The hegemonic ambiguity of big concepts in organization studies

Introduction

The use of concepts is a vital part of the research process. A scientific concept can be described as ‘a general idea which, once having been tagged, substantially generalized, and explicated can effectively guide inquiry into seemingly diverse phenomena’ (Merton, 1984: 267). It is indeed regarded as ‘one of the great tools of all scientific knowledge’ (Weber, 1946: 141) and a pivotal element of theorizing (Bacharach, 1989; Sutton and Staw, 1995), which typically includes activities such as ‘naming, conceptualization, constructing typologies . . . and so on’ (Swedberg, 2012: 15). The idea of using concepts and constructs is to sharpen thinking and communicate clearly, or, as Suddaby (2010: 352) puts it:

First, clear constructs facilitate communication between scholars. Second, improved clarity of constructs enhances researchers’ ability to empirically explore phenomena. Third, clear constructs allow for greater creativity and innovation in research.

Given their importance for scientific work, it is surprising how seldom we openly reflect upon and discuss the qualities and fundamentals of concepts and their use, as also pointed out by Suddaby (2010: 346): ‘The absence of an open discussion about theoretical constructs is somewhat surprising given their widespread use in and undeniable importance to management theory.’

In general, there is certainly no lack of development of ‘new’ concepts. This massive flow of concepts can cause confusion or, as Swedberg (2012: 20) writes:

[O]ne should avoid introducing too many names and also to give a new phenomenon some odd new name. It is rare to discover something really novel; and forcing the reader to remember new terms, without getting much for it, only creates irritation.

Indeed, new labels for the same old things and the repackaging of old ideas are both common and frustrating. Supervision and, to a large extent, managerial work has, for example, been replaced by the more appealing label leadership, used to denote any act of a person in a formal authority position (Bedeian and Hunt, 2006). Equally frustrating is the tendency of overexploiting a popular term by adding more and more (vague and poorly defined) meanings to it, thereby making its boundaries unclear and the concept increasingly unwieldy (Haslam, 2016). This leads to significant confusion and makes navigating in intellectual fields difficult. As a result, both theoretical thinking and communication between scholars (and others) become problematic. We will in this article address popular concepts used in varied and ambiguous ways.

We refer to these concepts as hembigs – an acronym for hegemonic, ambiguous, big concepts. This seemingly oxymoronic concept draws on Gramsci’s (1971) notion of cultural and/or linguistic dominance at the expense of other alternative expressions and vocabulary. Ambiguity refers to vagueness and uncertainty associated with multiple, incoherent meanings attributed to the phenomena in question. Big concerns the unhelpful broad application and usage of the concept, simply covering ‘too much’.

We realize that we move on a fine line between adding redundant and confusing new terms and constructively offering something that facilitates new thinking. We hope that we do the latter. Sometimes we need new concepts, partly because they allow the capturing of phenomena or a line of reasoning more sharply, and partly because they evoke new associations and inspire a different imagery.

Our primary ambition is to show that the use of dominant, ambiguous, broad-ranging concepts in our field obstructs good research and we address ways of dealing with this problem. We thus discuss how we can at least take some small steps from loose to sharper concept use in organization studies. Our hope is that we can facilitate clearer and more self-critical thinking in our field by illuminating problems with the too easy and vague use of popular and easily available terms. We are well aware that our article can be read as policing the field, creating constraints for ‘free’ language use and eliciting negative reactions from people eager to use their favourite and identity-supporting concepts as they see fit. As we will show, we enter contested terrain.

In this article, we will (a) introduce hembig as a concept and discuss its value, (b) empirically demonstrate the existence of hembigs, (c) show the substantive problems of hembigs and (d) discuss how hembigs can be addressed and their negative effects mitigated. The acronym ‘hembig’ might not be the most elegant one, but it captures the problems and its elements quite distinctly and we believe its introduction will serve a purpose. As we will show, it adds significantly to previously established terms like surplus meaning (MacCorquodale and Meehl, 1948), contested concepts (Gallie, 1956, 1964), concept creep (Haslam, 2016) and umbrella concepts (Astley, 1985; Hirsch and Levin, 1999). We will in a reflexive spirit also address the risk of our own concept potentially being or becoming just another hembig.

We focus on organization studies even if the discussion also is highly relevant for other fields within social science. Our main point is that the aggregate of meanings and ‘all-inclusive’ treatment of some popular concepts in organization studies make them appear as exceptionally important and broadly applicable. But this also makes the use of concepts intellectually problematic and a source of ambiguity and confusion. This affects research projects, communication between researchers within a field (hembig users) as well as those outside the field trying to get a grip of what researchers are referring to; what do I/we/they mean by X becomes an escalated problem when concepts get added hembig qualities. Additional problems will be addressed later in the article.

