Abstract
The concept of authentic leadership is increasingly the focus of much leadership scholarship, and many have called for a review of the basic assumptions that underpin it. Taking an interactional approach to authentic leadership (AL) and using naturally occurring workplace interaction as data, we seek to question two basic assumptions of AL scholarship, namely (1) that authentic leadership emanates from the atomized leader and (2) that there is a causal logic to it so that authentic leadership behaviours are the cause of follower outcomes. Addressing the research questions – what is the nature of the empirical phenomenon that is called AL and where can this be ontologically located? – our findings indicate that these two fundamental assumptions that underpin current AL research are not justified. Rather, what is taken to be AL is better understood as a collective and collaborative achievement, which can neither simply be attributed to the leader nor can the leader’s actions alone lead to follower outcomes.
Introduction
Over the past few years, the notion of authentic leadership (AL) has been the topic of considerable academic debate, elaboration and criticism (Gardner et al., 2011; Iszatt-White and Kempster, 2019; Sidani and Rowe, 2018). However, much of this research primarily targets conceptual and theoretical aspects, largely overlooking empirical concerns and sidestepping the important question of how – if at all – we can capture authenticity in action. This lack of engagement with these crucial issues is, of course, largely due to the fact that most empirical research studies on AL are quantitative in nature, focusing on reported instances of AL, and favouring measurements over in-depth analyses of what exactly it is that is measured (Iszatt-White and Kempster, 2019).
Following recent critical reassessments of both the construct and the emerging theory of AL, in this article, we aim to address these issues and respond to the call by Iszatt-White and Kempster (2019: 366) to provide a ‘rich and varied diet of qualitative research’ with the aim of contributing to ‘a more grounded, bottom-up, practice-based understanding of what it means to be authentic as a leader, and whether that is a useful defining characteristic’. We wish to contribute to such a grounded understanding by conducting a qualitative empirical study to explore the phenomenon that is called authentic leadership and its ontology. We do this by taking an interactional approach to leadership, which involves the close scrutiny of recorded workplace interactions. Thus, rather than analysing post hoc sensemaking or reported perceptions of authentic leadership (as in most previous research), our innovative empirical approach builds on the direct access to naturally occurring workplace interactions in which AL may be realized (Clifton et al., 2020). Through this approach, we can observe and empirically capture interactions reasonably seen as the performance of authentic leadership.
With its focus on AL in action, such an approach addresses some of the shortcomings of previous research which mainly focuses on locating (and attempting to capture) AL in individuals, behaviours and quantifiable measures. These shortcomings are also reflected in one of the most influential definitions of AL developed by Walumbwa et al. (2008: 94), who consider AL to be: [A] pattern of leader behavior that draws upon and promotes both positive psychological capacities and a positive ethical climate, to foster greater self-awareness, an internalized moral perspective, balanced processing of information, and relational transparency on the part of leaders working with followers, fostering positive self-development.
This definition encapsulates some of the issues with current AL research mentioned above. First, although it was derived a priori to any empirical work, this definition is often used as the basis on which certain behaviours are judged (often in hindsight) as AL. A second, albeit related, issue with this definition is that it is firmly based on the assumption that AL is located within one party to a leader–follower relationship – namely the leader – and that we can capture AL by asking people to evaluate their leader’s performance. The third problematic assumption underlying this definition is to view AL as a cause of outcomes that can often be observed in the followers. In contrast to this relatively linear and limiting conceptualization of AL, we argue that the notion of behaviour needs to be problematised and that these so-called (AL) behaviours need to be reconceptualized as collaborative achievements. Using conversation analysis (CA) as a methodological tool and recordings of naturally occurring workplace interaction as data, we critically engage with these problematic assumptions and provide empirical insights into the phenomenon that is called AL and its ontological location.
The article is structured as follows. First, we review existing theory on authentic leadership, and we draw attention to two of its central assumptions, notably that the locus of authentic leadership lies in the atomized leader and that there is a linear causal logic between leader behaviours and interactional consequences. Second, we review the operationalization of authentic leadership through questionnaires and clarify how we will draw on this in our study. Third, we present an interactional perspective on leadership and specify our methodological tool, namely conversation analysis (CA) which we use to analyse naturally occurring interaction. Fourth, we present the analysis of three interactional sequences, chosen on the basis of their resonance with items in the authentic leadership questionnaires, showing how these are collaboratively accomplished. This is followed by a discussion of the implications of our analysis, in which we question the assumptions of locating the phenomenon of authentic leadership in the individual atomized leader as well as the assumption that discrete individual leader behaviours are the cause of interactional consequences. We conclude our article by discussing the implications of our study for authentic leadership scholarship.
Locating authentic leadership in the leader
Although most definitions and previous studies of AL acknowledge that AL involves a relationship between leaders and followers and that context is important (Walumbwa et al., 2008), they nevertheless tend to locate the core of AL within the atomized leader. As noted by Iszatt-White and Kempster (2019: 356), authenticity is typically seen as ‘a property of that which is claiming to be authentic’, which, in this case, is the leader. Authentic leadership is described as being based on what the leader does and how he/she relates to him/herself. Similarly, Luthans and Avolio (2003: 243) claim that ‘the authentic leader is true to him/herself’, and Walumbwa et al. (2008) identify four central dimensions, all focusing on the leader as a person: self-awareness; relational transparency (‘presenting one’s authentic self to others’, p. 95); balanced processing and having an internalized moral perspective. Sparrowe (2005) develops a more interpretative perspective on authenticity, suggesting a narrative understanding of the self, inspired by Ricoeur. Being dynamic such a narrative has the capacity to integrate self-constancy and character, and according to Sparrowe (2005: 431), ‘a leader’s values and purposes are disclosed in relation to the changing events of a narrative life, rather than by what he or she says at a given moment in time’. Thus, while other organizational players are important for the development of the narrative, the resulting authentic self is nevertheless the leader’s self.
