Embracing essentialism: A realist critique of resistance to discursive power

Abstract

This article argues that, contrary to its detractors, essentialism is a necessary concept in understanding resistance to managerial discourses. The article first suggests that essentialism, under a critical realist framing, need not suffer from the reductionism or determinism found in many 19th and 20th Century essentialized accounts of the self, arguing instead that the concept adds analytical power to explanatory theorizing. Next, taking three common post-Foundational presentations of resistance to managerialist discourses, the article proposes that, despite protestations to the contrary, each relies on essentialist representations of both discourse and the self. The article then seeks to tackle the ‘problem’ of essentialism head-on by showing its potential for both framing resistance and building bridges between the post-Foundational, realist and natural worlds.

Post-foundational anti-essentialism

Whilst a strong and deterministic essentialism is always wrong and often dangerously misleading, a moderate, non-deterministic essentialism is necessary for explanation and for a social science that claims to be critical. (Sayer, 1997: 453)

Since Newton’s (1998) persuasive argument that neo-Foucauldian organization studies had failed to avoid essentialist positions, the post-foundational movement has made significant theoretical gains in seeking to explain, not only how normative forms of control impact the identities of employees (Holmer-Nadesan, 1996; Meriläinen et al., 2004; Musson and Duberley, 2007; Thomas and Linstead 2002), but also how individuals resist such processes (Ainsworth and Hardy, 2004; Roberts, 2005; Sveningsson and Alvesson, 2003; Thomas and Davies, 2005a; Whittle, 2005). In these accounts a number of complex and inter-related arguments are frequently deployed to illustrate the ways in which subjects resist, undermine and distance themselves from managerial, or capitalist, discourses. An adequate theoretical understanding of these arguments in toto is often difficult as they frequently combine a number of philosophical positions or do not specify the exact lineage of their thinking. Thus, to adequately understand and critique these arguments it is necessary to unpick the theoretical threads which are used to weave their arguments.

In seeking to clarify these accounts of resistance, therefore, this article identifies three dominant positions that theorists have drawn upon to conceptualize how employee resistance is possible from a post-essentialist perspective. The first focuses on the concept of ‘the other’ in enabling the ‘incompleteness’ of any discursive constructions (Carroll and Levy, 2008; Essers and Benschop, 2007; Stavrakakis, 2008); the second shows that ‘spaces of action’ and ‘contradictions’ between discourses develop space for resistance (Davies and Thomas, 2003; Holmer-Nadesan, 1996); finally, the third uses the concept of ‘dis-identification’ (Fleming and Sewell, 2002; Fleming and Spicer, 2004) to show how a ‘cynical’ distancing from managerial discourse can (and cannot) prevent the colonization of selfhood (Holmer-Nadesan, 1996). Such categorizations cannot, of course, be exhaustive or even mutually exclusive: whilst some authors adhere strongly to one perspective, others develop two or more of these threads, sometimes consecutively, to forge their own unique position on resistance. However, as I argue later, these three threads form the analytical bases upon which most post-foundational scholars in organization studies build their conceptualizations of resistance to discursive power in organizations.

In developing a post-foundational positioning, many authors explicitly claim a ‘rejection of an essentialist view of human Nature’ (Roberts, 2005: 620). Such a rejection comes in part from a movement away from the Marxist identification of ‘real interests’ (O’Doherty and Willmott, 2001) where individuals have been argued to be simply ‘the personification of economic categories’ (Marx, 1867: 90) and in part from the rejection of political forms of essentialism found in neo-colonial or patriarchal texts that derived characteristics of an individual (for example as oppressive, superior, maternal or lazy) from the group they belong to (such as a society, tribe, race or sex) (Tinker, 2002). The rejection of essentialism thus formed the basis of liberationist movements that rejected sexism, racism or colonialism (Fuss, 1989) and drove the deconstruction of binary oppositions such as male/female, white/black, rich/poor (Bacchi, 1990; Sayer, 2008):

if what we take ourselves and others to be are constructions and not objective descriptions, and if it is human beings who have built these constructions, then it is … possible to re-construct ourselves in ways that might be more facilitating for us. (Burr, 2003: 14).

Yet, politics aside, philosophically speaking, a rejection of essentialism is only ¹ logically necessary for ‘strong’ social constructionists, who cannot accept the non-discursive ontological status of essence (Sayer, 2000). However, many who are less catholic in their ontological claims also reject explicitly any notion of essences in their theorizing (see for example, Bergström and Knights, 2006: 353; Caldwell, 2007: 100; Fleming and Spicer, 2003: 161; Linstead and Thomas 2002: 75; Stratavakis, 2008: 1041; Sveningsson and Alvesson 2003: 1164; Thomas and Davies, 2005b: 727). In claiming this position, accounts frequently utilize discursive theorizations of identity, claiming to replace the essentialist view of the self with one which is ‘fragmented, partial and discontinuous’ (Taylor, 2004: 124) or ‘unstable, fragmented and susceptible to frequent rewriting’ (Webb, 2004: 724).

Such a movement has begged the question of whether resistance can be adequately conceptualized once essences such as interests have been disregarded (Ackroyd and Thompson, 1995; Fairclough, 2005; Thompson and Warhurst, 2003). Many critical realist scholars, for example, have long argued that post-foundational accounts of identity and resistance are inadequate without reference to a non-discursive reality (Archer, 2000; Reed, 1997; Sayer, 1997). With rare exceptions (Marks and Thompson, 2010), however, the critical realist critique of post-foundational accounts in organization studies has not considered recent developments in post-Foundational conceptualizations of resistance, focusing instead on the nascent period of discourse studies when it seemed justified to approaches as simply ‘neo-Foucauldian’ (Knights and Vurdubakis, 1994) or when the movement could be substantially critiqued by focusing solely on the work of David Knights and Hugh Willmott (Newton, 1998). Although, this is no longer the case, the realist critique, in organization studies at least, still often focuses on safer territory such as the lack of structural under-pinning to presentations of discourse (Ackroyd and Thompson, 1995; Fairclough, 1995; Thompson and Warhurst, 2003) or the perceived inadequacies in post-Foundational conceptions of agency (Archer, 2000; Fleetwood, 2005; Mutch, 2002; Reed, 1997). This article, therefore, argues that there has been a lacuna in the critique of recent post-Foundational accounts of resistance and their rejection of essentialism. Key questions that remain unanswered are: What are the ontological implications of recent post-foundational conceptions of resistance? To what extent is it possible to avoid essentialism in conceptualizing such forms of resistance?

