Participant Observation of Alcoholics Anonymous: Contrasting Roles of the Ethnographer and Ethnomethodologist
Seán O'Halloran

The Qualitative Report, Volume 8, Number 1 March 2003


This paper is an attempt to explore the possible research stances available to the researcher involved in participant observation of Alcoholics Anonymous (AA). It examines some ethnographic studies of AA, within both naturalistic and symbolic interactionism research paradigms. However, mindful of the constitutive nature of language in social interaction and also wishing to focus on AA as a discoursal process, ethnomethodological approaches are examined, particularly the insights available through Conversational Analysis (CA) and Institutional Interaction. The methods of scrutiny available through these approaches - the emphasis on data and fine-grained detail as well as the context sensitivity available to the acculturated observer - it is argued, make the ethnomethodological approach eminently appropriate in terms of exploring the reflexive relationship between AA discourse and its social organisation.

Key Words: Participant Observation, Alcoholics Anonymous, Ethnography, Ethnomethodology, Conversation Analysis, Institutional Interaction, Insider / Outsider


The purpose or this paper is to explore the positions of the insider and outsider in social research and relate these to ethnographic and ethnomethodological research methods involving participant observation of Alcoholics Anonymous (AA).

The paper will focus on method, which though determining what is known is itself determined by a particular way of seeing. It is useful to try to conceptualise and locate possible approaches according to the four paradigms Gubrium and Holstein (1997) refer to as four idioms of qualitative inquiry, namely naturalism, ethnomethodology, postmodernism and emotionalism. This paper will explore the first two of these approaches from the point of view of insider observation of AA. It will look at some existing research and argue that AA, being constituted mainly through discourse, may be effectively approached through ethnomethodology, focusing particularly on the form of Conversation Analysis (CA) known as institutional interaction.

The paper is also an exploration of what Collins' (1991, p. 53) calls a 'personal biography'. She sees 'personal and cultural biographies as significant sources of knowledge' for 'outsiders within the academy'.


In exploring the role of inside observer, it is useful to start by examining Naturalism, being the most fully established of the four idioms of qualitative inquiry referred to above. According to Gubrium and Holstein (1997, p. 6), 'naturalism seeks rich descriptions of people and interaction as they exist and unfold in their natural habitats'. Implicit in traditional forms of naturalistic ethnography is the notion of a coming together of two essentially discrete entities; one being the researcher and his/her methodology and the other the participants and their social world. Room (1993) in her studies of AA as a social movement takes a macroscopic perspective on AA as a structured social entity and tries to locate it in the context of history and other social movements in the United States. She looks at its formative influences, membership, impact on society and its organisational structure, remarking 'AA has succeeded in creating an organisation that breaks Michels' "iron law of oligarchy" (1958, p. 171) by building in structures and principles that minimize the professionalization of leadership and keep effective organisational power at the level of egalitarian face-to-face interaction". Using such a wide-angled lens, we are presented with an overview in the context of history and society in general. However, the voices involved in this 'face-to-face interaction'; are not heard nor their emotional world felt. For this, a more ethnographic approach is required. Neither does Room attempt to give insights into the possible constitutive reflexivity of the relationship between the organisational structures and that face-to-face interaction that an ethnomethodological approach may explore.

Robert Park and the Chicago School of field research provide a research model which involves accessing those voices and their worlds. One study of AA which broadly represents this approach is Milton Maxwell's The AA Experience (1984, p. ix). He describes his approach as, 'essentially anthropological - that of both participant observation and the use of knowledgeable informants'. By using AA members' accounts to describe the process of becoming an alcoholic and what active alcoholism entails, he allows the perceptions of the subjects under study to inform that study, thus to some extent breaking down the positivist dichotomy between the researcher and the objects of research. The following extract is instructive and cautionary. It involves an AA member's account of his alcoholism and Maxwell's own commentary on it:

I could neither control the drinking which had brought me to this state, nor could I leave it alone. But, at that time, it was very hard for me to accept the reality of all this - that this had really happened to me. Extremely difficult! What's more, it was impossible for me to picture a satisfactory life without alcohol.
How does A.A. meet such an alcoholic person 'where he is at'? How does A.A. help such a person to reconstruct his world - his total field? (p. 37)

Very interesting questions about the nature of representation, viewpoint and the resultant questions arise from this brief extract. First, it is interesting to note how the AA member represents the stage of his active alcoholism. It is done from a perspective of quite considerable distance. It is characterised, as are most AA stories, by persistent use of the past tense and past prefect aspect when referring to the period of active alcoholism, suggesting not only distance in time, but also in distance from the speaker's present reality. This is reinforced by phrases like 'at that time' and the strong suggestion of a person viewing their active alcoholism from the outside - 'it was very hard for me to accept the reality of all this - this had really happened to me', (Maxwell's italics). Such a stance is only acquired through time and recovery. AA stories disclose, according to AA's Preamble, what the alcoholic used to be like, what happened and what he is like now. As such, it can be argued, they are constructs - powerful tools that separate the perceptions and behaviours of the alcoholic in recovery from those the active alcoholic. They are distancing devices, refashioning the past as distant and fashioning a new view of reality for the narrator. They are also highly interpretive, in that AA stories are expressed in terms of the AA programme and concepts. Phrases like, 'I could neither drink nor leave it alone', 'hard to accept reality', 'a life without alcohol', are the stock and trade of AA discourse. As such they represent the view of an acculturated member of AA. It is unlikely that this informant would have described this particular period of his life in the same terms while he were still living it, or subsequently if he were not in recovery, or if he had achieved recovery through some other route. The telling of the past reveals perhaps more about the teller's current state of knowledge than it does about the events being referred to.

Thus Maxwell's question, 'how does A.A. meet such an alcoholic person 'where he is at?' is a problematic question. Maxwell goes on to ask, 'how does A.A. help such a person to reconstruct his world - his total field?', the answer must be that it already has. The 'person' who emerges from the narrative has been fashioned in recovery stories, has been reinterpreted through AA perspectives. AA stories tend to construct the teller's stories around an archetypical narrative. They are an interesting insight into how members have constructed their past. However, Maxwell does not use them as such but as data for describing the developmental stages and nature of actual, active alcoholism itself. Like other traditional sociologists, he uses use such material as a source of data. An ethnomethodologist would approach it as essentially indexical in nature (Garfinkel, 1969) and as itself the proper object of sociological enquiry.