Even more critical texts seem to embrace the idea that the phenomenon of AL is located in the leader. Ladkin and Taylor (2010: 72), for instance, suggest that the enactment of a true self is a complex question of embodiment in which: Enacting one’s ‘true self’ in such situations calls for leaders to balance how they might express something of the complexity of their competing emotional and bodily reactions in a way which is experienced as ‘leaderly’ for those looking for guidance in those situations.
In a psychoanalytically informed critique of authentic leadership, Ford and Harding (2011: 475) suggest that the dominant theory holds an inadequate and oversimplified view of the subject. They argue that the notion of authenticity in this theory denies the subject’s experience of suffering, tensions and more generally, the darker side of leadership. This, they argue, leads to paradoxical consequences: ‘[o]ur reading of authentic leadership through object relations theory is that the authentic leader is one who is so totally absorbed into the organization that s/he is an object lacking subjectivity’ Ford and Harding, 2011: 475). While object relations theory offers a far more complex understanding of subjectivity and brings stronger attention to intersubjectivity, Ford and Harding’s focus is, however, still largely on the psychic constitution of the subject/leader and how this finds expression.
In contrast to this emphasis on the leader’s internal authenticity, authors embracing a more follower-centric perspective argue that authenticity must be attributed to an individual by others. In other words, regardless of the level of internal authenticity, a leader does not qualify as an authentic leader unless he/she is perceived to be authentic (Černe et al., 2014; Goffee and Jones, 2005; Harvey et al., 2006). From this perspective, authentic leadership is a relational phenomenon, in the sense that it depends on follower perceptions. For example, Černe et al. (2014) argue that for authentic leadership to be truly present, both internal authenticity and the external behavioural demonstration of it need to be present.
However, despite giving the followers an important role in such a conceptualization of AL, even from a follower-centric perspective, authors tend to assume that the core of AL resides in the internal authenticity of the leader. For instance, Sidani and Rowe (2018: 623) re-conceptualize authentic leadership as being dependent not only on the leader but also on the follower: ‘AL is not a leadership style per se, but an outcome of a process co-created by leader-follower interaction’. Consequently, they argue that: ‘Authentic leadership represents legitimated follower perceptions of a leader’s authenticity which are activated by moral judgments’. The authentic leadership process thus becomes more complicated, but nevertheless it still starts with the leader’s authentic behaviours, which are essentially actions that demonstrate a range of inner characteristics (such as values, convictions and emotions). For authentic leadership to be present, it is thus the inner authentic core that has to be both demonstrated and perceived in authentic behaviours.
In this article, it is precisely this logic that we wish to question. More specifically, we examine how the assumption that AL is an expression of leader qualities through leader behaviours resonates with the actual enactment of those activities in workplace interactions – which have been identified as AL in the previous literature.
Authentic leadership as a cause of follower outcomes
The second assumption regarding authentic leadership that we wish to critically examine concerns the supposition that leader behaviour is the cause of follower outcomes. Essentially, theories of AL seem to subscribe to a linear, rather mechanistic logic, whereby the inner characteristics of one party are expressed through behaviours. Leader behaviours are understood to be separate from follower perceptions, and so, following this logic, it is possible for followers to observe, subsequently assess, and judge the leader’s behaviours. Logically, the leadership process starts with the leader’s values which are expressed through behaviours. These behaviours are then perceived and assessed by followers, in a way that causes particular follower outcomes.
This logic is also reflected in the definition by Walumbwa et al. (2008: 94), where AL is seen as ‘a pattern of behaviours’ that promote and foster a range of positive phenomena. As such, these behaviours are positioned as causes, and the assumption is that the more these behaviours are present, the more positive the outcomes will be. For instance, as illustrated in their model of variables, Walumbwa et al. (2008: 111) test how AL, understood as leader behaviours, influences organizational citizen behaviour, organizational commitment and satisfaction with one’s supervisor. Similarly, Sidani and Rowe (2018: 629) clarify their understanding of the AL relationship through a conceptual model in which the ‘leader’s moral system’ influences the ‘leader’s modelling behaviour’, and this, in turn, impacts on the followers.
Identifying authentic leadership
Turning from logic to empirical analysis, we are now faced with the question of how to identify authentic leadership. In the most prominent theories, AL is empirically identified through the use of questionnaires in which followers are asked to use their experiences and perceptions of their leaders (often equated with managers) to assess a range of qualities (for instance, regarding leader authenticity (Černe et al., 2014) or value congruence (Sidani and Rowe, 2018)). The theory subsequently treats these assessments as referring to behaviours. However, it is to be noted that the questionnaires per se do not explicitly call the assessments behaviours.
This operationalization of AL offers a useful point of entry for our empirical work. It helps us make certain we are examining the same empirical phenomenon as the dominant theories, and it allows us to critically challenge the theoretical understanding of this phenomenon. The two main instruments for measuring authentic leadership are the Authentic Leadership Questionnaire (ALQ; Walumbwa et al., 2008) and the Authentic Leadership Inventory (ALI; Neider and Schriesheim, 2011). The ALI is publicly available in its entirety, while only eight sample items from the ALQ are available from Walumbwa et al’s. (2008) article. However, a self-assessment version of the ALQ is fully available in Northouse (2010).
In many of these questionnaires (e.g. Walumbwa et al., 2008: 113), formal position is conflated with leadership (Bedian and Hunt, 2006), and subordinates are asked to assess their supervising manager. More specifically, based on their experiences of interactions with their leader, subordinates are asked to create a generalized assessment of her/him. Conversely, in the self-assessment version of the ALQ, it is the leader him/herself who is asked to rate their own authentic leadership on a Likert scale and to respond to each question ‘honestly’ and in a way that ‘most accurately characterizes their response to the statement’ (Northouse, 2010: 236).