The remainder of the article is structured as follows. We continue by defining hembigs and relate it to similar concepts. We then present three important and illustrative cases within organization studies, with hembig tendencies: leadership, strategy and institution. After that, we discuss why hembigs develop, why they are problematic, and what we think can be done in order to relieve the problems.

Hembigs defined

In this section, we briefly discuss the three keywords that together constitute hembig: hegemony, ambiguity and scope. Hegemony (originally from Greek politico-military dominance) refers to cultural and/or linguistic dominance at the expense of other alternative expressions and vocabulary. As Gramsci (1971) emphasized, hegemony is largely achieved through spontaneous consent by subordinate interests to the cultural domination they believe will serve their interests because it is regarded as ‘common sense’. People are dragged into this, take it for granted and embrace a consensus that hides potential conflicts and problems. ‘Hegemony refers to the ways in which subjects participate in their own subordination through socially constructing their identities, knowledge and institutions’ (Putnam and Boys, 2006: 557). Grint et al. (2016: 14) write that ‘Leadership is a salient example of a discursive hegemonic constellation’. A similar assessment of the domination of institutional theory views ‘modern institutionalism as a mythical beast that cannot be killed. When one stabs at it, it merely grows new appendages from its wounds’ (Kraatz, 2020: 255). It becomes almost impossible to resist. We will return to both leadership and institutional theory as well as address strategy later in the article.

Ambiguity refers to vagueness and uncertainty associated with multiple, incoherent meanings attributed to a phenomenon (Alvesson and Sveningsson, 2003; Levine, 1985; Meyerson and Martin, 1987; Sillince et al., 2012). It involves uncertainty that cannot be resolved or reconciled, and absence of agreement on boundaries, clear principles or solutions (Gray, 1977). Ambiguity means that a group of informed people are likely to hold multiple meanings and/or that several plausible interpretations can be made, without more data or rigorous analysis making it possible to assess them. Ambiguity is related to the notion of ‘essentially contested concepts’, being concepts which ‘inevitably involves endless disputes about their proper uses on the part of their users’ (Gallie, 1956: 169). Such disputes are assumed to be based on the nature of particular concepts (in Gallie’s article, abstract, qualitative and evaluative concepts) rather than on different (perhaps radically confused and/or hotly disputed) uses of these concepts (Clarke, 1979; Collier et al., 2006; Dworkin, 1972; Waldron, 2002). The latter is our main concern and focus in this article. We here address what Levine (1985: 8) refers to as ‘ambiguities of language and thought’, as opposed to the more fundamental and inevitable ambiguities of life and experience. Ambiguity is an important theme for investigations, but often scientific concepts are used in order to deny or repress ambiguity. They are applied in such a way as to give the appearance that a solid phenomenon is being studied, e.g. authors claim that there is such a thing as authentic leadership or an HR strategy and this is clearly captured by the study in question, but this often conceals ambiguity and diversity. The possibility that what appears to be one phenomenon may be seen as a variety of phenomena is not addressed. Levine (1985: 8) warns us of our ‘trained incapacity to observe and represent ambiguity as an empirical phenomenon . . . [and] . . . insufficient awareness of the multiple meanings of commonly used terms in social science’. As we will see, hembigs are usually not disputed on a larger scale, due to their hegemonic effect; we accept and use these concepts without problematization or acknowledging the ambiguity.

Scope refers to the range of meanings attributed to a concept. A broad scope (‘big’ in our acronym) involves a large number of more or less coherent meanings, which typically also leads to the concept being applied and used in a wide-ranging set of contexts and situations. Hirsch and Levin (1999: 200) talk about umbrella concepts, ‘used loosely to encompass and account for a set of diverse phenomena’. Concept use then implies that diverse phenomena are, in fact, the same phenomenon. According to them, organizational science is – or rather was – characterized by its attention to successive concepts, with different (but often related) terminologies that conform to a life-cycle of emerging excitement and expansion, followed by critique for vagueness and inconsistencies, and either transformation or decline. But as we will see, this cycle seems to come to a halt, as the hegemonic effect prevents or marginalizes effective critique, saving concepts from decline. Critiques of concepts like leadership and institutional theory for being too broad, vague and all-encompassing do not prevent their juggernauts from rolling on (Learmonth and Morrell, 2019; Lok, 2020). Haslam (2016) recently also identified the tendency of swelling concepts within the field of psychology and labelled it ‘concept creep’. He saw both vertical expansion, which occurs when ‘a concept’s meaning becomes less stringent, extending it to quantitatively milder variants of the phenomenon to which it originally referred’, and horizontal expansion, which refers to a concept that ‘extends to a qualitatively new class of phenomena or is applied in a new context’ (Haslam, 2016: 2).

Sometimes these types of scope expansions result in a research community divided in ‘camouflaged sub-tribes’ that seemingly use the same concept, but with very varied meanings (Levine, 1985). In the best case, these sub-tribes are well aware of their differences and are able to navigate within and between the various tribes (Abbott, 2001). However, clarity and understanding are often limited, and confusion prevails. A broad conceptual scope may or may not lead to ambiguity (above), but the risk usually increases if a concept covers a broad terrain in terms of variations and application.

