Women and Leadership in Higher Education in China: Discourse and the Discursive Construction of Identity

: Prior research indicates that just 4.5 percent of mainland China’s higher educational institution leaders are female. This article extends theory and research by drawing attention to identity and Discourse as an important, yet under-researched, aspect of the problem of women’s underrepresentation in higher education leadership. Drawing on in-depth qualitative interviews with nine female academics in Chinese universities and informed by discursive approaches to identity and constructionist views, we analyze how women construct multiple identities, the interplay of identities, and the inﬂuence of broader societal Discourses of gender and leadership. The ﬁndings highlight the interplay between competing multiple identities, and illustrate how the women’s identities are shaped and constrained by dominant historical and cultural Discourses in Chinese society, which results in identity regulation (Alvesson and Billing 2009), notably identity positioning that is congruent with social norms and conventions. A key ﬁnding is that the female academics reject the leader identity. This is true for those in middle management positions, as well as women in early career stages, who might otherwise aspire to leadership. Implications for the leadership pipeline in China’s universities is discussed and recommendations are made for future research directions.


Introduction
Although women around the world face many common barriers to higher education leadership, China presents a distinct case (Zhang 2005) Since the Han Dynasty (B.C. 207-A.D. 202), Confucianism has dominated social and official ideology and is said to have attributed to the subordination of women (Mak 2013).Traditionally, under this regime, women were asked to obey 'the three obedience's and the four virtues'.The three obedience's require a woman to obey her father before marriage, her husband during married life, and her sons in widowhood; the four virtues represent physical charm, fidelity, propriety in speech, and efficiency in needlework (Cheng 2008).Traditionally, the pursuit of a career and the possibility of a high position in society was a man's job.Men were likened to being the pillar or backbone of the family.Although many women undertook waged work, historically, they occupied roles of a lesser status in the workplace (Mak 2013).After several thousand years of Chinese feudal society, it has been difficult to eliminate old ways of thinking (Zhao 2008).Although in the past few decades, particularly since the 1990s, significant social transitions have taken place in China (Zhang 2005), Chinese society continues to attribute well-defined roles to men and women (Attané 2012).
Although nowadays women's employment rates in China are amongst the highest in the world, with almost three women out of four in the labor force (Attané 2012), with many women occupying management roles and taking responsibility for promoting the construction and development of universities, relatively few women progress to senior leadership (Angeloff 2010;Sheng 2009;Wang and Kai 2015).In higher education, a survey of presidents (the equivalent of Vice Chancellors in the UK context) in 1792 of mainland China's higher educational institutions, conducted by Renmin University of China, found that only 4.5% of leaders were female (cited in (Xue 2008)).Another study of the resumes of 7796 top-level university leaders at 1166 Chinese universities found a comparably low proportion of female leaders (Wang and Kai 2015), and a study of female leaders in a top university group, referred to as Project 985, found one or two female leaders in each institution, and some universities had none.Moreover, women were found to predominantly occupy deputy positions and serve as Communist Party leaders (Appendix A), rather than university leaders (Wang et al. 2013).
Explanations arising from prior research pinpoint a range of sociocultural barriers that may explain women's underrepresentation in leadership, specifically male-dominated networks, social stereotyping, gender discrimination in promotion and work overload due to dual work and family responsibilities (see (Zhang 2005(Zhang , 2010;;Huang and Aaltio 2014)).Further challenges arise for women because of dominant historical Discourses of gender and leadership.Even in contexts such as education where women often outnumber men, prototypes of leaders are predominantly male and so are the scripts that set the norms of associated attributes and characteristics for leaders (Eagly and Karau 2002).A situation of this kind fosters a masculine Discourse of leadership where leadership is viewed as male gendered, and this interplays with historical and cultural Discourses of gender that foster an idealized image of women as caring housewives and mothers (Fairhurst 2009).Since, gender plays a profound role in the formation of professional identity and acceptance within a community of practice (Nagatomo 2012), dominant Discourses of gender and leadership shape identities (Ford 2006).
It is against this background that our study, based on interviews with nine women sampled from two universities in China, analyzes how the women construct their (multiple gendered) identities, in relation to broader societal Discourses of gender and leadership.This is guided by three key questions: 1.
How do female academics define their identities?2.
How do they talk about the female leader identity more generally?

3.
In what ways is this talk (discourse) intertwined with larger societal Discourses of leadership and gender?
Following social constructionist perspectives (Vivien 1995), and informed by the classic literature on 'doing gender' (Butler 1999;West and Zimmerman 1987), we view (gendered) identities as talked into being and performed through interaction, influenced by historical, societal/cultural, and institutional factors, rather than being in there somewhere, an a priori identity that pre-exists talk (Van De Mieroop and Clifton 2016).Identity is treated as something we 'do' rather than something we 'have' (Fletcher 2004).Furthermore, individuals can have multiple identities throughout their life, related to social roles in professional and private spheres of life (for example, mother, wife, lecturer, leader) (Alvesson et al. 2008).