Despite the emphasis on behaviours in the theory, close inspection of the items of these scales reveals a certain complexity in the sense of going beyond observation of leader actions. Of the eight sample items presented in Walumbwa et al.’s (2008: 121) article, five explicitly ask the respondent to assess the internal characteristics or states of the leader. For instance, item three asks the respondent to assess to what degree the leader ‘says exactly what he or she means’ and item six, ‘makes decisions based on his/her core beliefs’. Similarly, the self-assessment version of the ALQ goes beyond behaviours and includes a certain amount of reflection. For instance, item one reads ‘I can list my three greatest weaknesses’ and item eleven reads ‘I do not emphasize my own point of view at the expense of others’. In other words, there seems to be a certain disconnect between what is identified and assessed by these questionnaires and the way this phenomenon is conceptualized in the dominant theories. We wish to emphasize that we do not take this disconnect to be a flaw in the questionnaires, but rather in the conceptualization and theoretical understanding of the phenomenon of interest.
In this study, we will take these operationalizations of AL in the questionnaires as a starting point for an empirical exploration of how AL is accomplished in everyday workplace interaction. Building on this operationalization of AL – rather than identifying another set of criteria – ensures that we are looking at the same phenomena as the studies based on the questionnaires. Moreover, identifying instances of AL based on these criteria enables us to critically examine how other theories operationalize and conceptualize AL as captured by the respondents’ answers to the questionnaires. Our question becomes assuming that what is targeted by these questionnaires is something reasonably called authentic leadership (wherever it is located, that is, in the leader or in the follower or between them), how then can that phenomenon be understood and theorized in a more nuanced and empirically grounded way?
Method: Interactional analysis of leadership
In our exploration of AL, we take an interactional approach to leadership which aims to show how leadership is achieved in and through naturally occurring workplace practices (Clifton et al., 2020; Larsson, 2017; Schnurr and Schroeder, 2019). The advantage of taking such an interactional approach to leadership is that, by analysing naturally occurring interaction, any claims about what (authentic) leadership is, or is not, how it is enacted and by whom are necessarily rooted in actual practice rather than in second-order interview- and survey-based accounts of what leadership is considered to be.
Within this broader interactional approach to leadership, we use conversational analysis (CA; Heritage and Clayman, 2010; Sacks, 1992) as a methodological approach to analyse sequences of interaction identifiable as containing authentic leadership. We stress that our use of CA diverges in significant ways from canonical CA. For the purpose of this study, following researchers such as Clifton (2006), Fairhurst (2007) and Svennevig (2008), we use CA as a methodological tool to analyse how (authentic) leadership, as defined in the ALQs and ALIs, is produced in interaction. We recognize that using a researcher-driven benchmark (i.e. the ALQs and ALIs) departs significantly from canonical CA. However, bracketing the epistemological and ontological assumptions that a canonical CA study would imply, we use the techniques, terminology and tools of CA to provide new and robust empirical grounding for claims about (authentic) leadership processes.
As a first step in our analyses, we identified instances of AL in our recordings of everyday workplace interactions based on the items in the ALQ and ALI. Once those instances were identified, we then used the tools of CA to analyse in more detail how these activities – such as ‘says exactly what he or she means’ (item 3 in Walumbwa et al., 2008) and ‘makes decisions based on his/her core beliefs’ (item 6 in Walumbwa et al., 2008) – were enacted on a turn-by-turn basis. Taking such an approach enables us to gain insights into how the phenomenon that is called AL is (interactionally) accomplished, and where it can be ontologically located.
Whilst it is beyond the scope of this article to provide an in-depth summary of conversation analysis (CA) (see, e.g. Hutchby and Woofit, 2008; Stivers and Sidnell, 2012; Ten Have, 2007 who provide excellent monographs summing up the essence of CA), suffice it to say that CA is ‘the study of how social action is brought about through the close organization of talk’ (Antaki, 2011: 1). Central to CA is the analysis of sequences of turns of talk, the most fundamental of which is the adjacency pair which can be defined as a pair of turns at talk that are • Physically adjacent to each other • Produced by two different speakers • Constructed in terms of first and second pair parts • Constructed such that speaker one’s first pair part makes it conditionally relevant for speaker two to respond with an appropriate second pair part Markee (2000: 56)
Thus, for example, ‘Hello, how are you?’ is the first part of an adjacency part which requires a conditionally relevant response (i.e. a reply) which could be something like ‘I’m fine thanks’. Consequently, it is through the sequential working of an adjacency pair that a social action (i.e. a greeting) is performed, and, as Goodwin (2000: 1491) points out, the ‘accomplishment of social action requires that not only the party producing an action, but also that others present, such as its addressee, be able to systematically recognize the shape and character of what is occurring’. In short, any social action (including AL) is a dialogic and not a monologic achievement.
Further, it is the functioning of adjacency pairs that renders the participants’ orientation to the doing of social action analysable. This is because, in what CA scholars call the next turn proof procedure, the participants’ understanding of a prior turn is displayed in the following turn (Sacks et al., 1974: 728). Consequently, the second pair part is taken as proof of understanding of the first pair part and is available to the participants as a way of constructing a mutual understanding of ‘what is going on’ and what behaviour is being performed. Moreover, such an understanding is also available to the researcher, who is able to understand and analyse what is demonstrably ‘going on’ through his/her own member’s knowledge (Peräkylä, 2011: 369). Significant here then is the fact that the analysis is limited to what is observable (i.e. talk). The analysis does not rely on psychological notions of intentions that are said to be ‘in there somewhere’. In the ethnomethodological tradition, such unobservable psychological notions are bracketed out of any analysis. In short, to paraphrase Sacks (1992: 11), CA does not try to second guess what people are thinking, rather it attempts to explicate how social action is achieved.
