Leadership,more or less? A processual,communication perspective on the role of agency in leadership theory

Introduction

Leadership has been extolled as ‘the’ solution to competitive disadvantage and the need for further economic growth for many years. Rather than ‘the Great Recession’ calling such notions into question, it largely continues to be assumed that ‘better’ or more refined models of leadership are needed to enable renewed growth and prosperity. Some scholars, recognising the limitations of previous approaches, have stressed the importance of ‘good’ followership as well as ‘good’ leadership. This article takes a more critical approach. It suggests that leadership, as widely envisaged, is a key part of the problems we now face rather than the solution. The widespread faith in such leadership may be viewed as another instance of ‘The God That Failed’. ¹ Drawing on process (Langley and Tsoukas, 2010; Langley, 2009) and communication (Putnam and Nicotera, 2009) based theories of organisation, it challenges dominant frames of reference within the debates about leadership. In particular, I argue that transformational leadership theory (Bass and Riggio, 2006), widely regarded as the dominant approach within leadership studies over several decades, suffers from an implicit model that puts ‘the leader’ at the centre of more or less solid hierarchies and stable networks in which greater agency is attached to the leader than to followers. Functionalist perspectives on followership have not overcome this problem.

Agency is defined here as the capacity to take action. A crucial aspect of such capacity is the actor’s sense of both their own power and that of others (Stones, 2005). When power is mobilised for action, it is a reflection of a deeper power that is embedded in the identities to which organisational actors subscribe, and in wider social relations of dominance, submissiveness and resistance (Clegg et al., 2006). Leadership theories that stress the agency of leaders but diminish that of others legitimise differential power relations between leaders and followers and endorse the associated concentration of decision making in the hands of managerial elites. ²

Utilising the analytic lens of dissent in organisations (e.g. Banks, 2008; Kassing, 2011), I therefore conceptualise leadership in terms of networks of interaction between organisational actors. This suggests that such networks are likely to reach more effective decisions when follower dissent is institutionalised into the theory and practice of leadership and offers greater space for voices of resistance. It follows that the premium often placed on organisational consensus is misplaced, since it tends to view follower dissent as ‘resistance to be overcome’ rather than useful feedback (Mumby, 2005; Tourish and Robson, 2006). This encourages excessive leader agency, including, often, when the perspective in question appears to advocate the opposite. A more processual and communication oriented perspective of leadership yields theoretical and applied possibilities beyond what the field has so far offered. In particular, this article explores the following propositions:

Leadership can best be viewed as a fluid process emerging from the communicatively constituted interactions of myriad organisational actors. It is not a finished category, standing apart from the complex organisational processes that produce it.

Leader agency (at expense of) followers

Extant models of leadership tend to unreflexively privilege leader agency over that of other organisational actors. As Banks (2008: 11) puts it:

Conventionally, leaders show the way, are positioned in the vanguard, guide and direct, innovate, and have a vision for change and make it come to actuality. Followers on the other hand conventionally track the leader from behind, obey and report, implement innovations and accept leaders’ vision for change.

Thus, Weber’s (1968) interest in charisma as a form of domination rooted in the relationship between a leader and the leader’s followers has been supplanted by a narrowly functionalist perspective, with most research focusing on its ‘effectiveness’ and how this can be measured (Conger and Kanungo, 1998). Transformational leadership theory, in particular, stresses how charismatic leaders can inspire, intellectually stimulate and radically reorder the values and actions of others (Van Knippenberg and Sitkin, 2013). Such leaders act on others rather than alongside them. They decide on a vision and then align followers’ mindsets with goals that are consistent with the vision (Hartnell and Walumbwa, 2011). It is an approach that has long been criticised for seeing organisational influence in unidirectional terms (i.e. flowing from leaders to more or less compliant followers), advocating the achievement of corporate cohesion and a monocultural environment to the detriment of internal dissent and exaggerating the role of charismatic visionaries in the achievement of corporate goals (Tourish and Pinnington, 2002; Tourish et al., 2010).

