Change and stability in academic agency in higher education curriculum reform

ABSTRACT This study explores academics’ changing agency in curriculum work in higher education. Bourdieu’s concepts of field, habitus and capital, followed by the metaphor of game, are used as tools to analyse stability and change in agency. The interview data collection from 17 academics was implemented twice over 3 years after two different processes of curriculum change at one multidisciplinary research university in Finland. Through narrative analysis, two storylines were identified. The storyline with changes in agency included transformative, sidelined and divided narratives. The storyline with stable agency included development-oriented, autonomous and opposition narratives. The lived experiences create habitus as it is, as internalization of social structures but also as unconscious enterprises to maintain old or develop new forms of capital through curriculum change. While competing for the capital, the habitus and the ‘feel of the game’ are shifting. The different narratives show how academics as players in the field of curriculum change have different access to compete for different types of capital. The results raise a question: who can legitimately become an agent in the curriculum process, and what qualities make for an academic ‘fit’ with curriculum change?


Introduction
Over the past decade, the scope of curriculum work in higher education has moved from subjectand discipline-based changes towards university-wide reforms (e.g. Blackmore & Kandiko, 2012;Karseth & Solbregge, 2016;Yates et al., 2017). Rowlands (2018) expressed concern about the diminution of academic voice, referring to the capacity of academics to contribute to decisionmaking in curriculum. Academics may have various levels of access to curriculum work but also respond in many ways to the request to change curriculum. They may feel pressured to take part in curriculum change processes urged by their institutions. The access is different if curriculum change is implemented as an engineering type of activity, where certain processes are followed to meet the expected results, or as a bricolage type of activity with more space to make creative decisions (Louvel, 2013). Curriculum-making is often regarded as a teaching-related activity, and because of that, academics' response to curriculum change may reflect the existing tensions between teaching and research duties (e.g. Mathieson, 2019). In particular, the question of time economy becomes vital in the decisions about where and how it is wise to invest time in order to gain valuable capital in the field of academia (Bourdieu, 1988, pp. 98-99;Osbaldiston et al., 2019).
In the present study, we aim to achieve a deeper understanding of academics' agency in different processes of curriculum change in higher education. To date, curriculum work as part of academic work has received scant attention. Several narrative studies have explored academic work and academic identities facing the changing university environment (e.g. Adler & Lalonde, 2019;Shams, 2019;Ylijoki & Ursin, 2013). Ylijoki and Ursin (2013) found that despite the changes in the university and in academic work, there is also continuity. In their study of academic identities in the changes within Finnish higher education, they found three fundamental storylines: regressive, progressive and stability. A regressive storyline represents the deterioration of work, filled with worries, fears and disappointments, and appeared, for example, as a narrative of resistance and a narrative of loss. With the progressive storyline, the authors refer to a movement towards optimism, emerging, for example, as a narrative of success and a narrative of change agency. A stability storyline included a neutral stance towards university transformation.
Academics have a variety of narrative identities-for example, a rebel, loser, winner, overloaded worker, bystander, change agent-all of which they can combine, rebuild and renegotiate (Ylijoki & Ursin, 2013). This is similar to the findings of Shams (2019), who uses a concept of 'elastic identities', as academics try to fit themselves into the different situations they face. Shams (2019) states that academics' identities are subjectively produced and reproduced by scholars in their local practice domains. However, according to Ylijoki and Ursin (2013), the narrative of loss and success, and correspondingly the identity as a loser and winner, are mutually exclusive. In their study, losses refer, for example, to situations where one's disciplinary field is marginalized, or work conditions deteriorate. Success is typical in the fields that have been chosen as strategic areas of a university, which are highly valued by society and where one's academic standing is strong, which 'benefits but often also determines and shapes the rules of the game ' (p. 1143). This gives a kind of immunity to reforms.
The present study aims to broaden the existing knowledge by looking at academics' agency from a longitudinal perspective and in a specific context of academic work, that is, work around curriculum. In relation to previous studies on identity narratives, curriculum as a context may enable a more nuanced perspective on the complex dynamics of agency in a higher education setting. Curriculum, as a key means of putting the idea of university into practice (Barnett, 2009), has become one of a university's core elements across disciplines and faculties. Due to its strategic importance, taking part in the curriculum development process is often part of an academic's duties. There are, consequently, at least two things that make curriculum development a particularly valuable and interesting context in this study. First, as it often includes debates and negotiations on good teaching, it very often touches on academics' personal values and identity formation. This makes it an interesting arena for agency negotiations. Second, there is usually only a very limited possibility of staying outside the process. There can either be rather straightforward instructions for participation or cultural norms make people take part in the process.
Curriculum change is implemented on a regular basis in universities, but the contexts differ. We investigate the same academics twice: first, during an intermission of curriculum change preceding strong disciplinary autonomy, and second, during a comprehensive curriculum change concerning the whole university. We approach curriculum change as an interactive, social process with various intentions, interests and dynamics (Pinar, 2004, pp. 185-187). Our objective is to explore the elements behind the stability and changes in academics' agency in different processes of curriculum change. In order to study closely the relations between different elements and shifts, we apply Bourdieu's concepts of field, capital and habitus as conceptual tools. Both field and habitus comprise an array of relations consisting of historical, mental and perceptual qualities and are anchored in certain forms of capital (Bourdieu & Wacquant, 1992, p. 16). With these tools, we also wish to shed some light on the value of curriculum work in the field of academia.
