Abstract
Although women in the United States make up about half of the workforce, only 14.6% of executive officer positions in the Fortune 500 and 16.9% of Fortune 500 board of director seats in 2013 were held by women, numbers that have remained flat for the past decade. Decades after the so-called “feminist revolution,” women are still struggling to be seen as leaders within organizations even though many have put in place hiring and recruitment policies to help eliminate this problem. Our study examines this disparity by observing how leadership emerges and is negotiated in discourse among male and female participants in decision-making groups in a masculine organizational culture. First, it identifies whether female participants randomly assigned to mixed-gender groups emerge as leaders. Second, it analyzes the discourse of those competing for leadership positions in mixed groups to identify the effects of leadership style on leader attribution by others. Of the 22 mixed-gender groups (N = 110) that took part in our study, no woman emerged as the unanimously chosen leader, even though women were identified as leaders by transcript coders. This article uses a case study approach to analyze leadership emergence in two mixed groups in which women were recognized by some members as demonstrating leadership. It then looks at a third case that demonstrates how some discourse behaviors that have been recognized as leadership may not be viewed as such in a masculine organizational culture. Study results illustrate how organizational culture can define accepted ways of “doing” leadership and affect who is and who is not recognized as a leader, particularly in terms of gender.
Keywords
Although women in the United States make up about half of the workforce, only 14.6% of executive officer positions in the Fortune 500 companies in 2013 were held by women, a number that has remained flat for the past decade (Soares, Bartkiewicz, & Mulligan-Ferry, 2013). That year, women held only 16.9% of Fortune 500 board of director seats, the same level as 2012 (Soares et al., 2013). Much has happened since the women’s movement that arose in the 1960s to better integrate women into the public sphere, but even after more than 50 years, they still lag behind men in leadership positions. This problem has been tackled by many organizations at the policy level, by putting in place programs to recruit and promote women, but as the numbers indicate, this approach has been far from successful.
Our study attempts to better understand this phenomenon by observing how leadership emerges and is negotiated in discourse among male and female participants in decision-making groups in a masculine organizational culture. It first identifies whether female participants randomly assigned to mixed-gender groups emerge as leaders. Second, it analyzes the discourse of those competing for leadership positions in mixed-gender groups to identify the effects of leadership style on leader attribution by others.
This study attempts to bring together research from two approaches to the study of leadership: what has been called the “psychological” approach and the discursive approach. It does so by first asking participants to identify the leaders of their group by identifying specific communicative traits they observed; we then look at the talk that is exhibited in the group interaction and how it creates certain leadership styles (Aritz & Walker, 2014). Ultimately, we are interested in looking at whether and how organizational culture affects the type of leadership that is recognized in mixed groups of men and women and how leadership is negotiated within a masculine organizational culture.
Discourse Studies in Leadership and Gender
An increasing body of research is studying leadership by looking at language and approaching the phenomenon as an act of social constructionism (Alvesson & Kärreman, 2000; Fairhurst, 2007, 2009). From this perspective, leadership is viewed in the context of what leaders do and is thus discursive in nature. According to Robinson (2001), “leadership is exercised when ideas expressed in talk or actions are recognized by others as capable of progressing tasks or problems which are important to them” (p. 93). According to Fairhurst (2008), this definition enables us to understand leadership as a process of influence and meaning management that advances a talk or goal, an attribution made by followers or observers, and a process, in which influence may shift and distribute itself among several organizational members. “To wit, leadership is co-constructed, a product of sociohistorical and collective meaning-making, and negotiated on an ongoing basis through a complex interplay among leadership actors, be they designated or emergent leaders, managers, followers, or both” (Fairhurst & Grant, 2010, p. 210).
This perspective contrasts with the psychological approach to leadership, which is prevalent in management studies, particularly in the United States, where a psychological lens and traditional empiricist methods still dominate (Alvesson & Sveningsson, 2003; Conger, 1998; Fairhurst, 2007; Knights & Wilmott, 1992). The concern of this approach is with the cognitive or social-cognitive origins of leadership and the perceptions they generate with weight given to the mental over the behavioral (Fairhurst, 2007). From this perspective, leadership is seen as residing within the individual and is often associated with certain personality traits divorced from the organizational culture.
More and more researchers, though, are treating language as a methodological question and a window into cultural meanings. A linguistic focus is also enabling scholars to rethink traditional approaches to business issues and in doing so, to reveal more nuanced details about how such issues as leadership are “brought off” (Fairhurst, 2007).
Like leadership, gender is a social construct. Gender is different from biological sex. It is a social patterning that has been created over time and that has been passed down from generation to generation within a culture. It is learned behavior that we enact each day to “create” our gendered selves. We do so through our clothing and accessory choices, our mannerisms, our vocal qualities, the way we walk, and talk, and the things we say and do, all activities that can be broadly thought of as communication if we understand that communication is a symbol system that conveys meaning to others through the process of perception. The meanings associated with specific communication displays over time have become coded as “male” or “female” and have thus created stereotypes that we rely on in order to make meaning of our environment.
Social constructionist theory contends that there are mainstream discourses of “gender difference” circulating in Western culture (e.g., Cameron, 2006; Sunderland, 2004), with the effect that the biological category of “men” is positioned to speak and behave in ways stereotypically coded as “masculine,” while “women” are positioned to speak and behave in ways coded as “feminine,” even though individuals can and do resist such stereotypical positioning.
Elements of Talk: Latching, Overlaps, and Questions
Discourse studies have thus focused on how features of talk are coded as feminine or masculine. For example, Coates (1996) found that women tend to construct talk jointly and that the group takes priority over the individual as women’s voices combine to construct a shared text. Utterances are often jointly constructed; in other words, speakers often cooperate to produce a chunk of talk. In addition, Coates observed that women friends often combine as speakers so that two or more voices may contribute to talk at the same time. This kind of overlapping speech is not seen as competitive, as a way of grabbing a turn, because the various contributions to talk are on the same theme.
