Understanding organizational realities through discourse analysis: the case for discursive pragmatism Dan Kärreman Talk, and discourse in general, is becoming an important object of study in social sciences. This trend is part of the linguistic turn that currently is making its mark on social science (Alvesson & Kärreman 2000a, 2000b). Among other things, the linguistic turn provides new ways of thinking about language and language use, and their relation to social reality. Thus, in various disciplines, including communications studies (Deetz 1992), feminist studies (Hollway 1987), organization theory (Boden 1994, Keenoy 1997 et al, Townley 1993), and social psychology (Potter & Wetherell 1987, Potter 1997), scholars stress the importance of what has been labeled “the level of talk” (Alvesson & Kärreman 2000a, 2011a) for an adequate understanding of social processes. Indeed, organizational discourse analysis (ODA) has become one of the most important approaches towards understanding organizational phenomena (Alvesson & Kärreman 2011a, Hardy & Grant 2012) It has also ignited debate and even controversy (Alvesson & Kärreman 2011a, 2011b, 2013, Bargiela- Chiappini 2011, Iedema 2011, Mumby 2011, Hardy & Grant 2012). Generally speaking, the trend towards a more sophisticated understanding of the meaning of talk in social settings is a good thing. It undoubtedly unsettles the conventional wisdom regarding language in Page 1 of 15 JBA 3(2): 201-215 Fall 2014 © The Author(s) 2014 ISSN 2245-4217 www.cbs.dk/jba Journal of Business Anthropology, 3(2), Fall 2014 202 most parts of social science―that language is complicated, philosophically speaking, but unproblematic, methodologically speaking. However, like most subversive movements in their most subversive moments, the focus on talk and discourse is currently more persuasive as a critique of conventional understandings in social science, than as a fully formed position for constructive research. It could be argued that discourse analysis―which perhaps is the most prominent point of view pursued from an understanding of the problematic nature of talk―has progressed further than “merely” function as a form of critique (cf. Potter & Wetherell 1987). Indeed, there is much promise in the various forms of discourse analysis currently available. There is also much confusion and hype. The confusion around the term “discourse” is telling regarding the difficulties to form a qualified opinion on what counts as a proper discourse analysis. Depending on who you ask, discourse can, for example, mean a) linguistic interaction, b) any form of talk, c) systematic exchange of utterances on a particular matter, d) epochal-defining historically rooted systems of ideas, e) any forms of texts and talk, and so on. The problem of the many meanings of “discourse” cannot be solved with better definitions. Discourse analysts are right in pointing out that there are good reasons why words seldom have one meaning, and that this rarely is a problem in everyday language use. Nevertheless, however helpful this argument might be to explain the confusing state of the field of discourse analysis, it doesn’t make it any less confusing. It should be clear that the only reason that the confusion around “discourse” as a problem is that it prevents and blocks disciplined thinking about the level of talk. The best, and perhaps the only, way to clarify the field of discourse analysis, and how it might contribute to the understanding of the level of talk, is to bypass the definitions of discourse. It is more important to focus on the core assumptions―and the analytical strategies thus enabled―that underlie the various meanings of the term. Thus, there are many strong arguments for close-up studies of talk in organizations. Discourse analysis provides one productive way of studying talk, together with, for example storytelling, narrative analysis and conversation analysis. In this discussion, I focus on discourse analysis. Put bluntly, the argument I will advance is that some forms of discourse analysis are more appropriate than others when studying talk in organizational settings. This essay attempts to demonstrate why, and what difference it will make. In general terms, it aims to suggest a framework for disciplined thought on the level of talk in organizational settings. It proceeds with a discussion of key dimensions in framing discourse in various versions of discourse analysis, and puts particular emphasis on how to proceed with disciplined thought on the level of talk, while dealing with the problem of discourse reductionism. Several attempts to organize the variants of discourse analysis are Kärreman / Understanding organizational realities 203 currently available. For example, Keenoy and his co-authors (1997) distinguish between discourse analysis as a sense-making tool and discourse analysis as a revealer of ambiguity and indeterminacy. Grant and his co-authors (2004:3) describe organizational discourse as “structured collection texts embodied in the practice of talking and writing.” Potter (1997) identifies five versions of discourse analysis, three of them crafted around the project of exploring psychological dispositions among individuals, and two of them of perhaps greater interest in this context: Potter & Wetherell’s (1987) version and the standard Foucauldian positions. These two latter versions will be discussed in some detail below. Variants of discursive reductionism: small d discourse and big D discourse First, allow me to elaborate on the different takes on discourse analysis―henceforth shortened to