Understanding academic agency in curriculum change in higher education

ABSTRACT This study explores the agency of academics and the structures that enable or impede agency in curriculum change. The study was conducted at a research university that experienced curriculum change in two waves: the first on a departmental level, and the second as a university-wide curriculum change. Interviews were conducted with the same academics after both waves. Thematic analysis generated six forms of agency. Progressive, oppositional, territorial, bridge-building, and accommodating agency appeared in both contexts. Powerless agency was identified in only the university-wide curriculum change. Individual, community, and institutional structures enabled or impeded agency. Instead of focusing on the micro or macro levels exclusively, this study highlights the interwovenness of structural-agentic processes, including the critical role of social cultures and relationships, and reveal how academic freedom is used in many ways as room for manoeuvring in curriculum changes.


Introduction
Educational and curriculum practices are being reconsidered in universities worldwide (Karseth and Solbregge 2016;Shay 2015;Yates et al. 2017). In policy papers and curriculum reforms, competency-based and interdisciplinary curricula are regarded as more forward-looking and innovative in addressing major societal problems and needs than traditional disciplineor subject-based curricula, despite the challenges of academic implementation (Caspersen, Frølich, and Muller 2017;Gantogtokh and Quinlan 2017;Millar 2016). Engagement in curriculum change may present challenges for both academics and their communities, given that universities can be characterised as collections of relatively non-hierarchical networks that resist strong top-down control and seek meaningful justifications for changes (Barman et al. 2016;Broström, Feldmann, and Kaulio 2019).
Besides top-down reforms, curriculum change is a staple of academic work in departments and faculties. As curriculum is a key element in defining research-based higher education for future generations, it is important to understand academics' role in curriculum change. This study focuses on the structural-agentic processes in curriculum change and contributes to a research area that aims to understand the wider dynamics of curriculum development.
Previous research on the response of academics to educational or organisational changes has focused on the relationships between the macro -(policy) and micro -(actors) structures in curriculum change. For example, in a study on the organisational effects of globalised policy-making and the Bologna Process 1 on curriculum change, Brøgger (2014, 537) found that reform 'is not one reality given from the 'outside', but consists of multiple 'reals' that 'trouble' the ongoing reforms'. She used Derrida's neologism, hauntology, to analyse how a history of disciplinary activities can fuel present actions, which Brøgger describes as 'ghost armies' and 'juntas'. She found that academics, as educational agents, did not simply cope as passive receptacles of suggested changes; instead, they negotiated, contested, adopted, and rejected (Brøgger 2014).
The response categories, often identified in the literature as proactive or compliant and passive (e.g. Di Napoli 2014; Oliver 1991), have been criticised as an overly simplistic view of academics' engagement in curriculum change (Louvel 2013). In contrast to an effective engineering type of activity, where a plan is followed to meet the expected optimal solution, Louvel (2013) found curriculum change to be a creative, bricolage type of activity: a regime of activity in which change and innovation are both creative and highly constrained, and where individuals or groups create something new out of the resources they have at hand-while they still respect certain rules or conventions, they don't follow any strictly predetermined plan. (Louvel 2013, 670) Such bricolage activities emerged in Louvel's (2013) study as academics used more or less proactive strategies to contribute to curriculum change, implemented by crafting specialised products, amalgamating existing elements and renewing existing programmes. Louvel (2013, 672) suggests that 'any change situationeven the most constrainedwill leave academics with a certain amount of room for manoeuvre'. This may also appear as a 'window-dressing' type of activity, where nothing really changes (Louvel 2013).
Within all control structures, it is people who essentially design and reform curricula and make choices based on the justifications they consider relevant. Biesta (2015) called these choices teachers' judgements, referring to their ability to make decisions about work and, in this case, curricula. An example of this is Roberts (2015) study of curriculum decisions in the design of a specific course. However, the present study takes a closer look at the academics' 'room for manoeuvre' in broader curriculum changes in higher education.