We explore identity from the perspective of discursive practices (Wiggins 2017), which allow the researcher to delve deep into the context, especially historical and cultural or political aspects and view this context as a medium for the social construction process (Fairhurst 2009).By taking a discursive and social constructionist perspective, we seek to provide a dynamic view of identity construction in the study context that reflects how individual, organizational, and socio-historical influences reflexively interrelate at a moment in time (Jaros 2012).This approach emphasizes the active role of language in the construction of reality (Lamsa and Sintonen), especially distinguishing discourse (little-d) and Discourse (big-D).Gee (1999) explains that little-d discourse refers to the micro-practices of talk, emerging in interviews, through which identities are talked into being, and big-D Discourse refers to 'the entire interlocking web of practices, structures, and ideologies: a system of understanding and exploration that prefigures which practices and interpretations are available, and how practices and structures are understood' (Kiesling 2006, p. 262).Put simply, discourses (little-d) and (big-D) are reflexively linked, since talk emerging in little 'd' discourse is affected by the larger societal Discourses available to an individual, which they enact and make relevant through talk (Jones and Jonathan forthcoming).Although individuals have some room to exercise agency, gender identity and related gendered roles are shaped and regulated by powerful Discourses in society (Alvesson and Billing 2009, p. 101;Lindsey 2015).
In this perspective, our study contributes to understanding the process by which women construct their identities, considering the multidimensionality of identities, such as the intersections between different social identities, within the context of dominant Discourses of gender and leadership pertaining to history and culture in China.
The paper begins with an outline of the theoretical framework underpinning the study, beginning with a discussion of universities, it draws on the literature on gender, identity, Discourses and leadership.Next, we explain the method adopted for the study, then we present segments of data from nine interviews with our analyses, followed by a discussion of the findings and conclusion.

Theoretical Framework
Historically organizations have been built upon the expectations and aspirations of men (Witz and Savage 1992), and universities are no exception (O'Connor 2014).It is said that universities are dominated by masculine Discourse (Harley 2003;Harding et al. 2010), and masculinized hegemonic structures, with the overwhelming majority of senior occupations occupied by men and definitions of excellence and merit (O'Connor 2014), and norms and values (Savigny 2014, p. 797) that benefit men but discriminate against women (Morley and Crossouard 2016).Masculinized cultures foster homosocial networks (Fletcher et al. 2007).These can resemble an exclusive gentleman's club (O'Connor 2014;Thomas and Davies 2002) and appear unfriendly (Morley and Crossouard 2016;Davies and Holloway 1995), or even hostile toward women (Williams et al. 2006).The higher up the hierarchy one goes, the more social capital appears to matter (O'Connor 2014).Precisely because Chinese society is permeated by relationships (guanxi), men benefit greatly from the social capital and opportunities that social networks provide (Lyness and Thompson 2000).Women, on the other hand, are constrained in taking part in important social networks by powerful socio-cultural barriers (Zhang 2005;Huang and Aaltio 2014;Zhang 2010).Female academics in Chinese universities are affected in many other often subtle ways (Zhang 2001).For instance, prior research suggests that women can be deterred from pursuing or accepting leadership positions because of the 'organizational culture of male-dominated senior management [ . . .], power politics, and exclusive networks among male colleague' (Zhao 2008, p. 78).In consequence, the opportunity structures for males and females in Chinese universities are different and 'social prejudice against women is very strong' (Zhang 2005, p. 105).
Studies confirm that gender is especially pertinent to cultural scripts for women (Lindsey 2015), such that gender Discourse is infused with historical and cultural patterns of thought regarding the category 'women' and this may shape and constrain the identity work of women who transgress social norms by entering professions not stereotypically associated with the category 'woman'.As such, the 'doing of gender' (West and Zimmerman 1987) involves enacting specific identities through a repertoire of practices (Martin 2003), associated with stereotypical and socially constructed masculine and feminine characteristics linked to biological categories of males and females (Holmes 2007).Exactly because individuals are held accountable to their sex category 'doing gender' (West and Zimmerman 1987), and because they are validated for 'doing gender well' (Mavin and Grandy 2012, pp. 3-4), 'gender is a powerful ideological device', which produces, reproduces and legitimates the choices and limits that are predicated on sex category' (West and Zimmerman 1987, p. 72).
The juncture between gender and identity is mediated by powerful historical and cultural scripts.Identity is shaped and constrained by social norms that one is required to abide by to be accepted by others (Ashforth 2000), shaped by the Discourses available, through culture, which individuals draw on in communications with other people (Vivien 1995, p. 123).From this perspective, Discourses and related discursive practices provide the means through which identities are crafted (Ford 2006, p. 79).
For instance, Discourses, such as patriarchy, influence the identity of an individual (Collinson 2003;Halford and Leonard 2006).In consequence of masculinized cultures within higher education, 'women's academic identities are often forged in otherness, as strangers in opposition to (privileged) men's belonging and entitlement' (Morley 2010, p.38).Although individuals have some room for agency in crafting their identity, this agency is shaped by available Discourses (Jaros 2012, p. 49).Since values, norms, and rules are expected to be adhered to if one is to behave in accordance with social expectations, identity work involves some degree of identity regulation (Alvesson and Billing 2009).In consequence, individuals are constrained, to some extent, by social structures, power relations and other factors in their social context (Alvesson and Billing 2009, p. 99).