Women’s talk is also characterized by the frequent use of questions whose main function is interactive rather than information seeking (i.e., the question, “there are limits aren’t there?” checks that a shared perspective obtains and does not expect an answer except perhaps for “yeah” or “mhm”; Coates, 1996).
While women’s voices combine and overlap, men take turns to hold court. Male friends prefer a one-at-a-time pattern of talking, with one speaker holding the floor for an extended period at any one time; overlapping speech is avoided and is viewed as contentious for seeking the floor. In terms of questions, men use them to seek information from each other, taking it in turns to play the expert. Table 1 below identifies widely cited characteristics of “male” and “female” talk.
Widely Cited Characteristics of “Feminine” and “Masculine” Styles.
Source. Adapted from Holmes and Stubbe (2003).
Elements of Talk: Amount of Talk
In more formal situations, the majority of studies find that men talk more than women. This outcome has been attributed to status characteristics theory, which focuses on how status differences organize interaction (Capella, 1985; Slater, 1966; Stein & Heller, 1979). According to this theory, individuals involved in social interactions evaluate themselves relative to the other individuals involved and come to hold expectations as to how and how well they will perform in relation to every other participant in the interaction (Capella, 1985; Slater, 1966; Stein & Heller, 1979). These “self-other performance expectations” provide the structure of the interaction, which then determines the subsequent interaction.
Research has shown that those with higher status participate more in task-oriented dyads or groups than those with lower status (Capella, 1985; Slater, 1966; Stein & Heller, 1979). Since men have traditionally held higher status than have women, one would expect men to talk more in task-oriented or instrumental situations.
Elements of Talk and Leadership Style
In our research on leadership styles (Aritz & Walker, 2014), we found that overlaps and questions were also used differently by different types of leaders (see Table 1). A directive leader uses questions to direct agreement on interaction participants, does not link his or her comments to the previous speaker’s statement, and makes abrupt topic shifts as well as uses minimal active listening techniques and tends to interrupt other speakers. Our research has indicated that those using the directive style often talk significantly more than others participating in the interaction. As such, the directive leadership style shares many of the common elements of what Coates has identified in masculine gendered talk.
In contrast, a cooperative leader uses questions to solicit information or participation from others, acknowledges the position or statement of previous speakers, avoids abrupt topic shifts, uses active listening techniques, and uses cooperative overlaps to show her support of other’s ideas. Our research indicates that those using a cooperative leadership style significantly reduce the imbalance of talk between leader and followers. Because of this as well as the use of questions and cooperative overlaps of this type of leader, this style is more in line with Coates’s description of feminine talk.
A third leadership style, collaborative, is also known as “distributed leadership,” which is defined as a property that emerges in team situations in which influence is distributed across multiple team members (Carson, Tesluk, & Marrone, 2007). In this style, participants use questions to frame the interaction and to check for agreement among members, acknowledge some of the contribution of others but more commonly build on other’s statements producing smooth topic shifts, even though these contributions may overlap with those of others.
Organizational Culture and Communities of Practice
Workplace settings play a critical role in the construction and enactment of members’ social identities. Organizations are “minicultures” that provide “sources and sites of identification for individuals” (Aaltio & Mills, 2002; Jenkins, 1996). More specifically, organizations contribute to the construction of member identities in at least two ways: They classify members into roles that have particular meanings and they develop discursive norms from which members draw to interact with others (Schnurr, 2009). Through these processes, organizations create leaders and subordinates.
Each organizational culture is different in the norms they provide to individuals to construct their roles. Hofstede (1980, 1998) describes masculine and feminine national cultures as representing the sex role pattern that is dominant in a given society and further suggests the masculinity-femininity dimension of a nation’s culture is reflected by organizations within that culture. Masculine cultures, such as Japan and Italy, emphasize the need for men to be successful breadwinners or be viewed as failures, and relatively few women occupy higher paying executive and top management positions. In Hofstede’s typology, American culture is considered moderately high in masculinity.
In feminine cultures, such as Sweden and the Netherlands, it is the norm for both men and women to pursue higher paying careers, and both males and females receive cultural support for prioritizing family time over time spent on the job. The women in higher level positions in these cultures are not necessarily expected to be assertive or to display the qualities and behaviors that are considered traditionally masculine (Hofstede 1980). Lyness and Kropf (2005) found that nations characterized as having feminine cultures tend to have organizational cultures that support work and family balance.
American organizations typically are characterized by a competitive, masculine organizational culture, which aligns with our “masculine” national culture. This organizational culture values respect for authority, competition, individualism, independence, and task orientation (Loden, 1985; Maier, 1999). Authoritarian management practices, respect for hierarchical structures, and adherence to chain-of-command are emphasized. Other values associated with a competitive organizational culture are assertive and aggressive behavior toward external or internal competitors and emphasis on individual, extrinsic rewards.
Supportive, feminine organizational cultures value and respect participation, collaboration, egalitarianism, and interpersonal relationships (Maier, 1999). There is less emphasis on hierarchical control; the supportive organizational culture focuses on group rather than individual rewards and places less emphasis on extrinsic rewards relative to intrinsic rewards (Loden, 1985). The cultural values associated with a supportive culture promote a balance of career and family roles, while competitive organizational cultures value commitment to the organization and the expectation that an employee’s career should be given priority over other roles (Maier, 1999).
In contrast to organizational culture, discourse researchers have used the concept of communities of practice as a means of identifying the linguistic strategies members use to negotiate organizational identity. A community of practice is an aggregate of people who come together around mutual engagement in some common endeavor. Ways of doing things, ways of talking, beliefs, values, power relations—in short, practices—emerge in the course of their joint activity around that endeavor (Eckert & McConnell-Ginet, 1992). A community of practice is different as a social construct from the traditional notion of community, primarily because it is defined simultaneously by its membership and by the practice in which that membership engages. It is the practices of the community and members’ differentiated participation in them that structures the community.