We approach curriculum from the perspective of curriculum theory as an interactive process with various ideas, intentions, interests, and dynamics, described as a 'complicated conversation' (Pinar 2004, 185-187). Curriculum is connected to different personal, institutional, and societal objectives and power relations that reflect a certain historical context (Annala, Lindén, and Mäkinen 2016). There has been little empirical research on what curriculum as a 'complicated conversation' or as an interactive social process in higher education is. Therefore, we use the concepts of agency and structure as an epistemological distinction to access and construct an understanding of the social processes that are not directly known (cf. Ashwin 2009). Following Ashwin's (2009, 18) approach, we see agency and structure not as 'things' but as processes. Agency in curriculum change can be characterised as a negotiation process or relationship between different structures that constrain or enable agency (cf. Ashwin 2009, 20). By analysing data from two different curriculum change processes, we aim to explore academics' agency and the structures that enable or impede agency in curriculum changes. The research questions are as follows.
(1) How does agency emerge among academics in curriculum change?
(2) How does structure shape agency in academic curriculum change?
(3) What is the relationship between agency and structure in different forms of curriculum change?
Approaching curriculum as a structural-agentic process Practically, curriculum is a platform in which shared understandings of 'educational ideas' are created, developed, and fostered in academic communities (Annala and Mäkinen 2017;Knight 2001). Curriculum provides a discursive and structural framework for negotiations on the principles and practices most suitable for the discipline and community (Barnett, Parry, and Coate 2001;Shay 2015). Consequently, it is important to understand and examine the space, factors, and mechanisms inside curriculum communities and the interaction processes behind negotiations. The analytical and theoretical conceptualisations around agency and structure offer a widely used and versatile framework for this purpose. The concept of agency was popularised by Giddens (1991) in structuration theory and has since been used widely (Eteläpelto et al. 2013). Over the last decade, teacher agency has been studied extensively in the context of curriculum change in schools (e.g. Alvunger 2018; Priestley et al. 2012) and vocational education (e.g. Vähäsantanen et al. 2008). Vähäsantanen et al. (2008) found that teachers are committed to their work organisations if they have enough individual agency and opportunities to practice their orientation towards the profession and if no major changes are imposed on their workplace from the outside. Their study reveals that weak or strong structures create possibilities and constraints for teachers' individual agency. This suggests that organisational, community, and individual levels should be considered when discussing agency.
However, there is limited research concerning agency in curriculum change in the context of higher education. Mathieson's (2012) research on a merger of two universities in South Africa highlights the potentially significant role of academics as agents of change instead of passive recipients or even resisters of external change forces. She identified different academic workgroup cultures that are characterised more by the differences in how academics mediate between disciplines and various contextual factors than by the structural characteristics of disciplinary epistemologies. Barman et al. (2016) examined teacher autonomy, which is often used in confluence with agency, in relation to educational reform; they illustrated how policies become meaningful for health teachers by reshaping local practice.
Previous research has often assumed overly individualistic views of agency, describing agency as one's capacity to act autonomously and shape responses to the environment, or overly socialised views, in which macrostructures determine agency (Biesta and Tedder 2007;Fuchs 2001). We position our approach in the middle, roughly following Priestley et al.'s (2012, 196) definition of agency as 'a matter of [personal] capacity to act, combined with the contingencies of the environment within which such action occurs' (see also Biesta and Tedder 2007). Our approach also borrows from Ashwin (2009, 20), who does not limit agency to individuals and instead acknowledges that groups may have agentic projects.
Agency refers to the projects of human agents, such as thinking, intending, and determining courses of action, that are performed differently in various situations (Archer 2003, 2), being relational in using power (Ashwin 2009, 22). Agents may not be fully aware of the factors that enable or constrain their projects (Ashwin 2009, 20;Archer 2003, 9). Structures refer to the social forms and cultural systems that enable or constrain different projects from groups of agents (Ashwin 2009). The power exercised here is systemic, referring to how agents are positioned (Ashwin 2009, 22). Emergent structural and cultural properties have the generative power to impede or facilitate projects. Archer (2003, 7) emphasises the importance of distinguishing between the existence of these properties and the exercise of their causal powers. She concludes that agents may also use causal powers of constraining and enabling to shape structural and cultural situations and processes (Archer 2003, 4).
Instead of distinguishing between the concepts of agency and structure, Ashwin (2009) uses the term 'structural-agentic processes.' He emphasises that 'structure and agency are not different kinds of processes but different ways of grouping or conceptualising complex social processes' (19). In structural terms, this study focuses on how particular sets of structural-agentic processes become situated in and shape curriculum change. In agentic terms, it focuses on how the dynamic relations between the projects of those involved in curriculum change can be characterised in terms of structuralagentic processes.