Agency, structure and Bourdieu's field theory Bourdieu and Wacquant (1992, p. 15) suggest that when doing research on social reality and social change, instead of looking at dualistic 'things' like the collective or the individual, structure or agent, system or actor, we should look at relations. Bourdieu's field theory, especially the notion of habitus; Giddens's (1991) theory of structuration; and Archer's (2003) realist social theory and notions of agency and structure all aim to cross the divide between micro and macro views of agency and give space to the local and the specific (Ashwin, 2009;Fuchs, 2001). Thus, agency in curriculum work is approached here as a negotiation process and relationship within different structures that constrain or enable agency (Ashwin, 2009). Agency refers to the projects of human agents and capacity to act, often combined with the contingencies of the environment within which such action occurs (Archer, 2003, p. 2;Ashwin, 2009, p. 22;Biesta & Tedder, 2007;Priestley et al., 2012). Akram and Hogan (2015) suggested that with Bourdieu's notion of habitus it is possible to achieve a more nuanced approach to agency, especially why it is difficult but not impossible to change our practices and why people are not able to control the process and rather prefer the status quo.
The next section of the article focuses on the basic principles of Bourdieu's field theory. The idea is to provide the reader with a contextual framework that is used in the analysis. In addition, the Bourdieusian framework is used to illustrate and characterize the specific features and mechanisms of the university as a context for the study later in the analysis. The key concepts in Bourdieu's field theory are field, capital and habitus.
According to Bourdieu (1993), a field is a network of objective relations between positions occupied by agents or institutions. For example, a single university or national or global higher education system is considered a field. Fields are relatively autonomous, objective structures with their own rules, regularities and forms of authority that have no value in other fields (Bourdieu & Wacquant, 1992, p. 97). All fields include the probability of rewards, gains, profits or sanctions (Bourdieu & Wacquant, 1992, p. 18). Fields are characterized by struggles between agents for control over power in the field (Bourdieu & Wacquant, 1992, p. 98).
Bourdieu compares social activity in the field to a game where different players compete in order to maintain and develop different types of capital that has power and value in the field (Bourdieu, 1990, p. 66;Wilkinson, 2010). The challenging part is that the rules of the game are constantly changing. For example, in the field of academia, teachers and researchers are 'players' who try to figure out the most useful moves to maximize their academic freedom and manoeuvring space, make their work meaningful but simultaneously meet what is valued (e.g. in publications, mobility, teaching). Consequently, they expect all these to offer them more prestige in the field. From the perspective of the present study, a topical question is whether investment in curriculum work is valued and whether it adds any forms of capital. Success in the field requires either possessing capital that is valued in the field or identifying new forms of capital that are valuable (Bourdieu, 1993, p. 72).
The three main forms of capital for which actors in a field struggle are cultural, economic and social. Cultural capital refers to knowledge, competences, skills and educational qualifications; economic capital to positions, possessions and wealth; and social capital to networks and ways of knowing social rules and practices (Bourdieu, 1986, p. 243). All three forms of capital may develop into symbolic capital, which entails prestige and renown. This is the most valuable form of capital, as it is often convertible to other forms of capital (Bourdieu, 2013, p. 179). The different forms of capital may be gained through inheritance, education or acting and struggling in the field. With regard to study on the academic field in France, Bourdieu (1988) also used the concepts of intellectual and academic capital as opposing qualities. Intellectual capital is primarily gained based on prestige in the scientific field, whereas academic capital is acquired within universities in senior management positions (Bourdieu, 1988).
With habitus, Bourdieu (2013, p. 72) refers to systems of durable dispositions that social agents develop through their past or new experiences. It is about how actors experience and 'have a feel for' the game, as a sense and meaning of the game which directs the interest and engagement for those who take part (Bourdieu, 1990, p. 66). Habitus can contribute to the reproduction or transformation of the fields. Habitus developed through past experiences is 'structured structure', a product of history that 'tend[s] to guarantee the "correctness" of practices and their constancy over time' (Bourdieu, 1990, p. 54). Ashwin (2009, p. 108) characterized this with 'people like us' who have a shared understanding of what to do and what not to do, with a tendency to reinforce the past practices instead of challenging them. Habitus may also develop through new experiences as 'structuring structure', bringing about an integration and transformation as an 'art of inventing' (Bourdieu, 1990, pp. 53-61). The practices produced by habitus enable agents to cope with unforeseen and ever-changing situations (Bourdieu, 2013, p. 72).
Extended from Bourdieu's work, habitus has been interpreted as not only an individual concept but also an organizational and institutional one (Ashwin, 2009, p. 112;Bourdieu, 1990, p. 60;Byrd, 2019). The university as a field is a place to compete for positions and status but also between faculties, departments and disciplinary fields and subjects (Bourdieu, 1988;Byrd, 2019). Subjects', departments', faculties' and universities' institutional habitus is shaped by their position in the field. Power struggles are not only within the field but also in existing and emerging fields (Deer, 2003).
