Higher Education 46: 167193, 2003. 2003 Kluwer Academic Publishers. Printed in the Netherlands.

167

Methodology for close up studies struggling with closeness and closure

MATS ALVESSON
Department of Business Administration, School of Economics and Management, Lund University, PO Box 7080, SE-220 07 LUND/Sweden (Phone: +46 46 222 42 44; Fax: +46 46 222 44 37; E-mail: Mats.Alvesson@fek.lu.se) Abstract. The paper critically reviews some themes and methods within qualitative research and argues for self-ethnography as an approach to study universities and other settings which the researcher is highly familiar with, and has direct access to. Advantages and disadvantages are discussed. In particular the struggle between utilizing closeness to empirically rich situations and avoid the closure following from being (staying) native is highlighted. Some ideas for self-ethnographic work in organization studies are developed. Keywords: ethnography, de-familiarization, higher education qualitative methodology, organizational research

Introduction It is rare that academics study the lived realities of their own organizations, e.g. their own universities or rather the more delimited settings in which they are active, such as departments, research groups, committees or interactions with students. There may be good reasons for this. It is difcult to study something one is heavily involved in. There may be anticipations of those targeted for study to experience breaks of trust. Some of the reasons for not studying the setting in which one is active may be bad, however, but still dominate due to their taken for granted character. Personal involvement should not necessarily rule out an inquiry, it may be a resource as much as a liability. Ideas about organizational loyalty requiring that one is not exposing backstage conditions may lead to, or be an excuse for, self-disciplination and subordination to conventions on proper behaviour which are taken for granted. Studying university settings as organizations may, of course, be carried out in a variety of ways. Much of the research work has addressed rather impersonal aspects of organizations. At a certain level of abstraction, acts, practices, relations, feelings and cognitions are totally lost to the benet of the correlation of variables. Here, the subjectivity of the researcher, or, and rather, the impression of it, has been minimized. My interest circles around

168

MATS ALVESSON

what really goes on in organizations: how people act, interact, talk and accomplish things. For me, the understanding of organizations calls for a considerable amount of micro-anchoring. This is typically understood as implying a qualitative approach. Normally such an approach means that the researcher tries to get relatively close to the meanings, ideas, discursive and/or social practices of a group of people that the researcher is approaching. Qualitative research is often seen in geographical terms as a movement when a researcher, initially at a distance, is coming closer and closer the lived realities of other people. Geographical terminology here is not used solely in a metaphorical manner. In anthropology the ethnographer may travel a long distance in order to get in physical contact and then make observations at, perhaps, 1020 yards from the natives. In qualitative social science addressing less exotic objects of study the researcher perhaps travels a few miles and then carry out eldwork at microphone-holding distance say 2 yards from the subjects. In this paper I will explore the possibilities of research avoiding the element of physical and metaphorical distance reducing activity, i.e. to study ones own setting rather than the setting of a group of other people. I will then introduce and explore the idea of a self-ethnography and elaborate on its possibilities and difculties. I will start, however, with a brief review of some problems in qualitative research indicating why we need to go more into the ethnographic direction and then discuss why we may want to explore new ways of doing so. I concentrate on these two methods (a) for space reasons, (b) as these two probably are the most popular ones, and (c) the points made about interviews also indicate some of the problems with methods such as focus groups and diaries.

Some problems with research interviews As will be addressed later, the difculties in doing research about ones own setting, per denition including (although not necessarily focusing) the researcher-author her-/himself, are great. There are intellectual as well as political problems involved. Why make research life so complicated and risky? One answer to this question is to point at problems with the alternatives. Awareness of the weakness of the alternatives may also, to the extent that ones critics are familiar with these, provide more space for and acceptance of efforts to experiment with new ways of doing research. Qualitative research often means conducting and interpreting interviews. The rationale is that a rich set of accounts of the interviewees experiences, knowledge, ideas and impressions may be produced and documented. Most of the literature on interviewing deal at length with how this practice may be

METHODOLOGY FOR CLOSE UP STUDIES

169

utilized as effectively as possible. While realizing some of the complexities involved, it assumes that skills may be developed and an approach taken in which errors are minimized and qualied empirical material be produced. Nevertheless, there are some serious problems with interviews that can not really be avoided through the use of techniques aiming to make interview work as rational as possible.1 There are always sources of inuence in an interview context that can not be minimized or controlled (Alvesson 2003). These go far beyond what may be seen as pure errors. As Silverman (1989, 1993) has stressed, the value of interview statements is in many cases limited in terms of their capacity to reect reality out there as well as the subjective world of the interviewee (beliefs, attitudes, psychological traits, etc.). This is partly the case because the statements are liable to be determined by the situation, i.e. they are related to the interview context rather than to any other specic experiential reality, and partly because they are affected by the available cultural scripts about how one should normally express oneself on particular topics (see also Dingwall 1997; Potter and Wetherell 1987; Shotter and Gergen 1989, 1994, etc.). An interview is a social situation and that which is said is far too contextdependent to be seen as a mirror of what goes on outside this specic situation in the mind of the interviewee or in the organization out there. Interviewees speak in accordance with norms of talk and interaction in a social situation. The research interview is thus better viewed as the scene for a social interaction rather than a simple tool for collection of data. Critics object to the nave and rather romantic view of research and research subjects, which believes that genuine experiences can be captured with the help of interviews. Most proponents of qualitative methods probably agree about the signicance of respondents expectations of what the researcher wants to hear and social norms for how one expresses oneself. Many would, however, believe that establishing close personal contact with respondents who then are seen as participants instead may minimize this problem. Fontana and Frey (1994), for example, suggest that the researcher may reject outdated techniques of avoiding getting involved or providing personal opinion and to engage in a real conversation with give and take and emphatic understanding. This makes the interview more honest, morally sound, and reliable, because it treats the respondent as an equal, allows him or her to express personal feelings, and therefore presents a more realistic picture that can be uncovered using traditional interview methods. (p. 371) Despite my sympathy for this kind of interviewing, I dont think it guarantees truthful interview statements that give a realistic picture. All

170

MATS ALVESSON

experiences and social phenomena may be represented in a variety of ways and there are always the element of arbitrariness, chance and availability of a particular mix of discourses guiding a specic interview statement. While an interview technique trying to maximize neutrality and minimize interviewer inuence may lead to shallow, convention-guided and not very honest accounts, closeness-maximizing approaches may lead to that the orientations of the researcher more strongly guide responses. An interviewer engaging in emphatic understanding may trigger sympathy from the interviewee and make him or her wanting to agree with and please the former. Even if the interviewer does not espouse his or her stand on the subject matter or even research interests thus deviating from the ideal of an honest, sound and reliable interview expressed in the quote above the interviewee develops an assumption of what the researcher is up to and this assumption frames the responses. In many cases the interview probably has clear constitutive effects. A feminist woman emphatically interviewing a woman may bring the latter to express a more pro-feminist opinion or feminism-supporting account of experiences than would otherwise be the case. A researcher interested in academic leadership may be successful in getting a department chairperson to constitute her- or himself as a leader in the interview, something that the chairperson may or may not do in everyday life. Even if an interviewer should manage to maximize honesty and minimize the want to adapt to the assumptions of the researchers anticipations and values or to comply to social norms for expressing oneself, honesty and independence do not restrict how one can represent the experiential or external world there are always a million of aspects, words and empirical illustrations to choose between when accounting for non-trivial issues. In addition one can never know for certain what expectations research subjects have, how honest they are, etc. To appear honest and not socially incompetent or odd is a social accomplishment and calls for impression management. Interviewees are frequently politically conscious actors. It seems reasonable to expect that interview accounts at least to some extent are driven by interests to held up specic versions of how social reality preferably should be understood as much as a neutral wish to tell the truth, as known by the interviewee. Even truth-telling may be selective and guided by ideas of the individual and collective interests of the interviewee. In academic contexts, people are typically aware of issues like personal, institutional and occupational prestige and reputation. Organizational and professional loyalties and fear of being perceived as illoyal are hardly bracketed when people respond to interview questions (Alvesson 2003). I am not against interviews and believe we can learn a lot from asking questions to people and with some skepticism listening to their answers.

