Abstract
Leadership scholars are beginning to understand leadership as a distributed phenomenon, produced in interaction and emerging in social situations. Although this perspective has contributed to understanding leadership processes in more detail, it has also been noted that its proponents have largely neglected power and asymmetrical hierarchical relations. In this paper, I address these issues by drawing on Erving Goffman’s notion of frame analysis. Through detailed analysis of the interactions in a core-values session, I show how leadership processes that appear to be distributed and emergent from the participants’ framework appear orchestrated when understood from the manager’s framework. The analysis reveals how power asymmetries operate in the framing of the situation, and how the experience of leadership differs among participants. Talk, text, tools, and movements in time and space all contribute to establish frameworks, and differences in access to these modalities show power asymmetries. The paper highlights how the experience of leadership is framed and how power asymmetries constitute this framing. It thereby contributes to multimodal, constructivist theories of distributed leadership by showing how leadership is simultaneously emergent, distributed, and orchestrated.
Introduction
Imagine a couple kissing in the street. At first you may appreciate the romance, but once you shift your gaze and notice the director and the camera, the situation changes abruptly and becomes less romantic. Two lovers are transformed into two actors simply doing their job.
Now, imagine a manager at a meeting, presenting the organization’s new ‘core values’ to subordinates. The new values are not well received, subordinates resist, and everyone at the meeting, including the manager, eventually agrees that the new core values are not good for the company after all. At first, this may appear to be a failed attempt at leadership. But what if you knew that the manager had orchestrated all of this beforehand? Perhaps it was a case of successful leadership after all?
My goal in this paper is to develop an understanding of leadership in which I consider the nuances and multiple layers of social situations, such as those described in the stories of the actors and the manager above, thereby contributing to the nascent literature of leadership as a distributed and emergent phenomenon.
The academic understanding of leadership and the everyday understanding of leadership have traditionally focused on individuals: on leaders and their traits, styles, or skills (Carroll et al., 2008; Tourish, 2014). But recent developments in leadership research have approached leadership as a collective effort of leaders and followers (Denis et al., 2012; Raelin, 2016a) – as a collaboratively produced, distributed, and emergent phenomenon highly dependent upon situation and context (Crevani et al., 2010; Ladkin, 2010). The roles of leader and follower are theorized as precarious, because there is room for identity distance, negotiation, and resistance (Collinson, 2006; DeRue and Ashford, 2010; Uhl-Bien et al., 2014). Moreover, the outcome of leadership processes is understood as a collective accomplishment, in which the outcome is the result of coordinated action, conflict, and tension (Crevani, 2018). Thus, leadership is understood as emerging from social interaction and distributed among several actors.
Although these perspectives have focused on leadership as emergent and distributed, most contributions ‘tend to overlook power relations and questions related to formal positions’ (Denis et al., 2012: 266). This neglect is particularly significant, as most leadership studies occur within established formal hierarchies (Sveningsson et al., 2012), and there is reason to assume that leadership processes that can be observed in everyday interaction depend on such situational circumstances as formal structures and routines (Larsson and Lundholm, 2010). In many situations ‘there exists an obligation or a perceived right on the part of certain individuals to define the reality of others’ (Smircich and Morgan, 1982: 258; emphasis in original). Thus, whereas emergent and distributed perspectives on leadership have emphasized that leadership is shaped in interaction, ‘not all organizational actors are equal when entering interactions’ (Denis et al., 2012: 266).
While appreciating leadership as a distributed and emergent phenomenon in this paper, I also consider how the asymmetrical nature of power relationships is reproduced in a bureaucratic context, by highlighting the opportunities for some individuals to frame the experience of others. By drawing on Goffman’s (1974) notion of frame analysis, I draw attention to the way leadership is simultaneously experienced as emergent, distributed, and orchestrated, and how time, space, tools, text, and talk establish and maintain these different frames.
I explore a session conducted in a firm I have called Allied, in which a new set of core values was being introduced. Managers were expected to present these core values and gain subordinate support for them – a context in which leadership is expected (by the manager at the focus of the study and by his subordinates), yet is constantly renegotiated and fluctuating. These fluctuations are not merely an effect of emerging social interaction, however; as this study demonstrates, the manager has orchestrated them to some extent. By analysing the core-values session in terms of frameworks (Goffman, 1974) – definitions of the situation – the distributed and emergent property of leadership is problematized. To theories of distributed and emergent leadership, I suggest that leadership can be understood by analysing how direction and identities emerge in ongoing interaction, and how the experience of these phenomena is frame dependent and differs between actors. Moreover, I suggest that power can be analysed in terms of asymmetrical access to framing. Leadership and the framing of the experience of leadership are accomplished in multiple ‘modalities’ (Fairhurst and Grant, 2010), by which actors use language (talk and text), material tools, and access to time and space (which I have referred to as time–space throughout this paper). From this, I argue that leadership should be understood as simultaneously distributed, emergent, and orchestrated.
The paper proceeds as follows. First, recent developments in leadership theories are reviewed, with particular focus on the way leadership has increasingly been understood as a distributed and emergent phenomenon within the leadership-as-practice approach. Then Goffman’s (1974) notion of frame analysis is explored in relation to its relevance for leadership research, and I specifically relate this perspective to recent developments regarding multimodal approaches to leadership practices. The case is then introduced, together with a discussion of methodology. In the ‘Findings’ section, the core-values session is explored in detail. The analysis proceeds by illustrating how the situation appears from a primary framework, yet how the participants contribute to a keying of the framework and how the leader fabricates a framework. Thus, leadership is distributed and emergent from one perspective, but from another perspective – that of the leader – it has been carefully orchestrated, and he reproduces the image of his leadership as involving, democratic, and laissez-faire. The analysis shows how talk, text, material tools, and movements in time–space produce framing and accomplish leadership. The paper concludes with a discussion of the contributions of frame analysis to theories of distributed leadership, primarily focusing on the value of acknowledging multimodality in terms of talk, text, tools, and time–space, and to power in terms of asymmetrical access to practices.
