How Dutch and Italian women’s networks mobilize affect to foster transformative change towards gender equality

This article contributes to the debate about the role of affect in transformative change towards gender equality, by comparing the building of affect in two recently founded women’s networks in Italian and Dutch universities. By conceptualizing networking as a social and cultural practice that organizes a collective body through the building of affect between specific groups of organizational stakeholders, we reveal the emotional, dynamic and context-dependent character of transformative change. We found that similar women’s networks build affect with organizational stakeholders in different ways, shaping boundaries of different collective bodies through emotions and adapting to the cultural meaning attributed to earlier actions. By bridging bottom-up and top-down approaches to gender equality and tailoring the building of affect to the local context they potentially contribute to transformational change towards gender equality.


Introduction
Many organizations installed women's networks to foster gender equality (Singh et al., 2006), considering these networks a compensation for the exclusionary effect of existing 'old boys' networks (Durbin, 2011;Kanter, 1977). Women's networks are both praised for their benefits for women's careers (Vinnicombe et al., 2004) and criticized for focusing on 'fixing the women' without transforming the gender relations of organization Meyerson, 2000a, 2000b). Although scholars increasingly recognize the critical potential of affect to transform organizations (Fotaki et al., 2017), they have focused on networks as sources of knowledge and means for knowledge creation (Durbin, 2011) rather than as affective bodies (Ahmed, 2010). In this article we call for understanding networking as a social and cultural practice that organizes a collective body (Ahmed, 2010) through the building of affect. We aim to contribute to the literature about transformative change towards gender equality, by revealing the role of the building of affect through women's networks in transformative change towards gender equality. The research question is: how do women's networks in universities build affect to foster transformative change towards gender equality? Firstly, we discuss the literature about transformative change, women's networks and the role of affect in networking. Secondly, we analyse two case studies of women's networks that were set up to foster gender equality in academia from the perspective of affect and embodiment. Finally, we discuss the building of affect in women's networks to foster transformative change and how we can understand this.

Theory
The literature assumes that gender inequalities in organizations are reproduced through informal processes and practices that set barriers in organizations (Acker, 1990(Acker, , 2006 thereby creating what Kanter (1977) defined as proper 'shadow structures'. Transformative change towards gender equality refers to 'addressing and trying to change gender as a structure' (Benschop and Verloo, 2011: 283, italics added to original text), thereby challenging power relations within organizations. Scholars consider transformative change more effective in fostering gender equality than strategies exclusively aimed at equipping individual women ('fixing the women') or revaluing gender differences (managing diversity), since the last two may keep unequal power relations in place (Benschop and Verloo, 2011;Meyerson, 2000a, 2000b). Transformative change happens when individual and collective agency becomes integrated with structural dynamics in organizations (Archer, 2003).
The literature increasingly discusses the role of networking in understanding and addressing gender equality in organizations (Avdelidou-Fischer and Kirton, 2016;Bleijenbergh and Van Engen, 2015;Cacace and Declich, 2016). Some scholars explain how gender inequalities in organizations affect the social networks of individual women (Berger et al., 2015;Bird, 2011;Gray et al., 2007;Timmers et al., 2010). Gendered social practices, such as emphasizing the social qualities of women, self-evidently expecting women in meetings to be the secretary, emphasizing the physical strength of men and making jokes at the expense of women, hamper women's individual networks (Berger et al., 2015: 19). Such social practices negatively affect the chances of individual women to have knowledge of or access to resources, to be recruited or get promoted (Bleijenbergh and Van Engen, 2015;Bird, 2011;Gray et al., 2007). The individual approach towards networking has been criticized for focusing on networking as a means to 'fix the women' without transforming the gender relations in an organization Meyerson, 2000a, 2000b).