Theories of Discourse acknowledge the fact that although occupational roles are given 'off the peg', they come with cultural 'scripts' attached-patterns of belief, values, attitudes, expectations, ways of thinking, and so on (Billington et al. 1998, p. 50).Since leadership is historically rooted in the symbolic sphere of the male and represented in Discourses of hegemonic masculinity, constructing a leadership identity can be challenging for women.In Chinese culture women are highly valued for being humanistic, people-orientated, and considerate to the feelings and needs of others.By contrast, women leaders are considered 'ambitious' and 'aggressive'-an identity that is not congruent with the 'virtuous wife and caring mum' (Xian qi liang mu) Discourse that is highly valued in Chinese culture (Zhao 2008, p. 78).In Chinese culture it is customary for leaders not to 'invite others to be involved in goal setting or decision making' and to use 'status and position to make independent decisions without the input of others' (Northouse 2012, pp. 350-51).Thus, powerful cultural scripts constructed in relation to gender (Lindsey 2015) clash with traditional hegemonic masculinist Discourses of leadership that characterize leaders controlling and self-reliant individualists (Ford 2006, p. 84).
Nevertheless, it has been suggested that post-heroic developments in leadership studies that emphasize so called feminine styles of leading that are more relational, local and shared (Ford 2006;Fletcher 2004), should create a feminine advantage for women (Eagly and Carli 2003;Helgesen 2008).Although this rhetoric has been around for several years (Fletcher 2004), dominant managerial practices in higher education persist to encourage masculine Discourses of competitiveness and individuality, not feminine Discourses of empathy, supportiveness, and nurturing associated with women (Thomas and Davies 2002).Buckinshaw (2015) likens higher education communities of practice to communities of practice of masculinities, whereby hegemonic masculinities shape the learning of leadership and the doing of leadership, and effectively perpetuate masculinities.Since success and productivity is measured according to masculine norms, this has a detrimental effect on female academics in Chinese universities (Zhang 2001), just as it does in other parts of the world (O'Connor 2014; Van den Brink and Benschop 2012).
Eagly and Karau (Eagly and Karau 2002) argue that because of perceived incongruity between the female gender role and leadership role women may be evaluated less favorably as potential occupants of leadership roles and leadership behavior may be evaluated less positively when enacted by a woman.Consistent with role incongruity theory, women are more likely to experience negative reactions, such as dislike and rejection, than men for showing dominance, expressing disagreement, or being highly assertive or self-promoting (Carli 1999;Copeland et al. 1995;Rudman 1998).This can result in backlash not only on the job in leadership evaluations, but also in hiring, promotion, and salary negotiations (Brescoll 2011;Rudman and Phelan 2008;Rudman et al. 2012).Research in Chinese universities suggests that role congruity is prevalent due to social stereotyping of women and gender discrimination in promotion (Zhang 2010).Research from other parts of the world suggests women may regulate their behavior out of the fear of backlash (Brescoll 2011).
Navigating and negotiating inhospitable cultures can be exhausting work.Research with senior women in higher education in the UK confirms that considerable energy and emotional labor is expended fitting into masculine gendered leadership cultures (Burkinshaw 2015).In consequence, Alvesson and Willmott (2002) contend that in order to achieve acceptance in the work place and maintain social solidarity, people develop self-images and work orientations that are deemed congruent with the culture of their organizational context.The achievement of this goal involves considerable identity regulation and 'identity work'.This prompted Burkinshaw (2015), citing Alimo-Metcalfe (2010), to jest that the job specification for women in higher education leadership should specify they will need chameleon-like features.Goffman's (Goffman 1959) classic theory of impression management provides another framework of analysis.Goffman shows how social order is maintained by individuals through the suppression of their own desires to maintain a working consensus.Through impression management and by presenting themselves to others as 'dramaturgical', they make the performance convincing to the audience.For example, in cultural contexts where caring for the needs of others is a valued identity, research shows that women tend to categorize themselves in that way (Kondo 1990).This is of relevance to our study since traditional ideology regarding gender roles promotes images of women as housewife's and mothers in China.The view that women's work is inside the home remains strong even among women (Zhao 2008;Yang 2011).
Studies show that because women experience a heavy dual burden of domestic and waged work, this can inhibit career development (Zhang 2001) and career ambitions (Zhao 2008).This inadvertently perpetuates the idea that a 'proper academic' is male (Skelton 2004).Those women that do pursue academic career success have been found to experience psychological conflict and work overload performing both work and family roles (Zhang 2010).In addition, they have to work hard to prove themselves as capable as men (Zhang 2005).Perhaps unsurprisingly, studies reveal women experience more stress, specifically due to conflicts between work and family than their male counterparts (Zhang 2003).This raises questions of how they construct their professional identities, since 'identities are embedded in social and family life, as well as hegemonic leadership Discourse' (Ford 2006, p. 89) and historical cultural Discourses of gender.Even for women unencumbered with caring responsibilities, 'the abjected maternal body is displaced onto all women (whether they are mothers or not)' and 'conflated with the feminine' (Fotaki 2013(Fotaki , p. 1257)).

