Skim03_Generating RQ through Problemaztization

஽ Academy of Management Review 2011, Vol. 36, No. 2, 247–271. GENERATING RESEARCH QUESTIONS THROUGH PROBLEMATIZATION MATS ALVESSON University of Lund and University of Queensland JO¨ RGEN SANDBERG University of Queensland It is increasingly recognized that what makes a theory interesting and influential is that it challenges our assumptions in some significant way. However, established ways for arriving at research questions mean spotting or constructing gaps in existing theories rather than challenging their assumptions. We propose problematization as a methodology for identifying and challenging assumptions underlying existing lit- erature and, based on that, formulating research questions that are likely to lead to more influential theories. As researchers, we all want to produce inter- lenging the assumptions underlying existing esting and influential theories. The dominant theories, therefore appears to be a central ingre- view is that a theory becomes influential if it is dient in the development of more interesting regarded as true. However, in his seminal study and influential theories within management Davis (1971) showed that what makes a theory studies. However, established ways of generat- notable, and sometimes even famous (Davis, ing research questions rarely express more am- 1986), is not only that it is seen as true but also, bitious and systematic attempts to challenge and more important, that it is seen as challeng- the assumptions underlying existing theories ing the assumptions underlying existing theo- (Barrett & Walsham, 2004; Bartunek et al., 2006; ries in some significant way. During the last four Clark & Wright, 2009; Johnson, 2003; Locke & decades, a large number of researchers within Golden-Biddle, 1997; Sandberg & Alvesson, management and the social sciences have con- 2011). Instead, they mainly try to identify or cre- firmed and elaborated Davis’s original thesis in ate gaps in existing literature that need to be various ways (e.g., Astley, 1985; Bartunek, Rynes, filled. It is common to refer either positively or & Ireland, 2006; Black, 2000; Campbell, Daft, & mildly critically to earlier studies in order to Hulin, 1982; Daft, 1983; Daft, Griffin, & Yates, “extend . . . this literature” (Westphal & Khanna, 1987; Daft & Lewin, 1990; Davis, 1999; Hargens, 2003: 363), to “address this gap in the literature” 2000; Lundberg, 1976; Miner, 1984; Mohr, 1982; (Musson & Tietze, 2004: 1301), to “fill this gap” Weick, 1989, 2001; Wicker, 1985). For example, (Lu¨ scher & Lewis, 2008: 221), to point at themes McKinley, Mone, and Moon (1999) showed that that others “have not paid particular attention for a theory to receive attention and establish a to” (Thornborrow & Brown, 2009: 356), or to “call new theoretical school, it must differ signifi- for more empirical research” (Ewenstein & cantly from, and at the same time be connected Whyte, 2009: 7). Such “gap-spotting” means that to, established literature in order to be seen as the assumptions underlying existing literature meaningful. Likewise, Bartunek et al.’s study of for the most part remain unchallenged in the what the board members of the Academy of formulation of research questions. In other Management Journal considered to be particu- words, gap-spotting tends to underproblematize larly interesting empirical articles provided existing literature and, thus, reinforces rather “support for Davis’s (1971) arguments regarding than challenges already influential theories. theory: empirical articles that challenge current assumptions are also particularly likely to be There are, however, an increasing number of viewed as interesting” (2006: 12). research orientations that directly or indirectly encourage problematization, such as certain Generating research questions through prob- versions of social constructionism, postmodern- lematizion, in the sense of identifying and chal- ism, feminism, and critical theory. Since the pri- 247 Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express written permission. Users may print, download, or email articles for individual use only.

248 Academy of Management Review April mary aim of many of these orientations is to relate to existing work (Alvesson, Hardy, & Har- disrupt rather than build upon and extend an ley, 2008). established body of literature, it could be ar- gued that they tend to overproblematize the re- The article is structured as follows. We begin search undertaken. In particular, these orienta- by placing problematization in its methodologi- tions tend to emphasize the “capacity to disturb cal context by discussing prevalent ways of gen- and threaten the stability of positive forms of erating research questions from existing litera- management science” (Knights, 1992: 533) as a ture. Against this background, we elaborate and way to highlight what is ”wrong” (e.g., mislead- propose problematization as a methodology for ing or dangerous) with existing knowledge generating research questions, in four steps: (1) (Deetz, 1996)—that is, ”negative” knowledge is we describe the aim and focal point of the meth- the aim (Knights, 1992). For a large majority of odology as challenging assumptions underlying researchers with a more ”positive” research existing literature; (2) we elaborate a typology agenda—with the aim of advancing knowledge consisting of five broad types of assumptions of a specific subject matter—such overprob- that are open for problematization in existing lematization is often seen as inappropriate and theory; (3) we develop a set of methodological unhelpful (Parker, 1991; Rorty, 1992). principles for identifying, articulating, and chal- lenging assumptions underlying existing litera- Our aim in this study is to integrate the posi- ture; and (4) we examine how the developed tive and the negative research agenda by devel- methodology can be used for generating re- oping and proposing problematization as a search questions by applying it to Dutton, Duke- methodology for identifying and challenging as- rich, and Harquail’s (1994) well-known article sumptions that underlie existing theories and, about organizational identity. Finally, we dis- based on that, generating research questions cuss what contributions the methodology can that lead to the development of more interesting make to theory development within manage- and influential theories within management ment studies. studies. To be more specific, (1) we develop a typology of what types of assumptions can be GAP-SPOTTING: THE PREVALENT WAY OF problematized in existing theories, and (2) we GENERATING RESEARCH QUESTIONS propose a set of methodological principles for how this can be done. A wide range of studies points to important ingredients involved in formulating good re- We focus only on problematizing assumptions search questions (e.g., Abbott, 2004; Astley, 1985; that underlie existing literature as a way to con- Becker, 1998; Davis, 1971, 1986; Frost & Stablein, struct research questions. We do not discuss 1992; Locke & Golden-Biddle, 1997; Mills, 1959; how other aspects of the research process, such Smith & Hitt, 2005; Starbuck, 2006; Van de Ven, as general interest, relevance for practitioners, 2007; Weick, 1989). However, few of these studies choice of case, and unexpected empirical find- have focused specifically on how researchers ings, may influence the research objective and, construct research questions by reviewing and thus, the formulation of research questions. criticizing existing literature. For example, There is also a large and overlapping body of while Abbott (2004) offers an array of heuristic literature on reflexivity dealing with key as- tools and Becker (1998) suggests a set of tricks of pects of research (e.g., Alvesson, Hardy, & Har- the trade for coming up with new research ley, 2008; Alvesson & Sko¨ ldberg, 2009; Hardy & ideas, these heuristics and tricks “are not spe- Clegg, 1997; Lynch, 2000; Westwood & Clegg, cifically aimed at any particular phase or aspect 2003). Since our emphasis is on how to work with of the research process” (Abbott, 2004: 112). reflexivity when formulating research ques- tions, we only marginally address other issues Prevalent Ways of Constructing Research of reflexivity in research, such as invoking Questions from Existing Literature awareness of the researcher him/herself, the role of rhetoric, and ongoing constructions of A study that comes close to how researchers reality in the research process. An exception is construct research questions from research texts the theme of the sociopolitical context of re- is Locke and Golden-Biddle’s (1997) investiga- search, which is a key issue for how researchers tion of how researchers create an opportunity for

2011 Alvesson and Sandberg 249 contribution in scholarly journals. They con- monly construct gaps by arranging existing ducted an empirical investigation of eighty-two studies in specific ways. For example, one way qualitative articles published in the Administra- to create a gap, identified by Locke and Golden- tive Science Quarterly (sixty-one studies) and Biddle, is to synthesize coherence in which the the Academy of Management Journal (twenty- researcher “cite[s] and draw[s] connections be- one studies) between 1976 and 1996. All of the tween works and investigative streams not typ- studies, except eight, created opportunities for ically cited together . . . [which] suggests the ex- contribution by arguing that existing literature istence of underdeveloped research areas” was either incomplete or had overlooked an im- (1997: 1030). A gap in existing literature may also portant perspective and that those were gaps be defined by specific negotiations between re- that needed to be filled. The remaining eight searchers, editors, and reviewers about what articles claimed that existing literature was studies actually constitute existing literature misleading in the way it produced knowledge and what is lacking from that domain of litera- about a specific topic. A contribution then de- ture (Bedeian, 2003, 2004; Tsang & Frey, 2007). pended on providing a superior study that was Moreover, gap-spotting is not something fixed; it able to correct faulty or inadequate existing lit- may differ in both size and complexity, such as erature. These findings by Locke and Golden- identifying or constructing fairly narrow gaps to Biddle (1997) have been confirmed in more re- more significant gaps, which can lead to impor- cent studies in the areas of information systems tant revisions and development of existing liter- (Barrett & Walsham, 2004) and marketing (John- ature (Colquitt & Zapata-Phelan, 2007). son, 2003). Nevertheless, regardless of variations in size In a more current study of management jour- and complexity, and regardless of the fact that nals, we specifically investigated how manage- researchers often creatively construct gaps in ment researchers constructed research ques- existing literature and criticize it for being defi- tions from existing literature (Sandberg & cient in some way (e.g., for being incomplete, Alvesson, 2011). In contrast to Locke and Golden- inadequate, inconclusive, or underdeveloped), Biddle’s, our study comprised a broader set of they rarely challenge the literature’s underlying journals and a mix of qualitative and quantita- assumptions in any significant way. Instead, tive studies. We analyzed fifty-two articles from they build on (or around) existing literature to eight randomly selected issues, between 2003 formulate research questions. In other words, and 2005, of Administrative Science Quarterly, whether researchers merely identify or cre- Journal of Management Studies, Organization, atively construct gaps in existing literature, they and Organization Studies. In all of the studies still adhere to the same purpose—namely, “gap- investigated, researchers generated research filling”—that is, adding something to existing questions by identifying or constructing specific literature, not identifying and challenging its gaps in existing literature. They tried to either underlying assumptions, and, based on that, for- identify competing explanations, to scan for mulating new and original research questions. overlooked areas, or to search for shortages of a particular theory or perspective in existing liter- The dominance of gap-spotting is not, as one ature. Then, based on those gaps, they formu- may assume, confined to quantitative or quali- lated their own research questions. tative hypothetico-deductive research; it is also prevalent within qualitative-inductive research. These studies suggest gap-spotting (i.e., iden- This is clearly the case in our earlier study tifying or constructing gaps in existing litera- (Sandberg & Alvesson, 2011) but particularly no- ture that need to be filled) is the most dominant ticeable in Locke and Golden-Biddle’s (1997) in- way of generating research questions from ex- vestigation of eighty-two qualitative studies, of isting literature in management. It is, however, which a large majority had an inductive re- important to note that gap-spotting rarely in- search design. The prevalence of gap-spotting volves a simple identification of obvious gaps in in qualitative inductive research is also evident a given body of literature. Instead, it consists of in Lee, Mitchell, and Sablynski’s (1999) review of complex, constructive, and sometimes creative qualitative research in organizational science processes. As both the Sandberg and Alvesson during the period 1979 to 1999, as well in Bluhm, (2011) and, in particular, Locke and Golden- Harman, Lee, and Mitchell’s (2010) follow-up Biddle (1997) studies show, researchers com- study of the period 1999 to 2008. And it is further

