INTRODUCTION
Leadership is traditionally viewed as a top-down process of influence in which those higher up in the hierarchy exercise decision-making (and other kinds of) power over those below them. However, these traditional perceptions of leadership have been challenged by so-called critical perspectives. These critical perspectives on leadership “challenge the hegemonic view that leaders are the people in charge and followers are the people who are influenced” (Jackson & Parry Reference Jackson and Parry2008:83). Rather than conceptualizing leadership as a one-way top-down influence process, critical perspectives claim that leadership may actually involve several people regardless of their formal position or hierarchical standing within their organization. In this article we take a closer look at one of the leadership constellations in which members share leadership responsibilities and as a consequence do leadership conjointly, namely co-leadership.
Co-leadership has been described as “two leaders in vertically contiguous positions who share the responsibilities of leadership” (Jackson & Parry Reference Jackson and Parry2008:82). The concept was first introduced by Heenan and Bennis (Reference Heenan and Bennis1999:6) who refer to co-leaders as “truly exceptional deputies—extremely talented men and women, often more capable than their more highly acclaimed superiors”. Research suggests that co-leadership is a highly successful practice that improves leadership effectiveness (Heenan & Bennis Reference Heenan and Bennis1999. O'Toole, Galbraith, & Lawler Reference O'Toole, Galbraith and Lawler2002, Sally Reference Sally2002), particularly in situations where corporations face severe challenges, and is more common than often thought. Although today's most well-known examples of co-leadership constellations concern the senior management level, sharing of leadership responsibilities “is not just an issue at the top of corporations” (O'Toole et al. Reference O'Toole, Galbraith and Lawler2002:79). Instead, various constellations of sharing leadership responsibilities and tasks have been identified and described in the literature. In particular, it has been suggested that these leadership constellations can be placed along a continuum with co-leadership at one end, shared leadership in the middle, and distributed leadership at the other end (e.g. Jackson & Parry Reference Jackson and Parry2008). Shared leadership refers to situations in which leadership responsibilities are shared and rotate among team members, while distributed leadership describes those constellations in which teams lead their work “collectively and independently of formal leaders” (Vine, Holmes, Marra, Pfeifer, & Jackson Reference Vine, Holmes, Marra, Pfeifer and Jackson2008:341). In this article we take a closer look at those co-leadership constellations where a leader and a co-leader (who are in “vertically contiguous positions” (Jackson & Parry Reference Jackson and Parry2008:82)) share a range of leadership responsibilities and tasks.
Using a sociolinguistic approach to co-leadership in three teams of co-leaders in different New Zealand workplaces, Vine and colleagues (Reference Vine, Holmes, Marra, Pfeifer and Jackson2008) illustrate different ways in which relational and transactional leadership responsibilities are shared among several individuals. The findings from their case studies indicate “certain similarities in the ways that co-leadership is achieved linguistically” (Vine et al. Reference Vine, Holmes, Marra, Pfeifer and Jackson2008:340). For example, they observed that the leaders and co-leaders in these workplaces shared task accomplishment and relationship maintenance in different ways, with typically one member of the co-leadership constellation being more in charge of one of these leadership behaviors. Some of the linguistic strategies through which leadership was enacted included giving approval, checking people's progress, outlining expectations, and easing tensions within the team. However, while the leaders and co-leaders drew on similar strategies to do leadership, Vine and colleagues (Reference Vine, Holmes, Marra, Pfeifer and Jackson2008:354) observed that “there are some important differences in practice, which are engendered by the individual leaders and the specific organizational contexts in which they act.”
In this article we explore in more detail how co-leadership is actually done by focusing on those situations where members disagree with each other. While previous research has convincingly shown that co-leadership may involve the effective sharing of various leadership responsibilities and tasks, other less harmonious aspects of co-leadership are often overlooked. We aim to address this issue by exploring another side of co-leadership by focusing on how members in different co-leadership constellations negotiate leadership responsibilities in situations where they disagree with each other.
DISCURSIVE LEADERSHIP
Our research can be placed within the tradition of discursive leadership. Recent developments in leadership research—most notably Fairhurst's work on discursive leadership (e.g. Fairhurst Reference Fairhurst2007)—are increasingly recognizing advantages of a discursive approach to leadership. Discursive leadership conceptualizes leadership as emerging and “as a co-constructed and iterative phenomenon, socially accomplished through linguistic interaction” (Tourish Reference Tourish2007:1733). This relatively new approach often positions itself in opposition to traditional leadership psychology. While leadership psychology is mostly concerned with leaders' perceptions and self-reflections, discursive leadership focuses on language in use, that is, on how leaders actually communicate and interact with the people they work with (e.g. Fairhurst Reference Fairhurst2007).
Discursive leadership draws on tools and methods developed by discourse analytic approaches (such as conversation analysis, interaction analysis, critical discourse analysis, and many more) in order to analyze leadership discourse. It conceptualizes communication as being at the heart of the leadership processes. More specifically, by analyzing interview discourse, actual dialogues, and other discursive formations, discursive leadership explores how people actually do leadership rather than describing how they think (and perceive) they do leadership. In other words, discursive leadership is interested in how leadership is actually accomplished in specific situations rather than attempting “to capture the experience of leadership by forming and statistically analyzing a host of cognitive, affective, and conative variables and their casual connections” (Fairhurst Reference Fairhurst2007:15).