Materials and Method
Adopting an inductive interpretative research strategy, the study sought to bring women in to 'dig deep' into their life-world (Berger and Luckmann 1966) to reveal aspects of their lives that may be invisible (DeVault 1996, p. 32).The data were collected through semi structured interviews because these are a form of interaction, as natural as any other (Atkinson and Silverman 1997), and benefit from following conventions participants are familiar with (Shakespeare 1996).In addition, interviews were deemed appropriate since 'talk produced in interviews can be analyzed as part of the ongoing project which is the speaker's identity work' (Taylor and Littleton 2006), and interviews can even provide an appropriate context for identity work (Taylor and Littleton 2005, p. 28).
Discursive research into professional identity is rooted in the role of language, meaning, and Discourse in the development of identification processes (Jaros 2012).As Burr (Vivien 1995) articulately puts it 'our identity therefore originates not from inside the person, but from the social realm, where people swim in a sea of language and other signs, a sea that is invisible to us because it is the very medium of our existence' (p.53).However, language is not just a by-product, people use language to do things, such as pinpoint blame, make excuses, project themselves in a positive light, and so forth (Gill 2000, p. 175).Therefore discourse (talk) incorporates not just language but social practice, too (Gill 2000, p. 73).Expanding on the social practice perspective, Discourse is described as 'a systematic set of beliefs, ideas or knowledge and practices specific to social situations or locations' (Billington et al. 1998, p. 33).These emerge in talk and we refer to these as 'little-d' discourses.'Big-D' Discourse can be described as systems of thought.Furthermore, professional identity is produced by Discourses, located both within the workplace and at a broader societal level (Foucault 1972;Foucault 1980).The purpose is not to report generalizable accounts of women and higher education leadership in China, but to reveal and challenge the so-called 'truths' located in cultural meanings and beliefs emerging in Discourses about women and leadership.
The sample was arrived at through a snowball sampling technique (Browne 2005).Two universities (University F and N) in the capital city of a northwest province in China were approached to begin this process.Two male deans, one from each university, provided introductions to potential participants.Nine female academics agreed to take part in the study.Following ethical approval, individual interviews were carried out in 2016, over the duration of a week, first at University F followed by University N. The researcher conducting the interviews is a native Chinese speaker and one of the authors of this paper.An interview schedule of semi-structured questions was used.This was translated into Chinese by the interviewer who is the first author of this paper.Respondents were not given a copy in advance, as this can result in rehearsed responses.The first interview was used to pilot the question schedule.No changes were required and the same method and questions were used in the interviews that followed.Nevertheless, if some words or sentences in interview questions caused comprehension barriers between languages, these words were explained thoroughly to the interviewees.The interview schedule began with initial questions about the participant's education and career trajectory to date.More in-depth questions followed to allow the participants to talk in depth about their role, how they view leaders and if they consider themselves to be a leader.This was followed by questions about their experiences and views on any dual professional and private family responsibilities they had.These interviews can be understood as social interaction between two people (Basit 2010), that allowed the researcher to approach the interviews as 'an open-ended input to identity work' (Alvesson and Willmott 2002, p. 640).
The interviews were recorded and transcription was undertaken by the researcher.They were then professionally translated into English.All personal details and information pertaining to the universities in which the participants work was anonymized.A process known as inter-rater reliability (Armstrong et al. 1997), involved reading the transcripts of all nine participants several times to assess consistency with regards to English translations and to provide a form of auditing (Lincoln and Guba 1985).This was repeated at subsequent stages of analysis to check reliability and validity, with regard to participant's understanding and coherence, for example (Gill 2000, p. 189).Through immersion in the data the researchers also gained familiarization with the material (Gill 2000).The material was then subjected to thematic coding (Yin 2014).The process was iterative as it involved cycles of inquiry back and forth between theory and the empirical data throughout the process (Strauss and Corbin 1990).Various dimensions of identities emerged, which reflect the different career experiences, life histories and attitudes of the women, which we explored further, guided by discursive analysis to pinpoint not only Discourses but also subject positions, positioning and subjectivity, in order to explore ideological aspects of Discourse and its influence on subjectivities (Wiggins 2017, p. 33).
Table 1 provides information about the sample.All the women began their careers as teachers (instructors/lecturers).They represent women at different career stages and with different levels of experience of leadership, ranging from very minimal early career committee roles to middle management level roles.Among these nine participants there is one Dean (head of school), one Associate Dean, two Department Directors, and one Section Chief.Four of the women are instructors/teachers with additional responsibilities as Secretary of the Youth League Committee (Appendix A).There are two systems in Chinese universities: the management or administration system, and the teaching system.These are independent of each other and teachers in the administration system and the teaching system undertake different types of work.First, the administration system focuses on the management of school administrative issues.Participants in this study working in the administrative system include the Section Chief and the four instructors-although they give a few lessons to students, their titles belong to the administrative level (for example, Section Chief, Division Chief, Director-General, and Minister, etc.).Second, participants working in the teaching system devote most of their energy to teaching.They include: the Dean, Associate Dean, and Department Directors.Their titles belong to Teacher professional titles (for example, Assistant, Lecturer, Associate Professor, and Professor).Regarding the mixed sample of women, the focus was placed on leadership experiences at different levels of higher education, ranging from teachers to a dean.Following Northouse, leadership is defined as: 'a process whereby an individual influences a group of individuals to achieve a common goal' (Eagly and Carli 2003, p. 5).The two universities are located in an important industrial city which acts as a transportation hub.The population of this city was approximately 3.69 million in 2015, and it has thirteen higher educational institutions (http://www.tjj.lanzhou.gov.cn).The two universities are typical of universities in the northwest of China.They have over fifty-five thousand students over three thousand teaching and administrative staff in total, so are well-known locally.