Bourdieu states that 'In order for a field to function, there have to be stakes and people prepared to play the game, endowed with the habitus that implies knowledge and recognition of the immanent laws of the field, the stakes, and so on' (Bourdieu, 1993, p. 72). However, Bourdieu (1993, p. 76) emphasizes that playing the game and struggling for capital is usually an unconscious activity. To access the field, newcomers must adopt a specific habitus and accept the rules of the game. Those who have the power aim to maintain the existing order and capital in the field, whereas newcomers or those least endowed with capital are inclined towards subversion strategies. This phenomenon could be explained as an unconscious struggle in the field in which new and old, subversion and conservation, and heretic and orthodox are competing for power relations, interests, benefits and the determination of what kind of capital is valuable and desirable in the field (Bourdieu, 1993, p. 73).
Bourdieu's field theory has been used extensively in analysing university education, most often from the macro perspective. Recently, it has been used to study curriculum policy implementation (Dirk & Gelderblom, 2017), academic voice in academic governance (Rowlands, 2018) and institutional change (Kloot, 2009). Deer (2003 claims that Bourdieu's framework, in its early stages, fits with analysing the reproductive social processes more than the transformative, especially if habitus is understood as inherited during the various stages of socialization. Bourdieu eventually revisited his understanding that social change is the result of changing relationships in the field, modified with forms of old and new capital. Recently, Ferrare and Apple (2015) developed Bourdieu's framework from the macro perspective towards analysing the experiences of educational actors and the affordances in the local field. In the present study, we also apply Bourdieu's field theory from a micro perspective to have a more nuanced understanding of how actors construct, experience and struggle over meanings in different curriculum change processes, as we aim to explore the game from the perspective of the 'players'. The present study offers a new context to apply Bourdieu's notions through examining curriculum work as a specific field of struggle and from the point of view of academics. The research questions are as follows: What kind of stability and changes in academics' agency can be identified in curriculum change? How could the academics' stable and changing agency be described in terms of Bourdieu's field theory?

Research design and data collection
Despite joining the harmonization of degrees in Europe through the Bologna Process, in Finnish research universities, the subjects and departments have maintained autonomy in developing their curricula in the way they consider relevant (Välimaa et al., 2007). 1 However, during the last decade, several universities in Finland have experienced university-wide curriculum reforms, following the global trend towards interdisciplinary and competency-based programmes (e.g. Karseth & Solbregge, 2016;Millar, 2016). A local curriculum change concerning a subject or department has a different size, scope and context compared to university-wide reforms, which indicates that the academics' agency differs as well.
The data collection started as part of a broader curriculum research project concerning one university in Finland. The research site is a multidisciplinary research university with about 15,000 students and 2,000 staff members. During the first data collection, 40 departments in six faculties had recently changed their curricula. At that time, the sizes of the departments and the number of members who joined the curriculum work varied, as well as the manner in which it was implemented. Most of the departments had a curriculum development group. Some had collaborative practices and had already accomplished major changes, and in some, one professor designed the subject-based curriculum. The guidelines from the university did not have a significant role, which was typical at that time in the university. In order to study curriculum and curriculum practices, we asked the departments to name an interviewee from their curriculum development group, resulting in 27 interviewees from six faculties.
Soon after the first interview round, without anticipating it and not associated with the curriculum research project, the new university management launched a comprehensive educational and organizational reform. As part of the change, nine schools replaced the faculties and departments. The number of degree programmes was reduced by roughly half, from 52 to 27, causing major changes in curriculum design and practice. It was expected that the old main subjects would find new partners to create a shared degree programme, but it was not clear how and where the decisive moves occurred. After formal decisions about the number and titles of degree programmes were made, curriculum work was supervised and coordinated by the university. The time reserved for curriculum change was about 10 months.
The key guideline was to create broad degree programmes with competence-based curricula. University management encouraged participation by students, researchers and staff members from different career stages. Guidelines were given by the teaching council, which had representatives from all the schools. Yet the schools created their own processes and practices for curriculum work and had the autonomy to make the final decisions about curricula. As such, it could be described as bricolage curriculum change (Louvel, 2013).
After the university-wide curriculum change and 3 years from the first interview, we took the chance to repeat the interviews, as we were interested in how academics experience curriculum work in two different processes. We reached only 17 of those originally interviewed for another interview. The data for the present study comprised 34 interviews with 17 people. Both semistructured interviews had the same themes concerning the practices, processes and experiences in the curriculum work in which they had been involved. Following Rowlands (2018), we agree that informants (agents) cannot meaningfully be asked to describe their habitus or capital during an interview, but through analysing their descriptions of their experiences in curriculum work processes, these elements become clear. The interviews were audiotaped and transcribed verbatim.
All 17 informants were involved in curriculum change in their curriculum development groups. Some of their positions changed between the interviews (for example, from a chair to a member or from a university lecturer to a professor). At the time of the later interview, there were nine professors; seven research associates, senior lecturers and university teachers; and one professional staff member with teaching duties. The informants had an average of 15 years of academic experience. Ten of them had participated in pedagogical training. Ten were men and seven were women. All participants gave informed consent to participate, and their anonymity was protected.