Speakers develop linguistic patterns as they engage in activity in the various communities in which they participate. In actual practice, social meaning, social identity, community membership, forms of participation, the full range of community practices, and the symbolic value of linguistic form are being constantly and mutually constructed (Eckert & McConnell-Ginet, 1992). The linguistic practices of any given community of practice are continually changing as a result of the many features that come into play through the interaction of its multiple members. In particular, organizations “provide a repertoire of procedures, contracts, rules, processes, and policies” that are then incorporated by the various communities of practice “into their own practices in order to decide in specific situations what they mean in practice, when to comply with them and when to ignore them” (Wenger, 1998, p. 245). Leaders and other organizational actors draw on this linguistic repertoire as well as the norms and values of their workplace culture to produce their discursive behaviors.
Workplace culture is thus a “communicative construction” that is “created and recreated as people interact over time” (Modaff & DeWine, 2002, p. 88). It is a system of shared meanings and values as reflected in the discursive and behavioral norms typically displayed by members that distinguishes the group or organization from others. It should be noted that organizations may be made up of multiple subcultures that may “co-exist in harmony, conflict, or indifference to each other” (Frost, Moore, Louis, Lundberg, & Martin, 1991, p. 8). Workplace culture contributes significantly to the establishment of norms and expectations about leadership by defining what competent and effective leadership means (Hickman, 1998; Schein, 1992). The relationship between workplace culture and leadership, though, is complex in that leaders themselves also play an important role in the creation, maintenance, and change of workplace culture (Neuhauser, Bender, & Stromberg, 2000; Parry & Proctor-Thomson, 2003; Schein, 1992).
Gender is also produced and reproduced in differential forms of participation in particular communities of practice. Women tend to be subordinate to men in the workplace; for example, women in the military have not traditionally engaged in combat, and in the academy, most theoretical disciplines are overwhelmingly male with women concentrated in descriptive and applied disciplines that “support” theorizing. The relations among communities of practice when they come together in overarching communities of practice also produce gender arrangements.
Holmes (2006) has shown that effective leaders are able to draw expertly on a repertoire of linguistic strategies stereotypically coded as masculine and feminine. The critical component, though, in determining their overall effectiveness is how actors are positioned by their community of practice. Like linguistic strategies, a community of practice can be either feminine, that is, supportive and team oriented, or masculine, competitive, and individualistic.
Mullany (2007) found numerous examples in her studies of management meetings whereby males use cooperative strategies and females use competitive strategies, depending on the community of practice in which they were situated. She argues that theorists should take greater account of the norms and conventions of different communities of practice, as well as institutional status, role, and corporate discourses, in order to achieve a more finely grained understanding of how different business communities “do leadership.”
This study thus takes up Mullany’s call by using a discursive lens to analyze the different communication patterns used in mixed-gender groups to negotiate leadership. It will identify the main leadership communication styles that emerged in small groups of business professionals and will then focus on the community of business professionals to see which styles were recognized as leadership attempts and which participants were identified as leaders by team members of these mixed groups.
Participants and Data Collection
All participants (N = 110) involved in the study were business professionals enrolled in an MBA program at a private university in Southern California with an average of 10 years of work experience. Participants were randomly assigned to groups of four to six persons each. This random division resulted in 22 mixed-gender teams.
The simulation used in the study, Subarctic Survival, asked each group to take the role of airplane crash survivors. Groups were then asked to discuss and ultimately agree on the ranking of items salvaged from the aircraft in terms of their critical function for survival. Although some may argue that this simulation will more likely call men to a leadership role because it involves an outdoor survival situation, we did not deem this a problem since the situation matches the masculine organizational culture in which the participants were involved, an MBA program located within a business school where both male faculty and students dominate, and aggressive, individualistic behaviors are encouraged and rewarded. The meetings were 20 minutes in length and were held and videotaped in an experiential learning laboratory equipped with professional facilities and technicians. The meetings were held in English, and the videotapes were then transcribed.
In order for participants to identify the leader of each group, we developed a communication style–oriented measure of leadership attribute preference using six global leader behaviors identified by the GLOBE Research Program—Charismatic/Value-based leadership, Team-Oriented leadership, Participative leadership, Autonomous leadership, Humane-Oriented leadership, and Self-Protective leadership (House et al., 2004). 1 Based on the definition of these six global leader behaviors (Dorfman, Hanges, & Brodbeck, 2004), we derived five communication styles that we used to measure leadership. We collapsed two separate GLOBE categories, Team-Oriented and Participative leadership, into one category “Involved other in decision-making process” based on the communicative moves that the leadership style would exhibit. The five communication styles that were derived were (1) decisive and task oriented; (2) involved others in the decision-making process; (3) modest, compassionate, and supportive; (4) independent and self-reliant; (5) status conscious and procedural. Participants were asked to complete this measure after participating in the simulation, since it was believed that a trait-oriented approach to leadership (as opposed to a discursive one) would be easier for them to apply. We used the results to identify the perceived leader of each team.
We also reviewed the transcripts to identify the leaders using discourse methods involving number of turns, turn length, and use of questions. Using these characteristics, three persons reviewed the transcripts and identified each member as a leader, a nonleader, or a transitory leader. (A transitory leader exhibits leadership behaviors at various times during an interaction and may share leadership with others. As such, this style of leadership more generally conforms with our collaborative leadership style.) The results were then compared to ensure intercoder reliability.
Methodology for Transcript Analysis
Two methods of analysis were used to interpret the transcript data: turn-taking patterns and interaction analysis. Both methods focus on a turn as the main unit of analysis to observe how contribution changes when multicultural groups involved in decision making are subject to different leadership styles. Turn taking is defined as the ordering of moves that involves the interchange of talking by speakers (Johnstone, 2002).
First, we used turn taking to analyze conversational interaction and to examine different leadership styles and group dynamics. Our specific method of analysis of turn taking is based on a model developed by Coates (1993) to analyze the management of naturally occurring interactions in which she describes cooperative and competitive conversation styles in gendered talk.