This article holds that avoiding such questions for fear of being labelled an essentialist is untenable as ‘the alternative to philosophy is not no philosophy, but bad philosophy’ (Collier, 1994: 17). It therefore seeks to tackle the essentialism question head on by arguing three points regarding identity and resistance. First, the article argues that that not all forms of essentialism associated with resistance are ‘bad’. Within a Critical Realist framework, some essences, necessary for an adequate understanding of the social world, can avoid the charges of determinism (that action and subjectivity are determined by essences) and reductionism (that the self is merely the sum of its parts) that are frequently made by those who reject the concept. Second, using the three dominant threads used by post-Foundational writers to theorize resistance, the article argues that essences are evident in many of these accounts and should be explicitly acknowledged rather than be denied or hidden in ambiguous language. Third, the article shows that the acceptance of a more sophisticated form of essentialism can be better supported by a (critical) realist, rather than a strong social constructionist, ontology, and that the commitments of critical realism are not only compatible with many constructivist accounts of resistance but also enhance their explanatory power.

In making these points, the article does not seek to contradict the empirical findings of discursive analyses of resistance, nor does it accept the incorrect, reductionist and determinist forms of essentialism found in patriarchal or neo-colonial accounts of the self (Tinker, 2002). Instead, the article aims to provide greater explanatory power and conceptual clarity to current conceptions of resistance to discursive power through a reclaiming of essences within an emergent and stratified ontology. Such a view, it argues, is fundamentally compatabilist, capable not only of building bridges between realist and constructivist notions of identity, but at a wider level, offering opportunities for dialogue between the natural and social sciences.

Essence: a critical realist perspective

Critical realism (CR) is an ontological perspective that distinguishes what exists (ontology) from what we know about its existence (epistemology). The ontology of CR is both stratified and emergent (Bhaskar, 1978, 1979) two concepts which require a brief explanation: ‘Stratified’ means that reality exists at different levels: the empirical (what is observed), the actual (the events that occur) and the real (the underlying processes that can generate these events). This distinction allows critical realists to accept that existence of mechanisms at one level (for example, class relations) may be mediated or obscured by the existence of cultural, social or discursive processes at a different level which might generate knowledge that is wrong—such as the earth being flat or women being less intelligent than men. A stratified ontology provides researchers with three useful distinctions. The first moves researchers away from the dangers of relativism where ‘a concept simply becomes true because of its origin within a discourse’ (Cruickshank, 2003:118). The second prevents the assumption that because something is invisible to researchers it can be said not to exist. The third is a distinction between the (real) generative effects of an entity or process and the (real) effects of the discursive concept used to describe or label it. To illustrate all three benefits, we can, for example, argue that discourse (or alternatively power, or class or exploitation) existed before its identification and labelling by philosophers and sociologists because it had real effects on the social world. The historical and empirical invisibility of discourse clearly does not mean it does not exist. Yet the word ‘discourse’ (and the discourse associated with that word) has its own trajectory which, whilst we might hope it corresponds to the actual processes of discourse, can be studied separately as an epistemological phenomenon. ² Thus the increasing prominence of the word ‘discourse’ and the changing theories about what discourse is, are interesting and important in their own right, yet do not necessarily reflect the actual behaviour of discourses in the social world (Sayer, 2000). These important and analytically powerful distinctions are necessarily impossible within an ontology where the reality of discourse is equated with the knowledge that constructs it.

The second concept, ‘emergence’, means that different levels of reality emerge from, but are irreducible to, others: this article is dependent on, but not reducible to the software on which it was written, and the software is dependent on, but irreducible to the hardware on which it runs. Each level is real but demands explanation at that level. Thus, the meaning of this article cannot be explained with reference to Microsoft Office any more than Microsoft Office can be explained at the atomic level. The dependence of memory or reflexivity on atomic and cellular, or social and discursive realities, in no way means that they can be reduced to, or explained by, these levels. The importance of emergence is evident when comparing to alternative ontological frameworks. Thus, compared to atomism, where wholes are merely the sum of their parts, or holism, which ‘asserts that parts of a system can only be explained in terms of the whole they together constitute’ (Danermark, 2002: 64), CR offers an explicitly unreductionist account of social phenomena.

Within CR, and many other realist philosophies, an essence is what makes something that thing and not something else, thus enabling a distinction between accidental properties of a thing, or process, and the important ones: ‘the essence of water is H2O … the essence of chess is its rules’ (Sayer, 1997: 456). Bhaskar differentiates between ‘real essences’ which are the ‘structures or constitutions in virtue of which [a] thing or substance tends to behave the way it does’ (Bhaskar, 2008: 209), and ‘nominal essences’ which are ‘those properties the manifestation of which are necessary for the thing to be correctly identified as one of a certain type’ (Bhaskar, 2008: 279). Such a distinction enables a separation of the essential generative or enabling processes of an entity and its epiphenomenal features. For example, the nominal essences of strikes may include balloting, union membership, and strike action, whereas the real essences will include the structured antagonism between workers and employers, the socio-legal evolution of strike legislation and the historically-embedded trajectory of labour relations within the organization. Thus, the identification of real essences necessarily refers to their contribution to a wider system and it is this that distinguishes essences from mere ‘parts’. The real essences of money, the vote or learning are not thus to be found in the items themselves (the coin, the voting slip or the individual) but in the systems they contribute to (Meikle, 1985: 158).