While the research presented in this article is firmly located within the tradition of discursive leadership, we utilize and apply two analytical concepts that have not yet been systematically applied to an analysis of leadership, namely face and identity. Using these concepts, we hope to provide further insights into how co-leadership is actually performed and which processes are at play. Our specific focus is how members of co-leadership constellations construct their (intertwined) professional identities through the process of doing face-work in situations where leader and co-leader disagree with each other. In particular, the analyses below illustrate how the identities of those involved in these co-leadership constellations are conjointly constructed throughout an interaction: more specifically, the identities of leader (i.e. the one in charge) and co-leader (i.e. the second-in-command) are fluid and are constantly negotiated among interlocutors in these constellations. By doing face-work, that is, by adhering to, reinforcing, or challenging their own and each others' face, interlocutors portray themselves as leaders and co-leaders in relation to each other.
CONSTRUCTING IDENTITIES AND DOING FACE-WORK
In line with discursive leadership, we take a postmodern stance in understanding professional (and other social) identities as “the social positioning of self and other” (Bucholtz & Hall Reference Bucholtz and Hall2005:586). According to this view, identities are constantly constructed, enacted, and negotiated in and through discourse (see Jenkins Reference Jenkins1996, Hall, Sarangi, & Slembrouck Reference Hall, Sarangi, Slembrouck, Sarangi and Roberts1999, Holmes, Stubbe, & Vine Reference Holmes, Stubbe, Vine, Sarangi and Roberts1999, Postmes Reference Postmes, Haslam, Knippenberg, Platow and Ellemers2003). As Fairhurst (Reference Fairhurst2007:107) puts it, “identities are multilayered, fluid and contingent” (see also Antaki & Widdicombe Reference Antaki, Widdicombe, Antaki and Widdicombe1998, Cranny-Francis, Waring, Stavropoulos, & Kirkby Reference Cranny-Francis, Waring, Stavropoulos and Kirkby2003). They can be conceptualized as intersubjective accomplishments and relational phenomena: they are always to some extent co-constructed among interlocutors and acquire their meaning in relation to other identities (Bucholtz & Hall Reference Bucholtz and Hall2005:598). Constructing identities, thus, involves multiple “different, often intersecting and antagonistic, discourses, practices and positions” (Hall Reference Hall, Gay, Evans and Redman2000:17). As a consequence, identities are not necessarily homogenous and consistent but may actually involve competing and sometimes contradictory aspects (Lytra Reference Lytra2009, Schnurr & Zayts Reference Schnurr, Zayts, Angouri and Marra2011).
One crucial aspect of identity construction is face-work. The concepts of face and identity have both received a good deal of attention in previous research, and several attempts have been made to link them explicitly (e.g. Spencer-Oatey Reference Spencer-Oatey2007, Locher Reference Locher, Antos and Weber2008). Most conceptualizations of face are based on Goffman (Reference Goffman1967:5), who describes face as “the positive social value a person effectively claims for himself [sic] by the line others assume he [sic] has taken during a particular contact”. This notion of face is closely related to notions of identity and subjectivity, as Geyer (Reference Geyer2008:50) explains: “an interactant's face manifests itself as his or her interactional self-image, which is determined in relation to others, discursively constructed during a particular contact, and closely aligned with the participant's discursive identity”. Tracey (Reference Tracey, Giles and Robinson1990:210) similarly maintains that “face references the socially situated identities people claim or attribute to others”. In this respect, face and identity seem to be cognitively similar (but see Spencer-Oatey Reference Spencer-Oatey2007:644 for further discussion). Locher (Reference Locher, Antos and Weber2008:515) even goes as far as claiming that “the notion of face can stand for identity construction in more general terms”. However, as Haugh (Reference Haugh, Bargiela-Chiappini and Haugh2009:3) points out: “[t]his move towards conceptualising face as concern for identity … raises the question of how such research on face can be distinguished in a meaningful way from broader work on identity”. And while we acknowledge that this issue warrants further discussion, in this study our focus is on how the two concepts of face and identity can be productively combined in order to shed further light into the complexities of co-leadership.
The specific focus of this article is on face-work and identity construction in those situations where the members of the co-leadership team disagree with each other and where as a consequence their face needs are particularly vulnerable. In these situations, interlocutors are faced with the difficult tasks of doing leadership and negotiating and constructing their professional identities, while at the same time maintaining harmony with the other members of the group (in particular the other member of the co-leadership constellation). Drawing on Bourdieu Reference Bourdieu and Nice1977, the strategies that leaders and co-leaders employ in these situations could be described as “symbolic struggles for positions” (see Stewart Reference Stewart2008:33). As Stewart (Reference Stewart2008:33) notes:
Within this perspective, the negotiation of face-wants is central to an appreciation of how individuals use language as a means of positioning themselves and others in interaction. This also allows us to conceive of ‘face-work’ not as a way of achieving social equilibrium in terms of interpersonal harmony but rather as an argumentative tool for the negotiation of self in society.
Thus, by doing face-work and orienting to their own as well as each other's face needs, they at the same time (co-)construct their own as well as each other's identities. But how exactly do interlocutors do face-work in and through discourse? Goffman (Reference Goffman1967:15ff) describes two “basic kinds of face-work”: avoidance-based and corrective processes. Avoidance-based processes include the use of mitigating strategies, such as hedges and humor, while corrective processes include apologizing and explaining or correcting an offence. However, as MacMartin, Wood, and Kroger (Reference MacMartin, Wood, Kroger, Robinson and Giles2001) caution, linguistic strategies are inherently multifunctional, which makes it impossible to ascribe specific strategies to face-work. Rather, face-work strategies need to be conceptualized more generally, for example, along the lines proposed by Geyer (Reference Geyer2008:51), who suggests that face-work “includes relational stances such as alignment and opposition”. As such, face-work describes how interlocutors relate and position themselves (and others) in relation to each other. According to her, on the level of discourse, face-work may be expressed through “[v]arious linguistic and discursive structures (e.g. preference organization, contextualization cues, membership categorization devices, discourse markers) and analytical procedures pertaining to face-related issues (e.g. solidarity, alignment, autonomy, competence)”. In a similar vein, Haugh (Reference Haugh, Bargiela-Chiappini and Haugh2009:7) remarks that face-work can also be accomplished “through drawing on more general features of sequential organisation of talk-in-interaction” (see also Lerner Reference Lerner1996).