The assumption of dominant leader agency is also a given in much of the literature on organisational failure. Amar et al. (2012: 69) assert that ‘The reason organizations fail to respond to market conditions is the inability of their senior leaders to manage their organizations to the complexity and dynamism of their business environment’. In then arguing for more distributed forms of leadership to deal with this problem, they suggest that

The right approach for responding to these changes requires vesting the authority to lead where it is needed for quick reading and responding. A management style consistent with this principle will allow employees at every level to assume leadership roles, make decisions and manage their part of the business like the top managers of the organization manage corporate strategy, compliance and other corporate tasks.

It is likely that theories which privilege the agency of those who hold formal, hierarchically based leadership positions above that of others will have an intuitive appeal for many actors, including business school students. The theory finds traction because it legitimises dominant power relations, which is appealing to those who either hold power or covet it. Those relations in turn further legitimise and promote a theory which appears to simply describe ‘what is,’ and which therefore (surely?) must lie beyond interrogation. Theory and practice may thus become mutually constitutive. In addition, by legitimising the concentration of power in elite hands, much leadership theory can contribute to the on-going crisis in society, however inadvertently. It is obvious by now that concentrating power in the hands of a few, whether in countries or organisations, has not been a successful experiment in decision making. Can we do better, in our theorising, pedagogy and practice?

The liberating potential of followership?

One answer to this question has been an increased focus on the notion of ‘followership’. However, the dominant, functionalist approach offers little conceptual clarity on how followership might offset the problem of excessive leader agency in organisations. While there has been some suggestion that followership is consistent with the notion of ‘participants’ or ‘collaborators’ (Uhl-Bien, 2006), the term is most often employed as a synonym for ‘subordinate’ (Crossman and Crossman, 2011). Kellerman (2008: xix) critically notes that subordinates are typically conceived as those ‘who have less power, authority and influence than do their superiors and who therefore usually, but not invariably, fall into line’. Thus, scholars frequently invoke the term ‘followership’, but in doing so they continue to reify and naturalise hierarchy, thereby reaffirming leader agency. Accordingly, ‘Followership is a relational role in which followers have the ability to influence leaders and contribute to the improvement and attainment of group and organizational objectives. It is primarily a hierarchically upwards influence’ (Carsten et al., 2010: 559). Asymmetrical power is taken for granted. It is simply assumed that ‘group’ and ‘organisational,’ as opposed to sectional, objectives exist, and that leaders are the prime arbiters of what they should be – albeit while remaining open to an unspecified degree of influence. Moreover, followership is viewed as being what assists in the ‘improvement’ and ‘attainment’ of such objectives, rather than what might fundamentally interrogate them.

While critical perspectives are much more nuanced on these issues (e.g. Chaleff, 2009; Collinson, 2006), functionalist approaches tend to presume that dissent and resistance are somehow incompatible with the notion of ‘good’ followership. Rather, ‘good’ followers are those ‘to whom a leader can safely delegate responsibility, people who anticipate needs at their own level of competence and authority’ (Kelley, 1988: 144). ³ Not only do they follow instructions from powerful others, they are endowed with the mind-reading skills that enable them to anticipate what these might be. Followers therefore ‘display competences that mirror those of their leaders’ (Cunha et al., 2013: 87), rather than develop contrary competences of their own.

There is a wholly imbalanced view in this literature of the nature of agency. On the one hand, leader agency is assumed to be absolute. On the other hand, ‘follower’ action is robbed of much of its agentic potential. Rather, the behaviour of followers is viewed as wholly dependent on the structural constraints that are determined by leaders. A process-oriented perspective would challenge the traditional separation in the literature between leaders and followers (Collinson, 2006). It can help to rebalance our view of agency in leader–follower interactions and therefore contribute to forms of organising that are less likely to inflict social, organisational and economic harm. Since critique bereft of an alternative often disables the urge to action, and therefore reinforces those power relations that it seeks to challenge, I see this article as answering the call for a ‘critical performativity’ approach to leadership (Alvesson and Spicer, 2012: 368). This recognises ‘the potentially positive value of functional exercises of authority’, and seeks to intervene in specific management debates with a view to improving both theory and practice.