Interview data from two different forms of curriculum change
To avoid overly individualistic and socialised views (Fuchs 2001;Priestley et al. 2012), we investigated the same group of academics in two different contexts, representing two different forms and levels of curriculum change. The research data were collected at a medium-sized multidisciplinary research university in Finland, 2 and three years passed between the two interview sessions conducted.
The first data collection was conducted during a situation in which departments and academic subjects had strong autonomy in curriculum making at the university in question. For years, the departments had changed their subjector discipline-based curricula in ways they themselves defined. There was a lot of variation in the curriculum practices. The sizes of the departments varied; some were big, while others had a small number of faculty and students. In the different departments, there was a disparity between the people who had access to curriculum work. For example, in some departments, professors conducted curriculum work with assistants, whereas other departments implemented collaborative curriculum practices.
In this study, we named this first context of data collection departmental curriculum change. We asked each department to select a member of its curriculum development group for an interview. We obtained 27 volunteers and conducted semi-structured, thematic interviews with them concerning their latest curriculum change process. The themes covered the practices, objectives, and significance of curriculum development. The informants were encouraged to share issues they found relevant and topical in the curriculum change process.
The second data collection took place after a comprehensive organisational and curriculum change that concerned the whole university. Six faculties and 40 departments and units were replaced with nine schools. There was a shift from 52 main subjects to 27 degree programmes, for which new curricula were needed. The reform was organised as a strategically led curriculum change towards broader degree programmes. We named this context university-wide curriculum change.
The change was initiated by the university itself. Nevertheless, its elements echoed the contemporary trends in higher education curriculum policy, especially in the European Union (e.g. Handala et al. 2014), which produced interdisciplinary and competencybased degree programmes. The university provided support through lectures and workshops that were conducted at the university and school levels, and guidelines were provided by the teaching council, which had members from all the schools. Many guidelines were based on experiences and suggestions from the schools or curriculum development groups.
Although the reform was strategically led and university-wide, the university management indicated that it did not want to dictate how the educational units organised their curriculum change processes. Instead, it encouraged a participatory approach that could be implemented by academics working in different career phases, including professors, lecturers, researchers, students, and professional staff. The new schools created their own practices for reform and had the autonomy to determine their curricula. Changes initiated in this way could be seen as an attempt to achieve the bricolage type of curriculum reform introduced by Louvel (2013).
After the reform, we conducted a second set of interviews, reaching only 17 of the original informants. The semi-structured interviews had similar themes to the first interviews but focused on the recent university-wide curriculum reform. Overall, the data for the present study comprised 34 interviews with 17 people who were members of the curriculum development groups. The interview data were audio-recorded and transcribed verbatim for analysis.
The informants represented a wide variety of disciplinary fields from seven schools and had 7-30 years of academic work experience. Some of the interviewees changed their posts (e.g. from university lecturer to professor) and positions (e.g. member or chair in the curriculum development group) between the interviews. At the time of the second interview, nine of the informants were professors; seven were senior lecturers, university teachers, and research associates; and one was a professional staff member with teaching duties. All informants were asked for their informed consent to participate in the study, and their anonymity was ensured. Therefore, we do not characterise their disciplinary fields or positions in detail when reporting the results.

Ethical considerations
University management gave permission to study the curriculum changes. The first and third authors collected the first data set and, at that time, were working as researchers in a faculty at the university in question. During the university-wide curriculum change, they acted as experts at the university level, because of having a background in curriculum research. The second data collection and data organisation was conducted by research assistants who were not involved in the changes. The second and fourth authors assisted in reducing insider researcher bias by providing investigator triangulation (Fusch, Fusch, and Ness 2018) during the analysis and writing process.

Thematic analysis process
To organise the data, we applied thematic analysis (Braun and Clarke 2006). First, the data were repeatedly read carefully, and notes were made to identify the basic units for analysis. A basic unit for coding was a sentence or a short notion containing a view expressing agency, i.e. how the informant describes and positions themselves or their group when discussing curriculum change. Next, we coded the expressions of agency identified in each informant's first and second interviews and collected the citations and codes into a comparative table. We reviewed the codes between the informants' interviews and generated themes characterising agency, 'identifying the "essence" of what each theme is about' (Braun and Clarke 2006, 92). Six agentic themes were generated. Five of them appeared in both interview data sets: progressive, oppositional, territorial, bridge-building, and accommodating agency. From the second interview data, we identified one new theme: powerless agency. We aimed to avoid simplifications, such that one informant would be assigned only one kind of agency. Therefore, we identified one to three different agentic themes in the interview data for each informant. We also identified individual differences in their first and second interviews (Table 1).