Finally, it should be recognized that identifying face-work in interaction is far from a straightforward issue. And one of the problems inherent in this analytical process is the question of whose notion of face is used in such an analysis—participants' own understanding or the analyst's conceptualization of face? In addressing this issue, MacMartin and colleagues (Reference MacMartin, Wood, Kroger, Robinson and Giles2001) and Haugh (Reference Haugh, Bargiela-Chiappini and Haugh2009) strongly recommend that any analysis of face needs to be grounded in participants' own interpretations: “we [as analysts] must see that they [the participants] orient to utterances as matters of face (threat, enhancement, mitigation, damage, attack, etc.)” (MacMartin et al. Reference MacMartin, Wood, Kroger, Robinson and Giles2001:229). In order to consider participants' views, then, in our study particular attention is paid to the discursive strategies through which they negotiate their differing viewpoints. More specifically, we aim to explore some of the face-work strategies through which interlocutors position themselves and each other as leader or co-leader, and thus as the “one in charge” and the “second in command”. We also take into account the comments participants made about their interactions in interviews, as well as insights gained during participant observation. These insights are used as valuable additional information that shed light on what is going on in an interaction.
DATA
Our analysis of co-leadership performance draws on a corpus of workplace data that comprises more than eighty hours of authentic workplace discourse collected in a variety of workplaces in Hong Kong, including large international corporations and small family-owned businesses. For this study we look at the co-leadership constellations in an IT consulting organization (ABC Consulting) and a paint manufacturing company (Rainbow).Footnote 1 In both workplaces we used video cameras and voice recorders to tape three to six regular meetings (for a more detailed description of the data collection see Chan Reference Chan2005, Reference Chan, Sun and Kadar2008, Schnurr & Zayts Reference Schnurr, Zayts, Angouri and Marra2011). Overall, we recorded more than twenty hours of meeting talk in these two workplaces. Although participants in both workplaces were native speakers of Cantonese, the meetings at ABC Consulting were held in English, while the Rainbow meetings were conducted in Cantonese, the most common language used in Hong Kong. Data was transcribed using a modified version of the LWP transcription conventions (http://www.victoria.ac.nz/lals/lwp/) and the Cantonese was translated by the second author and has been checked for correctness.
The meetings in both workplaces were attended by six to fifteen (Chinese) staff members including the CEOs of the organizations. However, although in both cases the CEOs (i.e. those highest in terms of organizational hierarchy and decision-making power) were present, the meetings themselves were run by the second-in-commands, namely the Managing Director (Danny) at ABC Consulting and the Production Manager (Anthony) at Rainbow. Thus, in both workplaces, the meeting chair was not the highest in the organizational hierarchy in terms of their position in the organization and their decision-making power and authority. In fact, as the analysis below demonstrates, although the meetings had assigned chairs, in reality they were co-chaired or co-led by the officially assigned chairs and the CEOs.
The primary discourse data is supplemented by interviews with participants. Insights gained from interviews are particularly valuable for an analysis not only of what is going on in an interaction but also, more specifically, in terms of interpreting the process of identity construction. In particular, the information gained through the interviews with the leaders, co-leaders, and some of the people they work with, was crucial for an understanding of how the co-leadership performance of the respective teams was actually perceived by those affected. The interviews also provided important explanations as to why certain behaviors were interpreted as more or less face-threatening.
In addition to collecting these discourse data, we have also undertaken participant observation and consulted a range of internal and external organizational documents. Employing such a multi-method approach involving diverse sources of data provides valuable additional information that facilitates and enriches the analysis of the data. Moreover, the information obtained through participant observation, consultation of organizational documents, as well as a workplace survey provided a detailed picture of the workplace cultures that characterise ABC Consulting and Rainbow.
We briefly describe the co-leadership constellations in both workplaces before exploring in more detail how the leaders and co-leaders at ABC Consulting and Rainbow construct their own as well as each others' professional identities through doing face-work in situations where they disagree with each other.
The co-leadership constellations
ABC Consulting is a Hong Kong-based consulting company that deals with clients all over the world. At the time of data collection the organization was strictly hierarchically structured with the Board of Directors at the top followed by the CEO and then the Managing Director. This hierarchical structure is also reflected on all levels of staff where rankings are translated, for example, into seating arrangements and size of offices. The co-leadership team at ABC Consulting includes Danny, the Managing Director, and QS,Footnote 2 the CEO and more senior person.