Within a CR framing real essences gain additional explanatory power in their description of the potential of an entity. Stating that water has the potential to quench thirst but not to cause sunburn or that discourse has the potential to structure knowledge but not to lubricate engines is an important component of the essence of discourse or water, regardless of whether these features are actualized in any empirical setting (Fleetwood and Hesketh, 2010: 25). At the level of the self, the identification of potentiality provides an enhancement of explanatory power: our potential to communicate, die or construct identities tells us important things about our essences as does our inability to be a dog, read each-other’s minds or think without a brain. The specification of potentiality within an open system, such as society, enables CR to avoid charges of determinism. The essential potential of an acorn to become an oak tree and not, say, a fir tree, is a crucial part of its definition, even though its actual manifestation may be prevented say, through fire, or may simply not be observed or described by researchers. In the social realm, the labour potential of workers or the emancipatory potential of social mobility are analytically important considerations that lead the academic towards a richer and more dynamic understanding of things and their conditions of possibility without determining what will happen.

It is important to stress that, from a CR perspective, the identification of essences and generative processes may be erroneous—we may create words for things that we incorrectly perceive to exist and entities may exist that we do not have words for. Nor should essences be seen as universal, fixed or timeless. This mistake is often made when the epiphenomena of interacting and complex processes are mistaken for essences. Essences can be, and frequently are, fragmented (shared with other entities), complex (existing dynamically at several levels), processual (changing over time), or ambiguous (difficult or impossible to discover or define)—features that are frequently held in contrast to caricatures of essentialism. Thus, given the rich depiction of essences within a CR framework it is understandable that analysts become frustrated when they read writing which claims:

Like many ideologies, essentialism has teeth that inflict pain and anguish. Essentialism is the ideological glue that enables people to engage in colonialism, racism, sexism etc. … that women are more nurturing … that gay guys really can decorate better than straight ones. (Baker, 2004: 18).

Such claims, however well intentioned, misunderstand essentialism almost as much as any neo-colonialist or patriarch has ever done. Such critiques are not of the philosophical category but of historical and political ideologies that bastardize the notion of essence. The errors made by some neo-Marxists (or neo-colonialists for that matter) in their presentation of a ‘real’ self were made, not because they were essentialists, but because they were simply incorrect in both the identification of essences (for example specifying incorrect nominal essences) and (where this was the case) the supposition that such nominal essences were determined. Yet, even more considered texts frequently link essentialism with ‘universalism’ (Thomas and Davies, 2005b: 711), ‘individualism’ (Stavrakakis, 2008: 1041), ‘fixed’ and ‘singular’ identities (Somers, 1994: 605), ‘fixed character traits’ (Ybema et al., 2009: 305) and an emphasis on worker’s ‘authentic experiences’ (Thomas and Davies, 2005a: 686). Yet, none of these claims have any philosophical relation to the notion of an essence. As Bhaskar’s real essences suggest, an essence may simply be an underlying generative process which creates an unpredictable phenomena. The potential to construct an identity, to learn or to be reflexive, may be essential features of the socialized self, but any specific identity (e.g. masculine) or any specific learning (e.g. the English language) is clearly not. Furthermore, it should not be assumed that CR is incompatible with the vast majority accounts of discourse and identity. In recent years there has been a growing movement arguing not simply that CR is compatible with notions of discourse but that it enhances the explanatory power of the concept (Bhaskar, 2002; Fairclough, 2005; Joseph and Roberts, 2004). A number of revisionist analyses have supported this contention by contending that many of the philosophers drawn upon to support constructivist arguments were realists, often by their own admission (Al-Moudi, 2007; Norris, 1987; Pearce and Woodiwiss, 2001).

For realists sensitive to the power of social construction, discourse clearly exists but as an effacious process, as it ‘operates on the world and is embedded in the world, and the world impacts on discourse’ (Bhaskar, 2002: 89) or, to put it another way, ‘discourse is always a construction of something and it is this something that both facilitates and constrains a discourse and prevents it from closure’ (Brown et al., 2001: 128). This specification means that, contrary to strong social constructivist claims, discourse is both analytically and really distinct from the entities that it constructs. This is not simply true of material realities (e.g. water) which have essences that are simply not reducible to discourse, but also entirely socially constructed concepts (e.g. the concept of ‘a game’) which, although emergent from, and dependent on discourse, exist at, and are therefore usually described at, a different level. Distinguishing discourse from the entities it constructs, or helps to construct, not only opens up analytical opportunities for describing the entity itself, but also helps understand the discourse itself by describing its essences, antecedent conditions and potential, as well as the ways in which it does, or does not, act on other entities.

At this point, it is also important to stress that a CR ontology does not simply distinguish between the materially real and the discursively real. Fleetwood (2005), for example, distinguishes between the material (e.g. rocks), artefactual (e.g. computers), social (e.g. unemployment) and ideal (e.g. discourse) modes of reality, whilst others emphasize the interplay between different strata of reality that exist at different levels of emergence (Bhaskar, 1998). It is this latter argument that is most useful in discussing complex entities such as the self (Martin and Sugarman, 1999) as it accepts that aspects of the mind, such as reflexivity and memory are real and emergent from, but not reducible to the materiality of the bio-chemical brain processes.

In seeking to identify the processes and mechanisms that generate actual events and nominal essences, Critical Realists often use retroduction ‘in which events are explained by postulating and identifying structures and causal powers that are capable of generating them’ (Leca and Naccache, 2006: 635). Retroduction is a powerful tool in methodological terms as it allows the researcher to move from an empirical observation to postulate reasons why that observation might have occurred: ‘for example, to move from the observation that all ravens are black to a theory of a mechanism intrinsic (and possibly extrinsic) to ravens which disposes them to be black’ (Lawson, 1999: 10). At its most basic, retroduction asks the ‘transcendental question’: what must the world be like for X to occur? For this article, the ‘X’ is resistance to discursive power. Thus, the next section examines the different claims for how discourse is said to be resisted in organization and seeks to retroduct information about what discourse, identities and individuals must be like in order for those claims to be true. It examines the three main approaches taken by scholars studying resistance to discourse and asks, what can we imply about individuals and discourse from claims about how discourse works?