Analysis and Findings
Segments of narrative from the interviews follow to illustrate our analysis of the data.Since the term manager/management appears interchangeably with the term leadership in the narrative, it follows that use of these terms may be called into question.It would be futile for us, as discursive scholars, to embark on a lengthy discussion on this, as leadership from a discursive perspective is in effect a set of 'language games' not in the strictest linguistic sense (Fairhurst 2009(Fairhurst , p. 1609), but grounded in the everyday realities of individuals who aspire to lead or be led (Kelly 2008).From this perspective, what is defined as leadership or management is not of concern, it is the context, especially historical and cultural or political aspects that are of interest to our analyses.In what follows, we begin to set out our analyses by exploring the interplays of multiple identities and Discourses of gender, followed by a more in-depth study of the 'leader' identity, then we present our analyses of respondents' discourse concerning female leadership and the broader societal Discourses of gender and leadership that are invoked.

How Do Female Academics Define Their Identities?
First, our analyses reveal the interplay and dynamics of multiple identities, derived from different social roles performed by the women in private and professional realms of their lives.Although the respondents do not share the same portfolio of identities (e.g., wives, mothers, teachers, managers or leaders), they all invested in the articulation of their own particular multiple identities, and the interplay and relative importance of each identity.This is illustrated in an excerpt from Respondent O, a Department Director: 'I am a good mother, then a good teacher, and last a good manager.' While each identity (e.g., mother, teacher, manager) appears to be distinctive to the other, there is a clear interplay between each of these identities.Notably, the presentation of such accounts in sequential order and within a particular temporal structure, highlights that certain things need to be known first (Wiggins 2017, p. 124).Most prominent in the data is the notion that to be a successful leader or manager, and to achieve legitimacy in that role, one must first be recognized as a good teacher, as illustrated in another segment from the previous respondent: I think I should be a good teacher, because in university, your ability in teaching matters a lot.Only if you have a good ability of teaching, other staff are willing to respect your management (Respondent O, Department Director).
Such utterances reflect the professional path shared by respondents, since all of them began their career in higher education as teachers (lecturers/instructors).However, identities should not simply be likened to job roles, as synergies, and interplays between identities arise, not least because the accomplishment of a legitimate identity in one realm is perceived to act as a precursor to success in another.As such, the accomplishment of legitimacy as a leader (or manager) interplays with other identities, such as that of a teacher.
The second theme that resonates with all the respondents to our study is a Discourse of gender that illustrates the strong role of socio-political ideology in shaping identities in ways that reflect idealized images of women as mothers and housewives (Nagatomo 2012).This is explored in detail in a segment from an interview with Respondent Y: I take care of the daily chores all by myself.Because my husband is an intellectual who needs much quiet time, it means that his thoughts cannot be easily interrupted.Hence, I have to do more housework.I think men should do great things, and women should get small things done.This is my principle of doing things [ . . .] At noon I go back home to cook, and then in the morning I get up at 6:00 a.m. to prepare for breakfast.For example, there are seven kinds of soup and a variety of dishes for breakfast in one week because I think breakfast is the most important.Then lunch and dinner.
There are a number of features to this narrative.First, use of a discursive strategy referred to as 'category entitlement' (Wiggins 2017) functions to produce categories to which women and men are bound.In this segment sex categories determined who might or might not be expected to engage in household chores.However, the segment is not simply about housework, rather, it accords with the 'virtuous wife and caring mum' (Xian qi liang mu) Discourse promoted through socio-political ideology in China, and it provides insight into gender hierarchies and identity.The husband is characterized as an 'intellectual' and his intellectual needs take priority over those of the woman who is portrayed as passive and lacking in agency.Specifically, the term 'I have' (to do more housework) indicates she has no choice in this matter.However, she resolves this with an assessment that 'men should do great things' and 'women should get small things done'.The detailed account of daily chores that follows discursively functions to support the speaker's entitlement to tell the account and reinforce her identity as accurate, since if a person claims an identity they must effectively do that identity to be believable (Wiggins 2017).Throughout the segment a male/female dichotomy is reinforced in line with broader Discourses of gender.Another example of traditional ideological Discourse pertaining to women's role in the private sphere of life is provided by Respondent D, a Dean, who explained: 'personally, I didn't want to do this job, because as a woman, family is the most important thing.'She reinforces her statement concerning the importance of family in a segment about domestic duties: I do all the housework, my husband was born in Gansu province, so he does not have the consciousness of doing housework, but if I arrange some work for him he will do it anyway.Generally, in my view, household is the most important thing for a woman.Happiness of the family is most important of all (Respondent D).