As alluded to earlier, Coates’s (1993) model of analysis focuses on the following areas: (1) The meaning of questions—are they direct in purpose or used indirectly to facilitate conversation? (2) Links between speaker turns—does the speaker acknowledge the contribution of the previous speaker or talk on the topic without acknowledging that contribution? (3) Topic shifts—are they abrupt or do speakers build on each other’s contributions? (4) Listening—is the speaker using backchannels or latching? and (5) Simultaneous speech—do the speakers overlap by elaborating on the previous contribution or does the contribution of the second speaker contradict or disrupt that of the first speaker?
These interactional elements are used to analyze how their combination affects the emergence of leadership within groups of business professionals. We followed Schiffrin (1994) and used her transcription conventions (Schiffrin, 1987; see the appendix) based on an earlier version of transcript notations by Jefferson (1979) to transcribe our data. Since we did not focus on gaze or vocal qualities in our analysis, we felt that Schiffrin’s conventions better served our needs. We did not include nonverbal clues, such as gaze and gestures, in our analysis because our focus was on language and the unit of analysis was limited to a turn as a vehicle to construct leadership in talk. Nonverbal elements may provide additional insights into our understanding of leadership, but they fell beyond the scope of this study.
Second, we used an interaction analysis approach, which involves the categorization of discourse units according to a predefined set of codes (Bakeman & Gottman, 1986). It is a quantitative approach to discourse analysis that draws from message functions and language structures to assess the frequency and types of verbal interaction. More specifically, we tracked member contribution by looking at three variables: the number of turns taken, number of words spoken, and the average turn length. We calculated the number of turns taken by looking at how many times a participant spoke in any given meeting. Turn length was used as another variable to measure member contribution. We measured an average turn length by dividing the total number of words spoken by each speaker by the number of turns they took. We selected a quantitative approach as a secondary method so as to more fully illuminate the findings from our qualitative analysis.
Findings and Analysis
In the 22 mixed-gender teams, no woman emerged as the unanimously chosen leader by team members. However, in two teams, women were selected by group members as showing leader attributes. This finding contrasts sharply with the results of the discourse analysis and the identification by three coders who identified eight women emerging as leaders of their respective groups, while in another six groups, women shared leadership with other women or men as transitory leaders. Of the remaining eight groups that were involved in the study, male leaders were identified in five and as sharing leadership in three by transcript coders.
The discussion that follows provides an analysis of leadership emergence in the two teams where at least one participant identified a woman as leader as well as a third situation in which a female unsuccessfully attempted to “do” leadership in her group but was not identified as a leader by any of participants. The first case illustrates the emergence of a more collaborative leadership style in which the two female participants identified the same woman as the leader of the group, whereas three male participants did not identify the female member as a leader in their group. The second case shows the interaction process of a group in which a majority of members were female—four of five—and in which the leader designation was divided between two women and the sole male. The third case illustrates an example of an unsuccessful attempt by a female to assert a leadership role using a cooperative leadership approach in a situation in which one male talks relatively infrequently but when he does, he uses a more direct style of leadership talk.
In the first two cases, the discussion will start with the results of the survey data showing leader attribution results. This discussion will be followed with the interactional analysis to reveal who talked the most and the discursive analysis of the leadership style used by each participant identified as a leader in the group.
Case Study 1
The first mixed-gender group was composed of five participants, two females and three males. Speaker 4, a female (S4F), identified herself as the leader and also was selected as the leader by the second female, S5F. Both participants marked the communication styles “decisive and task oriented” (#1) and “involved others in decision-making process” (#2) as helping them identify the leader. Speaker 1, S1M, a male participant, was selected as the leader by another male, S2M, who identified him as a leader because of the same qualities of being “decisive and task-oriented style” (#1) and “involving others in decision-making process” (#2). Notably, in both of these cases, the participants were looking for the same qualities in a leader but selected different persons, of different genders. S1M reported that there was no one leader and said that it was “leadership by consensus,” and a third male S3M was unable to select a clear leader, placing a question mark on the survey.
Table 2 displays the breakdown of the interaction analysis results by speaker in terms of amount of talk and turns taken.
Contribution to Decision Making by Group Member, Case Study 1.
As can be seen in Table 2, S4F, a female participant identified as a leader by herself and S5F, spoke the most words, while S5F, the second female, took the most turns. S1M, who was selected as the leader by another male, came in as the third highest in terms of both turns taken and words spoken, but was very close to another male participant S2M who identified S1M as a leader. The interaction analysis shows that despite the fact that different people were identified as leaders, the talk distribution among all participants in this group (number of turns taken and number of words spoken) was relatively comparable. The only person who was not contributing as much was a male participant S3M, who was not able to identify the leader and was not identified as a leader himself by anyone else. Thus, his inability to identify a leader might be caused by a relative lack of engagement with the group itself.
We will now look at the communication styles used in this group that led participants to identify the same leadership qualities and yet select different people as leaders.
Use of Questions
In this group, the questions in the very beginning are used to establish the collaborative nature of interaction in the group. The first couple of questions used by several members in the group are used to frame the type of discussion that will follow. It frames the type of discussion as collaborative as the group members are more actively engaged in coconstructing the rules and the process for discussion.
10. S2M: Do we wanna go around and just give like [our top 5? ]
11. S1M: [ What’s the best], what’s the [least]=
12. S5F: [Sure.]
Links Between Turns
In this group, the recognition of previous contributions is minimal but is present and positive. This can be seen in the “Okay” in Line 90 and the “Yeah” of agreement by S4F in Lines 87 and 93 in this example:
87. S4F: [Yeah, and I] figure if you can’t drink the streams, you can use the mirror to help.
88. you melt the water and then [you] just drink the snow.
89. S2M: [Or ]
90. S1F: Okay.
91. S2F: Or I was gonna say, you can, you can melt the snow in the metal can [from the 92. matches =
93. S4F: [Yeah, that’s
Topic Shifts
Speakers in this group tend to acknowledge and build on the previous speaker’s contribution and topic shifts are not abrupt but rather constructive. For example, in the same excerpt (given above), in Line 91, S2M elaborates on S1M’s idea and proposes a different variation that is introduced as an option by using a connector “or” that does not sound like an abrupt topic shift but more like a productive exploration of different alternatives.
