The Symbolic Order of Gender in Academic Workplace Ways ofReproducing Gender in Equality within the Discourse of Equality1

H is we l l known that universities are male dominated both in history as well as i n the dominant discourses. As knowledge producing organizations universities also carries the heritage of defending the scientific ethos of meritocracy and objectivity, these arc rules that many researchers still are trained to believe in. This makes often studying of gender inequality in academia a difficult task since it not only reveals the gendered structures of academia but also violates the norms of science i mplying that science is socially biased. This article explores how gender inequality is produced within the discourse of equality at Swedish universities. The underlying assumption is that gender inequality on the level of the academic departments i s produced within the broader discourses of gender, power, science and equality operating in everyday academic working lives and in society in general.


Introduction and Earlier Works
In 1995 the UN declared Sweden the best country in the world for women to live in according to a Gender and Development Index (GDI) (UNDP:1995). As a small Nordic welfare state Sweden offers a quite privileged setting when it comes to gender distribution in higher education. The total gender distribution within higher education was fo rty-eight percent women and fifty two percent men in 2000. However the gender distribution differs between different staff categories. Women represent a majority of the staff involved in administration and teaching. For positions that are more extensively designed fo r research such as research assistants and fu ll professors, women are still in the minority. Thirty eight percent of the research assistant positions and thirteen percent of the fu ll professorships are held by women (National Agency fo r Higher Education yearbook 2001). Thus, we can see that within the domain of higher education Sweden fo llows the inter national pattern of "The leaking pipeline", the higher up in the academic hierarchy the fe wer the women and the more the men. This has led to a wide range of efforts aimed at diminishing gender differences in academic career achievement. In some cases it has resulted in different equality positions such as, equal opportunities officers and equal opportunity committees. Now a days, fo rmal ways of excluding or diminishing women in academia are no longer accepted and it is highly unlikely that someone would openly declare that women are inferior to men when it comes to pursuing a research career. However, research has shown that gender differences in scientific careers still remain despite the introduction of equal opportunity policies. Gender marked inequalities are documented in the distribution of scholarships as well as in gender difference in career achievement (Winnifred and Hamilton 1988, Stolte Heiskanen 1991, Wold and Wenneras 81 1997. In order to increase our insight as to why gender differences in academic careers� pers i st, research about how the genderorder1 is produced and re produ ced in everyday academic working life, needs to be elaborated.
This article explores the production and reprod uction of the symbol ic order of gender at ditlcrent levels in the academic workplace. The underlying assumption is that gender relations exist within the broader discourses of gender, power, science and equality operating in the academic department and at a general level of society. The aim ofth is article is to explore "How gen der in equality on the level of the academic department is reproduced within the discourse ofcquality operating at a general level of society".
The iield of study is two academic departments in the same area of biology at two different Swedish universities.
Previous research about gender differences in scientific careers has produced numerous explanations. Ve ry briefly these can be categorized into three main groups; "gender differences in publication productivity", "impact of fa mily situation on scientific careers" and "the social organization of science, The Old Boys Network" (Fox and Faver 1985, Luukonen Gronow 1987, Davis and Astin 1990, Cole and Zuckerman 1991, Kyvik 1991, Long 1993, Sonnert and Holton 1995. The results have often been contradictory and cover a wide range of methodological and theoretical approaches. However, previous research has not gone without criticism and three main lines have emerged over the years. Firstly, the need fo r studying gender differences in scientific careers over a longer time perspective has been expressed, since women appear to have different career trajectories with more "winding tracks" than men (Elgquist-Saltzman 1994). Secondly, there is a risk that fo cusing on gender differences when studying scientific careers leads to an essentialist trap where differences are em phasized and reproduced. Thus, it is important to fo cus on the construction of fe mininity and masculinity and by which means differences are constructed, legitimized and reproduced. Thirdly, since a scientific career most probably depends on a cumulative effect of a multitude of factors interacting over time, further studies of 82 the complex interactions between and within individuals in their daily research milieu are called for. In the fo llowing discussion the theoretical fr amework of this study is fl.lrther outlined .

Theorizing Gender
Gender is often described as the social dimension of biological sex; this is particularly evident in the famous statement "you are not born a woman, rather you become one" (de Beauvoir 1949). Since these words were first written the social construction school has developed different theoretical lines and gained acceptance both within and outside of feminist readings'1• The fo llowing offers one definition of gender: "Gender, refers to patterned, socially produced, distinctions between fe male and male, fe minine and masculine. Gender is not something people are, in some inherent sense, although we may consciously think of ourselves in this way. Rather, fo r the individual and the collective it is a daily accomplishment (West & Zimmerman 1987) that occurs in the course of participation in work organizations as well as in many other locations and relations" (Acker 1992: 250).
In this article the social construction perspective also includes the body and physical appearance as well as sexuality, which are all part of the ongoing production of gender. The notion of power is central when conceptualizing gender since patterned differences between women and men, fe mininity and masculinity usually involve various expressions of the subordination of women (Acker 1992). Thus, it is important to remember that these power relations are fu rther complicated when fa ctors such as race and class are intertwined with gender. The gender order can be challenged if the rigidity of the male/ fe male dualism is challenged and the notion of difference is nuanced and contextualised (Gherardi 1995: 101-103).