Resisting discursive power: implied essences

Depictions of control and resistance in post-Foundational literature have moved a long way since they came under attack in the 1990s for representations of an ‘over determined’ compliant subject (Ackroyd and Thompson, 1995; Newton, 1998, 1999; Reed, 1997). For this reason, the article focuses on developments in the last ten years where post-Foundational writers have explicitly developed a variety of approaches to explain how resistance is possible within a position that prioritizes discursive power. Whilst such efforts encompass a wide range of positions and explanations of resistance, three themes are commonly drawn upon by writers to explicate resistance to management discourse: the concept of ‘the other’, the specification of ‘contradictions’ and ‘spaces’ between discourses, and, finally, the concept of ‘dis-identification’. These themes are often combined by writers but are more easily analysed in distinction as they each make different, if often cohesive, claims about the nature of resistance.

Before moving on to examine these in more detail, the selection of these three positions requires some unpacking. The vast majority of accounts of resistance to discourse draw on at least one of these three approaches in explaining resistance. To illustrate this quickly, there were 45 articles published concerning discourse, identity and resistance in top-ranked ³ management journals from 2003 to 2009. Of these, 28 used one or more of the concepts detailed above. Those that did not were more often concerned with organizational or social identities that drew on alternative perspectives such as institutional theory. These three positions were not only selected for their numerical prominence, but also, as we shall see, they are frequently associated with anti-essentialism by the authors themselves.

The three threads of resistance to discursive power identified here are neither mutually exclusive, nor exhaustive. Theoretically, authors tend towards pluralism, often using a mix of both terminology and philosophers without specifying either how they relate or what their ontological assumptions are. It is all but impossible, therefore, to critique individually the sheer variety and mix of sources, statements and approaches taken by different scholars. Instead, it is more beneficial to make an analytical distinction between the different threads that scholars have drawn on to develop their positions in relation to resistance. From this perspective, whilst there are a number of outliers, the majority of the articles use at least one of the approaches outlined below. Finally, the argument here does not require or suggest that any, or all, of the articles critiqued below are united by an ontological position—they are not. It seeks simply to show the extent to which the approaches they utilize imply different forms of essentialism.

The sections below, therefore, each focus on one of the three approaches to theorizing resistance to discursive power. Each section first attempts to describe and clarify the explanations and claims of each form of resistance. After doing so, the sections consider the extent to which the use of essences can or cannot be rejected in taking these positions.

‘The Other’ and other others

‘The other’ and ‘The Other’ have a variety of distinct, if related, philosophical origins (de Beauvoir, 1989; Foucault, 1995; Hegel, 1977; Lacan, 1988; Levinas, 1998; Sartre, 1956). To critique these within an article, let alone a sub-section, is clearly impossible. Fortunately, with few exceptions (for example Arnaud and Vanheule, 2007; Vanheule et al. 2003), studies that use ‘the other’ to conceptualize resistance to discourse in organization studies rarely delve deeply into which philosophical version they utilize (see for example Davies and Thomas, 2008; Meriläinen et al., 2008) and instead tend to gravitate around two uses for the use of otherness in their work.

The first extends traditional accounts of identity construction where ‘othering’ describes a process whereby the self is reflexively constructed though what it is not: ‘the forceful exclusion and exorcism of what is Other is an act of identity formation’ (Corbey and Leerssen, 1991: xii). Thus, studies, especially those rooted in Feminist or Foucauldian traditions frequently reference the production of dominant masculine, managerial or ethnic identities through the construction of difference with an ‘other’ (Pullen and Simpson, 2009; Thomas and Davies, 2005b). Dominant identities are generated because the conception of ‘the “other” is not only usually “different” but also less desirable’ (Ainsworth and Hardy, 2004: 157). Whilst within their own writing, many authors pursue emancipatory aims through the analytical deconstruction of such dualities, descriptions of resistance by those being studied have also been analysed through the conceptualization of the other. Such efforts frequently seek to illustrate how ‘instances of pitching of ‘self as other’ against the dominant position offered within discourses can be understood as moments of micropolitical resistance’ (Thomas and Davies, 2005b: 725). An example of this approach can be evidenced in Thomas and Davies study of Kate, a personnel manager in the police service (Davies and Thomas, 2004; Thomas and Davies, 2005a). After explicitly stating their belief that identities have a ‘non-essentialized nature’ (Davies and Thomas, 2004: 119) the authors seek to distance themselves from any non-discursive account of Kate (Thomas and Davies, 2005a: 690). The authors subsequently show how Kate draws on discourses of femininity, parenthood and motherhood to generate resistance to performative and ‘masculinist’ discourses through a presentation of ‘self as other’ (Davies and Thomas, 2004: 112–115). In doing so, we are told, female professionals draw on these alternative identities as discursive resources ‘while attempting to subvert and “wriggle out” of other attempts to classify, categorize and categorize them’ (Davies and Thomas, 2004: 119).

Yet, Davies and Thomas’ presentation of the self is at odds with their acceptance of Weedon’s (1999: 104) explicitly anti-essentialist definition of the individual as a ‘site for competing and often contradictory modes of subjectivity which together constitute a person’. At the most basic level, their analysis states or implies a number of nominal essences that Kate possesses. Over the space of three pages, Thomas and Davies (2005a) show us that Kate can speak (English), reflect upon her conditions, process information, have children, present an identity, draw on discursive resources and challenge (masculine) discourses. The authors also show that Kate has emotions such as stress and joy and experiences satisfaction with a secure identity. None of this is surprising but it is entirely incompatible with a notion that Kate is only discursively constructed and thus without essences. The power to speak, or be engaged with discourse, for example, implies, and I would argue pre-supposes, the potential to learn language (Archer, 2000), which is in turn an essence, if not of every person, then at least of Kate. There are many other essences implied in the work—the description of Kate’s employee status or her motherhood, for example, imply a number of real essences in relation to systems of employment and family.