Next, we looked for expressions of structure, the factors that enabled or constrained agency in curriculum change. Previous studies have examined enabling and restricting structures in organisational and management cultures (e.g. Vähäsantanen et al. 2008). We approached this in a data-driven manner around the agentic themes, looking not at formal or institutional structures but at features that we identified as impeding or enabling agency. These were coded and reviewed between the two interviews and added to the comparative table. We identified individual, community, and institutional structures that enabled or impeded agency. Within these, we identified nine sub-themes and two cross-sectional themes.
We reviewed the data multiple times, viewing the results from theme-specific, contextspecific, and interviewee-specific perspectives to construct a dense report of complex social processes in curriculum change (cf. Braun and Clarke 2006). In the following section, we discuss how agency emerged and was related to structural features. We use informant quotations to illustrate our findings. Each quotation features the sequential number of the interviewee and a roman numeral indicating the interview round (I or II). In a subsequent section, we look more closely at the structures that enable or impede agency in two different forms of curriculum change. In this report, we focus on agency and structure only; a more detailed analysis of individual changes and stability in agency is reported in another study (Annala et al. 2020).

Agency and structure in curriculum change
In both curriculum changes, academics' progressive agency was the most visible. It appeared as interest in student learning and teaching in general, readiness to take responsibility for curriculum change, and valuing education. Curriculum change was approached as a meaningful process that required persistent effort and participation. The focus was on cooperation and negotiation. Curriculum was emphasised as a joint, community enterprise. Curriculum was considered a forum for finding a shared understanding and creating new practices that would contribute to student learning, teaching, and research, as the following quotations exemplify: In our department, we have a very teaching-positive atmosphere. Our response to development, training, and things like that is positive … Before we knew what we wanted to do, well, we had some ideas for the new direction, so we presented our ideas to students, who had their own meetings on different subjects. So, they collected their worries and ideas and presented those to us in a bigger seminar, and we responded with our views, so here our students joined as a big group [in curriculum change]. (6, I) A good thing is that we have collaborated and tried to act benevolently towards all. It has been very positive at the faculty level in all disciplines; in the degree programme, there has been willingness to collaborate and to find shared lines, absolutely. (15, II) Academics who expressed progressive agency recognised the challenges of curriculum development but tried to turn it into meaningful work for the academic community. Structures that contributed to progressive agency appeared to be shared and aligned values and goals at the individual, community, and university levels. At the individual level, some of the informants admitted that they identified more as teachers than as researchers. If a curriculum change resulted in new responsibilities, those would be considered a position of trust. By contrast, some of the academics altered between responsive enthusiasm and being overburdened, especially those who were active in research. The university-wide curriculum change was experienced as a major and hasty process. Community support and shared values were essential, as shown by the following example: The schedule was tight for curriculum work, but it was nice that we had many meetings because those helped us create a shared vision well. (16,II) In the degree programme, we distribute work according to our specialities. Through that, we find our identity, which is connected to the research group and teaching related to it. (8,II) Academics who exhibited progressive agency appeared to leverage resources by being creative, thus overcoming the constraints of limited resources to reach the goal of having advanced curricula and top-quality university education while also focusing on research activities. During the university-wide curriculum change, the shared understanding of the key aims of the educational and curriculum reform appeared as a key factor in sustaining progressive agency. The working processes around curriculum reform were straightforward, and the tasks were approached as shared challenges rather than obstacles or problems.