METHODOLOGY FOR CLOSE UP STUDIES

171

Still, there are good reasons to be more restrictive than is presently the case in our reliance on this as a technique for getting knowledge of what goes on outside interview situations. It is simply too difcult to sort out scriptfollowing, the social dynamics of the interview situation, impression management and (other forms of) politically conscious language use from valid accounts about the employees feelings, thoughts and ideas, respectively, social practices out there.2 It is seldom possible to separate the distortions from authentic experiences or correct information (Alvesson 2003; Silverman 1993).

Ethnography An ambitious alternative to solely or mainly relying on a set of interviews is to carry out an ethnography. There are, of course, different opinions about what is included in the concept of ethnography. Some people dene it broadly as a study of an explorative nature, working with unstructured data, being case oriented and expressing an interest in meanings (Atkinson and Hammersley 1994) or any study involving observation of what is referred to as naturally occurring events (Silverman 1985). I think that the term serves us best if reserved for studies involving a longer period of eldwork in which the researcher tries to get close to the community (organization, group) being studied, relies on their accounts as well as on observations of a rich variety of naturally occurring events (as well as on other material, e.g. documents or material artifacts) and has an interest in cultural issues (meanings, symbols, ideas, assumptions). Unlike the authors just referred to, many authors view an interest in cultural analysis as a crucial criteria for an ethnography (e.g. Wolcott 1995). One element here is thick description, i.e. careful accounts of social phenomena in which layers of meaning are expressed (Geertz 1973). One may, however, also talk about ethnographies of institutional discourse (Miller 1997; cf Baszanger and Dodier 1997) in which it is not so much meaning as discourse in social settings that is investigated. The increasing awareness of problems with interviews, together with interests in relating macro and micro levels, as well as the current popularity of cultural analysis, have probably led to a greater interest in ethnography. Ethnography is looked upon as sophisticated - it involves more than just interviewing. First-hand experiences having been there offer a deeper level of understanding and a stronger authority-base than sending out questionnaires and listening to peoples stories in interview situations. Its positive aura seems to attract many people with diverse orientations. The variation amongst those wanting to associate themselves with the ethnography label means that it becomes diluted as a term, perhaps even as a methodology.

172

MATS ALVESSON

Sometimes people emphasize two major elements of an ethnography: the process of eldwork and the writing of a text (Van Maanen 1995). The term ethnography actually has two meanings: the empirical work and the completed study (text). Ethnographies have some clear advantages. Observations of naturally occurring events avoid or, more usually, reduce the researchers dependence of the accounts of respondents. The researcher may discover aspects which interviewees may be unaware of or which, for other reasons, they nd difcult to articulate. However, interviews or less formal, more spontaneous talks between researcher and informants are almost always an important complement to this method. Without the accounts of the people being studied, it is very difcult to say something about the meanings of and ideas guiding particular behaviours and practices. Interviews may provide richer results as the researcher over time develops his/her understanding, can ask better questions and may get better contact with the natives. The use of a multitude of methods sometimes referred to as triangulation is often to be preferred, not in order to zoom in the truth through different methods, but in order to create a richer picture (Denzin 1994). Disadvantages with ethnographies include them being time-consuming, often personally tiresome and stressful to carry out. The method of hanging around involves a lot of dead time. Ethnographies are too ineffective for most research purposes (Wolcott 1995). A somewhat more general problem concerns cultural studies in general as much as ethnography as a specic method. When studying exotic groups there is the risk that the researcher goes native, i.e. becomes caught in details and local understanding without being able to say something systematic of wider theoretical interest. The opposite problem concerns the difculties for a representative of one culture to fully or even adequately understanding another one. On the whole, students of organizational culture within ones own national context suffer from a lack of imagination making it possible to accomplish studies not caught up in the taken-for-granted assumptions and ideas that are broadly shared between the researcher and the researched (Alvesson 1993). Too much of organizational life is often too familiar. For academics studying other academics this is an especially strong problem. A more fundamental problem, which also characterizes other qualitative research, concerns the difculties in handling all the empirical material and in producing a text that does justice to it. Even if the ethnographer claims that his or her rsthand experiences of the object of study is a strong basis of authority, the text produced is not just a document mirroring something out there. The problems of writing the study up has received much attention in anthropology and other ethnographically oriented research areas

METHODOLOGY FOR CLOSE UP STUDIES

173

(Clifford and Marcus 1986; Geertz 1988; Marcus and Fischer 1986; Van Maanen 1988, 1995). Ethnography has gone from a relatively unreective, closed and general description of a whole way of life not too difcult to picture in a text - and instead emphasized a more tentative, open and partial interpretation, drawing attention to matters of uncertainty and style in writing (Geertz 1973, 1988). Recent critique has drawn attention to the ctional nature of ethnographies and, for that matter, all social research. The text is seen as central. It tells a story, it adapts a particular style, the author make all kinds of moves in order to create certain effects, e.g. trustworthiness, legitimacy, brilliance. All these go far beyond simply reporting data and describing objective reality. Things going on in social reality out there may inspire the author but puts highly uncertain imprints on the text. We have every reason to be wary of the realistic or naturalistic mode of writing in which the production of understanding and construction of the text are hidden by a form of account that purports to present what is described simply as it appeared; this being treated, with more or less conviction, as how it is (Hammersley 1990, p. 606). This aspect may be concretized through emphasizing the selectivity of the empirical material re-produced in the published text. The output of a set of (open) interviews and long-term observer participation aspiring to describe a complex reality is always difcult to transcribe in research texts. Along with a whole host of other problems, partly indicated above e.g. interview accounts as an outcome of the social situation and of the following of scripts the presentation of the material inevitably becomes a question of selection and discretion/arbitrariness. This is a particular problem of non-formal, broader studies utilizing rsthand experiences, extensive interview and observation material, much of this being collected in a non-structured and non-systematic way. Only a very small portion of all that which has been said by the interviewees and observed, usually during several weeks or months, can appear in a publication or even fully considered in analysis. In addition, of course, we have the problem that behaviours, meanings, etc can not really be mirrored in a text. Writing conventions typically prevent a text from appearing too contradictory and confusing for the reader. (The presentation of material pointing in different directions is acceptable, but only to a certain degree, and the author is supposed to get it all together at one level or another at some point.) To accomplish a text that gives a good account in the sense of mirroring a reality represented in all this empirical material is very difcult, if not impossible even if one disregards the fundamental problem of treating language as standing in a one-to-one relationship to other phenomena.