Within the social construction school some have fo cused on the relational aspect of gender; this interpretation emphasizes the mutual inter relational construction of fe mininity and masculinity as well as the importance of contextual and processu al a::;pect:-; on the con struction of ge nder (Gherardi 199fi, Davies 1996). Th e relational uspect of gender focuses on ihe "doing of' gender" and consequ ently on the meanings th at spring from the "doing of ge nder", a::; well a::; the condition::; and contexts surrounding thi::; proce::;::;. Thi::; raise::; the ques tion of how gender i::; represented in our daily live::; or, more precisely, how we give meaning to gender through l anguage, action and symbols. Th is has been elaborated in th eories of representation which, very briefly, can be said to deal with ihe processes by which subj ects of a specifi c culture and historical context use language, or any signifying system, to produce meaning (Hall 1997: 61 ).

Representing Gender
Theories of representation include a wide range of approaches, from semiotics to discourser.. In this article the broader concept of discourse is used rather th en a more narrow use oflanguage that a linguistic approach would require. In this article discourse not only includes what one says (language) but also what one does (practice).
In this sense discourses not only "defines" how we can talk about certain topics but also influences how ideas are put into practice and used to regulate the conduct of others (Hall 1997: 44). Foucault was one of the first to introduce the notion of discourse instead of language in the production of meaning. Stuart Hall defines discourse as "[ ... ] a group of statements which provide a language for talking about -a way of representing knowledge about -a particular topic at a particular historical moment" (Hall 1992: 291 in Hall 1997.
According to Foucault our perception of what constitutes the "truth" in certain historical moments and contexts contribute to the maintenance and internalization of dominant discourses in our everyday lives. This is what Foucault calls regime of truth. The regime of truth is constituted by a discursive fo rmation6 that legitimizes what is perceived as true or false, the means fo r doing so and the status of persons who arc in charge of thi s activi ty.
Foucault's puts it like this: "Each Society has it::; regime oftruth, it::; 'general politics' of truth, that is, the types of discourse::; which it accepts and makes fu nction as true, the mechanisms and instances which enable one to distinguish true and false statements, the mean::; by which each is sanctioned ... the statu::; of those who are in charge with saying what count::; as true" (Foucault 1980: 131).
Applied to gender studies this theoretical approach implies that a regime of truth, that i:-; sustained by discursive fo rmations and con sequently are relative to historical, situational and contextual aspects, makes it possible to internalized dominant discourses about gender. This influences how agents act and give meaning to gender in their everyday lives, in other words, how they discursively produce and reproduce gender in their everyday lives.

The Dual Presence
In general the history of science has been male dominated in numbers as well as in the dominant discourses (Keller 1985). This is also reflected in the symbolic order of gender. One set of qualities such as reason and public presentation, qualities associated with science as well as with activities in the public sphere in general have been associated with masculinity. Qualities associated with emotions and private activities have been associated with fe mininity, re production and the private sphere. In this perspective women will always be "lacking" important qualities when entering the pro fe ssional scene. Thus, women entering the pro fe ssional arena are still symbolically connected to qualities associated with the domestic sphere, and the gender order fr om the private sphere has been transferred to the public sphere (Marshall and Wetherell 1989, Wager 1994, Katila and Merilainen 1999. This phenomenon is often described as the "dual presence" of women which indicates a cross gender experience, more specifically the simultaneously presence of the private and the public, home and work, personal and political (Balbo 1979;in 83 Gherardi J 994 : 598). Managing dual presence rcf1uires dif'fercnt di::;cursive strategies.
Gherardi has identified two diflerent types of' strategies that arc u::;ed fi.>r handling "dual pre::;encc"; she call::; th em ceremonial and remedial work. In ceremonial work, diflerences between sexes arc recognized and celebrated which can be done in a number of ways, for exarn pic th rough gesture::;, tone::; or language. It is very diflicult to avoid ceremonial work since gender is one ofthe major social categorizations that we usc in our everyday life. It is also deeply imbedded in what we call "good manners"; to avoid celebrating gender is often seen as odd and sometimes rude behavior. Celebrating gender can also involve a sense of pleasure. When interacting with other people celebrating and responding to gender can create a sense of belonging to the "bigger" bodies, or with Foucault's words, to the "discursive fo rmations" of the fe minine and the masculine.
When the dual presence occurs there is a break in the gender order and this requires other rituals. This is where the remedial work enters the scene. Remedial work is "sim ultaneously supportive of the symbolic order of gender and remedial of the offence" (Gherardi 1994: 602). When women enter public organiz ational life they break the symbolic order of gender. Through remedial work women can enter public life and still celebrate conventional fe mininity. This can be done by working in fe male dominated areas or by adjusting gestures and language. Gherardi gives the fo llowing example of remedial work: "When women take the conversational initiative and apologize fo r doing so, when she expresses her doubts as to the importance of what she is about to say, when she minimizes her competence to speak on the subject -that is, when she requests authorization, protection and bene volence" (1994: 605).