Distributed and practice perspectives on leadership
Traditional approaches to leadership have focused on individual leaders and their traits, styles, or relationships with a group or team, with their adherents often taking formal hierarchies for granted and understanding power relationships as unproblematic. Contemporary studies of leadership have challenged this view. Leadership scholars have increasingly turned their attention to ‘leadership not as a property of individuals and their behaviors, but as a collective phenomenon that is distributed or shared among different people, potentially fluid, and constructed in interaction’ (Denis et al., 2012: 212). This literature is diverse and diverging, with many overlapping concepts, including leadership as practice, plural leadership, distributed leadership, shared leadership, collective leadership, and collaborative leadership. The roots of these perspectives can be found in mid-20th century studies: studies of group constellations, group interchange, and in the notion of empowerment (for an overview, see Fitzsimons et al., 2011). The term ‘distributed leadership’ has been attributed to Gibb, who argued in 1954 that leadership can be focused (the responsibility of a single individual) or distributed, (‘when two or more individuals share the roles, responsibilities, and functions of leadership’; Carson et al., 2007: 1218). Although seldom acknowledged in mainstream studies, distributed leadership and notions of relational interdependence have a long history, which Mary Parker Follett (1924) acknowledged nearly a century ago in her idea of ‘power with’ as opposed to ‘power over’.
Recent studies of distributed leadership have evolved in two main directions: post-heroic and constructionist. Post-heroic leadership (Eicher, 1997) takes several forms: notions of shared leadership (Pearce and Conger, 2003) and plural leadership (Denis et al., 2012), in which leadership is characterized as moving towards more inclusive and democratic forms and the ‘need to distribute the task and responsibilities of leadership up, down, and across the hierarchy’ (Fletcher, 2004: 650) is underscored. Similarly, Sveiby (2011) discusses collective leadership as distinct ‘models’ that can be applied in different contexts. Within this direction, distributed leadership is understood as a distinct form of leadership that can and should be implemented in organizations – by changing leaders’ behaviour, leaders’ attitudes, or organizational structures, for example.
A second direction taken by distributed leadership scholars assumes a constructionist stance. Social constructionist leadership researchers (Fairhurst and Grant, 2010) have challenged conventional theories of leadership on a more fundamental level. Rather than understanding distributed leadership as a new way of doing leadership, constructionist scholars encompass a more radical, ontological shift (Crevani et al., 2010). According to social constructionist views, leadership is, by definition, a process of co-construction involving leaders, followers, and context (Ladkin, 2010) and is therefore contested and fluid. Even in situations in which heroic leaders seem to arise and dominate, these scholars assume that these leaders depend on interactions, institutions, subordination, and on the reproduction of power relationships in social interaction. Thus, leadership is always distributed and dependent upon social processes.
The nascent leadership-as-practice approach explicitly engages in this theoretical development, recognizing leadership as a relational and collective phenomenon, dependent upon interaction and situated in a particular temporal, spatial, and cultural context (Carroll et al., 2008; Crevani et al., 2010; Raelin, 2016b; Uhl-Bien et al., 2014). Leadership-as-practice researchers explore how leadership emerges and develops in everyday experience, in which mutual interaction continuously shapes and reshapes relations and identities (Raelin, 2016a). Distinctive of this approach is a shift in interest from the actors to the process of interaction to the effects of interactional dynamics (Chreim, 2015; Simpson, 2016). Ladkin (2010) suggests, for instance, that leadership emerges in situations in which there is a confluence of understanding of the context, of the role taking of leaders and followers, and of a common sense of purpose. Larsson and Lundholm’s (2013) research demonstrates how everyday conversations produce collective identities that facilitate and direct future actions by those engaged in the interaction, and Simpson et al. (2017) show how ‘turning points’ in conversations recall past actions and provide an impulse to change, thereby indicating and prescribing a change in the flow of action. Similarly, Crevani (2018) identifies ‘direction’ as a core element of leadership processes. In conversational interaction, among group members collectively, and without decisive influence from any single actor, they produce a common sense of direction, clearing the way for future action. What happens in interaction is a product of the past and a way of opening possibilities for future action and closing others.
Leadership-as-practice approaches have thus provided rich insight into the ways in which leadership is distributed and how the experiencing of leadership emerges in interaction. Yet by focusing on fluidity and emergence, these approaches may neglect power and pre-existing organizational hierarchies. Denis et al. (2012: 274) argue that ‘scholarship in the field of leadership does not necessarily gain by moving from a view of leadership as individual heroism toward an equally naïve democratic ideal in which leadership is an organizational quality shared by all’. Thus, leadership researchers need to consider the inherent power asymmetries and hierarchy in social structures, because, ‘formal organization truncates the leadership process’ and provides ‘a blueprint of how the experience of organizational members is to be structured’ (Smircich and Morgan, 1982: 259).
This structuring of the experience is the focus of Goffman’s (1974) notion of frame analysis. Whereas approaches such as Smircich and Morgan’s (1982) have highlighted the way leaders frame and influence past and future actions of group members, Goffman (1974) accentuated the notion that experience is neither unambiguous nor entirely controllable. Different frames can exist simultaneously, rendering the nature of experience and power relations relative to particular frames, and the nature of leadership as frame dependent.