174

MATS ALVESSON

For these and other reasons, ethnography has lost some of its innocence and self-condence in terms of authority. As Van Maanen (1995, p. 2) says, ethnography is no longer pictured as a relatively simple look, listen and learn procedure but, rather, as something akin to an intense epistemological trial by re. Of course, these problems are not only valid for ethnography, although possible more pronounced and apparent there, but are relevant for all research, especially qualitative research. To sum up, there are many advantages with conventional ideas about ethnography. However, they are time-consuming and often uneconomical, they often mean that empirical material (piles of interview and observation notes) receive (too) much attention, at the expense of other virtues in research, such as reexivity at various levels of ones research and interpretations. The idea of covering large chunks of a social reality intensies the problem of producing a text faithful to this reality and the element of ction becomes important to attend to. This problem is somewhat less salient in more structured approaches such as interview inquiries. The problem of mastering closeness/distance learning the culture and being able to read it so that something of a broader/theoretical interests emerges out of the project is a general one for social science, but is of specic relevance for ethnographies, as they are more ambitious than other methods in terms of getting close to the natives and aim for cultural analysis.

Self-ethnography As a response to some of the problems of interview-based qualitative research and ethnographies what I call the self-ethnography may be considered. A selfethnography is a study and a text in which the researcher-author describes a cultural setting to which s/he has a natural access, is an active participant, more or less on equal terms with other participants. The researcher then works and/or lives in the setting and then uses the experiences, knowledge and access to empirical material for research purposes. This research is, however, not a major preoccupation, apart from at a particular time when the empirical material is targeted for close scrutiny and writing. The person is thus not an ethnographer in the sense of a professional stranger or a researcher primarily oriented to studying the specic setting. Participant observation is thus not a good label in this case, observing participant is better. Participation comes rst and is only occasionally complemented with observation in a researchfocused sense. One may also imagine versions of self-ethnography, when the monitoring of what goes on becomes a chief preoccupation. Or one may imagine certain moments in the research process that are more intense in

METHODOLOGY FOR CLOSE UP STUDIES

175

terms of observation the self-ethnographer becomes called to full-time duty. In this case, self-ethnography borders to participant observation. A signicant difference is that the conventional ethnographer uses any kind of active participation for an instrumental purpose the ethnographer working as a lumberjack does so in order to be able to produce research about lumberjacks, not because of an inner urge to cut down trees whereas the idea of a self-ethnography is to utilize the position one is in also for other, secondary purposes, i.e. doing research on the setting of which one is a part. The term self-ethnography may sound as if the researcher her- or himself that is focused. The intention is, however, as said to draw attention to ones own cultural context, what goes on around oneself rather than putting oneself and ones experiences in the centre. Self-ethnography then is a bit different from some recent work in which the deeply personal experiences of the researcher are in focus. This kind of work is often labelled as autoethnography (Ellis and Bochner 2000). Autoethnographies are highly personalized, revealing texts in which authors tell stories about their own lived experiences, relating the personal to the cultural (Richardson 2000, p. 931). There is a strong inward-looking element in this kind of work, even though the researcher goes back and forth between focusing outward on social and cultural aspects of their personal experience; then, they look inward, exposing a vulnerable self that is moved by and may move through, refract, and resist cultural interpretations. (Ellis and Bochner 2000, p. 739) A related stream, more common in Europe, is called memory work and is linked to feminism (e.g. Widerberg 1995). Memory work and ethnography share a focus on strongly personal experiences, e.g. and as a patient in the hospital system (e.g. Kolker 1996) or as a women in a male-dominated society (Widerberg 1995). These texts, that tend to be more autobiographical than observational in a conventional social science sense, then differs signicantly from the research method described here, which put emphasis on careful documentation and interpretation of social events that one is witnessing and do not necessarily emphasize personal meaning or the strong subjective aspects. In the kind of research that I propose, the work situation provides the viewpoint, but the aim is to carry out cultural analysis and not introspection, although it is important to not overstress this division as ones own feelings, thoughts and experiences may offer valuable material. As Van Maanen (1988) suggests ethnographies are sometimes written in a confessional style, i.e. the researcher-author writes from a personal position and proceeds from personal experiences of the encounter with those being studied. A self-ethnography

176

MATS ALVESSON

does not necessarily imply a confessional style; it can be written in a conventional (realist) or in any other way. One possibility is to provide detailed accounts of social events that the researcher observes but is not directly engaged in. Another issue is that a confessional style may be very appropriate in self-ethnographies. Alternative terms may be home-culture-ethnography or insiderethnography. (Brannick and Coghlan (2001), talk of insider research, but relates it primarily to action research.) I use the term self-ethnography in this paper, although with some worry as it may trigger associations to work focusing the researchers self or highly personal experiences, but as said this is not what I primarily have in mind. Self-ethnography is especially of relevance for research on universities and higher education. As mentioned, it is not, however, restricted to this. Other sites in which the researcher is engaged, e.g. neighbourhood, consultancy work, political organizations, other associations, commercial settings, families may also be targeted. This paper has primarily the context of higher education in mind, without very heavily focusing on it. A major difference between self-ethnography and conventional ethnography is that the home-base of the former is the setting being studied. When a conventional ethnographer may take a job or do something else in order to become a (temporary) member of community as the vital part of the empirical research process, to get data and learn the culture, the research then comes close to a socialization process viz. the community being studied. When the ethnography is a PhD project, the process may be seen as a double socialization process: (a) within the research community ending with full membership and internalization of the appropriate orientations, and (b) within the community being studied, ending with a good understanding and mastery of cultural rules, but also considerable distance to these. While conventional ethnography is basically a matter of the stranger entering a setting and breaking in, trying to create knowledge through understanding the natives from their point of view or their reading of acts, words and materia used, self-ethnography is more of a struggle of breaking out from the taken for grantedness of a particular framework and of creating knowledge through trying to interpret the acts, words and materia used by oneself and ones fellow organizational members from a certain distance. In the rst case, we have the researcher as burglar, in the second as a run-away. The burglarresearcher wants to overcome obstacles in order to get in contact with a target of interest, the run-away-researcher struggles in order to create sufcient distance in order to get perspective on lived reality. All researchers, in particular qualitative research and especially so ethnographers, struggle with the dilemma of closeness/distance, but in different studies different poles

METHODOLOGY FOR CLOSE UP STUDIES

177

represent the tricky issues to deal with. While the conventional researcher (with a anthropological orientation) may ask What in hell do they think they are up to? the self-ethnographer must ask What in hell do we think we are up to? I will address how this may be accomplished later on. It is of course often difcult to decide the precise territory of the home base: is it the department, a fraction of it, the school, faculty, university or the occupational community or perhaps the system of higher education (in a particular country)? I would say that those parts of the social setting with which one directly interacts and thus may observe with some regularity form a productive terrain for ethnographic work. Universities are not homogenous, the idea of exploring a university culture as a whole risks to encourage too supercial work. But even though universities or even departments are better conceptualized as multiple cultural congurations (Alvesson 1993; Trowler 1998), there are blind spots and intellectual closures shared by people in universities that makes also the project of a researcher studying a department in signicant ways different from his or her own difcult. All research of non-alien groups then must take the problem of breaking out seriously. The conventional ethnographer typically spends an enormous amount of time on note-taking. Scribbling down observations, informal interviews and perhaps ones thoughts and associations sometimes appear to be cornerstone of the entire project, central also for the researchers feelings (Jackson 1995). A self-ethnographer is more inclined to rely on familiarity with the setting as an empirical starting point. The trick is more a matter of accomplishing a description and insightful, theoretical relevant ideas and comments out of the material. It is a matter of thinking through what one may already has a good, although perhaps non-articulated and partly taking for granted understanding of. As with all research the trade-off between a good cultural description and an interesting theoretical idea, being abstracted from (theory-impregnated) interpretive description may vary. Denzin (1997) writes that ethnography is that form of inquiry and writing that produces descriptions and accounts about the ways of life of the writer and those written about (xi). This is a rather peculiar denition, at odds with almost all other views on ethnography. Denzin seems to have in mind any cultural analysis of a society. Normally ethnographies explicitly study a group that the writer is not a member of, which of course does not prevent the study from saying something of relevance for the self-understanding of the home community of the researcher. But the latter is not focused. Geertz (1973) famous study of the Balinese cockght is for example not about the ways of lifes of Princeton professors. Denzins denition has, however, unintentionally, I think, bearing on what I have in mind with the concept of self-ethnography.