Investigations into Rainbow's workplace culture, by contrast, indicate that the company views itself as “a big family”. And Liu, the CEO, founder, and sole owner of the company, is regarded as a father, as staff commented in the interviews. The co-leadership team at Rainbow consists of Liu and Anthony. According to the company's hierarchical chart, Liu holds the highest and most powerful position. All other meeting participants are placed on the same hierarchical level. Nevertheless, based on staff's comments and our knowledge of internal power relations and responsibilities, it is clear that Anthony has a special status. He is the chair of the Sales and Production meetings, and he is the head of the Production Department, which constitutes a major and crucial component of the company. Anthony thus has more legitimate and consultative power (French & Raven Reference French, Raven and Cartwright1959, see also Dwyer Reference Dwyer1993) than the other members.Footnote 3
In describing professional duos, Alvarez and Svejenova (Reference Alvarez and Svejenova2005:120) distinguish between “partnerships” and “hierarchical pairs.” They note that while “[p]artnerships consist of individuals … who share a position,” hierarchical pairs are those “relationships in which there is a clear authority line of superordination and subordination” (Alvarez & Svejenova Reference Alvarez and Svejenova2005:120). While we agree with this distinction in principle, we would suggest that in reality differences are often less clear-cut and more dynamically performed. In line with a social constructionist approach, it thus seems more productive to view the various types and constellations of professional duos as relationships that are dynamically enacted and negotiated rather than as fixed categories. The different ways in which these relationships are enacted can usefully be visualized as a continuum with “partnerships” at one end and “hierarchical pairs” at the other. Conceptualized in this way, the duos or co-leadership constellations at ABC Consulting and Rainbow would both have to be defined as “hierarchical pairs” with the team at ABC Consulting being placed closer to the “partnership” end than the team at Rainbow.
In the meetings at ABC Consulting, decision-making power and authority appear to be relatively equally shared between Danny and QS: most decisions are made conjointly and both of them have clearly defined areas where they are the expert and main decision maker. As Danny mentioned in the interview: “whenever business strategy arise, it's always he [QS] has come to question. When it's from the business operations, it's always me. Then he steps sort of step aside a little bit. I can have to step in, that doesn't stop him from his interest or whatever he's doing.”
At Rainbow, by contrast, the co-leadership constellation of Anthony and Liu is rather different: Liu is clearly the more powerful person in terms of authority and decision-making power. In the interview during data collection, Liu explained that he is generally in charge of managing and making decisions while Anthony is responsible for the production component of his team. Nevertheless, as our data indicate, Liu and Anthony also make some decisions conjointly and share some responsibilities regarding the various issues related to Productions, Anthony's area of expertise.
We have selected four representative examples here to illustrate how members of these co-leadership constellations construct and negotiate their identities through the process of doing face-work.
EXPLORING CO-LEADERSHIP PERFORMANCE IN DISAGREEMENTS
While most of the daily interactions between leaders and co-leaders, and specifically the ways in which they share leadership responsibilities, are accomplished harmoniously, we focus here on situations of potential conflict. In particular, we look at how the members of the co-leadership teams negotiate their intertwined professional identities through doing face-work in situations where they disagree with each other.
A great deal of research has been done on disagreements in various contexts, including political debates (Schegloff Reference Schegloff1988/89, Blum-Kulka, Blondheim, & Hacohen Reference Blum-Kulka, Blondheim and Hacohen2002), news interviews (Heritage & Greatbatch Reference Heritage, Greatbatch, Boden and Zimmerman1991, Clayman Reference Clayman2002), talk shows (Hutchby Reference Hutchby1996, Gruber Reference Gruber1998), focus group discussions (Myers Reference Myers1998), university seminars (Rees-Miller Reference Rees-Miller2000), family and friends (Goodwin Reference Goodwin1990, Goodwin, Goodwin, & Yaeger-Dror Reference Goodwin, Goodwin and Yaeger-Dror2002, Kakava Reference Kakava2002, Laforest Reference Laforest2002), academic discourse (Tannen Reference Tannen2002), online fora (Angouri & Tseliga Reference Angouri and Tseliga2010, Upadhyay Reference Upadhyay2010), and workplace contexts (Kangasharju Reference Kangasharju2002, Holmes & Stubbe Reference Holmes and Stubbe2003, Holmes & Marra Reference Holmes and Marra2004).
In spite of this large amount of research on disagreement there is no overall consensus in the literature on how disagreements are defined. We adopt Clayman's (Reference Clayman2002:1385) notion of disagreements as consisting of “an oppositional transaction between two primary participants”. This definition, we believe, is broad enough to capture a variety of disagreement phenomena while at the same time being specific enough to be applied to concrete examples in our data.
In a seminal study on disagreements, Pomerantz (Reference Pomerantz, Atkinson and Heritage1984) proposes two types of disagreement: strong and weak disagreements. In distinguishing between the two she notes that
[a] strong disagreement is one in which a conversant utters an evaluation which is directly contrastive with the prior evaluation. Such disagreements are strong inasmuch as they occur in turns containing exclusively disagreement components, and not in combination with agreement components. (Pomerantz Reference Pomerantz, Atkinson and Heritage1984:74)
Weak disagreements, by contrast, “are formed as partial agreements/partial disagreements” (Pomerantz Reference Pomerantz, Atkinson and Heritage1984:65). This distinction between strong and weak disagreements has subsequently been picked up by other researchers who observed that strong disagreements are relatively common between family members, close friends, and people who know each other very well (e.g. Tannen Reference Tannen2002, Habib Reference Habib2008). However, when interlocutors are not very familiar with each other, or when maintaining neutralism is required, weak disagreements that are typically mitigated by a range of attenuating discourse strategies tend to be used more frequently (e.g. Myers Reference Myers1998, Jacobs Reference Jacobs2002).