A processual, communication perspective

Process perspectives suggest that ‘the organization is constituted by the interaction processes among its members’ (Langley and Tsoukas, 2010: 4). This interaction is fluid, ongoing, evolving, creative and without any definite endpoint (Langley et al., 2013). It follows that organisations can be viewed as simultaneously differentiated and integrated, on a number of dimensions. In communication terms, this view is best captured by the work of what has become known as the Montreal School, and which is discussed by Fairhurst and Connaughton (2014) in the current issue of this journal. As Cooren et al. (2006: 2–3) express it:

organization emerges in the interplay of two interrelated spaces: the textual-conceptual world of ideas and interpretations and the practical world of an object-oriented conversation directed to action… The resulting image of organizational interaction is of an essentially fluid and open-ended process of organizing, in which inherited positions of strength are exploited creatively by the participants.

In line with this, the recognition that sensemaking, agency and the processes whereby co-orientation between organisational actors is mediated through language means to acknowledge that organising is ‘an act of juggling between co-evolutionary loops of discursive phenomena’ (Guney, 2006: 34). The metaphor of ‘juggling’ suggests tension, including between leaders and followers, and an omnipresent prospect of breakdown. Organisations therefore struggle to create shared meanings between organisational actors. Yet functionalist perspectives usually neglect to explore the ways in which different actors attribute varied meanings to leadership (Alvesson and Spicer, in press). But to acknowledge the reality of variegated meaning is to challenge the view of leadership as having a taken for granted existence, separate and apart from its discursive constructions. Rather, the organisational world can be depicted as a plenum of agencies (Cooren, 2006). What has been termed the ‘communicative constitution of organization’ (McPhee and Iverson, 2009: 49) in turns means the communicative constitution of leadership. ‘Leadership’ therefore emerges through the interaction of organisational actors and has a contested, fluid meaning for all of them, in a given social situation for determinate amount of time. It is fundamentally an on-going process rather than a finished accomplishment. Meaning is constructed, deconstructed and reconstructed between those in leadership positions and those that they lead (Fairhurst, 2007; Fairhurst and Grant, 2010; Smircich and Morgan, 1982). Leadership is not a discrete phenomenon with easily observable causal relationships, inherently powerful and charismatic leaders, measurable outcomes and clear demarcations between categories of meaning and behaviour.

Our perspective on leadership shifts, once we view organisation as dealing with how ‘socially constructed institutions are reproduced and transformed by the accounting activities of people in interdependent (joint) action as they make sense of what they do together’ (Varey, 2006: 191). It means accepting that ‘communication generates, not merely expresses, key organizational realities’ (Ashcraft et al., 2009: 2). Accordingly, the reputations of powerful leaders, particularly CEOs, emerge as a phenomenon that is co-produced and co-reproduced (within certain limits) by the discursive interactions between organisational actors (Sinha et al., 2012). This perspective draws attention to what has been described as ‘the dance between leader and led and its language of connectedness, temporalness, and embeddedness’ (Fairhurst, 2007: 24).

The study of dissenting or critical communication from followers to leaders is a useful illustration. Much of the literature on employee ‘voice’ focuses on those forms of expression most calculated to assist in the implementation of goals that, although determined by managers, are assumed to express a unitarist interest (Morrison, 2011a). When a broader approach is adopted scholars report that non-leaders typically find their opportunities for dissent from leaders constrained (Kassing, 2011). This has been described as the ‘hierarchical mum effect’ (Bisel et al., 2012: 128) and can be seen as a situation in which ‘moments of contestation are precluded by power imbalances’ (Deetz and McClellan, 2009: 446). But to stress only this misses crucial processes of co-construction that also occur. For example, ingratiating behaviour by followers, in which they exaggerate how much they agree with the opinions of leaders, contributes to exaggerated self-belief, narcissism and the adoption of ultimately destructive forms of leader action (Tourish and Robson, 2006). Leaders also often experience personally negative outcomes from the dynamic of ingratiation. One study, among many similar studies, has found compelling evidence that poor performance, resulting from the behaviour of CEOs who have internalised the overly positive feedback delivered by their followers, increases the likelihood of them being fired (Park et al., 2011).

