Theory,explanation,and understanding in management research

Abstract

Theory production has been a central focus of management research for decades, mostly because theory legitimizes both management research and, through its application, management practice as professional endeavors. However, such an emphasis on theory glosses over one of its constraining and particularized roles in scientific explanation, namely that theory codifies predictive knowledge. Committing to a ‘traditional’ or ‘critical’ understanding of theory thus amounts to embracing the view that prediction is achievable within a circumscribed field of study. Such an embrace is non-controversial in natural science. However, within the realm of management studies, it necessitates and smuggles in a strawman view of human existence, one which does not accommodate freedom and responsibility. This limitation of management theory explains its inadequate utility. This article argues that alternative avenues for management research exist.

JEL CLASSIFICATION: M10

Introduction

A common and long-established practice of leading management journals is that they require that authors make a theoretical contribution (Boer et al., 2015). Rabetino et al. (2020) note that such contributions are based on diverse ontological, epistemological, and methodological assumptions; embrace disparate conceptual approaches (behavioral, institutional, evolutionary, etc.); and seek to address differing organizational problems. Theoretical plurality also takes the form of development of new frameworks, testing of existing ones or extension of prevailing orthodoxy into new domains. Nonetheless, to reiterate, contribution to management theory (of one kind or another) is what editors of well-regarded scholarly journals typically expect from their prospective contributors. For example, the website of the Academy of Management Review states that the mission of the journal “is to publish theoretical insights that advance our understanding of management and organizations.” Similarly, as part of its stated objectives, the Strategic Management Journal “seeks to publish papers that . . . develop and/or test theory.” Another example is the Journal of Management, which (as of January 2021) is “committed to publishing scholarly empirical and theoretical research articles.”

The production of management theory has had two principal justifications. First, theory professionalizes researchers through creating a shared embrace of the scientific framework and commitment to disseminating findings using a common language. Second, theoretical knowledge acquired within the academy offsets the inexperience of novice managers and improves the practice of more seasoned ones (Van de Ven & Johnson, 2006). Scientific theory thus legitimizes both management research and practice as professional vocations (Shapira, 2011; Thomas & Wilson, 2011).

Despite its merits, the quest for management theory has not been without detractors. Indeed, three decades ago, Van Maanen (1989) lamented the ephemeral nature of the theories that management research produces. It is unclear whether substantial progress has been accomplished since such opining. For example, Academy of Management Journal editor George (2014) observed that the more management studies proliferate, the more they resemble “black cats in coal cellars” (p. 1). At least three of his predecessors held a similar view. Specifically, Bartunek et al. (2006) stated in one of their editorials that management research does not produce much, if anything, that is “really interesting” (p. 9).

A notable criticism of theory-driven management research is that it has little, if any, relevance for practice and education. This concern is traceable to Porter and McKibbin (1988), who stated point-blank that management research “does not produce practically relevant outcomes” (p. 30). Very recently, Joullié and Gould (in press) provided compelling evidence that the bulk of management research output cannot be applied in workplaces. Moreover, Pearce and Huang (2012) argued that management theories are unsatisfying to teach in that they seem disconnected from the realities and nuance of organizational life. More generally, Chia and Holt (2008) commented that “a preference for abstract causal explanation over practical knowledge, and for reason and truth over what works, has led to [management researchers] privileging . . . detached contemplation over involved action” (p. 473). Similarly, for Thomas and Wilson (2011), management research has become inward looking, producing theory for theory’s sake to boost business schools’ rankings and ensure their accreditation but with little concern for improving management practice. Perhaps, however, the problem is even more fundamental. Indeed, according to Kieser and Leiner (2009), the objective of improving practice was always illusory because output from management research is communicated in a way that is not reconcilable with the language of doing. Whatever the case, doubt about the value of theory-driven management research continues to plague scholarly commentary and there is evidence that the negative side (the skeptics) is winning (Joullié & Gould, in press; Kieser et al., 2015).

The purpose of this article is to reveal the nature of management theory and highlight the consequences of its creation. Specifically, in the following pages, it is argued that the academy’s lack of practical relevance stems mostly from a desire to explain management and organizational phenomena through exclusively proposing theories about such phenomena. Furthermore, irrespective of their ontological, epistemological, and methodological inclinations or the conceptual approach they follow, an obsession with theory blinds researchers to an insight long established in philosophy of science but not well ensconced within management literature. In a nutshell, the conundrum is that theory explains in a narrow sense but does not deliver understanding. Such difficulty arises because a scientific explanation is an inference from the general to the particular presented in terms of necessary or probabilistic relationships embedded in a theory, whereas understanding is arrived at through reflection about reasons and values in a way that marginalizes prescribed or probable outcomes. As such, considerations of individuals’ freedom, personal responsibility, and objectives are not compatible with scientific explanations. Accordingly, those who conceive of management theory as the sole foundation of management research should not have a monopoly within the academy. Indeed, alternative ways to propose meaningful and practical insights into management phenomena exist, namely, history and “poetry.” These approaches, as unorthodox and unsettling as they may seem, are adumbrated and defended in the later part of this article.

The focus of this article is the construct of scientific theory as instantiated in management studies, but not the process of theorizing. That is, this article offers a discussion about the underpinnings and consequences of the meaning of management theory, but not a comparative analysis of the “research paradigms” (or frameworks) that researchers use to produce it (although these latter matters create context for the argument presented and thus are touched upon). Specifically, the thesis starts from the observation that insofar as management research is concerned, two traditions have taken hold (Lincoln et al., 2005; March, 2007, pp. 166, 174–176; Rabetino et al., 2020, pp. 4, 22–24). First, research endeavors informed by positivist and postpositivist epistemologies have become, for the most part, indistinguishable because those working within these frameworks share a desire to propose explanations aimed at prediction and control of the phenomena they analyze. Second, scholars embracing interpretivist, constructivist, and critical studies paradigms have mostly set aside their initial differences and now view themselves as having in common that their mission is to create theories informed by subjective revelation. In the aftermath of such meetings of minds, management research has been mostly conducted either through using positivist (hereafter “traditional”) or postmodernist (hereafter “critical”) epistemologies, each broadly conceived to embrace minor and nuanced departures from orthodox protocols (Joullié, 2018; Shepherd & Shepherd, 2013).