The informants came from seven schools and represent various professional and generic disciplinary fields, including the social sciences, humanities, natural sciences and arts. To protect the anonymity of the informants, while presenting the results, we characterize only the general nature of their disciplinary field. Following the disciplinary categories introduced by Biglan (1973), 2 most of the informants were from soft/pure and soft/applied fields. Looking the pure and applied fields from the perspective of knowledge in curriculum, Bernstein (2003) calls the pure, established disciplines, with their own intellectual texts, practices and rules of entry, singulars, which are 'narcissistic, oriented to their own development rather than to applications outside themselves' (p. 135). The applied fields, called regions, recontextualize the discipline into larger units, operating in both the intellectual field of discipline and the field of practice (p. 135). Bernstein suggests that, for example, mathematics, history, economics and sociology are singulars, whereas engineering, education, medicine and management are regions. However, this categorization is not stable, as different disciplines may have strong or weak boundaries. Which disciplines enter a region, according to Bernstein (2003, p. 135), 'depends upon the recontextualizing principle and its social basis'. This is especially topical in a curriculum change where the aim is to create broad degree programmes.

Analysis of the data
This study is a follow-up to our earlier study, which aimed to identify academics' agency and the structures that enable or impede agency in two different processes of curriculum change at one site (Annala et al., under review). Based on thematic analysis (Braun & Clarke, 2006), we identified six agentic features, of which five, named as progressive, oppositional, territorial, bridge-building and accommodating agency, occur despite the size and scope of curriculum change. Powerless agency emerged only in the context of university-wide change. We identified individual, community and institutional structures that enable or impede agency. Even though five agentic features appeared in both contexts, they were not stable with regard to all individual academics. This was the rationale to continue the analysis to look at the data as individual accounts.
For the present study, we looked at the data via a narrative approach in order to expose participants' individual experience trajectories. As many narrative researchers (e.g. Bruner, 1990;Clandinin & Connelly, 2000) have claimed, at its best, the narrative approach can be a constructive tool that offers alternative ways to examine issues that are otherwise inaccessible using other qualitative methodologies. This method of choice thus was adapted to a narrative approach as a way of thinking about experience (Bruner, 1990;Connelly & Clandinin, 2006) and as a methodological response to the positivist paradigms (e.g. Lieblich et al., 1998). Bruner (1990) defines narrative as a system of understanding by which people organize their experiences in, knowledge about and transactions with their social worlds. Therefore, the narrative approach is here seen as a method to understand the complexity of the processes by which people interact, how they make sense of their experiences and how they structure them.
The narratives were perceived as individual storyline trajectories appearing through the interviews conducted in two contexts of curriculum change and were thus the results of confluences of social influence on participants' inner lives, social influence on their academic environments and their unique personal employment histories. We condensed an individual synopsis via focusing on the general stability and changes around the most dominant agentic features and in relation to two different contexts of curriculum reform. The narratives represent the most distinct features in individual agency as ideal types, referring to a mental construct or model to characterize and understand social reality (Swedberg, 2018). The analysis resulted in two storylines with six narratives (Table 1).
In the last stage of the analysis, the storylines with changing and stable agency were reflected with theoretical lenses from Bourdieu's field theory. In the following, we will report in more detail the narratives with changes in agency, where 'playing the game' and 'feel for the game' are prominent. Then we move on to present a concise account of the narratives with stable agency. The names used are pseudonyms. In parentheses, code I indicates the first interview and code II the second interview.

The transformative narrative
Peter and John's transformative narratives represent successful 'playing' in the field even though the emphases in their agency transformed in opposite directions: Peter from territorial towards progressive agency, and John from progressive towards territorial agency. Progressive agency refers here to development-oriented curriculum work with a focus on student learning and being open to implementing new ideas coming from outside, whereas territorial agency protects the full autonomy in academic and curriculum work (Annala et al., under review).
Peter had a leading role in both curriculum changes. For local curriculum change, Peter's expressions reflect an interest in protecting the boundaries of the discipline. He highlighted the discipline's non-professional nature as an established field of science that does not need to change. Following Bernstein (2003), it could be called singular. Peter was aware of interdisciplinary degree programmes to be implemented somewhere but expressed that ideas of this kind do not concern them. He admitted that they had some problems with heterogenous student quality and teachers' motivation to teach basic courses, but he did not see much value in curriculum work: I do not believe that any changes in curriculum work improve curriculum; the best way is as it is now: it is made within subjects and one person is responsible for it and discusses it with others. [. . .] I think a broad degree programme is a bad idea. (Peter, I) After three years, in the university-wide curriculum change, Peter's agency had changed. Instead of reproduction, he contributed to the transformation: We were the ones who made initiatives in different directions and got a very positive response straight away, and they have also wished for cooperation but not brought it about [. . .] this was the moment to take the first step, because during this reform, the transdisciplinary and thematic approach was encouraged. Peter spoke of 'we', which refers to a more collaborative mindset that may also be considered as a new form of social and cultural capital in curriculum work. The younger, 'enthusiastic' colleagues may have brought new competences and social practices to the field, and when changing the curriculum in collaboration, they benefited from the social, cultural and economic capital they mutually possessed. Thus, the practices produced by habitus contribute to the transformation of the field (Bourdieu, 2013). The university strategy encouraged new kinds of institutional working and communication cultures to emerge and supported forums for inter-and transdisciplinary discussions. Peter had understood that all the schools are making huge changes, which may provide an understanding that the habitus of the whole university is changing, as 'this is what we do' (cf. Ashwin, 2009, p. 108). Peter experienced that the curriculum was better outlined than ever and had better connection to research. He had some hope of finding more 'fit' students with the help of the more transparent curriculum and its implementation. Several cross-disciplinary themes based on shared research interests were introduced as part of the curriculum process, which focused the curriculum development more on intergrating research and teaching. Contrary to Peter's presumption in the first interview, the university-wide change raised the profile and status of this subject. This means that other subjects in the same broad degree programme might have the opposite situation.