250 Academy of Management Review April substantiated by Colquitt and Zapata-Phelan’s gap-spotting rhetoric when presenting their re- (2007) study of trends in the theoretical contribu- search in order to get published (Starbuck, 2003, tion and impact of theory-building research and 2006). According to Starbuck, “Authors can in- theory-testing research based on a sample of crease their acceptance of their innovations by 770 articles published in the Academy of Man- portraying them as being incremental enhance- agement Journal between 1963 and 2007. Their ments of wide-spread beliefs” (2003: 349). (See results indicated “that the typical [inductive re- also Bourdieu [1996], Knorr-Cetina [1981], Latour search] article published in AMJ during our five- and Woolgar [1979], and Mulkay and Gilbert decade span either examined effects that had [1983] for the difference between researchers’ been the subject of prior theorizing or introduced work and their publications.) a new mediator or moderator of an existing re- lationship or process” (2007: 1290). A closely related explanation of the wide- spread use of gap-spotting is the political con- The widespread activity of gap-spotting in text in which most management research takes qualitative inductive research is further con- place (Alvesson, Ashcraft, & Thomas, 2008; firmed in recent editorial advice in the Academy Bourdieu, 2004; McMullen & Shepard, 2006; of Management Journal to researchers and re- Sandberg & Alvesson, 2011). It is well known viewers about what characterizes high-quality that tenure, promotion, and funding decisions qualitative research. According to the editor, an are heavily dependent on being able to publish important feature of high-quality qualitative in- regularly in quality journals. Challenging as- ductive research is that it discusses “why this sumptions that underlie existing studies is often qualitative research is needed. . . . For inductive risky, since it means questioning existing power studies, articulating one’s motivation not only relations in a scientific field, which may result involves reviewing the literature to illustrate in upsetting colleagues, reviewers, and editors some ‘gaps’ in prior research, but also explain- and, thus, may reduce the chances of having an ing why it is important to fill this gap. The latter article published (Bourdieu, 2004; Breslau, 1997; is often forgotten” (Pratt, 2009: 858). In a similar Starbuck, 2003). Therefore, in order to increase vein, but more generally, based on her twenty- the chances of being published, many research- six years as Administrative Science Quarterly’s ers may carry out gap-spotting rather than more managing editor (and her reading of more than consensus-challenging research (McMullen & 19,000 reviews and more than 8,000 decision let- Shepard, 2006; Sandberg & Alvesson, 2011). ters), Johanson offers the following core advice to authors about what journal reviewers expect However, given the increased acknowledg- of the scholarly publication: “If you can’t make a ment that challenging the assumptions underly- convincing argument that you are filling an im- ing existing literature is what makes a theory portant gap in the literature, you will have a interesting, it seems odd if authors in general hard time establishing that you have a contri- deliberately choose to construct research ques- bution to make to that literature. You might be tions through gap-spotting, or if they try to surprised at how many authors miss this funda- downplay or conceal a strong contribution by mental point” (2007: 292). dressing it up in gap-spotting rhetoric. It is also likely that reviewers would pick up and chal- The above findings and studies showing the lenge a discrepancy between a research pur- prevalence of gap-spotting research in manage- pose that was presented in gap-spotting dis- ment studies can, of course, be questioned in course but produced results that challenged the various ways. For example, both the Locke and literature. Moreover, irrespective of how re- Golden-Biddle (1997) and Sandberg and Alves- searchers actually go about formulating and son (2011) analyses are based on how research- reformulating their research questions, and re- ers presented their studies in published articles, gardless of what social and political norms which might have deviated from how they “re- influence their presentation in journal articles, it ally” went about generating their research is, as noted in Sandberg and Alvesson, “in the questions. Rhetorical conventions may account crafting of the research text that the final re- for how authors present their research in pub- search question is constructed, which is the one lished texts. Perhaps some researchers prob- that specifies the actual contribution of the lematize the assumptions that underlie existing study” (2010: 25). In other words, assumption- theory to generate research questions but use a challenging research is of limited value if it is

2011 Alvesson and Sandberg 251 not clearly shown in the published research text. In a similar vein, the editors of the Academy of There are, therefore, strong reasons to take the Management Journal argued that while the jour- research questions as stated in the published nal is publishing “technically competent re- research text very seriously and not regard them search that simultaneously contributes to the- as less important than the research questions in ory . . . [it is] desirable to raise the proportion of operation during the early stages of the re- articles published in AMJ that are regarded as search project, which eventually lead up to pub- important, competently executed, and really in- lication. teresting” (i.e., assumption-challenging studies; Bartunek et al., 2006: 9). Gap-Spotting: An Increasingly Disturbing Problem in Management Studies The above editorial observations, along with others (e.g., Starbuck, 2006), suggest that the The dominance of research seeking the incre- scarcity of more interesting and influential the- mental gains of gap-spotting has, over the last ories is a serious problem in management stud- two decades, increasingly come to be seen as a ies, and to some extent also in social science as disturbing problem in management studies. For a whole (Delanty, 2005). There seems to be a example, in their editorial comments in the in- broadly shared sense in management that the augural issue of Organization Science, Daft and field is stronger in producing rigor than it is in Lewin observed a strong “need for reorienting producing interesting and influential theories [organizational] research away from incremen- (see also Sutton & Staw, 1995). It is unlikely that tal, footnote-on-footnote research as the norm for further efforts to develop existing or new gap- the field” (1990: 1). Reflecting back on the years spotting strategies will overcome the shortage since launching Organization Science, Daft and of high-impact research. This is not to say that Lewin (2008: 177) conceded that their original gap-spotting research is unimportant. It plays a mission had not been realized. They reempha- crucial role in developing existing management sized the need not to prioritize rigorous empiri- literature through systematic and incremental cal research methods but, instead, “new theo- additions, as well as through identifying and ries and ways of thinking about organizations, addressing more significant gaps in it. However, coupled with a plausible methodology that because gap-spotting does not deliberately try grounds the theory” (2008: 182). to challenge the assumptions that underlie ex- isting literature, it is less likely to raise the pro- The outgoing editors of the Journal of Manage- portion of high-impact theories within the man- ment Studies made similar observations in their agement field. It therefore seems vital to support concluding reflections on the management field. and strengthen attempts at more deliberate, sys- Based on their six years in office (2003–2008), tematic, and ambitious problematization, both they commented that while as a research ideal and as a methodology for constructing research questions. As an addition we along with many other journals have wit- to gap-spotting, we aim in this article to develop nessed a proliferation of articles submitted, it is problematization as a methodology for chal- hard to conclude that this has been accompanied lenging assumptions underlying existing litera- by a corresponding increase in papers that add ture and, based on that, to formulate research significantly to the discipline. More is being pro- questions that may lead to more interesting and duced but the big impact papers remain elu- influential theories. sive. . . . The emphasis on improving the rigour of theorizing and of empirical method . . . may have PROBLEMATIZATION AS A METHODOLOGY led to more incremental research questions being FOR GENERATING RESEARCH QUESTIONS addressed. . . . [And] the impact of the audit cul- ture and incentive system is likely to affect the In this section we develop problematization extent to which both junior faculty and, some- as a methodology for generating research ques- what surprisingly, highly competent senior fac- tions. We first describe the aim and focal point ulty (McMullen & Shepherd, 2006) engage in con- of the methodology. We then elaborate a typol- sensus-challenging research. The emphasis on ogy that specifies which assumptions are open “gap filling” seems to assume that we know what for problematization and follow this with a set of the boundaries of a field look like and tends to principles for identifying, articulating, and chal- dissuade examination of new areas outside this matrix (Clark & Wright, 2009: 6).

252 Academy of Management Review April lenging assumptions underlying existing litera- the assumptions underlying their own perspec- ture and, based on that, constructing research tive. Instead, our idea is to use problematization questions that will lead to the development of as a methodology for challenging the assump- more interesting and influential theories. tions that underlie not only others’ but also one’s own theoretical position and, based on that, to The Aim of the Problematization Methodology construct novel research questions. This is not to say that a problematizer is “blank” or position Although gap-spotting and problematization free. Any problematization necessarily takes its are two distinct ways of constructing research point of departure within a specific metatheo- questions from existing literature, it must be retical position (i.e., epistemological and onto- recognized that they are not mutually exclusive logical stance; Tsoukas & Knudsen, 2004: Chap- (Dewey, 1938; Foucault, 1972; Freire, 1970; Locke ter 1). The ambition is therefore not, nor is it & Golden-Biddle, 1997; Mills, 1959). Any prob- possible, to totally undo one’s own position; lematization of a literature domain calls for rather, it is to unpack it sufficiently so that some some scrutiny of particular debates, critiques, of one’s ordinary held assumptions can be scru- and possibly earlier challenges of assumptions tinized and reconsidered in the process of con- in the domain, and most gap-spotting efforts structing novel research questions. This unpack- involve some form of modest problematization ing is crucial because, as Slife and Williams (in the wider sense of the word—i.e., critical note, scrutiny). However, we do not see gap-spotting as a genuine form of problematization since it to truly evaluate and understand the ideas be- does not deliberately try to identify and chal- hind other ideas, we must have a point of com- lenge the assumptions underlying existing liter- parison. We must have some contrast with im- ature in the process of constructing research plicit ideas or they will not look like ideas. They questions. will look like common sense or truth or axioms rather than the points of view that they really are There are stronger elements of problematiza- (1995: 71). tion in debates between advocates of various schools and paradigms (Abbott, 2001, 2004; Bur- Hence, instead of spotting gaps within a liter- rell & Morgan, 1979; Donaldson, 1985; Reed, 1985, ature domain or applying a prepackaged prob- 2004), as well as within more radical orienta- lematization to challenge the assumptions of tions, such as postmodernism and critical the- others, the aim of the problematization method- ory. However, although many of the paradigm ology proposed here is to come up with novel warriors and proponents of more radical orien- research questions through a dialectical interro- tations forcefully critique existing theories, their gation of one’s own familiar position, other problematizations are often secondary in the stances, and the domain of literature targeted sense that they are more or less “ready-made” for assumption challenging. In such a method- by master thinkers, such as a Baudrillardian ology, paradigm and other broader debates, (Grandy & Mills, 2004) or a Foucauldian perspec- such as behaviorism and culturalism, contextu- tive on a particular field (e.g., Knights & Morgan, alism and noncontexualism, and choice and 1991; Townley, 1993). Similarly, countertexts, like constraint (Abbott, 2004: 162–210), and critical Donaldson’s (1985), typically aim to defend or frameworks, such as political (Alvesson & Will- reinforce a preferred position but do not offer mott, 1996; Burrell & Morgan, 1979; Foucault, new points of departure. As Abbott notes, per- 1977), linguistic (Grant, Hardy, Oswick, & Put- spectives with a ready-made stance toward so- nam, 2004), constructionist (Gergen, 1992; Sand- cial life often have “stock questions and puzzles berg, 2001), and postmodernist (Cooper & Burrell, about it (as in the feminist’s questions ‘what 1988; Deetz, 1992; Knights, 1992; Rosenau, 1992), about women and social networks?’ ‘what about as well as counterresponses to these (e.g., Don- a gendered concept of narrative?’ and so on)” aldson, 1985; Reed, 2004), are seen as important (2004: 85). methodological resources to open up and scruti- nize assumptions underlying established theo- We therefore do not see such prepackaged ries, including, to some extent, the favorite the- problematization attempts as genuine either, ory of the problematizer. Such a methodology because they apply rather than challenge the supports a more reflective scholarly attitude in literature they follow, thus mainly reproducing the sense that it encourages the researcher not