Bringing the situation back in: Frame analysis
Sociologist Erving Goffman is known mainly for his dramaturgical approach to understanding social interaction. In his later work, however, he shifted metaphors from the stability of the scene, with back and front stages, to flickering images reminiscent of cinema or television (or indeed YouTube), wherein various frameworks are layered upon each other – a more fragile and fluid image of the structuring of situations. To disentangle these frameworks and transformations is the imperative of frame analysis (Goffman, 1974). Goffman is rightly considered a theorist of the stability of social order, yet he does pose the question of how and why social order is (or is not) maintained – how and why frameworks hold up or break down. Goffman, albeit not generally recognized as a power theorist, thus offers valuable insight into power writ small: Power asymmetries in everyday interactions form the heart of his theorizing (Jenkins, 2008). As I discuss in what follows, an interest in the stability of social order, seen through Goffmanian lenses, is an interest in the maintenance of social order and thereby related to questions of power.
Frame analysis
Frame analysis takes as its starting point the notion that definitions of situations have a relative stability (Goffman, 1964). Frameworks – definitions of the situation – exist beyond the perception or sensemaking of individual participants. Rather, they are predefined by society: Participants have merely to assess the primary framework that applies to the situation and act correspondingly (Goffman, 1974). Members of any society must develop a basic social competence: the ability to assess situations correctly according to institutionalized conventions and accommodate their behaviour accordingly. Faulty assessments are met with sanctions, as anyone at a funeral, behaving as if it were a wedding, could attest. In a business context, different meetings and locations call for different behaviours (Schultz, 1991; Schwartzman, 1989): Boardroom meetings require different behaviour than the marketing department’s brainstorming sessions do.
This is not to say that assessments of situations are unambiguous. A practical joke, for instance, is based on the assumption that the victims will assess the situation in a particular way. To the audience and the prankster, the situation is entirely different. Once revealed, the victims will also, likely with a shock, snap out of their framework and reassess the experience from the new framework. In order to analyse how such transformations occur, Goffman (1974) introduced the notion of a primary framework, which can be understood as the immediate answer to the question, ‘What is it that’s going on here?’ (p. 25). Such immediate definitions are often correct, ‘and why not, since social life itself is often organized as something that individuals will be able to understand and deal with’ (p. 26).
Frameworks are possible to manipulate in various ways. Frame analysis also accounts for the fact that different frameworks can exist simultaneously and are not always apparent to all participants. Goffman discusses two main types of transformations of primary frameworks: keyings and fabrications.
Keying exists when an activity, ‘already meaningful in terms of some primary framework, is transformed into something patterned on this activity but seen by the participants to be something quite else’ (Goffman, 1974: 44). Children playing house is a keying of family life; boxing as a sport is a keying of people beating each other up for personal reasons. Teaching often involves keying: Take the solving of business cases, in which students simulate real life decision-making by placing themselves in the role of actual decision makers. The essence of keying is that all participants are consciously operating within the same framework; they are all ‘in on it’.
Fabrications, on the other hand, are ‘the intentional effort of one or more individuals to manage activity so that a party of one or more others will be induced to have a false belief about what it is that is going on’ (Goffman, 1974: 83). Fabrications can be more or less benign. A surprise party is likely a positive experience for the unknowing subject, whereas practical jokes are more ambiguous. A psychological study in which participants think they are part of one kind of study but are actually part of another kind, is an intricately developed case. In Stanley Milgram’s (2005) famous experiments, participants who thought they were involved in an experiment on learning were, in fact, subject to fabrication, as the experiment was about obedience to authority. Fabrications are, by definition, deceitful, as they rely on active manipulation of some agent. They normally require preparation and minimally require the fabricators to ensure that the other participants remain in their primary framework.
Frame analysis and distributed leadership
These observations on frameworks structuring experience have significant implications for social constructionist studies of distributed leadership. When experiencing leadership, participants in a situation (and leadership researchers) do so from a framework. (There has to be something making researchers believe that leadership is happening.) At times researchers adopt a different framework than the actors do; leadership as experienced by actors may amount to little more than fantasy in the eyes of researchers (Sveningsson and Larsson, 2006), but sometimes an actor may not experience leadership, but the researchers do (Sveningsson et al., 2012). This distinction becomes particularly salient in leadership-as-practice approaches, which are explicitly distanced from leadership reduced to something that is merely about participants’ perceptions (Simpson, 2016).
Several leadership researchers have highlighted the ways in which language use accomplishes leadership (Crevani, 2015; Larsson and Lundholm, 2013; Simpson et al., 2017). Talk initiates action, shapes directions, and is thus fundamental to the collective accomplishment of leadership. Most contributions to social constructionist leadership research remain ‘monomodal’, however, and primarily focus on language use. ‘Multimodal’ approaches engage with a broader repertoire of social phenomena, including language use, but also ‘the use of space, the body, clothing, technology, and so on’ (Fairhurst and Grant, 2010: 190). Fairhurst and Cooren (2009) highlight how materiality, from small objects (such as badges and clothes) to massive structures (such as New York’s Twin Towers), enables leadership and leaders to emerge. Objects and physical space actively co-constitute leadership and are ‘defining components’ of leadership practice (Spillane et al., 2001: 26).