Data and Method
The empirical data comes from two departments at two different universities specializing in the same area ofbiology. Biology was chosen because several studies indicates that women at biology 84 departments have passed a threshold that attenuate gender stratification (Long 199:.3, Sonnert and Holton 1995). Twelve interview::; of two types were conducted. The questions were loosely structured, with interviews lasting between 90 and 120 minutes. When needed, a second interview was arranged. All interviews were recorded, with the consent of the inter viewed person, and all have been transcribed . The interview quotes in this article have been translated into English by the author.
The first type of interviews was conducted with senior researchers who were asked questions about the research organization , policymaking, dissemination of information etc. My intention was to get an overview of the organization of the department as well as a glimpse at its specific culture. The second type of interview was conducted with "new re searchers" that rather recently had finished their PhD. The original ambition was to interview researchers that were within the first years of receiving their PhD degree. However, finding researchers matching these criteria was not easy and in some cases the time period had to be expanded. The "new researchers" were asked questions about personal background, ambitions and fu ture plans, fa culty advisor relationships, access to information, collaboration and support and other questions concerning how they perceived their everyday working life.
When working with a qualitative method and the material that comes along with it a researcher is often asked questions about he or she can be certain they are getting the "true story". Naturally, it is extremely difficult to answer such a question; it would require a more psychological approach and even then it would remain highly problematic. Rather than dwelling on the issue I have preferred to see it fr om a different perspective; people create different stories when they answer the questions and this is a way of affirming their identity/ies. I have acknowledged the diversity, ambiguity and fr agmented dimensions of the stories because they represent their lived experiences and are true fo r them. My interest lies in the fo rm of these stories, more precisely, how the interviewed researchers present the stories, rather than trying to explore which story is the "true" one7 . It is also importa n t to recogni ze th e power dim •nsion of' my po:.; ition as a researcher and that the interview situation is a rather extreme ::;ituation for both the re::;earcher interv i ew i n g and the one bei ng interviewed. However, the u nderlying assumption is that know l ed ge i::; ::;itu atedH a nd that identities are multiple and conti n uing constructs, which impl ie::; th at th ere is no "inner core" that represents a true version of what a person really th ink or believes.

Results and Analysis
Situating the Research ers9 At a fi rst gl ance it is very easy to find similarities between the two departments; they belong to the same academic discipline so it is natural to expect that they share some structural con straints and possibilities. The research field is characterized by a quite good financial situation, at least compared to the humanities and the social sciences. Also, it does not belong to the pure applied technological and medical research areas where the fu nding situation must be considered to be better. Findings in the field have short life span and research development and research innovations become old news very fa st. This most likely goes hand in hand with the use of/dependence on advanced technological equipment, something that is increasing. It is also reasonable to believe that the departments share the same research organization since most natural sciences are organized in research teams due to practical and economic reasons.
Peters and Va nraan's (1991) study fo und that the internal co-authorship networks of a chemical engineering department centered on a fe w productive scientists and fo rmed clusters. At our departments the researchers and PhD students were organized in teams based on research orientation. A "typical" research team consisted of a PhD research leader, sometimes a post-doc, and (at least) one or two doctoral students. In some cases the group also had their own laboratory assistant, which depended on their financial situation. However, a closer look at the departments revealed that they were differently structured by gender, academic position and research areas.
In Department A I found that four research groups were active. The department had one elderly male professor who was the research leader of one group, which did not appear to influence the power balance between the re search groups. Since academia is a hierarchical institution it is natural to expect certain hier archies such as those between PhD students and PhDs, however hierarchy between the research group leaders and the professor was not stressed in the interviews. The research groups appeared to be equally influential; no research orientation (or group) appeared to be more central or peripheral, or have more or less status than another. Below Howard, a senior researcher describes the seminar distribution and guest research distribution between the different research groups/ areas.
"We try to divide it up so that everyone has the same number. We are all very different but we try to divide it so that there is something for everyone. I mean, it would never happen that there would be a seminar series in my area only, that would be pointless. [ ... ] Even if we're all different I'd have to say that we complement each other in a way." Bill, a new researcher describes the department the fo llowing way: "( ... ] There's a good atmosphere between the senior researchers, they don't compete fo r the doctoral students, there's an alternating system. [ ... ] So there's no fr iction between them like there might be at other places. They've built up a pretty simple system fo r who gets what and then if there are any objections, it seems to work anyway. So the atmosphere is relatively, it's a pretty good group of people." Collaboration between the research groups did occur although the groups had their own distinct research areas and consequently the collabor ation mostly concerned methodological issues. Sex distribution appeared to be equal at both doctoral student level as well as research leader level; two of the fo ur research groups were headed by women.
In Department B research was similarly 85 organi:r.ed in resea rch groups. However, the depa rtment waR m uch more malo dominated both at doctoral student level and at research leader level. The department had two malo profes::;ors with ::;trong position::;, both as social and cognitive leader::;. The fl 1rmal structures for deci::;ion-makingwere well establ i shed but there also ::;ecmcd to be infc1rmal structures in th ose proce::;::;e::;. Edward, a new re::;earcher describes the department th e fi.Jllowing way: In this quotation we can see that although the department has become a bit more decentralized since the second professor has become the head of the department he still has a strong position.