2011 Alvesson and Sandberg 253 only to use his or her own favorite theoretical critical to grasp, because “if a theory is to be position but to start “using different standard properly used or tested, the theorist’s implicit stances to question one another . . . [and combin- assumptions which form the boundaries of the ing them] into far more complex forms of question- theory must be understood” (1989: 498). How- ing than any one of them can produce alone” (Ab- ever, understanding the assumptions that un- bott, 2004: 87). derpin existing theories is important not only for being able to use and test them but also for Thus, by elaborating and proposing prob- being able to develop new theories. In partic- lematization as a methodology for generating ular, without understanding the assumptions research questions, we do not take any particu- that underlie existing theories, it is not possi- lar paradigmatic stance more than we embrace ble to problematize them and, based on that, to the general and long-held metatheoretical as- construct research questions that may lead to sumption within academia that all knowledge is the development of more interesting and influ- uncertain, truths or theories cannot be accepted ential theories (e.g., Davis, 1971). as given, researchers tend to be conformist and paradigm bound (Kuhn, 1970), and theoretical Challenging Assumptions: The Focal Point in developments are partly based on rethinking Generating Research Questions Through and challenging fundamental assumptions un- Problematization derlying dominating theories (Tsoukas & Knud- sen, 2004). In other words, problematization, as But how can we problematize assumptions in we define it here, can, in principle, be applied to a way that generates novel research questions? all theoretical traditions or methodological con- Although problematization is featured in vari- victions and can be used within, and against, ous theoretical orientations, such as pragma- all, including the problematizer him/herself. tism (Dewey, 1916) and actor-network theory (Callon, 1980), Foucault’s conceptualization is a A Note on Theory good starting point (Castels, 1994; Deacon, 2000). According to Foucault, problematization is first Before elaborating problematization as a and foremost an “endeavour to know how and to methodology for generating research questions what extent it might be possible to think differ- more specifically, it is important to describe ently, instead of what is already known” (1985: what we mean by “theory.” Since there are many 9). Such an endeavor does not primarily ques- views on theories in the management field tion how well some constructs or relationships (Colquitt & Zapata-Phelan, 2007; DiMaggio, 1995; between constructs represent a particular sub- Sutton & Staw, 1995), and since these views are ject matter like “motivation” or “diversity.” In- in various ways part of what can and should be stead, it questions the necessary presupposi- targeted for assumption challenging, we are not tions researchers make about a subject matter asserting a strict view on theory. Bacharach’s in order to develop the specific theory about it. (1989) definition probably comes closest to the wide-ranging view of theory that we adopt here. As a range of scholars have noted (Bourdieu, He defines theory as 1996; Derrida, 1978/1967; Heidegger, 1981/1927; Husserl, 1970/1900 –1901; Merleau-Ponty, 1962/ a statement of relations among concepts within a 1945), assumptions work as a starting point for boundary set of assumptions and constraints. It is knowledge production since they always in- no more than a linguistic device used to organize volve some suppositions or, as Gadamer (1994/ a complex empirical world. . . . the purpose of a 1960) put it, prejudices about the subject matter theoretical statement is twofold: to organize (par- in question. For instance, leadership studies simoniously) and to communicate (clearly) (1989: presuppose a set of assumptions that enable us 496). to conceptualize “leadership” as something in the first place, such as trait theory, emphasizing Except for Bacharach’s broad and open defi- person-bound, stable qualities. Without such an nition of theory, what is particularly close to our initial understanding of leadership, we would own view is his notion that theories are not have no idea what to look for, how to design our free-floating statements but are always based study, what empirical material to collect, and on and bounded by researchers’ assumptions how to analyze and theorize leadership. The fo- about the subject matter in question. As Bach- arach notes, the boundary set of assumptions is

254 Academy of Management Review April cal point in problematization as a methodology lated, and challenged. Below we develop a ty- for generating research questions is therefore to pology of assumptions that specifies what types illuminate and challenge those assumptions un- of assumptions are available for problematiza- derlying existing theories about a specific sub- tion when generating research questions, fol- ject matter. lowed by an elaboration of a set of principles for how assumptions can be identified and prob- In order to develop problematization as a lematized. methodology for generating research questions, two key questions need to be answered regard- A Typology of Assumptions Open for ing assumptions. First, what types of assump- Problematization tions are relevant to consider? Second, how can these assumptions be identified, articulated, While there is a range of different assump- and challenged in a way that is likely to lead to tions within the scientific field, we find it pro- the development of an interesting theory? ductive to distinguish five broad sets of assump- Highly relevant here is the growing body of tions that differ in both depth and scope. These work that has focused on “interestingness” in are in-house, root metaphor, paradigm, ideol- theory development. Although many theorists ogy, and field assumptions. This categorization (e.g., Astley, 1985; Bartunek et al., 2006) have is partly inspired by Morgan’s (1980) differentia- described how a theory can be made more in- tion between puzzle solving, root metaphors, teresting by challenging assumptions, Davis and paradigms. The typology is also influenced (1971) has discussed this most fully, developing by the paradigm debate where some authors an ”index of the interesting.” The index de- claim to have an overview of various world scribes twelve different ways in which an audi- views (paradigms), thereby indicating the sig- ence’s assumptions can be challenged; these nificance of the wider arena held together by are subsumed in two main categories. The first some overall ideas and assumptions (Burrell & category (characterization of a single phenome- Morgan, 1979). An interest in ideology assump- non) includes those cases in which we assume tions proceeds from the observation that re- that a phenomenon is constituted in a particular searchers’ engagement in scientific fields like way, but in reality it is not, or vice versa; for management is in no way neutral regarding hu- example, a phenomenon that many assume to man interests and political positioning (Haber- be disorganized is, in fact, organized. The sec- mas, 1972). The notion of field assumption is ond category (relations among multiple phe- inspired by scholars who take a broader view of nomena) includes those instances in which we an academic area (e.g., Bourdieu, 1979; Foucault, assume that there is a particular relation be- 1972). tween multiple phenomena when there is not, or vice versa; for instance, phenomena that we as- In-house assumptions exist within a particular sume to be correlated are, in reality, uncorre- school of thought in the sense that they are lated. shared and accepted as unproblematic by its advocates. In-house assumptions differ from While Davis’s index provides a comprehen- puzzle solving in that they refer to a set of ideas sive account of ways in which a theory can chal- held by a theoretical school about a specific lenge an audience’s assumptions, the index subject matter, whereas puzzle solving refers to does not specify what types of assumptions can the particular way of conducting research stip- be problematized. It provides only a general ulated by that school. An example of in-house definition of assumption in the form of “what assumptions are trait theories within the ratio- seems to be X is in reality non-X, or what is nalistic school, which typically conceptualizes accepted as X is actually non-X” (Davis, 1971: leadership as a set of specific attributes, such as 313). In particular, such a general definition formal knowledge, skills, attitudes, and per- does not address how assumptions differ in both sonal traits possessed by the individual leader depth (Abbott, 2004; Schein, 1985) and scope (Yukl, 2006). If we were to question the trait the- (Gouldner, 1970), which are essential to under- ory assumption that leadership is defined less stand when constructing research questions by the trait of the leader than by the social through problematization. Nor does the index context, we would challenge an in-house as- provide any specific principles for how different sumption of leadership. types of assumptions can be identified, articu-

2011 Alvesson and Sandberg 255 Root metaphor assumptions are associated thought within a paradigm, and sometimes even with broader images of a particular subject mat- across paradigms and disciplines. Simon’s ter (Morgan, 1980, 1997). Within management (1947) work on bounded rationality can perhaps studies, for example, it is common to see orga- be seen as a mild but successful identification nizations as “cultures” in terms of a unitary set and challenge of a field assumption. His chal- of values and beliefs shared by organization lenge of the widely shared assumption that hu- members. However, at the root metaphor level mans are rational decision makers, and the al- (Smircich, 1983), authors have questioned as- ternative assumption of bounded rationality, sumptions around unity, uniqueness, and consen- opened up a range of new and interesting re- sus, and they have emphasized differentiation, search questions and theories. Field assump- fragmentation, discontinuity, and ambiguity as tions may also unite antagonistic schools, key elements in culture (e.g., Martin, 2002; Martin which, at one level, often present as different & Meyerson, 1988). and even oppositional but, at a deeper level, share a set of assumptions about their particu- The ontological, epistemological, and meth- lar field (cf. Bourdieu, 1979). For example, labor odological assumptions that underlie a specific process theorists and poststructural-oriented literature can be characterized as paradigmatic critical management scholars agree that there assumptions (cf. Burrell & Morgan, 1979; Kuhn, is something called “management” and an ide- 1970). The challenge of such assumptions is of- ology or discourse of managerialism, which ten a central ingredient for generating interest- should be critically addressed. However, in de- ing research questions. For example, by adopt- bates each of these schools of thought claims to ing an interpretive perspective on professional have privileged access to an insightful under- competence, Sandberg (2000) challenged the du- standing of management. alist ontology underlying the prevalent ratio- nalistic school, which conceptualizes profes- Taken together, the typology can be seen as a sional competence as consisting of two separate continuum of overlapping assumptions open for entities: a set of attributes possessed by the problematization, where in-house assumptions worker and a separate set of work activities. form one end and field assumptions the other However, from an interpretive approach, compe- end of the continuum. Challenging in-house as- tence does not consist of two separate entities; sumptions can be seen as a minor form of prob- instead, person and work form an inseparable lematization; questioning root metaphor as- relation through the lived experience of work. sumptions as a more middle-range form; and Such a questioning enabled Sandberg to pro- challenging paradigm, ideology, and field as- vide an alternative assumption ground and, sumptions as a broader and more fundamental based on that, to generate new research ques- form of problematization. It may seem that chal- tions about professional competence. lenging any of the three latter types of assump- tions is most likely to generate research ques- Ideology assumptions include various politi- tions that may lead to the development of more cal-, moral-, and gender-related assumptions interesting and influential theories. However, a held about the subject matter. Burawoy (1979), challenge of these broader assumptions may for example, suggested that researchers con- also be superficial, since it is difficult to achieve ducting studies of work should not proceed from depth when addressing broad intellectual ter- the question “Why don’t workers work harder?” rains. An insightful challenge of an in-house or and then investigate norms about a reasonable a root metaphor assumption can be a key part in work performance; instead, they should ask, the process of developing new theory. “Why do people work as hard as they do?” In a similar vein, Sievers (1986) challenged existing Methodological Principles for Identifying, theories of motivation by suggesting that in- Articulating, and Challenging Assumptions stead of asking how people can be motivated in organizations, they should ask why people need As described above, a key task in generating to be motivated at all if they experience their research questions through problematization is jobs as meaningful. to enter a dialectical interrogation between one’s own and other metatheoretical stances so Field assumptions are a broader set of as- as to identify, articulate, and challenge central sumptions about a specific subject matter that are shared by several different schools of

256 Academy of Management Review April assumptions underlying existing literature in a and rereadings. Identifying or constructing a do- way that opens up new areas of inquiry. To be main of literature provides the entrance to pick- able to problematize assumptions through such ing some texts, but careful reading of these may an interrogation, the following methodological inspire the revision of the literature domain that principles are central: (1) identifying a domain finally will be the research question target. One of literature, (2) identifying and articulating as- possibility is to focus on an exemplar—that is, a sumptions underlying this domain, (3) evaluat- path-defining study (Abbott, 2001; Kuhn, 1970)— ing them, (4) developing an alternative assump- that plays a key role in a literature domain. tion ground, (5) considering it in relation to its Given the significance of path-defining studies, audience, and (6) evaluating the alternative as- such a focus may be productive, although, of sumption ground. While we, for the sake of clar- course, later work drawing on the path-defining ity, present the principles in a sequential order, study needs to be identified and reviewed in the actual problematization process is consider- order to investigate whether all the assumptions ably more iterative than linear in character. that one finds potentially interesting to chal- Moreover, these principles should not be treated lenge are still in operation. Another option is to as a list of fixed ingredients in a recipe but, concentrate on one summary or a few authorita- rather, as important elements to consider in the tive summaries, given that they are not covering problematization process. As Deacon (2000) too much (which may mean that the clues to notes, problematization cannot be reduced to a assumptions are too vague). A third option is to mechanical or even strictly analytical proce- look at a few more recent, influential, and re- dure, since it always involves some kind of cre- spected pieces, covering some variation in a ative act. “It is a creation in the sense that, given particular domain of literature. Although these a certain situation, one cannot infer that pre- options need to be supplemented with broader cisely this kind of problematization will follow” readings, the in-depth reading of the selected (2000: 135). texts is the focal point for the problematizer. 1. Identifying a domain of literature for as- 2. Identifying and articulating assumptions sumption-challenging investigations. It is usu- underlying the chosen domain of literature. As- ally not obvious how to sort and delimit existing sumptions underlying a specific domain of lit- studies into a specific domain of literature and erature are rarely formulated as McGregorian then relate that literature to one’s own study theory X versus theory Y alternatives. Such (Locke & Golden-Biddle, 1997). This is the case explicitly formulated assumptions have more irrespective of whether one is using gap- the character of “postulations.” As Gouldner spotting or problematization. However, com- notes, postulations “contain a second set of pared to gap-spotting research, problematiza- assumptions that are unpostulated and unla- tion efforts are less concerned with covering all beled . . . because they provide the back- possible studies within a field than uncritically ground out of which the postulations in part reproducing the assumptions informing these emerge and . . . not being expressively formu- studies. Problematization research typically in- lated, they remain in the background of the volves a more narrow literature coverage and theorist’s attention” (1970: 29). It is the assump- in-depth readings of key texts, with the specific tions that mostly remain implicit or weakly aim of identifying and challenging the assump- articulated that are the main target in the tions underlying the specific literature domain problematization methodology. A key issue targeted. In this sense, the prevailing norm to here is to transform what are commonly seen relate one’s own study to all the relevant litera- as truths or facts into assumptions. ture works against problematization and needs to be resisted. However, it is important to make Drawing on the assumption typology outlined broad references to major or typical studies and above, we see a range of methodological tactics to scrutinize possible problematization in rele- available for identifying assumptions in exist- vant work. ing literature. In-house assumptions can be identified by scrutinizing internal debates and Two interrelated issues are important to con- the interfaces between a specific group of au- sider when identifying a domain of literature for thors who frequently refer to each other and problematization: the actual domain targeted neighboring areas, moderately relating one’s and the specific texts chosen for deep readings work to the focused group’s work, and mainly