Multimodal approaches thus suggest that language use is but one of several processes. Humans, institutional settings, and material objects co-produce the experience of leadership. Therefore, practices of all such modalities are involved in accomplishing leadership as social action, with directing effects on those involved and including a distribution of leader and follower identities (Crevani, 2018; DeRue and Ashford, 2010; Sergi, 2016). In Sergi (2016), a document plays a core role for assembling a development team and providing focus to their interaction, thereby directly affecting the direction of the group’s attention and subsequent actions. Similarly, Carroll (2016) demonstrated how leadership emerges in team interaction by showing, among other things, how physical space and the orientation of the furniture in an office contributes to the shaping of team relations and inter-team relations, and how a koosh ball becomes a cue for team members to disrupt their workflow and enter into a team meeting by moving their chairs.
A frame analysis approach to leadership deals with social situations in a similar manner, as actors, objects, and institutions all play a part in constituting the organization of experience (Goffman, 1974). It will also problematize the role of the primary framework involved, however. Such aspects as identities or direction (Crevani, 2018; DeRue and Ashford, 2010; Sergi, 2016) are always relative to a particular framework and must be analysed with potential transformations in mind – as in the imaginary situations of the kissing actors and the orchestrating manager presented at the beginning of this paper. Thus, central questions for a frame analysis are not ‘Did leadership happen?’ or ‘How did leadership happen?’ but ‘Why was it experienced as leadership?’ ‘What type of leadership was it experienced as?’ and ‘Who was experienced as leader or follower?’ Answers to these questions depend not only on primary frameworks, but also on keyings and fabrications.
Method
Developments in leadership research towards distributed conceptualizations of leadership present methodological challenges to researchers, as they force them to move beyond earlier gold standards of surveys and interviews to the study of actual interactions and leadership processes (Crevani et al., 2010; Larsson and Lundholm, 2010). Such qualitative methods as observation, conversation analysis, and ethnography become tools of the trade for researchers interested in leadership practices (Kempster et al., 2016; Sutherland, 2016).
The case presented in this paper is part of a larger study of Allied Technology Inc. 1 The general goal was to study managers and leadership in situ, by capturing everyday interaction and work processes. The fieldwork was undertaken over the course of approximately six months and consisted primarily of interviews, observations, and shadowing. Initial interviews with top management and people in such support functions as Human Resources provided general information about the firm. After these initial contacts, I approached five managers, three of whom I shadowed – a choice based on availability and timing. Each intense study consisted of an initial two-hour in-depth interview with each manager, two to four days of shadowing, interviews and informal talks with the members of each manager’s team during observation and in the days immediately following, follow-up observations of further meetings, and a concluding interview and debriefing session with each manager. Thus, the study generated first-hand data on interpersonal interactions and the flow of everyday work at Allied. I was also able to obtain information on the way managers and their subordinates experienced various situations. Through these studies, I developed a detailed understanding of Allied and the everyday work of each manager I shadowed.
The data in this paper were derived mainly from a specific situation: a four-hour core-values session held by Kel Varnsen – one of the managers I shadowed. The data comprise my observations before and during the session and the interviews and informal conversations before, during, and after the session. Field notes were taken as close to verbatim as possible, because audio recordings were not always allowed. Interviews and some meeting observations (including the discussion in Scenes 2–4 described later) were recorded and transcribed verbatim.
The first step in the data analysis was a description of the event based on the field notes. These data represent, as closely as possible, the unfolding of events and the actors’ language usage, although an ‘unavoidable aspect of fieldwork’ (Van Maanen, 1979: 543) is a transformation that naturally occurs because the wording used influences future interpretations. From this linear description of the events as they unfolded, a need for further empirical material addressing actors’ intentions, motivations, and interpretations was identified and described by drawing on other data – primarily interviews and informal conversations during shadowing. A narrative could then be constructed, which served as the main object of analysis. During the analytical process, data were revisited several times in order to complement the story with greater detail and to double-check the actors’ interpretations. The main analytical focus was four scenes, each illustrating key moments in the interaction; all other interactions have been edited out to accommodate journal space constraints. Detailed analysis was made with particular emphasis on the way the experiencing of leadership could be said to emerge from the situation. In this, theoretical concepts were employed in order to make sense of the data and to facilitate higher level theoretical interpretations. The notion of ‘direction’ (Crevani, 2018) was imperative in identifying leadership as it unfolded in the verbal interaction. Identifying different framing cues constituted a second focus and explicitly addressed Goffman’s (1974) question: ‘What is it that’s going on here?’ Examples of cues include an actor’s suggestion about certain interpretations, a challenge to interpretations, and an actor’s influence on the unfolding of events by introducing a new topic to a conversation. Moreover, contextual cues such as objects, body positioning, and office design were considered.
This research is conducted in an interpretive tradition, in which the researcher’s subjectivity plays a significant role. The empirical story told is a montage (Denzin and Lincoln, 2011), a narrative constructed and presented by the author (me). The empirical validity of the study is based not on statistical representativeness, but on observations of ‘a localized and time-specific practice and aspire to see it as part of – and as an illustration of – a more general pattern of practice’ (Kreiner, 2012: 402). Whereas the event, its context, and its participants are, of course, unique, the situation is by no means idiosyncratic. Rather, it is an example of the type of situation that often occurs in organizations, illustrative of processes from which we can learn something about these types of situations in general (Flyvbjerg, 2006), for understanding the particular can contribute to an understanding of the universal (Van Maanen, 2011) – in this case how framing is accomplished. Whereas this study involves thorough empirical work, comprising many hours of interviews, several days of intense shadowing (Czarniawska, 2007), and a carefully crafted empirical narrative, its relevance derives mainly from the theoretical insights provided by interpretations.
In the following, the findings of the study are presented. First, the firm and the main character are introduced, followed by a narrative encompassing four scenes. Throughout the narrative, initial interpretations of what it is that is going on are presented. These interpretations are subsequently developed in the ‘Discussion’ section.