One consequence of this discourse is that women may be perceived as less committed in professional spheres of life.Indeed, some respondents subscribed to this view, as illustrated by Respondent W: 'If the female is tethered by her family role, she would definitely cut down the time and efforts for her career.However, the male faces a different situation.' The third theme that resonates with all respondents is that they discursively distance themselves from leadership as a professional identity.Having explored this phenomena in more depth, we identified two key strategies used by respondents, which we refer to as 'downgrading leadership' and 'rejecting leadership'.For instance, although several of the respondents hold formal positions that involve leadership (e.g., giving direction, decision-making, social influence, etc.), and/or management (e.g., managing courses, staff, budgets, resources etc.), they frequently downgrade or reject that identity.The discursive strategies used to downgrade or reject leadership appear interchangeably and intermittently throughout the interviews.In what follows we illustrate this in segments from an interview with Respondent F, a Department Director.
Respondent F worked in a corporate role for ten years where she rose to secretary of the board.After gaining her Master's degree she decided to become a university teacher.That was over ten years ago, and she was promoted to her current position as a Director of a Department a year ago.She describes the promotion as 'not that complicated' [ . . .] there was a vacancy, so I became director'.The promotion, therefore, is described in oblique terms and its significance downplayed.By contrast, this is followed by an assessment offered not by the respondent herself, but by her family: 'they think that my present job is exceptionally great.' Extreme words such as 'exceptionally' discursively function to emphasize a point; in this instance the point is: 'my present job is exceptionally great'.However, she attributes this statement to her family, and in doing so evades taking ownership of the statement herself.Female modesty is congruent with idealized images of women in China and it has been documented elsewhere as an important factor in women's reluctance to self-promote (Rudman 1998).Interestingly, however, she is the only respondent to explicitly refer to herself as 'being a leader' but, thereafter, she denies herself this identity.In the following segment, she describes how she views her current role: As the director of the department of Human Resources Management, my responsibility is to do the tasks which are assigned by the superior, to arrange the courses for students, and to coordinate and manage the teachers' arrangements.That is not exactly management, more like coordination.Plus, I'm responsible for making and amending the syllabus.So far, I spend 20% of my energy on managing the whole department.
There are a number of features to this narrative.First, the respondent's professional status as a Director is reaffirmed, but she positions herself as lacking in agency in this role.Tasks are 'assigned', and she is passive in that process in relation to her 'superior'.The classification of tasks involved in Directing a Department is further downplayed in the assessment that follows.The term 'not exactly' (management) discursively functions to minimize the significance of this role.Following this the term 'management' is further downgraded to 'coordination'.Outright rejection of the leader identity follows again in response to a question about her personal experiences: Actually I don't need to answer this question, because I never treat myself as a leader.There is no such social status for me.I am a teacher, and only a teacher . . .all the women or female leaders I met, all of them, including me, have such feelings.I think it is normal.
In common with other respondents, she discursively constructs the teacher identity in relation to other social identities, but once more rejects the leader identity: I still prefer to be addressed as teacher or my name.Actually, we work for people, instead of being a leader . . .it's important to have a good attitude, and put you into others' shoes.Do not regard yourself as leader and [think] everyone must follow your leadership . . .There is no such social status for me.I am a teacher and only a teacher (Respondent F, Department Director).
Following the first assessment 'I still prefer to be addressed as a teacher', the evaluation ('I am a teacher and only a teacher') serves to strengthen the argument and preference structure for the teacher identity, which is upgraded above that of leader.Lists and contrasts serve to add rhetorical strength to an argument and can be used to manage accountability and identities.Furthermore, discursive devices such as minimization-'there is no such social status for me' function to downplay the significance or importance of something, in this instance the respondent's identity as a leader, which she rejects as 'there is no such social status for me' (Wiggins 2017).Shifting between author and narrator of this complicated discourse regarding various subject positions, she adds further rhetorical strength to the argument, by accounting for this assessment in oppositional terms-the importance of having 'a good attitude' and of putting 'you into others' shoes', is depicted as contrasting with the alternative unappealing reality that 'everyone must follow your leadership' which can be located within normative heroic leadership Discourse.Claiming dislike of this style of leadership, she claims a subject position that is more appealing in the traditional cultural context of China-'I am a teacher and only a teacher'.Alvesson and Billing (Alvesson and Billing 2009) suggest that by emphasizing the relative importance of one identity compared to another, an individual may not view themselves as a fully legitimate member of the social group to which they dis-identify (Alvesson and Billing 2009).In this instance, the teacher identity is given prominence and the leader identity is clearly rejected.O'Connor (2014), suggests that universities, with their homosocial male dominated organizational cultures, promote the view that leadership is a masculine pursuit, since not only is the language and voices of leadership male, but the working practices, networks and promotion routes are predisposed to benefit men.These processes can make it difficult for women to claim legitimacy in leadership roles.Women internalize negativity, discrimination, exclusion and othering as feelings of fraudulence.Added to this, women have fewer role models of the same gender to identity with (Singh et al. 2002), and this can lead to them feeling like an outsider in the organization (Bagilhole and White 2013).