Other scholars have focused on the role of so-called 'old boys' networks' in reinforcing patterns of exclusion and isolation (Durbin, 2011;Rees, 2011;Van den Brink and Benschop, 2014). This literature focuses on the group characteristics of networks, using the concept of homophily to connote individuals who share similar identities (Durbin, 2011;Kanter, 1977). Men gatekeepers easily identify with other men in their networks, so mobilizing masculinities (Van den Brink and Benschop, 2014). With white men being the dominant group in organizations, women and minorities as a group have less access to and need to invest more time and effort in informal networks that are based upon homophilous interactions. This literature considers networking between women a way for women to seek refuge from exclusion processes (Durbin, 2011: 97-99). Yet other scholars understand networking as a way to mobilize women and foster transformative change (Avdelidou-Fischer and Kirton, 2016;Nash, 2002). Via mobilizing groups of women, women's networks potentially counter the lack of widespread and bottom-up support for gender equality action (Timmers et al., 2010) and the connected risk that gender equality is reduced to a simple managerial issue (Castaño et al., 2010).
To summarize, scholars consider establishing women's networks a way to extend the social networks of individual women (Singh et al., 2006), to constitute an alternative towards the so-called 'old boys networks' (Durbin, 2011) and to mobilize women as a group (Avdelidou-Fischer and Kirton, 2016;Nash, 2002).
In this article we build on the third perspective by assuming that mobilizing a critical mass of people, empowering both women (and men) to take action for gender equality, and involving groups already active on the same or related issues is necessary for transformative change towards gender equality. We define networking as any kind of bottomup mobilization towards a common goal, which may, but not necessarily does result in the establishment of a formal network. We define a formal women's network as a collective body which publicly articulates to mobilize itself towards the common goal of fostering gender equality. In understanding the networking we focus on how emotions work to shape surfaces of bodies, 'which take shape through the repetition of actions over time, as well as through orientations towards and away from others' (Ahmed, 2010: 4). Although scholars increasingly recognize the critical potential of affect to transform organizations (Fotaki et al., 2017), until now the understanding of women's networks has focused on the role of knowledge (Durbin, 2011) rather than the role of affect in such mobilization. We assume that some things about change we can only grasp when we pay attention to affect. Inspired by the work of Ahmed (2010), this article contributes to this understanding by conceptualizing how affect operates to shape the collective body of a women's network, trying to understand how the initiators of such formal networks invite member and stakeholders to mobilize towards gender equality through the working of emotions.
This article examines networking as a social practice through which affect is built (Ahmed, 2010;Clough, 2007). Affect is here defined, after Gregg and Seigworth (2010), as a movement within 'a state of relations' between different things or people implying the 'passage of intensities' (Gregg and Seigworth, 2010: 1). We draw on the classical work of Elias and Scotson (1965) to understand the relation between the experiences of the actors that initiated networks (and in particular their effort to build a safe and 'intimate' space to support women's free expression of their own needs, problems, desires) and the change process within the larger organization. Elias and Scotson (1965) highlight the importance of networks in building social cohesion, sharing experiences and celebrating the value of women's experience and problems in male-dominated environments (cf. Fotaki, 2013). Networking may create egalitarian alliances with organizational stakeholders (which is more difficult for isolated members of less powerful groups in organizations), thus promoting transformative change.
Networks may increase the cohesion between people belonging to marginalized groups in organizations by strengthening their self-confidence and reciprocal recognition (Colgan and McKearny, 2012), creating conditions to link with other organizational stakeholders and promote cultural change (Dalal, 2001;Elias and Scotson, 1965). However, the literature suggests networking is a context-sensitive activity (Knappe and Lang, 2014). For instance, in some national contexts, promoting women-only groups (not to speak of feminist networks) historically was considered ideologically loaded and threatening for the status quo, while in others, women's networks flourish and are selfevident (Avdelidou-Fischer and Kirton, 2016). This article further unpacks how the building of affect shapes women's networks in different contexts and contributes to the debate about the role of affect in transformative change towards gender equality, by comparing the building of affect in two recently founded women's networks in Italian and Dutch universities.