We can see here that the identity (and fates) of leaders as leaders, and that of followers as followers, is the result of a mutually constitutive interaction between the two. For followers, the decision not to offer critical feedback is a demonstration of agency manifest in silence, based on an often justified calculation of self-interest. But silence remains a form of communication, and hence has a co-constructive impact on its recipients in formal positions of leadership. Organisational actors can never fully relinquish the power to manage meaning, since any attempt to abstain from communication itself becomes a form of communication. In Fairhurst’s (2007) terms, drawing on a Foucauldian perspective, this casts leaders as the subjects of influence attempts by others, as well as agents who make things happen to other people. In essence, this approach suggests that organisational phenomena, including leadership, are regarded ‘as (re)created through interacting agents embedded in sociomaterial practices, whose actions are mediated by institutional, linguistic, and objectual artefacts’ (Langley and Tsoukas, 2010: 9). Moreover, the attribution of leader and non-leader identities is something to which most actors attach considerable importance. Such attributions have implications for resource allocation, decision-making authority, voice and much else. It is difficult to see how they could be regarded as a secondary aspect of organisational life.

Leadership as a first-order communicative process

Recognising its importance, I therefore wish to revisit the proposition that leadership can be viewed as a first-order organisational process, to suggest that these ideas can be taken still further. Nicotera (2013), in looking at how it can be said that organisations are communicatively constituted, argues that what is constituted is

(a) the collectivity, (b) the social significance of the collectivity as an entity whose interests are represented in individual and collective activity, and (c) the distinct entitative being that transcends and eclipses any individual and the collective itself as it is attributed both identity and authority. (p. 67)

Such processes are both discursive and material, in that the tangible architecture and artefacts that we see in organisations are also employed to bolster entitative claims (Ashcraft et al., 2009). As Ropo et al. (2013: 379) have argued, ‘places and spaces construct and perform leadership,’ albeit in interaction with the non-material. Thus, the imposing symbolism of a CEO’s office transforms the discursive dynamics between its occupant and those in ‘subordinate’ positions who visit to report, account for themselves, engage in discussion – but rarely to critique. Non-leaders are complicit in the performance of leaders who assert entitative claims for social structures that facilitate their exercise of agency. This co-constructive complicity is manifest every time they follow instructions, embrace organisational rituals or acknowledge the primacy of formal leaders. Leadership is therefore a first order means whereby the entitative claims of organisational actors are both disputed and enacted, and by which their sense of agency is enabled and constrained.

Leadership as a dialectical nexus of fluid relationships

As outlined here, a communication and process perspective recognises what Collinson (2005: 1435) has described as the

deep-seated asymmetrical power relations of leadership dynamics…. From this perspective, control and resistance are viewed as mutually reinforcing, ambiguous, potentially contradictory processes. Followers’ resistance is one such unintended outcome. In its various forms, dissent constitutes a crucially important feature of leadership dialectics, requiring detailed examination by researchers.

By contrast, leaders are often tautologically presented as leaders simply because they are (charismatic) leaders, who act on relatively inert ‘others’. In such approaches, leadership is not a process of on-going co-construction. Thus, conventional leadership research explores causal mechanisms whereby leaders exert influence on others, or how followers exert influence on leaders (to a lesser degree). But ‘leadership’ and ‘followership’ are assumed to be dichotomous categories rather than complex, non-linear processes. From a processual, communication perspective, it is not possible to isolate causal variables in leadership research and produce much more than modest observations and small-scale predictions that only hold on a mini-canvas. To illustrate, I took three articles on transformational leadership at random. I found that they dealt with such issues as the impact of a rater’s personality on perceptions of transformational leadership (Bono et al., 2012), the impact of perceived androgyny on leader effectiveness (Kark et al., 2003) and how transformational leadership impacts on positive emotions and levels of customer service (Chuang et al., 2012). These may be worthy questions. But they simply take transformational leadership as a construct for granted, focus their efforts on resolving ‘gaps’ in incidental aspects of its theory and completely miss the larger issues of agency, power and control, and the processes whereby transformational leaders are co-constructed as such by varied organisational actors.

Such research overstates leader agency and diminishes the agency of others. It represents a preoccupation with leaders rather than leadership. This has real-world consequences, since it encourages the over attribution of agency to leaders by myriad organisational actors and hence has the tangible effect of reinforcing leader power at the expense of others. If, by contrast, we view organising ‘an on-going encounter with ambiguity, ambivalence and equivocality’ (Czarniawska, 2013:12), then this sensibility must be extended to leadership, with radical implications for both scholarship and practice.