The conception of theory which is focal in this article is that used within the scholarly frame of reference. As such, the analysis proposed in the following pages is not relevant to the term as it is applied loosely in everyday parlance to refer to conjecture or unsupported speculation. By way of context, this article will not describe or critique the theories that have been advanced in management and organization studies over the last decades (see Rabetino et al., 2020, for a classification of strategic management theories). Moreover, it will not provide commentary concerning the merits of the kind of data (quantitative or qualitative) and the level of analysis (micro, meso, or meta) used to generate management theories. Rather, for purposes of developing the current thesis and consistent with the view advanced by Hoskisson et al. (1999), it is assumed here that there is no consequential distinction between theory created from reflection on quantitative as opposed to qualitative data. Similarly, micro-, meso-, or meta-level contributions will be classified together as theory. Indeed, it will be argued that even if management theories differ in the nature of the data that support them, the level of analysis that led to their formulation, the conceptual approach along which they were developed, and the type of problems they are meant to address, they share a crucial characteristic. Specifically, theories are so classified because they prescribe, in one way or another, a stable relationship between variables (and thus constructs).

The structure of this article is as follows. Following the present introduction, first, the notion of theory in traditional management research is clarified using a discussion about the generic meaning of explanation in science. Salient conceptual and technical difficulties associated with the pursuit of theory in traditional management research are also presented. Second, noted challenges to theory and scientific explanation are explored through reference to exposition of the differences between natural and social science, completed with an examination of the construct of critical theory, which is often seen as (for reasons presented earlier) the alternative to traditional theory. This latter exercise will reveal that critical theory suffers from the same weakness that afflicts traditional theory. Third, non-theory-driven research avenues are highlighted. The article’s conclusion is that theory has consequential shortcomings that have placed practitioners and those in the academy at an unnecessary disadvantage.

Theory and traditional management research

Law, theory, and explanation

Since the last decades of the 19th century, philosophers have debated the degree to which science explains and its mode of so doing. A long-time dominant perspective on this matter was formalized by Hempel (1965). This view is known as the deductive-nomological model of scientific explanation for its deductive and lawful (inviolate principles) components (Hempel’s conception is also called the “covering law theory” or “received view”). Although Hempel’s model is no longer mainstream in philosophy of science, its continuing influence justifies a short exposition.

According to Hempel (1965), a scientific explanation is an account of a phenomenon consisting of two components (pp. 247–251). The first is the explanandum, a structured description of what is being explained. The second is the explanans, a series of statements which specifies particular conditions and includes one or more laws, from which the explanandum deductively follows as a conclusion. An example provided by Hempel (1965, p. 246) is the following: The level of the element mercury rises in a thermometer in proportion to the temperature of the liquid in which the thermometer is immersed (explanandum). This rise occurs because (1) mercury is contained in a glass tube (particular conditions, first part of the explanans), (2) bodies expand in proportion to their heat, and (3) mercury’s coefficient of expansion is greater than that of glass (two laws, second part of the explanans).

The deductive-nomological model of scientific explanation suffers from weaknesses the elucidation of which is beyond the scope of this article. Suffice to say that philosophers of science have proposed alternatives (for more details, see Salmon, 2006, or Godfrey-Smith, 2003). Two such substitutes dominate contemporary relevant literature: the so-called “causal” and “unificationist” modes of explanation (Godfrey-Smith, 2003, pp. 190–200). For proponents of the causal model (e.g., Salmon, 1998), to explain something in science is to provide detail about how that something is caused. In contrast, advocates of the unificationist perspective claim to approach the same problem through considering what scientists actually do, which (allegedly) is to develop explanatory schemes that can be widely applied. Unificationists thus hold that an explanation in science is an account that connects various facts and relationships by subsuming them under a set of general patterns and principles (Friedman, 1974; Kitcher, 1989). As such, unificationist science is best understood as an enterprise that strives to reduce the number of patterns and principles that must be accepted as fundamental.

Reconciling the causal and unificationist models of scientific explanation is not a necessary undertaking to advance this article’s argument. Rather, two focused comments suffice for the task at hand. First, as Godfrey-Smith (2003, pp. 196–197) noted, leading proponents of each perspective have converged in their analyses, each side conceding ground to the other. The reasons for this meeting of minds are somewhat intuitive. On the one hand, a causal process can be seen as the manifestation of a general principle operating within material reality; on the other hand, a general principle, if it is to explain material change, can be conceived of as causal in nature. The difference between the causal and unificationist models thus seems more a matter of presentation than of substance. Second and more consequentially, both models (causal and unificationist) use the same semantic structure to formulate a scientific explanation, a structure that is indebted to Hempel’s initial explanandum-explanans articulation. Namely, to explain something in science is to propose an account that goes from the general to the particular. Advanced on the basis of causal or unificationist logic, this shift rests on laws, be they universal (exception-less, that is, causal) or statistical (Friedman, 1974; Salmon, 1998; Woodward, 2017). ¹

In science, a law is a generalization that summarizes observations made under experimental, quasi-experimental, or fully natural conditions (Nagel, 1961, pp. 49–52; Thagard, 1992, pp. 225–227). A scientific law only reports on what is observed. For example, it is a law that ice floats on water. The converse (water floats on ice) is conceivable but has not been observed to be the case. In contrast, a scientific theory is an explanation for a law that has been successfully tested using accepted standards and protocols (an untested explanation for a law is hypothesis, conjecture, speculation, or proposition). For instance, Darwin’s theory of evolution by means of natural selection is an explanation for the law of evolution as evidenced by the fossil record (and other phenomena). Not all laws have been explained with theories. For example, no theory is available to explain why certain bird species form long-term dyads, whereas most do not, but conjecture exists.