Edward also highlights that decision-making often takes place in informal groups outside the fo rmal structures of the department. Status differences between research groups and research orientations were clearly expressed when describing the research activity at the department. The closer to the professor in terms of social and cognitive positions the higher status the researchers had. The department had six research groups but their research orientations were slightly more homogenous 86 th an at Department A. Only one research group appeared to have less collaboration with the others. That group also appeared to have l ess status than the others. The female re:,;carch leader of that group also reported th at she lacked collaboration partners at the department, fClt excluded fr om th e inner circles of the departments and had difficulties getting access to informal information although she was a senior researcher and member ofthe departmen t board.
Researchers at both departments were very keen on stressing their enthusiasm towards gender equality in academia. They were aware of the higher level of dropouts of female academics and offered several explanations for this They had also developed equality plans for achieving equal sex distribution at all levels of the departments. Formal ways of excluding or diminishing women in higher education are no longer accepted and the introduction of equality positions has among other things assured the discourses of equality between women and men in academia. Presently, it is unlikely that somebody would openly declare women as less suited than men for pursuing an academic career.
This does not mean that gender inequality does not exist in academia. The diminishing of fe male academics is expressed in more subtle ways that are imbedded in everyday working situa tions. This has led to a debate about the possibilities and practices of the rhetoric of equality. This will not be fu rther developed here. The main point is instead; how is inequality reproduced within a discourse of equality? Researchers were asked by me to describe what they thought the reasons fo r gender differences in career achievement in academia were and more particularly within the field of biology.

Turning Positive Stereotypes into Grounds fo r Exclusion
A male researcher, lets call him Andy, who is still active within academia doing research in Department A, which appeared to be less male dominated, less centralized and less hierarchical than department B. Andy is married and has no children. His wife is not a researcher. In the fo llowing quotation Andy a n s wers to th e fo llow ing question: "Do you think it is easier for men than f(>r women to pursue an academic career after receiving the PhD degree'?" "Within this area o fbiology we have fa irly equal nu mber:; of women an d men taking under grad u a te cou rses as well as at the doctoral student level. At the mom ent and at this department that's also the case at research assistant level. It':; is not statistically proven bu t l gue::;s that the ::; i tu ati on is fairly equal. They I women : my remark I have not been able to make it all the way to the top but it takes time. I don't think I there is a ditlerence: my remarkl .
One di fferencc is that women are away on parental leave longer, so the reasons should be looked for at home rather than at work. No, I can't sec that women have been discriminated against in academia, at least not here, I couldn't speak for other countries, but I haven't seen anything, at least not here. The difference is above all at home. Maybe women are not so manic as men are. [ . .. ] They're not prepared to sacrifice as much. I mean, very few have "normal" working days and it's very hard to advance if you have small children. That's just the way it is. I wouldn't be able to work as much if I had children. l ... ] Yo u could put it like this, you have to have the brains but there's nothing that can replace hard work. That's the way it is, so there's a tendency towards people who work a lot." Time plays a significant role in explaining gender differences in academic careers in this quote and is referred to at two levels. Firstly, time is fo und on a structural level: "They [[women] ] have not been able to make it all the way to the top but it takes time." This reasoning suggests that gender differences in scientific careers persist due to old values and perspectives, old ways of doing gender. Thus, it assumes that present ways of doing gender are unproblematic and leaves them intact. The stressing oftime as an explanatory fa ctor for gender inequality appears in other studies concerning gender equality in political representation as well (Tollin 1998:30). Secondly, it is present on a micro level, as a part ofthe labor division within fa milies: "One difference is that women are away on parental leave longer, so the reasons should be looked f(>r at home rather than ut work". This puts the focus on processes outside of academia; however, it also tends to indi vidualize gender differences in scientific careers. The fo cus is on situations within fa milies rather than on structural constrains in the system of' higher education. Although the concept of' "parental leave" is used, it soon becomes obvious that what is implied is in fact motherhood, "women are a way on parental leave longer". 'L'o refer to motherhood is to refer to women's reproductive capabilities and thus to stress the biological differences between female and male researchers. To stress gender differences as binary oppositions is a common way of producing and reproducing the gender order (Hirdman 1988). The ties between parents and children carry strong connotations regarding what is perceived as "natural"; however there are also has "positive" connotations. This legitimizes the differences in career achievement between researchers that have children and researchers without children but has an even stronger legitimizing effect on gender differences in scientific careers since motherhood is a major way of constructing fe mininity. Gender diffe rences in career achievement is thus, deferred to "natural" and "positive" processes of mother hood and leaves the structures of the academic career system intact and seen as unproblematic.
Later in the interview a second explanation is introduced: "maybe women are not as manic as men are". To be manic is to be in a negative mental condition or have a mental disease. At first glance this may appear flattering for fe male researchers: they are not as "manic" (sick) as men are. However, this quote also reveals that succeeding in academia is hard, in fa ct so hard that in order to succeed you must be able to break the "natural" and "positive" ties of parenthood. "Manic" is constructed negatively, breaking the ties of parenthood, but at the same time positively, being able to work hard and being ambitious, attributes that are normally used to describe good researchers. "Manic" appears to have been transformed and is now used as a positive indicator for being successful in academia. In the quote above we could see that "manic" was associated with men. However, 87 "manic" ab;o include:.; the separation between the private nnd the professional: "I wouldn't be able to work as much i fl had ch ildren". Manage ment studies have Hhown thai separating the private fro m the public and, th us, constructing women as a negation of the norm is a major obstacle fi>r women's cfloris to reach top positions within organizations (Kanter 1.977, Cockburn 1991, Wahl et al. 1998). This reasoning rests on the dualistic rel ation ship between femininity and masculinity, the private and the public, the irrational and the rational. In addition it can be said thai while bei ng a successful researcher reinforces male gender identity it implies a contraction fo r fe male researchers' gender identity.