For the purposes of this article, we are particularly interested in the notions of alignment and affiliation which are actions performed in the second part of an adjacency pair. Following Stivers et al. (2011: 21), alignment with a prior turn occurs at the structural level of talk by ‘facilitating the proposed activity or sequence by accepting the presupposition and terms of the proposed action or activity; and matching the formal design preference of the turn’. Affiliation with a prior turn, on the other hand, goes beyond such structural alignment and relates to how second speakers ‘match the prior speaker’s evaluative stance, display empathy and/or cooperate with the preference of the prior action’. Thus, as Stivers et al. (2011) point out, alignment can occur in any interaction (as a second part of an adjacency pair), but affiliation does not necessarily have to be present. For example, in the case of responding to a stranger who asks ‘Where is the elevator?’ simply aligning by saying ‘it’s over there’ is sufficient. An affiliating response would not be conditionally relevant. Expanding on this observation, and paraphrasing Oittinen (2018: 34), we note that alignment is about supporting the action/turn-in-progress, establishing mutual understanding (intersubjectivity) and accepting identities made relevant in the talk (e.g. speaker/hearer, chair/participants, manager/subordinate and (authentic) leader/follower). Conversely, disalignment is an action, such as topic shift, that interferes with the activity in progress. Affiliation, on the other hand, displays preference and empathy and supports the prior speaker’s stance (i.e. the displayed affective attitude to the action-in-progress). Disaffiliation is therefore an action that rejects the prior speaker’s stance.
Using these key CA concepts of adjacency pair, (dis)alignment and (dis)affiliation, we explore how some of the AL activities identified by the ALQ and ALI are enacted turn-by-turn throughout an interaction. We thus provide empirical (rather than reported) evidence of how AL is accomplished in everyday workplace practice. We thereby problematize the assumptions that authentic leadership is located within one party to a leader–follower relationship, namely the leader, and that authentic leader behaviour is the cause of an interactional outcome.
Analysis
We have chosen three sequences from our existing corpora of interactional data collected via audio recordings in two different organizations. In each organization, prior to recording, the permission of the participants was obtained. Further, in order to minimise any intrusion that may have affected the interaction, the researchers were not present during the meetings. The recordings were transcribed using a simplified form of the Jeffersonian transcription conventions as set out by Antaki (n. d.) (see Appendix for a list of transcription symbols used). The sequences have been selected as examples of what is taken to constitute authentic leadership in the major authentic leadership questionnaires outlined above. We identified these sequences based on a close reading of the descriptions of AL in these questionnaires. We acknowledge that the way we use these questionnaires differs slightly from how a typical respondent may treat them. While the typical respondent is asked to produce a generalized (averaged) assessment of their experiences and observations of their leader, we utilize the items in the questionnaires to look at specific interactions. However, we argue that the respondent’s score is necessarily based on a number of such interactions, and therefore, the close inspection of these is important if we want to capture what it actually is that is scored and measured, rather than to take the assumptions of leader behaviours for granted.
The selected examples of AL, as described in the questionnaires, were then analysed using the tools of CA to explore how AL is enacted throughout an interaction. The examples that we analyse below come from audio-recorded interactions from two different workplaces. Examples 1 and 2 occurred during a meeting of the senior management team at an IT company in New Zealand. Example 3 comes from interaction between a manager in a charity organization in the United Kingdom and her subordinates.
Analysis one: Producing authentic leadership
The meeting participants shown in examples 1 and 2 below are Victor (VN), who is the company’s CEO and one of the co-founders of the company, Neil (NL), an external HR consultant, Jacqueline (JB), who is responsible for the marketing of the company, and Shaun (SM), Joel (JL) and Chester (CS), who are senior managers at the company. In our first excerpt, we engage with the dimension of relational transparency, central to authentic leadership (Walumbwa et al., 2008). At this point in the meeting, participants are discussing the financial target of the company for the next few years.
Example 1
JB: this is Vic’s (.) this is Vic’s high level strategic plan↑ (.) so he’s SM: it’s gotta be three years at least↓ ??: yeah JB: ((drawls)):um: okay so the target [(.) (that) VN: [(is unlikely to) happen next year (.3) NL: mm [well JB: [do you want it to happen next year↑=
JL: I suspect <we’d better make a> try and make it happen next year
Example 1 and relational transparency.
Example 1 and consistency in relation to moral standards.
Through this benchmarking against the questionnaires, we make certain that we are exploring the same phenomenon that the dominant theories subsequently theorize as behaviours and leader qualities. However, rather than relying on such conceptualizations, we explore the interaction in its own right, empirically examining how what we as everyday observers could call AL (as benchmarked through the questionnaires) is produced as an empirical phenomenon in the first place.
Given that Victor’s behaviour here resonates with various items on the AL questionnaires, it seems reasonable to argue that Victor demonstrates authentic leadership. So the question then is, how is this action – seemingly easily understood as constitutive of authentic leadership and as such attributable to the leader’s (i.e. Victor’s) internal authenticity – produced in the first place? In order to address this question, we turn to a close examination of the interactional environment in which this utterance occurs. Consequently, in line with CA, we situate our investigation of authentic leadership in a naturally occurring environment, rather than in the context of Victor’s possible internal constitution and character.