Scientific theories explain all pertinent laws and more specific theories. For example, knowledge of fundamental physical theories affords the scientist a head start when seeking to explain fluid dynamics (Woodward, 2017). Like Russian dolls, scientific explanations are thus inclined to be cascading in nature. In all cases, however, a theory consists of statements about an aspect of the world (the laws subsumed by the theory itself) from which logical consequences derive (Bogen, 2017). These consequences can be necessary, merely probable or even comparative (as in “smoking increases the probability of developing lung cancer”—implicitly offering a comparison with not smoking), but in each case a scientific theory codifies predictive knowledge. ² In so doing, it allows for a measure of control or, in instances such as meteorology and astronomy, at least permits (probabilistic) prediction of the phenomena that the law describes and the theory explains. Indeed, to resurrect an adage, it is because theory captures predictive knowledge about a feature of reality that “there is nothing so practical as a good theory” (Lewin, 1952, p. 169). It is noteworthy that this predictive feature, at least according to orthodoxy, characterizes both social and natural science theories (Coleman, 2007; Taagepera, 2008).

The characteristic elements of scientific explanation were implicit in sociology and psychology even before philosophers of science attempted to formalize them. For example, Durkheim (1893/2014) sought to explain how societies maintain internal cohesion using such a schema. His theory is that social order arises because of the presence of either of two possible forms of solidarity, namely, mechanical for primitive societies or organic for industrialized ones. In psychology, Freud (1900/2005) proposed a theoretical system to explain adult behavior as an archetypical response caused by subconscious memories of childhood trauma. Later, Skinner (1953/2014), with his conception of operant conditioning, rejected the Freudian account and held that behavior is shaped in the course of interaction with an organism’s ambient stimuli. In each of these disparate cases, the phenomena of analytic interest are explained through proposing a theory. Management theorists have proffered equivalent contributions. An example from this latter domain which adheres to the typical aforementioned semantic structure of scientific explanation is as follows:

Sales representatives performed well last month because working conditions are adequate and financial incentives were attractive and because, according to Herzberg’s two-factor theory (Sachau, 2007), employees go beyond compliance when both “hygiene” (i.e., relating to baseline working conditions) and motivating factors (i.e., relating to commitment, dedication and pursuit of excellence) are in play.

Theory and explanation in management

In management (including strategic management) literature, inquiry into the nature of theory and theory building typically analyzes the epistemological and methodological assumptions that underpin research (Rabetino et al., 2020). Authors who contribute to such discussions reveal how they conceive of the process of management research as it is, or should be, practiced. In other words, those who speculate about management theory formulation propose meta-theoretical views, with an implicit or explicit acknowledgment that one unifying research paradigm does not exist (Calori, 1998). Examples of such management meta-theories abound: Hassard and Cox (2013), Kilduff et al. (2011), Lakomski and Evers (2011), Corley and Gioia (2011), Boisot and McKelvey (2010), Clegg & Ross-Smith (2003), Johnson and Duberley (2000), Deetz (1996), or again Gioia and Pitre (1990).

In management studies, discussion about meta-theory traces its origin to the work of Burrell and Morgan (1979), which heralds the starting point of the “paradigm wars” of the 1980s and 1990s (Bryman, 2008; Shepherd & Shepherd, 2013). In the meta-theoretical battles of this era, protagonists disagreed on the nature of organizational reality and on how best to study that reality. However, in spite of these controversies, a consensus exists among prominent (and to date mainstream) management meta-theorists concerning a central issue: Explanation in management studies follows the causal model of scientific explanation. Indeed, these meta-theorists hold that causal chains are present in organizational reality and a description of their modus operandi is a central element of what management theory is about. For instance, Sutton and Staw (1995) wrote, “theory emphasizes the nature of causal relationships, identifying what comes first as well as the timing of such events” (p. 378). Similarly, all four meta-theoretical models identified by Kilduff et al. (2011) “posit causal relationships” (p. 298). Shapira (2011, p. 1313), Davis (2015, p. 183), Colquitt and Zapata-Phelan (2007, p. 1281), Van de Ven and Johnson (2006, p. 806), and again Whetten (1989, p. 491) are broadly in agreement, each asserting or arguing in their own way that the objective of theoretical knowledge in management studies is the identification of causal chains to formulate predictions (the position of non-mainstream meta-theorists is discussed later). ³

In sum, for mainstream meta-theorists, management theories are statements that disclose an aspect of organizational reality, the knowledge of which enables prediction and possibly control of future work–related events through positing causal interactions, relationships, processes, or structures. While management meta-theorists have challenged Burrell and Morgan’s (1979) analysis of sociological research and disagreed about which paradigm is best suited for management research, the embrace of the causal model of scientific explanation in mainstream management studies has remained largely uncontested.

Explaining management laws

The world of managers and the resources they oversee offers an abundance of laws (in the descriptive sense of observed regularities). For instance, most employees arrive to and leave from a place of engagement at set times. When called to meetings, they typically attend them. Beyond such basic regularities, some employees consistently exceed compliance-related expectations. Managers are inclined to reward such conscientious workers and discipline or otherwise admonish their peers who, often in a somewhat relative sense, are identified as underperforming. Customers satisfied with the offerings provided by a firm tend to do repeat business with that entity. Moreover, they may recommend it. Executives generally look for and discuss opportunities to increase return on capital deployed under their stewardship. Behavioral laws of such kind are pervasive in the world of commerce and within workplaces. As noted, mainstream management and organizational researchers attempt to account for them by way of theories.