Whereas Peter's narrative moved from territorial towards progressive agency, John's narrative did the opposite. For local curriculum change, John had a leading role in curriculum work. In their department, they had already implemented a merger of subject-based curricula towards regionalized knowledge (cf. Bernstein, 2003). In that context, John featured progressive agency. He seemed to possess economic, social and cultural capital (Bourdieu, 1986) and emerged into a respected position in the department as a professor, leading the curriculum practices consciously, for example, with an enterprise to move from individual work practices towards team-based curricula and teaching. He was in a position to have an impact on the changes, and there were enough people enthusiastic about new ideas, but as in every reform, there were those who 'act as a brake on changes, which is an important role, too' (John, I).
In university-wide curriculum change, John no longer had the leading role. As his formal position in the field changed, this gave him more room to be critical. He saw the new change as a continuation of what they had already started in the previous change but was worried about the general trend to separate research groups and degree programmes. The emphasis in his agency shifted, and his views reflected a more inward-looking approach and individual perspective, as the priority was to unite his research interests and teaching: Within the degree programme, we have a distribution of work according to our specialities. Through that we find our identity, which is connected to the research group and teaching related to it. [. . .] From my personal perspective, the curriculum was changed so that now I can build my personal, cumulative teaching path through the degree; I can focus on my speciality from basic studies to an advanced level. (John, II) Even though the disciplinary fields of Peter and John were different, one having strong and the other weak disciplinary boundaries, and even though they had different emphases and changes in their agency, their accounts represent success in terms of achieving intellectual capital and prestige. In Peter's narrative, the curriculum change was taken as an opportunity to increase the subject's institutional habitus in the school's field and, through that, to gain more economic and symbolic capital. John benefited from the change by gaining more intellectual capital through aligning teaching and research activities. Bringing research and teaching together (research-teaching nexus; see, e.g. Brew, 2003) in curriculum change was capital that engaged both scholars in the process, but it is also capital that academic institutions value (cf. Mathieson, 2019). Here, their habitus emerged as 'structuring structure' (Bourdieu, 1990) and as elastic identity (Shams, 2019)-trying to make sense of the reform in a way that would be meaningful for them as scholars and focusing on what is thinkable and doable in their local communities.

The sidelined narrative
Susan and Kate were recognized as having progressive agency in local curriculum change, which changed to powerless agency in university-wide curriculum change. Powerless agency refers to an interest in contributing to the change, similar to progressive agency, but there is no access or space to carry out agency (Annala et al., under review). Susan and Kate represent sidelined narratives because the capital that was regarded as valuable and which they possessed in the first context did not have the same value in the second one. During the local curriculum change, Susan and Kate, though not professors, had a respected role in curriculum change: The professor leads, but we have had very intensive discussions together and are all engaged in the process. (Kate, I) As in John's department, similarly to Susan's department, when they understood that there was no future or resources for small programmes, they wanted to be the forerunners and had already merged old subject-based majors into a degree programme. They coped with the changes by having active agency and dialogue with students, researchers and teachers and ensuring that when implementing the curriculum, the most valuable pedagogical aspects remained, and the change was implemented according to their values.
In university-wide curriculum change, Susan took an active role but experienced that she was not heard or understood. She felt that their experiences of the recent curriculum change were disregarded, indicating an unequal distribution of power: Now we just go forward on the conditions of the masses and the powerful, structures that you have to adjust, structures you have to fill [. . .] when numbers are bigger, so it follows these types of pedagogical solutions that ruin everything we have thought to be of value. (Susan, II) University-wide curriculum change preferred big units, followed with big intakes and large groups. Both Susan and Kate experienced that in their schools, the technical-rational processes and 'organizational talk' defeated the discussions of the quality of curriculum change. There was a clash between dialogical and hierarchical organization cultures. As they tried to be active, they felt that they were not in a position to have an influence on the process. Their experience was that the status and power relations defeated competency and expertise.
Yet the formal position did not guarantee capital or success in the process, as the narratives of three professors, Eric, Tina and Alma, show. In the first curriculum change, Eric positioned himself as a critical person in the opposition. He pointed out the individual-and sector-specific working culture inhibiting real curriculum change and development, and the problem when no one really took a strong leadership role for more holistic curriculum work. At the end of the first interview, he stated that 'this interview stimulated me to think that we should renew everything' (Eric, I). However, in the second interview, his agency moved from opposition towards powerlessness. Because he was not personally addressed, he experienced that he was silenced: 'I was not invited to the meetings' (Eric, II). Yet he admitted that taking a marginal role was partly his own choice. As a result, the speciality he was responsible for was not included in the new curriculum.