2011 Alvesson and Sandberg 257 using a similar narrative style and vocabulary. titude of interests and values and the contradic- For example, various authors have challenged tions and dilemmas between these could also the idea that organizations typically form uni- be beneficial. The contradiction between values tary and unique cultures (e.g., Van Maanen & like autonomy and leadership or managerial Barley, 1984), or even clear and stable subcul- work as hierarchical control versus democratic tures (Martin & Meyerson, 1988), by seeing cul- accountability could exemplify this (Alvesson & ture as a process rather than as something sta- Willmott, 1996). ble (Alvesson, 2002). Field assumptions are difficult to identify be- Root metaphor assumptions can be explored cause “everyone” shares them, and, thus, they by (1) identifying the basic image or metaphor of are rarely thematized in research texts. One op- social reality informing a text or school and (2) tion is to search across theoretical schools and detecting or producing alternative possible con- intellectual camps to see whether they have frontational metaphors. Morgan’s (1997) Images anything in common regarding the conceptual- of Organization provides one well-known illus- ization of the particular subject matter in ques- tration of how metaphors can be used to become tion. Another option is to look at debates and aware of alternative conceptualizations and, critiques between seemingly very different po- thus, how they can inspire one to articulate sitions and focus on what they are not address- one’s own assumptions. Alvesson (1993) picks up ing—that is, the common consensual ground not this line, arguing that it is possible to carve out being debated. Looking at other fields may also assumptions by looking at the metaphors be- be valuable in getting some perspective. This is hind the metaphors used (i.e., second-level met- to some extent illustrated in this article, since aphors). For example, behind the metaphor that we identify and challenge gap-spotting as a conceptualizes organization as a political field assumption for how to generate research arena, one could imagine different views of this questions within management studies (in this arena, one being a parliamentary democracy regard, we acknowledge help from Davis [1971], (with rules of the game) and another being more a scholar outside our field). like a jungle, where the political battles are less democratic and rule bound. Although focusing on a specific type of as- sumption may be fruitful, it is often better to Identification of paradigm assumptions nor- vary one’s focus and, at least initially, consider mally calls for some familiarity with an alterna- what in-house, metaphor, paradigm, ideology, tive world view, without being stuck in the lat- and field assumptions underlie a particular do- ter. Some existing efforts to map and confront main of existing literature. It is also important to paradigms may be helpful (e.g., Astley & Van de focus on assumptions that may exist at different Ven, 1983; Burrell & Morgan, 1979; Deetz, 1996; theoretical levels within a targeted study. This Donaldson, 1985; Pfeffer, 1982). Although reading is because challenging an in-house assumption about paradigm debates can be useful, the chal- related to a broader theoretical perspective (e.g., lenge is not to be caught up in them or by the functionalist perspective, etc.) within the tar- positions expressed in those debates. Instead, geted study may facilitate the formulation of they should be used as important heuristic tools more interesting research questions than chal- to loosen up others’ as well as our own views lenging an in-house assumption underlying a (Abbott, 2004: 86). specific theory (e.g., trait theory, etc.) within the study targeted. It should also be borne in mind Ideological assumptions can also be explored that assumptions are not fixed but are, to some by being aware of positions very different from extent, an outcome of how one constructs the the focal one in terms of interests, focus, identi- nature and scope of the domain of literature fications, values, and ethical commitments. One targeted, and this can be narrowed or broad- tactic would be to read and interpret an exam- ened and can be interpreted in different ways. ple of what appears to be positive and worth Hence, the combination of hermeneutical in- taking seriously as a problem to be addressed or depth readings, creative efforts, some boldness, as a solution to be embraced. Another tactic patience, self-critique, support from theoretical would be to view something negative (e.g., re- stances other than one’s own, and sometimes pressive) as perhaps innocent or even positive even luck is important in order to identify and (e.g., laissez-faire leadership as a source of au- articulate assumptions. tonomy). Working with the recognition of a mul-

258 Academy of Management Review April 3. Evaluating articulated assumptions. Hav- lenged assumptions—a source of application ing identified and articulated assumptions rather than drivers for rethinking. Problemati- within the chosen literature domain, the prob- zating such assumptions may then be neces- lematizer needs to assess them. Certainly not all sary, either through informed defenses of the assumptions are worthy of being problematized problematized positions (e.g., Donaldson, 1985) and brought forward as significant research or through new or synthesized approaches like contributions— or as key steps in such an enter- skeptical partial affirmation (e.g., Newton, 1998). prise. The problematizer must therefore contin- ually ask him/herself, “What is the theoretical 4. Developing an alternative assumption potential of challenging a particular assump- ground. While the formulation of alternative as- tion?” As a general rule, challenging broader sumptions analytically marks a crucial “stage” assumptions, such as paradigm or field assump- in problematization, it should not be seen as tions, is likely to lead to greater impact theories, isolated from the other principles involved. The but these assumptions are often more difficult to (re)formulation part extends the earlier parts of identify and challenge successfully. the process: identifying assumptions calls for at least an intuitive idea of alternative assump- An overall but vague consideration for an tions, and success in the former means that the identified assumption to be problematized latter is likely to come through more clearly. should be that it does not contribute signifi- cantly to a “good” understanding of the subject Similar to identifying and articulating exist- matter but is still broadly shared within a re- ing assumptions, it can be useful to consult search area. “Truth” in any of the several avail- available critical and reflexive literature, repre- able senses is also an important criterion to sentatives of competing schools, and various consider—that is, an assumption that is seen as forms of heuristic tools, such as those offered by “untrue” is then targeted. Empirical evidence Abbott (2004: 110 –210), in developing new as- indicating that some assumptions are problem- sumptions. As emphasized above, a challenge atic is important here, even though assumptions of existing assumptions should include some seldom can be directly empirically investigated independence from these and should move be- or tested (Astley, 1985; Kuhn, 1970). yond already available counterassumptions. It may, for example, be tempting to use an inter- Something true can also be trivial, and a pretive stance against functionalist assump- strong insistence on proving that something is tions, or to replace interpretive humanism with true (where a hypothesis should be verified) can poststructuralism, but the purpose of this ap- be constraining (Becker, 1998: 20-24; Starbuck, proach is to avoid such moves. Producing new 2006: 99 –101). Theoretical fruitfulness, novelty, and good research questions means that there and provocative capacity can be equally impor- are no predefined answers available; new ques- tant to bear in mind—and are typically what tions offer starting points for new answers. Such makes a theory interesting (Astley, 1985). A a problematization is facilitated by temporarily closely related criterion is to what extent a chal- applying the dialectical interrogation between lenge of the identified assumptions can inspire different theoretical stances and the domain of new areas of research and research programs. literature targeted. The idea is to be inspired by The articulated assumptions may also be as- various theoretical stances and their resources sessed in terms of how they form the basis for and to use them creatively in order to come up other established knowledge areas or a domi- with something unexpected and novel. nant line of thinking that tends to produce main- stream effects (e.g., close alternatives). 5. Considering assumptions in relation to the audience. Assumptions to be targeted for chal- “Timing” is another consideration. An as- lenge must be considered in relation to the sumption may be productive and inspiring at a groups who hold them and the general intellec- specific time but may gradually become part of tual, social, and political situation of a research conventional wisdom and lose its power to gen- community. It is a complex issue because the erate new knowledge. Many critical perspec- ”audience” typically is not a unitary group— tives (poststructuralism, critical management primarily because there are often not one but studies, feminism, etc.) may, for example, be multiple audiences, and the assumptions held able to inspire problematization for some time by one audience may differ from the assump- but may later establish a new set of unchal- tions held by another audience. It is also likely

2011 Alvesson and Sandberg 259 that one particular audience consists of several the subject matter—it will be regarded as obvi- subgroups, which makes it even harder to spec- ous by many. ify the potentially relevant audiences. For in- stance, within a specific area, such as strategy It’s absurd! If, however, the alternative as- or leadership, there is an ambiguous mass of sumption ground denies all the assumptions overlapping groups, which are difficult to sepa- held by the targeted audiences, it is likely that it rate into clear segments. Layperson audiences will be regarded as unbelievable. Both of the may be even harder to identify and delimit since above responses indicate that the alternative they are usually not as well documented as ac- assumption ground is likely to be unsuccessful. ademic audiences. One option could be to re- view more popular business magazines that That’s interesting! This is the ideal response. practitioners read and perhaps also write for. According to Davis and other advocates of ”in- Apart from literature reviews, it is also impor- teresting theories” (e.g., Bartunek et al., 2006; tant to talk and listen to both academics and McKinley et al., 1999; Weick, 1989), the experi- practitioners in order to understand their views ence of ”this is interesting” occurs when the of the particular subject matter in question and alternative assumption ground accepts some the assumptions they hold about it. Sometimes and denies some of the assumptions held by the this leads to revisions of the literature domain targeted audiences. Because they are curious one started with. and willing to listen, the audiences may take the new idea or challenge seriously. Hence, the lit- It is important as well to recognize the politics mus test for being considered interesting is that involved when choosing the assumptions to be the alternative assumption ground should fall challenged. It is not only a matter of advancing somewhere between what is regarded as obvi- science but of understanding research politics— ous and absurd. who will lose or win when a specific assumption is challenged? Similarly, what type of challenge One could add to the intellectual response can an audience accept cognitively and emo- revolving around novelty, surprise, and excite- tionally? In other words, how can assumptions ment (Abbott, 2004) that it is important to con- be challenged without upsetting dominant sider the perceived fruitfulness or relevance of groups, which hold them so strongly that they the new research question for developing new ignore the critique or even prevent one’s study research programs and for contributing new from being published? Here problematization of knowledge having social relevance (Van de Ven, in-house and root metaphor assumptions prob- 2007). It is also important to consider its rhetor- ably will often be received more positively (less ical appeal (Golden-Biddle & Locke, 2007). A defensively) than problematization of ideology, commonly used rhetorical strategy is politeness paradigm, or field assumptions. (Locke & Golden-Biddle, 1997; Myers, 1993). For instance, all the authors in the texts investi- 6. Evaluating the alternative assumption gated by Locke and Golden-Biddle (1997) used ground. Following the body of work focusing on various politeness strategies (such as acknowl- interestingness in theory development (e.g., Bar- edging other researchers for their contribution tunek et al., 2006; Davis, 1971; McKinley et al., to the field) to reduce the risk of upsetting the 1999), the ultimate indicator of whether a prob- academics they were criticizing. Similarly, the lematization is going to be successful is not so aesthetic dimensions of the alternative assump- much rigor and empirical support—although tion ground are also central in composing an these qualities are part of the picture (since appealing and convincing argument (Astley, credibility is always important)—as it is the ex- 1985). For instance, to achieve the response of perience of “this is interesting.” Davis (1971) sug- “that’s interesting,” it is important to work with gests three responses that can be used to eval- metaphors that are appealing and concepts and uate to what extent an alternative assumption formulations that are challenging and provoca- ground is likely to generate a theory that will be tive. Examples could be March and Olsen’s regarded as interesting. (1976) garbage can model of decision making and Brunsson’s (2003) idea of organized hypoc- That’s obvious! If the set of alternative as- risy. It is important as well to test the alternative sumptions to a large extent confirms the as- assumption ground on various representatives sumptions held by the targeted audiences— from the targeted audiences. How do they react? what they already assume to be the case about