The domination of causal theory production in management research has an illustrious history that commenced (at least in the modern era) with the archetypal scientific theoretician, Frederick Taylor. Indeed, in insisting (and providing formulaic rationale) that there is “one best way” to do production-related tasks, the pioneering management scientist was advancing a theory of management, in his case of job design. However, despite Taylor’s enduring influence, to date no management theory has been advanced that can be implemented with the same reliability as those addressing the physical world’s interrelationships. This shortcoming represents something of an elephant in the room for the custodians of management studies. As Barker (2010) notes, the absence of a body of knowledge leading to predictable and reliable results undermines the efforts of commentators who argue that management is a profession.

As a rejoinder to Barker’s (2010) criticism, management researchers are justified in replying that the lack of progress in identifying domain-specific theory that can be impressively implemented arises principally because of validation difficulties. Such problems have been comprehensively documented (e.g., Crowther & Lancaster, 2008; Johnson & Duberley, 2000; Nagel, 1961, pp. 447–502; Starbuck, 2006). One is particularly noteworthy because, although it manifests essentially as a technical matter for those caught up in the machinations of quantitative analysis, it starkly exposes a crucial conceptual distinction between social and natural science. Specifically, unlike events observed in the physical world, organizational phenomena are influenced by an indeterminate number of variables. ⁴ Constructs of interest to social science researchers, although in principle often amenable to experimental control, are not typically investigated using the same laboratory protocols that are associated with the natural sciences, mostly for practical and ethical reasons. Furthermore, even in the case of field or natural experiments on work-related phenomena, delineating experimental and control groups is frequently not straightforward (Davis, 2015, p. 183).

Techniques exist in social sciences that are intended to compensate for the difficulties associated with being unable to conduct laboratory experiments. Most of these techniques attempt to offset problems arising as a consequence of not having a robust control group. They include manipulations such as statistical control (random sampling being the most widespread instantiation) and use of proto-control groups à la field or natural experiments. However, as Andreski (1969) long observed, unambiguous explanations invoking cause–effect relationships are not realistic in domains such as sociology and human-orientated psychology (pp. 47–59). Indeed, within the social sciences, conclusions based on experimentation protocols are beset by the possibility of competing explanation, arising owing to compromise of the principle of laboratory control. A generic case in point concerns the changing intentions of individuals, the behavior of whom is being studied. A manifestation of this problem is the recent “replication crisis,” wherein social science (including business studies) researchers have had difficulty reproducing findings of published research (Aguinis et al., 2017; Yong, 2016).

In summary, the main difficulties that management researchers face in conducting their studies, formulating their theories, and validating their results stem from a common origin. Specifically, malaise emerges from a desire to account for observed organizational regularities (laws) as natural scientists do, by way of necessary (typically causal) mechanisms: theory. However, looking for deterministic accounts of human interaction de facto assumes that these accounts exist. It is thus unsurprising that some scholars have taken issue with such an approach.

Challenges to theory and scientific explanation

Natural and social science

Scholarly interest in scientific explanation is understandable in light of 19th- and early 20th-century scientists’ purported ability to explain all phenomena, including psychological and social ones (Gould et al., 2017; Salmon, 2006, pp. 11–25). Not everyone was so easily convinced, however. For example, German sociologist and psychologist Wilhelm Dilthey (1833–1911) held that science would not account for the totality of human experience. Indeed, Dilthey (1883/1989, p. 58) pressed that whereas nature is characterized by “objective necessity” and causal continuity, “man finds within his self-consciousness a sovereignty of the will, a responsibility for actions, a capacity for subjecting everything to thought and for resisting from within the stronghold of his personal freedom any and every encroachment.” While “all science is experiential,” the “facts of consciousness” (Dilthey, 1883/1989, p. 50) such as willing, feeling, and thinking are not knowable through means of natural science. These phenomena have a different epistemological status because they are only accessible through reflective analysis, introspection, and exegesis of the records of human existence. In his late works, Dilthey (1900/1996) clubbed such methodological orientations under the umbrella term “hermeneutics.”

The difficulty inherent in studying human inner life generally and consciousness particularly through the same means that scientists use to study nature led Dilthey (1883/1989) to distinguish between Naturwissenschaften (the natural sciences) and Geisteswissenschaften (literally, the sciences of the mind or spirit), generally translated in English as the human or social sciences. In Dilthey’s view, the task of natural science is to propose cause-and-effect-based explanations of phenomena. As the notion of causality renders the construct of “choice” meaningless, such a kind of explanation is acceptable within natural sciences where researchers inquire into behavior that is amoral in nature and not associated with human-type agency. By contrast, the mission of the social sciences is understanding lived experience. For Dilthey (1883/1989), this undertaking entails identifying and articulating aspects of social and cultural life that highlight intrinsic reasons for, and provide meaning to, human action and relationships (pp. 72–79). In Dilthey’s conceptualization, understanding requires an analysis of the ideals, values, and norms that manifest in human choice. Such an approach marginalizes deterministic interactions and prescribed outcomes, which are what theories embed.

It is noteworthy that Dilthey (1883/1989, pp. 393–399; 1900/1996, pp. 263–266) did not deny the existence of structures governing human thought and socio-historical reality, phenomena that he sometimes referred to as “forces” or “laws.” For Dilthey, however, such structures do not determine human behavior. In his view, social scientists are less concerned with what people do individually or collectively, but more with what they should do. If individuals exhibit behavioral continuity, if they behave in regular and even predictable ways, it is because they have ultimately chosen to do so. Hence, the structures uncovered by the social scientist govern human reality only insofar as they are normative. Behavioral patterns emerge exclusively because humans, as an act of volition, conform to precedents that they have established for themselves purposefully, for idiosyncratic reasons or because they seek to comply with group norms. Human action therefore cannot be reconciled with cause-and-effect-type logic but rather is the product of choice, mostly made in constrained circumstances (Dilthey, 1883/1989, pp. 453–455).