The concept of"manic" appears to be central for the understanding of how Andy makes sense ofgender diflcrences in scientific careers. Let us therefore take a look at how he uses the concept in other situations. In the following quote Andy was asked ifhe had any intellectual role models in science. In the first quote we could see how "manic" was transformed to mean something positive, particularly in relation to women who were not as "manic" as men. In this latter quote the researcher is comparing himself to the scientific elite, and in relation to them "manic" is constructed negatively: "I wouldn't like to change places with them, because they really are totally manic about what they do". In addition, we can se how Andy uses the concept of manic to construct himself positively in relation to both groups; he paints a picture of himself as professionally and intellectually superior to women and emotionally superior compared to 88 the scientific elite. However, a second glance at th is reasoning suggests criticism against the academic career system. "Manic" is constructed negatively as is the scientific elite, which implieH that the researcher perceives the academic system as hostile towards those who can not put up with the harsh working conditions, parti cularly women and women and men with children.
From the quotations we can see that Andy is ambivalent towards gender in academic work places. In a daily working environment where fe males are present as high productive research leaders having both fa milies, children and an academic career, the symbolic order of gender is not easily legitimized and thus, requires other rituals fo r making sense of everyday working life. The construction of a researcher as someone that is totally devoted to academia and has no fa mily obligations can bee understood as one way of reproducing the gender order and making sense of gender in academic workplaces.
The interviewed researchers in this study most certainly perceive themselves, as "pro gender equality" and thus, their aim is to speak within the discourse of gender equality. As pointed out earlier it is unlikely that someone would openly declare women as less suited than men fo r pursuing an academic career. However, the changed discourse of "manic" from negative to a positive indicator fo r a successful researcher, and the construction ofwomen as "lacking" has done this in a more subtle way.
The quotes above were chosen because they eloquently provided answer to my research questions, however other women and men fr om both departments expressed similar opinions. The quotations below give fu rther insights into how the researchers understand gender diffe r ences in scientific careers. The fo llowing quote comes fo rm Andy's supervisor Howard who is a senior researcher at Department A. Below he gives his explanation as to why there are fe wer women pursuing an academic career.
"Of course it may in part be because women are less fo rward than men, even if they have the same qualifications but I also think that it boils down to society. What they hell are you going to do when that's standard, when women are expected to stay home te n months with the baby an d the m a n stay home two months? T m ean righ t there there's a d ifference. I mean, society has to solve the problem. Instead of the women doing as well at research as men and at the same time doing more at home. It doesn't make sense." Later he describes the success of the fe male senior researchers at his department in the fo llowing way.
"It's all by chan ce . Yo u can't have a small department and start drawing grand con clusions. It's just turned out so,itjust so happens that both Miranda and Isabel are doctoral students fr om here. They've succeeded, I guess you could say it's luck that we've had two bright doctoral students that have shown, in Miranda's case she had no children so in her case I guess you could say that she's been able to act like a man in that way. Isabel has succeeded with three kids and I really respect that. But I think she has a husband who does a lot at home. So I reaJly think, I blame society, no not society but the attitude in society. I don't think that it'll get any better before women have true equality, at home. How the hecks else are they supposed to succeed at the job if they don't, if they do everything, more than 80-90 percent at home." Howard suggests two explanations fo r women's poorer academic career output. First he suggests that women are less pushy than men and second he situates the problem to the private sphere. He uses the term society but it becomes clear that he is talking about the researchers' fa mily situation. It is also interesting to note that he refers to the notion of coincidence when referring to the success of his fe male colleagues. It is unclear whether he means that their success is due to coincidence or if it is a coincidence that they were trained at his department.
The fo llowing quote is fr om Amanda who is married to a researcher and has several children. Amanda has also been a doctoral student under Howard but she left the department after completing her PhD.
''Yes, but the problem is that they're cutting back and just when I was finished and was about to defend my th es i s, or just when I went on m a te rn i ty leave, and T didn't want to stay. I saw th ese women who struggled with their seventy-live percent work time and who had children and didn't have time fi>r it all and who were researchers too. And it was them who ended up staying at home when the children were sick and I mean, you fa ll behind on the research fr ont and it really is pretty tough being a woman, and being a woman and a researcher.
So I just felt like the pressure was too much both at home and at work. Ve ry Jew know that they have research money and whether they have a position. There are so few permanent positions at the university." Amanda has mainly been teaching during the past seven years. She is currently working as a teacher at a college and has ambitions to combine research and teaching. Amanda states that she did not want to stay at the department after she had her baby. We can see that Amanda is ref erring to women's double workload combining fa mily life with an academic career. Amanda is also critical of the structures of academia th at offer so fe w tenured positions. As we can see Amanda points at several processes influencing gender differences in academic careers as compared to Andy and Howard who mainly stressed the fa mily situation of researchers.