Findings
Allied is a medium-sized company in the high-tech industry. The workforce consists almost exclusively of engineers, working with advanced software and hardware applications in product development. The company has been highly successful and has grown rapidly. At the time of the study, top management was implementing a programme of cultural change, encompassing a new set of core values to be internalized by employees.
Kel Varnsen is the manager of a development team of 10 people. He has worked at Allied for over 10 years, primarily in leadership positions, predominantly in product development. He is talkative, charismatic, and outgoing, with a ready joke and a friendly remark and seems to carry his leader role lightly. Casually dressed in jeans and shirt (no tie and only occasionally a jacket), he blends easily with the style of other employees. When describing his leadership style, he emphasizes participative decision-making and the value of employee freedom: Last summer … I took some parental leave, and before I left I made up this plan over what should be done while I was gone. When I returned, none of that was done; they hade made other priorities. … I thought they had made good priorities, totally ignoring what I had said. … So I was happy about it.
His subordinates describe Kel as ‘friendly’, ‘sympathetic’, and ‘a good manager’. The general impression is that Kel is a popular manager, which can be attributed to his laissez-faire leadership style, in line with the Allied culture.
Scene 1: Preparation
The focal situation is a half-day session on core values. Top management has developed a new set of core values to replace the old, and middle managers are to lead sessions in which these values are presented, discussed, and hopefully firmly established. The day before the session, Kel sits in his office with the door closed and skims through the presentation he is supposed to give. He reduces it from 30 slides to eight, while directing comments at me and at his administrative sidekick, Louise: ‘What’s this? Am I supposed to present this?’ and ‘I refuse to show this one’. He admits that the presentation will not be well prepared: ‘I’ll do it from heart; it’ll have the same effect or even better than this rehearsed blah-blah-blah’. He adds: ‘[1]
2
Some of the questions we’ll talk about tomorrow will likely evoke discussion. One in particular, I don’t know if I’ll have to get rid of it, because I really don’t agree with it’.
Yet Kel does believe in having core values: They can ‘provide guidelines’, but they won’t work if they are trying to ‘brainwash’ the employees.
Right from the outset Kel has taken control of the content of the session. He is clear that he will not say anything to the group that he cannot defend. Several of the slides removed were ‘facts’ that he disagreed with; he is hesitant about certain questions they are supposed to discuss, and top management’s position on them. He has already started to transform the direction of the session into what he considers a more desirable one.
Scene 2: The session starts
The next day, Kel and I head off to the session. ‘Switching to bullshit mode!’ Kel jokes as we leave his office. He has assembled 14 employees, about half of whom are his subordinates, in a posh conference area. The session begins with Kel standing at the head of the oval table where the group sits. Slides are projected onto a screen. The three new core values – all together, go global, and be honest – are described as ‘deeply held beliefs which indicate the ideal expectation of how everyone in the organization behaves’. The presentation is over in five minutes. Even Kel seems surprised, pondering in front of the audience: [2] ‘Did I remove that many slides…?’ After the presentation, the exercise begins; it will consume the remainder of the four-hour session. The exercise is a Trivial Pursuit™-type game, played by the participants divided into four groups. The aim of the game is to familiarize the employees with the new core values, to make them understand their meaning correctly, and to encourage them to use them in decision-making. In preparation, each group goes to a separate room and works through the questions, chooses one of four answers, and motivates this with a core value.
The conference area comprises several small rooms for group work and one larger room, where the introduction and reassembly occurs. Activities are visible through glass walls and doors. The groups go to their rooms, while Kel and Louise go from room to room, listen to the discussions for a while, ask if everything is ok, and nod. There is little discussion; group members seem hesitant and uncomfortable.
About an hour later, the four groups reassemble in the main conference room. They are supposed to play the board game and score points by identifying the correct answers. The idea is to bring some dynamics to the discussions by generating variety in which questions to discuss and by having the groups compete. The rules are somewhat unclear, despite, as Kel says, that he and Louise ‘have received training’ on the game. And, in any case, no one has brought any dice. So Kel decides to drop the game and instead chooses a few questions for discussion that have seemed difficult during the group exercise.
Scene 3: Overtime
The discussion is initially hesitant; only a few words are exchanged across the table. The core values are discussed only when Kel reads aloud the ‘preferred answers’ (as predefined by top management). When there are controversies and discrepancies between the group’s answers and the preferred answers, Kel usually gives way to group opinion rather than enforcing his own or the preferred answer: [3] Kel: This one’s fun … ‘You and your team are under heavy workload. Yet you have agreed to introduce a new colleague. This extra task will increase the risk of your missing an important project deadline. How do you handle the situation?’ And what have we got here…? [Kel reads the answers from the groups:] 3 … 3 … 3 … 3 … All agree that the answer is 3, ‘Inform your manager.’ [4] Person A: Except those who have 1 put within brackets… [5] Kel [reading out answer 1]: ‘Keep working in the project, put in extra time and effort to do the introduction.’ Anyone want to guess the preferred answer? [6] Person B: Probably 1. [7] Kel: Yes. [8] Person B: But that doesn’t work in the long run. [9] Kel: No, it really doesn’t. [10] Person C: Then it ends up, like, you never accept to do anything out of the ordinary. [11] Kel: This is the one question I don’t like. I understand the point with answering 1: that we have to work, that if you’ve agreed to something you have to follow through. But what happens when you’ve said something? Well, everyone can be wrong, that’s my opinion.