178

MATS ALVESSON

One rationale for self-ethnography concerns its capacity to come up with novel and interesting empirical material. The insider is, potentially, better positioned than the one of an outside ethnographer to reveal the true story, although position alone is insufcient to realize the potential. The stranger-ethnographer will normally undoubtedly have some access problems, particularly on the level of depth access, i.e. stories on sensitive matters may not readily be produced. Covert studies may score better on this point, but here we face ethical problems. Another possibility is that the deeper and more profound knowledge of the setting may lead to theoretical development that is more well grounded in experiences and observations than is common. Compared to studying settings which the researcher, even after months of research work, still has a relatively limited knowledge of, the study of settings which the researcher really is familiar with may be productive. In particular, the excellent access possibilities of the self-ethnographer although within a limited eld may lead to accounts that give a good feeling for what goes on. There is much dissatisfaction with social research for its inability to portray everyday life. Representations are often remote, articial and clumsy. The reliance on procedures for data collection, categorizations, attaining high evaluator interreliability steers research away from giving valuable insights. So is especially the case with quantitative studies, but many qualitative research efforts do not score much better. They are boring to read, many people feel (Richardson 1994). Some researchers advocate ctional ethnographies in order to be able to produce sensitive and insightful accounts (e.g. Tierney 1993; Watson 2000). Without underestimating the importance of writing skills and putting ones own lived experiences in perspective, proceeding from the rich empirical material at hand in ones everyday life could give ample opportunities to produce good accounts. Being faithful to a specic empirical material, i.e. trying to stick to what one has seen and heard in a particular context, produces some constraints to ones fantasy and wish to portray the world in a certain way, which I think is motivated in social science, aiming at interpreting social reality. Of course, there are serious problems involved. I have so far indicated some potentials. Much is needed for these to be realized, as will be addressed later in this paper. Self-ethnography in universities There are a few additional reasons for why we should conduct studies of our own settings. One is about power and politics. The somewhat mixed pleasure of being a target of research is unevenly distributed among the population. It is seldom the elite that is being empirically studied, but more

METHODOLOGY FOR CLOSE UP STUDIES

179

often people at the other end of the social scale. So is especially the case when ethnographic and observational methods are used. When elites are being studied, it is normally in the form of interviews where they themselves control the situation and produce their own versions of the world. Workplace ethnographies are carried out among blue collar workers, not executives. (A few go up to middle-lever managers, though, e.g. Jackall 1988; Kunda 1992. In addition, there are a few observational studies of senior executives in mass meetings, e.g. Alvesson 1996; Rosen 1985.) Students are used for experiments in psychology. Conversational analysis carefully document and scrutinize what is being said more frequently in aids counseling and doctorpatient interaction than in boardroom meetings. In general, it is important that also relatively prestigious groups such as professionals get their share of in-depth exposure3 Despite degradation following from massication of higher education universities and the people working there (Trowler 1998), they belong to the upper echelons in society. Part of the research going in this direction includes, however, its own problems in terms of politics. The study of weak or marginal groups or people in vulnerable situations (such as in health care settings) involves a kind of mild vertical political-ethical dilemma. One may also talk about a horizontal political-ethical dilemma, associated with the conicts about power, prestige and cultural capital among groups in, broadly speaking, competitive relationships. Social scientists are, of course, caught in their (our) values, ideas, status aspirations and inclination to collective selfaggrandizement/ego-blowing. Social science can be understood as projects in which members of a community try to promote its position viz. other groups which compete in terms of prestige, power and resources, these groups being either other professionals or academics in other elds. This is inherent in the research project: the very idea is to show that we add something to the knowledge of others. This adding is, preferably, in some (certainly not all) ways superior to the knowledge and self-understandings of these others. Otherwise, there is no reason for tax payers nancing research. In sociology of science, sociologists study the practice and conicts among the natural science. The outcome is often less than attering for these disciplines. For example, one study found that proponents of different competing positions each explained their view in rational terms, the position of the competitor was accounted for in non-rational terms, e.g. speculation or personal loyalty with a master gure (Gilbert and Mulkay, ref. in Potter and Wetherell 1987). We may understand the outcome of this kind of research as contributing to changing the prestige balance between the natural and social science. Demystifying natural science then self-presents social science, at least sociology of science, as being, in some ways, more insightful than those

180

MATS ALVESSON

under study about what they really do. Within organization studies, we have a large industry of studies like this, targeted at organizations and management in general, although perhaps primarily at the public sector (e.g. Meyer and Rowan 1977; Perrow 1978). I am positive to this work, but the absence of similar studies of institutions of social science means a certain amount of asymmetry in the distribution of critical scrutiny. Good organizational ethnographies often portray their objects of study in non-attering terms. When reading them I often feel glad that I work at the university. (See for example Jackalls 1988 and Watsons 1994, sharp studies of corporate managers.) Actually, hardly any social setting comes out of an ethnographic study unblemished. Most well-done studies working beyond front-stage and the level of image-production produce some far from positive descriptions and analysis. This may be one reason for us being more inclined to study the Others rather than ourselves. Even though we within social science frequently internally engage in critique about method, science and the value of our knowledge-products, there is little doubt about the rationality of ones own position and the aws in this respect of the others (e.g. the positivists). Engaging in the struggle for the right kind of doing research or the superior view on social science, is something else than exploring what goes on in higher education institutions. From a horizontal political-ethical point of view it seems important for ethnographic researchers to address not only the organizations and works of groups of people that we, in some sense or another, are competing with, but also ourselves. There are a few research efforts going somewhat in an ethnographic direction studying home ground. I have no intention of providing a full review of such but will only mention some examples. Watson (1996) studied his students and his interaction with them, Tierny (1993) wrote about policy changes regarding homosexuals in universities using ctional form and Fairclough (1993) used his own application for promotion as an example of the marketization of universities while I have myself used my university department as a case for illustrating multiple cultural congurations in organization (Alvesson 1993). One person in a research team got data about a task force of which he was a member (Gioia et al. 1995), but this study engaged in coding procedures and abstractions and avoids any recognition of personal experiences, at the expense of rich descriptions of events, talk and actions. Studying ones own setting may not only be narcissistic and unethical in relationship to the people around people fully capable of inicting sanctions on the researcher but ethical in terms of wider social relations. A precondition is, of course, that the end product of such studies does not