Researchers that had small children seemed to be more inclined to alternate between different explanations that sometimes included processes in the fa mily situation of researchers, or pro cesses included in the everyday life of academia, when explaining gender differences in academic careers. Later in the interview Amanda also reported that she had fe lt like an outsider while pursuing her PhD. She had fe lt invisible and experienced that her supervisor ignored her in favor of a male PhD student which made her fe el discouraged.
Irene did her research training in Department B and is now working as a senior researcher at another department. Irene gives the fo llowing explanation fo r why there are gender differences in academic careers. Irene is married to a researcher and has several children.
"Well, often the women will say that they're not 89 p repa red to invest �o m uch , or don't think they have ti me, or that it'H not �o much f'u n that th ey have the energy for it all. I mean , it's very unu�ual that �omeone at that level will say they're �tupid. If' you've gotten your PhD you don't say you're �tupid becau se you're not, or else you would never have gotten your PhD. No, it's more th at, you don't say that you aren't good enough becau::;e you're ::;tupid, but because you don't want to spend so much time, or because it end::; up hurting the kids. Or becau::;e there are other things in life too. 1 fe lt that way, when I wa::; at the end of my rope, l thought, God it would be nice ifl didn't get any research fu nding, then I wouldn 't have to deal with this rubbish, an d maybe it'::; wrong to call it self-censorship, it's maybe not being prepared to pay the price. Not that the men have said that we shouldn't (laugh ) but because you actually don't want to. Because it's just not worth the high price. I guess that's being a negati ve role model, a fe male professor at the department who's a negative role model, hard on oneself and on th ose around her." Irene states that it is not due to intellectual inabilities that women leave academia. Instead she points at the harsh working conditions in academia and women being less willing to "pay the price". Irene also suggests that this is a common explanation among women. This makes sense since the construction of women as lacking can shift into an overvaluation of fe mininity, particularly when constructing motherhood. This can also be interpreted as criticism against the academic system. She refers to a fe male professor who is "hard" on herself and on the people in her surroundings, a negative role model.
In the quotes above we can see how women are constructed as lacking in a positive sense, "not being willing to pay the prize" or "valuing the fa mily higher than a career". Thus, the lacking in a negative sense, that is to devaluate women, tend to shift into overvaluation of the same, but both discourses are exclusionary. We can also see that researchers with children are more inclined to mention different fa ctors as obstacles varying fr om fa mily situation to the structures of academia when they talk about 90 gen der in academic workplaces.
One tentative explanation fi.>r why the diH course of the impossibility of combining motherhood with an academic career seem::; to be prominent when explaining gender dif ferences in academic careers is that in such a discourse tho problem is both individualised and placed outside academia. Focusing on internal processes in academia would imply criticism ofthe structure of academia and would therefore require collective action fo r change.
Internalization of Inferiority -Su periority through Paternalism This section offers different examples of how paternalism coexists with the discourse of gender equality when making sense of gender differences in academic careers. The first quote comes fr om Irene quoted above and the second is from Stanley who is also fr om Department B. Let us begin with Irene. Irene was asked: "Do you think it is easier for men than fo r women to pursue an academic career after PhD exam?" She answers the question by telling a story about a fe male colleague of hers.
"Where I work now, there is a fe male researcher. When she received tenure after her PhD, she hadn't planned to continue after PhD. [ ... ]but by chance a tenured position became available and her previous supervisor encouraged her to apply fo r it and she got it. I thought, a bit unfair, that she got it because she was a fe male but since I came to the department I've realized that she's one of the best in her generation. But she had given me those signals that she was not going to continue, that she didn't want to and that she didn't have the drive. But she had that [[drive: my ref.]] I think she's one of the best in her age group. She was putting herself down, I've tried to tell her 'you gave me those signals that you weren't good enough, but you are, there is no question about it' ".
The example should be studied in the light of the Swedish context which includes a recent initiative to raise the number of fe males at professor level as well as fa cilitating the transition fr om PhD exam to tenure (Bill 1994/ 95: 164). ThiH is to be done with the h elp of affirmative action measures and has led to a vivid debate about tho competence of women being emp loyed under these conditions. The fe ar of many i s that women will be seen as less competent and as h aving been oflercd positions or scholarships only because they arc women 1 11• Such argu ments rest on the notion of meritocracy and reveal that the dominant discourse offemale researchers is that they arc inferior researchers compared to their male counterparts. In other words, if meritocracy is the basis for ad vance ment in the academic system then women should be able to make it on their own as well as men.
If they do not make it on their own, it is because they are not good enough . The notion of meritocracy has been widely criticized during the last few decades and social scientific studies of science have highlighted tho social dimensions in the reward and evaluation systems of science.