Listening
The collaborative nature of the meeting can be seen in multiple backchannels that signal listening and agreement, even though they tend to be minimal. S3M and S2M use minimal positive acknowledgements of others’ contribution in the form of “Yeah” in Lines 110 and 113. S2M and S5F signal their agreement by using minimal responses, “true” in Line 118 and then “right” in Line 120. Lines 122 and 124 show multiple minimal responses, “right right” and then “okay,” that support the previous speaker and validate that the group is moving in the preferred direction in its decision-making process.
108. S5F: Well I put [as] my, my highest, uh the umm, the matches cause like you’re, you’re
109. wet =
110. S3M: [Yeah].
111. S5F: = [to the] waist, you’re heavily perspiring. You’re going to freeze to death
112. because it’s =
113. S2M: [Yeah. ]
114. S5F: = it’s almost certainly below freezing at that point and so you need to first, before
115. anything else get warm and dry.
116. S1M: I had [that,] I had that originally as matches. Actually I had, uh, matches and
117. Bacardi =
118. S2M: [True].
119. S1M: = because you can use Bacardi as fuel.
120. S5F: Right.
121. S1M: As lighter fluid.
122. S5F: Right. Right.
123. S2M: Okay.
Simultaneous Speech
In this group, there are frequent overlaps, but they are cooperative in the sense that they build on, expand, or productively question the previous speaker’s contribution. For example, S4F and S5F overlap in Lines 281 and 282 when they both elaborate on the same point that there must be wood if they choose to keep matches among their top-priority items. S1M and S2M overlap immediately following S4F and S5F in Lines 283 and 284 reinforcing the need to validate the assumption that there will be branches for them to use, “There’s gotta be . . . ” S4F uses an overlap in Line 285 to question this assumption but not as an abrupt interruption. Rather, she productively builds on the previous speakers’ contributions and introduces an element of doubt that is put out there for the group to discuss as they move forward with their decision making, “But will there be?”
277. S1M: You need . . .
278. S4F: z You need fuel. But, [what ] are we gonna do? Are we gonna burn a
279. tree =
280. S1M: [XXX].
281. S4F: down? [Are we gonna hope that there’s branches, right, right . . . ]
282. S5F: [There’s XXX dead wood on the ground ] [XXX XXX]
283. S1M: [We’re gonna XXX but there’s gotta be]
284. S2M: [There’s gotta be . . . ]
285. S4F: [But will there ] be?
In summary, this group’s method of decision making might be characterized as collaborative in terms of leadership style in which participants use questions to frame the interaction and to check for agreement among members. They acknowledge some of the contribution of others but more commonly build on other’s statements producing smooth topic shifts, even though these contributions may overlap with those of others. Because of this fast-paced interaction aimed primarily at collaborating to produce a solution, few active listening techniques are used. In this interaction, the women’s approach to the discussion was very similar to that of the males in the group. Regardless of this similarity, though, the collaborative leadership style used by both female and male participants in this group resulted in a split identification of leadership among group members. Obviously, something else is at work then that is affecting leadership attribution.
Case Study 2
In this team, four participants were female and one was male. Two persons, a male and a female participant, identified S3F as the leader, two females, including S5F herself, identified S5F as the leader, and one participant S3F identified S1M, the male participant, as the leader.
Reviewing the characteristics that members selected in identifying a leader, S5F was selected by herself for “involving others in decision making” and by another member as being “decisive and task oriented.” S3F was selected as the leader by the male member for being “status conscious and procedural”; the female who selected S3F as a leader did not identify the main characteristic that she used to determine the leader. S1M, the male team member, was identified as a leader by the fourth female for “involving others in decision making.” As these indicators show, not every member is looking for the same characteristic in a leader, and there is no pattern that emerges in terms of gendered preferences (see Table 3).
Contribution to Decision Making by Group Member, Case Study 2.
Looking at the results of the interaction analysis, S5F and S3F had very similar turn lengths (10.7 and 10.8 words per turn, respectively), but because S5F took more turns, 183 compared with 131 turns taken by S3F, she ended up with the largest number of words spoken, 1,962 compared with 1,414 by S3F. We can say, then, that if we look at the amount of talk as an indicator of leadership, S5F and S3F would likely be in the running. What amount of talk does not explain, however, is the selection of S1M by one participant as the group leader. He took only 82 turns and spoke only 740 words, half of what S3F and S5F did. This selection then may indicate a gender bias or that S3F saw something in his speech or manner that seemed more “leader-like” to her. This is particularly difficult to judge since she did not identify a specific leader characteristic.
We will now look at the leadership styles used by the three persons selected as the group leader to see whether this might influence leadership choice.
Use of Questions
S3F starts off by exhibiting the directive leadership style that corresponds to a masculine discourse style and is often marked by questions that are meant to direct and challenge members. For example, in Line 53 she uses a tag question to defend and reassert her decision.
53. S3F: Cause I would think, we gotta do something. Right? I mean we can’t just sit here.
54. and die. But on the other hand the way that they’re making it seem is that if we venture
55. out it’s almost sure death because you’ve got what ice, water that we can’t cross
56. S5F: I don’t know [about that]
In what follows, S3F uses a tag question in Lines 79 and 80 to further advance her decision to move rather than wait and in a way, challenges another option voiced by other team members.
77. S3F: That’s true. So they say that the streams and the lakes and all that stuff are
78. innumerable strands, lakes so we would have to cross over some of these really cold
79. wide rivers or whatever it is. So I think that we could die [doing that] I really do. Don’t
80. you?
81. S2F: [Oh, sorry]
82. S1M: Yeah
83. S4F: Want us to use the rope [or one another thing canvas]
What is interesting about S3F is that her attempt at leadership is not accepted by other team members but is contested by explicit disagreement, when S5F in Line 56 says, “I don’t know about that,” or by continued discussion of the previously introduced topic without acknowledging S3F (Lines 81-83).