METHODOLOGY FOR CLOSE UP STUDIES

181

portray the research object in more polished, uncritical or unfair ways than is the typical case in ethnographic work. While celebratory work tends to be uninteresting to read and the rationale for self-ethnography is that it may lead to realistic, lively and rich accounts going beyond the level of appearance, this risk is probably low. I think that is reasonable to expect an self-ethnography to deliver an engaging empirical account, going beyond the forestage, in exchange for scoring lower on some other possible virtues, e.g. a lot of empirical footwork and procedure-following. Still, the problem of blindspots and of missing or omitting some dark or tabooed aspect of the home culture must be taken very seriously. So must also the potential opposite problem of motivation coming from negative feelings and an urge to get even. On the practice of self-ethnography The idea of self-ethnography is consistent with a variety of different ways of creating and doing something with the empirical material: from a plannedsystematic kind of data collection to an emergent-spontaneous approach. In the rst case the researcher has a reasonable clear idea of what to study, plan the work (although in ethnographies never in a detailed sense) and want to have a pile of notes or interview statements to work with and from. This appears rational and scientic and looks good in methods section in research texts. A planned-systematic study tends to be bound to the empirical material gathered. To some extent the research interest must be decided upon in advance.4 One may for example be interested in consequences, interpretations or talk about policy issues at the local level, about students capacities or incidents involving gender and then make observations and take notes every time one encounters such material. Notes then typically work like an imperative. They are there as a result of the researchers tedious work therefore they should be used. The material may be interesting or it may be not. You cant know in advance, although with effort, creativity and writing skill empirical material that appears not too promising may lead to an interesting text. An emergent-spontaneous study is carried out when something revealing happens. In such a study the researcher waits for something interesting/generative to pop up. It may sound risky and not very ambitious. It may also lead to a limited representation of the world, chosen arbitrarily. I would recommend care in trying to publish studies based on emergentspontaneous research in US journals. There are, however, some advantages, the most signicant one is that it increases the likelihood of coming up with interesting material. The researcher does not nd the empirical material, it nds him or her. The researchers energy, on low blouse, is spent keeping eyes and ears open and the Macintosh switched on, ready for the ngers to hammer on it, while one is carrying out ordinary academic work: lecturing

182

MATS ALVESSON

for students, trying to get the copy machine to work, etc. What it is really all about is to develop a sensitivity for and preparedness to do something with the rich empirical material that one, at least occasionally, is facing. The idea is that consistent, long-term scanning of what one experiences produces a more extended set of incidents candidating for analysis. Technical issues such as should one rely on retrospective accounts, notestaking or tape-recording?, are of some but not profound interest. All modes can, of course, be used. Tape-recording does not necessarily imply a plannedsystematic approach, as one may have the tape-recorder in the pocket and, given (a priori or ex post) acceptance of other people (suddenly turned into research objects) put it on when something interesting is happening. Retrospective accounts are selective and does not allow really ne-tuned analysis of type discourse and conversation analysis (Potter and Wetherell 1987; Silverman 1993). Notes-taking and, in particular, tape-recording may disturb the situation, create irritation and domesticate participants. In particular if the researcher-author is active in the situation and in many interesting and engaging situation in ones workplace it is difcult to keep ones mouth shut detailed notes-taking is difcult. The reader may wonder what is interesting? This is of course to some extent according to ones personal taste and pre-understanding, but, according to my experiences, generally there is a relatively high level of intersubjectivity in the evaluation of what is interesting. An interesting account tells us something revealing about what goes on in a particular site. It explores something unexpected or allows seeing something familiar in a new light. A certain level of generality is therefore called for. Highly idiosyncratic stuff may be entertaining but is not necessarily intellectually interesting. Typically an interesting account touches upon a mix of familiarity/surprise. This mix assures some element of generalization (although of course not in any statistical sense) and some element of variation/uniqueness. All social situations contain both, not all trigger the right combination. One criteria is that it appeals to something in the experiences of readers (cf the concept of naturalistic generalization, Stake 1994). It involves some element of identication. In order to produce something interesting, but even more in order to avoid abstractions in which specic processes, acts and events are turned into unrecognizability, it is important to micro-anchor the account. This means that specic acts, events, situations are in focus. A good account then involves actors, acts (processes) and an institutional context. This may be referred to as a situational focus (Knorr-Cetina 1981). Instead of nding an average or something system-like through rigorous comparisons of a number of microsituations, the study concentrates on exploring the richness of one or a few situations and then rely on ones general knowledge for evaluations of what

METHODOLOGY FOR CLOSE UP STUDIES

183

is fairly typical compared to what is outside the mainstream of the organization one is exploring (Alvesson 1996). It is reasonable that the researcher can support the choice of a focused situation or event through being able to refer to at least a few other instances in which the same theme or process is present, thereby indicating that the focused situation is of some relevance for illuminating broader chunks of the institutional context in which the focal situation is played out. This would increase the chances of being fair in the portrait painted of the setting. Of perhaps greater interest than the empirical material is what the researcher-author may do with it. The material must then work in a productive and inspiring way and lead to interpretations of a more theoretical nature. One may, however, imagine different ratios in terms of the intrinsic/instrumental value of an empirical account. It may in itself be informative and revealing a thick description or it may work as a lever for the production of a more abstract, conceptual contribution. Some problems in self-ethnographic research Doing self-ethnography is difcult and I dont think this a method for everybody to use, at least not at any time in ones life and career. It is risky business from an intellectual point of view as it is more difcult to draw upon and hide behind an apparatus of techniques and procedures for controlling subjectivity and assuring the reader that Science has as much to say about the outcome as personal idiosyncrasies. Being personally involved in the object of study (the context in which one is studying) also means that one may be less able to liberate oneself from some taken for granted ideas or to view things in an open-minded way. The research situation is in certain ways also more politically complex than is common. The risk of producing a attering view of oneself and the site of which one is a member is perhaps not so great, but taken for granted assumptions, blind spots, taboos and the want to avoid upsetting colleagues may create difculties and/or self-disciplination. Social research in general can not avoid either supporting or questioning existing social institutions and can therefore not claim to be neutral (Burrell and Morgan 1979; Alvesson and Deetz 2000). If one has, or faces readers of ones research products with a strong and direct personal interest in the subject matter this dimension can be more difcult to handle in a sophisticated manner. If the researcher does so s/he may get more enemies at close distance (or give the enemies additional ammunition). Just the anticipation of what people will think and feel may lead to more careful and atter descriptions than a freer and bolder approach would imply. Of course, diplomacy is a part of all (qualitative) research efforts and this may contribute to the end products frequently being somewhat

184

MATS ALVESSON

watered down but it may be easier to cope with this aspect if the persons being less than happy about the research outcome is at some distance from the everyday life of the researcher. There is also an important ethical aspect of this respect fot the feelings and interests of those being studied must guide research and their acceptance of the study must at some point be attained for it being published or widely distributed. The approach forces the researcher to address subjectivity and preunderstanding as a complex mix of resource and blinder. Arguably, social research (and perhaps many other forms of research) are profoundly affected, indeed driven by, personal feelings and life histories, cultural frameworks, social belongings, etc. of the researcher. This does not mean that idiosyncrasy, personal opinions and arbitrariness necessarily rules or do so in a manner that means that the research text is of greater interest for ones shrink than somebody that wants to learn about university organizations. What is strongly personally engaging for the researcher may well be of great interest also for other persons. As Jaggar (1989) claims, emotions are often shared between people and are an important element in research work. Still, ways of dealing with the problematic side of closeness and personal involvement must be considered. Careful reections on it is crucial for doing self-ethnography. The major problem is not necessarily subjectivity, in the sense of highly individualistic biases. In general, research suffers from the inability of researchers to liberate themselves from socially shared frameworks (paradigms, cultures). That evaluators agree may not be a sign on objectivity as much as culturally or paradigmatically shared biases (Alvesson and Deetz 2000). The trick is to get away from frozen positions, irrespective if they are grounded in personal experiences or shared frameworks. A problem is that staying within socially shared frames and biases may make research life easier while what is seen as personal biases are sanctioned, proceeding from and reproducing socially shared biases may be applauded. The general problem can be formulated as accomplishing openness in relationship to a prestructured, xed line of focusing/interpretation. Struggling with closure: Creating breakdowns In ethnographic work within anthropology, the initial difference between the traditions involved (the researchers and that of the object of study) may produce a breakdown in understanding, a lack of t between ones encounter with a tradition and the schema-guided expectations by which one organizes experience (Agar 1986, p. 21). The researcher the professional stranger deals with this by investigating the cultural elements encountered triggering the breakdown and then adjusting her/his schema. Breakdowns continue to appear until the researcher fully given what is to be investigated under-