We can thus identify (what we suggest productively can be seen as) a hembig by its broad and diverse use, tendency to, over time, crowd out other concepts (even if they would be more precise or helpful for the purpose of representation), and the high and escalating level of ambiguity, typically due to the expansion of the use of the concept. A hembig scores high on all these three criteria.

There is a tension between a focus on concepts ‘as such’ and how concepts are used. As concepts are not disconnected from their actual use – all concepts change over time in a non-stable world, and few people use concepts in entirely idiosyncratic ways – this distinction should, however, not be overemphasized. We can point to three different possibilities or perspectives: (a) concepts having a long-term, stable, dictionary type of meaning – here we can refer to concepts ‘as such’; (b) concepts tending to be used in a broadly similar way in a specific field over at least a medium–long period and thus having a field-specific ‘quasi-dictionary’ meaning; and (c) concepts in use in specific texts by specific authors. (The last point may look at very detailed, context-specific meanings or how the text ‘on the whole’ uses a concept.) Our concept hembig primarily refers to (b), i.e. how concepts are broadly used by a community of researchers within a specific field. But this is, of course, not unrelated to (a) or (c). We cannot talk of concepts as such, as if they have a given, internal meaning outside their use in changing areas like (most parts of) social science. On the other hand, concepts in use typically reflect a broadly agreed upon and collectively enacted meaning that over time may change with concept use. Strategy in organization studies today, for example, is a very different concept from the one originally used by the military (Knights and Morgan, 1991). We are more interested in generalized or common ways of using the concepts – discourses in a Foucauldian (1976) – than the very specific language use revealed by close-up readings of texts that deconstructionists and discourse analysts focus on (e.g. Potter and Wetherell, 1987). The idea of hembig is to sensitize (Blumer, 1969) us to the variation and escalation of ambiguity within a knowledge area that follows from concept use, as the tolerance for ambiguity and big claims broadens and the pressure for using the concepts becomes stronger.

To sum up, concepts are not by nature, or some intrinsic property, hembigs, i.e. it is not words per se but how they are being used that is the problem. But when particular signifiers become broadly (over)used, they tend to be attributed to so many diverse, vague and inconsistent meanings and associations that confusion prevails and any author or text has problems escaping this. Hembigs are thus concepts that have become overloaded with cultural meanings that individual researchers become caught in. A ‘culture of ambiguity’ makes the consumption and use of texts vague and varied. It is not only the individual but, perhaps above all, the social context that turns concepts into hembigs – we are now witnessing an escalation of the use of hembigs in organization studies. We should add that our task here is not to join what Levine (1985: 2) refers to as the ‘modern assault on ambiguity’ – we are well aware of the poststructuralist emphasis on the fluidity of meaning – but to raise awareness of problems associated with tendencies to bypass rather than face problems with hembigs.

Three examples

We will now illustrate our point with three influential scientific concepts within organization studies: leadership, strategy and institution. Space constraints make it difficult to address additional concepts. Why these three? We could have picked almost any one being popular – culture, knowledge, identity, power, service, brand, transaction, resource, profession, organization, resistance, innovation, entrepreneurship, development, sustainability, and so forth. The main reason for picking the three mentioned concepts – leadership, strategy and institution – is that they illuminate the hembig problem very well. They belong to some of the most popular concepts in use in organization studies, and a scrutiny of them can hopefully encourage more careful work within the various camps. Equally important is that we have worked ourselves with these concepts and are well familiar with the literature. We therefore think that we can competently address them.

Not all concepts are best described as hembigs

Here we anticipate the reader associating in different directions and perhaps starting to wonder if all concepts (as typically used) in wide circulation are hembigs, i.e. having been addressed and used in hegemonic, ambiguous and big ways. Perhaps this is the case with management, human resources, trust, quality, diversity and all other popular terms? Hembigs are common, but there are concepts that, at least so far, have not on a grand scale become – or can productively be understood as – hembigs. As said, we are primarily interested in the general usage of specific concepts, and there are, of course, variations between and within areas. Even if it is difficult to break away from the collective ‘hembig problem’ – as the literature may bear imprints of this – individual authors and texts may be more or less able to handle the problem and avoid using concepts as hembigs, even in areas where this is common. In this article we focus on broader tendencies in terms of concept usage and do not go into individual variations and exceptions. Of course, within all (popular) areas, there are variations: some authors and texts use concepts in a clear and specific way; others score high on ‘hembig qualities’, including addressing phenomena in line with a dominant (but not necessarily the best) concept (e.g. talking about knowledge instead of information, innovation rather than product modification), within a wide scope of phenomena while being vague and ambiguous about what is referred to. In assessing the usage of concepts, one needs to consider all the components in the hembig: (a) is there a hegemonic element, in the sense that alternative, more precise, but less fashionable terms are not used; (b) are there signs of ambiguity (and not just clearly different definitions and variation); and (c) is the scope of what is addressed very large?