Such a post-modern approach reads texts not as 'findings' but as constructs to be interpreted (Rosenau, 1992). Participant accounts in such research are viewed not as sources for discovering the social reality of the member, where it is located and why, but as a resource to find out how the members, through their action and particularly their talk, construct their views of reality (Gubrium & Holstein, 1997).

Another ethnographic study of AA is that of David R. Rudy, Becoming Alcoholic (1986). Rudy opens his study with the following remarks:

One of the most anxiety-producing and yet exciting aspects of doing field research is the uncertainly of direction. Rather than beginning a study with a specific question, most field research starts with a setting, an arena of social life, and allows that setting, its participants, and the researcher's perceptions of these to forge themes, questions, hypothesis, and grounded theories. (p. 1)

Here the researcher emphasises the importance of entering into the participants' setting and observing them closely, keeping researcher preconceptions to a minimum. The 'setting' is represented as geographical - a location. Rudy chose to observe an AA group in a location which he named 'Mideastern City' to secure the anonymity of the AA participants involved. But the name is also surely honouring the traditions of naturalistic research, epitomised by William Whyte's (1945) Street Corner Society. This is a classic study of an Italian-American neighbourhood, Cornerville, and its inhabitants in 'Eastern City'. Like Whyte, Rudy chose to go among the group he was studying. He describes a great number of features of the settings in which AA members met and the types of meetings they held. He reports establishing rapport with fifteen to twenty persons, and choosing to do life histories of six of these.

Some forty years after Whyte's ethnographic research which attempted to faithfully reflect the informants' perspective, Rudy's approach to his participants' stories reflects postmodernist concerns about how the stories convey and constitute the informants' realities. He expresses a particular interest in explaining what Schur (1971) refers to as "retrospective interpretation" - the types of new explanations members of AA learn in order to reinterpret their lives. Unlike Maxwell (1984) he does not use participants' accounts as a source for data on the nature of alcoholism. In fact the entire study in framed within the idiom of symbolic interactionism, focusing on how participants construct and respond to their worlds based on the meanings they develop through interactional processes (Mead, 1934). Whereas Whyte's study of Cornerville is replete with rich descriptions of the physical settings, a symbolic interactionist approach emphasises the importance of the symbolic as well as physical environment; the main features of the symbolic environment being language, others, self and interaction. Rudy entered the 'field' with no 'specific question', he used members' accounts, careers, values and slipping (relapse) behaviour to formulate a sociological definition of alcoholism, not to describe the reality in the terms of those observed, but to generate data from which grounded theory may emerge. This is both a grounded (Glaser & Strauss, 1967) and an interactionist or subjective approach to alcoholism, presenting alcoholism as a 'relative phenomenon that is constructed', sharply different from a positivist approach which treats alcoholism as a medical entity with its own discoverable aetiology (Rudy, 1986, p. 95).

Insider Ethnographer

Rudy as an outsider used a 'grounded', approach to AA, attempting to avoid preconceptions prior to question formulation. An insider turned researcher could not hold such a position having being implicated, perhaps unconsciously, in the culture of AA prior to framing research questions. Pike (1954/1987) has usefully given us the distinction between the emic and the etic perspectives. The emic approach uses the language and the categories of the people studied; the emic approach involves the researcher using categories devised by the researcher and emanating through analysis. Tedlock (1991) gives methodological positions to both stances. He gives the term observation of participation where ethnographers both experience and observe their own and others' co-participation within the encounter, and the term participant observation to research where the aim is to be emotionally disengaged and coolly dispassionate.

All social researchers have their personal histories and stance. Messerschmidt (1981, p. 8) sees 'the extent of relative "insiderness" and "identity" between research and subjects is best conceived of as a continuum from virtual oneness to marginal nearness'. Indeed this continuum can be seen in the fact that AA is not a society one is born into. AA membership naturally coexists with other aspects of a member's identity and very member has as a resource the memory of the initial feeling of being new and the questions which arose from early participation in AA. For the ethnographer the point of at which questions are framed as research questions is on initial contact, for the AA insider researcher it is later and involves negotiating between two differing epistemologies, one arising within recovery and essentially spiritual in nature; the other located in the academy and manifest in research method. A bicultural stance must therefore be sought, where each epistemology informs the other.

An unconscious culture does not only affect the perceptions the insider (Aguilar, 1981). As researchers with our own histories and multi-faceted identities 'we both bring the self to the field and create the self in the field', (Reinharz, 1997, p. 3). Indeed, Denzin, (1989, p. 81) who has studied AA extensively, comments that an AA researcher is like the AA newcomer in that he must seek to become a knowledgeable member of the social structure being studied.

Rudy implicitly challenges the bipolar construction inherited from positivist science that a researcher is a deus ex machine, a privileged outsider looking in, which sets up a separation that neglects the interactive processes through which 'insiderness' and 'outsiderness' are constructed. He states, 'I moved from a tolerated intruder, an outsider to a near-member', suggesting that as ethnographer he was neither fully inside nor outside the community. He goes on to say, 'I not only participated and observed the processes in which individuals came to regard themselves as alcoholic, but I also experienced some of these processes myself.' (1986, p. 3). He claims a particular validity to his participation as he has experienced some of the process himself, thus acquiring some of the perspectives of his research subjects. Aguilar (1981, p. 16) points out that 'the covert culture of the insider has a heuristic value of lending psychological reality (or cultural validity) to ethnographic analysis.'. Denzin (1989, p. 120 ) adds that 'shared and shareable emotionality lie at the centre of the process of understanding'. It could be added that it lies at the centre of AA therapeutic discourse as well.