Oppositional agency is the opposite of progressive agency. Academics who exhibited oppositional agency did not believe in any curriculum change and did not necessarily see the idea of curriculum change as a part of scholars' work: Many professors like me think that we have so many other duties … supervising and reading doctoral dissertations, writing reviews and research plans, and having meetings. Thus, the curriculum is handled with the least effort. (9, I) I do not believe that what is written [in the curriculum] … has any effect on student motivation. (1, I) There have been discussions about uniting us with [x], so we thought it is not worth making any changes, just put the syllabus in a copy machine and change the date, and say this is our new curriculum. (3, I) Teaching, student learning, and research were loosely connected, if at all, to curriculum change. Academics who exhibited oppositional agency preferred the status quo. Bonding with like-minded people was a way to use power or to survive. Oppositional agency included resistant agentic projects regarding both departmental and universitywide curriculum initiatives. These projects were portrayed as active or passive resistance (for example, through non-participation) despite being required by certain professional positions and duties. As one informant expressed, In the end, I had quite a marginal role … It was partly my own choice, partly consciously knowing that they might not want me there because critical people are not wanted. (9, II) Institutional structures, such as administration, schedules, IT systems, and management, were considered to complicate rather than contribute to the work. The goals and values of oppositional agency seemed to differ remarkably from the goals and values of the leaders of the changes. In the absence of alternative ways forward, oppositional agency indicated cynicism, resentment, and tension, yet people seemed to balance empowering collective resistance and personal exhaustion.
Academics who exhibited territorial agency expressed an interest in protecting the high status and boundaries of disciplinary fields based on their experience with the sciences, exceptional features, or otherwise remarkable academic practices. These were used to justify not following the university-wide or general policy guidelines. The boundaries of disciplinary territories had to be secured or, if necessary, enlarged under their conditions. Compared with oppositional agency, territorial agency also had resistance, but the latter's focus was on providing excellent curricula for students, as in progressive agency. People who exhibited territorial agency had strong confidence in their superiority: I think our education is above average, and there is no need for such a change. (4, I) Although collaboration with others was possible, it was not needed or desired because there seemed to be competency, capacity, and sufficient resources. People who exhibited territorial agency had a strong commitment to certain professional and personal values. During the departmental curriculum changes, the internal practices remained untouched. However, during the university-wide curriculum change, the structures coming from the outside were a burden: [We] were given such instructions or restrictions that did not improve the work here. Rather, we were required to struggle against those to prevent or alleviate them so that we could have a territory where we can work reasonably. (2, II) In these cases, the university-wide guidelines were institutional structures, questioning and challenging the agency of the academics who were committed to developing their curriculum starting from their premises. Tension arose at the individual level with professional identities that could not be changed through administrative guidelines.
Academics who exhibited bridge-building agency described their role as mediators in curriculum change between the different interest groups. They strove to balance the intensified labour market needs and scientific traditions, research and teaching, university-wide expectations, and departmental preferences. They focused on fostering understanding between the different individuals and disciplinary groups. An enterprise that aimed to create a curriculum as a coherent entity required mediation and compromise and resulted in an unrewarding role in the community, as illustrated in the following: People have criticised strongly, and of course, I am the one who gets it over myself. We must work more with this issue. (14, II) I realised that I addressed many times and in many places like 'it will be all right, we will go on, let's wait a bit more, it won't result [in] a catastrophe'. (15,II) Academics who exhibited bridge-building agency focused on predicting frameworks, conditions, and limits and adjusting their curriculum development goals to meet the realm of possibility. This agency appeared as a personal choice but was often connected to tasks, such as a person with more responsibilities or the role of a leader in curriculum work in the community. In this role, there was no opportunity to question the curriculum change. Institutional structures were expected to contribute to the curriculum process, especially in the university-wide process, but they also required adjustment. As one informant explained, The starting points were quite clear. The university-wide reform had simple criteria, and we based our work on them. However, there were some dubious criteria, and they increased gradually throughout the year. It caused us a lot of unnecessary work because instead of clear standards and guidelines, there was this incredible conceptual jungle with which we were struggling for months. (11,II) While academics who exhibited bridge-building agency attempted to meet the aims of the university, department, and disciplines and to balance different groups and interests, the formal structures were expected to support the process. However, the fragmented time frames and the guidelines, which were simultaneously lenient and strict, made the change complicated. In some cases, IT systems limited innovation or the necessary backing from superiors was missing, and stronger contributive structures were desired.
Academics who exhibited accommodating agency focused on a rational process and a desire to avoid conflicts. They regarded curriculum work as a natural part of academic work, performed the given tasks as expected, and repeated process steps if needed. They were busy and exhausted but not hopeless. Chaos was accepted as part of the curriculum development process and was regarded as a driving force to create something new. There was neither considerable passion nor resistance, but everyone performed their parts when asked: In this work, people must have a tolerance for chaos and insecurity. (1, I) We do whatever comes from above. (13, I) Collaboration was seen as an elementary part of curriculum development, but collaborative practices were not advanced. Formal and technical guidelines were regarded as more dominant than pedagogical perspectives in curriculum work, but people with accommodating agency adjusted to the circumstances. As one informant expressed, [I] have a feeling that this is already the end because there are no more guidelines. This process has been supervised from the outside, the systems and regimes in the frontline. Now, we would need something like 'how to advance from here', something to exert effort. (17,II) The institutional structures, administrations and networks made accommodating agency more or less invisible. The individuals were serving big entitiesthe university and societybut there was still room for local subcultures in curriculum work, which could flourish if they fit in the broad framework.