In the Hong Kong context, research in organizational behavior and cultural studies suggests that Hong Kong people tend to compromise and avoid direct disagreement (Kirkbride, Tang, & Westwood Reference Kirkbride, Tang and Westwood1991, Pan Reference Pan2000). For example, in a study on disagreement in dyadic conversations from a discourse-analytic approach, Cheng and Tsui (Reference Cheng and Tsui2009) discovered no significant difference in the extent to which Hong Kong Chinese (HKC) and native speakers of English (NSE) disagreed with their interlocutors. Yet in terms of linguistic strategies, the HKC participants tended not to overtly disagree with their interlocutors but adopted more mitigated and redressive strategies than their NSE counterparts. The authors suggest that “for HKC, disagreeing with addressee involves a lot more face work and interpersonal management work” (Cheng & Tsui Reference Cheng and Tsui2009:2372).
The performance of disagreement has been described as being “by its very nature … a face-threatening act that jeopardizes the solidarity between speaker and addressee” (Rees-Miller Reference Rees-Miller2000:1089), and, we would like to add, that potentially challenges the subject positions that interlocutors have set up for themselves. Depending on the form of the disagreement (e.g. whether it is mitigated or aggravated) the face needs of the interlocutors are threatened and, especially in a workplace context, their professional identities are challenged. In the next section we identify and describe some of the strategies involved in these complex processes, and explore the link between face-work and identity construction in co-leadership constellations.
Co-leadership at ABC Consulting
We have chosen two examples here from each workplace that are representative of the ways in which the members of the co-leadership constellations disagree with each other.
(1) Context: Participants are discussing a particular project that is in its final stages. One more outstanding payment needs to be settled and participants argue over which library is responsible for this transaction.Footnote 4
Example (1) is a good illustration of an explicit or, to use Pomerantz's (Reference Pomerantz, Atkinson and Heritage1984) terms, strong disagreement between the members of the co-leadership constellation. In line 8 QS directly and overtly contradicts Danny's previous claim about the name of the library responsible for handling the award (lines 3 and 4). Since QS does not use any mitigation strategies to hedge his disagreement, its illocutionary force is relatively strong and potentially face-threatening to Danny. In particular, by repeating the disagreement markers “no not” (line 8), interlocutors' opposing views are emphasized. Directly and overtly disagreeing with Danny in such a way may be interpreted as a threat to Danny's professional identity, as it questions his knowledge and thus challenges his position as “the one in charge”.
Interestingly, in replying to QS, Danny issues an equally strong disagreement (the structure of which almost exactly mirrors QS's previous utterance): like QS, Danny does not use any mitigation strategies when insisting on his initial claim about which library is responsible for the outstanding payment (line 9). This direct disagreement with his boss without any mitigation strategies could also potentially be perceived as challenging QS's face—in particular, since this disagreement is issued “upwards”. In disagreeing with each other Danny and QS not only challenge each others' face but they also construct and negotiate their professional identities. More specifically, by claiming to “know better” than the other and by arguing about who is right, the interlocutors at the same time seem to fight for positions and thereby actively construct themselves and each other as the experts.
After some attempts to give explanations by another meeting participant (Francis, line 10) and QS (line 11), Danny repeats his initial claim (line 12) and provides further (albeit rather short) explanations as to why this particular library is indeed responsible for the outstanding payment. This time, however, QS signals agreement and understanding (“oh” in line 13). “Oh” is a typical “change of state” (from unknown to known) token in English (Heritage Reference Heritage, Atkinson and Heritage1984), and using it QS indicates that Danny's information is new to him. He thereby seems to implicitly acknowledge that in this case Danny knows better than him. This kind of agreeing after disagreeing is referred to by Kotthoff (Reference Kotthoff1993) as concession, which she argues “can be very face-threatening, since it could be viewed by others as the inability to defend one's own opinion” (1993:201). However, this potential threat to QS's own face is mitigated to some extent by his subsequent humorous remark and laughter (line 16). Nevertheless, by accepting and (implicitly) acknowledging that Danny was right, QS to some extent threatens his own face and at the same time reinforces Danny's professional identity as the leader and the one in charge in this instance. QS thereby positions himself, at least temporarily and with regards to this particular project, as less expert than Danny. This short extract thus shows how identities are not only co-constructed between interlocutors but also shift and change throughout an interaction: in line 8 QS attempts to construct himself as the expert, which is then challenged by Danny (line 9). This positioning of themselves is subsequently negotiated amongst Danny and QS.
In the end the disagreement is resolved and the participants move on to discuss other questions (lines 17 and 18) without returning to this issue again in the meeting. It is important to note here that the issue over which the interlocutors disagree is of relatively minor importance. It is thus perhaps not surprising that the disagreement is rather direct and strong. The next example illustrates how the members of the co-leadership team at ABC Consulting disagree with each other on a more serious issue.
(2) Context: After a weekly meeting at ABC Consulting. Participants are starting to leave the room, but several people continue their discussions. QS, Danny, and Lilly (one of the other meeting participants) are discussing an upcoming marketing event that Lilly is supposed to prepare.
The disagreement in this example revolves around the question of whether Lilly, a staff member and subordinate to QS and Danny, should attend a particular marketing event. When QS brings up the issue (line 1), Danny remarks that Lilly “is not ready” (line 2), which is further supported by Lilly herself (line 4). Even QS himself agrees with this judgement “no I know you're not ready” (line 5). However, after having signalled his agreement in general (line 5), he qualifies it subsequently by mentioning that there are several marketing events coming up (thereby implying that it would be good if she could attend some of them). And while we would not classify QS's remark in line 6 as a disagreement, nonetheless, it seems to prepare the way for the subsequent much more explicit disagreement (in lines 11 and 12). In particular, the discourse marker “but” (line 6), which functions here as a “forewarner telling that a disagreement can be expected” (Kotthoff Reference Kotthoff1993:208) together with the structure of the utterance (which resembles a “yes but” structure typical for weak disagreements; Pomerantz Reference Pomerantz, Atkinson and Heritage1984), already indicates that QS is not entirely convinced. The illocutionary force of QS's contribution, however, is mitigated to some extent by his laughter accompanying his utterance (line 5) and his initial explicit agreement with the previous speakers. Danny then agrees with QS (“yeah” in line 7; see also his use of the inclusive pronoun “we”) and orients to QS's concern by suggesting to consult the “marketing event er calendar” (line 7) in order to confirm the dates of the upcoming events.