These positive aspects of the role of insider as researcher are discussed by Adler and Adler (1997) who exploit their role as parents in their ethnographic studies of children and schools. They point out distinct advantages in the ability to capitalise on the 'complete membership' role of parents have as researchers of children. They discuss many scholars who have fruitfully made their personal arena a focus of research. The researcher who is a true member does not have to negotiate either formal or informal entrée with other members and will be already intimately familiar with many members and the meeting venues. An AA insider has full access to the meetings closed to non members. Reinharz (1997) in viewing a researcher in terms of self and role, points out that the researcher when seen as 'outsider' is identified in terms of his role, not in terms of his self. The social attributes that affect the researcher's self, the perspectives and the attitudes of the researcher - issues of race, age, gender, class - are likely to be the main concern of those being researched. In the case of AA groups, the researched are likely to be interested in the researchers' attitudes to and experience of alcohol and alcoholism and tolerance of a spiritual view of life. Thus the insider has an advantage in terms of access, in that those being researched are less likely to fear being appraised or judged and thus withhold parts of themselves or alter their behaviour in accordance to what they perceive as researcher's expectations.

Collins' (1991, p. 53 ) sees 'personal and cultural biographies as significant sources of knowledge' for 'outsiders within the academy'. Her insights are particularly pertinent within the context of which she is talking (i.e., of African American woman within academe). However, they are not altogether misplaced for the AA member seeking to account for personal experiences and histories in a framework of academic enquiry. The bipolar construction 'insider' and 'outsider' fails to capture the additive nature of what happens when insider becomes researcher. Western academic traditions are imbued with the legacy of objectivity, positivism and respect for the empirical. AA members, however, are implicated in a highly subjective, seemingly intangible process which had wrought verifiable change in their lives. They use concepts like God or a higher power to give meaning to their experience which may appear intellectually inexplicable. The acquisition of the role of researcher may not be change in the self but may involve the application of an additional epistemology.

The post-modern assault on the assumption of ethnographic objectivity and research subject subjectivity means that an AA member turned researcher does not have to be so heedful of the reverse of the ethnographer's exhortation not to go native, as theirs is the case of the native going ethnographer. Much the same can be said of the native going ethnomethodologist. The latter position is less fraught with methodological problems. The ethnomethodologist relies more on tape recordings of actual language used in interaction than field notes, thus using empirical, verifiable and incontrovertible data. Furthermore, to be able to interpret such data the analyst must in some sense be an insider to have an understanding of the commonsense view of the acculturated members being studied as well as display cultural competence in recognising how talk-in-interaction is being carried out.

Discoursal Practice and Social Context

In turning attention from naturalistic ethnography to a more language orientated ethnomethodological approach, it can be seen how the methodology of naturalist ethnography misses a significant aspect of 'setting'. How one frames the setting of the social group under investigation impacts on the nature of the data generated and how it is represented. Whyte was able to locate his Italian-American social group firmly in its neighbourhood Cornerville, and though Rudy does not attempt to locate AA solely within a particular geographical setting, he seems unconcerned about the discoursal setting of his data. It is often not clear if the spoken data he refers to derives from interviews or meetings. This neglect of the discoursal framing of data in observation of AA reflects an indifference to one of the key aspects of how AA is constituted.

Such an approach implies context is a container where pre-existing hierarchical features within that context exert causal forces over available actions. A more ethnomethodological approach would explore how participants' interaction constitutes a context which enables them to accomplish discoursal events for their own particular ends.

I have argued elsewhere (O'Halloran, 2000) that part of the effectiveness of AA is that it not only nurtured a particular type of talk but that its founders showed a profound awareness of the constitutive nature of discourse on social settings in developing the social structures of AA. AA's Twelve Traditions and discourse conventions preclude displays of personal status, distinction or authority (Room, 1993). The resulting organisation is polycephalous, egalitarian and anarchistic, effectively resisting the development of oligarchy and professionalisation of leadership (Mäkelä, 1996) and thus preserves the original mutual features of AA talk. AA would appear to offer a spiritual home to alcoholics of the type described by Bateson, (1971/1985) who describes the alcoholics' basic fault as an erroneous self-concept that is experienced as a hierarchical, compartmental structure in which interests are pitted against one another. Further consideration will be given to this in the next section.

AA as a Dialogical Discourse Process

The exploration of such reflectivity is central to ethnomethodology. Reflexivity refers to the practices that at once describe and constitute a social framework (Coulon, 1995). To clarify this in relation to AA it is necessary to look at some aspects of AA itself. The founding of AA is attributed to two conversations - dialogues. One was in late 1934 when Bill W, a cofounder of AA, was approached after years of career-ruining and marriage-threatening drinking by a former drinking partner Ebby T who had 'got religion' (Anon, 1939/1985, p. xx) and managed to stay sober through his contact with the Oxford Group, a Christian group which aspired towards the spirit of early Christianity. Bill was deeply impressed by Ebby's sobriety, but more so by their conversation, 'I could not forget what he said. In the kinship of common suffering, one alcoholic had been talking to another,' (original italics). Ebby had not tried 'to pressure or evangelise,' (Anon, 1986, p. 59), and though suspicious of the talk of religion, Bill was impressed and went to Oxford Group meetings. He was deeply fascinated by what happened when one alcoholic talked to another. He felt certain such talk supported his sobriety. Therefore, a few months later, when in a strange town, having just failed in a major business negotiation, and feeling vulnerable, he sought out another alcoholic through contacts within the local Oxford Group. Thus he met Dr Bob, regarded as the other co-founder of AA, who had joined the local Oxford Group to overcome his drinking problems. Their conversations are seen by both men as crucial to their recovery and the formation of AA itself. Bill later wrote:

You see, our talk was a completely mutual thing. I had quit preaching. I knew I needed this alcoholic as much as he needed me. This was it. And this mutual give and take is at the very heart of all of AA's Twelve Step work today. This was how to carry the message. The final missing link was located right there in my first talk with Dr Bob. (original's italics) (Anon, 1986, p. 70)

Later Bill W wrote, 'when one alcoholic had planted in the mind of another the true nature of his malady, that person could never be the same again,' (Anon, 1952, p. 23). In studying AA, therefore, it is necessary to approach it not only as a social entity but as a discoursal process, which itself is somehow constitutes the social entity in which it occurs.