Dilthey’s analysis of natural and social science has been influential (Cohen, 1994, pp. xxxi–xxxii). For example, his dichotomy between scientific explanation and understanding was not lost on Max Weber (1920/2012), who tried to combine both approaches when developing interpretive sociology (Giddens, 1982; Khurana, 2007, pp. 394–395). For Weber, the interpretive sociologist’s job is to reconstruct the meaning of social events from the perspective of those who live them, rather than exclusively through the detached and objective standpoint of the natural scientist (in organization studies, this reconstructive stance has notably been adopted in Weick, 1995). That is, the Weberian sociologist is to account for the actions of other human beings not as manifesting impersonal causal processes but by attributing to them the feelings and thoughts which he would have if he carried out the same actions. However, Weber acknowledged the merit of scientific explanation where social and historical phenomena have no known origin or when actions have unintended consequences. As he put it, in these circumstances, “the interpretive understanding of social action” is to arrive “at a causal explanation of its course and effects” (Weber, 1920/2012, p. 88).

Critical theory

Dilthey’s contention that human endeavor and relationships should be received in light of their social and historical context is also noticeable in Horkheimer’s (1937/2002, pp. 10–17) work and in particular his critique of theory. For Horkheimer (1937/2002), theory as it is understood in natural science (“traditional theory” in his parlance) follows unificationist logic (pp. 188–190). Indeed, in Horkheimer’s view, theory consists in propositions about a subject of inquiry from which predictions derive. To the extent that such predictions are compatible with facts, theory is validated. Furthermore, the “smaller the number of primary principles in comparison with the derivations, the more perfect the theory” (Horkheimer, 1937/2002, p. 188).

While recognizing the merit and successes of natural science, Horkheimer (1937/2002) argued that its underlying conception of theory is inadequate and in fact counter-productive in social science (pp. 190–196). This inadequacy arises because, for Horkheimer (1937/2002), traditional theorists deliberately study reality only as it is given in experience, in isolation from broader moral, social, or historical considerations (pp. 133–134). Thus, even when they are social scientists, traditional theorists do not appreciate that their work is morally, socially, and historically contingent. In this sense, the knowledge they produce is not only superficial but also incomplete. Indeed, because they ignore their work’s social (moral and historical) underpinnings, traditional social scientists are ignorant of its social (moral and historical) consequences.

Oblivious to their work’s historical, social, and moral context, traditional theorists do not recognize that the problems they address have only come to be identified as such in particular circumstances. In other words, traditional scientists do not realize that they work only insofar as society legitimates their professional activity. It follows that traditional theory, by way of the scientific knowledge it produces and despite claims of neutrality and objectivity, supports rather than challenges the established social order. Moreover, since the triumph of the Enlightenment, that social order is for Horkheimer (1937/2002) quintessentially bourgeois: It oppresses the laboring masses to the benefit of a capitalist elite and certain of its more privileged servants (pp. 197–207). Willingly or not, knowingly or not, traditional scientists, including social scientists, belong to the latter of these categories.

Horkheimer (1937/2002) insisted that the purpose of social science is to remake society as more humane and just (pp. 208–210). To achieve such an objective, that is, to liberate the masses from bourgeois oppression and capitalist exploitation, he averred that a new kind of theory is required in social science: critical theory. Critical theorists differ from traditional theorists (Horkheimer, 1937/2002) in their conviction that social reality is not a given but the product of human activity and interpersonal relationships (pp. 210–215). They approach their subject matter with an underlying motive: to expose the subordination of traditional science to bourgeois society. Indeed, unlike traditional theory, critical theory does not pretend to be neutral or objective. The critical theorist is aware of his own partiality and superordinate objective of changing history through a process of critique. Such a stance does not imply superficiality or one-sidedness; on the contrary and again in contrast to traditional theory, Horkheimer (1937/2002) maintained that critical theory reconciles subject with object or man with society to produce humane knowledge as opposed to knowledge per se (pp. 236–243). ⁵ Irrespective of their merits (some of which are examined later), Horkheimer’s arguments have had a distinguished legacy in social science (Berendzen, 2017), including organization studies (Steffy & Grimes, 1986). In management studies, their influence is most visible in the work of critical management studies scholars.

Although mostly associated with the seminal collection of essays edited by Alvesson and Willmott (1992), the research agenda of critical management studies has splintered into several lines of inquiry (Adler et al., 2007; Fournier & Grey, 2000). Notwithstanding this splintering, with varying emphases, critical management authors typically consider that individuals in general and employees in particular are prevented from attaining their full potential by social and bureaucratic structures. Such constraining elements are found, for example, in government agencies and the private sector, including hospitals, prisons, universities, and corporations. Institutions of these kinds are thus perceived of as vehicles of mass oppression operating under the guise of welfare provision, health preservation, crime prevention, higher education dispensation, and economic development. Business schools are notoriously at the receiving end of this indictment in that they are charged to purvey corporate interests, managerialist assumptions, and neoliberal ideologies under the guise of common economic sense and good managerial practice (Alvesson & Willmott, 2003; Bowden, 2018, pp. 208–235; Grey & Willmott, 2005).

For critical management studies scholars, not only is knowledge power, but also, as Foucault (1995) pointed out, power itself is knowledge because it supports mainly the production of the kind of expertise that consolidates social control. Such a view renders logic, rationality, and the scientific enterprise suspect, in that it posits that oppression underlies the production of conclusions purported to be universal, ahistorical, and unquestionable. The objectivity claimed by scientific research (particularly when undertaken in the social sciences) is therefore revealed as a delusion, exposed as such through a rigorous hermeneutics of suspicion. Of more pragmatic concern, the research conclusions of scientists are better thought of as ideologically driven distortions of reality serving institutional interests. However, emancipation, although difficult, remains possible. It requires analysis of social structures and managerial as well as scientific discourse which explicitly exposes oppressive agendas (Fournier & Grey, 2000; Grey & Willmott, 2005). Or so it seems.