The critical reader may ask whether a hembig is a hembig in itself, i.e. referring to ‘everything’ in an ambiguous way? As the concept hembig is not widely used (actually it is probably so far only used by us), it is far from being hegemonic. (We also predict that it may be, if not neglected, at least heavily contested, as many have an interest in the continuous use of their preferred concepts and may resist problematization in the references to hembigs.) The lack of aesthetic and seductive quality of the word hembig in combination with the critical edge further reduce the risk. In that sense, hembig may actually work more as an ‘anti-hembig’, not likely to reinforce hegemony, and only used to illuminate major problems. But we realize that the concept, if becoming part of social studies vocabulary, may be applied to address ‘everything’ and used more ambiguously. Of course, our use is not razor-sharp; this is an area where a very high level of precision is not possible or optimal, and hembig can here be seen as a sensitising concept (Blumer, 1969) rather than an operationalized construct ready for empirical testing. We still, however, try to avoid a problematically wide and ambiguous use. Through restricting the scope by pointing to exceptions, we hope to counter the risk of the hembig becoming/being used as an ambiguous and big concept. We give two examples: identity work and isomorphism.

Brown (2015) suggests that identity work is the most significant metaphor among many that may be useful in the analysis of identities in and around organizations. He refers to ‘perhaps the most widely cited formulation’, by Sveningsson and Alvesson (2003: 1165), who suggest that ‘ . . . identity work refers to people being engaged in forming, repairing, maintaining, strengthening or revising the constructions that are productive of a sense of coherence and distinctiveness’. In his review article, Brown (2015) identified 31 texts in the management literature with identity work in the title, and noted that the majority of researchers come close to this definition and use of the concept (e.g. Down and Reveley 2009; Petriglieri and Petriglieri, 2010), and also more recent uses of the word follow this route (e.g. Watson, 2020), so the concept seems broadly to be used in similar and fairly consistent ways. There are, of course, variations; for example, a few see identity work as something that occurs more or less continuously in the course of organizational life, triggered by run-of-the-mill events, as Brown (2015) puts it. However, most scholars address identity work that occurs in particularly demanding situations or at times of significant transition. The latter view scores lower on the ‘big’ element: there are fewer instances and less scope if identity work is viewed as episodic rather than on-going and covering more or less everything within organizational behaviour, both everyday routine issues and more extraordinary or routine-breaking exposures to identity. But as said, most authors seem to avoid the ‘bigness’ problem of identity work.

Some researchers suggest and work with alternative framings, downplaying or denying the ‘work’ element. For example, Brown (2015) points at identity practicing as well as identity play. While work is often associated with compliance, rationality, logic, effort and a means–ends orientation, play implicates a different set of potentially generative ideas relating to enjoyment, discovery, intuition, imagined others, spontaneity, and fantasy (Pratt, 2012). A more recent concept is teflonic identity manoeuvring, suggesting that what could be seen as challenges to identity concerns are taken rather lightly and in an impersonal way (Alvesson and Robertson, 2016) and not triggering identity work.

This indicates that current use of the identity work concept is probably not something that motivates being seen as an hembig. It is not hegemonic, it is not particularly ambiguous, and most authors use it to address fairly distinct terrain. It is, we think, quite different from concepts like institutional work, strategy in practice or discursive practice, where the combination of two individually hegemonic terms does not lead to much ‘zooming in’ or reduction of ambiguity or scope. Identity work, then, is less of a hembig than identity (without ‘work’ used as a specifying element), a concept used very broadly and loosely.

Another popular concept that we do not see as a hembig, even if it is an offspring of institutional theory, is isomorphism. The definition by DiMaggio and Powell (1983: 149) is fairly distinct: ‘isomorphism is a constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions’. They follow Meyer and Rowan (1977: 347), who argued that ‘[o]rganizations both deal with their environments at their boundaries and imitate environmental elements in their structures’. As Boxenbaum and Jonsson (2017) write, organizations became similar not through adaptation to an external or technically demanding environment or through the ‘weeding out’ of technical and social misfits, but through adaptation to a socially constructed environment. The concept of isomorphism seems to be applied in organization studies in a fairly clear way and it is not over-used, i.e. put in operation all over the place or working as a seductive, hard-to-avoid term, with a clear loss of feeling of being on the right track or missing the ideological benefits of appearing attractive and gaining extra attention, in the way that, for example, leadership, strategy, institution, knowledge, diversity or identity (as opposed to identity work) have. We would therefore hesitate in addressing the general use of isomorphism as a hembig.