Bateson (1971/1985) argues that 'the 'sobriety' of the alcoholic is characterised by an unusually disastrous variant of the Cartesian dualism, the division between the Mind and Matter - between conscious will, or 'self' and the remainder of the personality'. It is in perpetual intoxication that a subjective correction of this is made. He feels that Bill Wilson's 'stroke of genius' was that in AA's first step a declaration of personal powerlessness broke that dualism, thereby changing the alcoholics' false epistemology', (p. 313). AA, he argues, changes this contextual structure by permanently placing alcoholism within the self, and therefore something to be lived with.

Using the phraseology of religion rather than AA, Bateson (1971/1985, p. 326) claims 'the religious conversion' of the alcoholic when 'saved' by AA can be described as a dramatic shift from a symmetrical habit, or epistemology, to an almost purely complementary view of his relationship to others and the universe or God. The alcoholic no longer battles against addiction or life but accepts addiction as integral of self and self as part of life. Hitting bottom, Bateson sees, as a spell of panic which provides a favourable moment for change. It is the 'double bind' described by Dr Silkworth in the foreword to Alcoholics Anonymous, (Silkworth, 1939/1985): the obsession of the mind which compels the alcoholic to drink and the allergy of the body that condemns him or her to go mad or die. This discovery forces the alcoholic to a 'point at which only an involuntary change in deep unconscious epistemology - a spiritual experience - will make the lethal description irrelevant', (Bateson, 1971/1985, p. 331).

Bateson relates his discussion to aspects of AA literature, particularly terms like 'higher power', and the slogans 'hitting bottom and 'an alcoholic can't drink like other people'. His approach is to explicate the dynamics of cybernetic systems. He makes only limited use of data derived from AA literature and from interaction between AA members. An ethnomethodological approach informed by an analysis of Institutional Interaction can use naturally occurring spoken data derived in situ during AA meetings to explore how AA members display and therefore constitute their alignment with their past, each other, the AA programme and processes.

Researching AA Talk

Denzin (1989) in his interpretative interactionist research into AA refers to it as a distinct linguistic community and describes the steps necessary to learn its language. This, he says, involves isolating and noting how different members use key terms, reading AA literature and studying how language is connected to personal experience. He approaches AA language as a means whereby it is understood and interpreted. He talks of bringing 'lived experience before the reader,' (p. 83) by creating a richly descriptive text, a 'thick description', which allows the reader to share vicariously the experiences that have been captured from which the reader can naturalistically generalise. Furthermore, the researcher uses language to describe this reality and make it accessible to the reader. In other words, the researcher explores the language of AA in order to understand and interpret its nature i.e. language is a seen as a representation and reflection of a social reality. For him 'language structures and creates the process of understanding and interpretation' (p. 72). Powerful as such an approach may be, it does not analyse how language itself constitutes the very social phenomenon under study.

Field Notes

At the start of the meeting there were only a few regular male members (Stan, Ted) present, and the female newcomer. During the opening readings a regular lady member (Lynn) arrived. The theme for the reading was 'Learning to love ourselves'. During the reading the newcomer cried quietly and apologised with conflicting suppressed tears and nervous laughter in her voice.

After the reading the chair (Stan), discovering the presence of a newcomer, spoke of his own experience of drinking and early recovery, particularly in relation to resentment and developing relationships. Another member (Ted) also spoke, addressing his remarks more directly to the newcomer. This is a stylistic variation, as usually in AA meetings sharing is not addressed to a specific audience, except in specific cases like the presence of a newcomer, or if some one has brought up a specific problem early in the meeting.

The newcomer then, without the usual self-introduction, asked about religion and spirituality, saying she knew in AA people said it was not a religious programme but a spiritual one, but she had not much of an idea about either. Her talk was marked by frequent pauses and eye contact indicating that she wanted a response. After about three of pauses, transition points which were not taken up, Ted responded in a light vein about how religion was full of rules but in AA there were none. He took up an explanatory tone, again usually avoided, except when the audience is a newcomer, prefacing them with 'you see' and 'you know', but tempering the implicit authority role with humour and self-deprecating comments, 'no one will know how insane I am unless I pick up a drink.'

At this point the she asked, "Should I get a sponsor?" The chair, avoiding giving direct advice as the request required, shared about his own experience in finding a sponsor and the importance of that relationship to his recovery, but did not respond directly to the question, indicating a reluctance to respond directly to a request for advice but instead used the conventional AA illustrative personal recount.

The new-comer again stated that she had been to only two other meetings and had never shared, that she just sat at the back of the meetings she had been to which had about 20 people in them. She said she was pleased that this was a small meeting where she felt comfortable. Again her eyes and pausing indicated a desire for others to take a turn, which did not happen so the chair asked her if she would like to share.

At this point she entered her first sustained episode of monologue, recounting how she had recently been told bluntly by her doctor on her admission to hospital that she was an alcoholic, that her alcohol blood level was 'ridiculously high' and that she had been to an AA meeting a few years previously but did not go back. She also recounted how going to meetings recently (in the last few days) had helped her cut back on her drinking and how supportive her husband was, saying he was in the Al-Anon meeting next door, indicating this with her eyes, head and hands. She added that nonetheless he could not understand why she drank so much. She made explicit reference to the earlier reading and sharing by saying how she was incapable of self-love. She also made explicit reference to the chair's talk in referring to her inability to stop drinking once she had had one drink.

Throughout this episode she cried quietly, frequently drying her eyes and nose with a tissue offered by Lynn, which she accepted without apology or thanks. This episode of talk, with its strong elements of self-revelatory narrative and its focus on her drinking experience, indicated a clear shift toward the characteristics of AA talk.

This was also indicated by her references to the initial reading and other members' sharing. There were no longer invitations to others to take a turn either through pausing or eye contact. However, on at least two occasions she did say she had never talked like this before, her eyes moving upwards to make direct contact with a particular person as her audience.