In the lines preceding Victor’s exclamation in line 8, the team is discussing the timeline for accomplishing their goals. In line 1, Jacqueline attributes the goals to Victor by saying that this is ‘Vic’s high level strategic plan’. By specifying that ‘it’s [the plan] gotta be three years at least’, Shaun adds to this turn in progress. An unidentified speaker aligns with this projected timeframe in the next turn (line 3: ‘yeah’). Signalled by ‘so’, Jacqueline then begins a turn that projects an upshot of the preceding talk. However, as the turn is in progress, it is completed by Victor who gives the upshot that it is unlikely to happen next year, thereby conveying a critical attitude towards Shaun’s suggestion in line 2. Such an upshot/assessment makes a second assessment a conditionally relevant next action (Pomerantz, 1984), and Neil’s turn, prefaced by the pragmatic particle ‘well’, projects a second assessment.
However, Neil is overlapped by Jacqueline (line 7) who asks ‘do you want it to happen next year’. This discontinues the ongoing assessment of the timeframe for the strategic plan and inserts a question and answer adjacency pair which projects that Victor should reveal his wants. In line 8, Victor provides a conditionally relevant response to this question and states that ‘I would love it to happen next year’. Choosing the verb ‘love’ rather than ‘want’, he upgrades Jacqueline’s question and presents the answer as his own view. Thus, despite assessing it as unlikely to happen next year, he nevertheless displays his desire for it to happen next year and so, by being transparent about his wants and views, does authentic leadership (following the items in the questionnaires).
Significant here is the dialogic achievement of this display of authentic leadership. In other words, Jacqueline’s question sets up a slot in which expressing Victor’s wants becomes an expected and relevant next action, and the rather strong emotional display of his wants, expressed by the word ‘love’, offers a smooth and efficient way to respond to the interactional requirements of the moment. In other words, we can see how Victor’s expression of his wants – his relational transparency – aligns with, and is occasioned by, an interactional environment set-up both by Jacqueline’s question and, to a lesser extent, the team members’ speculation around the duration of the strategic plan (lines 1–6).
In line 9, we see the uptake of this display of authentic leadership in which Jacqueline affiliates and aligns Victor’s display of relational transparency. This is because, orienting to Victor’s displayed stance that despite the fact that he thinks it is unlikely to happen next year (line 5), he would love it to happen (line 8), she makes a suggestion to resolve this dilemma, that is, to ‘make a try and make it happen next year’. Thus, she not only aligns with prior talk by moving the talk forward and by proposing a solution to this dilemma but she also affiliates by proposing future action that is in accordance with Victor’s displayed opinion (line 8: ‘I would love it to happen next year’). Moreover, through the use of hedge (line 9: ‘make a try’), Jacqueline further affiliates with Victor by displaying a similar stance to Victor in relation to the difficulties that such a course of action might face (cf. line 5).
In sum, a close examination of the interactional context of what we took to be a display of Victor’s enactment of authentic leadership (based on several items in the AL questionnaires) (line 8) reveals that this is not a monologic achievement. Rather, it is set up, on the one hand, by the interactional environment, notably Jacqueline’s question which requires a relatively affirmative, or upgraded, response. And, on the other hand, this emotional response is treated as appropriate and is confirmed, rather than contested, through Jacqueline’s affiliation. Thus, we argue that Victor’s seemingly ‘authentic leader behaviour’ is clearly occasioned by, and as such is part of, a particular interactional environment. It is not an independent cause of subsequent follower outcomes as the logic behind the questionnaires would suggest. Further, we argue that the leader’s contribution is occasioned by the structure of the conversation and as such it is an element in an ongoing, collaboratively produced social setting. In other words, the interactional analysis showed that the leader’s contribution is at least as much occasioned and ‘caused’ by the interactional environment (Heritage and Clayman, 2010; Sacks et al., 1974) as by the individual’s internal qualities. This is not to deny that the individual leader might have internal moral standards, rather it demonstrates that AL is a collaborative achievement and that the interactional context is important.
Analysis two: Attributing authentic leadership to others
In the second analysis, we focus on how the follower is active in attributing authentic leadership to others and how this is also a collaborative achievement. The exchange below occurred shortly after the discussion above. At this point in the meeting, participants have moved on from the discussion of the timeframe and are now talking about what profit the company should aim for. We use this excerpt to illustrate that it is not necessarily only the leader who performs authentic leadership, but that perceptions of authentic leadership may also be expressed by others who thereby attribute and project authentic leadership onto their leader.
Example 2
CS: but but we know three of our (directors) will be asking for two million↓ VN: we have to be aggressive (.) CS: yeah VN: but we need to bear in mind what that means it’s it’s easy writing two million on a white board (.) and put after tax <straight after it> that means we’re gonna have to earn (.) over three million (.) before tax (.) (so) three million before tax out of thirty six mill NL: mm VN: is close on nine percent↑ (.) which means if we keep our cost levels at thirty six↑ (.) we margin of forty five percent (.) (out of) that money (.) to achieve that (.) and we’re not running anywhere near that at the moment and we’re talking (.) in terms of the market putting pressure on SM: we’re talking about another twelve months before the market starts showing any (.) return ((several lines omitted where the discussion continues)) JL: we’ve got over a hundred percent in the ((name)) team for instance but (.) NL: yeah and I I I mean <I mean I mean> there’s a key point isn't it (.) I mean (.) when you start putting up er two two million bucks and VN: [(that’d be great) NL: yeah I mean < issue is VN: [( ) we’re either going to achieve (.) something like eight and a half nine percent on thirty six mill (.) to get to get to that number or we can (.) set a more realistic target if we’re running around four or five the thirty six has to go (above) the forty NL: mm
Example 2 and authentic leadership questionnaires.