After a few more unsuccessful attempts at leadership using a directive style, S3F moderates and shifts to a more collaborative leadership style. Her questions become more group oriented, checking for agreement, consolidating multiple contributions, and framing interaction, as in Line 181, “So what have we so far? Compass, book, alarm clock.” Her links between turns become collaborative rather than competitive by acknowledging previous contributions as in Lines 158-159:
158. S3F: That’s fine, yeah, that’s fine. If we are going to stay, there are a couple of things that
159. we don’t need. Northstar navigation I have as 15.
160. S1M: [Agree]
161. S2F: [I do too]
S3F shows active listening by using backchannels, such as “yeah,” “o.k.” in Line 232:
231. S5F: I was just saying seems like the things we’re keeping [so far]
232. S3F: [Yeah] Um. Ok [so]
244. S4F: [It’s important] but we have other things that are more important.
S3F switches to using simultaneous speech not as an interruption but either as cooperative overlaps or as a productive way to express a disagreement with the previous contribution as in Line 154; S3F questions the decision to rank the items without having agreed on the purpose of these items. This cooperative overlap prompts S5F to revisit the premise of decision making. This exchange is representative of the features of a collaborative leadership style.
151. S2F: Either way we have to rank them.
152. S1M: If we move we need more things basically because
153. S2F: But [we can’t
154. S3M: [But I’m just] saying we can’t rank them and then say oh this isn’t working
155. Out. Let’s go back to go
157. S5F: Let’s go back to staying and see what you guys are thinking.
S5F, on the other hand, uses a relatively uniform leadership style that includes elements of the collaborative style but is also marked by significant contributions and that does not rely exclusively on other team members’ contributions to the decision-making process, as seen in Case Study 1. S5F often times changes the direction of the meeting but does so in a different way than the directive leadership style initially used by S3F. She is constructing her authority by offering significant contributions to the discussion that are not driven by a more self-centered need to control the outcome of the discussion by centering her claims in “I” statements, but instead, she offers solid grounds for her claims, providing a logical basis for her arguments and ultimately marking her as a leader in some participants’ eyes. For example, from Lines 60 to 65, S5F questions the rationale of staying but in a less challenging and directive way than S3F does at the beginning of the meeting, and, as a result of her assertiveness, keeps the floor and has other members of the team seriously entertain her observation:
58. S5F: I don’t know [about that]
59. S3F: [Can’t cross]
60. S5F: I’m not trying to be difficult. I’m just saying I don’t know about that. I am concerned
61. about you won’t die there is plenty of water, water purification there’s maple syrup
62. which is plenty of sugar to get you through the number of days we have lots of
63. navigation but what we don’t know is when will people come by. If we’re just sitting 64. there you’re still going to have to use your syrup. Still going to have to get water and
65. use those other things
66. S4F: But you’ll use a lot less energy by just sitting there.
S5F uses questions to clarify statements made by other participants, as shown in Line 7 in the following excerpt:
5. S4F: [We just] say what we do individually [but as a] group z
6. S3F: [Right]
7. S5F: Why do you say two different things we have to do?
8. S2F: Because we also gotta decide if we are going to stay at the site [or go]
Or to check for agreement as shown in Line 74 below:
71. S3F: Ok.
72. S5F: I mean.
73. S3F: So you wanna vote?
74. S5F: Are you guys still all for staying?
75. S1M: Yeah
In all instances, the questions used by S5F are elicitations that help build an all-inclusive context for decision making. In contrast, S1M asked few questions; instead, his part in the discussion consisted almost entirely of responses to questions, many of which were aimed at the group as a whole. He often took the initiative to answer them before others had a chance as if he used them rhetorically rather than as actual questions to team members.
Links Between Turns
S5F showed strategies for linking to previous speakers. The links were acknowledging previous contribution and building on it, even if it means offering an alternative view that the group does not agree with (e.g., S5F in Lines 24 below):
21. S1M: You need to know a lot more if you’re going to move.
22. S4F: That’s a lot.
23. S3F: So, one more time.
24. S5F: So I mean normally I would say staying is a good idea if you think that people are coming by. What we don’t know is how often people are coming by. Fifty miles
SF5 handles the disagreement in a constructive way, acknowledging previous contributions but not contesting or challenging them. Similarly, in Line 41 and then again in Line 43, S5F disagrees with S3F, but does so in a way that moves the joint decision making forward without directly challenging other participants:
40. S3F: Cause I would think, we gotta do something. Right? I mean we can’t just sit here and die. But on the other hand the way that they’re making it seem is that if we venture out it’s almost sure death because you’ve got what ice, water that we can’t cross
41. S5F: I don’t know [about that]
42. S3F: [Can’t cross]
43. S5F: I’m not trying to be difficult. I’m just saying I don’t know about that. I am concerned about you won’t die there is plenty of water, water purification there’s maple syrup which is plenty of sugar to get you through the number of days we have lots of navigation but what we don’t know is when will people come by. If we’re just sitting there you’re still going to have to use your syrup. Still going to have to get water and use those other things.
This way, S5F demonstrates a leadership style that is different from the directive one. We have labeled this style as assertive leadership. Rather than dominating the meeting, S5F stands strong for her position, finally conceding and becoming an active participant in the discussion that follows. For her, the priority is to make sure that her opinion is taken into full consideration before making the final decision. It is not about the group simply accepting her point of view (directive leadership style); it is about stating her point of view with supporting explanation as well as considering other perspectives to reach the best decision given all available information.
In contrast, S1M did not link to other’s contributions, but primarily made assertions about his preferences or opinion on the matter at hand, commonly in response to a question.
Topic Shifts
Topic shifts exhibited by both S3F and S5F are smooth and relevant to the discussion at hand. For example, in Line 70, S5F introduces a new topic, which is her concern over the group’s prioritized list items, but she does it in a way that enhanced the decision-making process by opening a discussion on alternative scenarios that were not previously considered by the group.