We thus suggest that identity work and isomorphism can be seen as often used concepts that so far are not necessarily good candidates for being described as hembigs. We do not want to make sharp statements, such as this is a hembig and this is not, or claim that this concept scores 3 on a 5-point hembig scale. The point of this article is to encourage critical reflection of what we see as the hembig problem and encourage researchers and evaluators of research to carefully consider and work to reduce the potential problems. As emphasized, we do not refer to concepts ‘as such’ nor specific details of how individual texts use the concepts, but to the general ways concepts are used within a research area during a particular episode, e.g. over a couple of decades or longer. Over time, hembig qualities may vary, typically going up when a concept is fashionable and the temptation to follow the flow becomes so strong that it is used without much care or discipline. As we will discuss in the next section, as time passes, some hembig concepts may run out of fashion and the hegemonic pressure and tolerance for breadth and vagueness are reduced.

The acceptance of hembigs of organization studies

About 20 years ago, Hirsch and Levin (1999) identified a struggle between umbrella advocates, supportive of broad concepts, believing that without them our field risks becoming disconnected and irrelevant, and a validity police who called for narrower perspectives with rigour, precision and clarity in concept use. Without the police, Hirsch and Levin (1999: 200) argued, ‘our field risks becoming too sloppy and scattered’. They suggest the following simple development stages for umbrella concepts: emerging excitement, followed by critique, and either transformation or decline.

Hirsch and Levin’s (1999) model is supported by an examination of the development of the concept ‘organizational effectiveness’. They show that it became popular in the 1970s, but definitions rarely overlapped, and so no consensus ever emerged on what exactly constituted organizational effectiveness. Organizational effectiveness as a concept (and a research topic) then gradually went out of fashion, so between 1977 and 1994 the proportion of articles published on organizational effectiveness dropped from 0.3–4% to 0.01%. Hirsch and Levin (1999) then postulate that umbrella constructs will eventually be seriously challenged and collapse and then possibly be reborn with a new and different name.

Hirsch and Levin (1999) departed from a positivistic empirical tradition, where problems in finding consensus on definitions and measurements lead to the demise of studies of effectiveness. We are less concerned about the exact relationship between concepts and empirics and have doubts about correspondence theory. We find their term validity police unfortunate, as neither validity nor police are necessarily the best ways of conceptualizing what we find to be better addressed as ‘hembig critique’. We do, however, see their main point about the vague and scattered use of concepts, calling for some efforts to critically scrutinize these and increase clarity in language use, as important.

We made a similar review as Hirsch and Levin (1999) by looking at the publication pattern in 20 of the most prestigious journals for management research. For a complete list of journals included in the study, see Appendix 1. Compared to Hirsch and Levin (1999), a different pattern emerged. During the period 2000–2017, we found a steady rise of articles with ‘leader’ or ‘leadership’ and ‘institution’ or ‘institutional’ in the title, abstract and/or keywords, while the number of articles with ‘strategy’ and ‘strategic’ increased modestly (in absolute as well as relative numbers). For a graphical overview of the developments described above, see Figures 1–3 in Appendix 1. We do not make claims about the exact distribution of articles on various themes – some articles also overlap as they use more than one of our three hembigs in signalling their content – but we are only interested in developments over time. Our data only show the frequency of particular terms and not how they are defined, used, embraced or criticized in the actual texts. This means that the sample of articles might include some that actually criticize mainstream conceptualizations of, for example, ‘leadership’.

Compared with Hirsch and Levin’s (1999) model, the situation has changed, at least for some concepts. The considerable critique of the three ‘umbrella concepts’ here addressed has not led to signs of decline, collapse or re-labelling. Strategy shows roughly the same percentage over the time period 2000–2017, while leadership and institution have significantly increased in terms of proportions of all published studies. Established concepts seem to go on undisturbed, despite critique for being too broad and vague. For example, despite the critique of institutional theory for being all over the place, advocates still see promise in integration, as Lok (2020) points out. There appears to be little of the fruitful dialectic between hembig (umbrella) advocates and critique (validity police): as umbrellas have turned to hembigs they are difficult to rock. Palmer and Hardy (2000: 233) remark that strategy as well as leadership are ‘difficult to escape’, indicating the terms’ hegemonic status (see also Crevani et al., 2010: 81–82, trying to motivate their use of the term leadership).

Our study indicates the need to revise Hirsch and Levin’s (1999) model: ‘Emerging excitement, followed by critique, and either transformation or decline’. The development seems instead to be: ‘Emerging excitement, further expansion, a hegemonic position with marginal and neglected critique and continued (but increasingly confused and confusing) full scale use’. Umbrella advocates seem to have an ability to fend away and neglect critics. In contrast to Hirsch and Levin’s (1999) model, critique seems here no longer to be a salient and effective part of a productive dialectic process.

What drives the development of hembigs?