The second use of ‘the other’ to conceptualize resistance to discursive power follows a more Lacanian approach (1977a, 1977b, 1988) which focuses on the necessary incompleteness of any constructed self due to a fundamental and unavoidable ‘lack’ in totality of subjectivity. It is not desirable or possible here to critique Lacanian psychoanalysis generally, ⁴ but it is important to focus on how a Lacanian use of the other has been deployed in organization studies to better understand resistance. In the organization studies identity literature, this lack in the self (and in the other) is represented as driving employees to seek organizational identities that have the semblance, though not the actuality, of security and completeness (Arnaud and Vanheule, 2007; Driver, 2009; Vanheule et al., 2003). Yet, it is the necessary incompleteness of self which means ‘that subjection never totally succeeds in eliminating acts of resistance and silencing socio-political imagination’ (Stravrakakis, 2008: 1040). Managerial ideology and discourse, whilst inherently seductive to the insecure individual can never be complete and ‘the attempt to resolve this ambiguity provides the permanent, and permanently failing, engine for many of the dramas of organizational control and resistance’ (Roberts, 2005: 632). Thus, the ever-present otherness implied by ones own identity always signifies its own incompleteness.

Yet, whilst these Lacanian interpretations of self are often presented as complementary to discursive accounts, none of the authors that use ‘the other’ as an explanation for discourse adhere to an essentialist position. This is problematic because the ‘lack’ identified within the self by Lacan and others not only explicitly identifies a nominal essence (a defined self, albeit in reference to other(s)), but by specifying that self is ‘incomplete’ infers the theoretical position of an unfulfilled completeness (a real essence). Both the specification of a self, especially in relation to (and therefore distinct from) another, and the potentiality of completeness (or incompleteness), are fundamental essentialisms. The system of relationships between the self, the other and (in)completeness are necessarily properties of that system which are not, for example, shared by a laptop. That the completeness is never achieved is neither here nor there: it implies a teleology that has causal effects, one of which is the experience of a ‘lack’ which, by generating discomfort, drives the individual to seek stable identities, which itself is a nominal essence. That the Lacanian concept of self should be essentialized should not prove surprising. His location within a fundamentally psychoanalytic perspective makes him ‘at bottom, a philosophical realist’ (Mellard, 1991: 12). As Cahalan (1985: 448–449) argues:

if a thing is not related to another by its identity with itself, does the non-existence of the other fail to require that the thing itself does not exist? … [In these cases] causality is a relation to the other … Otherness is a logical relation but a relation between existents.

It should also be noted that a ‘lack’ in relation to the other is an acceptance, rather than a rejection, of essence. ‘Lack’ does not refer to an ex nihilo absence of anything but a specific referential dynamic—the lack being whatever the ‘other’ is not. Such a specification not only generates nominal essence(s) in reference to the characteristics of ‘the other’ but also a real essence—the identification of a system or process by which these essences are identified in relation to that other.

Spaces of action

A complementary approach taken to theorize resistance to managerial discourses is to be found in the idea of ‘spaces of action’ between discourses. Following Laclau and Mouffe (1985: 111) analysts argue that signifiers such as a worker or ‘quality’ are ‘over-determined’, thus opening a discursive struggle for meaning in the appropriation of any terms. Laclau and Mouffe follow Lacan in emphasizing the symbolic nature of over-determination and thus rejecting essences in both the subject and the social (Lewis, 2005: 6). In doing so, identities are represented as ‘accepted, refused and negotiated in discursive processes … identity is thus something entirely social’ (Jørgensen and Phillips, 2002: 43). Such over-determination, for subjects, we are told, forces discourses into conflict which generates ‘spaces of action’ for resistance and agency which, in turn, hold employees ‘suspended betwixt and between…subject positions’ (Musson and Duberley, 2007: 160; see also Whittle, 2005: 1302).

An example of this positioning is to be found in Holmer-Nadesan’s ‘Spaces of Action’ article which seeks to demonstrate how ‘a group of women service workers … identify, counter-identify and dis-identify with managerial formations of their identity (1996: 50). Holmer-Nadesan first explicitly rejects the notion that ‘identities exist in stable, self-contained form[s] across time … [or that] identities are constituted in essences that have trans-historical potential’. She then moves on to build on Laclau and Mouffe to argue that managerial discourses and their signifiers are overdetermined and thus cannot be fixed. Using the word ‘woman’ as an example, she suggests that ‘individuals that occupy the subject position ‘woman’ are faced with the ambiguities and possibilities that stem from the concept’s instability’ (p. 51). She goes on, ‘with each articulation (in speech and practice) the subject invokes her/his identity by drawing upon discursive forms but always/already partially … inability to fully determine the identities of self and practice has the effect of engendering space for contingency and for choice’ (p. 52) [emphasis added].

Such an analysis, however, cannot be made without reference to essences, either of the self, or indeed of discourse. The descriptions of discourses, to take this first, state that they can that they can over-determine a subject, but not completely construct a subject, that they can be in conflict and that there are discourses of patriarchy and class. Such statements imply much about the metaphysical properties and dynamics of discourse which themselves must be non-discursive: we know, for example, that a patriarchal discourse is not a capitalist discourse, presumably because it has some features which are unique to it. We know that these discourses can construct a subject, in a way that a toaster or gravity cannot. We know that one signifier (for example ‘woman’) can relate to several discourses. We can also infer much from what is not said. For example, the fact that discourses appear to ‘over-determine’ a subject and do not, it seems, dissolve each-other or allow one discourse to ‘win’ and the other to ‘die’ (or create rain or nuclear fission). All of these statements are specifications about either the (potential) causal powers of discourses (real essences) or their epiphenomenal properties (nominal essences).