Academics who exhibited powerless agency appeared only in the university-wide curriculum change context. Powerless agency shares many similar features with progressive agency, such as an interest in student learning and collaborative work. Academics who exhibited powerless agency had a strong will to contribute to the university community, which was considered a site of research and teaching, but they had no status or space to execute agency. This situation caused frustration and disappointment, as illustrated in the following quotations: It is like a huge ship that just sails on. The direction has been chosen. Inside, you can twiddle around, but you cannot influence the direction. (6, II) The culture at this university has changed so much, so that many have a feeling that [they] have to keep an eye out for the surroundings, that what is happening and the previous sense of security has disappeared. (9, II) Powerless agency represents a conflict between personal values and understanding connected to academic teaching quality versus the values in new curriculum work practices and structures. Some academics felt very competent but realised that they had no access to curriculum change because of their low status. Alternative suggestions based on experience and knowledge seemed pointless because these voices were not heard, just as before. However, status or position did not guarantee a space to implement agency when new competent people took the lead. A paradox emerged: while there were formal guidelines to include people from different positions, some academics experienced the 'open invitation' as exclusion, resulting in a lack of agency. The formal structures emphasised participation, but democracy was considered artificial because the power structures were hidden and unclear.
The relationship between structure and agency in different forms of curriculum change With regard to structure and agency in two forms of curriculum change, we identified institutional, community, and individual structures, with nine sub-themes and two cross-sectional themes, which were found to enable or impede agency (Figure 1).
The institutional structures consisted of institutional design, management practices, and resources. The departmental curriculum change featured an institutional design with strong autonomy, and management practices provided space for diversity, which benefited oppositional, territorial, and progressive agency. The administrative guidelines issued by the university were lenient or not relevant enough to mention. This feature of universities can be described as 'loosely coupled' systems (Weick 1976). Yet, these systems did not always enable bridge-building agency when the responsibility for decisions had to be assumed without institutional backup. While some fields, like those expressing territorial agency, had stable positions and resources within the institution, some others, like those featuring oppositional and progressive agency, were insecure about their future. The latter ended up being either passive or creative in making their limited resources work within the curriculum change. The limited resources here refer not only to funding and staff members' contracts but also to stability and the time available for all the work required.
In the university-wide curriculum change, the institutional design was characterised as being both a backup resource and an irrational systemstrict and loose, hierarchical and obscure, and dynamic and unpredictable at the same time. The most emergent structure that restricted agency was related to the simultaneous organisational change. Because of the new organisation, the management practices became unclear such that there appeared to be neither leadership nor responsibility, which made implementing and accommodating other forms of agency difficult. Additionally, Clavert et al. (2018) found that formal leadership is important in pedagogical development efforts, especially in university-wide reforms.
Competition emerged between the actors and local units. The unbalanced resources restricted agency because there was no time to share histories and future aims on which to ground the new curricula for the degree programmes. By contrast, for those expressing progressive agency, the university-wide change was merely a continuation of the ongoing curriculum change. For them, curriculum was clearly a process. This result reveals the dissimilarity of curriculum changes between the fields. This finding is consistent with Brøgger's (2014) notion of how such change can be described not as one reality given from the outside but as multiple realities.
The community structure consisted of disciplinary, pedagogical, and social cultures and relationships. Compared with the institutional structure, the community structure played a similar key role in supporting or restricting agency in both forms of curriculum change organisation. Barman et al. (2016) stated that the local management and community involved in reforms can reshape policy and local practice. In this study, the disciplinary, pedagogical, and social cultures were intertwined, and they either supported or restricted agency. Finding connections between research and teaching, strong disciplinary status, and strong pedagogical culture supports progressive and territorial agency. In comparison, a need to balance professional and scientific aims and living through pedagogical 'shadiness' (i.e. unclear work processes) does not, emerged especially in bridge-building agency. This finding on the impact of a direct work environment is consistent with that of van Lankveld et al. (2017), who found it to be strengthening or constraining, depending on whether teaching is valued in a department.