260 Academy of Management Review April The outlined problematization methodology is offers a good opportunity for in-depth explora- summarized in Figure 1 and further elaborated tion of assumptions, it can also lead to limited in the next section by applying it to the litera- results. Therefore, in order to accomplish a ture domain of identity constructions in organi- broader relevance, we also consider a few other zations. Again, while the actual problematiza- influential studies in the domain with a some- tion process is considerably more organic, for what different approach (i.e., Ashforth & Mael, illustrative purposes we follow the six prob- 1989; Gioia, Schulz, & Corley, 2000; Pratt, 2000; lematization principles outlined above sequen- Pratt & Foreman, 2000). There is also a wealth of tially. other studies that, to various degrees, are rele- vant in problematizing Dutton’s et al.’s text (e.g., AN ILLUSTRATION OF THE Alvesson, Ashcraft, & Thomas, 2008; Brown, 2006; PROBLEMATIZATION METHODOLOGY Collinson, 2003; Deetz, 1992; Elsbach, 1999; Fou- cault, 1977, 1980; Haslam, 2004; Jenkins, 2000; 1. Identifying a Domain of Literature for Knights & Willmott, 1989; Shotter & Gergen, 1989; Assumption-Challenging Investigations Weedon, 1987). However, in order to focus on the elements in the problematization methodology, In order to illustrate our problematization with the exception of a few occasions, we avoid methodology, we choose to focus primarily on looking into how others have raised points of Dutton et al.’s (1994) path-setting study, “Orga- relevance for discussing the various issues that nizational Images and Member Identification,” we address in our problematization of Dutton within the domain of identity constructions in et al.’s text below. organizations. Although focusing on a key text FIGURE 1 The Problematization Methodology and Its Key Elements Aim of the problematization methodology Generating novel research questions through a dialectical interrogation of one’s own familiar position, other stances, and the literature domain targeted for assumption challenging A typology of assumptions open for problematization In-house: Root metaphor: Paradigm: Ideology: Field: Assumptions that Broader images of Ontological, Political-, moral-, Assumptions exist within a a particular subject epistemological, and gender- about a specific school of matter underlying and related specific thought subject matter existing literature methodological assumptions that are assumptions underlying shared across underlying existing literature different existing literature theoretical schools Principles for identifying and challenging assumptions 1. Identify a 2. Identify and 3. Evaluate 4. Develop 5. Relate 6. Evaluate domain of articulate articulated alternative assumptions alternative literature: assumptions: assumptions: assumptions: to audience: assumptions: What main What major Are the What What major Are the bodies of assumptions identified alternative audiences alternative literature underlie the assumptions assumptions hold the assumptions and key literature worthy to be can be challenged likely to texts make within the challenged? developed? assumptions? generate a up the identified theory that domain? domain? will be regarded as interesting by the audiences targeted?

2011 Alvesson and Sandberg 261 The particular subject matter in Dutton et al.’s ple’s sense of membership in an organization study is how individuals are attached to social shapes their self-concept, very few assumptions groups, which they conceptualize as “member on which they base their argument are high- identification.” They explain it as follows: lighted in this way. Instead, the text creates the impression that its argument and logic are Members vary in how much they identify with grounded in specific factors reflecting self- their work organization. When they identify evident truths. For example, the authors claim strongly with the organization, the attributes they that a perceived organizational identity exists use to define the organization also define them. in the sense of a member’s having beliefs about Organizations affect their members through this the distinctive, central, and enduring attributes identification process, as shown by the comments of the organization (reflecting Albert and of a 3M salesman, quoted in Garbett (1988: 2): “I Whetten’s [1985] definition), and that an organi- found out today that it is a lot easer being a zational member sometimes defines him/herself salesman for 3M than for a little jobber no one by the same attributes that he or she believes has ever heard of. When you don’t have to waste define the organization. But these statements time justifying your existence or explaining why contain assumptions that conceptualize their you are here, it gives you a certain amount of subject matter of how individuals are attached self-assurance. And I discovered I came across to organizations in a particular way and are not warmer and friendlier. It made me feel good and necessarily correct or productive. enthusiastic to be ‘somebody for a change.’” This salesman attributes his new, more positive sense Let us first consider the statement “a mem- of self to his membership in 3M, a well-known ber’s beliefs about the distinctive, central, and company. What he thinks about his organization enduring attributes of the organization” (1994: and what he suspects others think about his or- 239). One of its assumptions is that people see ganization affects the way that he thinks about themselves as members of an organization, as if himself as a salesperson (Dutton et al., 1994: 239). the latter is like a club or an association, which people join as a positive choice. Another is that Dutton et al. try to understand member iden- members have (1) beliefs (2) about attributes of tification by investigating how “a member’s cog- the organization and (3) that these attributes are nitive connection with his or her work organiza- distinctive, central, and enduring. Similarly, the tion . . . [derives] from images that each member statement “the degree to which a member de- has of the organization” (1994: 239). The first im- fines him- or herself by the same attributes that age (what the member believes is distinctive, he or she believes define the organization” (1994: central, and enduring about the organization) is 239) is also underpinned by a range of assump- defined as perceived organizational identity. tions. One is that individuals and organizations The second image (what the member believes are constituted by a set of inherent and more or outsiders think about the organization) is called less stable attributes. Another is that the attri- “the construed external image” (1994: 239). Dut- butes of the individual are comparable with the ton et al. develop a model of member identifica- attributes of the organization through a mem- tion that suggests that the two organizational ber’s cognitive connection. Based on those as- images “influence the cognitive connection that sumptions, Dutton et al. conceptualize person members create with their organization and the and organization as externally related to each kind of behaviors that follow” (1994: 239). Their other through an individual’s images of his or model proposes that “members assess the at- her organization and what outsiders think about tractiveness of these images by how well the the organization. This reasoning carries a range image preserves the continuity of their self- of paradigmatic assumptions, such as the dual- concept, provides distinctiveness, and enhances ist ontological assumption that a person and the self-esteem” (1994: 239). Based on the model, they world exist independently of each other (Sand- develop a range of propositions about organiza- berg & Targama, 2007: Chapter 2). tional identification. These can be tested, but we here look at the assumptions behind the propo- Let us briefly compare the Dutton et al. text sitions. with the other selected texts in the domain. Pratt, drawing heavily on Dutton et al., investi- 2. Identifying and Articulating Assumptions gated “how organizations attempt, succeed, and Underlying the Chosen Domain of Literature fail to change how members view themselves in Although Dutton et al. point out explicitly that a central assumption of their study is that peo-

262 Academy of Management Review April relation to the organization” (2000: 457). His work 3. Evaluating Articulated Assumptions departs from the emphasis in the literature that “most research [should] focus on how organiza- The assumptions identified above (on mem- tions successfully engender strong ties with bership, fixed perceptions of the individual and members” and instead should “look at organi- the organization as a thing-like phenomenon, zational conditions that lead to positive, nega- and a perceived similarity between individual tive, ambivalent and broken identifications” and organizational attributes) need to be as- (2000: 457), and at how identification manage- sessed to determine if, and to what extent, they ment is “associated with a variety of identifica- are worthy of further problematizations. For ex- tion types” (2000: 458). ample, the assumption that people regard them- selves as members of their work organizations While sharing similar assumptions as Dut- can be challenged with the more instrumental ton et al., Pratt adds to the literature by point- and often darker aspects of employment. One ing out that the individual can change identi- can thus question Dutton et al.’s ideological as- fication states. His claim resonates to some sumption of an “organizational man” view of a extent with Ashforth’s claim that “identity is positive and strong link between an employer perpetual work in progress” (1998: 213), further and a compliant employee with a limited inde- underscored by Ashforth and Mael’s observa- pendent self, using the employment situation as tion of “the often unique and context-specific a natural and significant source of identity. demands of an identity” (1989: 147). In a similar Pratt’s (2000) work opens this up to some extent vein, Gioia et al. argue that the “apparent by pointing out less positive identifications, but durability of identity is somewhat illusory” it still adheres to the assumption that “mem- (2000: 64), because it is mainly a matter of “the bers view themselves in relation to the orga- stability used by organization members to ex- nization” and that issues around identity “can press what they believe the organization to and should be managed” (Pratt & Foreman, be” (2000: 64). Hence, while still sharing Dutton 2000: 18). et al.’s assumptions that organizational mem- ber identification is a “distinctive and endur- The assumption that members have (1) beliefs ing characteristic” (Ashforth & Mael, 1989: 154), (2) about attributes of the organization and (3) the above authors express a more dynamic that these attributes are distinctive, central, and and less organization-focused view of organi- enduring can also be further questioned. Are zational identification. people’s ways of relating to organizations typi- cally so thing-like? Using an alternative meta- The assumptions held by Dutton et al. (and to phor, the organization can perhaps be seen as a a significant degree also by Ashforth & Mael, broad and complex terrain where perceptions Gioia et al., and Pratt) can be further elabo- and sentiments are shifting, depending on as- rated and articulated with the help of the as- pects, moments, and contexts. For example, “or- sumptions typology. For example, their as- ganization” may sometimes refer to colleagues sumption that members may have beliefs or to top management; at other times to one’s about the specific attributes of the organiza- own department or work or one’s future career tion can be regarded as an in-house assump- prospects, rewards, and fringe benefits; and, tion among these authors. The assumption on other occasions, to mass medial represen- that individuals are carriers of beliefs can tations, products, and HR policies. As Ashcraft also be targeted at a paradigmatic level. The and Alvesson (2009) show, people construct “natural” and potentially harmonious rela- and relate to a seemingly straightforward ob- tionship between individuals and the (human- ject like “management” in highly shifting and like) organization indicated by the overlap of varied ways. As an identification target, “the characteristics can be further explored in organization” may be best conceived as mul- terms of ideology. The very idea that there is tiple and moving. This is also to some extent something— constructed or not—such as “or- pointed out by Gioia et al. (2000) and Pratt and ganizational identity” or “individual identity” Foreman (2000), but these authors still assume and that they are worthy of investigation may the existence of beliefs about the organization indicate some field-level assumptions. as a whole (and its central, distinct, and endur- ing characteristics), while a counterassumption

2011 Alvesson and Sandberg 263 could be that such an entity is not what most organizations as multidimensional, shape shift- people primarily relate to. ing, and discursively constituted—a domain ex- hibiting multiple and varied social identities The assumption that individuals and organi- (Chia, 2000). This assumption is different from zations hold similar attributes and generate a positions mainly pointing out changes over time “fit” appears to be as problematic and can be (as expressed, for example, by Gioia et al., 2000, further questioned. The possible connection and Pratt, 2000). may be considerably more frictional, volatile, and fluid. Ideas of varied identification types The above problematizations, associated with (Pratt, 2000), pluralistic beliefs about organiza- (two versions of) critical theory, economic man tional identity (Pratt & Foreman, 2000), and iden- thinking, and radical process thinking, offer ref- tity changes reflecting image changes (Gioia et erence points for alternative assumptions. We al., 2000) are also relevant to consider here, since selectively use all in order to develop novel re- they give some clues about what assumptions search questions. As emphasized, problematiza- are worthwhile to problematize further. tion is best accomplished through using (but not directly applying) a broad set of theoretical 4. Developing an Alternative Assumption stances, offering resources for unpacking and Ground rethinking. We now arrive at the task of developing as- The assumption that postulates a stable and sumptions counter, or at least alternative, to the robust degree of perceived similarity between ones identified and articulated through the individual and organization could be related to problematization above. Similar to the identifi- ideas on variation, process, and dynamics cation and articulation of the above assump- around self-definition and construction of the tions, we can here draw on different theoretical organization. The possible meeting points— positions to play up reference points and re- spaces for establishing a possible “perceived sources for problematization. One possible similarity”—may be rare, since most parts of stance is critical theory, which provides at least people’s working lives may go on without them two alternative assumptions. One proposes that comparing themselves to the employing organi- the organizational membership assumption is a zation at a more abstract and holistic level. Still, naive idealization of contemporary work experi- these meeting points may be important. Rather ences in flexible capitalism, strongly downplay- than seeing the similarity between individual ing lasting relationships and commitment (Sen- and organization as static (or only gradually nett, 1998) and thereby making organizational dynamic, as Pratt and Gioia et al. do), one can identification a rare or fragile phenomenon— regard organization and individual as differ- perhaps a managerial dream rather than some- ent traffic of stories (of self and organization), thing existing on a broader scale. Another and and sometimes these stories may converge— quite different critical theory assumption is that that is, organizational identification temporar- the possibility of strong identification with the ily occurs. organization may mean people become cultural dopes and lose a clear sense of independence in One possibility here could be that employees relation to the employer, who wins the minds articulate a positive link between themselves and hearts of employees (Kunda, 1992; Willmott, and their organizations when the context im- 1993). plies certain advantages but not when it implies disadvantages. Identification is, thus, self- A quite different route would be to proceed interest driven, a discursive act and typically from the economic man assumption about ra- temporal and situation specific, sometimes op- tional maximization of self-interest (Camerer & portunistic. The citation of the 3M employee by Fehr, 2006; Henrich et al., 2005), leading to a view Dutton et al. above illustrates this. Since it can of identification as a tactical resource for self- be an advantage to be a representative of a promotion. A third alternative would be to be large and well-known firm in a certain sale sit- influenced by a poststructuralist stance, in uation, making presentation easier, a positive which the assumption of the organization as a link between individual and organization is em- fixed and one-dimensional object can be chal- phasized in that situation. Whether the same lenged by a hyperprocess or fluidity view of positive link—and identification—is expressed when corporate bureaucracy or hierarchy (often