From this point onwards the interaction takes on a very different tone and the disagreement between Danny and QS becomes more explicit and aggravated: without acknowledging or making any reference to Danny's previous utterance QS states that he is going to talk to the VP of marketing (line 9). He thereby seems to indicate that he questions Danny's authority or decision and intends to consult an expert on that matter. In a similar vein, Danny then responds by saying that he himself is going to talk to somebody else (line 10). However, Danny's reply is cut off by QS who in an agitated tone of voice repeats his initial statement about calling the VP of marketing. At this stage the disagreement is very severe and aggravated as is shown, for example, by the agitated tone of voice of both speakers (lines 11–13) and by QS' comment “I don't care” (line 11). After Danny's unintelligible reply, which overlaps with QS's previous utterance (line 13), the sequence ends with a relatively long silence, which is only interrupted by the researchers coming into the room to collect the recording material as another meeting was scheduled to take place. Silence has been identified as one of the means through which interlocutors can effectively express their disagreement while at the same time avoiding potential confrontation (Laforest Reference Laforest2002, Kjaerbeck Reference Kjaerbeck2008). Thus, by remaining silent for a remarkably long period, QS and Danny seem to acknowledge their disagreement without wanting to further elaborate on it in this situation. Whether the discussion between QS and Danny continued after the meeting we do not know. But in the interview after data collection Danny explained that he and QS can be of different opinions but they do not always have their discussions in the meetings. Rather, they “hold it until the next part” (i.e. until after the meeting), as Danny put it, indicating that many important decisions are actually made outside the meeting room.
This disagreement sequence is rather complex and, we believe, illustrates well how face-work contributes to constructing and negotiating interlocutors' intertwined professional identities. It is clear from our analysis above that in disagreeing with each other in these rather explicit and aggravated ways, QS and Danny severely threaten each others' face. And in doing so, they at the same time challenge each others' professional identity or, more specifically, their relative positions within the co-leadership constellation. For example, by insisting on consulting the VP of marketing, and by explicitly stating that he does not care about Danny's views, QS challenges Danny's identity as the leader and the one in charge of Lilly's marketing activities. However, by refusing to accept his boss's criticism and disagreements and by insisting on doing things his way, Danny at the same time constructs himself as the leader and positions QS as the second-in-command in this area. Thus, doing face-work and constructing professional identities are closely related processes. This example also demonstrates how identity construction is accomplished dynamically and may change from moment to moment in an interaction: by challenging each others' suggestions and views interlocutors position themselves as the more powerful person while at the same time assigning less powerful positions to each other. Starting with QS's “yes but” comment in lines 5–6 this moment by moment construction of identities is particularly obvious: with every utterance QS and Danny challenge each others' position and thereby construct themselves as “the one in charge.”
In the next section we look at how the members of the co-leadership constellation at Rainbow negotiate their intertwined professional identities through doing face-work in situations of disagreement.
Co-leadership at Rainbow
In contrast to ABC Consulting where members of the co-leadership team regularly and openly disagree with each other, at Rainbow Anthony very rarely disagrees with his boss Liu. In this company, hierarchical structures are more pronounced and decision-making power and authority are much less negotiable and are closely associated with hierarchical positions.
(3) Context: In a meeting of the Production team and Sales team at Rainbow. The Sales team complains that the Production team does not provide them with an accurate estimation about the length of time for production and asks the Production team to work out possible solutions to this problem. The topic has been discussed for quite a while when Anthony, the chair of the meeting and head of the Production team, attempts to recount some of the difficulties his team encountered regarding this issue.Footnote 5
The disagreement between Anthony and Liu evolves around the question of what should be discussed at that moment in the meeting. In lines 1–6 Anthony starts outlining some of the difficulties his department (and he as its head) encountered in estimating the length of production time. Attending to the prosodic features indicated that Liu's “why” in line 7 sounds more like a request for more information than a challenge to Anthony's explanations. However, when Anthony continues with his account (seemingly contradicting his previous utterances in lines 8 and 9) he is interrupted by Liu (line 10) who makes explicit what he wants him to focus on, namely on Anthony's “difficulties” rather than his “explanations”. Interrupting Anthony, speaking at a fast pace and repeating the disagreement markers “no no” (line 10) Liu's utterance is potentially face-threatening to Anthony. In particular, interrupting not only Anthony's turn but actually attempting to prevent him from providing explanations, as well as repeatedly using the pronoun “you” (line 11) to address Anthony directly, and the frequent use of “no” (e.g. lines 10 and 13) make this rather strong and explicit disagreement particularly threatening as it challenges Anthony's professional identity as the person in charge of the meeting.
However, rather than accepting Liu's suggestion that there is “no need to explain” (line 13), Anthony insists on providing further explanations (lines 14 and 15). And by ignoring Liu's recommendation, he at the same time threatens Liu's face (in particular by directly disagreeing with him (by saying the opposite of what Liu has proposed in his previous utterance), and by speaking faster and louder). In disagreeing with Liu, Anthony thus challenges Liu's attempt to position himself as the more powerful person and instead tries to reinstate himself as the chair and the one in charge of the meeting. Interlocutors' identities as leader and co-leader are dynamically constructed and who is in charge of the meeting seems to change throughout this sequence.