In the messy and complex world that we see around us, as opposed to that depicted in most leadership research, leaders and followers are interacting organisational actors, whose identities as leaders and followers are simultaneously constructed and deconstructed by the force of their on-going respective struggles to realise their agentic potential. Leadership therefore cannot be depicted as a force that stands separate and apart from complex systems, neutrally exerting varied degrees of influence and control to achieve putatively positive outcomes – the approach envisaged by Solow and Szmerekovsky (2006) and McKelvey (2010), among others. Rather, it is a process that is itself an integral component of the complexity that it is increasingly recognised constitutes organisational action. In communication and process terms, it is a twofold category mistake to use ‘leader’ as a synonym for ‘leadership’ and to describe complex systems while positioning leaders/leadership as independent agents standing apart from organisational complexity. Rather, we need to step back and see how leaders are themselves complexly constructed and deconstructed over time.

This standpoint offers a dynamic conception of power dynamics, since from a communication perspective, ‘power is conceptualised primarily as a struggle over meaning: the group that is best able to “fix” meaning and articulate it to its own interests is the one that will be best able to maintain and reproduce relations of power’ (Mumby, 2001: 601). Recognising this, I advocate an approach in which leader agency is acknowledged to exist, but in which it is balanced by a view which takes fuller account of the agency of other organisational actors, and the degree to which this agency is complicit in the construction of leader agency and action. Greater attention is therefore placed on the positive value of dissent and resistance, and on the notion of followers as knowledgeable and proactive agents with multiple prospects for action and deep vestiges of power at their disposal.

The view of leadership that emerges from a processual communication perspective is more inclined to see it as an unstable, continuously evolving social construction embedded in what Gergen (2010: 57) has characterised as ‘turbulent streams or conversational flows’. There are no finished, static entities: rather, there is only endless process (Hernes, 2008). Once leadership is conceived in these terms it ceases to be a discrete ‘event’, an observable interaction within clearly bounded organisational structures or a uni-directional flow of influence in which A has a causal impact on B. Rather, it is a communicatively organised, fluid process of co-orientation and co-construction between myriad organisational actors, whose ‘essence’ varies of necessity between each occasion of its occurrence. There is no essence of leadership waiting to be discovered, and then summarised in formal definitions, or lists of competencies and desired behaviours that are torn from particular social, organisational and temporal contexts (Ford and Harding, 2011). It follows that discursive closure is neither a desirable outcome of leadership practice nor of leadership theorising. Theories and research methods (such as laboratory experiments) that seek intense precision in their approach can become so constrained by the need for ‘tightness’ in the specification of dependent and independent variables that they may be blindsided to leadership’s most distinctive feature of all: it is inherently protean, and therefore lacking in any clearly defined and context-free essence.

This suggests that leadership theories and practices should place more stress on the promotion of dissent, difference, and the facilitation of alternative viewpoints than the achievement of consensus, or the promotion of an organisational view wholly originating in the perspectives and values of formal leaders. This is not to imply that those in formal leadership positions should avoid seeking consensus, or refrain from advocating particular points of view. But it is to argue that they should also recognise the productive potential of dissent, be wary of excessive agreement (particularly at an early stage of decision making) and value critical ideas that emanate from non-leaders. It follows that followership could be conceived in terms of differentiation and alternative positioning, while leadership is seen as those practices that facilitate such creative expression. Overt consensus is likely to mark covert dissent, since it is unlikely that followers will ever feel completely free to express the full range of their disagreements with leaders. The illusion of such consensus can therefore be held to denote leadership practices that are insufficiently sensitive to follower feedback, rather than a rational endpoint of healthy information exchange processes. While leadership actions that only facilitate the creative expression of divergent viewpoints would be likely to undermine cohesion in a manner at least as destructive as that delivered by excessive conformity, the approach outlined here simply argues that leadership should be rebalanced so that more (but not exclusive) emphasis is placed on communication processes that validate dissent. This would help to ‘promote empathy for differing ideas, opinions, and worldviews’ (Eisenberg, 1994: 282). Eisenberg goes on to suggest that ‘Perhaps the greatest obstacle to progress in most organisations is the stubborn belief on the part of those in power that their view of organizational reality is the one, correct view’ (p. 282). I would argue here that this view of the ‘correctness’ of a particular organisational view is also often found among non-leaders, and that the invitation to embrace dissent and show empathy therefore needs to encompass actors at all points of the power-powerless continuum.