Case description
The comparative case study in this article is embedded in a larger action research project on gender equality in academia between 2012 and 2015, funded by the European Seventh Framework Programme (STAGES, 2011). As part of this action research we set up women's networks in two universities, the Università degli Studi di Milano (UMIL) and Radboud University (RU) in Nijmegen. UMIL is a public teaching and research university with eight faculties and two schools and an academic staff of 2140 professors. With 64,000 students, it is one of the largest universities in Italy and a leading institute in Europe. The proportion of female full professors at UMIL was 27.7% in 2013. RU is a student-oriented research university with seven faculties and 20,000 students. Like the UMIL, it has a wide variety of disciplines. The RU had more than 2900 full-time equivalent (FTE) academic staff 1 in 2013 (Radboud Universiteit, 2014: 65). With 20.8% female full professors in 2013, RU had the largest proportion of female full professors of all Dutch universities, although this proportion is 7% lower than at the UMIL.
We consider the women's networks in these two universities most similar cases (George and Bennett, 2004: 50), since these networks were both established in 2012 by research teams of interdisciplinary universities (thus not technical universities, or arts/ humanities universities), with the aim to support transformative change towards gender equality and supported by European funding. In both universities, the creation of women's networks was aiming at early-career female academics and was part of action plans to support transformative change towards gender equality (STAGES, 2011). In Milan (UMIL) the action plan linked the creation of a network of early-career female academics at the Faculty of Agricultural and Food Sciences with the implementation of two other actions -developing a Course on European Project Drafting and Management and a Course on International Publishing. The two courses addressed both men and women academics. By creating the network, the researchers aimed at mobilizing and including a specific group of beneficiaries in the planning and implementation of the courses, while at the same time creating an informal space for further developing activities on gender equality. The target audience of the women's network were female PhD candidates, postdoctoral researchers and research assistants and the meetings were mainly organized around preparing the two courses.
The action plan in Nijmegen aimed at creating a network for early-career female academics to support the empowerment of this group (STAGES, 2011). The action researchers in Nijmegen cooperated with colleagues from their professional (disciplinary) networks to set up a university-broad network, which -although men were allowed entrance to the meetings on invitation -primarily targeted women (PhD candidates, post-docs, assistant and associate professors) and their specific needs. With this target audience, the action researchers addressed faculty that until then were not covered by the existing network of female full professors (NVH). They named the network the Halkes Women Faculty Network after Catherina Halkes, the first full professor in women's studies at the university. Their meetings addressed themes like work-life balance, publication strategies, expat experiences and communication of research results.
In both cases, the action researchers connected the new women's networks to existing institutes for gender studies at their universities. At UMIL an interdepartmental centre 'Women and Gender Differences' had existed since 1995. The centre, which coordinated the implementation of the STAGES Project Action Plan at UMIL, was reorganized in 2015 as GENDERS -Gender and Equality in Research and Science -thereby becoming the first centre in Italy with an explicit focus on gender in science. The action researchers in Milan were formally affiliated to the centre. Like the Italian university, the RU had an Institute for Gender Studies, installed in 1985 and transformed into a network of Gender and Diversity Studies in 2016, supporting the development and integration of gender and diversity scholarship in research and teaching. The action researchers in Nijmegen had strong informal connections with the institute, and one of them was formally affiliated to it.
The action researchers in Milan involved the women's network in the design of general activities supporting the career path of academic staff, namely the aforementioned courses on European Projects Drafting and Management and International Publishing.
They had become aware of the need to involve early-career female academics during preliminary research. When drafting their action plan, they had gathered sex segregation data, discovering that women had more or less equal chances as men to obtain European funding, but applied less than men did. These findings matched European research results, especially with regard to the field of agricultural sciences (European Commission, 2012: 126-129). Subsequent semi-structured interviews with female academics at the Faculty of Agricultural and Food Sciences revealed that most of them perceived access to European funding schemes as very difficult, considered applications to be time consuming, complicated and offering limited possibilities of success. This perception, which did not match actual success rates, was widespread. The action researchers thus decided to develop activities that would sensitize female academics about the importance of European funding and of international publishing for their careers. The overall strategy for transformative change was to involve early-career female academics in setting up facilities that were relevant and open to both men and women. During network meetings, the action researchers invited participants to share their perceptions of barriers in getting project funding and publishing. They discussed and agreed upon the contents, speakers, timing and logistics of the two courses with participants to enhance their participation in the courses.