Theory and critical management studies

The research agenda of critical management scholars is potentially seductive for researchers wary of mainstream research. Indeed, it is a small step from believing that traditional (theory-driven) management research rests on and conveys a socially harmful ideology to believing that the two central pillars of scientific research that theory embeds (objective observation and ensuing logical reasoning) should be abandoned. However, a closer inspection of critical management authors’ assumptions and objectives reveals that the kind of research they advocate cannot in fact make good on its stated promise. This inability arises from weaknesses inherent in critical theory, exacerbated when such theory is being applied by critical management advocates.

The current authors will not question whether Horkheimer’s confidence that adopting critical theory will help deliver a more just and humane society is reasonable (or even whether Horkheimer’s envisioned society really is more just and humane). More relevant to this article is to evaluate the degree to which critical theory differs from the general construct of theory described earlier, that which Horkheimer labeled “traditional.” Perhaps surprisingly, scrutiny of Horkheimer’s arguments reveals that critical theory, as the name in fact plainly suggests, is still theory. Indeed, if critical theory differs from traditional theory for being historically situated and directed to a particular social end, it shares with traditional theory crucial characteristics. Specifically, critical theory (like some traditional theory) uses unificationist logic ostensibly to provide explanatory knowledge through embrace of the same language forms (in particular its semantic structure) when offering predictions.

Critical theory follows unificationist logic because it rests on a general explanatory principle. This general principle is apparent in Horkheimer’s (1937/2002) view that “the discrepancy between fact and theory [must be overcome because behind them] lies a deeper unity, namely the general subjectivity upon which individual knowledge depends” (p. 203). From this “deeper unity” or ultimate explanatory relationship derives Horkheimer’s general contention that traditional theoretical knowledge is not objective but rather is contingent on historical and social events and his more specific view that traditional theory serves the interests of bourgeois society.

Horkheimer (1937/2002) acknowledged that social science based on traditional theory delivers knowledge of society and, as such, is sometimes useful (p. 197). For example, in the arena of management studies and industrial sociology, social science as traditionally conceived has revealed that division of labor improves productive efficiency in factory contexts (Horkheimer, 1937/2002, p. 216). He (Horkheimer, 1937/2002) insisted, however, that traditional social science is flawed in that its methods do not recognize that “reality is itself the product of a society’s work” (p. 203). As such, knowledge derived from traditional theory contributes to maintaining existing hierarchical power structures in general and exploitation of the masses by the capitalist elite in particular. By contrast, the objective of critical theory is to deliver knowledge that will drive societal change. Whatever the case, it is noteworthy that explanatory knowledge of society is an objective pursued by both critical and traditional theory.

In summary, as Horkheimer (1937/2002) himself admitted, if the works of the traditional and critical theorists differ in their specific interests, they “manifest the same logical form” (p. 216). Indeed, from the perspective of unificationist logic, they are indistinguishable. In other words, management scholars looking for an alternative to (traditional) theory, its particular mode of scientific explanation and its accompanying deterministic portrayal of human agency will not find solace in Horkheimer’s critical theory, irrespective of the merit of the research program such theory makes possible. Similarly, the work of critical management scholars offers little to the management researcher wary of the limitations of scientific explanation. There are two main reasons for such a letdown.

First, critical management authors, despite their wariness of science, still embrace its central objective. As Alvesson and Willmott (2003) wrote in their review of the body of scholarship they initiated, mainstream management research promotes the view that

knowledge of management [is] knowledge for management, [whereas] critical perspectives on management share the aim of developing a less managerially partisan position. Insights drawn from traditions of critical social science are applied to rethink and develop the theory and practice of management. (p. 1)

What Alvesson and Willmott are saying here is that although the intentions of critical management authors are at odds with those of the orthodox researchers they criticize, critical authors, just as much as traditional ones, pursue professional legitimation through improved managerial practice by way of theory (on this point, see also Spicer et al., 2009, and Adler et al., 2007, p. 2). Understandably so, because adopting a purely critical, non-performative stance would ultimately condemn critical management authors to solipsism and thus irrelevance. As Parker and Parker (2017) noted, in critical management studies, “critique has always been in tension with a desire for influence” (p. 1367). The tension to which they refer was first highlighted by Habermas (1998) with regard to critical theory in general and remains unresolved (pp. 126–130). In fact, it has recently intensified (King & Land, 2018; Parker and Parker, 2017; Spicer et al., 2009, 2016). A more pressing dilemma is that insofar as they embrace pursuit of theory and the legitimation (notably by way of contribution to managerial practice) it affords, critical management authors do not propose a counter-perspective to mainstream management studies.

Second, critical management studies cannot deliver definitive outcomes, be they in the form of emancipation or theory. Indeed, if social structures are by definition alienating as the critical management community purports them to be, then so are critical management studies, because this body of research, with its dedicated journals, conferences, and institutional grants, has created for itself a social structure. Besides, if rationality is misleading as critical management authors propose, then rational arguments (including defense of the proposition that rationality is misleading) lose their potency—or do they? On these matters, critical management advocates appear to lack the reflexivity they trumpet as being one of the strengths of their epistemology (cf. Fournier & Grey, 2000, p. 282).

The aforementioned comments imply that in the absence of knowledge of ultimate truth (or ways to obtain it), definitive theoretical conclusions or practical courses of action cannot be proposed. Critical management inquiry will only reveal that underneath every alienating institution or managerial layer lies another which is equally estranging. As such, theoretical abyss and practical paralysis, not emancipation, result from critical management research (a sustained discussion on this theme is available in Joullié & Spillane, 2020; see also Bowden, 2018, pp. 250–252 and Donaldson, 2003). In the final analysis, critical management authors play the role of their academic orthodoxy’s good conscience. They speak from self-claimed higher moral and epistemological grounds but their embraced epistemology denudes them of the ability to influence orthodox management research and managerial practice.