The contribution of Habermas

Here, I wish to consider how this stress on constitutive interaction and co-construction between leaders and non-leaders can be advanced by considering the work of Habermas. I do so because of Habermas’s theoretical focus on communication, and because other leadership theorists (e.g. Fryer, 2011) have suggested that his approach can resolve some of the issues I discuss here. As argued below, this is not a conclusion I would share.

In key writings, Habermas has focused on communicative action and discourse ethics, developing such notions as ‘ideal speech acts’ and considering how these might promote more rational human behaviour and forms of communication. What he saw as ‘the autonomy of the individual, with the elimination of suffering and the furthering of concrete happiness’ (1974: 254) was central to this preoccupation, and became even more manifest in his later work on communicative action (1984, 1987). For Habermas, communication is bound up with attempts to create the shared understandings and cooperative relationships that underpin enduring forms of social organisation. Morrison (2011b: 159) sees such an approach as a means whereby leaders sensitised to complexity theory can re-energise ‘employees by valuing them as human with freedoms, voice, equality and openness to participation’.

Habermas’s (1984, 1987) notion of communicative action is particularly pertinent for this discussion. Central to this is the idea of the ideal speech situation. This puts a particular stress on how validity claims are raised and the degree to which they may be challenged. All ‘speech acts’ invite a listener to accept a person’s authority to raise issues, put some trust in the accuracy of the speaker’s content and have some conception of what the speaker hopes to achieve by it. But it also follows that people have the right to query such claims. In leadership terms, shared understanding could be envisaged as a goal of collaborative action. But this will only be achieved to the degree that followers respect the speech acts of leaders in the terms described here. They therefore have the right to query the validity content of a leader’s communication, thereby embarking on a process of negotiation to construct some kind of shared meaning. From this perspective, a leader’s ‘vision’ is there to be openly challenged rather than blindly acclaimed (Clifton, 2012). Habermas would acknowledge that disagreement inevitably results from such debate, but then assumes that it will be mediated constructively through the normal processes of human communication. Critically,

Rationality carries with it connotations based ultimately on the central experience of the unconstrained, unifying, consensus-bringing force of argumentative speech in which different participants overcome their merely subjective views and, owing to the mutuality of rationally motivated conviction, assure themselves of both the unity of the objective world and the intersubjectivity of their life world. (Habermas, 1984: 10)

Implications for practice

Among the major implications for practice of this approach, I would highlight the following key points:

The context in which leadership is practiced is critical. There is little point in implementing generic (and usually expensive) forms of ‘leadership development’ that fail to recognise this. Development must always be situation and organisation specific, if it is to have any real impact on an organisation’s fortunes (Tourish, 2012).

Conclusion

I suggested at the outset of this article that traditional models of leadership can be viewed as further examples of ‘The God That Failed’. Much was promised, but little delivered. A continued stress on unbridled leader agency is likely to produce further imaginary Gods who fail to meet the impossible expectations of their followers. Arguably, the greater the euphoria that greets a new leader’s appointment the more quickly disillusionment is likely to set in, as fantasy clashes with reality. Yet this does not mean that the potency of leader-centric visions has diminished. As Lipman-Blumen (2008: 40) observes, it is often the case that ‘crisis provokes followers to turn to God or human leaders willing to play God… Followers, shaken to their foundations by a crisis for which they have no ready answers, seek protection from an all-knowing, strong leader’. The dangers are considerable and include the potential for a swing to authoritarian forms of leadership in both organisations and wider societies.

A different view of agency is central to any reimagining of leadership that can help avert such destructive outcomes. It is one that confronts the tendency to see almost all power and agency as vested in the hands of a few leaders and that depicts other organisational actors as the more or less powerless recipients of leader action. This challenges the hope that a leader will emerge who offers a transcendental sense of purpose and promises to save the planet, while somehow also doubling GDP. In doing so, it recognises how crucial agency is when it is vested in non-leaders, acknowledges the productive potential of dissent and sees leadership and followership as co-constructed phenomenon embedded in fluid social structures that we have barely begun to understand.