The action researchers in Nijmegen set up the women's network to empower earlycareer female researchers. In collaboration with female academics from their professional networks who had also articulated the intention to set up a network, they decided the aims, target group and activities they would pursue. They formed an official network board of six female academics, formulating the formal aims of connecting women of different career levels, supporting their academic career development, supporting the empowerment of women in general and the enhancement of an inclusive academic work environment (internal email 1 October 2012). The first activity in September 2012 was the launch of the Halkes Women Faculty Network website, presenting the network's aims, the identity of its board members and its planned events. After its kick-off meeting in November 2012, the network board organized a series of 10 meetings (round tables, seminars and conferences and meet-and-greet meetings with senior academics) during the research period (2012)(2013)(2014)(2015). After successfully involving the University Board in the kick-off meeting and having several formal meetings with them, the network acquired financial and administrative support from the Board (budget of 5000 Euros per year). The Halkes Women Faculty Network became recognized as a dialogue partner in setting university equality policies. Together with the network of female full professors (NVH), the network mobilized a critical mass of women faculty that supports and scrutinizes the development of gender equality policies at the university. To illustrate, the Halkes Women Faculty Network worked together with the NVH to influence university policies via letters and policy reports. Some months after they jointly wrote a formal letter to call for a female University Board member, the first female member in the history of the university was installed in 2013.
A first comparison between the Dutch and Italian women's networks suggests that they organize different types of activities: the Dutch network organizing women-only activities aimed at empowering academics, the Italian network involving female earlycareer academics in setting up activities accessible to both women and men.

Data collection and analysis
All authors of this article were active members of the larger action research (STAGES, 2011). Three of them were directly involved in setting up the networks in their universities. The other two authors were not part of the networks but of the research on the networking. The three authors setting up the networks (first, third, fourth author) were able to make field notes through their direct involvement. During the research period (2012)(2013)(2014)(2015), the second author bimonthly gathered these field notes, additionally half yearly interviewing the three authors about their experiences. This made data collection and data management similar for both case studies. The field notes and transcripts of interviews were stored in a database ultimately consisting of 64 individual texts and analysed and coded by the two authors that examined the networking (second and fifth author). The sensitizing concept during the coding procedure (Corbin and Strauss, 1990) was transformative change, while the role of emotion, embodiment and context emerged during an abductive process of interpretation (Reichertz, 2009). The analysis helped to identify how affect operates to shape transformative change. In order to compare the two women's networks, the dynamics of transformative change were analysed in relation to the mobilization of affect towards different actors: from the action researchers themselves to other individuals and groups pursuing similar objectives and the university leadership.

Creating a space: Affect between network members
Our analysis showed how the action researchers in both universities created space for the building of affect with and between network members. In Milan, the researchers built affect with potential network members at the Faculty of Agricultural and Food Sciences via face-to-face contacts. Affect emerged because the action researchers, like the target group, were also female academics with a temporary employment position; they literally embodied being one of them which enabled feelings of 'trust'. As was recorded in the field notes, a researcher in Milan observed that, . . . the core [action researchers AUTHORS] team consisting of young, non-permanent researchers [with a temporary contract, AUTHORS] favors the relation with young women researchers in the faculties involved, which are an important part of the project, particularly through the network. We are a reference point for them, which they trust. (29 January 2014) The Faculty of Agricultural and Food Sciences had participated in other activities on gender equality promoted by the centre 'Women and Gender Differences'. Contacts with gender researchers were thus already in place. The action researchers established contacts with early-career female academics by interviewing them when drafting their action plan. By visiting and listening to them, they embodied concern with the situation and issues of the target group. They confirmed the contacts during the launch meeting of the network (with informal refreshments) attended by over 20 researchers. The action researchers observed that participants came in small groups, part of the same research group or friends. They organized further meetings, in this vein, during lunchtime (to avoid overlap with other meetings and classes) which were informal as the participants had lunch together (the team provided food and drinks). Participants reported feeling that they could have a constructive and safe 'exchange among peers' (Avdelidou-Fischer and Kirton, 2016) and at the same time meet experts in the field of gender equality who listened to them and provided them with support. During further meetings, the number of participants grew to 25-30 as some participants involved friends and colleagues.