The same occurred with the specialities of Tina and Alma, although they took proactive approaches to work on the new curriculum. The subject specialities they were responsible for were marginalized in the curriculum change in their schools: Tina and Alma did not have the roles they wanted and instead 'had to agree with a compromise that was not what I was wishing for' (Alma, II). In the sidelined narrative, the subject's-here including both singulars and regions-institutional habitus was downgraded in the school's field and either combined with others, marginalized or closed. The access to curriculum work was open and could be interpreted as an enterprise to broaden democracy, but the location of power was hidden. The democracy appeared to have hidden rules for its information flows, and decisions were made through building alliances with the right people. The 'right' kind of habitus entails access to the shared cultural, academic and linguistic resources that support development and give access to powerful knowledge (see Clegg, 2016, p. 459).
Those who had formal positions, like Eric, realized that the old rules of using power by just having the habitus of professor were not enough in this new game. Yet those who were proactive, like Susan, Kate, Alma and Tina, did not have a 'fit' habitus to have agency in the process. When thinking of habitus as 'structuring structure' (Bourdieu, 1990), new experiences can contradict values. These individuals' habitus did not fit the logic of the new game, and they lost the 'feel for the game'. Their old local communities had supported dialogue, diversity and collegial decision-making, but symbolic capital of this kind did not work in the new contexts of curriculum change. The implicit rules of the game seemed to favour particular forms of capital (economic, competitive, hierarchical) compared to others (social, cultural, dialogical).
The question of gender may also be one hidden rule in the game even though for Bourdieu, the gender perspective is hidden under other types of categories (Skeggs, 2004). Susan, Kate, Tina and Alma were all recognized as student-centred and pedagogically oriented female academics, experience being sidelined. The gendered power relations emerge when women in academia are positioned or position themselves as 'more for the others' focusing on caring responsibilities (of students, teaching, curriculum work) than as 'more for themselves', prioritizing competition for their subject or career (McNay, 1999). In addition, Deem and Lucas (2007) state that women are more likely than men to identify with a holistic approach to academic work. Gender division inside the academy cannot easily be reshaped, as 'certain conventional arrangements of gender have not necessarily been dismantled and indeed may have become more entrenched' (McNay, 1999, p. 103). Habitus could disrupt the routine adjustment of subjective and objective structures and change what is valued in the field (Adkins, 2004), but in this case, in the sidelined narratives, there emerged a lack of fit between habitus and field.

The divided narrative
Paul, Roy and Lisa featured progressive agency in the local curriculum change, but in the universitywide change, the emphasis moved towards bridge-building agency. This refers to balancing, compromising and mediating between different interest groups (Annala et al., under review). Paul had a leading role in curriculum work in both contexts and Roy in the latter context. In the local curriculum change, both Paul and Roy had ambitious ideas for developing curriculum and, through that, making the subject's profile better than before. Lisa saw the heterogenous views among academics and a multidisciplinary approach in light of the potential for curriculum development as a continuous process. One idea was to see the existing tensions between theory and practice as a pedagogically productive phenomenon: If we think of curriculum as a broad process, it has a cultural power. I mean, like the expectation that we are producing professional qualification [. . .] in our subject, we resist this thinking because we cannot produce qualifications, but we can provide preconditions that will strengthen students' own identification. (Roy, I) In the university-wide change, they all declined their pedagogical development ambitions, expectations and wishes. Both Paul and Roy had contradictory 'feels for the game' because they experienced that the focus had moved to secondary issues and to more personal struggles in the field: This structural principle, that different subjects are put together, it blurs the focus of pedagogical thinking, because then our energy is used for arguing, bargaining, sometimes raw force, power; it was needed to make decisions, and it takes energy from thinking of curriculum as a pedagogical entity. (Paul, II) We had for the first time a possibility to create something new, to cross borders, but we already knew that it is so difficult, and proved to be difficult and impossible, because the stricter the advice, the more people started to defend their territory. (Roy, II) Paul was relieved to give away the leading role. Curriculum work seemed to be a new area of leadership in academic work, where the position and power relations were unclearly situated between the local and university-wide interests. Paul and Roy had to adjust their broader intellectual approaches to curriculum work towards academic capital as management and coordination. Lisa, as a member of a curriculum work group, took a stand on leadership in curriculum work, questioning how to unite intellectual capital and academic leadership in curriculum work: Personally, I wished that some people who have really extensive views and so on, I wished they had contributed, but they focused on something else-some really had research leave and so. If I were in charge, I'd never give any leave during a year like this. (Lisa, II) Lisa's account shows how academics, even when they have interest, prefer to exit curriculum work, as it does not include similar value or prestige to research tasks. According to the narratives of Paul, Roy and Lisa, in university-wide change, instead of aiming towards the best possible curriculum to support student learning, the focus moved towards making compromises between the scholars and adjusting to the structures and already existing power relations framing the process. Ylijoki and Ursin (2013) and Rowlands (2018) identified positions between the extremes as an existence of blurred and in-between spaces, especially when academics needed to struggle with a combination of teaching, research and service tasks. Those representing divided narratives needed to adjust their intellectual capital towards academic capital (Bourdieu, 1988). These changes in agency could be described with Bourdieu's notion of the divided habitus, which refers to 'a habitus divided against itself, in constant negotiation with itself and with its ambivalence, and therefore doomed to a kind of duplication' (Bourdieu, 1999, p. 511).