We have presented a segment to illustrate through fine grained analyses the common theme of downgrading or rejecting a leadership identity.In what follows we shift attention from respondents' descriptions of their own identities to explore how they talk about male and female leaders more generally.

How Do They Talk about the Female Leader Identity More Generally?
Respondents' descriptions of leaders (Table 2), drawn from across the interviews, reflect gender stereotypes-normative essentialist definitions of gender and masculine and feminine leadership.The dichotomy between male and females is one dimensional-men are described as unemotional whereas women are the direct opposite-emotional.According to this stereotype, men are rational and women sensitive, and so forth.Evidently, this dichotomy leaves no room for a man to display emotions or for a woman to be rational.Naturally, this stereotypical view can constrain men, as well as women, to do gender (West and Zimmerman 1987) and enact leadership in a way that is congruent with gender stereotypes.In each of the interviews numerous accounts of male and female leaders highlight the influence of dominant Discourses of leadership and gender.The discourse (little-d) is more complex since it reveals many contradictions and tensions.For instance, although the respondents unanimously claim that 'men and women are equal' (Respondent D), a key theme in the interviews is that their descriptions invoke difference and inferiority in relation to female leaders.This discourse reflects dominant historical and cultural Discourses concerning gender and leadership that prevail in China, and elsewhere.In the majority of cases, difference is attributed inherent traits or personality characteristics.For example, Respondent R, said: 'the different genes result in male's rationality and female's sensibility, which no one can change'.Although, generally, there is consensus among respondents that women can be successful leaders, especially with 'training' and 'experience', this came with the proviso that: When a female takes a role in management, it's important to control her emotion well.From my work experience in past few years, I don't think females can often manage it well.If a female can do it [manage her emotions] better, she can be an excellent manager (Respondent F).
A second theme is that of invoking difference in a way that highlights women's relative advantage over male counterparts.For example, Respondent X said: Males are much more vigorous, resolute, and decisive.Because they consider things from a different angle, they focus on the overall situation and solving problems more quickly.Females are meticulous and patient, and they focus on details.They solve the problem with systematic guidance and they are very good at dealing with everything in a soft way.
Another similar example, from Respondent S is that men are 'rigorous, decisive, and bold at work' and women are 'more gentle, tender, and more considerate'.While this discourse emphasizes feminine styles of leadership (Ford 2006;Fletcher 2004) and difference, women are not inferior.
Many other examples of this permeate the transcripts.In each instance, female leaders are typically considered 'more delicate and more considerate' (Respondent Y).In addition, 'Female leaders have their own methods of dealing with work and might in some aspects work better than male leaders' (Respondent L).Overall, by emphasizing Discourses of gender that promote the essentialist view of women as possessing unique feminine qualities (Eagly and Carli 2003;Helgesen 2008), respondents emphasized that: 'women have their own advantages.Being sensitive and being rational have their respective advantages' (Respondent O).Indeed, many believed a female leader 'should exert her leadership charisma through some female traits' (Respondent W).
These discursive strategies function to highlight women's unique qualities in leadership, but still men are believed to have a relative advantage: I worked over twenty years, and met more male leaders.On the one hand, male leaders are objective [ . . .] There are some advantages for men to be leaders.Firstly, he has more time and energy to do work; secondly, he is rational and reasonable instead of easily being influenced by emotions.Thirdly, [ . . .] it is easier for men to make orders.In our university, there are more female teachers, usually speaking, they are more likely to follow the male leaders' instructions.It is true, though I don't know the exact reasons.This discourse about women leaders, although not specifically relating to themselves, provide insights into self-understandings of how to be a woman (Alvesson and Billing 2009).Indeed, some respondents provided examples that reflect a Discourse of women's difference, and in some instances inferiority, in relation to themselves.For instance, Respondent D revealed: 'I don't have the courage and willpower like a man, as a woman I am more emotional and soft hearted.' Providing a different insight an excerpt from Respondent R reveals how gender and leadership Discourse influences not only her views, but also her actions: Even in my student management work, I prefer to assign a job to a boy student instead of girls.Our college has more than 30% boy students and more than 60% girls.Girls are obviously in the majority, but I still would rather assign a job to a boy student.The reason is that I think boy students can do a more satisfying job than girls.
The comparisons between men and women leaders (and boys/girls) in these segments take a grammatical form that involves setting up a hierarchy of who 'does' difference, when they 'do' difference and what 'doing' difference means.Once more, not only does this theme in the data highlight the negative aspect of being a woman, it reinforces the male/female dichotomy.On the one hand, the discourse brings out the so-called caring side of women while, on the other hand, it maintains a deficit version of women; that is, women are somehow deficient in relation to male norms of leadership.

Discussion and Conclusions
This article has explored how female academics in higher education in China construct their (gendered) identities.We considered the multidimensionality of identities important to our analyses, as many of the women in our study simultaneously performed roles as teachers, managers, or leaders while also being wives and mothers.In addition, we considered the influence of powerful historical and cultural Discourses.