Curriculum change is essentially linked with pedagogy. In the present study, progressive agency was present in communities that cherished pedagogical cultures and focused on student learning. However, the pedagogical culture was dismantled in some fields during the university-wide change because of the reformulation of the curriculum structures into more interdisciplinary entities. This was most visible in powerless agency. Progressive agency in curriculum change is hindered if the pedagogical culture and activities related to it, like curriculum work, are not considerably valued. This finding reflects a trend in higher education in which research and teaching are increasingly separated as two distinct activities (Leisyte, Enders, and de Boer 2009).
Some felt that the university-wide reform was a shift towards schooling and thus limited agency and autonomy in teaching and pedagogy. Those who expressed powerless agency found that technical, rational, and administrative processes dominated. They were experienced as separate from the pedagogical development processes and research in the field, thereby impeding the teachers' opportunities for agentic projects. Meanwhile, some, especially those who exhibited progressive or bridge-building agency, felt that different processes could be united while still giving academics autonomy in teaching practice.
Social cultures refer to (1) the relationships between people within a local community, its subcultures, and groups and (2) the connection and networks across the faculty borders and the world outside the university. Networking and a collaborative, flexible, and attentive work culture supported progressive and bridge-building agency. These findings align with the findings of Burrell et al. (2015), who found that collaborative team culture may form the basis for transformative cultural change. Additionally, collective resistance created a positive group cohesion and agentic project against the mainstream enterprises, as indicated in oppositional and territorial agency. According to Brøgger (2014), this kind of collective agentic project could be described as 'ghost armies' and 'juntas', which are rooted in history. In the departmental curriculum change, different interest groups evidently struggled between subjects and other subgroups, but in the university-wide change, they also struggled with the university. One strategy to proceed with the change was to avoid disciplinary, pedagogical, and social conflicts, but this restricted the agency one wished to use.
In both forms of curriculum change organisation, tension existed between participatory practices and 'lonely riders'. During the departmental curriculum changes, there was sizeable room for individuality, and tension existed between subcultures. The universitywide change's emphasis on a participatory approach was regarded by some as a threat to their autonomy and power positions within the community, which emerged in oppositional and territorial agency. Academic autonomy is usually connected to limitations in political and interest groups' influence over academics' teaching and research but is broadly understood as academics' right to autonomous behaviour (Broström, Feldmann, and Kaulio 2019).
The individual structure consisted of engagement, position, and identity. In both forms of curriculum change, engagement and contribution or disengagement and resistance supported the agentic projects. The main difference between the departmental and university-wide curriculum changes was that the latter offered no space for all to engage, even if they wished to contribute, as emerged in powerless agency. The dispositions and qualities that featured oppositional agency were connected to exhaustion, pessimism, cynicism, and a decision to withdraw from the processes. Furthermore, lack of knowledge of the processes seemed to play a role as an emergent disinterest in such knowledge, especially in oppositional agency. Yet the bridge-building agency that accompanied the responsibility role in the change process included more or less forced engagement. This finding agrees with that of Barman et al. (2016), who suggest that unwanted policies become meaningful through collaborative learning if both academics and local managers are included. Here, a professor's position or leading role has meaning in either supporting or restricting their own or other groups' agentic projects.
Variations in academic identities appeared in both forms of change and involved either enabling or impeding agency, such as when people regarded themselves more as teachers than as researchers, as special and superior, as professionals who were training new professionals, as development-oriented persons, as ethical and responsible individuals, or as rebels and critical representatives of the opposition. However, the universitywide change yielded an identity crisis experienced by informants who exhibited powerless agency. Forming a new identity was a solution exhibited in a study by Barman et al. (2016). However, the new academic identity in their study was built on knowledge about student learning processes, whereas the crisis in the present data emerged because the massive degree programmes seemed to lack a student-centred approach, and the curriculum culture changed significantly.