Neither were there any specific questions or requests for advice - which had characterised her earlier talk. The abandoning of these mechanisms inviting other turns with the attendant seeking of a specific interactive partner allowed her to sustain her talk for about five minutes and recount events and display emotions normally avoided. However, when she stated she did not want to admit she was an alcoholic and that the idea of never drinking again seemed impossible to her, her eyes sought particular partners with whom to interact; suggesting the need for reassurance, advice or acknowledgement, and as such at variance with much AA interaction.

The resulting pause and failure to close her turn resulted in brief responses from the others. As direct responses to her unclosed turn, they were seen to be directly contingent on the content of the previous turn - again, not typical of AA interaction. Stan suggested she need not think of drinking for the rest of her life but to take it 'a day at a time'. Ted recommended she went to meetings regularly. The first attempt to interact by Lynn was immediately seized on by the chair who remarked, 'would you like to share?' This giving of the turn to Lynn resulted in a return to an episode of sustained AA sharing, where Lynn oriented her remarks to the new-comer's concerns by relating her own particular experiences.

When the new-comer again interrupted with a direct question, the resulting responses showed another return to adjacency pairing, the chair suggested they close the meeting early so they could simply 'chat' for the rest of the time. This suggestion was responded to by vigorous nodding of the head by Lynn. After the usual closing rituals - reciting the Serenity Prayer in a ring while holding hands - the meeting was closed and the group engaged in interactive conversation involving questions and explanations, rather than single turn monologues.

The chair, it appears, had been unwilling to direct and prescribe the rules of discourse; however, he also appeared to be reluctant to allow the meeting to proceed with a lot of cross-talk and with direct questions taking place. He had steered the meeting to extended sharing on two occasions, this had broken down. Everyone had in fact shared so the meeting was brought to an early end, not, it would appear, because the interaction had finished but because the chair had decided the nature of the interaction required was not the type which could take place within an AA meeting. This poses two questions:

  1. What had taken place in the interaction that was considered should not take place in the frame of an AA meeting?
  2. What are the typical features of AA interaction during a meeting?

There are 2 possible ways of answering these questions.

  1. The ethnographic way - ask the members about what they do (i.e., seek out their explicit knowledge of their own categories)
  2. The ethnomethodological way - examine the discourse itself in great detail to unravel what the members do to seek out the implicit rules which govern the practice. This being a 'breach' situation is illuminating but actual recorded data is required. Is it possible to record a newcomers meeting? Probably not.

In order to clarify this methodological distinction, it is worth examining a description of an AA meeting I made some time ago (see boxed Field Notes). The field notes in question are 'a thick description' based on observation. In them a newcomer is observed and described as persistently breaching what seem to be largely unnoticed 'rules' which govern AA talk. Though four of the turns involved sustained monologues, including one such turn by the newcomer, each sustained turn was followed by adjacency pairs involving questions and advice seeking from the newcomer. It appears that the elements that sustain an AA meeting were not established as the newcomer was engaging in speech acts not usually associated with an AA meeting.

Speech Acts

The following account of an AA member's response to the talk of a group from a treatment clinic patients participating in an AA meeting further illustrates that members have shared expectations of discursive practices in an AA meeting.

I can't take that fucking group. The members don't know how to talk. They interrupt. They talk out of turn. They gossip, they give advice. (Denzin, 1997, p. 334).

Clearly norms were felt to have been violated. As in the case of the meeting under discussion, they involve breaching norms about permissible illocutionary acts - gossiping and giving advice, as well as how to take turns. Austin (1976) has given us tools to explore the former. He has indicated that we do certain things with language. Searle's (1960) concept of the speech act is a powerful tool to help discover what exactly people are doing with words in an AA meeting. Similarly, in Conversation Analysis (CA) utterances are seen as objects which speakers use to accomplish certain things (Hutchby & Wooffitt, 1998). The field notes and the above comments from an AA member indicate that AA members display and orientation to certain illocutionary acts and avoid others. The newcomer's questions and advice seeking were not responded to and more experienced members tried to establish other forms of discourse. From this it may be inferred that the newcomer was not complying with certain establish norms of interaction which constitute an AA meeting.

Conversation Analysis

Though such close observation, note taking and a thick description of the interaction may supply good prima facie evidence of the existence of order and that the order can be located, it does not supply us with the machinery for securing co-ordinated talk or discovering how the 'co-ordinating machinery' actually works (Lee, 1987), or, as in this case, why it breaks down. Sacks (1984, p. 22), using recorded interaction, has shown that talk is subject to formal description and that social activities are methodological occurrences. He sees talk as its own social process governed by its own regularities; he has demonstrated there is 'order at all points' in talk-in-interaction.

The idea of regularities suggests that one is dealing with the deductible, predictable and therefore verifiable (i.e., rules). Searle (1960, p. 18) had proposed two types of rule, one that regulates antecedently existing forms of behaviour, such as etiquette, which can be expressed in imperatives. Rules governing the content of AA meetings seem to be of this type in that AA traditions dictate that meetings be confined to discussions of alcoholism. The other type of rule regulates and defines new forms of behaviour and is therefore constitutive, as with the rules of football. These rules however describe overt regularities. Neither of these categories of rules seems to capture what Sacks was referring to. Button (1990, p. 79) refers to 'rules that in their conduct people display and orientation to … the ethnomethodological rule does not precede the action. Rather, the rule is discoverable in the action.' A challenge for the observer of AA discourse is to deduce the type of rule use from situated practice in order to describe it. A participant observer taking field notes may detect prima facie evidence of the operation of such regularities but would not be able to uncover them fully. To do this an analysis of tape recording is required.