Neil’s description of Victor as somebody who makes it clear where he stands on controversial issues and also as someone who expresses his thoughts clearly to others resonates with these AL items. At the beginning of the extract, we already see some evidence of Victor’s stance towards the controversial topic of the company’s anticipated profit (e.g. his declaration that ‘we have to be very aggressive’ in line 3), which is then responded to with minimal feedback from Neil (line 10) expressing alignment. After some more discussion among all the members of the team (deleted here for reasons of space), Neil’s portrayal of Victor as an authentic leader starts in line 20 where he signals affiliation with the utterance initial ‘yeah’ before summing up the gist of the talk (‘I mean there’s the key point’) (Heritage and Watson, 1979). With the deictic marker ‘there’s’, he indexes Victor’s (and others’) prior talk and sums it up. He thereby puts particular emphasis not only on the controversial issue of the company’s predicted profit but he also emphasizes Victor’s anticipated reaction to it (lines 22 and 23: ‘Vic is sitting there going shit you know how the hell are we going to achieve that’). The formulation of the talk takes the form of reported speech in which the words hypothetically uttered by Victor are produced by others who speak on his behalf. In line 24, Victor assesses the proposition to put up two million bucks positively (‘that’d be great’). Neil, incorporating the same word ‘great’, aligns and affiliates with this assessment in his next turn, and in the continuation of the turn, he repeats the projected displayed stance of Victor (‘how the hell are we going to achieve that’). Neil’s utterance thus anticipates and projects Victor’s position on the controversial issue – it is not caused by Victor’s actual behaviour as the logic behind the questionnaires would suggest. Further, the emotional element of ‘shit’ is hearable as reflecting a consideration of the obligations inherent in his position, both towards the organization (in delivering on the goal) and towards his team (in placing high demands on them). In short, not only is Victor’s position on the issue displayed (as being committed to the goal) but his awareness of the challenges and demands of this is also displayed.
The subsequent lines illustrate that Victor aligns with this attribution. This is because Neil, reporting Victor’s speech, has set up the question ‘how the hell are we going to achieve that’ and in lines 27 ff. Victor begins to give the answer to the question, thus aligning with the prior turn by moving the talk forward and by providing a conditionally relevant next turn, that is, an answer to a question. The point to emphasize here, however, is that this is not about whether Neil’s voicing of Victor’s evaluation is, or is not, what Victor ‘truly’ thinks – rather, we argue that letting others know where one stands (which is an activity that the questionnaires associate with AL) is jointly produced here by both Victor and Neil, with Neil playing an active part and actually being the one formulating and expressing it. It is thus Neil, rather than Victor himself, who portrays Victor as an authentic leader.
Analysis three: Failed authentic leadership
In authentic leadership theory, leader behaviour is seen as expressing internal authenticity. In other words, the meaning of what is considered to be a behaviour is given by the internal constitution of the leader. More follower-centric versions of authentic leadership emphasize that these behaviours need to be perceived as such by followers and that the follower’s assessment is central to subsequent consequences and effects. Nevertheless, the meaning of leader behaviours is given by the expression of inner authenticity, rather than either being constituted by the followers or the interaction and the relationship. In essence, a follower might misperceive an authentic behaviour as non-authentic (for instance, through a perception of value incongruence (Sidani and Rowe, 2018)). However, such misperception does not, in theory, affect the meaning of the leader behaviours in the first place. We will utilize the following excerpt to analyse a case of failed authentic behaviour and discuss how such a situation can occur.
The data discussed in this extract come from a meeting in a charitable organization in which the manager (Adele), who is relatively new to the organization, is trying to influence her subordinates (Kate and Diana) to accept a new system for managing time off which is taken in lieu of work (‘toil’) that they have done at charitable/fundraising events. However, tightening up the system could be seen as checking up on the charity workers because maybe they are taking extra days off. Indeed, the talk prior to the extract analysed below is replete with turns in which Diana and Kate account for the amount of time off in lieu that they have, therefore warding off any possible inference that they are ‘fiddling’ the amount of time off they take. We join the transcript in the final turns of the discussion.
Example 3
K: I think what it is that (.) we are acc-like on a day in November we’ll work ten hours↓ and then I’ll take a day [off D: [ °yeah° K: >so< I’ll remove seven↓ A: yeah yeah I’ve got that [I understand that K: [ ( ) so get too much added up↑ A: I-I understand that↓> and < I’m not trying to not I’m not trying to say you’re not entitled to the toil I’m just trying to say that (.).hh I need to sign it off to know what the amount of work that you’re doing↓ (.5) K: and in the summer it has been just [°kinda° A: [well you know we’ll talk about the summer when summer’s over
Example 3 and authentic leadership questionnaires.
In what follows, we will have a closer look at this utterance in its interactional context and particularly examine what work it does in the sequential unfolding of the interaction. In other words, rather than considering how an external observer might make sense of this as a decontextualized utterance, we will look at how the participants make sense of it as it unfolds in the interaction.
In line 1, Kate continues justifying the amount of time that she has to take off in lieu of work she has done attending charity events, and she explains that if she works 10 hours (at a charity event), she then takes a day off and takes seven (hours) as time off in lieu. This can be seen as a form of troubles-telling talk (Jefferson, 1988), whereby people share various kinds of experiences which are treated as something that bothers or troubles the teller (such as, having physical pain or having a car that has broken down or, in this case, having a problem at work).
Kate and Diana, who by showing the same stance (line 3: ‘°yeah°’), align and affiliate with this troubles talk, thus implicitly accounting for, and justifying, the amount of time they take off and presenting it as a ‘trouble’. They therefore display that they consider that Adele’s interest in the way in which time off in lieu is managed and the desire that she has expressed to tighten up this system is not innocent and may be motivated by the suspicion that they are in some way exaggerating the amount of time off that they take.