Power relations played a key role in enabling and restricting structural-agentic processes at the individual, community, and institutional levels. The locus of power could rest on strong individuals and strong disciplinary communities, which enabled or restricted agency in both forms of curriculum change. In the departmental curriculum change, students and teachers had power through their belonging to local units and their decision-making. In the university-wide change, individuals and the department did not have similar powers as they did before. In the departmental curriculum change, agency was impeded because of fear of losing personal or disciplinary power within the local community, but it could still be characterised as an optimistic struggle compared to the struggles in the university-wide change. Therefore, the scale of the struggle for power changed when institutional-level management and administration played reinforced roles in steering the change. Some struggled for their position, some for the students, and some for the university educational culture. The meaning of power was thus highly relevant, and the power struggles and relations between people, disciplines, and professions were remarkable and visible.
The emergence of intentions related to curriculum change from the individuals' or community's needs enabled agency. During the departmental changes, the difficulty of defining the boundaries for the changes created some challenges. In comparison, during the university-wide change, many obstacles could be identified. First, the intentions of the university had a risk of leading to conflicts at the local and community levels. Second, the fundamental intentions behind the university-wide change were not articulated well enough, and the object of the actions was not always shared. Third, the theoretical understanding of curriculum and its importance in defining the highest education were various or limited at the school and community levels.

Discussion
The results of our study showed that the emergent body of agency remained the same regardless of whether the curriculum change was organised at the departmental level or as university-wide. Some of the six agentic themes identified in this study bore similarities to Oliver's (1991) typology of different strategies towards institutional change pressures. His typology represents 'negative' categoriesfrom active organisational resistance to passive conformity to proactive manipulation. In this study, agency appeared to be more multifaceted and ultimately highly engaged. In many cases, the curriculum development deeply concerned the academics' areas of expertise, pedagogical philosophies, and preferred research fields, thereby compelling them to commit to the curriculum change (see Priestley et al. 2012). This may illustrate how the university as an institution differs from other types of organisations, even though universities have become more 'organisation-like' than before (Broström, Feldmann, and Kaulio 2019).
In both forms of curriculum change, the local community and its networks formed a vital structure that supported agency. The collective approach and orientation in the local community seemingly affected the individuals' engagement with curriculum change. Therefore, having the agency to make justified decisions at the grassroots level is an important experience. However, the results show the difficulty of implementing university-wide reform with a participatory approach, as required in previous research (Shay and Peseta 2016). In this case, such an approach was introduced to the university community as a reform initiated from the inside. The participatory approach with loose institutional structures shifted the emphasis on power struggles to the community and individual levels, causing tensions.
In curriculum change, there is a danger of interpreting agency and structure as social phenomena too narrowly. Reliance on overly socialised, macro views of the relationship between agency and structure could lead to managerial macro structures being blamed because of changes that limit academic autonomy. However, there is also a risk of relying on overly individualised, psychological views of agency and judging teachers' dispositions, human capacity, and willingness to change. Between these extremes, the results of this study highlight the importance of structural-agentic processes, including social cultures and relationships. This can be characterised by the concept of relational agency, which was introduced by Edwards (2005, 172) as 'the capacity to work with others to expand the object that one is working on and trying to transform by recognising and accessing the resources that others bring to bear as they interpret and respond to the object'.
With curriculum being the object of activity, collegial sharing and understanding are vital. As Trowler and Cooper (2002;Trowler 2019) suggested within their context of teaching and learning regimes, an academic's identity needs to be considered in relation to others. This study indicated the importance of the links between the individual and community. Social cultures have a connection to how academic freedom is used as a room for manoeuvring during curriculum changes.
The agency in this study was connected to how curriculum and curriculum change were understood as parts of academic work. This finding is similar to the results of previous studies that found that curricula do not have a shared meaning in academic communities (Annala, Lindén, and Mäkinen 2016;Bovill and Woolmer 2019). This is also related to the question of how curriculum change is understood and implemented as practice as either an engineering or a bricolage type of activity (Louvel 2013), or as a thing or process (cf. Ashwin 2009). Consciousness of the visible and hidden intentions and dynamics in the structural-agentic processes in curriculum change may help members of the academic community find their role in these processes meaningful. In this study, we provided a general account of academics' agentic projects in curriculum changes, but more research is needed from the perspective of academics with different positions and with more variation in the forms of curriculum change.

Notes
1. An aim to create a European Higher Education Area and educational harmonisation through several initiatives. 2. The university in question was a public university with about 15,000 students and 2,000 staff members. In research universities in Finland, students are admitted to pursue bachelor and master degrees as a single entity; therefore, curriculum change concerns both degrees.