Implications for management research

In a widely quoted essay, Ghoshal (2005) pressed that management researchers do little else but convey liberalism dressed up as “pretense of knowledge” (p. 77). In Ghoshal’s view, such pretense arises because researchers adopt the approach of natural science when they study organizational reality. In so doing, they develop a sort of physics in which “causal and testable theories” (Ghoshal, 2005, p. 86) operate on employees (and others) considered as non-sentient input in the process of productive transformation. Such an impoverished view of organizational reality, insisted Ghoshal (2005, pp. 88–89), is at its core ideological and socially harmful in that it has a tendency to be self-fulfilling.

To move from pretense to substance, Ghoshal (2005) urged management researchers to recognize that no single idea captures social complexity and in particular that human beings do not exclusively pursue personal gain, as alleged in much current management scholarship (pp. 86–87). He further held that management research requires a unifying framework able to accommodate simultaneously different views of human nature. As a first step on the journey toward developing such a framework, Ghoshal (2005, pp. 88–89) called for the embrace of intellectual pluralism in the academy (a call echoed by Morell & Learmonth, 2015). Within such a pluralism, multiple perspectives sit beside each other with equanimity and the currently dominating research framework is downgraded. However, and despite his open-minded inclinations, Ghoshal (2005) did not give up on theory (p. 87). In fact, he encouraged management researchers to produce more of it. He apparently did not appreciate all the implications of this emphasis, one of these being that a single-minded quest for theory puts researchers on the path to natural science and its particular mode of explanation.

Management researchers seeking a substitute to their theory-driven orthodoxy but suspicious of critical management studies have refuge in at least one alternative. Indeed, they need only return to Dilthey’s core arguments. Specifically, rather than looking for scientific explanation for management laws, scholars have the option to attempt to understand such phenomena. That is, if researchers want to account for the regularities they observe within workplaces, then theory production is not an optimal strategy because, as explained, the accounts that theory generates leave little scope for, or at least marginalize, the roles played by reason, choice, and values.

Since antiquity, humans have developed pictures of themselves in which they choose and consequently attract merit or blame for their actions. In particular, people have expended energy in maintaining systems of thought, like religion, to guide them in making sense of their existence. Modern justice systems also assume, as their most elementary characteristic, that individuals are responsible for their actions, insofar as they understood and could have reasonably predicted those actions’ consequences before undertaking them. It is also noteworthy that at least insofar as U.S. Federal Policy for the Protection of Human Subjects (the so-called “Common Rule”; U.S. Department of Health & Human Services, 2019) is concerned, human subjects should be dealt with by researchers as autonomous and responsible (i.e., freely choosing) individuals.

A defining feature of human societies is that they assume that behavior is chosen and as such attracts responsibility. On this basis, rather than trying to discover causal relationships underlying the phenomena they observe, management researchers have the option to understand and explain deliberate actions, situational choices, ambiguities, and constraints. If they take this latter path, scholars would study why managers have made the decisions they have, but, crucially, embed such analyses in a context of reasons, values, biases, and idiosyncrasies. So-inclined researchers would stop acting as natural scientists and abandon the claim to predictive knowledge. Rather, they would conduct themselves as historians because their analyses will be after-the-fact reconstructions of events, situations, decisions, and intentions. As far as the business and activities of humans are concerned, the past helps understand the present and this understanding is the principal guide for future actions. As Dilthey (1900/1996) pressed, in the social sciences, historical experience and its retrospective analysis, not controlled experiments, are the main source of knowledge. The existence of such publications as Business History and the Journal of Management History and the inclusion of history in organization studies, pioneered by Kieser (1994) and to which the Academy of Management Review dedicated a special issue (Godfrey et al., 2016), rest on this contention, broadly understood.

In Poetics, Aristotle proposed that metaphor, alliteration, and imagination-fuelled speculation are superior to history because the former techniques invoke the universal and as such embrace that which could be true, whereas the latter merely records the particular, what is true. If Aristotle’s insight is accepted, then, to complement their historical analyses, management scholars looking for an alternative to theory-driven research should engage, for lack of a better word, in “poetry.” That is, rather than trying to identify predictive theories leading to certain results, management authors should start from the premise that human agency is not constrained in a deterministic sense and that the future is thus best conceived of as a smorgasbord of possible scenarios. Such scholars would explore what could have happened if different values had been held, other objectives pursued, other ambiguities faced, other language used, other emotions felt, other problems resolved, and other decisions made. Although not in these words, this view of organizational research as “social poetics” or as “situational dialogical action research” has been notably defended by Schön (1983), Cunliffe (2002), and Shotter (2010).

Readers smiling at the idea that management scholarship should be a form of “poetry” should pause and recognize that even in organizational contexts, ambiguity is the norm and decisions have always an emotional dimension. Furthermore, if they want to guide managers in replicating the successes of Apple’s Steve Jobs or Facebook’s Mark Zuckerberg, management scholars cannot only be scientists and content themselves with understanding the world as it is or seek to identify “best practice,” the replication of which may buy time but spells extinction when (not if) market conditions change. Rather, they need also to emphasize their imagination. Indeed, focusing on what there is, as theorists do, comes at the expense of missing out on what there is not yet, on what there could be. Management scholars have thus the choice of proposing new ideas, keeping in mind that if they want to help practitioners make sense of the present and its ambiguities, they will seek to reconcile the empirical with the imaginative, the rational with the emotional, and the technical with the moral. This coming to terms with the necessary and the possible is not a scientific endeavor; it has been variously called “good judgement” (Tichy & Bennis, 2007; Vickers, 1983), “practical intelligence” (Sternberg, 2000), or again (after Aristotelian phronesis) “phronetic leadership” (Shotter & Tsoukas, 2014; see also Antonacopoulou, 2010).