In the second pair part of the adjacency pair, Adele provides a conditionally relevant response and displays that she understands this (line 5: ‘yeah yeah I’ve got that [ I understand that’). However, Kate overlaps this turn in progress and continues her explanation of how and why she has so much toil to take (line 6: ‘>so< get too much added up’). Thus, she pursues the troubles telling as the relevant interactional project for the moment. As shown by Jefferson (1988), this projects an expectation of a display of empathy and affiliation by the recipient (rather than, for instance, suggestions for improving or solving the situation). Yet, Adele does not produce much of a display of empathy and affiliation. Instead, starting in overlap with Kate (line 7), she displays what can be heard as alignment (‘I-I understand that’) but without affiliating with the emotional stance displayed by Kate. She then accounts for the reasons why she wants to tighten up the system. Significantly, this is framed in terms of what she is, and is not, ‘trying to do’ (i.e. lines 7–9). Through justifying her actions, she therefore displays an exteriorization of her motivations. She can be heard to attempt to display that she is not presenting a ‘false front’ to Diana and Kate in the sense that she is not tightening up the system because she does not believe that they are entitled to the toil.
However, her orientation towards accounting for her actions and thus displaying that she is not presenting a ‘false front’ does not fit with the previous turn’s projection of sequential relevance (i.e. some kind of affiliation and sympathetic response to troubles talk). Adele is thus failing to show any empathy, rather she disaligns with the prior talk and closes down an opportunity to address these troubles. Consequently, her display of what, taken by itself could be heard as an authentic behaviour (not presenting a ‘false front’), turns out to be more of a failure to both produce relational transparency and to be guided by moral standards, as expressed by aligning with the projected expectation of affiliation and empathy.
Adele’s account projects a relevant next turn that displays alignment and affiliation, that is, a next turn that treats her utterance precisely as a display that she is not presenting a ‘false front’ and that she is not suspicious or distrustful. However, this is not what happens. Instead, after a micro-pause which often prefaces a dispreferred turn (Pomerantz, 1984), Kate then skip-connects, (i.e. produces an utterance which is not related to the immediately prior utterance (Sacks, 1992: 349)) and, using ‘and’ to explicitly link this turn to her prior turn, continues the topic of accounting for the amount of toil she has (line 11: ‘and in the summer it has been just °kinda°’). In short, Kate’s turn disaligns with Adele’s prior turn and so does not affiliate with Adele’s emotional stance, expressed in her account. By skip-connecting to her and Diana’s troubles, she disaligns with Adele’s attempted display of authenticity (i.e. display of not putting on a false front), and she again opens up a space in which Adele could affiliate with her subordinate’s troubles telling.
In sum, this analysis suggests that considering authentic leadership to be ontologically located in the leader and expressed by leader behaviours becomes problematic when confronted with the actual production of those behaviours. While the external observer might see Adele’s utterance in lines 7–9 as resonant with behaviours constituting authentic leadership, this is not how the participants visibly treat it. In other words, even though it might seem to be authentic behaviour when taken out of its interactional context, when taking this context into consideration, it does not contribute to the production of an authentic leadership relationship. Following Sidani and Rowe (2018), follower perceptions are central to the production of authentic leadership relationships. However, while Sidani and Rowe see this as a cognitive process, the analysis of the interaction reveals that it is a collaborative accomplishment, visible and negotiated in the sequential organization of talk.
Discussions and conclusions
In this article, we specifically engaged with two assumptions central to the theory of AL. The first is that AL essentially emanates from qualities inherent to one party of the leadership relationship, that is, the leader, expressed through his/her behaviours (Ford and Harding, 2011; Sparrowe, 2005; Walumbwa et al., 2008). The second assumption is that AL is basically a linear process in which behaviours are distinct from subsequent perception and assessment by followers and that such behaviours causally influence interaction and the leadership relationship (Černe et al., 2014; Sidani and Rowe, 2018; Walumbwa et al., 2008). We engaged with these assumptions by identifying exemplary actions, resonating with how AL is operationalized in the major questionnaires (Neider and Schriesheim, 2011; Northouse, 2010; Walumbwa et al., 2008). Rather than exploring the practice of AL in a particular context, our intention was thus to utilize a small number of illustrative examples to challenge current AL theory and more specifically to address the questions of both the nature of the empirical phenomenon that is called AL and where this can be ontologically located.
Our first example contained a sequence that – following the items in the AL questionnaires – could be described as AL behaviour, namely showing relational transparency. A fine-grained examination of the interaction then showed that what AL theory conceptualizes as leader behaviour is in fact dialogically and collaboratively produced through the way in which participants align or affiliate with prior turns in this sequence of talk.
In example two, we further questioned the nature of AL as proposed by the dominant theories. Here, we demonstrated that AL is not necessarily a behaviour performed by the leader, but that AL can in fact be constructed by others who then assign and project authenticity onto their leader. In this instance, as in the previous one, AL is collaboratively produced: it is performed, in this case, through a formulation by a follower which aligns and affiliates with the leader’s prior talk and which is subsequently aligned and affiliated with by the leader as the interaction unfolds.
Example 3 shows failed authentic leadership, in the sense that what on the surface might look like an expression of leader authenticity is not treated as such by the other participants. In other words, an action, that taken by itself seems to resonate with how authentic leadership is identified in the questionnaires, fails to be treated as such by the participants in the interaction. This failure is explained by the lack of alignment with the situated expectations of a sequence of troubles telling (Jefferson, 1988). In this interaction, despite resonating with the description in the questionnaires, something seems to be missing if authentic leadership is to be accomplished. We suggest that one important element that is missing from the conceptualization of authentic leadership in terms of discrete behaviours performed by one of the participants to an interaction is the importance of the sequential structure of interaction. This example shows how the sequential positioning of the utterance within the evolving interaction is as important as the way in which the utterance could be seen to reflect the internal qualities of the leader. The failure to ‘bring off’ authentic leadership, on account of disalignment and disaffiliation in subsequent turns, thus underlines the dialogic and interactionally situated nature of doing authentic leadership. It also underlines the failure of the notion of discrete behaviours to tap into, and adequately capture, this phenomenon in its complexity.