Social science is, of course, susceptible to academic fads and fashion (Miller, 2009; Starbuck, 2009), and scholars sometimes use power and politics to strengthen group positions and individual careers. Therefore, labels sometimes have more of social and/or political functioning than representing a distinct intellectual content (Astley, 1985). Our hembig critique is, of course, not free from socio-political dynamics, but we do not have a specific agenda of promoting a specific stream of research or supporting a distinct sub-community of academics. However, on a confessional note, we should perhaps mention that we have published work in the areas we here critically discuss, using/benefiting from the hembig labels ourselves, so there is also an element of self-critique in this article. We acknowledge the temptation to engage in ‘big’ issues and use terms loosely.

The increased use of a favourable, seemingly important label can furthermore be understood in the light of the massification of higher education and research. With tens of thousands of organization researchers trying to publish every year, it is difficult for researchers to develop really new ideas or remarkable findings. There is a tendency to change focus from generating knowledge to generating careers (Alvesson et al., 2017b). Researchers focus on what seems publishable and what generates citations. With an expansion – perhaps even an over-supply – of studies aiming to find a place in a crowded terrain, a quick signalling of something recognizable becomes important. A brand logic takes the upper hand (Mehrpouya and Willmott, 2018). Locating oneself firmly in a sub-literature and showing a full grasp of this pushes researchers away from broader scholarship. Furthermore, complicating work by adding reflections around the basic terminology involved would sometimes only make things more difficult. Many researchers adapt a gap-spotting logic, and the ‘find and fill the gap in the literature-mentality’ gives limited space for broader reflections (Sandberg and Alvesson, 2011). Mastery of a specific hembig here becomes central. In addition to this, opportunism makes us associate ourselves with seemingly ‘winning’ concepts that increase the chances of grants, publication and citations. There is a strong tendency to go for what is popular, creating self-reinforcing developments. A well-known concept gets attention and is easy to remember and relate to.

As more and more are defined as a strategic issue and with the accumulation of claims about the centrality of leadership (not management), then the ‘non-strategic’ or what is not ‘leadership’ become insignificant. Given how many things are referred to as leadership, strategy or institutions, there appears to be little point in considering ‘non-leadership’, ‘tactics’ and ‘non-institutions’, as they are not much ‘outside’ these concepts the way they are normally used. But, as emphasized earlier, as it is the actual usage of the concept that matters, we can only talk about a hembig Ex Post, not Ex Ante, based solely on a definition.

Related to the above is that big concepts also make it possible to scale up rather small and mundane empirical studies, making them appear not only socially significant but as contributions to great theory. Interviews with supervisors become studies of leadership, small modifications of a product portfolio are addressed as strategic change and almost anything can be characterised as institutional work or institutional logic. It makes it easier to create the impression of intellectual work being organized through schools, journal special issues and handbooks, covering up diversity and fragmentation through a seemingly unifying label.

Often, career investments and habits stop us from questioning the usefulness or the relevance of a particular concept. If a large part of one’s career has been built upon a certain concept, it is hard to ‘kill a darling’ or ‘the cow that provides the milk’. Identity issues also matter and become fused with social and instrumental concerns of hembig protection. Academics sometimes appear to live and die with their favoured label. Although some are into hero-worshipping and see themselves very much as followers of Foucault, Freud, Lacan, Heidegger or another guru, others are into ‘labelism’, carefully guarding the right concept. Repeating the same symbolically loaded vocabulary becomes a part of who one is. Leadership, strategy, institution etc. become emotionally charged signifiers, making hembig critique unwelcome (without denying that there are elements of valid, intellectual reasoning in the defence of some concepts that may not warrant being called hembigs). An illustrating example of ‘labelism’ is the recent debate on current problems with the understanding of ‘institution’ in Journal of Management Inquiry, where a proposal to put a (temporary) moratorium on the use of the term ‘institution’ (Alvesson et al., 2019) is met very negatively (Kraatz, 2020; Ocasio and Gai, 2020).

All these deeply human factors described above (by which we, the authors, of course also are affected) tend to coexist and reinforce each other. Sometimes a ‘cognitive economy’ makes us stick to a certain concept and its tail of a well-packaged set of literature. Here, unwillingness to think through issues and terminology make us overuse overloaded and perhaps less relevant concepts in order to make sense of a certain phenomenon.

So what?

We are, of course, not the first ones that have observed the fragmented use of broad and vague concepts (see, for example, Clarke, 1979; Dewey, 1979; Dworkin, 1972; Gallie, 1956; Garver, 1978; Haslam, 2016; Hirsch and Levin, 1999). Common responses when faced with broad and contested concepts include academic dogmatism (my take on the concept is the right one and all others are wrong), scepticism/nihilism (everyone has a right to his/her own truth), and eclecticisms/pluralism (each view adds new perspectives) (Garver, 1978: 168). Eclecticism and pluralism seem to be common (e.g. Abbott, 2001; Levine, 1985). A fatalist – or even Darwinist – approach to this is sometimes taken where it could be argued that ‘just as successful species increase their territory, invading and adapting to new habitats, successful concepts and disciplines also expand their range into new semantic niches. Concepts that successfully attract the attention of researchers and practitioners are more likely to be applied in new ways and new contexts than those that do not’ (Haslam, 2016: 12). In a generous and pluralistic spirit this can be seen as unavoidable, acceptable or even good.