DA―discussed above. I will start with the small d discourse approach (Alvesson & Kärreman 2000a, 2011). Potter and Wetherell (1987) provide a relatively pure, and well- reputed, version of it, which they develop from the assumption that the social world is created bottom-up: people create and construct the social world through linguistic interaction. According to Potter & Wetherell, this is a three-stage process: first, people actively create accounts on a basis of previously existing linguistic resources; second, they are continually and actively involved in selecting some of the infinite number of words and meaning constructions available, and in rejecting others; and. third, the chosen construction has its consequences: the mode of expression has an effect, it influences ideas, generates responses and so on. DA starts from the following: 1. Language is used for a variety of functions and has a variety of consequences; 2. Language is both constructed and constructive; 3. The same phenomenon can be described in several different ways; 4. Consequently there will be considerable variations in the accounts thereof; 5. There is no foolproof way yet of handling these variations, or of distinguishing accounts which are “literal” or “accurate” from those which are rhetorical or incorrect, thereby avoiding the problem of variation which faces researchers working with a more “realistic” language model; and 6. The constructive and flexible ways in which language is used should themselves be a central subject of study (Potter and Wetherell, 1987:35). Journal of Business Anthropology, 3(2), Fall 2014 204 Thus, from these authors’ perspective, discourse analysis means studying conversations, interview statements and other linguistic expressions, without drawing any conclusions that are clearly “beyond” the micro situations constituting the contexts in question. In other words, the conversations or the interviews themselves become the context of the accounts. In their book, Potter & Wetherell (1987) demonstrate how surprisingly much mileage there is in such an approach to discourse. The most powerful demonstration of the potency of their framework lies, perhaps, in their analysis of how some forms of social science does away with variation in its accounts. They claim that the failure of scholars to pay serious attention to variations in people's accounts, particularly in psychology, is an artifact of particular analytical strategies. These strategies have favored the restriction, gross categorization and selective interpretation of utterances. Restriction means that the scientist “locks” subjects by applying various techniques that force them into certain reaction patterns, as is the case in experiments, questionnaires and, to some extent, structured interviews. Gross categorization means that all accounts are arranged in a scheme of broader categories, which then operates as the primary empirical material. Thus, variation and subtle differences are purged from the material. Selective interpretation means that a predetermined idea is function as an ordering device that privilege one interpretation from other, equally possible, interpretations. Again, variation and differences are explained away or overlooked. Potter & Wetherell clearly point to processes that eclipse and distort our understanding of discursive interaction. However, their approach, although legitimate, has severe limitations. Elsewhere, together with Mats Alvesson, I have provided a detailed critique of these limitations (see Alvesson & Kärreman 2011a). Allow me to summarize our main points. First, their exclusive focus on language use is far too exclusionary. For example, studying strategy from this approach means taking an exclusive interest in the talk and communication of strategy produced in particular contexts; actual strategies are defined as being out of the scope of analysis. Second, Potter & Wetherell’s approach seems not only to separate the speaker from the spoken: it seems to eliminate him or her. Potter & Wetherell explain utterances from their effects, by what arises from them, and not by the intentions or cognitive processes lying behind what is said or written. Third, they mystify agency. For example, it could be argued that their claim that contextual contingencies are more important in determining how the conversation is going to develop, than in―say―people’s psychological dispositions. The problem is that in their approach, almost everything―excluding (some) utterances, but including people’s psychological dispositions―could count as contextual contingencies. This also means that Potter & Wetherell are unable to provide a theory that tells us what’s not included in the context. Kärreman / Understanding organizational realities 205 Big D discourse analysis draws heavily on Foucault’s ideas on the matter. Foucault understood discourses as bodies of knowledge, and thus expressions of power/knowledge-relations, “that systematically form the object of what they speak” (Foucault 1980:49). Discourse in the Foucauldian sense is less about everyday linguistic interaction, and more about historically developed systems of ideas that form institutionalized and authoritative ways of addressing a topic, to “regimes of truth.” Discourse in this sense not only shapes our particular ways of talking about a subject matter, but also shapes and constitutes our understanding of the real on the experiential level: it informs us of what is “normal,” “natural,” and “true.” Thus, both subjective and objective reality, in Berger & Luckmans (1967) sense, is constituted, constructed, and