264 Academy of Management Review April mentioned as negative aspects of very large poral, fragile, and possibly rare position rather firms), or the possible harsh performance pres- than a fixed trait? sure from management, provides the context is perhaps more doubtful. Possible identifications 5. Considering Assumptions in Relation to the may therefore be more area specific and dy- Audience namic, existing in a space that also includes salient moments of alienation or opportunism. The four previous principles indicate reasons Research questions on the perceived unity or to reconsider some of the assumptions underly- multicontextuality of an organization (if that cat- ing not only Dutton et al.’s approach but also egory is relevant for people) and how individu- broader parts of the organizational identity and als may couple/decouple themselves at various identification domain. A key assumption in this times and in various domains (settings) may large and expanding literature domain (Haslam then be suggested. & Reicher, 2006) is that most employees define themselves as organizational members, or they Let us sum up alternative assumptions and may, given proper (identity) management, do so. research questions. First, people working in or- This can, of course, motivate various forms of ganizations more commonly see themselves as problematization—from a strong (paradigmatic) employees with varying degrees of experiences one, aiming at undermining the key belief that of organizational membership. An employee’s people define themselves partly or mainly way of defining him/herself may be more or less through belonging to an organization (in terms congruent, nonrelated to or antagonistic to of central, distinctive, and enduring traits), as meanings used to portray and refer to the orga- indicated by the organizational identity and nization. Do people see a similarity between identification industry, to milder ones, suggest- themselves and their organization, and if so, ing revisions through more limited (in-house) how often and when? Perhaps the (rare?) situa- problematizations. tions where statements of self and organization seem to be related can be explored as situation- On the one hand, given the heavy investments specific construction processes, offering sites for and the structuring of organization studies identity work. partly around identity as a key subfield and a key variable, a strong problematization case Second, employees do not necessarily have may be seen as irrelevant (absurd) and become fixed or enduring beliefs only slowly changing marginalized. On the other hand, a radical chal- over time as an effect of radically new circum- lenge of conventional identity research may be stances, as proposed by Gioia et al. (2000) and applauded by various groups that hold more Pratt (2000). Instead, employees take temporary process-sensitive social constructionist assump- positions on their organizational affiliation, tions about identification, although they may such as variation in feelings about membership, not regard it as particularly novel. However, be- being part of an employment contract, and be- ing taken seriously by the majority of manage- ing subordinated to an organizational structure. ment scholars and practitioners probably im- Perhaps situation, event, and process matter plies a less extreme version than that favored by more than static or enduring images about at- poststructuralists, which we think our alterna- tributes? Do people have/express consistent and tive assumption ground expresses. Also, within united or shifting and fragmented beliefs/ the group whose assumptions are challenged, a images about self and organization? One can variety of responses can be expected. Some of here imagine a garbage can–like situation, these will no doubt be political, since research- where the individual and various social identi- ers have vested interests in and identify with ties and identification options (organizational their theories (Bourdieu, 2004; Bresleau, 1997). but also group, occupational, ethnic, gender, and age) plus various subject positions (e.g., 6. Evaluating the Alternative Assumption opportunism, alienation, sense of belonging) are Ground in circulation and sometimes come together in a variety of combinations. Occasionally, a posi- The main task of the sixth problematization tive construction of organizational identity be- principle is to assess to what extent the alterna- comes linked to a positive self-conception tive assumption ground can lead to new re- through identification, but perhaps this is a tem-

2011 Alvesson and Sandberg 265 search questions that have the potential to gen- audiences, we are in a position to leave the erate more interesting identity theories. A first problematization process and begin to formu- step in such an evaluation is to further explore late new research questions. For example, do which major audiences are related to the iden- employees construct/perceive their employing tity field within organization theory and, per- organizations in stable ways? And, if so, when haps, also more broadly in the social sciences. and in what ways, if any, would the personal While it is not possible to do so in this article, a meaning be related to (varieties of) self-identity review of existing literature on identity in orga- of these possible constructions/perceptions? nizations would be central for identifying major One could possibly sharpen this question fur- audience segments, since it would offer mate- ther. Rather than assuming that employees are rial for how to fine-tune the message. Even with- members with clear and, over at least a short out reviewing existing literature in detail, an time period, fixed beliefs about organizational important audience in our example is likely to distinctiveness and endurance, one could pro- be those who broadly share (consciously or un- ceed from the idea that they are (normally) not consciously) the cognitive psychology perspec- best conceptualized as members and could tive on which Dutton et al.’s work is based, to- study if, when, why, and how people construct gether with those favoring a view of the world themselves as members having fixed beliefs made up by perceptions of stable entities. about their employing organizations in relation- ship to themselves. The study of the circulation When the major audiences are known, we are of self and organizational representations/ in a position to use the criteria suggested by identity possibilities and garbage can–like con- Davis: will they regard the alternative assump- nections and disconnections could be an inter- tion ground as absurd, irrelevant, or interesting esting research task. For example, do people and promising? Although the alternative as- move and, if so, how— between identification as sumption ground suggests that individuals’ a positive and a negative source of social identification with organizations is far more identity—and to what extent are such moves weak (or even nonexisting), fluid, and volatile driven by calculative and exploitative motives than assumed by Dutton et al. (and, on the and experience of skeptical distancing (de- whole, by many other influential organizational identification)? identification researchers as well), it does not strongly question the conceptualization of the Studying how employees arrive at and main- subject matter, member identification, as such. tain beliefs that their organizations have traits Nor does the alternative set of assumptions pro- that are distinctive, central, and enduring could vide a deliberate ground attack on the paradig- also be a good research task. Being able to pro- matic assumptions underlying the cognitive duce a coherent set of such beliefs would not be perspective adopted by Dutton et al. It is there- seen as unproblematic and typical but as a true fore possible that the alternative set of assump- accomplishment, facilitated by an ability to tions will be found as potentially interesting by block out the changing, ambiguous, and frag- many of the audiences addressing organization- mented nature of contemporary organizational al identity and identification from a functional- life. Assuming a fluid and nonreified nature of ist view. social reality, organizational identity and self- identity, as well as alignment constructions (“I The extent to which more radical social con- am similar to my organization”), could be structionist audiences will find our alternative viewed as defragmentation and deprocessual- assumptions interesting is questionable, since ization of organizational life, countering the they already embrace some of them. If they were multiple and moving constructions of the targeted, the task would be to avoid the “that’s themes included. Interesting, problematization- obvious” response, perhaps by emphasizing the based research questions would then be as continuation and development of a particular follows. Do people stabilize themes like orga- line of thought (not in itself targeted for prob- nization and self and organizational/self- lematization). For this audience the problemati- identification? What are the (rare) conditions and zation of a quite different set of assumptions operations under which experiences of self and than those of the Dutton et al. text is relevant. organization can be cognitively frozen and sym- bolically merged? Alternatively expressed, If the alternative assumption ground is likely to be regarded as interesting by our targeted

266 Academy of Management Review April when and how do positive stories of self and gaps in existing literature with the aim of filling organization happily merge? The production of them, we think there is considerable room for an organizational identity as a topic and the more increased use of problematization as a method- or less taken-for-granted phenomenon of such ology for constructing novel research questions identification are then placed in a dynamic and that can lead to the development of more inter- fluid context. And the specific construction pro- esting and influential theories within manage- cesses involved are then opened up for inquiry. ment studies. Would the above-generated research ques- The proposed methodology seems particu- tions lead to more interesting and influential larly relevant in situations of political domina- research than a study building positively on tion and cognitive closure that easily follow Dutton et al.? There are no guarantees, but if all from a dominant and established tradition. The the research on this topic is right (e.g., Astley, political situation refers to cases where a social 1985; Bartunek et al., 2006; Black, 2000; Daft et al., interest bias and/or political factors govern 1987; Davis, 1971, 1986, 1999; Hargens, 2000; knowledge production rather than good ideas. Weick, 1989, 2001), one could expect that the re- But also the domination of a particular school of search questions generated through the prob- thought can stifle new ideas and call for politi- lematization of assumptions underlying Dutton cally motivated problematizations. The situa- et al.’s approach are more likely to lead to an tion of cognitive closure is especially salient in interesting theory than the use of a gap-spotting research areas where a particular world view strategy to identify or create a gap in their ap- has colonized the researchers. In such situations proach that needs to be filled. there is often limited critical debate and there are few counterideas because deviant voices When and Why Problematization in are silenced and people have to come up with Generating Research Questions? alternative views. It seems particularly impor- tant to avoid a gap-spotting, extend-the-litera- Given its potential to generate more interest- ture logic here. The benefits of rejuvenating the ing theories, it may be tempting to advocate the field may be high, although the task is not an problematization methodology as the key ingre- easy one. dient in formulating research questions. There are, however, often good reasons to also con- CONCLUSION sider various forms of gap-spotting routes, such as supplementing and enriching other studies This study makes two interrelated contribu- and clarifying issues where there are diverse tions to theory development within the manage- opinions and incoherent results. Sometimes em- ment field. First is the identification and dem- pirical findings play a major role in the formu- onstration of how gap-spotting as the prevalent lation of the purpose of a study, such as in cases way of constructing research questions from ex- when one (re)formulates the research task quite isting literature leads to a shortage of really late in the process (Alvesson & Ka¨ rreman, 2007). interesting and influential studies within man- Combinations of various elements/tactics for se- agement science. In the vocabulary developed lectively building upon and partially problema- in this study, the prevalence of gap-spotting tizing established literature by challenging its across intellectual traditions suggests that it underlying assumptions are probably more pro- constitutes a field assumption within manage- ductive than “purist” approaches. We may also ment studies. It provides researchers with a remind ourselves of the risk of perpetual prob- shared, and to a large extent taken-for-granted, lematization— overproblematization—leading norm for generating research questions from ex- to a sense of fatigue and a deficit of positive isting theory (at least as it is presented in pub- results, as in the case of postmodernism (e.g., lished texts, guiding the actual research contri- deconstruction and partly critical theory). There bution). However, while gap-spotting plays a is a problem if more energy goes into challeng- significant role in developing existing manage- ing assumptions than into working out and re- ment literature, it reinforces rather than chal- fining or testing well-founded and productive lenges the assumptions underlying established ideas. Having said this, given the strong main- theories and, thus, actually reduces the chances stream tradition of identifying or constructing of producing really interesting theories. Our