In Nijmegen, the action researchers 2 created space to build affect with and between network members by organizing roundtables during which participants could exchange experiences with colleagues from other departments. The action researchers chaired such exchanges around topics on which participants had signed in, providing them with bread and soup at the same time. An action researcher recorded this process in the field notes in March 2013: Women from the whole university were invited six weeks in advance and they could tell the team if they wanted to talk about work-life balance, visibility or publishing. Women wanting to discuss the same issue were put together, and three almost equal groups were formed. We started the meeting by asking them why they were interested in the subject, what were their experiences, which challenges they faced and what did or did not work well for them. The meeting was really nice, with a lot of very young women, but also senior women, full professors and women in their middle career, and they were so surprised that so many women had the same issues. (4 March 2013) The analysis suggests that the exchange of experiences in a physical setting of sitting around tables supported feelings of recognition and comfort (Colgan and McKearny, 2012). The physical setting fostered the idea that all women (and some men) participated on an equal level despite different disciplinary backgrounds and hierarchical positions. The roundtable meetings were repeated four times during the research period, with an average of 49 participants attending. Other meetings more indirectly fostered affect. For example, female and male full professors gave lectures during seminars about being a woman or a change agent in academia, functioning as role models by sharing personal experiences with the audience. The action researchers encouraged them to reflect upon their personal development and private situation, so inviting feelings of recognition in the audience. Finally, during meet-and-greet meetings senior academics functioned as an individual conversation partner of participants, allowing to bring personal doubts and questions into the conversation. Creating such a safe space for exchange was another way to build affect.

Participating on equal terms: Affect with male network members
Building of affect between network members not only involved women, but also some men. In Milan, whilst the meetings of the network were only open to female academics, the activities derived from them (the Course on European Project Drafting and Management and the Course on International Publishing) were open to both sexes. This practice reverses the more common situation that general services in universities are set up from a masculine perspective to which women have to adapt (cf. Durbin, 2011;Fotaki, 2013). The men who attended the activities were, in particular, young male PhD and post-docs, joining their female colleagues and so physically embodying a new generation that would be open for transformative change.
In Nijmegen, building affect with men as network members was a side-effect. The network sent email invitations for meetings only to female PhD candidates, post-docs, assistant and associate professors. So, male academics were not directly invited, but the network board decided they were welcome when invited by female colleagues as guests. During the research period, on average two or three men participated per meeting, compared to on average 30 women. The few men participating in the meetings became easily part of the group, exchanging experiences in roundtables and being part of the audience in seminars and conferences, so embodying their willingness to participate on equal terms.

Legitimizing positions: Affect between generations of feminists
Our analysis showed that both women's networks build affect between junior women and older generations of feminists. Existing women's groups consisting of feminists of an older generation were the first actors the new women's networks had to relate to. In both cases, this relation was not self-evidently affective from the beginning. This resonates with literature about feminist generational conflict (Dean, 2012). In Milan, the action researchers identified that existing groups of feminist scholars often have a leftwing political character, which would estrange the target group of early-career female academics from the start. The field notes show the perspective of an action researcher in Milan talking about a feminist generational conflict: This is also one consequence of the typical hostility of the Italian leftist movements towards academic institutions. The fact that a middle generation [women in their 40s] is missing is worsening the situation, so that you prevalently have women in their 60s and women in their 20s. (13 September 2012) Italian feminism, indeed, traditionally had a strong autonomous and extra-institutional character, which delayed the introduction of gender studies as a proper field in academia (Barazzetti and Di Cori, 2001). Italian universities have few politically active and committed scholars in the field of gender equality (mostly in the social sciences, law, philosophy, history, etc.), causing an intergenerational gap with feminists in early academic careers. While establishing the network, the action researchers in Milan thus did not have to face existing formal women's networks, but rather faced informal groups of feminist scholars at the university. The fact that the women's network was created in the framework of a European funded project on gender equality in science and that it started from the specific Faculty of Agricultural and Food Sciences, however, helped in legitimizing its existence and activities.