A number of researchers have noted how in AA meetings the turn taking of mundane conversation is suspended, (Mäkelä, 1996). However, this is not the case in the meeting discussed above where things have gone wrong. This methodological focus on 'breaches" is ethnomethodological in nature, being similar to the 'breaching' experiments used by Garfinkel (1986) to gain analytical access to the taken-for-granted, common sense knowledge which he sought to analyse through the examination of violated norms. It is important to note here that these are members' norms. They are recognised as such because they are procedurally consequential (Schegloff, 1992, p. 110) to the participants themselves and this is displayed in their interaction. It is though this display that the analysis discovers them. In this quest for members' norms, CA takes an etic approach (Pike, 1954/1987), however, the approach is also essentially emic and empirical. The mechanism of conversational interaction is what Schegloff refers to as an 'enabling institution'; the machinery that underlies the construction of conventions and is thus basic and primordial, not conventional or contingent (Zimmerman & Boden, 1993). The analyst discovers the norms displayed by participants in the details of recorded naturally occurring interaction rather than though eliciting members' categories through research specific activities. Norms may be so pervasive they are hardly detected, thus the methodological focus on violations. The violations noted in the newcomer's meeting were, however, naturally occurring, and therefore free from the criticism levelled at Garfinkel of contriving them. Contriving such an experiment in an AA meeting would be highly unethical.

CA provides useful insights into the nature of turn taking in interaction, (Hutchby & Wooffitt, 1998, p. 147). Using recorded data, it demonstrates that mundane conversation is characterised by the use of adjacency pairs which set up trajectories in subsequent talk. Here it appears that the newcomer, by engaging in illocutionary acts which required adjacency or near adjacency responses, is determining the content of the sharing; allowing one participant a degree of 'control' over the response of another, an authority role not usually taken in AA discourse.

In some AA groups, but not in the group observed above, turn taking in meetings is pre-allocated (Mäkelä, 1996). Thus the rules appear to be discoverable in antecedent prescriptions rather than in current action, as are ethnomethodological rules. This may be the reason Denzin (1987) felt that 'AA meeting talk does not admit easy analysis within the framework of CA'. However, Arminen (1998) has applied the methods of ethnomethodological conversation analysis to AA interaction and demonstrated, using materials from recorded meetings, how parties orient to the on-going interaction and compose their turns. He has also explored the formal interactional design of extended turns and the members' methodological ways to deal with the distribution of knowledge between participants.

Transcribing talk

The actual practice of recording and transcription is fundamental to examining any sort of talk-in-interaction. It is the necessary initial step which makes analysis possible as the production of the transcript is an analytical process in itself. The transcript is not viewed as the data, but a 'representation' of it (Hutchby & Wooffitt, 1998). The data is the actual interaction; audio or video tape recording is merely a 'good enough' reproduction of it, (Sacks, 1984, p. 26).

This is not to imply a definitive system of transcription exists. Transcription is a selective process based on theoretical goals and definitions. Most are variations of a common system devised by Gail Jefferson, (Have, 1999) which indicates as much detail as is theoretically desirable. With mundane conversation, the analytical concerns revolve around turn taking and involve marking precise beginning and end points, the duration of pauses and intakes of breath associated with an attempt at a turn. With the suspension of conversational turn taking in AA meetings, such a detailed study of turn taking may be unnecessary. However, much of AA talk has been shown to be marked by intertextual references to AA literature and slogans, (O'Halloran, 2000). This is an important aspect of how speakers orientate their experiences, past and present, to the AA programme and the experience of other recovering alcoholics. Marking such intertextual features, a practise derived from discourse analysis, would help explicate text and intertextual coherence and cohesiveness. Such however would be a departure from normal CA practice where only audible data is indicated through transcription and the focus is exclusively on language form.

Institutional Interaction

In Arminen's study (1998), the principles of CA have been taken beyond mundane talk to institutional talk. Institutional talk itself is seen as centrally and actively involved in the 'institutional' nature of the institutions themselves. What characterises interaction as institutional is the special characteristics of the speech exchange systems that participants orient to. As Arminen (2000, p. 436) points out, 'the analysis of talk-in-interaction in institutional settings aims at disclosing and specifying the verbal practices and interactional arrangements though which institutional practice is brought into being.'

CA, in its ability to explore how conversational order is accomplished is an invaluable methodology in this context in that it takes a comparative perspective whereby the turn taking of mundane talk is treated as a benchmark against which other forms of talk-in-interaction can be distinguished. It therefore enables talk in various institutional settings to be characterised by either a reduction of certain turn taking features or a systematic specialisation in the range of practices usually available in mundane conversation, thus revealing what is distinctive about interaction in different types of institutional contexts. Thus the institutional character of the interaction is embodied first and foremost in its form - most notably in the turn-taking systems which depart substantially from the way in which turn-taking in managed in conversation.

Like most institutional interactions, AA interaction is 'systematically asymmetrical' (Hutchby & Wooffitt, 1998, p. 160) when set against an idealised mundane conversation between equals. CA demonstrates how asymmetry in most institutional interaction show-oriented to structural patterns - for example in questions and answers - which furnish participants with differential resources, thus putting some participants in a more powerful position discursively. This asymmetry largely stems from the social and professional roles of members. Frankel (1990) has shown how patients are complicit in allowing doctors to determine the topics to be discussed and define their outcomes. Drew and Heritage (1992) similarly have shown how patients display a particular orientation to the expert status of a doctor. Turn type preallocation in court cases studied by Atkinson and Drew (1979) show how participants' interaction is constrained according to institutional roles imposed by the courtroom setting. These studies illustrate how CA offers the researcher data whereby the systematic deviance of a particular form of institutional talk from the norms of mundane conversation can be observed and analysed. By selectively reducing or transforming the scope of conversational practices found in mundane conversation, and by concentrating on some and withholding others, participants can be seen to display an orientation to particular institutional contexts.

Initial impressions from observation of the meeting referred to above suggest that abandoning adjacent turns in AA discourse produces such asymmetry. This would suggest that the asymmetries of AA interaction can be observed and analysed and inferences made about the resources with it furnishes participants in terms of solidarity building, avoiding power roles and expert status.