However, all these responses generate problems or challenges for researchers engaged with the particular concept in question and even more so for others without deep expertise in the area:

What can be done?

How, then, can we mitigate the development of hembigs and its negative effects? Are there any antidotes or remedies?

Specifying the key concept in question is important, but perhaps less common than one would assume. A formal definition is seldom sufficient. That an institution is ‘cultural rules’ or ‘a procedure’, a strategy is about ‘shaping the future’ or leadership is ‘having followers’ or ‘exercising influence’ do not say much. Longer definitions may as well increase as reduce vagueness as the words used may call for additional definitions and clarifications. As Weick (1987) shows, some definitions of culture and strategy look almost the same, despite the concepts typically referring to quite different phenomena or aspects of organizations. Geertz (1973) recommends people to cut concepts down in size so that they cover less and reveal more. A possibility here is to illuminate what a concept is not covering. To make sense of leadership, strategy, institution we need to understand what is not leadership, institution, strategy etc. This ‘not’ needs to be reasonably specific. Again, a good example is Geertz (1973) pointing at what culture is not through contrasting it with structure. The argument above does not imply that we can control the use of a specific signifier. Language use tends to be complicated, situational, metaphorical, performative etc., and does not simply reflect reality or work as a device for the transport of meaning. But authors can be more or less careful and try to sharpen the meanings of favourite concepts. Creativity in research is very much a matter of using language in a clever way. Hembigs often stand in opposition to this.

An additional possibility is to unpack hembigs through clear differentiation of categories. Interesting research sometimes means demonstrating that what appears to be one phenomenon is in fact two or more phenomena (Davis, 1971). This is more than just saying that there are different types of leadership, strategy, institutions or anything else. It means that what appears to be variation of one and the same category, conveniently labelled through the use of one overall term, is better viewed as fundamentally different phenomena, calling for different labels. The distinction between planned and emergent strategy (Mintzberg and Waters, 1985), indicating sameness through the common word ‘strategy’, could, for example, be viewed as quite diverse and labelled, e.g. centralized planning versus emerging and converging direction paths. Lok (2020) suggests five clearly delimited versions of what is conventionally labelled (dumped together) as institutional theory. He shows the variation and indicates that these versions could be treated explicitly and distinctively. These suggestions do not necessarily mean that we stop using the master signifier altogether, but that more moderate, mid-range concepts are brought to the front. In a more radical spirit, Grint et al. (2016) suggest that we replace leadership with alternative signifiers: efficient organizing, rhetorical flourish etc. They ask whether the term leadership is needed at all: ‘Wouldn’t we inhabit more transparent and accountable organisations were these (and other) organisational and social constructs not poorly expressed, or concealed entirely, under a single signifier’ (Grint et al., 2016: 16). See also Alvesson et al. (2017a) and Learmonth and Morrell (2019) for similar ideas.

The research community could also support a hembig critique function in various organized ways. Journals could, for example, use less narrowly specialized reviewers – not socialized into hembigs – to encourage authors and reviewers to reflect and critically scrutinize major concepts. Rather than integrative or even critical reviews one could have anti-reviews, i.e. aiming at radical unpacking of what reviewers normally try to integrate or find common themes within. The principle would then be to look for divergence rather than convergence. Reviews often proceed from the assumption of a common theme associated with the key concept, point to different sub-areas, but seldom ask if the key concept offers a good overall representation of the actual work being covered, sometimes only having the label in common (Alvesson and Sandberg, 2020).

Another way to support a hembig critique function could be to encourage authors to self-critically discuss the disadvantages of a given concept – what may be problematic with this concept, how it is generally applied and how it is used in the actual enterprise of a specific study or text. This will call for the active support and protection of editors, as specialized reviewers are unlikely to promote the questioning of labels central for their careers.

Discussion and conclusion

This article hopefully contributes to an increased awareness and perhaps deeper understanding of the problem with vague, all-embracing and omnipotent scientific concepts. By launching a new concept – hembig – we hope to stimulate thinking and reflexivity within the organization studies community around the use of scientific concepts. If scientific concepts are supposed to help us ‘effectively guide inquiry’ (Merton, 1984: 267), it is important that they are reasonably distinctive and do not cover everything and thereby nothing (Geertz, 1973); opportunistic or less well thought-through ‘jump-on-the-bandwagon’ tendencies need to problematized and mitigated. As emphasized, this is not only an intellectual issue, but a social and political one as well.