In Nijmegen, the installation of a network for early-career female academics at first met with some reluctance among the existing network of women full professors (NVH). The NVH had already achieved an established position at the university, involving financial and administrative support from the University Board. Actively seeking contact, the action researchers had to wait months until the board of the NVH had time to meet and discuss the plans for a new network with them. When a meeting was finally scheduled, the secretary of the NVH board warned them in advance that they would only get 10 minutes to present their ideas. To the action researchers this felt like a lack of enthusiasm from the senior women towards a new women's network. They adopted the strategy of being respectful and open to the full professors, explicitly asking them for advice. Finally, the physical presence of the younger generation in the meeting with the board of the NVH broke the ice. The new network was represented by female PhD candidates in their 20s and 30s, while the NVH was represented by full professors in their 50s and 60s. That the new network was visibly represented by a new generation increased the legitimacy of the network towards the NVH.
The action researchers in Nijmegen provided an interpretation of the reasons behind the initial reserve on the part of the full professors: I think that the women in the full professor women's network consider themselves feminist. At least the ones who are part of the board of the network. They are women in their 50s or 60s and they do not hesitate to call themselves feminists. . . . They have the feeling that the younger generation doesn't understand how much structural gender inequality is there and they also feel a little bit undervalued in their efforts to allow women's entrance to science. (3 February 2014) The transcript suggests that older generations of feminists do not self-evidently consider younger female academics to be feminists. This illustrates the tension between the different generations of female academics active on the issue of gender equality and resonates with literature about the women's movement in the United Kingdom (Dean, 2012). The action researchers were able to build affect by recognizing the position of the older feminists and respecting them for their achievements, actively seeking their approval.

Becoming a conversation partner: Building affect with the leadership
Both women's networks built affect with the leadership of the university, although in different ways. In Milan, women in academic leadership positions in the Faculty of Agricultural and Food Sciences were heavily involved in the STAGES project since the beginning. The person scientifically responsible for the project was the former (and first female) dean of the Faculty of Agricultural and Food Sciences. Moreover, the action researchers installed an ad hoc board of female and male full and associate professors and researchers at the faculty with advisory and support tasks. Male heads of departments were also very supportive, as an action researcher from Milan noted. The words 'extremely receptive and responsive' literally refer to an exchange of intensities, as is part of the definition of affect (Gregg and Seigworth, 2010).
The launch of a network of women researchers was firstly proposed to the Board of the Faculty of Agriculture, where the action is more viable, since they are extremely receptive and responsive, also thanks to leadership support. (12 March 2013) Women in academic leadership functioned as gatekeepers. It was often through them that action researchers made their first contacts with early-career academics. Female leaders created space for affect by supporting, legitimizing and disseminating the activities of the women's network, without disturbing the perception of meetings as a 'free informal space' among peers.