Turning to the subject, it is clear from the presentation of discourse that most accounts, including Holmer-Nadesan, not only infer a (partially) non-discursive self, but also tell us something about the properties and powers of that self. We are told, for example, that the conflicting ‘discourses do not completely colonize processes of meaning-making, nor determine how they are acted upon’ (Whittle, 2005: 1302; also see Musson and Duberley, 2007; Thomas and Davies, 2005b). This implies several things about the capability of subjects. First, that they are, or have the potential to be, reflexive. This is often explicitly stated by authors (e.g. Thornborrow and Brown, 2009) but, in any case, is implicit in the assertion that there is some form of consideration of, and choice between, discourses by the subject—a statement necessary for the rejection of the charge of determinism). Furthermore, the preference for resisting dominant managerial discourses must be based upon a perceived preference for one or the other. For example, in Sveningsson and Alvesson’s (2003) study of ‘H’, a manager in a high-tech company, the manager seeks to avoid low status janitorial work in favour in high value strategic tasks. Whilst such a preference is obviously constructed in the social realm, it has clearly been internalized as an interest. This is not (necessarily) an interest that can be read from H’s class position, but it is an interest nonetheless. As Marks and Thompson (2010) argue, such interests are central to the generation of identity:

interests are fundamental. Even those who write about identity seldom explicitly disavow interests as a motive force. Most of the time … interests are a hidden script in identity explanations.

To reiterate, whilst any specific interest is clearly not a generalizable property of individuals, the capability of developing interests is a necessary condition of being able to chose between discourses and thus being capable of resistance and emancipatory action. The possibility of having interest raises questions about the processes by which these are learned and how the individual develops some judgemental capability by which they can be weighed against competing or alternative interests.

Dis-identification and cynicism

Many writers have shown how workers, at the local level, dis-associate themselves from managerial discourses through humour (Collinson 2002), scepticism (Fleming and Sewell, 2002) or cynicism (Fleming and Spicer, 2004). These micro-level activities taken together, demonstrate how workers ‘secretly voice statements of disbelief about the official culture and are rather jaundice-eyed towards management. They could ‘see through’ the hollow promises of human resource departments and did not really ‘buy into’ the hype of corporate culture pundits’ (Fleming and Spicer, 2003: 159). Dis-identification frequently means acting out the role that management expect but maintaining a ‘psychological distance’ which avoids the total identification with management ideology. Whilst some writers, such as Judith Butler (1993) point to the destabilizing nature of dis-identification, recent applications in organization studies have argued that the ‘acting out’ involved in dis-identification simply reproduces the exploitative nature of work regardless of how workers feel about it on the ‘inside’ (Fleming and Spicer, 2003): ‘I can resist others in defence of my autonomy but then find myself curiously bound up with those I resist. I can through dis-identification find ways to deny that it is really ‘me’ who is conforming but still conform’ (Roberts, 2005: 632).

To be fair, not all of those who describe ‘dis-identification’ with management power explicitly reject an essentialized view of the self. An earlier set of studies, set apart from post-Foundational movements describe, for example, ‘resistance through distance’ (Collinson, 2000: 163), ‘a defensive siege mentality’ (Delbridge, 1998: 189) or ‘role distancing’ (Kunda, 1992: 163). However, in recent years, dis-identification has been used specifically to describe how resistance to discourse is possible within an ontology that rejects essentialism (Holmer-Nadesan, 1996: 59; Roberts, 2005: 620; Whittle, 2005; Willmott, 1993). Holmer-Nadesan’s definition of dis-identification is a prime example: ‘dis-identification marks the limits of organization because in the absence of some generalizable form of identification with a dominant discourse there are no common organizing principles—no centres’ (1996: 59). However, despite these claims, it is difficult to envisage how dis-identification is possible without postulating an individual capable of choice.

The very word (dis)identification implies something that (dis)identifies and thus is not simply painted over by discourses in a completely passive manner. Indeed, the way in which many authors describe the process of dis-identification illustrates an individual whose rejection of managerial discourse is based, not simply on being an inert canvass but instead, a reflexive being that considers the validity and consequences of the cultural imperatives which managers foist upon them. In virtually every empirically-informed account of dis-identification authors depict individual workers reflecting upon and assessing the validity of managerial discourse as it is passed to them in training sessions, newsletters, meetings and speeches. ‘Workers’ here cannot simply be dismissed as a metaphorical device for subject positions because their reflexivity, consideration, actions and communication are presented as essential components in the dynamics of dis-identification. In other words, the process, as described, is, in part dependent on essences of a self that are non-discursive.

The counter argument, put by Roberts (2005: 630) is that:

dramaturgical conformance and cynical dis-identification seem to depend upon such a sense of interiority as a space within which some vestige of choice and self-determination can be preserved. But it is also clear from Lacan that this is an imaginary space that, as Fleming and Spicer argue, makes subjection as conduct more possible.

As stated earlier, an essentialist view of the self or discourse does not equate to a continuous (universal, singular or knowable) self and so the argument that a ‘sense’ of interiority is imagined rather than an essence (either real or nominal) is not incompatible with essentialism. However, the obvious, and answered question is what is it that is doing the ‘imagining’ or ‘sensing’ the interiority? If the entity that can imagine and sense, as well as reflect, decide and act, then these can only amount to essences that interact with, but are not entirely constituted by discourse.