67. S2F: And then also isn’t it going to be darker more because [of XXXX]
68. S5F: [It’s October]
69. S1M: Yeah
70. S5F: Alright my biggest concern is that no planes are going to come by and we’ll be just
sitting there and we don’t have food and things like that. If it’s a matter of I mean, I
recognize the rest of the water and things like that. I don’t see 50 miles or even 30 miles
as being that difficult of a walk. Even with the terrain I mean it’s several days it’s not
weeks and weeks.
In contrast, S1M changes the topic by directly and emphatically stating his disagreement as in the following statement:
S1M: I’m not, I’m not buying it. I’m not dying dehydrated. I’m putting my foot down. I’m not dying dehydrated z
Listening
S3F uses a moderate amount of backchannels to signal listening and agreement, as show in Line 232 and Line 246 below:
231. S5F: I was just saying seems like the things we’re keeping [so far]
232. S3F: [Yeah] Um. Ok [so]
244. S4F: [It’s important] but we have other things that are more important.
245. S2F: [Yeah]
246. S5F: [Yeah] If you’re going to say that we have matches and an axe to get more wood. How many different things do we need to burn? We’re in a place that has abundant amounts of wood.
In contrast, S5F and to a much greater extent S1M use minimal backchannels in the form of “yeah.”
Simultaneous Speech
In this group, participants use simultaneous speech either as cooperative overlaps or productive way to express a disagreement with the previous contribution, for example, in Line 154, S3F questions the decision to rank the items without having agreed on the purpose of these items. This cooperative overlap prompts S5F to revisit the premise of decision making. It is representative of the features of cooperative leadership style.
151. S2F: Either way we have to rank them.
152. S1M: If we move we need more things basically because
z
153. S2F: But [we can’t
154. S3M: [But I’m just] saying we can’t rank them and then say oh this isn’t working 155. Out. Let’s go back to go
156. z
157. S5F: Let’s go back to staying and see what you guys are thinking.
In summary, the group interaction shows some aspects of the collaborative leadership style, particularly as used by S3F in acknowledging previous contributions, using smooth topic shifts, demonstrating active listening, and using cooperative overlaps. In contrast, S1M uses elements that might be more characteristic of the directive leadership style. S5F, on the other hand, is more assertive than S3F but differs from S1M. She uses elements of the directive style, which is characterized by minimal listening acknowledgment and few links between turns. However, she is more collaborative, especially in her use of questions. She uses them not to direct members (directive style) but to check for agreement and frame interaction. As we mentioned earlier, for her, the interaction is not about forcing her preferences on others but making sure her opinion and expertise are taken into full consideration before a final decision is made. We labeled this leadership style “assertive leader” and included this type in Table 4 to represent an expanded list of leadership styles.
Differences and Similarities Exhibited in Discourse Styles by Leadership Types.
Case Study 3
This example illustrates how choosing a leadership style that is not recognized or preferred by the group may result in a failed attempt at leading. In the example below, S6, a female participant, attempts to establish herself as an active participant in the decision-making process using discursive moves characteristic of the cooperative leadership style.
Simultaneous Speech
In the following example, S6F tries to take the floor by overlapping to acknowledge the contribution of the previous speaker and elaborating on the topic initiated by S4M.
S4M: Alright. See the thing is, I, I think it’s more important to survive first before ou start moving, cause 50 miles, no matter how fast you walk will still take you like about, um, if you’re with all that stuff maybe 5, 6 days. So you have to make sure you can survive first cause no matter what you’re gonna have to stop and rest.
[So] [I figured it would be more important to . . . ]
S3F: [XX]
S6F: [Yeah, especially it’s, uh . . . ] close to arctic.
S4M: [Yeah.]
As can be seen in this interaction, the willingness to put forth an argument by S4M is similar to the exchanges of S3F and S5F in Case Study 2. S4M acknowledges S6F’s contribution with a brief “yeah” but does not apparently feel the need to build on it.
Links Between Turns
S6F does not speak again until a few minutes later in the meeting, when she talks on the topic directly connected to what has been said before: “[I think, ] so it’s October, close to Arctic, cause I’ve been to Alaska in the August, uh, the, uh in August the day, there, the day is really long . . . ”
Just as in her first attempt, S6F uses a cooperative overlap and contributes additional content by showing some expertise relevant to the situation with her statement, “I’ve been to Alaska in August.”
30. S3M: z Yeah, that’s true. [I was thinking of that] too.
31. S6F: [You have to ] z take it into consideration.
In her next turn, S6F receives validation from S3M in Line 30 (given above), at which point she overlaps again in Line 31.
Listening
S6F uses minimal responses in the form of yeah and mhm to signal listening.
Later on in the discussion, S5F challenges S6F by noting, “September is the fall, um, is the end of September is the equinox. So you’re getting about a 12 to 12 at, by October 5th.” S6F attempts to defend her statement “October, the day should be much longer than the night,” but after S5F continues to pursue her point, S6F finally concedes, “Oh, right.” At this point, S3M changes the topic, “Maybe, maybe we should determine what is most useless before what’s most important” and S6F overlaps again in agreement, “right.” She makes no further attempt to contribute in the meeting.
As shown, S6F uses some of discursive moves that are more characteristic of a cooperative leadership style. She acknowledges the contribution of the previous speaker and connects to and builds on the previous speaker’s topic on all occasions. But she did not use some of the more assertive methods of talk, such as questioning and changing the topic of discussion. This more stereotypically feminine style of leadership as enacted by S6F does not gain much traction in this particular group. The more directive style exhibited by S5M and S3M essentially serves to eventually silence S6F, even though she has demonstrated some expertise relevant to the discussion.