Kel is supposed to lead the group in the direction prescribed by top management; after all, this is about reinforcing the new core values and what they imply in terms of action. But from the beginning, Kel signals that he may not agree with the preferred answer ([3] ‘This one’s fun’). When he reads out the answers, the group members (given voice by B) realize they have all given the wrong answer [4, 6] – and they seem to be aware of this [6], for what they see as good reasons [8, 10]. Kel immediately agrees with the group [9] and emphasizes that his position was clear at the outset [11]. Thus, the direction emerging from the interaction is not the one prescribed. Moreover, by emphasizing that he already held this position Kel ensures that this change of direction is not interpreted as being created solely by the other group members. Thus, the emerging direction is a keying of the primary framework: Instead of enforcing and gaining support for the company line, the participants together reinterpret the new core values in a way that makes sense to them. Consistent with his leadership image, Kel follows the group. The emerging leadership direction is thus partly influenced by Kel, while simultaneously constructed in interaction as Persons A, B, and C contribute to the direction with their comments.
However, while the groups were holding their discussions in separate rooms behind closed doors, Kel and Louise were sitting alone talking about the questions, including the one described here. Kel concluded that neither he nor the groups could subscribe to the preferred answer. Thus, whereas leadership emerges in the group interaction in a direction inconsistent with the preferred answer, this ‘emergence’ is well prepared. Although Kel cannot predict the exact direction, of course, the framing he has established allows the group to influence the direction. His opening statement [3] invites the keying, and his concluding statement [11] confirms it.
Scene 4: Skunkworks
‘Skunkworks’ are development projects unsanctioned by management and undertaken during working hours – an established way of working at Allied, and highly valued by the engineers. Although unofficially tolerated until now, it is clear in the core-values presentation and the game that top management is no longer going to accept skunkworks. When the group reaches the skunkworks question, the debate heats up. Person C, who has been employed at Allied for many years, is defiant: one core value should be ‘skunkworks are great’. Kel notes that [13] ‘this is a big change in how Allied is supposed to work … Skunkworks aren’t accepted any more’. This statement is met with several critical comments: [14] Person D: There’s always some skunkworks. Sometimes you have to spend extra time on something in order to show … [15] Person E: We wouldn’t have the 1001 model if it hadn’t been for skunkworks. It’s fatal to think that all ideas can come from the product managers. [16] Person F: You can’t ban inventiveness. Then people will quit, and ideas will suffocate. I don’t know how this is supposed to work. [17] Person C: Without skunkworks, I don’t think there will be much innovation. It’s like, what people really care about; they want to do them … usually it’s good stuff. [18]: Someone intervenes: But if it puts projects at risk? [19] Person C: That’s what I said; you can’t formalize this. [20] Kel: So it’s okay to put projects at risk as long as you don’t formalize it? [21] Person C: No, on the contrary! It’s okay to do skunkworks as long as the project isn’t at risk. [22] Kel: So, how do we handle this, Louise? We’re supposed to lead this core-values session, after all. [Laughs followed by some jokes.] It’s not okay with skunkworks, that’s what we’re saying now. [23] Person C: But some is okay … [24] Kel: Oh no! It wasn’t me who did that Visual Basic program to demo that 202 model. Oh no! [26]: Any second now I’ll throw you out of this room! [All laugh.]
Here the direction in which events unfold seems highly dynamic and fluid; leadership emerges without clear leader and follower roles. Kel’s initial remark [13] is met with a barrage of protests [14–17, 19], and his attempts at maintaining the new company line, questioning what is meant by ‘putting projects at risk’ [20], receives no approval from the group. After trying to summarize the discussion [22], and again repeating the new company policy, Kel suddenly caves in and starts bragging about a skunkwork he had engaged in some years earlier [24]. This turn of events triggers the discussion again, and a number of successful skunkworks are mentioned [25]. Kel’s final comment [26] ends the discussion, but it is more for Kel to signal his surrender and conclude the discussion than it is a real threat.
As in Scene 3, the discussion is fluid, and the direction is initially unclear. Kel seems to try to maintain a leader identity in relation to the group but is challenged by C. As C pushes, Kel gives way and the discussion – and its conclusion – ends where C initially pushed it: Skunkworks are acknowledged as vital to the firm. When Kel tries to reinforce the company line [22] and C again intervenes [23], Kel chooses to follow C’s lead [24]. Eventually, then, Kel is apparently ‘run over’ by the group, particularly by C, and assumes a follower identity. Similar to Scene 3 and the keying of the framework, leadership emerges through interaction in a dynamic process; the main role seems to be played by C. It seems that Kel is forced to give in to the group and follow rather than lead. The direction that emerges is one of questioning and undermining a fundamental intent of the new core values.
Again, however, there is a backstory. In the earlier discussion with Louise in Scene 2, Kel realizes that the group will never accept the new stance on skunkworks [27]: ‘Not a chance that we can convince them that this is right’, he says. Thus, for Kel, the situation is framed differently. He initially plays a game he knows he is unlikely to win, performing the leader identity as expected from the primary framework. When C and others push in a different direction, he initially resists, re-acknowledging to the group what should be his formal message, but he then sides with the group. Thus, from Kel’s framework, his initial statements were largely an act; he was well aware of the likely direction of the ensuing discussion. As Kel plays out the situation in terms of the keyed framework, however, he seems to allow the group, through C, to decide the outcome. But it is Kel who sets the agenda; he could have chosen not to go into the skunkworks issue, just as he ignored many other questions. The game itself here plays a key role: If the game had been played as designed, the questions would be drawn randomly from a card pile. When Kel decided not to play the game (Scene 2), he created a way of avoiding questions, as he could now pick and choose which ones to discuss. Thus, he deliberately chose to bring up the question he did not ‘agree with’ [1].