2011 Alvesson and Sandberg 267 identification and articulation of gap-spotting ing the alternative line of inquiry in a dialogic as a field assumption within management can form to increase the likelihood that readers will therefore be seen as an important contribution respond positively to it. in itself. It offers a strong signal to the field that the grip of gap-spotting as the main way of It is important to emphasize that the proposed constructing research questions needs to be methodology in itself does not guarantee a suc- loosened. At the same time, it encourages re- cessful problematization outcome. A whole searchers to go beyond the logic of gap-spotting range of other factors, such as creativity, imag- and to work with alternative ways of generating ination, reflexivity, range of knowledge mas- research questions that may lead to the devel- tered, and a broad understanding of different opment of more interesting theories. metatheoretical standpoints, is also critical. However, taken together, the methodology pre- Second, and the main contribution of this sented here offers a systematic approach for study, is the proposed problematization method- generating more novel research questions ology, which provides a comprehensive and sys- through problematization of existing literature. tematic addition to gap-spotting and prepack- aged problematization. Instead of providing An important inspiration for this paper was different strategies for identifying or construct- Davis’s (1971) seminal insight that challenging ing gaps in existing literature (and then filling assumptions is what makes a theory interesting, them) or a prepackaged problematization to elaborated in his “index of the interesting.” Our challenge the assumptions of others, this meth- problematization methodology extends and odology enables us—through a dialectical inter- goes beyond Davis’s index in two significant rogation of our own familiar position, other the- ways: (1) compared to Davis’s general definition oretical stances, and the literature domain of assumption (“We thought it was X but it is targeted—to identify, articulate, and challenge really Y”), the typology of assumptions elabo- different types of assumptions underlying exist- rated within the problematization methodology ing literature and, based on that, to formulate provides a more nuanced and enriched specifi- research questions that may facilitate the devel- cation of what types of assumptions are avail- opment of more interesting and influential the- able for problematization, and (2) in contrast to ories. Davis, the methodology offers a set of specific principles for how to identify, articulate, and It does so in two ways. First, it offers specific challenge assumptions underlying existing lit- heuristic support for identifying and challeng- erature and, based on that, to construct interest- ing assumptions in existing literature through ing and novel research questions. its typology, consisting of five broad types of assumptions: in-house, root metaphor, para- More generally, the problematization method- digm, ideology, and field assumptions. Second, ology also contributes to more reflective schol- it provides a set of specific principles for how arship in the sense that it counteracts or supple- assumptions in existing theory can be prob- ments the domination of gap-spotting as a lematized and, based on that, can generate research ideal. As a methodology, it encourages novel research questions: (1) identifying a do- us to produce more novel research questions main of literature for assumption-challenging and theories by actively questioning and criti- investigations; (2) identifying and articulating cally scrutinizing established knowledge in ac- the assumptions (in-house, root metaphor, para- ademia and in society at large. It does so by digm, ideology, and field assumptions) under- offering a distinct alternative to the dominant pinning existing theory as clearly as possible; mode of using the literature in a field for formu- (3) assessing them, pointing at shortcomings, lating research questions. Given the current problems, and oversights; (4) developing new shortage of interesting and influential theories assumptions and formulating research ques- in management studies, the proposed problema- tions; (5) relating the alternative assumption tization methodology seems much needed. ground to an identified audience and assessing the audience’s potential resistance and respon- REFERENCES siveness to it; and (6) evaluating whether the alternative assumptions are likely to generate a Abbott, A. 2001. Chaos of disciplines. Chicago: University of theory that will be seen as interesting and craft- Chicago Press.

268 Academy of Management Review April Abbott, A. 2004. Methods of discovery: Heuristics for the social Bedeian, A. G. 2004. Peer review and the social construction sciences. New York: Norton. of knowledge in the management discipline. Academy of Management Learning & Education, 3: 198 –216. Albert, S., & Whetten, D. A. 1985. Organizational identity: Research in organizational behavior. Greenwich, CT: Black, D. 2000. Dreams of pure sociology. Sociological The- JAI Press. ory, 18: 343–367. Alvesson, M. 1993. Cultural perspectives on organizations. Bluhm, D. J., Harman, W., Lee, T. W., & Mitchell, T. R. 2010. Cambridge: Cambridge University Press. Qualitative research in management: A decade of prog- ress. Journal of Management Studies, accessed at http:// Alvesson, M. 2002. Understanding organizational culture. onlinelibrary.wiley.com/ doi/10.1111/ j.1467-6486.2010. London: Sage. 00972.x/abstract. Alvesson, M., Ashcraft, K., & Thomas, R. 2008. Identity mat- Bourdieu, P. 1979. Outline of a theory of practice. Cambridge: ters: Reflections on the construction of identity scholar- Cambridge University Press. ship in organization studies. Organization, 15: 5–28. Bourdieu, P. 1996. The rules of art: Genesis and structure of Alvesson, M., Hardy, C., & Harley, B. 2008. Reflecting on the literary field. Cambridge: Polity Press. reflexivity: Reappraising practice. Journal of Manage- ment Studies, 45: 480 –501. Bourdieu, P. 2004. Science of science and reflexivity. Chi- cago: University of Chicago Press. Alvesson, M., & Ka¨ rreman, D. 2007. Constructing mystery: Empirical matters in theory development. Academy of Breslau, D. 1997. Contract shop epistemology: Credibility Management Review, 32: 1265–1281. and problem construction in applied social science. Social Studies of Science, 27: 363–394. Alvesson, M., & Sko¨ ldberg, K. 2009. Reflexive methodology (2nd ed.). London: Sage. Brown, A. 2006. A narrative approach to collective identities. Journal of Management Studies, 43: 731–754. Alvesson, M., & Willmott, H. 1996. Making sense of manage- men: A critical introduction. London: Sage. Brunsson, N. 2003. Organized hypocrisy. In B. Czarniawska & G. Sevon (Eds.), The northern lights: Organization theory Ashcraft, K. L., & Alvesson, M. 2009. The moving targets of in Scandinavia: 201–222. Copenhagen: Liber and Copen- dis/identification: Wrestling with the reality of social hagen Business Press. construction. Working paper, University of Colorado, Denver, and Lund University. Burawoy, M. 1979. Manufacturing consent. Chicago: Univer- sity of Chicago Press. Ashforth, B. 1998. Becoming: How does the process of identi- fication unfold? In D. Whetten & C. Godfrey (Eds.), Iden- Burrell, G., & Morgan, G. 1979. Sociological paradigms and tity in organizations: 213–222. Thousand Oaks, CA: Sage. organisational analysis. Aldershot, UK: Gower. Ashforth, B., & Mael, F. 1989. Social identity theory and the Callon, M. 1980. Struggles and negotiations of what is prob- organization. Academy of Management Review, 14: 20 – lematic and what is not: The socio-logics of translation. 39. In K. Knorr, R. Krohn, & R. Whitley (Eds.), The social process of scientific investigation: 197–214. Dordrecht, Astley, W. G. 1985. Administrative science as socially con- Netherlands: Reidel. structed truth. Administrative Science Quarterly, 30: 497– 513. Camerer, C. F., & Fehr, E. 2006. When does “economic man” dominate social behavior? Science, 6: 47–52. Astley, W. G., & Van de Ven, A. 1983. Central perspectives and debates in organization theory. Administrative Sci- Campbell, J. P., Daft, R. L., & Hulin, C. 1982. What to study: ence Quarterly, 28: 245–273. Generating and developing research questions. Beverly Hills, CA: Sage. Bacharach, S. B. 1989. Organizational theories: Some criteria for evaluation. Academy of Management Review, 14: Castels, R. 1994. “Problematization” as a mode of reading 496 –515. history. In J. Goldstein (Ed.), Foucault and the writing of history: 237–252. Oxford: Blackwell. Barrett, M., & Walsham, G. 2004. Making contributions from interpretive case studies: Examining processes of con- Chia, R. 2000. Discourse analysis as organizational analysis. struction and use. In B. Kaplan, D. P. Truex III, D. Was- Organization, 7: 513–518. tell, A. T. Wood-Harper, & J. I. DeGross (Eds.), Information systems research: Relevant theory and informed prac- Clark, T., & Wright, M. 2009. So farewell then . . . Reflections tice: 293–312. Boston: Kluwer Academic. on editing the Journal of Management Studies. Journal of Management Studies, 46: 1–9. Bartunek, J. M., Rynes, S. L., & Ireland, D. R. 2006. What makes management research interesting, and why does it mat- Collinson, D. 2003. Identities and insecurities. Organization, ter? Academy of Management Journal, 49: 9 –15. 10: 527–547. Becker, H. S. 1998. Tricks of the trade: How to think about your Colquitt, J. A., & Zapata-Phelan, C. P. 2007. Trends in theory research while doing it. Chicago: University of Chicago building and theory testing: A five-decade study of the Press. Academy of Management Journal. Academy of Manage- ment Journal, 50: 1261–1303. Bedeian, A. G. 2003. The manuscript review process: The proper roles of authors, referees, and editors. Journal of Cooper, R., & Burrell, G. 1988. Modernism, postmodernism Management Inquiry, 12: 331–338. and organizational analysis: An introduction. Organiza- tion Studies, 9: 91–112.

2011 Alvesson and Sandberg 269 Daft, R. L. 1983. Learning the craft of organizational research. Foucault, M. 1980. Power/knowledge. New York: Pantheon Academy of Management Review, 8: 539 –546. Books. Daft, R. L., Griffin, R. W., & Yates, V. 1987. Retrospective Foucault, M. 1985. The use of pleasure: History of sexuality, accounts of research factors associated with significant vol. 2. New York: Vintage Books. and not-so-significant research outcomes. Academy of Management Journal, 30: 763–785. Freire, P. 1970. Pedagogy of the oppressed. New York: Herder & Herder. Daft, R. L., & Lewin, A. Y. 1990. Can organization studies begin to break out of the normal science straitjacket? An Frost, P. J., & Stablein, R. E. 1992. Doing exemplary research. editorial essay. Organization Science, 1: 1–9. Newbury Park, CA: Sage. Daft, R. L., & Lewin, A. Y. 2008. Rigor and relevance in orga- Gadamer, H.-G. 1994. (First published in 1960.) Truth and nization studies: Idea migration and academic journal method. New York: Continuum. evolution. Organization Science, 19: 177–183. Garbett, T. 1988. How to build a corporation’s identity and Davis, M. S. 1971. That’s interesting! Towards a phenomenol- project its image. Lexington, MA: D.C. Heath. ogy of sociology and a sociology of phenomenology. Philosophy of Social Sciences, 1: 309 –344. Gergen, K. 1992. Organization theory in the postmodern era. In M. Reed & M. Hughes (Eds.), Rethinking organizations: Davis, M. S. 1986. That’s classic! The phenomenology and 207–226. London: Sage. rhetoric of successful social theories. Philosophy of Social Sciences, 16: 285–301. Gioia, D., Schulz, M., & Corley, K. 2000. Organizational iden- tity, image, and adaptive instability. Academy of Man- Davis, M. S. 1999. Aphorism and cliche´ s: The generation and agement Review, 25: 63– 81. dissipation of conceptual charisma. Annual Review of Sociology, 25: 245–269. Golden-Biddle, K., & Locke, K. 2007. Composing qualitative research. Thousand Oaks, CA: Sage. Deacon, R. 2000. Theory as practice: Foucault’s concept of problematization. Telos, 118: 127–139. Gouldner, A. W. 1970. The coming crisis of western sociology. New York: Basic Books. Deetz, S. 1992. Democracy in an age of corporate coloniza- tion: Developments in communication and the politics of Grandy, G., & Mills, A. J. 2004. Strategy as simulacra? A everyday life. Albany: State University of New York radical reflexive look at the discipline and practice of Press. strategy. Journal of Management Studies, 41: 1153–1170. Deetz, S. 1996. Describing differences in approaches to or- Grant, D., Hardy, S., Oswick, C., & Putnam, L. 2004. The Sage ganizational science: Rethinking Burrell and Morgan handbook of organizational discourse. London: Sage. and their legacy. Organization Science, 7: 191–207. Habermas, J. 1972. Knowledge and the human interest. Lon- Delanty, G. 2005. Social science. Buckingham, UK: Open Uni- don: Heinemann. versity Press. Hardy, S., & Clegg, S. 1997. Relativity without relativism: Derrida, J. 1978. (First published in 1967.) Edmund Husserl’s Reflexivity in post-paradigm organization studies. Brit- origin of geometry: An introduction. New York: Harvester ish Journal of Management, 8: 5–17. Press. Hargens, L. L. 2000. Using the literature: Reference networks, Dewey, J. 1916. Essays in experimental logic. New York: reference contexts, and the social structure of scholar- Dover. ship. American Sociological Review, 65: 846 – 865. Dewey, J. 1938. Logic: The theory of inquiry. New York: Holt. Haslam, A. 2004. Psychology of organizations (2nd ed.). Lon- don: Sage. DiMaggio, P. 1995. Comments on “What theory is not.” Administrative Science Quarterly, 40: 391–397. Haslam, A., & Reicher, S. 2006. Social identity and the dy- namics of organizational life. In C. Bartel, S. Blader, & A. Donaldson, L. 1985. In defence of organization theory. Cam- Wrzesniewski (Eds.), Identity and the modern organiza- bridge: Cambridge University Press. tion: 135–166. New York: Lawrence Erlbaum Associates. Dutton J., Dukerich, J., & Harquail, C. 1994. Organizational Heidegger, M. 1981. (First published in 1927.) Being and time. images and member identification. Administrative Sci- New York: SCM Press. ence Quarterly, 43: 293–327. Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, Elsbach, K. 1999. An expanded model of organizational iden- H., McElreath, R., Alvard, M., Barr, A., Ensminger, J., tification. Research in Organizational Behavior, 21: 163– Henrich, A. S., Hill, K., Gil-White, F., Gurven, M., Mar- 200. lowe, F. M., Patton, J. Q., & Tracer, D. 2005. “Economic man” in cross-cultural perspective: Behavioral experi- Ewenstein, B., & Whyte, J. 2009. Knowledge practices in de- ments in 15 small-scale societies. Behavioral and Brain sign. Organization Studies, 30: 7–30. Sciences, 28: 1– 61. Foucault, M. 1972. The archaeology of knowledge. New York: Husserl, E. 1970. (First published in 1900 –1901.) Logical inves- Pantheon Books. tigations, vol. 2. London: Routledge. Foucault, M. 1977. Discipline and punish: The birth of the Jenkins, R. 2000. Categorization: Identity, social process and prison. New York: Random House. epistemology. Current Sociology, 48: 7–25. Johanson, L. M. 2007. Sitting in your readers’ chair: Attending