Discussion

Essences of discourse and the self

The section above sought to clarify the main forms of recent theorizing around resistance to discursive power and to show that, despite protestations to the contrary, this theorizing often rests upon essentialist notions of both discourse and the self. With regard to the self, not only do post-Foundational writers provide rich empirical descriptions of embodied selves that think, feel, imagine and sense, but their own theoretical positions infer that such entities are, at least partially, non-discursive. Whilst discourses may conflict, and spaces may open up, something ultimately needs to engage with the possibilities that are generated. The texts clearly show individuals reflecting, deciding and acting in relation to discourse, and it is this engagement where the non-discursive meets the discursive: ‘human beings must have a particular make-up or nature for them to be conditioned by social influences in consistent ways’ (Sayer, 1992: 121). This does not mean, of course, that social identity, subjectivity or even self-identity are not, to a great extent, discursively produced or that extra-discursive essences are not discursively mediated, simply that there is no reason to sacrifice the human in order to analyse this.

At the discursive level, writers similarly tell us much about the essence(s) of discourses: that they have the potential to construct meaning, that they come into conflict, emerge from language and semiotics, effect power relations and affect (or effect) subjectivity. The latter point is important because it tells us that individuals have the potential to be socialized through discourse (Townley, 1994) not only changing how they are constructed in society (social identity) but how they construct themselves (self identity). This in turn implies further powers of individuals that other entities such as tables do not possess. These powers enable individuals to both access discourse (for example, through experience) and to internalize it (for example, through socialization) which, if one wanted to go further, might imply the capacity for socialization, learning and memory.

The elephant in the room here is the mismatch between the espoused ontology of many post-Foundational writers and the philosophical consequences of their empirical findings. Whilst many authors, in the last ten years, have moved away from a strong social constructivist position (that everything is discursively constructed) to a weak social constructivist position (that there may be things which are not discursively constructed but we cannot know about them) this is often not reflected in their claims. It is increasingly common to witness articles that begin and end with statements that reject the possibilities of both essentialism and a knowable world (for example, Caldwell, 2007: 100; Linstead and Thomas, 2002: 75; Prasad and Prasad, 1998: 251), but empirically describe people that think, feel, act, learn, reflect and resist. Such findings not only imply that individuals have causal powers, and therefore essences, but also that the researcher can know about them.

To some extent, none of this should be surprising. It is impossible to discuss, let alone theorize, the dynamics of anything, without implying that something exists and has real effects. The reaction against essentialism, for the most part, seems not to be prompted by philosophical strictures, after all even Foucault ‘in no way privileges the discursive or cultural realm as that within which the remainder of sociality subsists’ (Pearce and Woodiwiss, 2001: 61). Instead it seems driven by a rejection of the forms of essentialism that were inherent in neo-colonial, patriarchal or Marxist accounts of the individual. Yet, from a CR perspective, it is entirely plausible to reject these incorrect and sometimes damaging forms of essentialism promoted by these movements but to accept the forms of essentialism that allow a coherent and consistent discussion of discourse, subjectivity and resistance. So, if this is the case, what would a CR framing of resistance to discursive power look like?

Discourse, essence and resistance: a critical realist perspective

From a CR perspective, resistance to discursive power necessitates an ontological distinction between the discourse and the self, with an acceptance that both have causal powers that operate within a stratified and emergent ontology. In this section we will examine the realist repositioning of both discourse and the self and examine how this enables a more plausible conceptualization of resistance. Within this viewpoint, discourse is now framed, less as an all-encompassing, amorphous entity with no essential powers or structures, and instead as stratified across the real, actual and empirical domains. It is real because it has causal powers (Bhaskar, 1979: 12) such as the capacity to construct a social identity. Yet, as with all powers, its potential, and its manifestation in the domain of the actual, is dependent and contingent on a number of factors which include interactions with other discourses, different emergent levels (such as semiotics) and action. Thus ‘hermeneutics by itself cannot provide an adequate explanation of social phenomena … there is always an extra-semiotic context to the operation of hermeneutics’ (Fairclough et al., 2004). The operation of discourse in the domain of the actual does not necessarily mean that researchers have access to it, but if they do, it enters the domain of the empirical. Thus, our identification and analysis of discourse (i.e. our discourse regarding discourse) could be mistaken. This presentation of discourse is one with both powers and essences identical to those described above but held within an ontology which is consistent with its empirical claims.

Such a definition allows us to begin to answer how discourse comes into being and how it can change: a CR perspective rejects a discursive answer to this question as self-referential and instead points to the possibilities of emergence. Discourse is dependent on, though irreducible to materially, artefactually, socially and conceptually real phenomena. (Fleetwood, 2005). Whilst discourse may affect these things it does not effect them because they both exist at a different ‘level’ of reality and rely, in part, on non-discursive conditions. Power, for example, is discursive in part, but inter alia, it is also material, social, conceptual and artefactual, as anyone who has been imprisoned will experience. In CR terms even semiosis cannot be reduced only to discourse (Collier, 1998; Sayer, 2000: 37-40). Thus discourse changes, in part, when those entities which it is emergent from change. Amongst other things, this incorporates changes in the realms of power, materiality and social interaction, whether intentional (such as 9/11), accidental (such as Chernobyl), technological (such as Facebook) or structural (such as the growing gap between the rich and the poor). Emergence, therefore prevents self-referential dynamics, whereby discursive change can only be explained through discourse.

Second, we need to examine the implications of a stratified and emergent ontology with regard to the self (Archer, 2000, 2003). From this perspective, aspects of the self, for example social identity, are, of course, constructed primarily through discourse, but also interact with other phenomena, such as self-identity, action and materiality (Watson, 2008). In addition, emergence, with reference to the self, shows that the self is emergent from, but irreducible to, a number of complex and inter-related essences, many of which are specified by the post-Foundational writers that have been reviewed here: reflexivity, feeling, memory, imagination, learning, action, identity construction (O’Mahoney, 2011). Other realists have specified these in more detail (Archer, 2000; Smith, 2010) but for the purposes of this article there is no need to go beyond those already implied in the reviewed articles. These essences, in turn, are emergent from, but irreducible to, a number of embodied (physical) and discursive (social) phenomena.