Fem inists have pointed at the impact ofthe "old boys network" fo r fu ture career development and concepts such as "glass ceiling" or "subtle discrimination" are well established in the field of gender in higher education (Harding 1986, Caplan 1994, Long and Fox. 1995, Mackinnon 1997, Husu 2000. In this quote Irene is very keen on stressing her position in relation to her colleague. It becomes obvious that she perceives herself as the older and more experienced researcher in relation to her colleague "[ ... ] I've realized that she's one ofthe best in her generation." Or "[ ... ] I think she's one of the best in her age group".At a first glance she appears to be very supportive ofher colleague. "She was putting herself down, I've tried to tell her 'you gave me those signals that you weren't good enough, but you are, there is no question about it'. These lines also reveal that she constructs herself in the position as the wiser, more experienced researcher who en courages her younger colleague to pursue an academic career. It is tempting to stop the analysis here and conclude that she is supporting her colleague but at the same time constructing her as less competent and inferior to herself. That would be to ignore the issue of subject positions11 and the complexities of it. If con sidering the lines again it becomes obvious that the tone is quite paternalistic, consequently the fe male researcher has take n tho subject posi ti o n of a (paternalistic) m a le rm;carcher wh e n ex plaining gender diflerences in sci enti fic careers. She constructs herself as superior to her fe male colleague and constructs her co l l eagu e as feminine and inferior to her. This is n ot a:.; contradictory as it sounds; it can be interpreted as one strategy fo r constructing h erself m o re positively in a discourse where femininity is devalued. In addition it is important to remember that since academia is male dominated, both historically and in its dominant discourses, there are no subject positions available to women academics unless they construct them th emselves. This can be hard to accomplish without support.
In many areas the demands for equality between women and men in contemporary western societies have stimulated the processes of a reformulation of gender identities of both women and men. The interview excerpts above illustrate that gender diflerences in scientific careers arc still explained in ways that reproduce gender hierarchy. Stanley fr om department B was asked the same question about gender differences in academic careers as Irene. Stanley has several children and his wife is also a researcher. He is ambivalent towards the effe ct of gender on scientific careers.
"Well, it's a world of old men, doing research, or at least has been. Women who continue as researchers, they have to show "fighting spirit", be really, [good: my ref.] they have to work even harder [than men: my ref.] , it's sad but that's the way it is. I mean, it's not the way I want it to be. One can see how many PhDs there are as compared to doctoral students at the depart ment. If considering sex distribution. It's pro bably fifty-fifty [[women and men: my ref.]] at doctoral student level but not at PhD level. But I believe we have a lot more fe male doctoral students compared to chemistry and organic chemistry, I'm not sure they have any fe males at all, well, maybe a few anyway. Why is it like that? It shouldn't be like that." gender differences in scientific ca reers are thus deferred to processes within academia, which can be :;eon in light of the prev ious quotes. This i:; even 1nore intere:;ting when con::;idering the reseurcher::;' fa mily :;ituation; re:;earchers without ch ildren were more inclined to stress the importunce of having ch ildren as an explanation than wa::; tho researcher who had children and who pointed at several diflcrent difliculties, the family situation being one of them. The quote also reveals that Stanley is part of the category that he refers to as oppressive, however, in a eli ftcrent age and power position. He is also keen to emphasize his dislike, which makes his nrgument n bit defensive: "it's sad but that's the way it is. I mean, it's not the way I want it to be". The first pnrt ofthe sentence also reveals an element of acceptance: "it's sad but that's the way it is". The concept of "a world old men" appears several times when discussing reasons for gender diHcrence:; in scientific careers.

The following quote oilers fu rther insights into
what is implied in this concept.
"It used to be easier fo r men to get tenure, that was what the study by Agnes Wold showed [ ... ] I don't know whether it' stills true but that's what it showed. I don't know, but it seems as if women who want to have an academic career have to work so hard, on the other hand it should be like that fo r us too. One can hope that the problems are due to generational differences, but I'm afraid they're not. I fe ar that it's something you learn fr om older colleagues and you take after their values. Yo u become like them even though you didn't think like that fr om the beginning[ . . . ]. It is a world of old men, they have all the high positions. From the beginning you're so self-centered and deter mined to do your own business but then you attend meetings, meet people and hear how they talk. Then you see how it is." In this quote we can see that Stanley refers to a study, which states that nepotism, and sexism effects grant distribution in the life sciences. He expresses doubts about the relevance of the study "I don't know whether it' stills true but that's what it showed." A bit later in the quote he states: "[ ... ] on the other hand it should be 92 like that fo r u s too". This implies that he i::; aware of the inequalities ofthe system, though he is ambivalent towards his position in it. He also refers to socialization processes and there is a bit of self-victimization in it implying that he himself is exposed to the patriarchal structures "You become like them even though you didn't think like that fr om the beginning! . .. I It is a world of old men, they have all the high positions". On the other hand we can see that he is in the process of internalizing their views: "You become like them even though you didn't think like that fr om the beginning". Stanley is ambivalent throughout the interview. He appears to be in a transitional period between, on the one hand keeping his distance to the system and being able to reflect critically over it On the other hand he is on his way to being socialized into the system, realizing that he is inferior to the privileged group. He also realizes that in order to belong to this group he has to pay the price, to "become like them". To "become like them" also appears to include a denial of gender inequality in academia, since it is highly unlikely that someone in the privileged group would openly declare that they had reached their position due to positive discriminating practices. His position, between criticizing and internalizing the dominant discourse of gender in academia, may be one explanation fo r his ambivalence towards the relevance of gender inequality in academia.