First and foremost, the findings show that the women describe a multiplicity of identities drawn from private and social realms of life.Notably the findings highlight the interplay and tensions between competing multiple identities, and reveal how those identities are shaped and constrained by dominant Discourses concerning gender in Chinese society.Notably this results in identity regulation (Alvesson and Billing 2009, p. 99), notably identity positioning that is congruent with social norms and conventions (Lindgren and Packendorff 2008).A key theme is that the women do not identify as leaders, despite undertaking roles that involve leadership.This is an important finding since the rejection of an identity is an indication of not belonging to a particular social group (Alvesson and Billing 2009, p. 98).This may be due to the leader identity breaking too closely with traditional ideas of femininities, which can be problematic for women (Billing 2011).The extent to which we view identity as free choice or as an unjust social outcome is debatable.On the one hand, personal identity is not fixed once and for all (Alvesson et al. 2008;Alvesson 2010;Stewart and McDermott 2004), and there is scope for resistance and change within higher education in China.However, it has long been established that social expectations shape the doing of gender (West and Zimmerman 1987) in ways that are congruent with social and cultural expectations (Eagly and Karau 2002), and those women that deviate from social expectations are likely to experience repercussions (Brescoll 2011;Rudman and Phelan 2008;Rudman et al. 2012).Fundamentally, even though identity should not be reduced to simply to gender variables, 'the male norm indicates that gender identity is a major identity signifier' (Billing 2011, p. 300).This means the process of doing gender and doing leadership does not simply involve individual micro-level identity work; rather, identities are shaped by a range of factors at meso (organizational) and macro (societal) levels (Billing 2011, p. 303), since constructions of 'femininities and masculinities permeate social life, and guide and constrain people's behavior' (Alvesson and Billing 2009, p. 18).On the point of meso-level factors, it has been said that the academy is a 'source of oppression and a location for exploring liberation and empowerment' (Morley 1994, p. 202).Our study shows that the Chinese higher education context continues to be marked by gendered norms.This has implications for the career advancement of female academics and consequences for reversing the underrepresentation of women in senior leadership in universities in China.However, this situation should not be viewed as irreversible.Senior leaders in higher education can potentially shape the internal culture of their organizations and influence the wider higher education context (O'Connor 2014).
A second key contribution of this article is that the data provides unique insights into the discursive construction of male and female leaders more generally in the higher education context in China, specifically illustrating how culture and language both construct and reflect gendered meanings (Alvesson and Billing 2009).Most cultures have 'systems of meanings and norms prescribing different activities and characteristics for women and men (Alvesson and Billing 2009, p. 50).It is of note that the findings highlight the constructs of feminine and masculine are positioned as opposites in the descriptions of female and male leaders, and the masculine is generally privileged (Alvesson and Billing 2009), while the feminine is given a lesser or deficit status.Traditional Discourses of gender and leadership that permeate Chinese culture were found to be strongly reflected in the discourse invoked as part of the respondents' descriptions of female leaders as sensitive and emotional and male leaders as rational, decisive, unemotional, and so forth.In terms of practical implications, the findings suggest that initiatives that seek to address the underrepresentation of women in leadership are likely to be undermined in the absence of understanding how career opportunities are shaped by discursive practices (Tannen 1995;Sturges 1999;Lämsä and Sintonen 2001) and dominant Discourses (Ford 2006).
While this article contributes to understanding key issues surrounding women and leadership in China's universities, which may be relevant in other cultural contexts, we acknowledge there are limitations to this study.First, we interviewed nine women from two universities.A larger sample from a broader range of higher education contexts could reveal a diversity of views regarding male and female leaders.Second, the interviews were conducted at a single point in time and do not capture the temporal nature of identity work.A further limitation of our study is that only four of the women were in formal leadership positions, although the remaining five respondents had some experience of leadership through their work with the Youth League Committee.Future research could involve longitudinal research, perhaps following newly appointed female academic leaders.Comparative studies with other national contexts could also prove fruitful.Lastly, although our study concerned women in higher education in China, future studies could explore men and masculinities to understand gendered structures and open up opportunities to challenge the masculine norm.
In conclusion, an individual acquires and learns various identities, by way of their sense of belonging or membership to social categories and groups (Alvesson and Billing 2009, p. 98).Although the women in this study reject or downgrade their identity as a leader, they could be viewed as mediating larger societal Discourses of gender and leadership, since they have learnt how to position themselves within those Discourses in ways that are congruent with social and cultural norms and expectations.From this perspective, it could be argued these women are forging new ways of doing leadership and doing identity work.While they may appear to be complicit in perpetuating dominant Discourses, they have found ways to navigate those Discourses.As such the circumstances through which they have forged an academic career may be viewed as emancipatory and part of the change process, at least as a first step on a longer journey.

Table 1 .
Profile of research participants.

Table 2 .
Words used by respondents to describe male and female leaders.

Table 2 .
Cont.These discourses of gender stereotypes significantly emphasize traditional societal Discourses of gender and leadership that permeate Chinese culture.We explore how this discourse (little-d) and Discourse (big-D) is intertwined in greater detail in what follows.4.3.In What Ways Is This Talk Intertwined with Larger Societal Discourses of Leadership and Gender?