follows:

Ethnomethodology proposes that the properties of social life which seem objective, factual and transsituational, are actually managed accomplishments or achievements of local processes.  ...  The aim of ethnomethodological inquiry is to analyze the situated conduct of societal members in order to see how ‘objective’ properties of social life are achieved.  (West and Fenstermaker, p. 152). 

While there are commonly expected norms regarding appropriate social action and, in particular, actions consistent with gender expectations, “the ‘meaning’ of these conceptions is dependent on the context in which they are invoked – rather than transsituational” (WF, p. 153).  WF argue that conventional approaches to gender, “including its interpretation as an individual attribute and as a role” (WF, p. 151) are mistaken. These approaches result in reducing gender to sex and thus imply that gender roles may be natural ones, or at least ones that are relatively fixed after approximately age five (WF, p. 154). In contrast, West and Fenstermaker argue that gender should be understood in an ethnomethodological manner

as an accomplishment: an emergent feature of social situations that is both an outcome of and a rationale for the most fundamental division of society. (p. 151). ... gender is not merely an individual attribute but something that is accomplished through interaction with others. (WF, p. 155)

In order to begin this analysis, West and Fenstermaker begin with three distinctions – sex, sex category, and gender.  These are as follows (WF, p. 152):

·        sex is biologically determined and ascribed at birth

·        sex category is the social classification of individuals into men and women

·        gender is the conduct that is accountable to normative conceptions of womanly or manly natures.

They reject concepts of sex or gender roles because (i) actions appropriate for roles cannot be specified in advance of their occurrence, (ii) almost any action can be regarded as displaying a sex role, (iii) roles imply complementarity of males and females and ignore power differences (WF, p. 154).  In particular, they object to the view that “masculinity or femininity [is] inherent in a person” (WF, p. 153).  Instead, they argue that gender is not a characteristic that is merely attached to individuals, but gender is something accomplished through interactions with others.  This makes their focus of investigation the activity, the types of interaction, and the institutions and structures within which these interactions take place.

While individuals are born with a particular sex – male or female – sex category is socially constructed by others in the new born male or female, often soon after birth and continued through life.  West and Fenstermaker note that “we treat appearances (e.g. deportment, dress, and bearing) as indicative of underlying states of affairs.” (p. 156). Many could pass for members of the other sex, but in our interactions we use visible characteristics as means of identifying the sex of individuals.

In terms of accomplishing gender, they draw on the idea of accountability.  That is, members of society take notice of activities and place them in a social framework.  These involve both activities that conform to accepted norms and those that might be considered deviant.  In the ethnomethodological framework, all of this is in social interaction, so any accomplishment is “interactional accomplishment” (WF, p. 156) within a social situation and in context (WF, p. 157).  That is, “doing gender consists of creating differences between girls and boys and between men and women – differences that are neither natural nor biological” (WF, p. 159).  But once such differences become established, they are often regarded as natural, become expected of social actors, and are difficult or slow to change. 

b. Examples

i. Conversational work

In conversational analysis, the issue of allocation involves “opening and closing a state of talk, maintaining an orderly exchange of turns at talk, and providing for a steady stream of conversational topics” (WF, p. 159).  In their analysis of exchanges among unacquainted persons, one problem was to maintain the flow of conversation – often the conversation may have stopped without initiation by one or other parties.  In terms of changes in topic, they report that in conversations between men and women men were more likely than women to unilaterally changed the topic of conversation.  This was especially the case when women appeared ready to report their point of view on particular topics.  WF regard this as evidence that “the enactment of women’s and men’s essential natures conditions what gets talked about (and by whom)” (WF, p. 161). 

ii. Household work

Another example they discuss is household labour.  In the case of couples where both husband and wife are in the labour force, the wife may still do the bulk of household work. This does not indicate a rational or efficient division of household work, or a division based on who is most skilled.  “Rather, it is determined by a complex relationship between the structure of work imperatives and the structure of normative conceptions of that work as gendered” (WF, p. 162).  That is, “household labor is regarded as women's work, but that for a woman to do it and a man not to do it draws on and affirms what people conceive to be the essential nature of each” (WF, pp. 162-163).  This is shown by the actual division of labour and the justifications for it.  A wife may say that it is not the job of the husband, or that the husband is not very good at the job.  In turn the husband may not become skilled at this work, or may even deskill himself.  West and Fenstermaker point out that household work thus is two processes together – production of household goods and services and accomplishment of gender.  See quote, bottom of p. 162.

iii. Parenting

Another example they cite is parenting in cases where parenting is shared more equally. The normal expectation, of course, is that the mother is the primary care giver.  For those couples where parenting was more or less equally divided as a responsibility, both the mothers and the father argued that each parent could provide emotional closeness to the child.  For couples where the women took greater responsibility, the explanation they used was that there are essential sex differences in ability to parent.  In addition, all couples regarded themselves as accountable for the parenting arrangements, that is, accountable to others.  One father who took his young children to public places was told that it was so good to see fathers taking on these responsibilities – thus making him a generic liberated father rather than an individual.   See quotes on p. 164.

iv. Emotional labour

West and Fenstermaker also cite Hochschild’s findings about airline attendants, noting that “the job of a flight attendant is something quite different for a woman that it is for a man.  … women flight attendants served as airline ‘shock absorbers,’ placating and soothing mishandled passengers” (WF, p. 165). “By contrast, men flight attendants were used as authority figures” (WF, p. 165).  

c. Conclusion

What West and Fenstermaker argue is that “Gender ... is a situated accomplishment: the local management of conduct in relation to normative conceptions of appropriate attitudes and activities for particular sex categories” (West and Fenstermaker, p. 156). Each situation we find ourselves in means we may relate to these norms, or accountable to these norms, but “we cannot determine the relevance of gender to social action apart from the context in which it is accomplished. ... While sex category is potentially omnirelevant to social life, individuals inhabit many different social identities that may be stressed or muted, depending on the situations.” (West and Fenstermaker, p. 157).  In this approach, there are no fixed male and female roles, but rather people have the view that “women and men possess essentially different natures, for which they will be held accountable in human affair.” (West and Fenstermaker, p. 159).

In light of this, the ethnmethodologist studies particular situations and the activities associated with them.  This study is not so concerned with the meanings of the situation, as the interactionist would, but with the activities and work that make meanings possible.   The actual set of actions and behaviour that is undertaken is what is important in this approach.  An ethnmethodologist may thus study conversation, interactions between individuals, the manner in which people “do walking” (Ritzer, p. 400), or generating applause (Ritzer, p. 404). 



Garfinkel, Harold, Studies in Ethnomethodology, Englewood Cliffs, New Jersey, Prentice-Hall, Inc., 1967.

Heritage, John, Garfinkel and Ethnomethodology, Cambridge, Polity Press, 1984.

Hochschild, Arlie Russell, The Managed Heart: Commercialization of Human Feeling, Berkeley, University of California Press, 1983.

Sharrock, W., "Individual and Society," in R. J. Anderson,  J. A. Hughes and W. W. Sharrock, Classic Disputes in Sociology.  HM51 C54.

Sharrock, W. and B. Anderson, The Ethnomethodologists  HM24 S424

Turner, Jonathan H., The Structure of Sociological Theory, fifth edition, Belmont, Ca., Wadsworth, 1991.  HM24 T84

West, Candace and Sarah Fenstermaker, “Power, Inequality, and the Accomplishment of Gender: An Ethnomethodological View,” in Paula England, editor, Theory on Gender / Feminism on Theory, New York, Aldine de Gruyter,1993, pp. 151-174.

Last edited on January 27, 2003

Return to Sociology 319.