If we explore the concept of "old men" in the quote we can see that masculinities, seniority and academic position construct the group that holds the regime of truth of gender differences in academic careers in academia. This con struction also implies that its binary opposition is different types of fe mininities. Thus, we can say that the groups of older men and younger men mutually reinforce each other since a major way of constructing masculinities is to compare with other fo rms of masculinities and to negate fe mininity (Collinson and Hearb 1994). For women to belong to this group they would have to change subject position; however, the price they would have to pay would most certainly be much higher since they would have to break considerably more normative rules as compared to men that enter this group.

Conclusions
As poi nted out earlier it is h i ghly unlikely that someone would openly d ed are women a::; le::;s suited fo r pursu i ng an academic career. Still, gender diflerences in scicnt i li e careers based on gender i nequal i ty per::;ist. In thi s article I have studied ways of reproducing gender inequality with in the d i::;course of equality at two difl"crent Swedish university departments in the same area ofbiology. The results ofthe study indicate that the interviewed researchers at tho depart ments arc very keen to speak within the dominant discourse of gender equal ity operating on a general level of society. Ye t, when making sense of gender diflcrenccs in scientific careers, they tend to reproduce gender hierarchy, though in more subtle ways. One example is the exclusionary practices that seem to be at work. Women tend to be excluded, either by being constructed as problematic and lacking im portant qualities fo r succeeding in academia or by being constructed as different and sometimes overvalued as compared to men, however, both ways are exclusionary. There is also a tendency to individualize gender differences in scientific careers. Focus is often put on processes outside academia such as inequalities within fa milies.
Another finding was that different fo rms of paternalism coexisted with the discourse of gender equality. By internalizing superiority as in the case of Irene in relation to her fe male colleague or by internalizing inferiority as in the case of Stanley in relation to the "world of old men" the gender order was reproduced when making sense of gender differences in academic careers. Thus, it appears to be more rewarding fo r researchers (male and fe male) to internalize the dominant views of gender rather than contest them and this seems to reinforce the "self reproductive" tendencies in the academic system. By internalizing the regime of truth of fe mininity and masculinity the researchers can reproduce the symbolic order of gender in academia and still speak within the dominant discourse of gender equality. Thus, they construct themselves as "gender neutral", "good" re searchers. Gender differences based on gender inequality in scientific careers becomes some thing that is outside the norm of gender equality in th e Swedish u n ive rs i ty systcm 1 � . Gender inequality i::; then constructed as ::;omothing· "abnormal" or at least something that happens at othe r department::; or universities.
The lack of avai l able subject positions fiu· women academics, and the hard work of" co nstructi ng such alternative positions, is an obstacle for women academics to support each other. The absence ofsuqject positions for women, combined with the often unreasonably high expectations on women to support other women, arc factors that fu rther complicate how women researchers perceive other women researchers. As pointed out earlier, it is very hard to avoid making gender since it is a major way of categorize and making sense of our daily lives.
In order to stop producing and reproducing hierarchical relations between fe mininities and masculinities we need to fu rther examine how we create and maintain the symbolic order of gender when making sense of different situations and relations related to everyday academic working life.

Acknowledgments: I thank Nora Rathzel and
Elin Kvist (Department of Sociology) and Marin Carbin (Department of Political Science, Umea University, Sweden) for valuable comments during the process of writing this work. 2. In this article "scientific", "academic" and "research" career are used synonymously. They imply that the researcher has completed his/her PhD and continued to work with research or research related work within academia, or fo und research related work outside academia. 3. In this article "gender order" implies an asym metrical relation between fe mininity and masculinity where the fe minine is devalued in relation to the masculine. 4. There is an ongoing debate about the usefulness and implications of the distinction between (social) gender and (biological) sex. This will not be fu rther elaborated in this article, however, fo r an introduction to the debate, please consider Sit.uat.ed Kn owled,.;e's (19!:16). !:l. The depa rtme n t!; have prcviOt.JRiy been deRcribed a nd a n alywd by sociological and bibliomelric rncnnR in "MappinggendPrdil'lo re ncPH in :-�cientilic career!; in :-�ocial and bibliomelric space" Science l.ech n ology and human va lt(('.'l. Vo l. 26. No. 2 S p r· i n g 2002: 167 190. 10. An n n a l yRi R oflhc med ia diRcou rRc that lilllowed al't er the bill was pa:o;sed revea led that these ty pe::; o l' argu ment!; were very common and that male academic::;, with one exception, represented those who were negative io the bill. On the positi ve side wer·e women ami men as well as academics and non-academic!; ( Bonde!;iam 1 999).
11. Di sco urRes produce subject s (for exa m pl e the madman or the cri minal), however they also p roduce p l aces lilr the su�jeci where the discourse makes most sense. In other words discourses producc d ifleren t su�j cct pos i ti ons thro ugh which th e s u bject can make meaning ( H all 1997: 56). 12. Gender inequality as a contradiction to the Swed i sh "norm ol' equality" has p rev i o u s ly been elaborated by Ron n blom et al. in Jti mstiilldh.et, retorik som Pra.ktik? ( 1999).