METHODOLOGY FOR CLOSE UP STUDIES

185

stands the culture under study, and therefore ethnography can be described as a process of coherently resolving breakdowns (p. 39). When studying relatively familiar phenomena such as the workplaces and educational institutions of ones own country, the problem is not so much resolving breakdowns but creating them. In the study of foreign cultures breakdowns occur automatically, but in ones own they are mostly marginal. The trick then is to locate ones framework (cultural understanding) away from the culture being studied, so that signicant material to resolve emerges. The problem and rationale for organizational culture studies is to turn the well known and self-evident into the exotic and explicit to raise and answer the question What does it mean (apart from the obvious)? (Asplund 1970). This is of course to a large extent a matter of creativity, but it is also a matter of aspiring to accomplish anthropological rather than technical/pragmatic results. To some degree it is a matter of using the critical strategy of defamiliarization: Disruption of common sense, doing the unexpected, placing familiar subjects in unfamiliar, even shocking, context are the aims of this strategy to make the reader conscious of difference (Marcus and Fischer 1986, p. 137). The art of producing breakdowns in settings only too familiar is not easy to master. The more familiar setting, the fewer breakdowns. It is facilitated by a rich and generative kind of empirical material, plenty of time to consider what it means and access to a broad set of different resources theories, vocabularies, experiences to challenge self-evident forms of understanding. To discover or, to emphasize the need for effort and construction, to create revealing situations is one vital element. I prefer micro anchoring and rich descriptions rather than covering broader areas in a thin way, although the challenge is to see how chunks of social reality are writ small in specic events, these showing some similarity with a reasonable number of other events in the myriad of micro-situations making up social reality. To work with a situation or a process bounded in time and space and thus possible to grasp gives the researcher energy and mindpower to illuminate it from different angles (Alvesson 1996). Considering a variety of perspectives, and shifting these, is always important in research, but perhaps especially so in the type of research here suggested. As said, the trick is to try to get away from the inclination to see things only in a specic light as this means that ones personal and paradigmatic-cultural blinders tend to shadow other aspects than those preferred. I will briey indicate ve ways in which one may improve the prospect that this will be accomplished. One way of creating distance towards one self and ones cultural inclinations is to try to embrace positions of irony and self-irony (e.g. Brown 1977; Woolgar 1983). These must not necessarily dominate the nal text but may be taken temporarily in order to create a certain distance to more serious argu-

186

MATS ALVESSON

ments put forward. The modern higher education, by its advocates perhaps described as a triumph for equality and knowledge for everybody and/or being a motor for economic development, may be described as the McUniversity (Parker and Jary 1995), where standardization, rationalization and managerialization take over according to McDonald-like principles. Instead of describing oneself or, in the case of modesty, ones ideal as the scholarly champion of academic freedom and knowledge one may sketch a selfportrait of a high-brow egocentric in the ivory tower. Such alien and ironic representations may be used as thought experiments or explicitly in analysis. A second way is the use of theories which challenges common sense, not only for the direct application but also for encouraging perspective on ones own lived reality and thus facilitating looking upon things in a more all-sided way than is spontaneously the case or if one tries to adopt a grounded theoryapproach. Foucault (1976) and Bourdieu (1979) belong to those authors helpful in this respect. Both may help shake around xed preunderstandings, if these authors are not used solely against positions disliked by the researcher, but also in a self-questioning way. (This presupposes, of course, that the researcher is not a great devotee of the theory in question it is difcult to use a theory highly favoured to get a perspective on ones own thinking.) A third way is to build up an interpretive repertoire sufciently broad in order to read empirical material in a variety of ways, thereby considering and perhaps developing themes not too closely tied to ones personal-political tastes. This third way may overlap the two just mentioned. Mind-shaking theories such as those of Foucault and Bourdieu may be part of the interpretive repertoire, as may be (other) ideas of an ironic nature. A fourth way may be to work systematically with a notion of reexivity, in which one tries to change level of interpretation so that ones favored interpretations in a rst instance is the target of interpretation from a meta-level position, inspired by another standpoint. This other standpoint then functions in a metatheoretical way, i.e. it addresses ones interpretations and not directly reality out there (Alvesson and Skldberg, 2000). An effort to develop a particular point may be challenged by seeing this point of view from for example a feminist perspective exploring the (false) gender-neutral nature of the prior interpretation or from a poststructuralist position exploring the fragility behind the claims of authority of a preferred interpretation. Through incorporating challenges from counterperspectives framing the issue and line of interpretation in other ways, the researcher is forced to work through the preferred wisdom and unfreeze the position associated with personal history and shared taken for granted meanings. A fth way may be to explicitly work with the processual nature of the self. Instead of assuming a xed essence in terms of orientations and pre-

METHODOLOGY FOR CLOSE UP STUDIES

187

understandings associated with a single, predened self, the researcher can work with different self-concepts. Weedon (1987) and Deetz (1992) talk about process subjectivity in order to indicate that we can and actually do move between different ways of conceiving ourselves and our orientations. As a researcher, one is not only a student of education, but also perhaps a woman, middle-aged, senior (to some), junior (to others), a teacher, union person, etc. Attitudes are not xed of absolute, but contextdependent. The same person who is in favor or accountability and performance evaluations, likes autonomy and freedom from dysfunctional measurements. Drawing upon different orientations taking different positions, with different perspectives may facilitate avoidance of getting caught in a certain set of blinders. This way, of course, to some extent overlaps with and is assisted by the other approaches suggested above. Much more can be said about this issue and many versions of breaking out of constraining frameworks are possible. The important thing is that selfethnography, in order to escape the specic traps facing it, calls for some additional efforts in this direction compared to conventional modes of doing research. The who and when of self-ethnography Doing self-ethnography is probably not for everybody. Certain social situations, experiences and periods in ones career may facilitate it. Having been in a certain job and organization for a long time and having limited access to other work organizational experiences would do the research difcult. Being a newcomer and thus having to learn local culture makes the job in some ways intellectually easier, but politically more risky and possibly emotionally more stressful. Being active in other spheres of the working and public life, e.g. consultancy and political work, may also be helpful. In connection to career changes and the adaptation to new situations, e.g. becoming a department chairperson, self-ethnographic work may also be less tricky to carry out. The transitions contingent upon a sabbatical may also offer not just time but also perspectivating experiences facilitating self-ethnography. Doing such research may thus partly be a matter of timing. Some periods may be more suitable for it than others. There are also variations between people, partly related to familiarity with other sites than universities, that are helpful in self-ethnography. Multiple work-related social identities university professor, consultant, afliations to more than one academic discipline make it easier to avoid being caught in a staying native position. Being an outsider rather than a mainstream person may work in the same way if research is not only used to justify an outsiders position against the inferior mainstream. Co-authored work may in which an insider and an outsider collaborate may be one option.