Taken together, our analyses thus challenge the basic logic of AL theory as formulated by Walumbwa et al. (2008) and others (Černe et al., 2014; Neider and Schriesheim, 2011; Sidani and Rowe, 2018). More specifically, they challenge the assumption that authentic leadership is located in the discrete behaviours of authentic leaders. Rather, we suggest that what are taken to be AL behaviours are essentially collaboratively produced interactional outcomes. Therefore, they are not easily reducible to the discrete psychological qualities or cognitive processes of the participating individuals (Garfinkel, 1967; Rawls, 2008). Utterances and actions are successful, in the sense of contributing to an experience of authentic leadership, when they adhere to an expectation framework set up collaboratively over a number of turns. Second, they might be performed by others than the identified leader (as in example 2). Third, as shown by the contrast between examples 1 and 3, their quality as authentic leadership actions is primarily dependent how they fit with what previous turns make relevant as a next contribution (Heritage and Clayman, 2010) and how they are responded to by others (Sacks et al., 1974).
Our analyses also challenge the assumption of a linear logic whereby followers are impacted by leader actions. As shown in our examples, followers actively contribute to the evolving interactional environment in which AL is constructed, enacted and negotiated. Thus, rather than conceptualizing leader behaviours as impacting on followers, we see followers actively co-producing them. This collaborative relationship and the crucial role of followers were particularly obvious in Example 3 where, due to the setup of the interactional environment and sequential expectations, the leader’s attempts to display AL were rejected by the followers. Examples like this thus illustrate that followers play a fundamental role in co-producing AL as an interactional accomplishment, rather than being its targets.
In sum, our analysis shows that what the questionnaires take to be AL can indeed be found in workplace interactions. However, the assumptions that the predominant theories – building on these questionnaires – make about AL do not hold up to close (qualitative) empirical scrutiny. Formulated another way, the questionnaire, as an operationalization of the predominant theories, turns out to be highly problematic. This is because the empirical phenomenon thus identified does not resonate well with the theoretical assumptions.
Instead, we find that what might be perceived as an individual’s AL behaviour emerges as a dialogic achievement, where interactants collaborate on a turn-by-turn basis to produce particular interactional outcomes. This is not to say that interactants might not later make sense of, and describe, these and other interactions through everyday and common sense categories of authentic behaviours and leader qualities. However, we argue that a deeper and more analytical understanding of the phenomenon of leadership needs to be built on an exploration of how such experiences (of AL) were produced in the first place.
Consequently, we argue that looking closely at how what is taken to be authentic leadership is actually produced in workplace interaction suggests a need for relocating the phenomenon. Rather than relying on the assumptions that (1) authenticity is an intrapersonal phenomenon and that (2) behaviours are discrete actions that both express, and are determined, by internal qualities, it seems to make more sense to see authentic leadership as a collaborative accomplishment. We thus follow Sidani and Rowe’s (2018) suggestion that scholars should consider AL to be an outcome which is contingent on the follower as much as on the leader (see especially extract 3 in which we claim that authentic leadership is not achieved because there is no alignment or affiliation in response to a so-called authentic leadership behaviour). However, while Sidani and Rowe treat this relationship as a cognitive process that is internal to the follower (assessing value congruence), we suggest that, from an ontological perspective, AL should be located in the actual interaction (Clifton, 2006; Fairhurst, 2007; Larsson, 2017).
A reformulated concept of authentic leadership would thus need to reserve AL for particular types of interactional outcomes that resonate with the central elements of an AL relationship. Taking inspiration from Sidani and Rowe (2018), such aspects would be related to affective and normative affiliation between the parties, in a process where influence is accomplished (Clifton, 2006) and actions organized (Larsson and Lundholm, 2013). As a preliminary definition, we can offer the following. We suggest that the term authentic leadership refers to the interactional accomplishment of a leadership relationship characterized by a mutual affective and value-oriented stance. This mutual stance may be conceptualized by researchers and (co-)constructed among interlocutors as sincerity, transparency and value congruence. It can be captured empirically by identifying and describing the interactional processes through which sincerity, transparency and value congruence are mobilised and oriented to throughout an interaction by both leaders and followers. A leadership relationship is in turn understood as an asymmetrical influence relationship, where a shared commitment to actions oriented towards achieving organizational goals emerges.
It is to be noted that such a conceptualization of AL allows for experiences of what can be talked about as sincerity, transparency and value congruence, without locating these in one of the parties. Instead, such experiences are seen as interactional accomplishments, emerging from the interaction to which both parties contribute. Further, such a conceptualization does not locate authenticity as a causal factor impacting on interaction, but instead treats it as a collaboratively achieved outcome. However, that does not deny the possibility for the relationship thus established to be consequential, in terms of subsequent organizing (Meschitti, 2019), commitment (Arvedsen and Hassert, 2020) and obligations (Larsson and Lundholm, 2013).
However, this definition is only preliminary and needs further development. Consequently, echoing Iszatt-White and Kempster (2019: 366) call for a more ‘bottom up, practice-based understanding’ of authentic leadership, we end this article with a call for more research that seeks to locate the phenomenon of authentic leadership in the empirically observable domain of everyday interaction.
Footnotes
Acknowledgements
The authors would like to thank to thank Erika Darics from Aston University for allowing us to use the data presented in Example 3.
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.
Author biographies
Appendix. Transcription symbols used
(·): micro pause
(.6): pause in tenths of a second
[: beginning of overlapping talk
((activity)): description of activity
( ): untranscribable text
(word): word unclear, transcriber’s best guess
↑word,↓word : onset of noticeable pitch rise or fall
owordo: spoken more quietly than surrounding talk
>word word<: word spoken more quickly than surrounding talk
<word> : word spoken more slowly than surrounding talk
word=: word latched to following word