Such theorising may be unwarranted without more extensive recorded data, however it does suggest that the methodological approaches outlined are amenable to being framed in terms of the wider paradigms of Critical theory and Constructivism. Issues of power and discourse suggest the paradigms of Critical Discourse Analysis (Fairclough, 1995). Issues related to solidarity, mutuality and the spiritual have less clearly defined paradigms. Sidorkin (1999), exploring the works of Mikhail Bakhtin and Martin Buber, has attempted to outline the laws of the dialogical. Both Bakhtin and Buber claim the essence of true humanity is to be found in the dialogical. Buber states 'all real living is meeting'. The very fact of human existence is contingent upon engagement in dialogic relations. Furthermore, Buber sees the world as two-fold. The I-Thou refers to the realm of the dialogical relations, while I-It to the realm of the subject-object experiences.

In Buber's terms (2000), to impose authority, expertise or special knowledge on someone involves an I - it relationship. The I - thou relationship is the dialogic one and the one that appears to be sustained by AA discourse, where participants communicate free from distinctions of rank or knowledge, confining their talk to the sharing of experience. 'It (the dialogic) is about subjects only … there is no medium between the subjects; it is a direct relation.' (Sidorkin, 1999)

the only thing that matters is the fact that for each of the two (in the dialogic)… the other happens as the particular other that each becomes aware of the other and is related to him in such a way that he doesn't regard him and use him as an object, but as his partner in a living event. (Bakhtin, 1984, p. 40)

Little research has been done to exploring how to frame discourse in dialogical as well as critical paradigms, and I think none on how the dialogical is constituted through AA talk in meetings.


To relate this discussion back to the issue of the role of the insider it should be said that working within constructivist and critical paradigms requires the researcher to demonstrate the trustworthiness and credibility of the data, (Denzin, 1997). An ethnomethodological approach involving CA, with is emphasis on recorded data, supplies the element of confirmability required to establish this. Furthermore, CA requires the analyst display cultural competence in recognising how mundane talk-in-interaction can be carried out, (Arminen, 2000). The very concept 'mundane' suggests its very ordinariness and accessibility. An analysis of mundane conversation should illuminate the common understandings that are relevant to the participants and the practices that provide for those understandings, in that it describes both the knowledge that the participants use, and when and how they use it. It requires an 'insider' with intimate knowledge of the language system and its illocutionary and pragmatic subsystems as well as the culture and context in which it situated. The same 'insider' cultural competence is required in the analysis of institutional settings. Arminen has remarked, 'in institutional contexts, the disclosure of the context-sensitive meaning of the activities may depend on access to participants' knowledge or organisational particulars without which the analysis may remain insufficient' (Arminen, 2000, p. 437).

Naturalists have treated members' talk as expressing an underlying, shared, cognitive order, but have not explicated the ways in which talk is in itself and essential feature of the setting it describes. They have seen language as essentially descriptive and interpretive rather than constitutive (Gubrium & Holstein, 1997). Postmodernism has challenged the assumption that the researcher can act as neutral observer of things social. Sacks (1984) has shown that nonetheless, hard empirical data can be found in the language of social interaction. Using this hard data, ethnomethodology can look at language events, 'under the microscope' so to speak and explore how they effectively constitute the very events in which the language takes place. They can demonstrate the circumstances that provide context for meaning are themselves reflexively generated through talk and interaction. In Garfinkel and Sack's words (1970, p. 353) 'social facts are the accomplishments of the members'.

Whereas the naturalist tradition in ethnography typically asks what the social reality of the member is, where it is located and why, the ethnomethodologist seeks to ask how the members through their action and particularly their talk, construct their view of reality (Gubrium & Holstein, 1997). It has given us the tools to re-examine the taken-for-granted assumptions about what constitutes 'indigenous' knowledge. Within AA, as elsewhere, it is the view of the insider, either acquired through ethnographic immersion or through the role of true participant turned ethnographer who has particular access to those taken-for-granted assumptions.


      Adler, P. A., & Adler, P. (1997). Parent as researcher: The politics of researching in the personal life. In R. Hertz (Ed.), Reflexivity and voice (pp. 21-44). Thousand Oaks, CA: Sage Publications.

      Aguilar, J. L. (1981). Insider research: An ethnography of a debate. In D. A. Messerschmidt (Ed.), Anthropologists at home in North America (pp. 15-26). London: Cambridge University Press.

      Anon (1952). The twelve steps and twelve traditions. New York: Alcoholics Anonymous, World Services, Inc.

      Anon (1985). Alcoholics Anonymous - The story of how many thousands of men and women have recovered from alcoholism. New York: Alcoholics Anonymous World Services. (Original work published 1939)

      Anon (1986). Alcoholics Anonymous comes of age. New York: Alcoholics Anonymous World Services.

      Arminen, I. (1998). Therapeutic interaction: A study of mutual help in the meetings of Alcoholics Anonymous. Helsinki: The Finnish Foundation for Alcohol Studies.

      Arminen, I. (2000). On the context sensitivity of institutional interaction. Discourse and Society, 11(4), 435-498.

      Atkinson, J. M., & Drew, P. (1979). Order in court: The organisation of verbal interaction in judicial settings. London: Macmillan.

      Austin, J. L. (1976). How to do things with words. Oxford: Oxford University Press.

      Bakhtin, M. (1984). Problems of Dostoevsky's poetics. Minneapolis: University of Minnesota Press.

      Bateson, G. (1985). The cybernetics of 'self': A theory of alcoholism. In Steps to an ecology of mind (pp. 309-337. New York: Ballantine Books. (Original work published 1971)

      Buber, M., & Smith, R. G. (2000). I and thou. New York: Scribner.

      Button, G. (1990). Going up a blind alley: Conflating conversation analysis and computational modelling. In P. Luff, N. Gilbert, & D. Frohlich (Eds.), Computers and conversation (pp. 67-90). London: Academic Press.

      Collins, P. H. (Ed.). (1991). Learning from the outsider within: Sociological significance of black feminist thought. In M. M. Fonow & J. A. Cook (Eds.), Beyond methodology: Feminist scholarship as lived research (pp. 35-59). Bloomington: Indiana University Press.

      Coulon, A. (1995). Ethnomethodology. Thousand Oaks, CA: Sage Publications.