maintained through the particular discourses available in any given epoch. From an organizational analysis point of view, the perhaps biggest drawback with Foucauldian discourses is that they are claimed to constitute reality―not only in their ideational dimension, but also in their practical-behavioral dimension―yet without being able to spell out how this actually happens. In other words, the Foucauldian take on discourse lacks a clear idea on how it influences people to act in a prescribed way. On the one hand, it seems as if individuals are only embodied appendices of various discourses that have constituted the subjectivity the observer may think that s/he observes. On the other hand, Foucault empathetically underscored the possibility for the individual to exercise resistance, thereby implying the possibility of choice (cf. Foucault 1979). To sum up, although the mega-discourse approach provides an economic and elegant way of thinking on the productive and formative forces that shape social reality, it does so only by eliminating those who actually produce and form it. This is less problematic if one studies the history of the development of social forces―as Foucault primarily did―since history always starts from a perspective already chosen by actors, and when their actual choices can be recorded and interpreted. History can be reinterpreted, but it cannot be undone. However, such reduction is unacceptable if one studies history in the making. Without an adequate understanding of those performing history―their options, their choices, and their ways of reaching conclusion―their actions will always look predetermined, if only by the fact that it is possible to construct an intelligible and patterned reconstruction of them. It will come to seem as if discourse determines this or that when, in fact, the only reason things looks decided from discourse is because they are assumed to be so. Does discourse analysis afford language too much agency? The interest in organizational communication and discourse analysis is fueled by the claim that communication has a significant stake in Journal of Business Anthropology, 3(2), Fall 2014 206 constituting social reality. Since this claim is central for organization communication it seems wise to stop and think a bit about this particular claim: what it means and to what extent it matters. On one level, the claim that communication constitutes social reality is very strong. It is an inescapable fact that we see something as something. While the first something may or may not be linguistic in character, the second something always is. The second something, in the sense of as something, highlight the irreducible dimension of meaning. To make sense, reality is covered in layers of meaning. Thus, reality always reveals itself in a figurative manner. The layer of meaning is very powerful in many ways, but we should not be carried away. The layer of meaning is not all there is. Yes, it is always important, and it may in some instances be the most important thing or even the only thing, but this is far from always the case. When I was a Ph.D. student, one of my fellow students practised a running joke where he invited people to his socially constructed balcony. The punch-line was, of course, that his apartment lacked a balcony. (I didn’t say the joke was funny). Beneath the silliness of the joke lurks an important truth. Although all balconies obviously are socially constructed―after all, they hardly grow on trees―they rely on more things than imaginary meaning: for example, on concrete, steel or other materials. They also rely on performances for their actual construction: on welding and brick-laying and other forms of work. Arguably they matter more for the constitution of the balcony―in the first something sense―than the layer of meaning that embeds it. Cars, trains, bridges, houses, machines, meals―I can go on. The point is not that they exist independent of meanings and words. They clearly don’t. My point is rather that, apart from the cultural significance of these objects, their constitution is more a matter of matter and performances than of meanings and words. This is also true for certain less matter bound entities, such as the performance of services and rituals. Take for example, the cleaning of a house or a wedding. Words are important for understanding what cleaning is all about but the actual performance is likely to be performed without a word. And a wedding is clearly more dependent on performances of the bride, the groom, the officiant and whoever pays the bills, than on any words. Actually, let’s roll back the tape on the wedding. This is not really a clear type of performance (or something in the first sense) where words are not enough. To a certain extent words clearly are not enough. As I pointed out above, there is more to a wedding than words. However, words are doing important, if not the most important, work during this ceremony. A wedding is not just a party were people meet and congregate. It is also a mechanism for changing the social status among certain people. This change is almost exclusively done by words. It is accomplished Kärreman / Understanding organizational realities 207 though elaborate sequences of words that are uttered by the bride, the groom and the officiant. Without these words a wedding would just be a very expensive party. With them, complex social relations are constituted and legitimized. The wedding is not the only ritual or performance where words are playing the perhaps most significant role. In getting married or divorced, getting