Discussion
Examining this situation scene by scene reveals intricate leadership interactions, in which leader and follower identities (DeRue and Ashford, 2010) and the direction of leadership (Crevani, 2018) changes. The framework is both keyed and fabricated in the four scenes, as discussed later in terms of language (talk and text), tools, and time–space together constituting the different frameworks, and the effect on the experience of leadership (see Table 1).
Frameworks in the core-values session.
As Goffman (1974) argues, frame analysis is difficult, given that any framework depends on situational cues and the observers’ or participants’ observations. Question: ‘What is it that is going on here?’ The answer, to participants and the researcher: ‘a core values workshop in which participants are supposed to embrace and internalize the new core values’. The stage is set for the manager to exercise leadership, enabled by time–space arrangements. The posh conference area and the fact that Allied has invested heavy resources in this project indicates that it is an occasion of some magnitude. The physical layout and the way it is used contribute to this framing. The introduction occurs in a main conference room where Kel has positioned himself standing at the head of the table by the presentation screen, which creates a spatial configuration whereby everyone focuses on Kel. The presentation is a tool that co-produces this common focus, and the text on the slides provides content as it introduces the new core values, what they mean, and why they are important. The primary framework here appears to provide a staging of heroic leadership wherein the manager implements the new core values into the organization. This framework is invited by a distinct time–space (a half day in a posh conference area), by tools (a game and a presentation), and by providing specific language use (text) for new core values.
From the beginning of the session, however, Kel invites a keying to this primary framework (Scene 2). By running a surprisingly short presentation and explicitly mentioning that he changed the slides [2], he shows that he has diverged from the company line, as reinforced by his dropping the game: Tools and time management underline Kel’s casual attitude to the primary framework. Keying is sustained in Scenes 3 and 4, as talk flows and the direction and leader/follower roles change. Kel sides with the group, which eventually takes the lead, and Kel follows. Leadership is collectively accomplished and follows a ‘post-heroic’ model (Fletcher, 2004): Leader and follower identities are co-constituted and negotiated (DeRue and Ashford, 2010), and direction emerges through participant interaction, mediated by talk (Crevani, 2018). In Scene 3, for example, the verbal protests of the participants [8, 10]; Kel’s agreement [11]; and his conclusion, from which they move on [12], are all accomplished by talk. Just like the primary framework, this keying, ‘collective leadership process’ (see Table 1), occurs primarily in the time–space of the large meeting room.
Although the breakout sessions convey the idea that participants are supposed to talk and discover the meaning of the new core values, the whole idea of the main tool for this game is to find the correct answers. The tool thus attempts to limit the range of possible meanings to the core values, but through talk these meanings are questioned and subverted; the keying is accomplished in interaction. Kel and the group collectively diverge from the prescribed meanings of the preferred answers, and co-produce a new direction, which counters that of the new core values. The leadership experience in the keyed framework is one of emergence, in which roles shift, power is distributed, and talk is dominant in accomplishing this direction in terms of new meanings to the core values (cf. Crevani, 2018; Larsson and Lundholm, 2013). Whereas time–space and tools help shape the process, it is in talk that keying primarily becomes manifest: participants’ protests [14–17], alternative interpretations [8, 10, 11], and Kel’s ‘caving in’ [24] are all accomplished predominantly through verbal interaction.
But simultaneously, the situation is a fabrication. Whereas all the workshop participants are part of the keyed framework, the fabrication is, by definition, asymmetrical. In Scenes 1 and 2, Kel takes advantage of time–space in order to prepare the situation the day before and away from the participants. Some participants are sitting close to his office, and his door is closed – not Kel’s usual policy. He changes the tools that are supposed to contribute to the implementation of a new direction for the company (the new core values) and ensures that he is prepared to act on reactions from the group. The changes in his presentation (Scene 1) facilitate his display of independence from top management (Scene 2). Furthermore, in Scenes 3 and 4, Kel takes advantage of his ability to move in time–space and listen in on the groups and prepare the questions. The other participants are prevented from listening in on other groups, as it could be considered cheating in the game. The game thus contributes to defining the situation and maintaining frame, and Kel does not drop the game until after this part of the session is over. In Scene 4, Kel chooses to include the controversial question on skunkworks [1], a discussion he correctly predicts will go against the company line. He adheres to the company line for some time before caving in and siding with the group, positioning himself as a successful skunkworker. Thus, the way Kel actively contributes to shaping the tools (game and the presentation) subtly contributes to shaping the leadership experience. Whereas the direction that emerges is accomplished collectively, and whereas all participants influence its emergence, the process is also orchestrated. In accomplishing this, tools and time–space actively exclude most of the participants from experiencing the fabrication of ‘reproducing Kel as a laissez-faire leader’ (see Table 1). Kel has access to time–space and tools, and he proactively uses them to shape direction and leader/follower roles. Louise and I (the researcher) passively contribute by not exposing Kel’s orchestration. The resulting leadership experience is a reproduction of Kel’s leadership image: He appears his usual casual, inviting, listening, and laissez-faire self, an image appreciated by his subordinates, which plays well into Allied’s culture. Kel may even feel compelled to do this – as noted by Crevani (2018), leadership must be understood as an effect of past actions, and for Kel this is a way of ensuring that he does not diverge from his usual image, despite the primary framework implicating heroic leadership. In DeRue and Ashfords’ (2010) terms, the leadership schema – the co-construction of leader and follower identities – is successfully reinforced by Kel’s framing activities.