270 Academy of Management Review April to your academic sensemakers. Journal of Management Mills, C. W. 1959. The sociological imagination. Oxford: Inquiry, 16: 290 –294. Oxford University Press. Johnson, M. S. 2003. Designating opponents in empirical Miner, J. B. 1984. The validity and usefulness of theories in an research: The rhetoric of “interestingness” in consumer emerging organizational science. Academy of Manage- research. Marketing Theory, 3: 477–501. ment Review, 9: 296 –306. Knights, D. 1992. Changing spaces: The disruptive impact of Mohr, B. 1982. Explaining organizational behavior. San Fran- a new epistemological location for the study of manage- cisco: Jossey-Bass. ment. Academy of Management Review, 17: 514 –536. Morgan, G. 1980. Paradigms, metaphors, and puzzle solving Knights, D., & Morgan, G. 1991. Corporate strategy, organi- in organization theory. Administrative Science Quar- zations, and subjectivity: A critique. Organization Stud- terly, 25: 605– 622. ies, 12: 251–273. Morgan, G. 1997. Images of organization. Thousand Oaks, Knights, D., & Willmott, H. 1989. Power and subjectivity at CA: Sage. work. Sociology, 23: 535–558. Mulkay, M., & Gilbert, N. G. 1983. Scientists’ theory talk. Knorr-Cetina, K. 1981. The manufacture of knowledge: An Canadian Journal of Sociology, 8: 179 –197. essay on the constructivist and contextual nature of sci- ence. New York: Pergamon Press. Musson, G., & Tietze, S. 2004. Places and spaces: The role of metonymy in organizational talk. Journal of Manage- Kuhn, T. S. 1970. The structure of scientific revolutions. Chi- ment Studies, 41: 1301–1323. cago: University of Chicago Press. Myers, G. 1993. Making enemies: How Gould and Lewontin Kunda, G. 1992. Engineering culture: Control and commit- criticize. In J. Selzer (Ed.), Understanding scientific prose: ment in a high-tech corporation. Philadelphia: Temple 256 –275. Madison: University of Wisconsin Press. University Press. Newton, T. 1998. Theorizing subjectivity in organizations: Latour, B., & Woolgar, S. 1979. Laboratory life: The social The failure of Foucauldian studies? Organization Stud- construction of scientific facts. London: Sage. ies, 19: 415– 447. Lee, T., Mitchell, T., & Sablynski, C. 1999. Qualitative re- Parker, M. 1991. Post-modern organizations or postmodern search in organizational and vocational behavior. Jour- organization theory? Organization Studies, 13: 1–17. nal of Vocational Behavior, 55: 161–187. Pfeffer, J. 1982. Organizations and organization theory. Cam- Locke, K., & Golden-Biddle, K. 1997. Constructing opportuni- bridge, MA: Ballinger. ties for contribution: Structuring intertextual coherence and “problematizing” in organizational studies. Acad- Pratt, M. 2000. The good, the bad, and the ambivalent: Man- emy of Management Journal, 40: 1023–1062. aging identification among Amway distributors. Admin- istrative Science Quarterly, 45: 456 – 493. Lundberg, C. C. 1976. Hypothesis creation in organizational behavior research. Academy of Management Review, 1: Pratt, M. 2009. From the editors: The lack of a boilerplate: 5–12. Tips on writing up (and rewriting) qualitative research. Academy of Management Journal, 52: 856 – 862. Lu¨ scher, L. S., & Lewis, M. W. 2008. Organizational change and managerial sensemaking: Working through para- Pratt, M., & Foreman, P. 2000. Classifying responses to mul- dox. Academy of Management Journal, 51: 221–240. tiple organizational identities. Academy of Manage- ment Review, 25: 18 – 42. Lynch, M. 2000. Against reflexivity as an academic virtue and source of privileged knowledge. Theory, Culture & Reed, M. 1985. Re-directions in organizational analysis. Lon- Society, 17: 26 –54. don: Routledge. March, J., & Olsen, J. 1976. Ambiguity and choice in organi- Reed, M. 2004. Getting real about organizational discourse. zations. Bergen: Unversitetsforlaget. In D. Grant, C. Hardy, C. Oswick, & L. Putnam (Eds.), Handbook of organizational discourse: 413– 420. London: Martin, J. 2002. Organizational culture: Mapping the terrain. Sage. Thousand Oaks, CA: Sage. Rorty, R. 1992. Cosmopolitanism without emancipation: A Martin, J., & Meyerson, D. 1988. Organizational culture and response to Lyotard. In S. Lash & J. Friedman (Eds.), the denial, channeling and acknowledgment of ambigu- Modernity & identity: 59 –72. Oxford: Blackwell. ity. In L. R. Pondy (Ed.), Managing ambiguity and change: 93–125. New York: Wiley. Rosenau, P. M. 1992. Post-modernism and the social sciences: Insights, inroads and intrusions. Princeton, NJ: Princeton McKinley, W., Mone, M. A., & Moon, G. 1999. Determinants University Press. and development of schools in organization theory. Academy of Management Review, 24: 634 – 648. Sandberg, J. 2000. Understanding human competence at work: An interpretive approach. Academy of Manage- McMullen, J., & Shepard, D. 2006. Encouraging consensus- ment Journal, 43: 9 –25. challenging research in universities. Journal of Manage- ment Studies, 43: 1643–1670. Sandberg, J. 2001. The constructions of social construction- ism: In S. E. Sjo¨ strand, J. Sandberg, & M. Tyrstrup (Eds.), Merleau-Ponty, M. 1962. (First published in 1945.) Phenome- Invisible management: The social construction of lead- nology of perception. London: Routledge and Kegan ership: 29 – 48. London: Thomson. Paul.

2011 Alvesson and Sandberg 271 Sandberg, J., & Alvesson, M. 2011. Ways of constructing re- Townley, B. 1993. Foucault, power/knowledge, and its rele- search questions: Gap-spotting or problematization? vance for human resource management. Academy of Organization, 18: 23– 44. Management Review, 18: 518 –545. Sandberg, J., & Targama, A. 2007. Managing understanding Tsang, E. W. K., & Frey, B. S. 2007. The as-is journal review in organizations. London: Sage. process: Let authors own their ideas. Academy of Man- agement Learning & Education, 6: 128 –136. Schein, E. 1985. Organization culture and leadership. San Francisco: Jossey-Bass. Tsoukas, H., & Knudsen, C. (Eds.). 2004. The Oxford handbook of organization theory. Oxford: Oxford University Press. Sennett, R. 1998. The corrosion of character. New York: Norton. Van de Ven, A. H. 2007. Engaged scholarship. A guide for organizational and social research. New York: Oxford Shotter, J., & Gergen, K. (Eds.). 1989. Texts of identity. London: University Press. Sage. Van Maanen, J., & Barley, S. R. 1984. Occupational commu- Sievers, B. 1986. Beyond the surrogate of motivation. Orga- nities: Culture and control in organizations. Research in nization Studies, 7: 335–351. Organizational Behavior, 6: 287–365. Simon, H. A. 1947. Administrative behavior: A study of deci- Weedon, C. 1987. Feminist practice and poststructuralist sion-making processes in administrative organization. theory. Oxford: Blackwell. New York: Macmillan. Weick, K. E. 1989. Theory construction as disciplined imagi- Slife, B. D., & Williams, R. N. 1995. What’s behind the re- nation. Academy of Management Review, 14: 516 –531. search? Discovering hidden assumptions in the behav- ioral sciences. Thousand Oaks, CA: Sage. Weick, K. E. 2001. Gapping the relevance gap: Fashions meet fundamentalist in management research. British Journal Smircich, L. 1983. Concepts of culture and organizational of Management, 12: 71–75. analysis. Administrative Science Quarterly, 28: 339 –358. Westphal, J., & Khanna, P. 2003. Keeping directors in line: Smith, K. G., & Hitt, M. A. (Eds.). 2005. Great minds in man- Social distancing as a control mechanism in the corpo- agement: The process of developing theory. New York: rate elite. Administrative Science Quarterly, 48: 361–398. Oxford University Press. Westwood, R., & Clegg, S. 2003. The discourse of organiza- Starbuck, W. H. 2003. Turning lemons into lemonade: Where tion studies: Dissensus, politics, and paradigms. In R. is the value in peer reviews? Journal of Management Westwood & S. Clegg (Eds.), Debating organization: Inquiry, 12: 344 –351. Point-counterpoint in organization studies: 1– 43. Oxford: Blackwell. Starbuck, W. H. 2006. The production of knowledge: The chal- lenge of social science research. Oxford: Oxford Univer- Wicker, A. W. 1985. Getting out of our conceptual ruts. Amer- sity Press. ican Psychologist, 40: 1094 –1103. Sutton, R., & Staw, B. 1995. What theory is not. Administrative Willmott, H. 1993. Strength is ignorance; slavery is freedom: Science Quarterly, 40: 371–384. Managing culture in modern organizations. Journal of Management Studies, 30: 515–552. Thornborrow, T., & Brown, A. 2009. Being regimented: Aspi- ration, discipline and identity work in the British para- Yukl, G. 2006. Leadership in organizations (6th ed.). Upper chute regiment. Organization Studies, 30: 355–376. Saddle River, NJ: Pearson/Prentice-Hall. Mats Alvesson ([email protected]) is professor of business administration at the University of Lund and the University of Queensland Business School. He received his Ph.D. from the University of Lund. His research interests include critical theory, gender, power, professional services firms, organizational culture, leadership, iden- tity, organizational image, qualitative methods, and philosophy of science. Jo¨ rgen Sandberg ([email protected]) is a reader in management in the School of Business at the University of Queensland and leads its research program, Knowledge in Organizations. He received his Ph.D. from the University of Gothenburg. His research interests include competence and learning in organizations, leadership, practice-based theories, qualitative research methods, and the philosophical under- pinnings of organizational research.

Copyright of Academy of Management Review is the property of Academy of Management and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.