188

MATS ALVESSON

Apart from making wise choices if and when one is the right person and has the right kind of background and experiences in order to intellectually look through ones own workplace culture, it is also a good idea to think about the politics involved. I guess that three issues matter here: one is the tolerance and openness of people at the workplace (the victims), a second is ones own position a prestigious, tenured researcher has more leeway and the third relates to the extent one constructs oneself in terms of integrity and inner-directness. Self-ethnography is not for the mainstream, organizational (wo)man, eager to conform to workplace norms and to be very loyal. My idea of raising the issue of the when and who of self-ethnography is only partly to aid people in terms of putting or omitting self-ethnography project work in their career planning calendars or to assist recruiters of selfethnographers to develop personality proles for candidates, but to repeat and vary the point about the need for working through the problem with a constraining, taken for granted cultural framework and a too myopic personal view on the subject matter. Cultural belongingness means a high degree of closure to the rich variety of potential ways of interpreting ones organization. These problems can be coped with, but there may be a trade off between the efforts called for and the output of the project. The advantage of selfethnography in terms of saving energy may not be available for everybody to exploit at any given moment. Of course, apart from the personal side, periods of great transformation and/or specic organizational changes in which the on-going production of social reality is disrupted offer good occasions for self-ethnographic work.

Conclusions As a social scientist interested in how universities or knowledge-intensive organizations in general work, one is, of course, not obliged to leave ones home base in order to encounter productive empirical material. This paper argues for the self-ethnography as an alternative or a complement to other ways of doing research. Arguably, there are some major advantages with the approach here suggested, of course conditioned upon an ability to deal with the considerable traps involved in this kind of work. Self-ethnography offers good research economy. In particular if we accept the rather basic and profound problems with interview material (and other methods assuming that the individual simply reports experiences, insights and meanings) as reections of what goes on out there and view conventional ethnography as time-consuming and uneconomical, it appears motivated to search for new ways of proceeding.

METHODOLOGY FOR CLOSE UP STUDIES

189

Self-ethnography may facilitate the production of rich empirical accounts. Given that one is not deciding a priori thereby constraining the options that a particular time and space area is to be researched, but is modestly engaged in scanning ones lived reality for research options for some time there is a good chance that one sooner or later runs into events making a good account possible, providing a feeling for what goes on and facilitating productive interpretations. I also believe that there are a few minor advantages with self-ethnography: Self-ethnography may develop reexivity in relation to ones own organizational practice, thus combining theory and practice, and transcend the border between doing research and being an organizational member in other capacities (teacher, committee member, administrator). For the committed researcher, it may reduce some of the frustrations of administrative work. We avoid the problem of the Other, i.e. constructing the natives as somebody clearly different from Oneself, as it is we rather than they that are the targets of research. We may also reduce the political-ethical problem of solely doing research downwards or targeted at groups of people which we, as social scientists, may have a, in a broad sense, competitive relationship with. (e.g. other professions). The problems and difculties are, however, substantial. These are both of intellectual and political character. In order to cope with the taken for granted assumptions and blind spots and utilize the potential the researcher needs to engage in an ambitious struggle with his/her personal and cultural framework. The paper has pointed at a number of possibilities in this respect. While a challange for the ethnograper is to avoid going native, the self-ethnographer must make strong efforts to avoid staying native. It is, of course, hardly sufcient that the researcher-native avoids excessive idiosyncrasies and gets approval of the other natives for a particular version of the world. The problem of a closed mind may be less a matter of personal bias than about belongingness to the tribes shared cultural frame. While the other natives are inclined to stay native they have no (research-motivated) reasons to escape cultural closure some difference, or tension, between the researcherrunaway and the researchers-buddies-the inmates may emerge regarding how lived social reality is best made sense of. The self-ethnographers efforts may well involve demystications and questioning of basic ideas and assumptions. Self-ethnography makes the politics of research more complicated (if one avoids painting a rosy picture of those being studied) it can not be held at arms length as is perhaps common in studies of other kinds of people. The ethical problems involved here call for careful attention.

190

MATS ALVESSON

The challenge of ethnography, and of most qualitative work, is to be close and avoid closure. The rst element is no problem in self-ethnography, the second calls for struggling. Unlike conventional ethnography the challenge is one of creating breakdowns, as one may not spontaneously experience them. Doing self-ethnography at the right moment, when the mix of local familiarity and access to other frameworks is a good one, may reduce problems. Working with theories perspectivating academic social practice in a somewhat radical mindshaking fashion may also be productive. The idea of self-ethnography pushes for intellectual curiosity not only deliminited and wellpackaged into specic projects focusing specic objects of study at safe distance from ones everyday life. Self-ethnography implies a mindset to some extent in opposition to a more technocratic-bureaucratic approach in which procedures, rules and techniques dene and legitimize the scientic project. It calls for a more reective approach in which data management matters less than a revealing, insightful account and interpretation. Self-reection is thus crucial. Self-ethnography is indeed a risky project, but may offer an interesting alternative to other approaches.

Appendix. Work methodology: Elements in self-ethnography

The following offers a very brief sketch I am no fan of cook books in research for doing self-ethnography: 1. Routinely scan what turns up in everyday life for interesting options for thick descriptions. 2. When something happens, document it, e.g. tape-record it or write it up as remembering it. Evaluate this specic material compared to other observations. 3. Acknowledge that data are constructions: revise the text so that the ctional nature of the write up becomes claried (for the author and the reader). Making explicit that an account are not simply the reporting of data, but is made up of constructions, selective reporting and of various means of framing the story line, with various (unavoidable) effects on the reader. 4. Rethink and possibly rewrite the account from an (self-)ironic position (or with inspiration of a theory running partly against ones preferences/biases): one option is to get self-distance through considering the author as a bit narcissististic; caught in his/her own subjectivity, fusing everything with aspects of one-self. 5. Coping with ethics and politics: getting input from the victims in order to balance ones ideas and biases with those of ones fellows. Consensus is not the goal, but varying viewpoints, opinions and interpretations need to be considered and worked through. If there are varied views on the subject matter, these should come through in the text and the author should motivate why s/he sticks to a particular claim despite the views of the others. Of course, those being studied

METHODOLOGY FOR CLOSE UP STUDIES

191

must approve of the study being published or widely distributed. The selfethnographer may delay publishing or take pains in order to camouage the setting being studied as part of the negotiation with the people involved. 6. Developing some tentative conclusions on the empirical material, having considered it in relation to other empirical impressions and a variety of interpretive possibilities. 7. Engage in reexive lopes: repeat parts of the process (36) after having got some new input (readings or personal experiences), either through a broad(ened) interpretative repertoire or the changes of interpretative levels, putting the tentative interpretations and conclusions in a metatheoretical perspective (see Alvesson and Skldberg 2000, for development).

Notes
1. Problems here are seen in the context of the typical purpose with interviews of getting empirical material saying something accurate about either social reality out there strategies, decisions, goals, organizational structures, interorganizational relations, etc. or subjective meanings values, intentions, ideas about students. Interviews may be used for other purposes, e.g. studying language use (discourse in operation). Interviews in themselves are, of course, not problematic. Problems emerge in the context of various uses of interview material for making different kind of empirical claims. 2. In addition to these methodological problems, one can also raise doubts of an ontological nature, debating the assumptions of the existence of stable feelings, thoughts, ideas and meanings and confront these with the idea of these aspects being discursively constituted in social interaction. While tending to agree with a processual understanding of social reality, I am here focusing on methodology rather than ontology, even though no strict separation is possible. 3. In the case of health care or social workers and patients/clients to some extent the practices of the former, relatively prestigious groups, are highlighted. As the focus is on conversations and one part is located in a weak situation, this type of study does not fully break with the tendency to focus on weak or socially low-status groups. 4. As we are talking about self-ethnographies and not questionnaire studies, still a high level of openness are characterizing the planned-systematic approach. It is planned and systematic in relationship to emergent-spontaneous research, not in relationship to for example half-structured interview studies.