Our ambition is, of course, not to suggest a strict language regime, uniting all under a centralized authorized correct conceptualization. This is neither possible nor desirable. Sometimes loose, intuitive use of language may be difficult to avoid or even called for. Individual researchers engaged in specific projects may have good reasons to use what otherwise may be seen as hembigs in distinct and clear ways; as argued earlier, hembigs are not the concepts in themselves, but are created through the uses to which they are put. Words have no fixed, essential meaning and, as poststructuralists have taught us, we can never fully control the signifier. We can always expect some variation and fluidity in language use between – and also within – different sub-camps and authors. However, acknowledging the impossibility of fixing terminology once and for all does not prevent us from trying to specify meaning, unpack overloaded concepts and counteracting the hembig problem.

This does not deny the familiar problem of the steady flow of new concepts for phenomena that already have established and useful terms (and we have carefully thought of this issue when launching a new concept in this article), but this problem does not legitimate the widespread use of hembigs. Good academics should be able to deal with both problems.

Our concept hembig has some specifics and advantages when compared to partly similar expressions such as umbrella concepts, contested concepts, concept creep and knowledge brand, in that it emphasizes not only variation and expansion, but also the hegemonic impact and the ambiguity. The three components hegemonic, ambiguous and big refer to important characteristics and problems. They point to the combined effect of field domination with vagueness and a far too broad scope in language use. They link the socio-political and intellectual (cognitive-linguistic) problems of much contemporary research, where conformism and lack of distinctiveness go hand in hand. For space reasons we can only give brief indicators of how hembig is different from and adds to other terms: compared with umbrella concepts, it emphasizes in particular the power and scope aspect (while umbrella simply points to loose variation and diversity, an umbrella term becomes a hembig when it takes a strong grip over a community and the pressure to use it becomes hard to resist); with contested concepts, the understanding that there is often within the field a consensus and wide scope (many contested concepts become through hegemony only marginally contested and are successful in neglecting critique); and with concept creep, once again power, but also ambiguity (not only the expansion of the range of concepts) are highlighted. Compared with knowledge brands, hembigs are not (more than to a modest degree) centrally controlled, distinct and specific, but broadly emergent, have a large scope and are ambiguous. Hembigs show, we think, the full spectrum of problematic aspects of the phenomenon here discussed, partly and meritoriously addressed by other scholars.

So, a main contribution is the hembig concept itself, to the extent it is productively used – its capacity to evoke a particular imaginary and function as an antidote to contemporary problems. Is hembig a hembig itself, the sceptical reader may ask? Of course, it depends on its use, individually as well as collectively. The point of the concept is, however, not in labelling or offering a ‘truth’, but in inspiring critical reflection and providing an intellectual countermeasure when concepts have expanded and tend to be used in such a way that people have problems avoiding their ambiguity and power effects. The hembig issue calls for assessment in the same way as researchers need to think about and assess what is ‘good enough’ in terms of literature review, saturation of data, sufficient link to theory, clear writing and the strength of the contribution. In terms of hembig tendencies, the issue is to raise critique when the situation has become ‘bad enough’. Arguably this is the case within parts of contemporary management and organization studies.

A second contribution is the empirical investigation of the publication pattern of our three examples, and our revisions of Hirsch and Levin’s (1999) model of umbrella concepts. If we compare Hirsch and Levin (1999) with our findings 20 years later, the critique appears to have been severely weakened in the sense that the three concepts (even if there has been some critique, also in the journals we included in our study) in question are used more frequently and broadly than ever. Rather than development, expansion, critique and diminishing interest or re-labelling, we find development, expansion, marginalization of critique and a continued strong presence or even further expansion. What Hirsch and Levin label the ‘validity police’ and we call the hembig critique function has been seriously weakened and the dialectic dynamics thereby put out of play. We believe this is an effect of mass research, increased pressure for publications and a strong wish for return on investments, making most academics disinclined to support or accept radical critique of favoured key concepts; the hembigs function as important career enhancers. But also, less material and structural elements, such as identity projects – where researchers are into labelism, i.e. seeing the right concept as crucial for their work and themselves – play a significant role.

A third contribution is the analysis of problems following from the hembigs. This includes generally loose and vague thinking, social amnesia, and blocking of creativity and novel thinking through pulling people into ‘research boxes’. This creates difficulties for non-experts (and sometimes also experts) to orient themselves in an increasingly messy academic landscape, camouflaged under seemingly well-ordered categorizations.

A fourth contribution concerns remedies, including more careful work with specifications, unpacking, clarification of what is not included, and discussions of the disadvantages of each particular concept. We have also pointed to the need for a reinforced anti-hembig function. Critical and self-critical work is much needed in order to reduce the popularity of terms that are used to cover everything and nothing. This is a task for individual researchers but is also a responsibility of our organization studies community, our conferences and journals.