      Denzin, N. K (1987). The recovering alcoholic. Newbury Park, CA: Sage.

      Denzin, N. K. (1989). Interpretive interactionism. Newbury Park, CA: Sage Publications.

      Denzin, N. K (1997). The alcoholic society. New Brunswick: Transaction Publishers.

      Drew, P., & Heritage, J. (1992). Analysing talk at work. In P. Drew & J. Heritage (Eds.), Talk at work: Interaction in institutional settings (pp. 3-65). Cambridge: Cambridge University Press.

      Fairclough, N. (1995). Critical discourse analysis: Papers in the critical study of language. London: Longman.

      Frankel, R. (1990). Talking in interviews: A dispreference for patient-initiated questions in physician-patient encounters. In G. Psathus (Ed.), Interaction competence (pp. 231-262). Washington D.C.: University Press of America.

      Garfinkel, H. (1969). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.

      Garfinkel, H. (1986). Ethnomethodological studies of work. London: Routledge & K. Paul.

      Garfinkel, H., & Sacks, H. (1970). On formal structures of practical actions. In C. McKinney & E. A. Tiryakian (Eds.), Theoretical sociology: Perspectives and developments (pp. 338-366). New York: Appleton-Century -Crofts.

      Glaser, B. G., & Strauss, A. (1967). The discovery of grounded theory. Chicago: Aldine.

      Gubrium, J. F., & Holstein, J. A. (1997). The new language of qualitative method. New York: Oxford University Press.

      Have, P. T. (1999). Doing conversation analysis. London: Sage Publications.

      Hutchby, I., & Wooffitt, R. (1998). Conversation analysis; Principles, practices, and applications. Cambridge: Polity Press.

      Lee, J. R. E. (1987). Prologue: Talking Organisation. In G. Button & J. R. E. Lee (Eds.), Talk and social organisation (pp. 19-53). Philadelphia: Multilingual Matters.

      Mäkelä, K. (1996). Alcoholics Anonymous as a mutual help movement: A study in eight societies. Wisconsin: The University of Wisconsin Press.

      Maxwell, M. (1984). The AA experience: A close up view for professionals. New York: McGraw-Hill.

      Mead, G. H. (1934). Mind, self and society. Chicago: University of Chicago Press.

      Messerschmidt, D. A. (1981). On anthropology "at home". In D. A. Messerschmidt (Ed.), Anthropologists at home in North America (pp. 3-14). London: Cambridge University Press.

      Michels, R. (1958). Political parties: A sociological study of the oligarchical tendencies of modern democracy. Glencoe, IL: Free Press.

      O'Halloran, S. (2000). Power and solidarity in the discourse of Alcoholics Anonymous. Paper presented at Research & Practice in Professional Discourse Conference, Hong Kong City University.

      Pike, K. (1987). Language in relation to a unified theory of the structure of human behavior. The Hague, the Netherlands: Mouton. (Original work published 1954)

      Reinharz, S. (1997). Who am I? In R. Hertz (Ed.), The need for a variety of selves in the field: Reflexivity and voice (pp. 3-13). Thousand Oaks, CA: Sage Publications.

      Room, R. (1993). Alcoholics Anonymous as a social movement. In B. S. McCrady, & W. R. Miller (Eds.), Research on Alcoholics Anonymous: Opportunities and developments (pp. 167-188). New Brunswick, NJ: Rutgers Centre for Alcohol Studies.

      Rosenau, P. M. (1992). Post-modernism and the social sciences: Insights, inroads, and intrusions. Princeton, NJ: Princeton University Press.

      Rudy, D. R (1986). Becoming alcoholic: Alcoholics Anonymous and the reality of alcoholism. Carbondale and Edwardsville: Southern Illinois University Press.

      Sacks, H. (1984). Notes on methodology. In J. M. Atkinson & J. Heritage (Eds.), Structures of social action: Studies in conversation analysis (pp. 21-27). Cambridge: Cambridge University Press.

      Schegloff, E. (1992). On talk and its institutional occasions. In P. Drew & J. Heritage (Eds.), Talk at work: Interaction in institutional settings (pp. 101-134). Cambridge: Cambridge University Press.

      Schur, E. M. (1971). Labeling deviant behavior. New York: Harper and Row.

      Searle, J. (1960). What is a speech-act? In P. Giglioli (Ed.), Language and social context (pp. 136-154). London: Penguin.

      Sidorkin, A. M. (1999). Beyond discourse: Education, the self and dialogue. New York: State University of New York Press.

      Silkworth, W. D. (1985). The doctor's opinion. Alcoholics Anonymous - the story of how many thousands of men and women have recovered from alcoholism. New York: Alcoholics Anonymous World Services. (Original work published 1939)

      Tedlock, B. (1991). From participant observation to the observation of participation: The emergence of narrative ethnography. Journal of Anthropological Research, 47, 69-94.

      Whyte, W. F. (1945). Street corner society. Chicago: University of Chicago Press.

      Zimmerman, D. H., & Boden, D. (1993). Structure-in-action: an introduction. In D. Boden & D. H. Zimmerman (Eds.), Talk and social structure: Studies in ethnomethodology and conversation analysis (pp. 3-21). Cambridge: Polity Press.

Author Note

+Seán O'Halloran currently lectures in the Education Department of the Ulster University, Northern Ireland. Most of his professional life has been involved in English language teaching and teacher education in England, Ireland, Singapore and Hong Kong. His previous publications have been related to teacher education and the cultures of learning in Asia. He is currently doing a Doctorate in Education in Leicester University, England, on the discoursal practices of members in AA meetings. He can be contacted at 3 Drumavoley Park, Ballycastle, BT54 6PE Co. Antrim, Northern Ireland and his email is:

Article Citation

      O'Halloran, S. (2003, March). Participant observation of alcoholics anonymous: Contrasting roles of the ethnographer and ethnomethodologist. The Qualitative Report, 8(1). Retrieved [Insert date here], from

Seán O'Halloran
2003 copyright

Return to the top of the paper.

Return to the Table of Contents.