In Nijmegen, the Halkes Women Faculty Network regularly involved academic leaders in their events. For example, the vice-rector gave an opening speech at the kick-off meeting of the network, and deans and full professors participated in meet-and-greet events with young female researchers. Moreover, a yearly meeting of the network board with the rector to discuss the planning and activities allowed for building a structural relationship with the University Board. In Nijmegen, these academic leaders were mainly men, until in 2013 the first female vice-rector was appointed. The action researchers felt increasingly recognized as a conversation partner on diversity policies, able to actively influence policies in the direction of more gender equality, as reported in field notes: The new strategic plan 2015-2020 has been presented in January 2015 and, thanks to lobbying by different stakeholders like the STAGES project team, the two women's networks and other stakeholders, the plan pays large attention to increasing diversity (in a broader sense than gender equality). Target figures are set for 25% women, 25% men and 25% people from abroad for all positions by 2020. (7 August 2015) The action researchers refer to their own lobbying activities, the efforts of the boards of the two women's networks and other stakeholders as the influence behind the introduction of the target figures, the 'thanks to' expressing pride about this result. Mentioning the different actors explicitly shows their awareness that mobilization was a common effort (Fotaki et al., 2017).

Networking as an embodied activity
Considering networking an embodied activity helped identify the importance of the physical presence (Pullen and Rhodes, 2015) of the network initiators at the same location for a longer period. Early-career female academics (had to) change academic position and geographical location very often due to their precarious working conditions. The network board in Nijmegen experienced a relatively high turnover. During the research period, three out of six board members left, partly due to the termination of a temporary contract. Another board member was absent for a year due to burn out. As an embodied activity, physical health affects the opportunities for networking. Therefore, the high mobility rate of early-career researchers and health problems typical of professionals with a work overload forced the network board to literally look for new bodies. The action researchers decided to increase the board size from six to ten members. So, more board members could share the work and counterbalance long-term illness and changes of location.
The problem of mobility and turnover arose in Milan as well when initial members of the network with temporary contracts changed position or moved away. The action researchers took this issue into account and decided to also include academics with permanent positions who were sensitive towards issues of gender equality in meetings and activities. This granted continuity in the network activities.

Discussion
The analysis shows the emotional, dynamic and context-dependent character of fostering transformative change via the building of affect in women's networks. First, comparing the two cases suggests how emotions shape transformative change, grounding the effort for change in sharing experiences between individuals and groups and so triggering a bottom-up mobilization towards gender equality, which counteracts merely managerial and top-down approaches to gender equality in organizations. Initiators of women's networks built affect with members of marginalized groups like female academics with precarious working conditions by involving them in setting up general activities. This empowered the participants and shaped general activities and services to their needs. Moreover, the women's networks supported members to exchange experiences. This provides feelings of recognition for individual members, but also allowed network boards to recognize shared problems and articulate them to the university boards. By building affect between network members, older feminist groups and the leadership of the universities, initiators of women's networks increase the legitimacy of their claims and commitment.
Second, we show that the building of affect is context-dependent. The two networks, although similar in aim, academic context and target group built affect in different ways, namely involving early-career female academics in organizing general activities in Milan, contrasted to organizing activities for early-career female academics particularly while simultaneously influencing university policies in Nijmegen. This shows how much networking is a practice which 'is basically constituted by people's actions and thus is inherently socially constituted' (Berger et al., 2015: 40). The action researchers adapted to the particular context in Milan to effectively tailor actions to the actual needs of earlycareer female academics. They used a 'reversal practice' (Cacace and Declich, 2016) or gender mainstreaming strategy (Booth and Bennett, 2002), singling out the contents and aims of general services on the basis of women's preferences, opening them up to both sexes afterwards. The design of the services, where the network provided a safe environment to express problems and possible solutions, later inspired new programmes at the university. This highlights the importance of creating an informal atmosphere and shows how building links between network members allows them to voice their feelings and needs, to the benefit of the organization at large. The context dependency of networking is also illustrated by the choice of the women's network in Nijmegen to organize womenonly activities. This approach had already been successfully adopted by the senior women in the organization. The fact that the existing network organized a limited target group of female full professors made it easier for the researchers to legitimize organizing an additional target group, namely women faculty below full professor level. That the existing network of female full professors already received financial support from the University Board, made it easier for the new network to claim additional resources and become a dialogue partner on diversity policies. So, the existence and cultural meaning attributed to existing feminist actors was decisive for how the new women's networks mobilized affect in the two different universities.