The development-oriented narrative
The development-oriented narrative of Eva and Sam features progressive agency in both curriculum change contexts. They had key responsibilities in curriculum work, as they were academics recognized as having pedagogical knowledge. The pedagogical knowledge and broad understanding of curriculum could be considered cultural capital that was also valued by their local communities (cf. Bourdieu, 1986). The narratives of Eva and Sam imply that in order to make a coherent curriculum, enhancing the collaborative work culture is important. Both Eva and Sam represent disciplinary fields that were established singulars, and they told about the history of gradual change in their work culture: We have had an enormous change during the last ten years, I mean radical. [. . .] it used to be very split, every teacher took care of their own sector and own courses and the professors said the last word that the structure was either good or not [. . .] now it is dialogic, and because of that we have reached good results. (Sam, I) In the university-wide curriculum change, there were tensions, but Eva and Sam both describe the importance of finding a shared language and repertoire (see Wenger, 2003), which has a role in mutual understanding and working towards a shared goal. The conceptual knowledge about curriculum could be seen as cultural capital, which has also followed with social capital in their direct work communities: We all had familiarized ourselves with the issue. We were talking about the same topics with similar concepts. I think in curriculum development groups it may be a problem that we use different concepts. (Eva, II) In addition to collaboration and shared language, leadership was raised as one element for sustaining the progressive agency in the community. In Sam's case, the school's dean had a key role as mediator, whereas Eva's narrative indicates the importance of professors as formal leaders to engage in the meetings and a need to prepare the agenda carefully. Similarly, an earlier study showed that academics' agency in change is highly dependent on formal structures and policies for supporting all faculty in pedagogical development, particularly the supportiveness of formal leadership strategies (Clavert et al., 2018).
Both Eva and Sam reflect on the challenges they had in their curriculum work but believe in gradual change towards a better curriculum and education. It seems that the work culture in the disciplinary community and the agents' professional values were aligned with each other and with the organizational initiative. This reflects the relationship between habitus and the field, as if habitus had contributed to the transformation of the fields (cf. Bourdieu, 1990). Because of this, there was no need to use energy for questioning, but academics were more able to focus on the development in the long run and as a part of broader structures than before.
The narratives of Eva and Sam indicate that they did not regard the curriculum change as easy but coped with it by promoting a dialogic atmosphere and being creative. It may be interpreted that Sam and Eva possess social and cultural capital and represent a new type of academic that unites intellectual and academic capital (cf. Bourdieu, 1988) in their work. In their case, this seemed to be valued in their schools and disciplinary fields as well.

The autonomous narrative
The autonomous narrative of Mary and Gary represents territorial agency in both curriculum change contexts. In local curriculum change, their disciplinary fields had already undergone significant changes in curriculum, and they saw it as an important tool to develop education in collaboration with working life. However, there was an expectation of doing it autonomously. Especially for university-wide curriculum change, Mary and Gary protected their existing practices and social rules in the field. Mary emphasized that curriculum work is a continuous process that cannot follow the schedules or guidelines given for the whole university. Even though Mary's and Gary's disciplinary fields did not have similar status, they were able to make their own rules for the game: This reform was necessary to some other fields, but not for us. Somehow, the goals were wrong or maybe already in use. For example, this competence-based thinking or whatever it was called, it sounded absurd because I feel we have followed it already for 20 years. (Gary, II) Curriculum work is extremely important, and I think it is great that it is done now in a new way in the degree programmes, that is OK [. . .] but we hold on with our own frames [. . .] so often, I need to say that this does not concern us because our education is different. (Mary, II) These disciplinary fields followed the university-wide process but pursued changes that seemed more superficial (see Louvel, 2013). For example, by joining the formal meetings, they were able to show that they were 'in the game', but did not do much, as they were already one step ahead of others. Adler and Lalonde (2019) explain that the conformist 'façade' may appear as a strategy for avoiding direct confrontation with the institution and maintaining mental distance to protect academics from potential threats to their identity. Playing the game may appear as a way to protect the moral foundations of the academic profession (Adler & Lalonde, 2019).
The autonomous narrative features strong disciplinary status and possession of symbolic capital, which is different from Susan's field. She also experienced that they were forerunners, but experienced powerlessness in the university-wide reform. This indicates how unequal people, subjects and disciplinary fields are in the reforms. Gary's and Mary's attempts to hold on to disciplinary territorial borders seems to have a connection with the prestige they and their subjects had in their schools and in the university-they were the only hard/applied fields in the present data. The individual and institutional habitus had power over the university-wide initiatives, which was also a feature identified by Ylijoki and Ursin (2013) in their study where some people, research groups or fields of science had such a valued position that they could ignore or even shape the rules of the game.