hired and fired, in making war and peace, in getting promotion or demotion, words are making their magic and get the things done. They may not be the only ingredients but they are surely making a critical contribution in creating and constituting reality, not just by being an irreducible dimension but by being the actual tools or agents of change and creation. Finally, for a certain class of somethings in the first sense. meaning is all we have. In a way, the something in the first sense here collapses with the something in the second sense. Take, for example, weddings again: this time not the type of weddings that you and I participate in, but rather the pure idea of the institution as it operates in, for example, a particular culture. Here we are not interested in brides, grooms, and officiants in flesh and blood. Rather, we are interested in what they represent from a symbolic point of view, and how the institution of marriage and weddings operate in, for example, the political economy, shaping gendered opportunities and division of labor, and distribution of wealth. From this point of view, the somethings we take an interest in are institutions or Discourses, and, in particular, how different institutions (in the imaginary form) interrelate and interacts in society in general. These kinds of somethings exist as words only―as abstractions or generalities. As such they shape how we think and interact in reality in very thorough and ultimately powerful way. In many ways the inert and taken for granted character of encultured institutions tend to fool us into believing that they are far less fragile and malleable than they actually are. In this sense, the emergence of gay marriage, for example, will not change the actual ritual and performance much, if at all, but it will change cultural patterns of understandings and social relations in a profound way. The discussion above has some implications, as we already have hinted, on the distinction between discourses, with a small d, and Discourse, with a capitalized D. In cases where words are not enough, discourses arguably tend to be of interest from a discourse (with a small letter d) analysis point of view. The proper place for communication analysis in these cases is about the structure and content of the conversations that occurs around these objects and performances. Although we would argue that it is a stretch to think that these objects and performances primarily are constituted through communicative processes, the analysis of communicative processes embedding them certainly would tell us important things about how they operate in social reality. Journal of Business Anthropology, 3(2), Fall 2014 208 Small d discourse analysis would also generally be helpful for understanding cases where words are doing the main job. In many ways, the value and importance of discourse analysis has been established through the analysis of the performative aspects of language use (see, for example, Potter & Wetherell 1987): when language clearly is used to accomplish things. In fact, I think that this is probably the area of analysis where small letter discourse analysis has the greatest potential, and also few competing analytical tools. Discourse analysis with a capital D, on the other hand, is eminently suitable for the analysis of ideational phenomena on an abstract level. As Foucault has demonstrated numerous times, this analytical perspective is powerful way of understanding the history and the sociology of ideas (although Foucault’s analysis of power probably is more telling regarding the sociology of ideas). In a way, it could be argued that the pure meanings of words also have performative effects, that they in fact also do things in the world. I think that this is correct to some extent, but I think the pure meaning side of words has a different type of agency in comparison to words that accomplish things. In the first case the agency lies in the framing effect of the meaning the word carries. In the second case, meaning is secondary, and the agency lies in actually uttering the words, thus connecting events rather than framing them. The key point of much communication and discourse studies is the idea that language use does not simply mirror reality out there but rather creates or constitutes it. As said above, this sometimes is insightful, but sometimes it is a less relevant or productive point. “Constitutes” covers a wide spectrum of various analytical options. It may be seen as the vital, powerful element in reality construction. “Reality”―behaviours, practices, meanings, talk―out there is ambiguous, “weak,” and offers only a general and soft input to the discourse user. It may be seen as medium- strong―social reality (as practices) or cultural and individual mindsets (or deeply anchored meanings) may show considerable resistance to most efforts to constitute reality in specific ways, although there are some variations and options. Or reality may be quite robust―we may want to constitute knowledge society as much as we want, but most part of working life may put effective stoppers to do most work mainly through using brains and intellectual analysis. Discursive pragmatism It is tempting, on the basis of the critique above, to claim that the small d discourse and big D discourse approaches are examples of bad methodology. However, it is important to remember that this is only true in a relative sense. Whether a particular methodology is good or bad is not exclusively dependent on its inherent qualities. Rather, it is highly dependent on the particular phenomenon under study. If one