This disagreement sequence ends with Liu apparently abandoning his opposition (lines 16 and 17): using a softer voice he agrees to let Anthony provide his explanations and hands back the floor to him (line 17). However, although Liu's utterance in lines 16–17 could be interpreted as a concession (Kotthoff Reference Kotthoff1993), Liu nevertheless positions himself as the more powerful interlocutor in particular by uttering “it doesn't matter” (line 17), thereby downplaying the importance of the disagreement. Moreover, in our data there is abundant evidence of examples where Liu has the last word in discussions and decision making: he is generally the one who ultimately decides who gets to speak on what topic and for how long (Chan Reference Chan2005, Reference Chan, Sun and Kadar2008). In these instances, Liu undermines Anthony's authority and position and challenges his professional identity as the “one in charge” of the meeting. He at the same time constructs himself as the “one in charge” and as the leader of the team and the meeting.
The next example provides another instance where Anthony and Liu disagree. However, in contrast to (3), this time the disagreement is resolved rather implicitly and covertly.
(4) Context: Regular Production and Sales meeting at Rainbow, which is chaired by Anthony. Immediately prior to the example, after a long discussion on how to make use of a large quantity of some leftover materials, Liu decides that a particular staff member in the Production Department will be responsible for working out a solution. Anthony does not agree with this suggestion as he thinks the employee in question already has a lot of work to do.
This example occurred at the end of a discussion in another meeting of the Production and Sales teams. Immediately prior to the example, Liu, the CEO, decided that he would like a technician in the Production Department (i.e. somebody who directly reports to Anthony) to work on an urgent task in addition to his normal duties. However, the silence in line 1 indicates that Anthony is reluctant to accept this proposition (Laforest Reference Laforest2002, Kjaerbeck Reference Kjaerbeck2008) because he is worried that “{this} would affect his other duties” (line 7). Liu's first reaction to Anthony's concern is that he does not want to “discuss it now” (line 4). Although his utterance is mitigated to some extent by the restart and the cut off in the beginning, his disagreement with Anthony is rather strong and thus potentially face-threatening.
Interestingly, as in (3), Anthony does not adhere to Liu's proposition not to discuss this issue here but rather disagrees with Liu's decision by referring to the heavy workload of the employee in question (line 5). Disagreeing with Liu and insisting on talking about this issue in the meeting are both activities that potentially threaten Liu's face and challenge his professional identity. By so doing, Anthony challenges Liu's attempt to construct himself as the more powerful (i.e. as the one who decides about work allocation and meeting management). Moreover, Anthony's behavior also challenges traditional Chinese norms, which are a crucial component of Rainbow's culture, and which expect subordinates to be submissive to and respectful towards their boss (Redding Reference Redding1990, Westwood Reference Westwood and Westwood1992). In challenging Liu, then, Anthony reinstates and emphasizes his position as the head of the Production team who should be in charge of assigning tasks to his subordinates. He thereby at the same time portrays himself as the leader and more powerful member in the co-leadership constellation.
In order to convince Anthony and to gain his compliance Liu seems to initially agree with him (line 6: “yes”). However, as becomes clear a few turns later, this initial agreement is actually a forewarner (Kotthoff Reference Kotthoff1993) to Liu's subsequent disagreement (line 8–11) in the form of a “yes but” structure. This disagreement with Anthony, however, is rather weak as it is mitigated not only by the “yes but” structure but also by smiling (line 13), a tag question to gain further compliance (line 13), and relatively detailed explanations for the reasons underlying Liu's decision (e.g. lines 9–11). Only when Anthony signals some agreement (“mm” in line 15) does Liu explicitly state his point: “there is no other way” (line 16) leaving no room for renegotiation or further disagreement. In particular, by repeating this utterance (line 22), it becomes clear that his decision is final and not to be contested any further. Liu's relatively explicit exercise of power here is in line with previous research, which observed that in workplaces where power differences between members are rather pronounced, such as in Rainbow, those with more power “can exploit their influential positions and adopt an assertive style, pressuring others to change and accept their viewpoint” (Holmes & Stubbe Reference Holmes and Stubbe2003:72).
Anthony seems to understand Liu's point as his final agreement indicates (line 27). Moreover, by elaborating on potential negative consequences of not dealing with the task in question (lines 23–26), Liu implies that he is making the decision based on the company's interest. He thereby constructs himself as someone who is concerned about the company as a whole (rather than about individual employees, as Anthony's contributions indicate). Together with a typical preclosing marker “anything else” (line 28) (Boden Reference Boden1994) and passing the chairing duties back to Anthony (line 29), Liu further portrays himself as the leader and the “one in charge” of making decisions. As in the previous examples, this excerpt shows how interlocutors' identities are constructed conjointly and how they shift throughout the interaction: by negotiating their opposing viewpoints Liu and Anthony construct and deconstruct their own and each others' professional identities as leader and co-leader as their discussion unfolds.
Although Anthony and Liu's behaviors in this example seem to threaten each others' face, their disagreements are actually embedded in several face-work strategies. For instance, in disagreeing with Liu, Anthony skillfully avoids direct confrontation while still managing to exhibit his disagreement (e.g. by responding to Liu's decision with silence (line 1)). In contrast, by organizing the rationale for his initial decision in a sequence and by smiling (line 13), Liu is able to attenuate the imposition of his disagreement and to mitigate the threat to Anthony's face. This example thus demonstrates how face-work is often done in an extended and negotiated sequence rather than a single turn (Holtgraves Reference Holtgraves1992, cited in Lerner Reference Lerner1996), and how both interlocutors effectively attend to each other's face needs while at the same time saving their own face, even in disagreements (Zhang Reference Zhang and Kasper1995, cited in Cheng & Tsui Reference Cheng and Tsui2009).