The opposition narrative
Oppositional agency approaches curriculum work as valueless in the academy and as a possible threat (Annala et al., under review). The opposition narratives of Jack, Mark and Fred represent resistance in both curriculum changes but feature inconsistent talk. In the first interview, Jack was not happy with the last curriculum reform, in which programmes were split into subject-based minors. Small subjects had small resources and threats of closing. In university-wide change, the subjects merged back into new degree programmes. Jack's subject achieved a stronger position than before, but the change in its institutional habitus did not change Jack's agency. He was critical about the merger: There were so many meetings that if you want, you can avoid real work [. . .] teaching doesn't give any boost; I have been teaching for 30 years, and it is as routine as it can be. (Jack, II) For Mark and Jack, curriculum development was not 'real' and profitable work. For Fred, it seemed more valuable, but like Jack, his talk was inconsistent. In the first interview, he talks about how to pay attention to the idea of democracy from the point of view of students, but in the second interview he opposes the reform because of how it implemented democracy: Professors used to have a certain responsibility included in their position and traditional university functions through professors. [. . .] now there is the ideal of democracy that all can join and have the freedom to speech [. . .] as if all had similar experience and responsibility, which is not true. (Fred, II) Academic work is often connected to enthusiasm or passion towards research or teaching, and it can vary from time to time (Shams, 2019;Ylijoki & Ursin, 2013). Jack, Mark and Fred seemed to lack habitus of this kind, as they showed passive resistance and seemed to look at the academic game from a distance. Following Bourdieu (1993), the habitus is not a conscious decision to resist, but the opposition agency may manifest itself when there are no other forms available to use autonomy and power. The game may have changed so much from the old days that they do not see it as worth playing. This could be called academic nostalgia, an idealization of the past that functions to describe the problems and tensions of the present and to mediate coping (Ylijoki, 2005). Ylijoki (2005) suggests that when organizational change is glorified, nostalgia can also be seen as a positive and progressive force to keep core morals and values alive. However, in this data, this did not appear. Instead, it is similar to a 'victimized plot', connected with passivity, as found by Juntrasook et al. (2013), and a 'narrative of loss', described by Ylijoki and Ursin (2003), where the end result always seems very dark. A combination of an idealized past, fears and mistrust in the present, and the absence of a shared reflexive capability within the academic community may create a defensive stance that limits agency (Adler & Lalonde, 2019). Habitus appears as 'structured structure', as developed through past experiences, and there is no room for making sense of new experiences or building new meanings ('structuring structure') (cf. Bourdieu, 1990).

Discussion
Curriculum work is an arena for struggles that has a role not only for individual academics but also for the position of disciplines and fields of science in academia. The present study shows how academics as players in the field of curriculum work struggle for different types of capital and strive for a habitus that has power in the game. The lived experiences create habitus as it is, as the internationalization of social structures and as enterprises to maintain old or develop new forms of capital. Power structures became visible when all did not have real access to join the game, and some withdrew from it. The results raise a question: who can legitimately become an agent in the curriculum process, and what qualities make an academic 'fit' to curriculum work?
The lived experiences create habitus as it is, as actors 'act according to a social logic that directs them towards particular social destinies already present through their own historical and cultural background' (Deem & Lucas, 2007). Curriculum work is a field in which the struggle for disciplinary status and position affects both individual and institutional habitus: they remain, improve or weaken from earlier, and in relation to the new faculties and institution as a whole. This creates polarization among academics and disciplinary fields.
For some, the strategic guidelines and structural frames work as an asset and backup, and they manage to implement changes in curriculum and improve their position in the field, whereas for some other subjects and disciplinary fields, the consequences were the opposite. The capital that was valuable in university-wide change favoured more the functional, hierarchical and hard qualities than the dialogical, inclusive and pedagogical. These rules were hidden and new for some of the academics. Institutional habitus was on the move and found its place between the old and new ideas of academic work and traditional disciplinary cultures, between the struggles of heresy and orthodoxy. In particular, those who had intellectually and pedagogically ambitious visions of curriculum change failed in the struggles in the field. Those who had established positions and status or possessed a range of different forms of capital in use, partly because of a collaborative work approach, seemed to be successful.
The stability and moves in agency emerged through the lived experiences, successes and losses in the game in the field of higher education. Success, when defined as maintaining or achieving valuable capital, was achieved through both stable and changing agency. In addition, various ways of carrying out agency could lead to a successful result from the perspective of academics and their disciplinary fields. Stable agency is portrayed as either possessing symbolic capital that is valid despite the changes (autonomous narrative) or gradually bringing new forms of capital into the community (development-oriented narrative). Those were a result of a long-term process in the field and show that curriculum development is possible when abiding by one's own principles and traditions.
In the social reality of the university, the enhancement of the research-teaching nexus and shared conceptual tools are key for meaningful processes and engaging in curriculum work, and these could be capital valued both by the university and scholars. This implies that curriculum change would benefit from framing it more holistically, not only as teaching-related but also as a research-related activity and approaching it with more theoretical understanding. If the local field or 'community of practice' with mutual interaction and shared enterprise (e.g. Wenger, 2003) has space for different voices, shared repertoire and flexibility in its practices, it gives room for meaningful agentic projects in curriculum work. This requires curriculum leadership, which appears topical in many of the narratives. It is a new area of academic work that needs more elaboration in future studies.

Notes
1. In Finland, research universities and the labour market consider the master's degree as the 'basic degree', and the BA is more like a stage in one's studies (Välimaa et al., 2007). It is common that students are admitted to study BA and MA degrees at the same time. Thus, curriculum changes concern both degrees simultaneously. 2. Biglan's (1973) classification of academic subject matter by its nature includes hard and pure (H/P), hard and applied (H/A), soft and pure (S/P), and soft and applied (S/A) sciences, characterizing the difference in their paradigms and applicability.