wants to study linguistic interaction and language use among individuals, the small Kärreman / Understanding organizational realities 209 d discourse approach will serve eminently. And although the Foucauldian notion of big D discourse might not quite cut it as Grand theory, explaining everything, it has been successful in many other respects, to say the least. In this essay, I want to propose discursive pragmatism as an alternative strategy to understanding organizational phenomena, informed by the linguistic turn. This version of discourse analysis draws on the idea that discourse and texts are important in their own right but also important as clues to extra-linguistic issues. The study of discourses (i.e., verbal interactions or written accounts) provides an opportunity to gain insight on issues more or less loosely connected to discourse. Such issues may include symbolic aspect of organizational realities (corporate culture), structural aspects of organizational work (division of labor) and interactive aspects of the construction and reproduction of particular selves (social identity). As hinted above, the discursive pragmatist approach builds on the assumption that organizational phenomena can heuristically be understood and studied in terms of talk, meaning and practices. I will, following Alvesson & Kärreman 2000a, henceforth talk about the levels of talk, meaning, and practice. Allow me some space to develop the specific character of each level, and to suggest some ideas on how to study them. The level of practice To study this level means to take interest in and address what people do and what means of employment they use to accomplish what they doing: the procedures, routines and practical arrangements activated by organizational habitants to accomplish their various tasks at hand. As Sandelands & Drazin (1989) have pointed out, to name the achievements is not enough when studying this level, since the achievements themselves contains no information on how they where achieved. Obviously, much organizational activity take place at this level, but, as Sandelands & Drazin points out empathically, few have paid enough attention to provide convincing evidence on how organizational habitants practically performs such key organizational phenomena as, for example, “decisions” and “adaptions”―if they perform them at all. To make convincing claims of studying the level of practice clearly demands some evidence of first-hand experience: participate observation and/or longitudinal contact with people performing the practices claimed to be described. Sandelands & Drazin may overestimate the possibility of making unambiguous observations, and the possibility of unambiguously communicating these observations, but they are right in their insistence on observational content. We might not have the means to resolve the different interpretations that are possible to make of the “same” observation but it is difficult to see why this must stop us from making Journal of Business Anthropology, 3(2), Fall 2014 210 observations. Making facts might be a contestable business, but this does not prevent “facts” from being informative and, in one sense or another, potentially illuminating. Ultimately, the price of making contestable observations is a lesser burden to bear, compared to the price of not making observations at all. The level of talk We are all more or less aware of “the capacity of betrayal” in language – that words might tell lies, might be used incompetently and might for various reasons be misleading. In that sense, everybody probably agree that truth-claims in utterances may be problematic to a greater or lesser extent. Discourse analyst, however, also points out that the representational validity of utterances―the extent to which what is verbally claimed also corresponds with realities―might be the least interesting thing to study in regard to language use. This is so, they argue, not only because the validity of certain forms of speak might be very difficult, if not impossible, to assess―as it is in the case of talk on mental states, since our only way to find out about peoples mental states is through what they tell us―but also because the simpleminded focus on representational validity obfuscates the constructive and contextual character of language use. Proponents of a micro-discursive approach emphasize that language use is a worthy object of study in its own right (cf. Potter & Wetherell 1987). Thus the expressive modes used, the discursive moves deployed, and the interpretive repertoires utilized in attempts to construct particular realities, and to respond to various contextual pressures becomes the focus of investigation. At the level of talk, what matters is the effects created by utterances, and how various utterances connects to create similar effects, and not what utterances stands for. While it is counter-productive to constrict all organizational research exclusively to the level of talk, it seems quite clear that this level of study is all too easily overlooked in organizational analysis, although much less so since the arrival of organizational discourse analysis as a robust methodology for studying of organizational phenomena. A focus on discourse provides insight that facilitate and enriches reflective thinking on organizational phenomena in general. It also provides ideas that points out underdeveloped domains of research (e.g. how do organizational members speak in certain situations, what do they achieve with these forms of speak. etc), and ways and means that makes it possible to account for extensive