DISCUSSION AND CONCLUSION
In this article we have explored another side of co-leadership constellations by focusing on those instances where leaders and co-leaders disagree with each other. In particular, we have analyzed how the members of two co-leadership constellations construct their (intertwined) professional identities by drawing on a range of strategies to attend to as well as to challenge and threaten each others' and their own face. Four representative examples have shown how interlocutors portray themselves as leader and co-leader in relation to each other by doing face-work in situations of strong and weak disagreements.
Using the analytical concepts of face and identity in exploring co-leadership constellations we have identified and described some of the complex processes through which co-leadership is actually enacted on the micro-level of everyday interactions. Our analyses have demonstrated how interlocutors constantly negotiate their professional identities and subject positions in dynamic and interrelated ways thereby taking “relational stances” (Geyer Reference Geyer2008): by orienting to or challenging each others' face, members of co-leadership constellations at the same time portray themselves (and each other) as more or less powerful and in charge, and thereby construct their intertwined professional identities as leader and co-leader. In particular, while one member is constructed as the leader (and thus as the more powerful actor and the one in charge), the other at the same time is positioned, at least for the moment, as the co-leader (i.e. the less powerful and thus second-in-command). These subject positions are fluid and change frequently as an interaction unfolds. As Clifton (Reference Clifton2006:209) maintains “[l]eader, like any other identity, is not an a priori label that participants carry with them. Identities are made procedurally relevant through the talk and participants literally talk themselves into being as the leader” and, we would add, as the co-leader and other related identities.
These processes are particularly obvious in situations where interlocutors disagree with each other. Interlocutors in our examples employed a range of different linguistic strategies when disagreeing with each other and negotiating their positions. They more or less explicitly contradicted each other, and depending on the severity of the disagreement they did or did not use mitigation strategies in their disagreements. Most of the linguistic strategies our participants used to do face-work when disagreeing with each other can be classified as “avoidance based” using Goffman's (1967) term: the disagreements were typically characterized by a “yes but” structure, the use of forewarners to signal disagreement (Kotthoff Reference Kotthoff1993), and the use of several mitigation strategies including laughter, silence, attempts to postpone the discussion and pronoun use. However, they also included the (sometimes repeated) use of disagreement markers (e.g. “no”, “not”) and considerable changes in speaker's pace and volume. Interruptions, cutting each other off, and comments like “I don't care” also characterized some of the disagreements.
So, how then do interlocutors do co-leadership when disagreeing with each other? As our examples have shown, co-leadership (just like leadership itself) is a very complex process that involves the moment-by-moment negotiation of power relations: interlocutors constantly negotiate who is “the one in charge” and who is the “second-in-command” in relation to specific issues. As a consequence, their positions within the co-leadership constellation change dynamically as the interaction unfolds. Even if one member of the co-leadership constellation is more senior and has a more powerful status (for example, in terms of seniority or position within the organizational hierarchy), our examples have illustrated that with regards to how co-leadership is actually enacted, these differences are often open to negotiation. Rather, co-leadership is a dynamic and ongoing process in which both members position themselves and each other as leader and co-leader at different moments throughout an interaction. This dynamic nature of co-leadership, we believe, can be captured and described particularly well by employing the analytical concepts of face and identity, and by exploring how face-work and identity construction are accomplished in and through interlocutors' everyday workplace talk.
These insights thus provide further support for the claim made by many researchers that leadership is a dynamic process that is enacted and reflected at the level of discourse. And while the specific ways in which this is achieved often remain hidden in more traditional approaches to leadership, they can be productively observed through micro-analyses of leadership discourse. Discourse-analytical approaches to leadership, like the one pursued in this article, thus seem to be promising additional resources with the potential to enhance an understanding of the complexities of leadership. In particular, our proposition to approach leadership phenomena through the analytical concepts of face and identity provides further specific tools and conceptual models that may be used to make sense of the complexities of leadership within the increasingly popular tradition of discursive leadership. As we have shown, such an approach could be particularly rewarding for those situations where leadership is enacted less harmoniously, such as disagreements. We would thus like to suggest that future research should increasingly look at these “other sides” of leadership that are largely under-researched but may reveal new insights into the complexities of leadership processes.
As Clifton maintains, leadership “researchers and practitioners require fewer ‘grand theories’ of leadership … . Rather, they require a better understanding of the everyday practices of talk that constitute leadership and a deeper knowledge of how leaders use language to craft ‘reality’ out of the ‘hustle and bustle’ of events that surround them” (2006:203). Following this claim, we hope that our analyses will provide further insights into the everyday practices of co-leadership. Clearly, more discourse-analytic studies like this one are necessary to start understanding the complexities of leadership phenomena and how leadership is actually done on the micro-level of everyday interactions.
Appendix: TRANSCRIPTION CONVENTIONS
- +
pause up to one second
- -
incomplete or cut-off utterance
- … // ….. \ …
- … / ….. \\ …
simultaneous speech
- ( )
indecipherable speech
- (hello)
transcriber's best guess at an unclear utterance
- (.)
untimed brief pause
- {it was}
words added in English translation to help comprehension
- […]
section of transcript omitted
- [laughs]
paralinguistic features in square brackets
- [laughs]: no:
laughter throughout the utterance of the word in between the colons