Discussion
As this analysis shows, the results of our survey provide little evidence that men and women are looking for different characteristics in their choice of a leader; both tend to prefer task-oriented leaders and those that involve others, to varying degrees. Regardless, our survey results indicate that both genders tend to select males as the leader, even though both genders may exhibit these characteristics during group interactions. This finding indicates there is likely a gender bias in leadership perception, and this is supported by our use of discourse analysis techniques to identify the leader in which our coders found that women were using some discourse techniques typically associated with a masculine style. Specifically, these involved taking more turns, producing more words, and using questions to move the task forward or illicit involvement from others. The results of this analysis found that women were leaders in the majority of groups, even though they were not generally recognized as such by their members, regardless of member gender.
In the three case studies that have been reviewed here, there does seem to be a preference for a more aggressive style of leadership, the directive style, exhibited by the male participants. The two females in Case Study 1 tended to be more assertive and direct and less supportive in terms of recognizing others’ contributions. This led to a more collaborative approach to leadership within that group, even though only one member noted this style—consensus or collaboration—on his survey. In the second case study, a fourth style of leadership emerged, one in which a female member used some aspects of the cooperative style to recognize others but also put forth sustained arguments to support her position, while a second female began by using a more directive style but then changed to a collaborative approach. The male in this group tended to favor the directive approach, even though he contributed much less than the two previously mentioned women. In this case, we can see at least two women adapting their discourse behaviors to better suit the group’s decision-making process. The women appeared to be more responsive to the group dynamic, or context, and thus at least intuitively recognized that leadership is coconstructed by other group members in the sense that they must accept or adapt to the discourse that is being produced. This observation illustrates how a trait approach to leadership oversimplifies the process of leadership production in the sense that it does not necessarily reside in a single individual but is negotiated among participants, depending on contextual factors.
The third case study shows a female using a cooperative style without success in a group in which two males are using the directive approach. Clearly in this case, the female participant did not adapt her preferred leadership style to that of the group. Her style of leadership, the cooperative one, apparently was not recognized as such, at least in this case, by members of this particular community of practice.
Conclusion and Implications
As this study shows, women participants were not generally recognized as leaders in decision-making groups, at least not in the traditionally masculine cultural setting of an MBA program in a business school where only 20% to 30% of the students are female. 2 This study attempts to illuminate one possible cause of this stereotyping: The type of leadership style that women tend to exhibit may not be recognized by others as “doing” leadership by members of this particular type of organizational culture or community of practice. No women were unanimously selected as “the leader” by any group, but in the few groups in which some women were recognized as demonstrating leadership characteristics, the females used either a collaborative or assertive style of leadership talk. These results also suggest that regardless of whether women know how to demonstrate expected leadership behaviors, they may still be ignored or not recognized by others as a leader. The results of our survey show a clear bias toward men as leaders, regardless of whether women demonstrated leadership skills, at least in a masculine organizational culture.
As such, it may do little good to teach or train women in leadership skills if they work in an organizational culture that values masculine leadership traits over others. As this study indicates, it is likely they will not be recognized as a leader regardless of their ability to lead. If true, organizations need to devote more resources to proactively developing organizational cultures that are more supportive and collaborative, that is, more feminine, if they hope to create environments where bias against women is reduced. In fact, a recent study by the Ketchum Leadership Communication Monitor (2014) predicts that the new global workplace will require a new, more “feminine” leadership communication archetype because, according to the study, female leaders outperform male leaders on most important leader attributes. It predicts the end of the traditional macho leadership style era, strong, silent, imperious, domineering, and invariably arrogant—in favor of transparency, collaboration, genuine dialogue, clear values, and the alignment of words and deeds.
Business schools thus need to do a much better job of educating and training students and employees about the importance of developing organizational cultures that are more inclusive. In doing so, leadership needs to be treated, not simply as a collection of traits, but perhaps more accurately as a set of communication practices that can be contested and that construct leadership jointly and temporally. This focus on communication makes the need to discuss organizational cultures using the framework of community of practice obvious. That is, understanding how organizational culture is created through talk means that business schools should educate students to better understand this concept as well as the elements and value of ways of speaking in creating more inclusive organizations. To put this more bluntly, teaching women to enact masculine leadership behaviors will likely not aid in improving their numbers in management. Instead, business schools need to focus on the value of changing organizational cultures to become more supportive and inclusive and to understand the value of talk in creating those realities.
As this study indicates, discourse analysis can provide a more finer grained approach to study both the production of organizational cultures and leadership. For example, it is likely that students and employees hold unexamined stereotypes about the type of talk that may be recognized as “leadership” in certain organizational settings. Education on the elements of talk and their use in creating different styles of leadership may help open up the possibilities for what is considered leadership and in turn, better enable women, and others who do not fall within a certain stereotypical field, to be seen as leaders within organizations. Understanding that discourse also provides the building blocks of culture can enable managers to more proactively and confidently engage in creating more supportive, inclusive work environments.
A key step is to understand culture not as something an organization has but as something an organization is. In this latter view, “an organization is an expressive form, a manifestation of human consciousness. The social world is seen as constructed by people and reproduced by the network of symbols and meanings that people share and that make shared action possible” (Alvesson, 2013, p. 22). Exerting greater efforts to identify the types of leadership communication behaviors that are recognized in organizational cultures or particular communities of practice and then educating members to be more aware of and more accepting of various leadership styles and ways of talking “reality into being” may help resolve some of the “glass ceiling” issues faced by women that have been well documented elsewhere.
This article provides educators and managers with the tools to begin looking at how leadership manifests in discourse in very explicit and concrete ways and its reception (or not) by others. The outcome may provide significant potential for improving organizational cultures and communities of practice in terms of the greater recognition that women and others might receive with a broader view of what leadership is and a resulting payoff in productivity and knowledge production and dissemination. But this change will not come easily. As Gordon (2010) and Reardon and Reardon (1999) have noted, efforts to change traditionally masculine cultures where masculine forms of power are unquestioned and entrenched are difficult, at best. Even so, this knowledge should be widely distributed so that women who strive to take leadership roles can make more informed choices about the organizations they choose to join.
Footnotes
Appendix
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.