study of the level of talk. The level of meaning The level of practice and the level of talk constitute important areas of Kärreman / Understanding organizational realities 211 research but any study stays incomplete without the incorporation and consideration of the level of meaning. In one sense, the level of meaning is always present in any attempt to make sense of what people are doing or saying, since we cannot see something without seeing it as something (Asplund 1970, Geertz 1973). Seeing people engaged in verbal exchange as talking to each other already is in itself an interpretation―and this remains true if one chooses to see them as carrying out a particular practice. The meanings produced by researcher to make sense of the phenomena under investigation is but one aspect of the level of meaning. In the context of empirical research, the meanings produced by the people belonging to the field is perhaps of even greater significance. To consider the level of meaning means ―apart from being aware of that all observations to some extent also are interpretations (Geertz 1973)―to take interest in what people mean with expressions they are using and what meaning they ascribe to the practices they, and others, are deploying. In Geertz’s (1973) famous formulation, it means to engage in thick, rather than thin, descriptions: extracting, explicating, and translating the many layers of meaning in cultural material, hidden for the uninitiated. To study the level of meaning in terms of the discourse produced in the settings studied includes, for example, efforts in figuring out both particular meaning and more stable and invariant meaning. And in terms of the practices employed by the inhabitants, any serious interest in the level of meaning includes attempts to figure out what these practices mean for those who engage in them―what sense they make of them. Thus, both the approach advocated by Sandelands & Drazin (1989), and the approach suggested by Potter & Wetherell (1987) demonstrate a certain “thinness,” in the Geertzian sense. Extensive study of the practices employed in organizational settings may reveal what inhabitants actually are doing and how they accomplish things, at least from the researchers point of view. But it will provide little, if any, information on what count as significant practices in the setting (with the exception of what the researcher deem as significant) and it will say even less on what the practices mean to inhabitants and possible consequences thereof. The same is true if one exclusively sticks to the level of talk. Studies on the discourses deployed by inhabitants in organizational settings may reveal the many―or, for that matter, the few―ways people use language to produce certain effects. But as long as one restricts oneself to the level of talk, those who actually are talking becomes curiously fugitive. They become receiver/transmitters of discourse, language game prosumers stripped off any intentional capacity, engaged in games whose structure is clear, but whose point is not. Journal of Business Anthropology, 3(2), Fall 2014 212 Conclusion This essay critiques the reductionism inherent in many version of discourse analysis. It proposes an alternative take: that of discursive pragmatism. Nevertheless, discursive pragmatism cannot escape being reductionist. It is, after all, unsustainable to make claims of a complete and exhaustive understanding of the phenomena under investigation; to provide the complete story. The problem is not to exclude the “false” stories from the “true” ones, but to exclude unsupported or trivial stories from empirically grounded and telling stories. It makes much more sense, from this vantage point, to focus on particular situations―a meeting, an interaction, an event in the flow of organizational life – and elaborate its many facets, rather than to try to say something―almost certainly something thinly―on all the events, interaction and events spotted during fieldwork. This is only possible if one engages in a careful construction of the field work, the particular field, and the field situations, that make it possible to defend claims on the level of practice with observational evidence, claims on the level of meaning with ethnographic evidence, and claims on the level of talk with conversational evidence. The task is to construct illuminating and manageable sections of the realities at hand in fieldwork situation. This also makes it possible to raise defendable claims about actual interrelations between talk, meanings, and practices, since this makes it possible to ground these claims on actual empirical situations. This type of deep grounding enables the researcher to say something revealing, not only about the situation explicitly analyzed, but also about matters stretching out of the situation. Discursive pragmatism may delimit the space available for claims on general and broad patterns in the setting under investigation. But it is premature to decide whether this is a genuine difficulty with the approach or not. Actually, it can be argued that discursive pragmatism provides a better foundation for claims on general patterns than conventional approaches do. The empirical generalizations distilled in the typical case study depend on decontextualization. Starting off from context dependent empirical findings the researcher suddenly and mysteriously finds himself or herself in the decontextualized world of theory. Claims of general relevance in the discursive pragmatist approach depend rather on microanchored theory (Knorr-Cetina 1981) and on the inter- and intracontextual