Resistance, then, occurs at different levels within a CR ontology, but the very specification that resistance to discourse is possible necessitates non-discursive essences: ‘if humans were totally plastic they could not be compliant with particular managerial or other discourses (even leaves-in-the-wind need to have certain properties, including resistance, to be blown around)’ (Fleetwood, 2004: 18). Many of the essences that humans possess, through their very existence, or lack of existence, create resistance. To give a few examples, at a physiological level, humans need to eat, sleep, drink and rest to survive. These essences by their very existence create boundaries to the exercise of corporate power, even within extreme conditions such as sweatshops. Conversely, the incapability of employees to work indefinitely and be in two places at once, or for managers to be omniscient and omnipotent, again, creates boundaries to the exercise of managerial power.

At a discursive level, however, a conceptualization of resistance necessarily involves powers such as reflexivity and action: ‘for changes in discourse to be causally efficacious … and not just accidental we must know something about how the determinations we want to avoid work and how they can be subverted’ (Fleetwood and Ackroyd, 2004: 15). Thus, a critical realist perspective enables a number of obvious opportunities for resistance to discourse that are simply not possible within an ontology where such essences ‘do not have an existence outside socially constructed discourses’ (Caldwell, 2007: 100). For example, that the discourse has not has requisite time to socialize the individual (such as a newcomer to a company); that the person does not speak /read the relevant language; or that an individual notices important contradictions within a discourse which prompts its rejection (‘employees are our greatest asset’).

This ontological distinction between the individual and the discourse enables the three forms of resistance outlined earlier to be accepted but without the awkward pretence that no essences are being implied. First, the imagination of an ‘other’ is acceptable, not only because imagination is an essence of a person but also because it is plausible to specify that the process of constructing an identity may include a causal process by which one’s identity is contrasted with a real or imagined other. Moreover, it is acceptable to suggest that the nominal essences of that other define, by their absence, some of the nominal essences of the identity being constructed. Second, if we wish to specify that discourses come into conflict and open up ‘spaces for action’ this is nothing more than a specification of the properties and causal powers of the dynamics of discourse. The non-discursive existence of a proactive, reflexive self enables an individual to choose between conflicting discourses in constructing a resisting identity. Third, dis-identification is clearly labelled as a causal process by which an individual chooses not to accept the validity of a discourse and takes actions to resist it. Thus, an ontology which accepts a non-discursive self enables a much clearer and less inconsistent presentation of how resistance is possible in the three ways frequently used by post-Foundational writers. Moreover, the location of that self within an emergent and stratified ontology avoids the reductionism or political degradation associated with essentialisms found in alternative philosophies.

Conclusion

How is a theory of resistance to discursive power possible without a concept of essentialism? This article has argued that it is not. It has shown not only that the theoretical devices used by post-Foundational writers to achieve such a goal contain essentialist notions of both the self and of discourse, but also that any theorization that denied this would be out of step with depictions of humans as reflexive agents capable of resistance and action. Essentialism provides a theoretical basis, not only to discuss how the powers of discourse construct social identity, but equally how the powers of a reflexive agent can resist these. An important subsequent question asks, how can we use essentialism without incurring the consequences of reductionism evident in neo-colonial or patriarchal texts? The article proposes the use of the emergent and stratified ontology found in critical realism. Thus, it is unnecessary to suggest that individuals are fixed, universal, single or determined—the temporal and processual understanding promoted by the notion of real essences generates identities which are consistent with the claims of many post-Foundationalists as emergent, processual, fragile and multiple. These realist notions do not contradict the existence or operation of discourse but provide a complimentary and richer explanatory framework for understanding how it works in relation to the self. The article has not sought to make any empirical assertions about the self but has drawn, for the most part, on the assertions of post-Foundational writers, attempting a retroductive analysis of their findings in an attempt to provide greater clarity about what must be the ontological case for their assertions to be valid.

In taking this position, the article joins a growing number of texts that seek to move debates within CR closer to those of post-Foundationalism (Archer, 2000, 2003; Cruickshank, 2003; Fairclough, 2005; Sayer, 2005; Smith 2010). Critical realists in the sociological realm, perhaps more fearful of the charge of essentialism when discussing individuals, have traditionally focused instead on structural matters such as class, professions or unions. Yet this aversion is unnecessary. The ontological clarity and conceptual power that CR can deliver can help clarify and bolster our understandings of identity and the self. In this case the article sought to show how many writings that reject essentialism actually make considerable inferences and pre-suppositions of essentialism in their theoretical and empirical claims. Such a position does not deny these empirical claims but denies the ontological claim that ‘the nature and form of resistance are discursively produced’ (Thomas and Davies, 2005a: 701). As such the argument taken here may well be extended to other areas, such as discursive psychology, where the relativism generated by discursive approaches has prompted analysts to note that ‘an attempt to locate our understanding of psychological processes in political context without sliding into full-blown relativism have unfortunately been compelled to articulate arguments that are compatible with CR on the terrain of ‘discourse’ (Parker, 2007: 391).

By using an emergent ontology, critical realists seek to build bridges, not only with post-Foundational authors, but also with the natural sciences whose bogey-man status is becoming an unsustainable embarrassment to progressive sociology. The promotion of a non-reductionist, anti-positivist, emergent ontology accepts the importance of discourse, power and identity in mediating or knowledge of the natural world, but can equally accept that material and immaterial realities constrain and enable such dynamics (Archer, 2000: 87). Such an acceptance opens possibilities for future theorizing—critically engaging with, rather than dismissing, disciplines such as neurology, biology and psychology, whose own ontological strictures might limit their own understanding of identity and the self. The alternative to the proposed acceptance of essences is (still), I would argue, an increasingly idiosyncratic and self-referential ontology that is incapable of describing, and therefore promoting, resistance.