The body of knowledge that will emerge from management history and “poetry” will not qualify as a science with its scaffolding of predictive theories buttressed by causal mechanisms. Rather, it will consist of a repository of actual and potential management practices underpinned by psychological and moral stances. However, such a reconceived management scholarship will remain faithful to the requirement to generate practically relevant knowledge. It will achieve it not by claiming it has discovered theories of management, but by educating its students about past managerial decisions grounded in the study of recorded victories, defeats, and missed opportunities. The critical evaluation of the past brings about awareness of alternative futures: Innovation and responsible decision-making have no other possible foundation. Management scholars convinced by these principles will also remind their students of the fleeting nature of successes and the attractive if elusive character of possibilities. Such a reminder will be a lesson in skepticism and in humility directed to those who aspire to business leadership—one that can do no harm.

Conclusion

Management research is predicated on the assumption that there exists within workplaces and their environment phenomena (concerning structures, forms, processes, capabilities, and practices) that are stable enough to be discovered, analyzed, and formalized. Researchers are expected to reveal and codify such a body of knowledge, educators to teach it, and managers to apply it. Commencing with Taylor and his ilk, the building block of this corpus has been, or at least has been alleged to be, management theory.

In science, laws summarize observations and theories explain laws. More precisely, scientific theories formalize, by way of statements from which logical consequences derive, the processes or structures that operate within reality as described by laws. Knowledge of theory allows for prediction and possibly control of the phenomena that the theory explains and that the law summarizes. These elements characterize conventional models of scientific explanation.

For decades, it has been standard practice that researchers seeking to be published in leading management journals make a theoretical contribution. In support of this agenda, meta-theorists have debated ways to conduct the sort of research that produces management theory. In so doing, they have missed opportunities to critique prevailing orthodoxy. However, over the same period, there have been detractors who have pointed out in increasingly strident terms that management research has limited practical relevance (e.g., Joullié & Gould, in press). As with meta-theorizing, however, the view that the raison d’être of management research is to produce theory has remained unchallenged. Even critical management authors, who aspire to disrupt orthodox research and remedy its purported alienating effects, have espoused its agenda insofar as they have sought to contribute to theory. Thus, for all involved, be they traditional or critical management scholars, the pursuit of theory has been the central preoccupation.

This article has argued that a single-minded quest for management theory does not serve researchers’ interests. Indeed, an obsession with theory undermines management as a research field because, in science generally and management studies particularly, appeals to theory involve causes. The language of causality, however, does not accommodate reflection on reasons and objectives. Hence, in conventional parlance, as well as in more technical applications, to be caused to do something is to have no reason to do it, but rather be forced to. Exclusively pursuing theory production in management studies requires that the world of managers and workplaces is reconcilable with natural science and describable using a deterministic language of forces, stimuli, reactions, pressures, causes, and effects. However, the language of natural science cannot give voice to constructs such as morality, imagination, choice, purpose, creativity, freedom, and responsibility, for centuries considered characteristic hallmarks of human agency. In any case, if valid, the view of humanity implied by natural science’s vocabulary (i.e., the language used to present a research question and describe a methodology for answering it) applies not only to an established object of study, but also to management researchers themselves, perhaps unbeknownst to them.

In certain respects, work settings are microcosms of human societies. For example, they typically reproduce, in simplified form, the dilemmas that individuals confront in other contexts (Alvesson, 2002). Hence, ontological assumptions in which workplaces are populated with individuals seen as bereft of choice, self-selected purpose and other distinctively human forms of agency represent questionable research foundations. For reasons presented earlier, management theory smuggles in such assumptions. Furthermore, attempting to validate a management theory as part of a research project de facto contradicts the basic ethical principles of the Common Rule, as it requires assuming that the individuals, whose behavior is being compared with what the theory predicts, are denuded of autonomy, decision-making capacity, and, ultimately, responsibility. In this respect, techniques resting on a vision of human existence that contradicts how managers and their subordinates understand themselves are destined to be either ignored or, when applied, typically yield underwhelming results. The limited practical relevance of theory-driven management research is a visible manifestation of this problem.

After surveying the difficulties that management researchers were facing three decades ago, Van Maanen (1989) called for a 10-year moratorium on theory generation (p. 32). The current authors go further and challenge management studies’ theory-driven research agenda. In so doing, they resurrect Gellner’s (1986) argument that if science is characterized by its ability to generate consensual, cumulative knowledge capable of improving human existence by way of predictions, the so-called “social sciences” (to which management studies belong) are not scientific (pp. 126–127). Winch (1958/1990) did not say otherwise when he argued that social science is better approached not as science but as a branch of philosophy. Indeed, according to Winch, explanations of human behavior are not of the same order as (and are not reducible to) the kind of explanations produced by natural scientists.

The expression “physics envy,” which captures sarcastically the established central role of theory in management research, should be received, if not as an alarm bell, at least as a wake-up call. Kieser and Leiner (2009) were justified in arguing that the gap between contemporary management research and practice is unbridgeable because the language of management research is at least partly incompatible with that of management practice. Thus, as long as researchers attempt to formulate answers to management problems in the (scientific, deterministic) language of their academy’s current orthodoxy, they will fall short on producing practically relevant output.

To the extent that language is the principal means of communication among researchers and between researchers and the individuals the behavior of whom they study, management research has a crucial linguistic component. Critical management authors are thus justified in directing attention to the constraining and possibly alienating role of language. The construct of management theory is here a case in point: Such theory embodies, indeed requires, a deterministic picture of human existence that is typically unacknowledged, presumably because it is unrecognized. In management research as in other aspects of life, words and phraseology convey a vision of human nature and social reality. In this sense, management research requires appreciation of the spectrum of conventional usages of the terms (and their consequences) to which one subjects others and to which one is subjected. Such an observation inspires the researcher to think about the possibilities afforded by history and poetry to the enterprise of inquiry into management and work-related phenomena. Indeed, the observation lays bare several of the unacknowledged flaws of theory (traditional and critical) when applied to management research.