character of the findings. Discursive pragmatism admits for systematic comparisons within the setting where field work has been carried out, and thus, for example, for claims of findings on the level of meaning that appears to be stable across various contexts. As I pointed out above, discursive pragmatism necessarily includes reductionist moves. However, the reduction inherent in discursive pragmatism is less problematic, from an organizational analysis point of Kärreman / Understanding organizational realities 213 view, than the reductionism in other forms of discourse analysis. This is because discursive pragmatism targets the levels of interaction and social structure. As Wiley (1988) points out, these levels are of particular concern from an organizational point of view. Discursive pragmatism relates particular talk (interaction) to particular contexts (social structure), and also takes an interest in how talk and meanings migrate between contexts. In effect discursive pragmatism take particular interest in the intersection between interaction and social structure, i.e. precisely where organizational phenomena occur (Wiley 1988). As a consequence, the levels of the individual and culture are less emphasized and visible. This is, of course, an inevitable cost, and if the research project aims to say something significant about individuals and culture per se, discursive pragmatism obviously is not an optimal choice, from a methodology point of view. However, individuals and culture per se is rarely the main focus in studies of organizational phenomena, and if they are, this is probably due to cognitive mistakes in the design of the study (i.e. denying of failing to grasp the layered “nature” of social reality). References Alvesson M. & Kärreman, D. (2000a). Varieties of discourse: On the study of organizations through discourse analysis. Human Relations, 53(9), 1125–49. Alvesson M. & Kärreman, D. (2000b). Taking the linguistic turn in organizational research: challenges, responses, consequences. Journal of Applied Behavioural Science, 36(2), 134-156. Alvesson M. & Kärreman, D. (2011a). Decolonializing discourse: Critical reflections on organizational discourse analysis. Human Relations, 64(9), 1121-1146. Alvesson M. & Kärreman D. (2011b). Organizational Discourse analysis – well done or too rare? Human Relations, 64(9), 1193-1202. Asplund, J. (1970). Om undran inför samhället. Argos. Bargiela-Chiappini, F. (2011). Discourse (s), social construction and language practices: In conversation with Alvesson and Kärreman. Human Relations, 64(9), 1177-1191. Boden, D. (1994). The business of talk: Organizations in action. Deetz, S. (1992). Democracy in an age of corporate colonization: Developments in communication and the politics of everyday life. SUNY Press. Foucault, M. (1980). Power/Knowledge. New York: Pantheon. Journal of Business Anthropology, 3(2), Fall 2014 214 Geertz, C. (1973). The Interpretation of Cultures. New York: Basic Books. Grant, D., Hardy, C, Oswick, C & Putnam, L. (2004). (Eds.). Handbook of Organizational Discourse. London: SAGE. Hardy, C & Grant, D. (2012). Readers beware: provocation, problematization and … problems. Human Relations, 65(5), 547-566. Hardy, C., Lawrence, T & Grant, D. (2005). Discourse and collaboration: the role of conversations and collective identity. Academy of Management Review 30, 58-77. Iedema R (2011) Discourse studies in the 21st century: A response to Mats Alvesson and Dan Kärreman’s “Decolonializing discourse’. Human Relations 64. (9): 1163-1176. Keenoy, T., Oswick, C., & Grant, D. (1997). Organizational discourses: Text and context. Organization, 4(2), 147-157. Knorr-Cetina, K. (1981). The micro-sociological challenge of macro- sociology: towarda reconstruction of social theory and methodology. In Knorr-Cetina, K., & Cicourel, A. V. (Eds.), Advances in social theory and methodology: Toward an integration of micro-and macro-sociologies pp. 1- 48). Routledge. Mumby DK. (2011). What’s cooking in organizational discourse studies? A response to Alvesson and Kärreman. Human Relations, 64, 1147-1161. Phillips, N & Oswick, C. (2012). Organizational discourse. Academy of Management Annals. Potter, J. (1997). Discourse analysis as a way of analysing naturally occurring talk. Qualitative research: Theory, method and practice, 144-160. Potter, J. & Wetherell, M. (1987). Discourse and Social Psychology. Beyond Attitudes and Behaviour. London: SAGE. Sandelands, L., & Drazin, R. (1989). On the language of organization theory. Organization Studies, 10(4), 457-477. Townley, B. (1993). Foucault, power/knowledge, and its relevance for human resource management. Academy of Management Review. 18, 518– 545. Wiley, N. (1988). The micro-macro problem in social theory. Sociological Theory, 6, 254-261. Kärreman / Understanding organizational realities 215 Dan Kärreman is Professor in Management and Organization Studies at Copenhagen Business School, and Professor in Management at Royal Holloway, University of London. He is also affiliated to the Lumos group at Lund University. His research interests include critical management studies, knowledge work, identity in organizations, leadership, organizational control and research methodology, and he has published his work in Academy of Management Review, Human Relations, Journal of Management Studies, Organization, Organization Science, and Organization Studies, among others. His most recent book is Qualitative Methodology and Theory Development: Mystery as Method (SAGE 2011, with Mats Alvesson). He may be reached at dk.ikl@cbs.dk mailto:dk.ikl@cbs.dk