References
Agar, M.H. (1986). Speaking of Ethnography. Beverly Hills: Sage. Alvesson, M. (1993). Cultural Perspectives on Organizations. Cambridge: Cambridge University Press. Alvesson, M. (1996). Communication, Power and Organization. Berlin/New York: de Gruyter. Alvesson, M. (2003). Beyond neo-positivism, romanticism and localism. A reexive approach to the research interview, Academy of Management Review 28, 1. Alvesson, M. and Deetz, S. (2000). Doing Critical Management Research. London: Sage.

192

MATS ALVESSON

Alvesson, M. and Skldberg, K. (2000). Reexive Methodology. London: Sage. Asplund, J. (1970). Om undran infr samhllet (On Reecting on Society). Lund: Argos. Atkinson, P. and Hammersley, M. (1994). Ethnography and participant observation, in Denzin, N. and Lincoln, Y. (eds.), Handbook of Qualitative Research. Thousand Oaks: Sage. Baszanger, I. and Dodier, N. (1997). Ethnography: relating the part to the whole. in Silverman, D. (ed.), Qualitative Research. London: Sage. Bourdieu, P. (1979). Outline of a Theory of Practice. Cambridge: Cambridge University Press. Brannick, T. and Coghlan, D. (2001). In Defense of Being Native: The Case for Insider Research. Paper. University College, Dublin. Brown, R.H. (1977). A Poetic for Sociology. Chicago: University of Chicago Press. Clifford, J. and Marcus, G. (eds.) (1986). Writing Culture. The Poetics and Politics of Ethnography. Los Angeles: University of California Press. Deetz, S. (1992). Democracy in an Age of Corporate Colonization. Albany: State University of New York Press. Denzin, N. (1994). The art of interpretation, in Denzin, N. and Lincoln, Y. (eds.), Handbook of Qualitative Research. Thousand Oaks: Sage. Denzin, N. (1997). Interpretive Ethnography. Thousend Oaks: Sage. Dingwall, R. (1997). Accounts, interviews and observations, In Miller, G. and Dingwall, R. (eds.), Context and Method in Qualitative Research. London, Sage. Ellis, C. and Bochner, A. (2000). Autoethnography, personal narrative, reexivity, in Denzin, N. and Lincoln, Y. (eds.), Handbook of Qualitative Research, 2nd edn. Thousand Oaks: Sage. Fairclough, N. (1993). Critical discourse analysis and the marketization of public discourse: the universities, Discourse and Society 4(2), 133168. Fontana, A. and Frey, J. (1994). Interviewing: the art of science, in Denzin, N. and Lincoln, Y. (eds.), Handbook of Qualitative Research. Thousand Oaks: Sage. Foucault, M. (1976). The History of Sexuality, Vol. 1. New York: Pantheon. Geertz, C. (1973). The Interpretation of Cultures. New York: Basic Books. Geertz, C. (1988). Work and Lives. The Anthropologist as Author. Oxford: Polity Press. Gioia, D. et al. (1995). Symbolism and strategic change in academia: The dynamics of sensemaking and inuence, Organization Science 5(3), 363383. Hammarsley, M. (1990). What is wrong with ethnography? The myth of theoretical description, Sociology 24, 597615. Jackall, R. (1988). Moral Mazes. The World of Corporate Manager. New York: Oxford University Press. Jackson, J. (1995). Dj Entendu: The liminal qualities of anthropological eldnotes, in Van Maanen, J. (ed.), Representation in Ethnograhy. Thousand Oaks: Sage. Jaggar, A. (1989). Love and knowledge. Emotion in feminist epistemology, Inquiry 32, 151 176. Knorr-Cetina, K. (1981), Introduction. The micro-sociological challenge of macro-sociology: Towards a reconstruction of social theory and methodology, in Knorr-Cetina, K. and Cicourel, A. (eds.), Advances in Social Theory and Methodology. Boston: Routledge and Kegan Paul. Kolker, A. (1996). Thrown overboard: The human costs of health care rationing, in Ellis, C. and Bochner, A. (eds.), Composing Ethnography. Walnut Creek, CA: AltaMira Press. Kunda, G. (1992). Engineering Culture. Control and Commitment in a High-Tech Corporation. Philadelphia: Temple University Press.

METHODOLOGY FOR CLOSE UP STUDIES

193

Marcus, G. and Fischer, M. (1986). Anthropology as Cultural Critique. Chicago: University of Chicago Press. Meyer, J. and Rowan, B. (1977). Institutionalized organizations: Formal structure as myth and ceremony, American Journal of Sociology 83, 340363. Miller, G. (1997). Towards ethnographies of institutional discourse: Proposal and suggestions, in Miller, G. and Dingwall, R. (eds.), Context and Method in Qualitative Research. London: Sage. Mills, C.W. (1940). Situated actions and vocabularies of motives, American Sociological Review 5, 904913. Parker, M. and Jary, D. (1995). The McUniversity: Organization, management and academic subjectivity, Organization 2(2), 319338. Perrow, C. (1978). Demystifying organizations, in Sarri, R. and Hasenfeld, Y. (eds.), The Management of Human Services. New York: Columbia University Press. Potter, J. and Wetherell, M. (1987). Discourse and Social Psychology. Beyond Attitudes and Behaviour. London: Sage. Richardson, L. (1994). Writing: A method of inquiry, in Denzin, N. and Lincoln, Y. (eds.), Handbook of Qualitative Research. Thousand Oaks: Sage. Richardson, L. (2000). Writing: A method of inquiry, in Denzin, N. and Lincoln, Y. (eds.), Handbook of Qualitative Research, 2nd edn. Thousand Oaks: Sage. Rosen, M. (1985). Breakrst at Spiros: Dramaturgy and dominance, Journal of Management 11(2), 3148. Shotter, J. and Gergen, K. (eds.) (1989). Texts of Identity. London: Sage. Shotter, J. and Gergen, K. (1994). Social construction: knowledge, self, others, and continuing the conversation, in Deetz, S. (ed.), Communication Yearbook, Vol. 17. Newbury Park: Sage. Silverman, D. (1985). Qualitative Methodology and Sociology. Aldershot: Gower. Silverman, D. (1989). Six rules of qualitative research: A post-romantic argument, Symbolic Interaction 12(2), 2540. Silverman, D. (1993). Interpreting Qualitative Data. London: Sage. Stake, R. (1984). Case studies, in Denzin, N. and Lincoln, Y. (eds.), Handbook of Qualitative Research. Thousand Oaks: Sage. Tierney, W. (1993). The cedar closet, Qualitative Studies in Education 6(4), 303314. Trowler, P. (1998). Academics Responding to Change. Milton Keynes: Open University Press. Van Maanen, J. (1988). Tales of the Field. On Writing Ethnography. Chicago, IL: The University of Chicago Press. Van Maanen, J. (ed.) (1995). Representation in Ethnograhy. Thousand Oaks: Sage. Watson, T. (1994). In Search of Management. London: Routledge. Watson, T. (1996). Motivation: Thats Maslow, isnt it?, Management Learning 27(4), 447 464. Watson, T. (2000). Ethnographic ction science, Organization 7, 493510. Widerberg, K. (1995). Kunskapens kn (Gender of knowledge). Stockholm: Norstedts. Wolcott, H. (1995). Making a study more ethnographic , in Van Maanen, J. (ed.), Representation in Ethnograhy. Thousand Oaks: Sage. Woolgar, S. (1983). Irony in the social studies of science, in Knorr-Cetina, K. and Mulkay, M. (eds.), Science Observed: Perspectives on the Social Study of Science. London: Sage.