Thus, the primary framework of a ‘core-values session’ from the keyed framework of the participants appears as a ‘collective leadership process’. But for some of the participants, it is also experienced as an effect of a fabrication, reproducing Kel as ‘a laissez-faire leader’. The main contribution of frame analysis is not the observation that different frameworks exist in parallel. That insight, however, sets the stage for asking how such frameworks and layers of frameworks are accomplished. Here, talk, text, tools, and time–space are of paramount significance. Notably, both the primary framework and the fabricated framework draw heavily on time–space and tools, whereas the keying depends largely on language use (talk).
Talk becomes the medium for negotiation and for the experience of emergence and shifting leader/follower roles. Talk is a modality more easily accessible to all participants as they co-produce leadership in interaction. As noted in empirical studies of distributed leadership (Crevani, 2018; Crevani et al., 2010; Larsson and Lundholm, 2013), talk contributes to role positioning and to the experience of emerging directions. The fluid and emerging character of leadership is highlighted in talk, and monomodal approaches generally support this view.
The multimodal approach draws a number of significant asymmetries into the picture. Objects and temporal and spatial configurations all play significant roles in the accomplishment of leadership (Carroll, 2016; Sergi, 2016). Kel’s bureaucratically defined role allows him to operate in time–spaces different from those of the other participants. Kel can also influence the tools involved in the production of leadership. What is from one framework a collective process with open-ended outcome, in which identities and processes are fluid, is, from another framework, an orchestrated situation producing an experience of participation and consolidating the image of a laissez-faire leader. In this, there is clear asymmetry, but asymmetry not of the outcome of the processes – the direction produced – but the way the experience of leadership is framed. Here, non-talk modalities, particularly time–space, play a prominent role in shaping the leadership experience.
Frame analysis helps to unveil different answers to the question ‘What is it that’s going on here?’ In the core-values case, participants’ will have one answer to the question; Kel will provide another answer. There is a danger, in frame analysis just as in leadership studies in general, of privileging the point of view of the lead actors in the story (Tourish, 2014). In this case, Kel’s experience gains some prominence, as he was the one I shadowed. But we must not see Kel’s orchestration as explaining the totality of the situation; it is but one of several possible understandings. To the situation at Allied we could add the framing activities of other actors. What about top-level managers? How would they understand the keying, and what are the potential consequences of that frame breaking down? Perhaps the participants conspired to put pressure on Kel as they’d heard rumours about the new core values limiting skunkworks? Perhaps they used tools such as smartphones to coordinate their intervention, overriding the time–space constraints of the separate group rooms? Moreover, Kel caves in during the skunkworks discussion. But the other participants contribute in shaping this ‘opportunity’. There is no way of knowing if caving in was merely Kel’s backup plan, forced into realization by the group as they protested. The group is also aware of Kel’s general leadership image and from Person C’s perspective, for example, this could be utilized as a way of forcing Kel into keying the situation to a direction that C preferred.
We will never know any of this, but what we do know is that we can never be sure that the leadership we experience – as managers, workshop participants, or even researchers – is the definite answer to what it is that is going on.
Conclusions
Frame analysis sensitizes us to the need for acknowledging the fluid nature of leadership and to the power asymmetries of formal organizational structures in our understanding of leadership processes. In this paper I have given a detailed analysis of a particular, but not idiosyncratic, leadership situation and of how framing simultaneously makes it appear distributed and emergent, yet orchestrated. I thereby make two distinct contributions to multimodal, constructivist theorizing on distributed leadership: first, regarding the framing of leadership experience and second, regarding power asymmetries in the leadership process.
First, the modalities in in the framework presented here – talk, text, tools, and time–space – all contribute to the accomplishment of leadership. The type of leadership experienced primarily depends on the framing of the situation. Situations have a degree of stability (Goffman, 1964, 1974). Power over the way situations are experienced is anchored in practices of various modalities. Whereas some modalities (talk in particular) enable the fluidity of leadership, others (such as tools and time–space) serve to stabilize the leadership experience. So, whereas the ‘identity (and fates) of leaders as leaders, and that of followers as followers, is a result of a mutually constitutive interaction between the two’ (Tourish, 2014: 85), and whereas direction can be understood in terms of the ‘ongoing evolution of the relational configurations they form’ (Crevani, 2018: 90), we must not lose sight of how keying and fabrication shape and stabilize the leadership experience. The way in which the situation presents itself to the actor will strongly influence why was it experienced as leadership, the type of leadership it was experienced as, and who was experienced as leader or follower.
Second, we are not all equal when we enter a situation (Denis et al., 2012). Access to practices is partly dependent on formal roles and relationships. Leadership, then, is not an organizational quality shared democratically, but a question of contestation and negotiation, depending on power asymmetries. Yet social processes are not determined by such asymmetries. There is variability in the modalities that come into play, and different modalities show different levels of plasticity. Thus, the multimodal framework of talk, text, tools, and time–space does not offer a deterministic view on power, but one in which power is negotiated yet constrained and the different modalities make different contributions to the experience of leadership. Power in leadership processes is, to a significant extent, also the power of framing – of shaping the leadership experience. Different modalities carry different weights, and this paper suggests that time–space and tools are particularly susceptible to orchestration activities, in turn affecting framing. In this, managers have a distinct advantage by their privileged access to time–space and tools.
Much of the debate around distributed and emergent perspectives on leadership has been on issues of ontology and epistemology, and whereas such debates have value, the time is ripe for challenging leader-centric understandings of leadership through in-depth empirical studies. The goals behind frame analysis are that certain actors (e.g. managers) not be privileged and that distributed and multimodal leadership practices be acknowledged. It sensitizes our theoretical lens to the multilayered nature of reality and to the myriad possibilities of producing transformations of that reality.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.