Third, we show the dynamic character of transformative change towards gender equality. Women's networks -depending on their evolving composition and agenda, and their dynamic interaction with the organization -are flexible in adapting to changing circumstances, even when they achieve some form of institutionalization. Moreover, they can pursue different objectives at the same time, for example empowerment of their members and a more inclusive work environment; supporting the implementation of existing gender equality policies and promoting new actions; investigating the current situation with regard to gender equality and raising gender awareness among key stakeholders. Finally, by building affect they support reciprocal recognition and cohesion among women and support them to more successfully mobilize and negotiate change with the leadership. This illustrates the dynamic character of women's networking in bridging top-down and bottom-up mobilization towards gender equality, creating a new space for dialogue and negotiation within the organization.
We reflect upon our own position in directly setting up the women's networks (first, third, fourth author) and supporting the setting up at a distance while being responsible for the data collection and analysis (second and fifth author). All authors are women with an academic background, with a master's or PhD in social sciences. All authors qualify as feminist scholars, having received national and European research grants for gender research before becoming involved in setting up women's networks within the context of the larger action research on gender equality. This meant we easily identified with the target audience of female early-career researchers in our countries and the problems and issues they faced. Our theoretical background helped us to see the structural dimension of gender inequality processes, allowing us to identify the many issues the network members faced which were also shaped by the (international) neoliberal organization of academia. This made us willing and determined to articulate the issues of the network members to the university leadership, supported by our position as researchers that received an EU grant to pursue gender equality in our institutions.
As practical recommendations, we suggest university boards to involve women's networks in the design and implementation of gender equality policies. In fact, networks may be able to involve large groups of employees and may serve as a platform for drawing both staff and leadership into gender equality policies. The safe space of the networks supports women academics in precarious positions to express their needs freely, enabling network boards to build coalitions with other stakeholders and articulate these needs in meetings with the organizational leadership. This may, at least, put precarious work on the agenda of university boards. We suggest researchers setting up women's networks take the physical context into account. Create a setting that emphasizes equivalence of participants, like for example roundtables or informal lunch settings. Make explicit appointments about confidentiality of the experiences exchanged to support a safe space. Make sure organizers or speakers can function as role models for the target audience, for example by being in the same (precarious) position or by being willing to reveal personal experiences and private life situations. Provide generous food and drinks that fit with the local context, to give the meetings pleasant incentives.

Conclusion
Our comparative case study of women's networks in academia identifies the emotional, dynamic and context-dependent character of transformative change towards gender equality. Researchers involved in similar actions, setting up a women's network, built affect between network members and organizational actors in different ways, based on the dynamic and context-dependent character of transformative change, the most important being the existence and type of organizational actors involved in gender equality, the cultural meaning attributed to earlier actions of these actors and the (lack of) formal recognition of these actors. Via building affect with members, older feminist groups and leadership initiators shape women's networks towards the local context, dynamically adapting the working of emotions towards the different needs of their members and the organization. In a context where an existing network of full professors had earned financial support and organizational legitimacy for organizing women-only activities, actors mobilized affect between generations of feminists when setting up a network for earlycareer female academics. In a context where feminist organizing was politically loaded, actors mobilized emotions of recognition with other women academics in precarious positions to develop services that benefit career development of both sexes. Comparing dynamics in the two cases highlights the need to reinterpret and adapt gender equality practices tested in different contexts as part of continuous reflection on local needs.
Both cases show how women's networks were shaped through the building of affectboth between the network members and with other organizational stakeholders -by promoting exchange of experiences and mutual recognition and by building commitment from and involving the leadership in their activities. Both women's networks function as organizational actors that are able to negotiate gender equality policies with the leadership by mobilizing affect with their female and male network members and groups of older feminists and so potentially contribute to transformative change towards gender inequality.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and publication of this article: the authors received financial support from the European Union seventh framework programme under grant ID 289051.